From 7eac0081a8e958106ed3aea402c8105f30fad6d9 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Wed, 10 Aug 2022 09:59:15 +0100 Subject: [PATCH 001/121] riscv: dts: microchip: add qspi compatible fallback The "hard" QSPI peripheral on PolarFire SoC is derived from version 2 of the FPGA IP core. The original binding had no fallback etc, so this device tree is valid as is. There was also no functional driver for the QSPI IP, so no device with a devicetree from a previous mainline release will regress. Link: https://lore.kernel.org/linux-spi/7c9f0d96-2882-964a-cd1f-916ddb3f0410@linaro.org/ Signed-off-by: Conor Dooley Acked-by: Krzysztof Kozlowski --- arch/riscv/boot/dts/microchip/mpfs.dtsi | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/riscv/boot/dts/microchip/mpfs.dtsi b/arch/riscv/boot/dts/microchip/mpfs.dtsi index 499c2e63ad35..45e3cc659882 100644 --- a/arch/riscv/boot/dts/microchip/mpfs.dtsi +++ b/arch/riscv/boot/dts/microchip/mpfs.dtsi @@ -330,7 +330,7 @@ spi1: spi@20109000 { }; qspi: spi@21000000 { - compatible = "microchip,mpfs-qspi"; + compatible = "microchip,mpfs-qspi", "microchip,coreqspi-rtl-v2"; #address-cells = <1>; #size-cells = <0>; reg = <0x0 0x21000000 0x0 0x1000>; From c530a3c716b963625e43aa915e0de6b4d1ce8ad9 Mon Sep 17 00:00:00 2001 From: Chengming Zhou Date: Fri, 26 Aug 2022 00:41:02 +0800 Subject: [PATCH 002/121] sched/psi: Fix periodic aggregation shut off We don't want to wake periodic aggregation work back up if the task change is the aggregation worker itself going to sleep, or we'll ping-pong forever. Previously, we would use psi_task_change() in psi_dequeue() when task going to sleep, so this check was put in psi_task_change(). But commit 4117cebf1a9f ("psi: Optimize task switch inside shared cgroups") defer task sleep handling to psi_task_switch(), won't go through psi_task_change() anymore. So this patch move this check to psi_task_switch(). Fixes: 4117cebf1a9f ("psi: Optimize task switch inside shared cgroups") Signed-off-by: Chengming Zhou Signed-off-by: Peter Zijlstra (Intel) Acked-by: Johannes Weiner Link: https://lore.kernel.org/r/20220825164111.29534-2-zhouchengming@bytedance.com --- kernel/sched/psi.c | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-) diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c index ecb4b4ff4ce0..39463dcc16bb 100644 --- a/kernel/sched/psi.c +++ b/kernel/sched/psi.c @@ -796,7 +796,6 @@ void psi_task_change(struct task_struct *task, int clear, int set) { int cpu = task_cpu(task); struct psi_group *group; - bool wake_clock = true; void *iter = NULL; u64 now; @@ -806,19 +805,9 @@ void psi_task_change(struct task_struct *task, int clear, int set) psi_flags_change(task, clear, set); now = cpu_clock(cpu); - /* - * Periodic aggregation shuts off if there is a period of no - * task changes, so we wake it back up if necessary. However, - * don't do this if the task change is the aggregation worker - * itself going to sleep, or we'll ping-pong forever. - */ - if (unlikely((clear & TSK_RUNNING) && - (task->flags & PF_WQ_WORKER) && - wq_worker_last_func(task) == psi_avgs_work)) - wake_clock = false; while ((group = iterate_groups(task, &iter))) - psi_group_change(group, cpu, clear, set, now, wake_clock); + psi_group_change(group, cpu, clear, set, now, true); } void psi_task_switch(struct task_struct *prev, struct task_struct *next, @@ -854,6 +843,7 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, if (prev->pid) { int clear = TSK_ONCPU, set = 0; + bool wake_clock = true; /* * When we're going to sleep, psi_dequeue() lets us @@ -867,13 +857,23 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, clear |= TSK_MEMSTALL_RUNNING; if (prev->in_iowait) set |= TSK_IOWAIT; + + /* + * Periodic aggregation shuts off if there is a period of no + * task changes, so we wake it back up if necessary. However, + * don't do this if the task change is the aggregation worker + * itself going to sleep, or we'll ping-pong forever. + */ + if (unlikely((prev->flags & PF_WQ_WORKER) && + wq_worker_last_func(prev) == psi_avgs_work)) + wake_clock = false; } psi_flags_change(prev, clear, set); iter = NULL; while ((group = iterate_groups(prev, &iter)) && group != common) - psi_group_change(group, cpu, clear, set, now, true); + psi_group_change(group, cpu, clear, set, now, wake_clock); /* * TSK_ONCPU is handled up to the common ancestor. If we're tasked @@ -882,7 +882,7 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, if (sleep) { clear &= ~TSK_ONCPU; for (; group; group = iterate_groups(prev, &iter)) - psi_group_change(group, cpu, clear, set, now, true); + psi_group_change(group, cpu, clear, set, now, wake_clock); } } } From 58d8c2586cedb8a67f6f0dffa5eaed0f89135b39 Mon Sep 17 00:00:00 2001 From: Chengming Zhou Date: Fri, 26 Aug 2022 00:41:03 +0800 Subject: [PATCH 003/121] sched/psi: Don't create cgroup PSI files when psi_disabled commit 3958e2d0c34e ("cgroup: make per-cgroup pressure stall tracking configurable") make PSI can be configured to skip per-cgroup stall accounting. And doesn't expose PSI files in cgroup hierarchy. This patch do the same thing when psi_disabled. Signed-off-by: Chengming Zhou Signed-off-by: Peter Zijlstra (Intel) Acked-by: Johannes Weiner Link: https://lore.kernel.org/r/20220825164111.29534-3-zhouchengming@bytedance.com --- kernel/cgroup/cgroup.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index 718a70c01c04..96aefdb064bb 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -3780,6 +3780,9 @@ static void cgroup_pressure_release(struct kernfs_open_file *of) bool cgroup_psi_enabled(void) { + if (static_branch_likely(&psi_disabled)) + return false; + return (cgroup_feature_disable_mask & (1 << OPT_FEATURE_PRESSURE)) == 0; } From e2ad8ab04c5cdfc8dc2f382c45d248ab01dee991 Mon Sep 17 00:00:00 2001 From: Chengming Zhou Date: Fri, 26 Aug 2022 00:41:04 +0800 Subject: [PATCH 004/121] sched/psi: Save percpu memory when !psi_cgroups_enabled We won't use cgroup psi_group when !psi_cgroups_enabled, so don't bother to alloc percpu memory and init for it. Also don't need to migrate task PSI stats between cgroups in cgroup_move_task(). Signed-off-by: Chengming Zhou Signed-off-by: Peter Zijlstra (Intel) Acked-by: Johannes Weiner Link: https://lore.kernel.org/r/20220825164111.29534-4-zhouchengming@bytedance.com --- kernel/sched/psi.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c index 39463dcc16bb..77d53c03a76f 100644 --- a/kernel/sched/psi.c +++ b/kernel/sched/psi.c @@ -201,6 +201,7 @@ void __init psi_init(void) { if (!psi_enable) { static_branch_enable(&psi_disabled); + static_branch_disable(&psi_cgroups_enabled); return; } @@ -950,7 +951,7 @@ void psi_memstall_leave(unsigned long *flags) #ifdef CONFIG_CGROUPS int psi_cgroup_alloc(struct cgroup *cgroup) { - if (static_branch_likely(&psi_disabled)) + if (!static_branch_likely(&psi_cgroups_enabled)) return 0; cgroup->psi = kzalloc(sizeof(struct psi_group), GFP_KERNEL); @@ -968,7 +969,7 @@ int psi_cgroup_alloc(struct cgroup *cgroup) void psi_cgroup_free(struct cgroup *cgroup) { - if (static_branch_likely(&psi_disabled)) + if (!static_branch_likely(&psi_cgroups_enabled)) return; cancel_delayed_work_sync(&cgroup->psi->avgs_work); @@ -996,7 +997,7 @@ void cgroup_move_task(struct task_struct *task, struct css_set *to) struct rq_flags rf; struct rq *rq; - if (static_branch_likely(&psi_disabled)) { + if (!static_branch_likely(&psi_cgroups_enabled)) { /* * Lame to do this here, but the scheduler cannot be locked * from the outside, so we move cgroups from inside sched/. From d79ddb069c5257a924456eb99b53fc1ea715c0a3 Mon Sep 17 00:00:00 2001 From: Chengming Zhou Date: Fri, 26 Aug 2022 00:41:05 +0800 Subject: [PATCH 005/121] sched/psi: Move private helpers to sched/stats.h This patch move psi_task_change/psi_task_switch declarations out of PSI public header, since they are only needed for implementing the PSI stats tracking in sched/stats.h psi_task_switch is obvious, psi_task_change can't be public helper since it doesn't check psi_disabled static key. And there is no any user now, so put it in sched/stats.h too. Signed-off-by: Chengming Zhou Signed-off-by: Peter Zijlstra (Intel) Acked-by: Johannes Weiner Link: https://lore.kernel.org/r/20220825164111.29534-5-zhouchengming@bytedance.com --- include/linux/psi.h | 4 ---- kernel/sched/stats.h | 4 ++++ 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/include/linux/psi.h b/include/linux/psi.h index dd74411ac21d..fffd229fbf19 100644 --- a/include/linux/psi.h +++ b/include/linux/psi.h @@ -18,10 +18,6 @@ extern struct psi_group psi_system; void psi_init(void); -void psi_task_change(struct task_struct *task, int clear, int set); -void psi_task_switch(struct task_struct *prev, struct task_struct *next, - bool sleep); - void psi_memstall_enter(unsigned long *flags); void psi_memstall_leave(unsigned long *flags); diff --git a/kernel/sched/stats.h b/kernel/sched/stats.h index baa839c1ba96..c39b467ece43 100644 --- a/kernel/sched/stats.h +++ b/kernel/sched/stats.h @@ -107,6 +107,10 @@ __schedstats_from_se(struct sched_entity *se) } #ifdef CONFIG_PSI +void psi_task_change(struct task_struct *task, int clear, int set); +void psi_task_switch(struct task_struct *prev, struct task_struct *next, + bool sleep); + /* * PSI tracks state that persists across sleeps, such as iowaits and * memory stalls. As a result, it has to distinguish between sleeps, From 65176f59a18d888684525658a1d0b8bf749d24f3 Mon Sep 17 00:00:00 2001 From: Chengming Zhou Date: Fri, 26 Aug 2022 00:41:06 +0800 Subject: [PATCH 006/121] sched/psi: Optimize task switch inside shared cgroups again Way back when PSI_MEM_FULL was accounted from the timer tick, task switching could simply iterate next and prev to the common ancestor to update TSK_ONCPU and be done. Then memstall ticks were replaced with checking curr->in_memstall directly in psi_group_change(). That meant that now if the task switch was between a memstall and a !memstall task, we had to iterate through the common ancestors at least ONCE to fix up their state_masks. We added the identical_state filter to make sure the common ancestor elimination was skipped in that case. It seems that was always a little too eager, because it caused us to walk the common ancestors *twice* instead of the required once: the iteration for next could have stopped at the common ancestor; prev could have updated TSK_ONCPU up to the common ancestor, then finish to the root without changing any flags, just to get the new curr->in_memstall into the state_masks. This patch recognizes this and makes it so that we walk to the root exactly once if state_mask needs updating, which is simply catching up on a missed optimization that could have been done in commit 7fae6c8171d2 ("psi: Use ONCPU state tracking machinery to detect reclaim") directly. Apart from this, it's also necessary for the next patch "sched/psi: remove NR_ONCPU task accounting". Suppose we walk the common ancestors twice: (1) psi_group_change(.clear = 0, .set = TSK_ONCPU) (2) psi_group_change(.clear = TSK_ONCPU, .set = 0) We previously used tasks[NR_ONCPU] to record TSK_ONCPU, tasks[NR_ONCPU]++ in (1) then tasks[NR_ONCPU]-- in (2), so tasks[NR_ONCPU] still be correct. The next patch change to use one bit in state mask to record TSK_ONCPU, PSI_ONCPU bit will be set in (1), but then be cleared in (2), which cause the psi_group_cpu has task running on CPU but without PSI_ONCPU bit set! With this patch, we will never walk the common ancestors twice, so won't have above problem. Suggested-by: Johannes Weiner Signed-off-by: Chengming Zhou Signed-off-by: Peter Zijlstra (Intel) Acked-by: Johannes Weiner Link: https://lore.kernel.org/r/20220825164111.29534-6-zhouchengming@bytedance.com --- kernel/sched/psi.c | 21 +++++++++------------ 1 file changed, 9 insertions(+), 12 deletions(-) diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c index 77d53c03a76f..d71dbc2356ff 100644 --- a/kernel/sched/psi.c +++ b/kernel/sched/psi.c @@ -820,20 +820,15 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, u64 now = cpu_clock(cpu); if (next->pid) { - bool identical_state; - psi_flags_change(next, 0, TSK_ONCPU); /* - * When switching between tasks that have an identical - * runtime state, the cgroup that contains both tasks - * we reach the first common ancestor. Iterate @next's - * ancestors only until we encounter @prev's ONCPU. + * Set TSK_ONCPU on @next's cgroups. If @next shares any + * ancestors with @prev, those will already have @prev's + * TSK_ONCPU bit set, and we can stop the iteration there. */ - identical_state = prev->psi_flags == next->psi_flags; iter = NULL; while ((group = iterate_groups(next, &iter))) { - if (identical_state && - per_cpu_ptr(group->pcpu, cpu)->tasks[NR_ONCPU]) { + if (per_cpu_ptr(group->pcpu, cpu)->tasks[NR_ONCPU]) { common = group; break; } @@ -877,10 +872,12 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, psi_group_change(group, cpu, clear, set, now, wake_clock); /* - * TSK_ONCPU is handled up to the common ancestor. If we're tasked - * with dequeuing too, finish that for the rest of the hierarchy. + * TSK_ONCPU is handled up to the common ancestor. If there are + * any other differences between the two tasks (e.g. prev goes + * to sleep, or only one task is memstall), finish propagating + * those differences all the way up to the root. */ - if (sleep) { + if ((prev->psi_flags ^ next->psi_flags) & ~TSK_ONCPU) { clear &= ~TSK_ONCPU; for (; group; group = iterate_groups(prev, &iter)) psi_group_change(group, cpu, clear, set, now, wake_clock); From 71dbdde7914d32e86f01ac1f6e54e964c9dfdbd9 Mon Sep 17 00:00:00 2001 From: Johannes Weiner Date: Fri, 26 Aug 2022 00:41:07 +0800 Subject: [PATCH 007/121] sched/psi: Remove NR_ONCPU task accounting We put all fields updated by the scheduler in the first cacheline of struct psi_group_cpu for performance. Since we want add another PSI_IRQ_FULL to track IRQ/SOFTIRQ pressure, we need to reclaim space first. This patch remove NR_ONCPU task accounting in struct psi_group_cpu, use one bit in state_mask to track instead. Signed-off-by: Johannes Weiner Signed-off-by: Chengming Zhou Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Chengming Zhou Tested-by: Chengming Zhou Link: https://lore.kernel.org/r/20220825164111.29534-7-zhouchengming@bytedance.com --- include/linux/psi_types.h | 16 +++++++-------- kernel/sched/psi.c | 41 ++++++++++++++++++++++++++++----------- 2 files changed, 37 insertions(+), 20 deletions(-) diff --git a/include/linux/psi_types.h b/include/linux/psi_types.h index c7fe7c089718..54cb74946db4 100644 --- a/include/linux/psi_types.h +++ b/include/linux/psi_types.h @@ -15,13 +15,6 @@ enum psi_task_count { NR_IOWAIT, NR_MEMSTALL, NR_RUNNING, - /* - * This can't have values other than 0 or 1 and could be - * implemented as a bit flag. But for now we still have room - * in the first cacheline of psi_group_cpu, and this way we - * don't have to special case any state tracking for it. - */ - NR_ONCPU, /* * For IO and CPU stalls the presence of running/oncpu tasks * in the domain means a partial rather than a full stall. @@ -32,16 +25,18 @@ enum psi_task_count { * threads and memstall ones. */ NR_MEMSTALL_RUNNING, - NR_PSI_TASK_COUNTS = 5, + NR_PSI_TASK_COUNTS = 4, }; /* Task state bitmasks */ #define TSK_IOWAIT (1 << NR_IOWAIT) #define TSK_MEMSTALL (1 << NR_MEMSTALL) #define TSK_RUNNING (1 << NR_RUNNING) -#define TSK_ONCPU (1 << NR_ONCPU) #define TSK_MEMSTALL_RUNNING (1 << NR_MEMSTALL_RUNNING) +/* Only one task can be scheduled, no corresponding task count */ +#define TSK_ONCPU (1 << NR_PSI_TASK_COUNTS) + /* Resources that workloads could be stalled on */ enum psi_res { PSI_IO, @@ -68,6 +63,9 @@ enum psi_states { NR_PSI_STATES = 7, }; +/* Use one bit in the state mask to track TSK_ONCPU */ +#define PSI_ONCPU (1 << NR_PSI_STATES) + enum psi_aggregators { PSI_AVGS = 0, PSI_POLL, diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c index d71dbc2356ff..4702a770e272 100644 --- a/kernel/sched/psi.c +++ b/kernel/sched/psi.c @@ -212,7 +212,7 @@ void __init psi_init(void) group_init(&psi_system); } -static bool test_state(unsigned int *tasks, enum psi_states state) +static bool test_state(unsigned int *tasks, enum psi_states state, bool oncpu) { switch (state) { case PSI_IO_SOME: @@ -225,9 +225,9 @@ static bool test_state(unsigned int *tasks, enum psi_states state) return unlikely(tasks[NR_MEMSTALL] && tasks[NR_RUNNING] == tasks[NR_MEMSTALL_RUNNING]); case PSI_CPU_SOME: - return unlikely(tasks[NR_RUNNING] > tasks[NR_ONCPU]); + return unlikely(tasks[NR_RUNNING] > oncpu); case PSI_CPU_FULL: - return unlikely(tasks[NR_RUNNING] && !tasks[NR_ONCPU]); + return unlikely(tasks[NR_RUNNING] && !oncpu); case PSI_NONIDLE: return tasks[NR_IOWAIT] || tasks[NR_MEMSTALL] || tasks[NR_RUNNING]; @@ -689,9 +689,9 @@ static void psi_group_change(struct psi_group *group, int cpu, bool wake_clock) { struct psi_group_cpu *groupc; - u32 state_mask = 0; unsigned int t, m; enum psi_states s; + u32 state_mask; groupc = per_cpu_ptr(group->pcpu, cpu); @@ -707,17 +707,36 @@ static void psi_group_change(struct psi_group *group, int cpu, record_times(groupc, now); + /* + * Start with TSK_ONCPU, which doesn't have a corresponding + * task count - it's just a boolean flag directly encoded in + * the state mask. Clear, set, or carry the current state if + * no changes are requested. + */ + if (unlikely(clear & TSK_ONCPU)) { + state_mask = 0; + clear &= ~TSK_ONCPU; + } else if (unlikely(set & TSK_ONCPU)) { + state_mask = PSI_ONCPU; + set &= ~TSK_ONCPU; + } else { + state_mask = groupc->state_mask & PSI_ONCPU; + } + + /* + * The rest of the state mask is calculated based on the task + * counts. Update those first, then construct the mask. + */ for (t = 0, m = clear; m; m &= ~(1 << t), t++) { if (!(m & (1 << t))) continue; if (groupc->tasks[t]) { groupc->tasks[t]--; } else if (!psi_bug) { - printk_deferred(KERN_ERR "psi: task underflow! cpu=%d t=%d tasks=[%u %u %u %u %u] clear=%x set=%x\n", + printk_deferred(KERN_ERR "psi: task underflow! cpu=%d t=%d tasks=[%u %u %u %u] clear=%x set=%x\n", cpu, t, groupc->tasks[0], groupc->tasks[1], groupc->tasks[2], - groupc->tasks[3], groupc->tasks[4], - clear, set); + groupc->tasks[3], clear, set); psi_bug = 1; } } @@ -726,9 +745,8 @@ static void psi_group_change(struct psi_group *group, int cpu, if (set & (1 << t)) groupc->tasks[t]++; - /* Calculate state mask representing active states */ for (s = 0; s < NR_PSI_STATES; s++) { - if (test_state(groupc->tasks, s)) + if (test_state(groupc->tasks, s, state_mask & PSI_ONCPU)) state_mask |= (1 << s); } @@ -740,7 +758,7 @@ static void psi_group_change(struct psi_group *group, int cpu, * task in a cgroup is in_memstall, the corresponding groupc * on that cpu is in PSI_MEM_FULL state. */ - if (unlikely(groupc->tasks[NR_ONCPU] && cpu_curr(cpu)->in_memstall)) + if (unlikely((state_mask & PSI_ONCPU) && cpu_curr(cpu)->in_memstall)) state_mask |= (1 << PSI_MEM_FULL); groupc->state_mask = state_mask; @@ -828,7 +846,8 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, */ iter = NULL; while ((group = iterate_groups(next, &iter))) { - if (per_cpu_ptr(group->pcpu, cpu)->tasks[NR_ONCPU]) { + if (per_cpu_ptr(group->pcpu, cpu)->state_mask & + PSI_ONCPU) { common = group; break; } From 52b1364ba0b105122d6de0e719b36db705011ac1 Mon Sep 17 00:00:00 2001 From: Chengming Zhou Date: Fri, 26 Aug 2022 00:41:08 +0800 Subject: [PATCH 008/121] sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Now PSI already tracked workload pressure stall information for CPU, memory and IO. Apart from these, IRQ/SOFTIRQ could have obvious impact on some workload productivity, such as web service workload. When CONFIG_IRQ_TIME_ACCOUNTING, we can get IRQ/SOFTIRQ delta time from update_rq_clock_task(), in which we can record that delta to CPU curr task's cgroups as PSI_IRQ_FULL status. Note we don't use PSI_IRQ_SOME since IRQ/SOFTIRQ always happen in the current task on the CPU, make nothing productive could run even if it were runnable, so we only use PSI_IRQ_FULL. Signed-off-by: Chengming Zhou Signed-off-by: Peter Zijlstra (Intel) Acked-by: Johannes Weiner Link: https://lore.kernel.org/r/20220825164111.29534-8-zhouchengming@bytedance.com --- Documentation/admin-guide/cgroup-v2.rst | 6 ++ include/linux/psi_types.h | 10 +++- kernel/cgroup/cgroup.c | 27 +++++++++ kernel/sched/core.c | 1 + kernel/sched/psi.c | 74 ++++++++++++++++++++++++- kernel/sched/stats.h | 2 + 6 files changed, 116 insertions(+), 4 deletions(-) diff --git a/Documentation/admin-guide/cgroup-v2.rst b/Documentation/admin-guide/cgroup-v2.rst index be4a77baf784..971c418bc778 100644 --- a/Documentation/admin-guide/cgroup-v2.rst +++ b/Documentation/admin-guide/cgroup-v2.rst @@ -976,6 +976,12 @@ All cgroup core files are prefixed with "cgroup." killing cgroups is a process directed operation, i.e. it affects the whole thread-group. + irq.pressure + A read-write nested-keyed file. + + Shows pressure stall information for IRQ/SOFTIRQ. See + :ref:`Documentation/accounting/psi.rst ` for details. + Controllers =========== diff --git a/include/linux/psi_types.h b/include/linux/psi_types.h index 54cb74946db4..40c28171cd91 100644 --- a/include/linux/psi_types.h +++ b/include/linux/psi_types.h @@ -42,7 +42,10 @@ enum psi_res { PSI_IO, PSI_MEM, PSI_CPU, - NR_PSI_RESOURCES = 3, +#ifdef CONFIG_IRQ_TIME_ACCOUNTING + PSI_IRQ, +#endif + NR_PSI_RESOURCES, }; /* @@ -58,9 +61,12 @@ enum psi_states { PSI_MEM_FULL, PSI_CPU_SOME, PSI_CPU_FULL, +#ifdef CONFIG_IRQ_TIME_ACCOUNTING + PSI_IRQ_FULL, +#endif /* Only per-CPU, to weigh the CPU in the global average: */ PSI_NONIDLE, - NR_PSI_STATES = 7, + NR_PSI_STATES, }; /* Use one bit in the state mask to track TSK_ONCPU */ diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index 96aefdb064bb..b46d39b66214 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -3763,6 +3763,23 @@ static ssize_t cgroup_cpu_pressure_write(struct kernfs_open_file *of, return cgroup_pressure_write(of, buf, nbytes, PSI_CPU); } +#ifdef CONFIG_IRQ_TIME_ACCOUNTING +static int cgroup_irq_pressure_show(struct seq_file *seq, void *v) +{ + struct cgroup *cgrp = seq_css(seq)->cgroup; + struct psi_group *psi = cgroup_ino(cgrp) == 1 ? &psi_system : cgrp->psi; + + return psi_show(seq, psi, PSI_IRQ); +} + +static ssize_t cgroup_irq_pressure_write(struct kernfs_open_file *of, + char *buf, size_t nbytes, + loff_t off) +{ + return cgroup_pressure_write(of, buf, nbytes, PSI_IRQ); +} +#endif + static __poll_t cgroup_pressure_poll(struct kernfs_open_file *of, poll_table *pt) { @@ -5179,6 +5196,16 @@ static struct cftype cgroup_base_files[] = { .poll = cgroup_pressure_poll, .release = cgroup_pressure_release, }, +#ifdef CONFIG_IRQ_TIME_ACCOUNTING + { + .name = "irq.pressure", + .flags = CFTYPE_PRESSURE, + .seq_show = cgroup_irq_pressure_show, + .write = cgroup_irq_pressure_write, + .poll = cgroup_pressure_poll, + .release = cgroup_pressure_release, + }, +#endif #endif /* CONFIG_PSI */ { } /* terminate */ }; diff --git a/kernel/sched/core.c b/kernel/sched/core.c index ee28253c9ac0..7d1ea9240af0 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -708,6 +708,7 @@ static void update_rq_clock_task(struct rq *rq, s64 delta) rq->prev_irq_time += irq_delta; delta -= irq_delta; + psi_account_irqtime(rq->curr, irq_delta); #endif #ifdef CONFIG_PARAVIRT_TIME_ACCOUNTING if (static_key_false((¶virt_steal_rq_enabled))) { diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c index 4702a770e272..2545a78f82d8 100644 --- a/kernel/sched/psi.c +++ b/kernel/sched/psi.c @@ -904,6 +904,36 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, } } +#ifdef CONFIG_IRQ_TIME_ACCOUNTING +void psi_account_irqtime(struct task_struct *task, u32 delta) +{ + int cpu = task_cpu(task); + void *iter = NULL; + struct psi_group *group; + struct psi_group_cpu *groupc; + u64 now; + + if (!task->pid) + return; + + now = cpu_clock(cpu); + + while ((group = iterate_groups(task, &iter))) { + groupc = per_cpu_ptr(group->pcpu, cpu); + + write_seqcount_begin(&groupc->seq); + + record_times(groupc, now); + groupc->times[PSI_IRQ_FULL] += delta; + + write_seqcount_end(&groupc->seq); + + if (group->poll_states & (1 << PSI_IRQ_FULL)) + psi_schedule_poll_work(group, 1); + } +} +#endif + /** * psi_memstall_enter - mark the beginning of a memory stall section * @flags: flags to handle nested sections @@ -1065,6 +1095,7 @@ void cgroup_move_task(struct task_struct *task, struct css_set *to) int psi_show(struct seq_file *m, struct psi_group *group, enum psi_res res) { + bool only_full = false; int full; u64 now; @@ -1079,7 +1110,11 @@ int psi_show(struct seq_file *m, struct psi_group *group, enum psi_res res) group->avg_next_update = update_averages(group, now); mutex_unlock(&group->avgs_lock); - for (full = 0; full < 2; full++) { +#ifdef CONFIG_IRQ_TIME_ACCOUNTING + only_full = res == PSI_IRQ; +#endif + + for (full = 0; full < 2 - only_full; full++) { unsigned long avg[3] = { 0, }; u64 total = 0; int w; @@ -1093,7 +1128,7 @@ int psi_show(struct seq_file *m, struct psi_group *group, enum psi_res res) } seq_printf(m, "%s avg10=%lu.%02lu avg60=%lu.%02lu avg300=%lu.%02lu total=%llu\n", - full ? "full" : "some", + full || only_full ? "full" : "some", LOAD_INT(avg[0]), LOAD_FRAC(avg[0]), LOAD_INT(avg[1]), LOAD_FRAC(avg[1]), LOAD_INT(avg[2]), LOAD_FRAC(avg[2]), @@ -1121,6 +1156,11 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, else return ERR_PTR(-EINVAL); +#ifdef CONFIG_IRQ_TIME_ACCOUNTING + if (res == PSI_IRQ && --state != PSI_IRQ_FULL) + return ERR_PTR(-EINVAL); +#endif + if (state >= PSI_NONIDLE) return ERR_PTR(-EINVAL); @@ -1405,6 +1445,33 @@ static const struct proc_ops psi_cpu_proc_ops = { .proc_release = psi_fop_release, }; +#ifdef CONFIG_IRQ_TIME_ACCOUNTING +static int psi_irq_show(struct seq_file *m, void *v) +{ + return psi_show(m, &psi_system, PSI_IRQ); +} + +static int psi_irq_open(struct inode *inode, struct file *file) +{ + return psi_open(file, psi_irq_show); +} + +static ssize_t psi_irq_write(struct file *file, const char __user *user_buf, + size_t nbytes, loff_t *ppos) +{ + return psi_write(file, user_buf, nbytes, PSI_IRQ); +} + +static const struct proc_ops psi_irq_proc_ops = { + .proc_open = psi_irq_open, + .proc_read = seq_read, + .proc_lseek = seq_lseek, + .proc_write = psi_irq_write, + .proc_poll = psi_fop_poll, + .proc_release = psi_fop_release, +}; +#endif + static int __init psi_proc_init(void) { if (psi_enable) { @@ -1412,6 +1479,9 @@ static int __init psi_proc_init(void) proc_create("pressure/io", 0666, NULL, &psi_io_proc_ops); proc_create("pressure/memory", 0666, NULL, &psi_memory_proc_ops); proc_create("pressure/cpu", 0666, NULL, &psi_cpu_proc_ops); +#ifdef CONFIG_IRQ_TIME_ACCOUNTING + proc_create("pressure/irq", 0666, NULL, &psi_irq_proc_ops); +#endif } return 0; } diff --git a/kernel/sched/stats.h b/kernel/sched/stats.h index c39b467ece43..84a188913cc9 100644 --- a/kernel/sched/stats.h +++ b/kernel/sched/stats.h @@ -110,6 +110,7 @@ __schedstats_from_se(struct sched_entity *se) void psi_task_change(struct task_struct *task, int clear, int set); void psi_task_switch(struct task_struct *prev, struct task_struct *next, bool sleep); +void psi_account_irqtime(struct task_struct *task, u32 delta); /* * PSI tracks state that persists across sleeps, such as iowaits and @@ -205,6 +206,7 @@ static inline void psi_ttwu_dequeue(struct task_struct *p) {} static inline void psi_sched_switch(struct task_struct *prev, struct task_struct *next, bool sleep) {} +static inline void psi_account_irqtime(struct task_struct *task, u32 delta) {} #endif /* CONFIG_PSI */ #ifdef CONFIG_SCHED_INFO From 57899a6610e67ba26fa3251ebbef4a5ed21efc5d Mon Sep 17 00:00:00 2001 From: Chengming Zhou Date: Fri, 26 Aug 2022 00:41:09 +0800 Subject: [PATCH 009/121] sched/psi: Consolidate cgroup_psi() cgroup_psi() can't return psi_group for root cgroup, so we have many open code "psi = cgroup_ino(cgrp) == 1 ? &psi_system : cgrp->psi". This patch move cgroup_psi() definition to , in which we can return psi_system for root cgroup, so can handle all cgroups. Signed-off-by: Chengming Zhou Signed-off-by: Peter Zijlstra (Intel) Acked-by: Johannes Weiner Link: https://lore.kernel.org/r/20220825164111.29534-9-zhouchengming@bytedance.com --- include/linux/cgroup.h | 5 ----- include/linux/psi.h | 6 ++++++ kernel/cgroup/cgroup.c | 10 +++++----- 3 files changed, 11 insertions(+), 10 deletions(-) diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h index b0914aa26506..80cb970257be 100644 --- a/include/linux/cgroup.h +++ b/include/linux/cgroup.h @@ -673,11 +673,6 @@ static inline void pr_cont_cgroup_path(struct cgroup *cgrp) pr_cont_kernfs_path(cgrp->kn); } -static inline struct psi_group *cgroup_psi(struct cgroup *cgrp) -{ - return cgrp->psi; -} - bool cgroup_psi_enabled(void); static inline void cgroup_init_kthreadd(void) diff --git a/include/linux/psi.h b/include/linux/psi.h index fffd229fbf19..362a74ca1d3b 100644 --- a/include/linux/psi.h +++ b/include/linux/psi.h @@ -7,6 +7,7 @@ #include #include #include +#include struct seq_file; struct css_set; @@ -30,6 +31,11 @@ __poll_t psi_trigger_poll(void **trigger_ptr, struct file *file, poll_table *wait); #ifdef CONFIG_CGROUPS +static inline struct psi_group *cgroup_psi(struct cgroup *cgrp) +{ + return cgroup_ino(cgrp) == 1 ? &psi_system : cgrp->psi; +} + int psi_cgroup_alloc(struct cgroup *cgrp); void psi_cgroup_free(struct cgroup *cgrp); void cgroup_move_task(struct task_struct *p, struct css_set *to); diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index b46d39b66214..772b35d65d1f 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -3689,21 +3689,21 @@ static int cpu_stat_show(struct seq_file *seq, void *v) static int cgroup_io_pressure_show(struct seq_file *seq, void *v) { struct cgroup *cgrp = seq_css(seq)->cgroup; - struct psi_group *psi = cgroup_ino(cgrp) == 1 ? &psi_system : cgrp->psi; + struct psi_group *psi = cgroup_psi(cgrp); return psi_show(seq, psi, PSI_IO); } static int cgroup_memory_pressure_show(struct seq_file *seq, void *v) { struct cgroup *cgrp = seq_css(seq)->cgroup; - struct psi_group *psi = cgroup_ino(cgrp) == 1 ? &psi_system : cgrp->psi; + struct psi_group *psi = cgroup_psi(cgrp); return psi_show(seq, psi, PSI_MEM); } static int cgroup_cpu_pressure_show(struct seq_file *seq, void *v) { struct cgroup *cgrp = seq_css(seq)->cgroup; - struct psi_group *psi = cgroup_ino(cgrp) == 1 ? &psi_system : cgrp->psi; + struct psi_group *psi = cgroup_psi(cgrp); return psi_show(seq, psi, PSI_CPU); } @@ -3729,7 +3729,7 @@ static ssize_t cgroup_pressure_write(struct kernfs_open_file *of, char *buf, return -EBUSY; } - psi = cgroup_ino(cgrp) == 1 ? &psi_system : cgrp->psi; + psi = cgroup_psi(cgrp); new = psi_trigger_create(psi, buf, res); if (IS_ERR(new)) { cgroup_put(cgrp); @@ -3767,7 +3767,7 @@ static ssize_t cgroup_cpu_pressure_write(struct kernfs_open_file *of, static int cgroup_irq_pressure_show(struct seq_file *seq, void *v) { struct cgroup *cgrp = seq_css(seq)->cgroup; - struct psi_group *psi = cgroup_ino(cgrp) == 1 ? &psi_system : cgrp->psi; + struct psi_group *psi = cgroup_psi(cgrp); return psi_show(seq, psi, PSI_IRQ); } From dc86aba751e2867244411adda1562f6664747019 Mon Sep 17 00:00:00 2001 From: Chengming Zhou Date: Fri, 26 Aug 2022 00:41:10 +0800 Subject: [PATCH 010/121] sched/psi: Cache parent psi_group to speed up group iteration We use iterate_groups() to iterate each level psi_group to update PSI stats, which is a very hot path. In current code, iterate_groups() have to use multiple branches and cgroup_parent() to get parent psi_group for each level, which is not very efficient. This patch cache parent psi_group in struct psi_group, only need to get psi_group of task itself first, then just use group->parent to iterate. Signed-off-by: Chengming Zhou Signed-off-by: Peter Zijlstra (Intel) Acked-by: Johannes Weiner Link: https://lore.kernel.org/r/20220825164111.29534-10-zhouchengming@bytedance.com --- include/linux/psi_types.h | 2 ++ kernel/sched/psi.c | 49 +++++++++++++++------------------------ 2 files changed, 21 insertions(+), 30 deletions(-) diff --git a/include/linux/psi_types.h b/include/linux/psi_types.h index 40c28171cd91..a0b746258c68 100644 --- a/include/linux/psi_types.h +++ b/include/linux/psi_types.h @@ -151,6 +151,8 @@ struct psi_trigger { }; struct psi_group { + struct psi_group *parent; + /* Protects data used by the aggregator */ struct mutex avgs_lock; diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c index 2545a78f82d8..9a8aee80a087 100644 --- a/kernel/sched/psi.c +++ b/kernel/sched/psi.c @@ -772,27 +772,12 @@ static void psi_group_change(struct psi_group *group, int cpu, schedule_delayed_work(&group->avgs_work, PSI_FREQ); } -static struct psi_group *iterate_groups(struct task_struct *task, void **iter) +static inline struct psi_group *task_psi_group(struct task_struct *task) { - if (*iter == &psi_system) - return NULL; - #ifdef CONFIG_CGROUPS - if (static_branch_likely(&psi_cgroups_enabled)) { - struct cgroup *cgroup = NULL; - - if (!*iter) - cgroup = task->cgroups->dfl_cgrp; - else - cgroup = cgroup_parent(*iter); - - if (cgroup && cgroup_parent(cgroup)) { - *iter = cgroup; - return cgroup_psi(cgroup); - } - } + if (static_branch_likely(&psi_cgroups_enabled)) + return cgroup_psi(task_dfl_cgroup(task)); #endif - *iter = &psi_system; return &psi_system; } @@ -815,7 +800,6 @@ void psi_task_change(struct task_struct *task, int clear, int set) { int cpu = task_cpu(task); struct psi_group *group; - void *iter = NULL; u64 now; if (!task->pid) @@ -825,8 +809,10 @@ void psi_task_change(struct task_struct *task, int clear, int set) now = cpu_clock(cpu); - while ((group = iterate_groups(task, &iter))) + group = task_psi_group(task); + do { psi_group_change(group, cpu, clear, set, now, true); + } while ((group = group->parent)); } void psi_task_switch(struct task_struct *prev, struct task_struct *next, @@ -834,7 +820,6 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, { struct psi_group *group, *common = NULL; int cpu = task_cpu(prev); - void *iter; u64 now = cpu_clock(cpu); if (next->pid) { @@ -844,8 +829,8 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, * ancestors with @prev, those will already have @prev's * TSK_ONCPU bit set, and we can stop the iteration there. */ - iter = NULL; - while ((group = iterate_groups(next, &iter))) { + group = task_psi_group(next); + do { if (per_cpu_ptr(group->pcpu, cpu)->state_mask & PSI_ONCPU) { common = group; @@ -853,7 +838,7 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, } psi_group_change(group, cpu, 0, TSK_ONCPU, now, true); - } + } while ((group = group->parent)); } if (prev->pid) { @@ -886,9 +871,12 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, psi_flags_change(prev, clear, set); - iter = NULL; - while ((group = iterate_groups(prev, &iter)) && group != common) + group = task_psi_group(prev); + do { + if (group == common) + break; psi_group_change(group, cpu, clear, set, now, wake_clock); + } while ((group = group->parent)); /* * TSK_ONCPU is handled up to the common ancestor. If there are @@ -898,7 +886,7 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, */ if ((prev->psi_flags ^ next->psi_flags) & ~TSK_ONCPU) { clear &= ~TSK_ONCPU; - for (; group; group = iterate_groups(prev, &iter)) + for (; group; group = group->parent) psi_group_change(group, cpu, clear, set, now, wake_clock); } } @@ -908,7 +896,6 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, void psi_account_irqtime(struct task_struct *task, u32 delta) { int cpu = task_cpu(task); - void *iter = NULL; struct psi_group *group; struct psi_group_cpu *groupc; u64 now; @@ -918,7 +905,8 @@ void psi_account_irqtime(struct task_struct *task, u32 delta) now = cpu_clock(cpu); - while ((group = iterate_groups(task, &iter))) { + group = task_psi_group(task); + do { groupc = per_cpu_ptr(group->pcpu, cpu); write_seqcount_begin(&groupc->seq); @@ -930,7 +918,7 @@ void psi_account_irqtime(struct task_struct *task, u32 delta) if (group->poll_states & (1 << PSI_IRQ_FULL)) psi_schedule_poll_work(group, 1); - } + } while ((group = group->parent)); } #endif @@ -1010,6 +998,7 @@ int psi_cgroup_alloc(struct cgroup *cgroup) return -ENOMEM; } group_init(cgroup->psi); + cgroup->psi->parent = cgroup_psi(cgroup_parent(cgroup)); return 0; } From 34f26a15611afb03c33df6819359d36f5b382589 Mon Sep 17 00:00:00 2001 From: Chengming Zhou Date: Wed, 7 Sep 2022 17:03:32 +0800 Subject: [PATCH 011/121] sched/psi: Per-cgroup PSI accounting disable/re-enable interface PSI accounts stalls for each cgroup separately and aggregates it at each level of the hierarchy. This may cause non-negligible overhead for some workloads when under deep level of the hierarchy. commit 3958e2d0c34e ("cgroup: make per-cgroup pressure stall tracking configurable") make PSI to skip per-cgroup stall accounting, only account system-wide to avoid this each level overhead. But for our use case, we also want leaf cgroup PSI stats accounted for userspace adjustment on that cgroup, apart from only system-wide adjustment. So this patch introduce a per-cgroup PSI accounting disable/re-enable interface "cgroup.pressure", which is a read-write single value file that allowed values are "0" and "1", the defaults is "1" so per-cgroup PSI stats is enabled by default. Implementation details: It should be relatively straight-forward to disable and re-enable state aggregation, time tracking, averaging on a per-cgroup level, if we can live with losing history from while it was disabled. I.e. the avgs will restart from 0, total= will have gaps. But it's hard or complex to stop/restart groupc->tasks[] updates, which is not implemented in this patch. So we always update groupc->tasks[] and PSI_ONCPU bit in psi_group_change() even when the cgroup PSI stats is disabled. Suggested-by: Johannes Weiner Suggested-by: Tejun Heo Signed-off-by: Chengming Zhou Signed-off-by: Peter Zijlstra (Intel) Acked-by: Johannes Weiner Link: https://lkml.kernel.org/r/20220907090332.2078-1-zhouchengming@bytedance.com --- Documentation/admin-guide/cgroup-v2.rst | 17 ++++++ include/linux/cgroup-defs.h | 3 ++ include/linux/psi.h | 2 + include/linux/psi_types.h | 3 ++ kernel/cgroup/cgroup.c | 70 ++++++++++++++++++++++--- kernel/sched/psi.c | 70 ++++++++++++++++++++++--- 6 files changed, 152 insertions(+), 13 deletions(-) diff --git a/Documentation/admin-guide/cgroup-v2.rst b/Documentation/admin-guide/cgroup-v2.rst index 971c418bc778..4cad4e2b31ec 100644 --- a/Documentation/admin-guide/cgroup-v2.rst +++ b/Documentation/admin-guide/cgroup-v2.rst @@ -976,6 +976,23 @@ All cgroup core files are prefixed with "cgroup." killing cgroups is a process directed operation, i.e. it affects the whole thread-group. + cgroup.pressure + A read-write single value file that allowed values are "0" and "1". + The default is "1". + + Writing "0" to the file will disable the cgroup PSI accounting. + Writing "1" to the file will re-enable the cgroup PSI accounting. + + This control attribute is not hierarchical, so disable or enable PSI + accounting in a cgroup does not affect PSI accounting in descendants + and doesn't need pass enablement via ancestors from root. + + The reason this control attribute exists is that PSI accounts stalls for + each cgroup separately and aggregates it at each level of the hierarchy. + This may cause non-negligible overhead for some workloads when under + deep level of the hierarchy, in which case this control attribute can + be used to disable PSI accounting in the non-leaf cgroups. + irq.pressure A read-write nested-keyed file. diff --git a/include/linux/cgroup-defs.h b/include/linux/cgroup-defs.h index 4bcf56b3491c..7df76b318245 100644 --- a/include/linux/cgroup-defs.h +++ b/include/linux/cgroup-defs.h @@ -428,6 +428,9 @@ struct cgroup { struct cgroup_file procs_file; /* handle for "cgroup.procs" */ struct cgroup_file events_file; /* handle for "cgroup.events" */ + /* handles for "{cpu,memory,io,irq}.pressure" */ + struct cgroup_file psi_files[NR_PSI_RESOURCES]; + /* * The bitmask of subsystems enabled on the child cgroups. * ->subtree_control is the one configured through diff --git a/include/linux/psi.h b/include/linux/psi.h index 362a74ca1d3b..b029a847def1 100644 --- a/include/linux/psi.h +++ b/include/linux/psi.h @@ -39,6 +39,7 @@ static inline struct psi_group *cgroup_psi(struct cgroup *cgrp) int psi_cgroup_alloc(struct cgroup *cgrp); void psi_cgroup_free(struct cgroup *cgrp); void cgroup_move_task(struct task_struct *p, struct css_set *to); +void psi_cgroup_restart(struct psi_group *group); #endif #else /* CONFIG_PSI */ @@ -60,6 +61,7 @@ static inline void cgroup_move_task(struct task_struct *p, struct css_set *to) { rcu_assign_pointer(p->cgroups, to); } +static inline void psi_cgroup_restart(struct psi_group *group) {} #endif #endif /* CONFIG_PSI */ diff --git a/include/linux/psi_types.h b/include/linux/psi_types.h index a0b746258c68..6e4372735068 100644 --- a/include/linux/psi_types.h +++ b/include/linux/psi_types.h @@ -152,6 +152,7 @@ struct psi_trigger { struct psi_group { struct psi_group *parent; + bool enabled; /* Protects data used by the aggregator */ struct mutex avgs_lock; @@ -194,6 +195,8 @@ struct psi_group { #else /* CONFIG_PSI */ +#define NR_PSI_RESOURCES 0 + struct psi_group { }; #endif /* CONFIG_PSI */ diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index 772b35d65d1f..fa1cf836b66a 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -3708,8 +3708,8 @@ static int cgroup_cpu_pressure_show(struct seq_file *seq, void *v) return psi_show(seq, psi, PSI_CPU); } -static ssize_t cgroup_pressure_write(struct kernfs_open_file *of, char *buf, - size_t nbytes, enum psi_res res) +static ssize_t pressure_write(struct kernfs_open_file *of, char *buf, + size_t nbytes, enum psi_res res) { struct cgroup_file_ctx *ctx = of->priv; struct psi_trigger *new; @@ -3746,21 +3746,21 @@ static ssize_t cgroup_io_pressure_write(struct kernfs_open_file *of, char *buf, size_t nbytes, loff_t off) { - return cgroup_pressure_write(of, buf, nbytes, PSI_IO); + return pressure_write(of, buf, nbytes, PSI_IO); } static ssize_t cgroup_memory_pressure_write(struct kernfs_open_file *of, char *buf, size_t nbytes, loff_t off) { - return cgroup_pressure_write(of, buf, nbytes, PSI_MEM); + return pressure_write(of, buf, nbytes, PSI_MEM); } static ssize_t cgroup_cpu_pressure_write(struct kernfs_open_file *of, char *buf, size_t nbytes, loff_t off) { - return cgroup_pressure_write(of, buf, nbytes, PSI_CPU); + return pressure_write(of, buf, nbytes, PSI_CPU); } #ifdef CONFIG_IRQ_TIME_ACCOUNTING @@ -3776,10 +3776,58 @@ static ssize_t cgroup_irq_pressure_write(struct kernfs_open_file *of, char *buf, size_t nbytes, loff_t off) { - return cgroup_pressure_write(of, buf, nbytes, PSI_IRQ); + return pressure_write(of, buf, nbytes, PSI_IRQ); } #endif +static int cgroup_pressure_show(struct seq_file *seq, void *v) +{ + struct cgroup *cgrp = seq_css(seq)->cgroup; + struct psi_group *psi = cgroup_psi(cgrp); + + seq_printf(seq, "%d\n", psi->enabled); + + return 0; +} + +static ssize_t cgroup_pressure_write(struct kernfs_open_file *of, + char *buf, size_t nbytes, + loff_t off) +{ + ssize_t ret; + int enable; + struct cgroup *cgrp; + struct psi_group *psi; + + ret = kstrtoint(strstrip(buf), 0, &enable); + if (ret) + return ret; + + if (enable < 0 || enable > 1) + return -ERANGE; + + cgrp = cgroup_kn_lock_live(of->kn, false); + if (!cgrp) + return -ENOENT; + + psi = cgroup_psi(cgrp); + if (psi->enabled != enable) { + int i; + + /* show or hide {cpu,memory,io,irq}.pressure files */ + for (i = 0; i < NR_PSI_RESOURCES; i++) + cgroup_file_show(&cgrp->psi_files[i], enable); + + psi->enabled = enable; + if (enable) + psi_cgroup_restart(psi); + } + + cgroup_kn_unlock(of->kn); + + return nbytes; +} + static __poll_t cgroup_pressure_poll(struct kernfs_open_file *of, poll_table *pt) { @@ -5175,6 +5223,7 @@ static struct cftype cgroup_base_files[] = { { .name = "io.pressure", .flags = CFTYPE_PRESSURE, + .file_offset = offsetof(struct cgroup, psi_files[PSI_IO]), .seq_show = cgroup_io_pressure_show, .write = cgroup_io_pressure_write, .poll = cgroup_pressure_poll, @@ -5183,6 +5232,7 @@ static struct cftype cgroup_base_files[] = { { .name = "memory.pressure", .flags = CFTYPE_PRESSURE, + .file_offset = offsetof(struct cgroup, psi_files[PSI_MEM]), .seq_show = cgroup_memory_pressure_show, .write = cgroup_memory_pressure_write, .poll = cgroup_pressure_poll, @@ -5191,6 +5241,7 @@ static struct cftype cgroup_base_files[] = { { .name = "cpu.pressure", .flags = CFTYPE_PRESSURE, + .file_offset = offsetof(struct cgroup, psi_files[PSI_CPU]), .seq_show = cgroup_cpu_pressure_show, .write = cgroup_cpu_pressure_write, .poll = cgroup_pressure_poll, @@ -5200,12 +5251,19 @@ static struct cftype cgroup_base_files[] = { { .name = "irq.pressure", .flags = CFTYPE_PRESSURE, + .file_offset = offsetof(struct cgroup, psi_files[PSI_IRQ]), .seq_show = cgroup_irq_pressure_show, .write = cgroup_irq_pressure_write, .poll = cgroup_pressure_poll, .release = cgroup_pressure_release, }, #endif + { + .name = "cgroup.pressure", + .flags = CFTYPE_PRESSURE, + .seq_show = cgroup_pressure_show, + .write = cgroup_pressure_write, + }, #endif /* CONFIG_PSI */ { } /* terminate */ }; diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c index 9a8aee80a087..9711827e31e5 100644 --- a/kernel/sched/psi.c +++ b/kernel/sched/psi.c @@ -181,6 +181,7 @@ static void group_init(struct psi_group *group) { int cpu; + group->enabled = true; for_each_possible_cpu(cpu) seqcount_init(&per_cpu_ptr(group->pcpu, cpu)->seq); group->avg_last_update = sched_clock(); @@ -696,17 +697,16 @@ static void psi_group_change(struct psi_group *group, int cpu, groupc = per_cpu_ptr(group->pcpu, cpu); /* - * First we assess the aggregate resource states this CPU's - * tasks have been in since the last change, and account any - * SOME and FULL time these may have resulted in. - * - * Then we update the task counts according to the state + * First we update the task counts according to the state * change requested through the @clear and @set bits. + * + * Then if the cgroup PSI stats accounting enabled, we + * assess the aggregate resource states this CPU's tasks + * have been in since the last change, and account any + * SOME and FULL time these may have resulted in. */ write_seqcount_begin(&groupc->seq); - record_times(groupc, now); - /* * Start with TSK_ONCPU, which doesn't have a corresponding * task count - it's just a boolean flag directly encoded in @@ -745,6 +745,23 @@ static void psi_group_change(struct psi_group *group, int cpu, if (set & (1 << t)) groupc->tasks[t]++; + if (!group->enabled) { + /* + * On the first group change after disabling PSI, conclude + * the current state and flush its time. This is unlikely + * to matter to the user, but aggregation (get_recent_times) + * may have already incorporated the live state into times_prev; + * avoid a delta sample underflow when PSI is later re-enabled. + */ + if (unlikely(groupc->state_mask & (1 << PSI_NONIDLE))) + record_times(groupc, now); + + groupc->state_mask = state_mask; + + write_seqcount_end(&groupc->seq); + return; + } + for (s = 0; s < NR_PSI_STATES; s++) { if (test_state(groupc->tasks, s, state_mask & PSI_ONCPU)) state_mask |= (1 << s); @@ -761,6 +778,8 @@ static void psi_group_change(struct psi_group *group, int cpu, if (unlikely((state_mask & PSI_ONCPU) && cpu_curr(cpu)->in_memstall)) state_mask |= (1 << PSI_MEM_FULL); + record_times(groupc, now); + groupc->state_mask = state_mask; write_seqcount_end(&groupc->seq); @@ -907,6 +926,9 @@ void psi_account_irqtime(struct task_struct *task, u32 delta) group = task_psi_group(task); do { + if (!group->enabled) + continue; + groupc = per_cpu_ptr(group->pcpu, cpu); write_seqcount_begin(&groupc->seq); @@ -1080,6 +1102,40 @@ void cgroup_move_task(struct task_struct *task, struct css_set *to) task_rq_unlock(rq, task, &rf); } + +void psi_cgroup_restart(struct psi_group *group) +{ + int cpu; + + /* + * After we disable psi_group->enabled, we don't actually + * stop percpu tasks accounting in each psi_group_cpu, + * instead only stop test_state() loop, record_times() + * and averaging worker, see psi_group_change() for details. + * + * When disable cgroup PSI, this function has nothing to sync + * since cgroup pressure files are hidden and percpu psi_group_cpu + * would see !psi_group->enabled and only do task accounting. + * + * When re-enable cgroup PSI, this function use psi_group_change() + * to get correct state mask from test_state() loop on tasks[], + * and restart groupc->state_start from now, use .clear = .set = 0 + * here since no task status really changed. + */ + if (!group->enabled) + return; + + for_each_possible_cpu(cpu) { + struct rq *rq = cpu_rq(cpu); + struct rq_flags rf; + u64 now; + + rq_lock_irq(rq, &rf); + now = cpu_clock(cpu); + psi_group_change(group, cpu, 0, 0, now, true); + rq_unlock_irq(rq, &rf); + } +} #endif /* CONFIG_CGROUPS */ int psi_show(struct seq_file *m, struct psi_group *group, enum psi_res res) From fd6dd9584ed3ee6debf2e7f9c9e69ef09b368277 Mon Sep 17 00:00:00 2001 From: Bernardo Rodrigues Date: Sun, 5 Dec 2021 18:00:49 -0300 Subject: [PATCH 012/121] leds: pca963x: fix blink with hw acceleration LEDs would behave differently depending on the blink hardware acceleration configuration. This commit will make LEDs respond exactly the same independently of the hardware acceleration status. In other words, if you had two pca963x, side by side, one with blink hardware acceleration "ON" and the other "OFF; and performed some arbitrary sequence of API calls (e.g. turn on/off, change brightness, change blink mode, etc.) you probably would end with not matching LED states. 'pca963x software blink' and 'leds-gpio' behavior were used as reference. Actual chip used to validate this change: pca9634 Some of the unmatched behaviors being fixed are (when hw blink was "ON") - Leds would stop blinking when the brightness was changed. - Leds would persist their blinking mode even after being turned off (brightness = 0). - Leds would only blink if another led was solid (pca963x will be forced out of low power) Signed-off-by: Bernardo Rodrigues Signed-off-by: Pavel Machek --- drivers/leds/leds-pca963x.c | 22 ++++++++++++++++++++-- 1 file changed, 20 insertions(+), 2 deletions(-) diff --git a/drivers/leds/leds-pca963x.c b/drivers/leds/leds-pca963x.c index 00aecd67e348..d8d866bcda19 100644 --- a/drivers/leds/leds-pca963x.c +++ b/drivers/leds/leds-pca963x.c @@ -101,6 +101,7 @@ struct pca963x_led { struct pca963x *chip; struct led_classdev led_cdev; int led_num; /* 0 .. 15 potentially */ + bool blinking; u8 gdc; u8 gfrq; }; @@ -129,12 +130,21 @@ static int pca963x_brightness(struct pca963x_led *led, switch (brightness) { case LED_FULL: - val = (ledout & ~mask) | (PCA963X_LED_ON << shift); + if (led->blinking) { + val = (ledout & ~mask) | (PCA963X_LED_GRP_PWM << shift); + ret = i2c_smbus_write_byte_data(client, + PCA963X_PWM_BASE + + led->led_num, + LED_FULL); + } else { + val = (ledout & ~mask) | (PCA963X_LED_ON << shift); + } ret = i2c_smbus_write_byte_data(client, ledout_addr, val); break; case LED_OFF: val = ledout & ~mask; ret = i2c_smbus_write_byte_data(client, ledout_addr, val); + led->blinking = false; break; default: ret = i2c_smbus_write_byte_data(client, @@ -144,7 +154,11 @@ static int pca963x_brightness(struct pca963x_led *led, if (ret < 0) return ret; - val = (ledout & ~mask) | (PCA963X_LED_PWM << shift); + if (led->blinking) + val = (ledout & ~mask) | (PCA963X_LED_GRP_PWM << shift); + else + val = (ledout & ~mask) | (PCA963X_LED_PWM << shift); + ret = i2c_smbus_write_byte_data(client, ledout_addr, val); break; } @@ -181,6 +195,7 @@ static void pca963x_blink(struct pca963x_led *led) } mutex_unlock(&led->chip->mutex); + led->blinking = true; } static int pca963x_power_state(struct pca963x_led *led) @@ -275,6 +290,8 @@ static int pca963x_blink_set(struct led_classdev *led_cdev, led->gfrq = gfrq; pca963x_blink(led); + led->led_cdev.brightness = LED_FULL; + pca963x_led_set(led_cdev, LED_FULL); *delay_on = time_on; *delay_off = time_off; @@ -337,6 +354,7 @@ static int pca963x_register_leds(struct i2c_client *client, led->led_cdev.brightness_set_blocking = pca963x_led_set; if (hw_blink) led->led_cdev.blink_set = pca963x_blink_set; + led->blinking = false; init_data.fwnode = child; /* for backwards compatibility */ From 31fd7108302388d732973c58470d4be559d352ec Mon Sep 17 00:00:00 2001 From: Marek Vasut Date: Thu, 17 Feb 2022 18:43:57 +0100 Subject: [PATCH 013/121] dt-bindings: leds: Document mmc trigger The mmc subsystem supports triggering leds on card activity, document the trigger value here. The value is a pattern in this case. Signed-off-by: Marek Vasut Cc: Jacek Anaszewski Cc: Pavel Machek Cc: Rob Herring Cc: devicetree@vger.kernel.org To: linux-leds@vger.kernel.org Signed-off-by: Pavel Machek Reviewed-by: Rob Herring --- .../devicetree/bindings/leds/common.yaml | 39 ++++++++++--------- 1 file changed, 21 insertions(+), 18 deletions(-) diff --git a/Documentation/devicetree/bindings/leds/common.yaml b/Documentation/devicetree/bindings/leds/common.yaml index 328952d7acbb..3c14a98430e1 100644 --- a/Documentation/devicetree/bindings/leds/common.yaml +++ b/Documentation/devicetree/bindings/leds/common.yaml @@ -79,24 +79,27 @@ properties: the LED. $ref: /schemas/types.yaml#/definitions/string - enum: - # LED will act as a back-light, controlled by the framebuffer system - - backlight - # LED will turn on (but for leds-gpio see "default-state" property in - # Documentation/devicetree/bindings/leds/leds-gpio.yaml) - - default-on - # LED "double" flashes at a load average based rate - - heartbeat - # LED indicates disk activity - - disk-activity - # LED indicates IDE disk activity (deprecated), in new implementations - # use "disk-activity" - - ide-disk - # LED flashes at a fixed, configurable rate - - timer - # LED alters the brightness for the specified duration with one software - # timer (requires "led-pattern" property) - - pattern + oneOf: + - enum: + # LED will act as a back-light, controlled by the framebuffer system + - backlight + # LED will turn on (but for leds-gpio see "default-state" property in + # Documentation/devicetree/bindings/leds/leds-gpio.yaml) + - default-on + # LED "double" flashes at a load average based rate + - heartbeat + # LED indicates disk activity + - disk-activity + # LED indicates IDE disk activity (deprecated), in new implementations + # use "disk-activity" + - ide-disk + # LED flashes at a fixed, configurable rate + - timer + # LED alters the brightness for the specified duration with one software + # timer (requires "led-pattern" property) + - pattern + # LED is triggered by SD/MMC activity + - pattern: "^mmc[0-9]+$" led-pattern: description: | From 7ab72c597356be1e7f0f3d856e54ce78527f43c8 Mon Sep 17 00:00:00 2001 From: Andrew Bresticker Date: Thu, 15 Sep 2022 15:37:01 -0400 Subject: [PATCH 014/121] riscv: Make VM_WRITE imply VM_READ RISC-V does not presently have write-only mappings as that PTE bit pattern is considered reserved in the privileged spec, so allow handling of read faults in VMAs that have VM_WRITE without VM_READ in order to be consistent with other architectures that have similar limitations. Fixes: 2139619bcad7 ("riscv: mmap with PROT_WRITE but no PROT_READ is invalid") Reviewed-by: Atish Patra Signed-off-by: Andrew Bresticker Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20220915193702.2201018-2-abrestic@rivosinc.com/ Signed-off-by: Palmer Dabbelt --- arch/riscv/mm/fault.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/riscv/mm/fault.c b/arch/riscv/mm/fault.c index f2fbd1400b7c..d86f7cebd4a7 100644 --- a/arch/riscv/mm/fault.c +++ b/arch/riscv/mm/fault.c @@ -184,7 +184,8 @@ static inline bool access_error(unsigned long cause, struct vm_area_struct *vma) } break; case EXC_LOAD_PAGE_FAULT: - if (!(vma->vm_flags & VM_READ)) { + /* Write implies read */ + if (!(vma->vm_flags & (VM_READ | VM_WRITE))) { return true; } break; From 9e2e6042a7ec6504fe8e366717afa2f40cf16488 Mon Sep 17 00:00:00 2001 From: Andrew Bresticker Date: Thu, 15 Sep 2022 15:37:02 -0400 Subject: [PATCH 015/121] riscv: Allow PROT_WRITE-only mmap() Commit 2139619bcad7 ("riscv: mmap with PROT_WRITE but no PROT_READ is invalid") made mmap() return EINVAL if PROT_WRITE was set wihtout PROT_READ with the justification that a write-only PTE is considered a reserved PTE permission bit pattern in the privileged spec. This check is unnecessary since we let VM_WRITE imply VM_READ on RISC-V, and it is inconsistent with other architectures that don't support write-only PTEs, creating a potential software portability issue. Just remove the check altogether and let PROT_WRITE imply PROT_READ as is the case on other architectures. Note that this also allows PROT_WRITE|PROT_EXEC mappings which were disallowed prior to the aforementioned commit; PROT_READ is implied in such mappings as well. Fixes: 2139619bcad7 ("riscv: mmap with PROT_WRITE but no PROT_READ is invalid") Reviewed-by: Atish Patra Signed-off-by: Andrew Bresticker Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20220915193702.2201018-3-abrestic@rivosinc.com/ Signed-off-by: Palmer Dabbelt --- arch/riscv/kernel/sys_riscv.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/arch/riscv/kernel/sys_riscv.c b/arch/riscv/kernel/sys_riscv.c index 571556bb9261..5d3f2fbeb33c 100644 --- a/arch/riscv/kernel/sys_riscv.c +++ b/arch/riscv/kernel/sys_riscv.c @@ -18,9 +18,6 @@ static long riscv_sys_mmap(unsigned long addr, unsigned long len, if (unlikely(offset & (~PAGE_MASK >> page_shift_offset))) return -EINVAL; - if (unlikely((prot & PROT_WRITE) && !(prot & PROT_READ))) - return -EINVAL; - return ksys_mmap_pgoff(addr, len, prot, flags, fd, offset >> (PAGE_SHIFT - page_shift_offset)); } From 4d1632151bde847230a0bd2318806380d309655f Mon Sep 17 00:00:00 2001 From: Pavel Machek Date: Mon, 26 Sep 2022 23:16:37 +0200 Subject: [PATCH 016/121] leds: pca963: fix misleading indentation I'm getting warnings: /tmp/next/build/drivers/leds/leds-pca963x.c: In function 'pca963x_register_leds': /tmp/next/build/drivers/leds/leds-pca963x.c:355:3: error: this 'if' clause does not guard... +[-Werror=misleading-indentation] 355 | if (hw_blink) | ^~ /tmp/next/build/drivers/leds/leds-pca963x.c:357:4: note: ...this statement, but the latter is +misleadingly indented as if it were guarded by the 'if' 357 | led->blinking = false; | ^~~ cc1: all warnings being treated as errors Fix the indentation to make them go away. Signed-off-by: Pavel Machek --- drivers/leds/leds-pca963x.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/leds/leds-pca963x.c b/drivers/leds/leds-pca963x.c index d8d866bcda19..a7e052c1db53 100644 --- a/drivers/leds/leds-pca963x.c +++ b/drivers/leds/leds-pca963x.c @@ -354,7 +354,7 @@ static int pca963x_register_leds(struct i2c_client *client, led->led_cdev.brightness_set_blocking = pca963x_led_set; if (hw_blink) led->led_cdev.blink_set = pca963x_blink_set; - led->blinking = false; + led->blinking = false; init_data.fwnode = child; /* for backwards compatibility */ From a0d49a8f77f26609036a05e7832393b6279554db Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 27 Sep 2022 12:19:13 +0100 Subject: [PATCH 017/121] dt-bindings: riscv: microchip: document icicle reference design The icicle kit reference design's v2022.09 release made some changes to the memory map - including adding the ability to read the fabric clock controllers via the system controller bus & making the PCI controller work with upstream Linux. While the PCI was not working in the v2022.03 design, so nothing is broken there in terms of backwards compatibility, the fabric clocks used in the v2022.03 design were chosen by the individual run of the synthesis tool. In the v2022.09 reference design, the clocks are fixed to use the "north west" fabric Clock Conditioning Circuitry. In the v2022.10 release, the memory map on the DDR side is also changing, so to avoid making a breaking change here twice, jump over the v2022.09 release and straight to the v2022.10 one. Make use of a new compatible to denote that v2022.{09,10} reference design releases are not backwards compatible. Acked-by: Krzysztof Kozlowski Signed-off-by: Conor Dooley --- .../devicetree/bindings/riscv/microchip.yaml | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) diff --git a/Documentation/devicetree/bindings/riscv/microchip.yaml b/Documentation/devicetree/bindings/riscv/microchip.yaml index 1aa7336a9672..5c1ad2108049 100644 --- a/Documentation/devicetree/bindings/riscv/microchip.yaml +++ b/Documentation/devicetree/bindings/riscv/microchip.yaml @@ -17,12 +17,18 @@ properties: $nodename: const: '/' compatible: - items: - - enum: - - microchip,mpfs-icicle-kit - - microchip,mpfs-icicle-reference-rtlv2203 - - sundance,polarberry - - const: microchip,mpfs + oneOf: + - items: + - enum: + - microchip,mpfs-icicle-reference-rtlv2203 + - microchip,mpfs-icicle-reference-rtlv2210 + - const: microchip,mpfs-icicle-kit + - const: microchip,mpfs + + - items: + - enum: + - sundance,polarberry + - const: microchip,mpfs additionalProperties: true From 0ebdc51787dbb8ef8d259daa98b8fd35babf8970 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 27 Sep 2022 12:19:14 +0100 Subject: [PATCH 018/121] dt-bindings: riscv: microchip: document the aries m100pfsevp Add a compatible for the Aries Embedded M100PFSEVP SOM + EVK platform. Link: https://www.aries-embedded.com/polarfire-soc-fpga-microsemi-m100pfs-som-mpfs025t-pcie-serdes Signed-off-by: Conor Dooley Acked-by: Rob Herring --- Documentation/devicetree/bindings/riscv/microchip.yaml | 1 + 1 file changed, 1 insertion(+) diff --git a/Documentation/devicetree/bindings/riscv/microchip.yaml b/Documentation/devicetree/bindings/riscv/microchip.yaml index 5c1ad2108049..681cedc5578f 100644 --- a/Documentation/devicetree/bindings/riscv/microchip.yaml +++ b/Documentation/devicetree/bindings/riscv/microchip.yaml @@ -27,6 +27,7 @@ properties: - items: - enum: + - aries,m100pfsevp - sundance,polarberry - const: microchip,mpfs From db3d481698efe8a7a943ecbba9491648c5a49ef3 Mon Sep 17 00:00:00 2001 From: Shravan Chippa Date: Tue, 27 Sep 2022 12:19:15 +0100 Subject: [PATCH 019/121] dt-bindings: riscv: microchip: document the sev kit Update devicetree bindings document with PolarFire SoC Video Kit, known by its "sev-kit" product code. Link: https://onlinedocs.microchip.com/pr/GUID-404D3738-DC76-46BA-8683-6A77E837C2DD-en-US-1/index.html?GUID-065AEBEE-7B2C-4895-8579-B1D73D797F06 Signed-off-by: Shravan Chippa Reviewed-by: Krzysztof Kozlowski Signed-off-by: Conor Dooley --- Documentation/devicetree/bindings/riscv/microchip.yaml | 1 + 1 file changed, 1 insertion(+) diff --git a/Documentation/devicetree/bindings/riscv/microchip.yaml b/Documentation/devicetree/bindings/riscv/microchip.yaml index 681cedc5578f..2b8c6a695e99 100644 --- a/Documentation/devicetree/bindings/riscv/microchip.yaml +++ b/Documentation/devicetree/bindings/riscv/microchip.yaml @@ -28,6 +28,7 @@ properties: - items: - enum: - aries,m100pfsevp + - microchip,mpfs-sev-kit - sundance,polarberry - const: microchip,mpfs From f890e67f292db46c9bd5b5c004ba0f98761d1a33 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 27 Sep 2022 12:19:16 +0100 Subject: [PATCH 020/121] riscv: dts: microchip: add pci dma ranges for the icicle kit The recently removed, accidentally included, "matr0" property was used in place of a dma-ranges property. The PCI controller is non-functional with mainline Linux in the v2022.02 or later reference designs and has not worked without configuration of address-translation since v2021.08. Add the address translation that will be used by the v2022.09 reference design & update the compatible used by the dts. Since this change is not backwards compatible, update the compatible to denote this, jumping over v2022.09 directly to v2022.10. Signed-off-by: Conor Dooley --- arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi | 7 ++++++- arch/riscv/boot/dts/microchip/mpfs-icicle-kit.dts | 3 ++- 2 files changed, 8 insertions(+), 2 deletions(-) diff --git a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi index 0d28858b83f2..eec5aba43436 100644 --- a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi +++ b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi @@ -2,7 +2,8 @@ /* Copyright (c) 2020-2021 Microchip Technology Inc */ / { - compatible = "microchip,mpfs-icicle-reference-rtlv2203", "microchip,mpfs"; + compatible = "microchip,mpfs-icicle-reference-rtlv2210", "microchip,mpfs-icicle-kit", + "microchip,mpfs"; core_pwm0: pwm@41000000 { compatible = "microchip,corepwm-rtl-v4"; @@ -37,3 +38,7 @@ fabric_clk1: fabric-clk1 { clock-frequency = <125000000>; }; }; + +&pcie { + dma-ranges = <0x02000000 0x0 0x00000000 0x0 0x00000000 0x1 0x00000000>; +}; diff --git a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit.dts b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit.dts index 044982a11df5..42d350fe6c6b 100644 --- a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit.dts +++ b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit.dts @@ -11,7 +11,8 @@ / { model = "Microchip PolarFire-SoC Icicle Kit"; - compatible = "microchip,mpfs-icicle-kit", "microchip,mpfs"; + compatible = "microchip,mpfs-icicle-reference-rtlv2210", "microchip,mpfs-icicle-kit", + "microchip,mpfs"; aliases { ethernet0 = &mac1; From 99d451a7db1624308bc9eb94b7befb3722f67b10 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 27 Sep 2022 12:19:17 +0100 Subject: [PATCH 021/121] riscv: dts: microchip: move the mpfs' pci node to -fabric.dtsi In today's edition of moving things around: The PCIe root port on PolarFire SoC is more part of the FPGA than of the Core Complex. It is located on the other side of the chip and, apart from its interrupts, most of its configuration is determined by the FPGA bitstream rather. This includes: - address translation in both directions - the addresses at which the config and data regions appear to the core complex - the clocks used by the AXI bus - the plic interrupt used Moving the PCIe node to the -fabric.dtsi makes it clearer than a singular configuration for root port is not correct & allows the base SoC dtsi to be more easily included. Signed-off-by: Conor Dooley --- .../dts/microchip/mpfs-icicle-kit-fabric.dtsi | 32 +++++++++++++++++-- .../dts/microchip/mpfs-polarberry-fabric.dtsi | 29 +++++++++++++++++ arch/riscv/boot/dts/microchip/mpfs.dtsi | 30 ----------------- 3 files changed, 58 insertions(+), 33 deletions(-) diff --git a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi index eec5aba43436..688ef0fc5a64 100644 --- a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi +++ b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi @@ -37,8 +37,34 @@ fabric_clk1: fabric-clk1 { #clock-cells = <0>; clock-frequency = <125000000>; }; -}; -&pcie { - dma-ranges = <0x02000000 0x0 0x00000000 0x0 0x00000000 0x1 0x00000000>; + pcie: pcie@2000000000 { + compatible = "microchip,pcie-host-1.0"; + #address-cells = <0x3>; + #interrupt-cells = <0x1>; + #size-cells = <0x2>; + device_type = "pci"; + reg = <0x20 0x0 0x0 0x8000000>, <0x0 0x43000000 0x0 0x10000>; + reg-names = "cfg", "apb"; + bus-range = <0x0 0x7f>; + interrupt-parent = <&plic>; + interrupts = <119>; + interrupt-map = <0 0 0 1 &pcie_intc 0>, + <0 0 0 2 &pcie_intc 1>, + <0 0 0 3 &pcie_intc 2>, + <0 0 0 4 &pcie_intc 3>; + interrupt-map-mask = <0 0 0 7>; + clocks = <&fabric_clk1>, <&fabric_clk1>, <&fabric_clk3>; + clock-names = "fic0", "fic1", "fic3"; + ranges = <0x3000000 0x0 0x8000000 0x20 0x8000000 0x0 0x80000000>; + dma-ranges = <0x02000000 0x0 0x00000000 0x0 0x00000000 0x1 0x00000000>; + msi-parent = <&pcie>; + msi-controller; + status = "disabled"; + pcie_intc: interrupt-controller { + #address-cells = <0>; + #interrupt-cells = <1>; + interrupt-controller; + }; + }; }; diff --git a/arch/riscv/boot/dts/microchip/mpfs-polarberry-fabric.dtsi b/arch/riscv/boot/dts/microchip/mpfs-polarberry-fabric.dtsi index 49380c428ec9..67303bc0e451 100644 --- a/arch/riscv/boot/dts/microchip/mpfs-polarberry-fabric.dtsi +++ b/arch/riscv/boot/dts/microchip/mpfs-polarberry-fabric.dtsi @@ -13,4 +13,33 @@ fabric_clk1: fabric-clk1 { #clock-cells = <0>; clock-frequency = <125000000>; }; + + pcie: pcie@2000000000 { + compatible = "microchip,pcie-host-1.0"; + #address-cells = <0x3>; + #interrupt-cells = <0x1>; + #size-cells = <0x2>; + device_type = "pci"; + reg = <0x20 0x0 0x0 0x8000000>, <0x0 0x43000000 0x0 0x10000>; + reg-names = "cfg", "apb"; + bus-range = <0x0 0x7f>; + interrupt-parent = <&plic>; + interrupts = <119>; + interrupt-map = <0 0 0 1 &pcie_intc 0>, + <0 0 0 2 &pcie_intc 1>, + <0 0 0 3 &pcie_intc 2>, + <0 0 0 4 &pcie_intc 3>; + interrupt-map-mask = <0 0 0 7>; + clocks = <&fabric_clk1>, <&fabric_clk1>, <&fabric_clk3>; + clock-names = "fic0", "fic1", "fic3"; + ranges = <0x3000000 0x0 0x8000000 0x20 0x8000000 0x0 0x80000000>; + msi-parent = <&pcie>; + msi-controller; + status = "disabled"; + pcie_intc: interrupt-controller { + #address-cells = <0>; + #interrupt-cells = <1>; + interrupt-controller; + }; + }; }; diff --git a/arch/riscv/boot/dts/microchip/mpfs.dtsi b/arch/riscv/boot/dts/microchip/mpfs.dtsi index 45e3cc659882..79fd8dfce96f 100644 --- a/arch/riscv/boot/dts/microchip/mpfs.dtsi +++ b/arch/riscv/boot/dts/microchip/mpfs.dtsi @@ -464,36 +464,6 @@ usb: usb@20201000 { status = "disabled"; }; - pcie: pcie@2000000000 { - compatible = "microchip,pcie-host-1.0"; - #address-cells = <0x3>; - #interrupt-cells = <0x1>; - #size-cells = <0x2>; - device_type = "pci"; - reg = <0x20 0x0 0x0 0x8000000>, <0x0 0x43000000 0x0 0x10000>; - reg-names = "cfg", "apb"; - bus-range = <0x0 0x7f>; - interrupt-parent = <&plic>; - interrupts = <119>; - interrupt-map = <0 0 0 1 &pcie_intc 0>, - <0 0 0 2 &pcie_intc 1>, - <0 0 0 3 &pcie_intc 2>, - <0 0 0 4 &pcie_intc 3>; - interrupt-map-mask = <0 0 0 7>; - clocks = <&fabric_clk1>, <&fabric_clk1>, <&fabric_clk3>; - clock-names = "fic0", "fic1", "fic3"; - ranges = <0x3000000 0x0 0x8000000 0x20 0x8000000 0x0 0x80000000>; - msi-parent = <&pcie>; - msi-controller; - microchip,axi-m-atr0 = <0x10 0x0>; - status = "disabled"; - pcie_intc: legacy-interrupt-controller { - #address-cells = <0>; - #interrupt-cells = <1>; - interrupt-controller; - }; - }; - mbox: mailbox@37020000 { compatible = "microchip,mpfs-mailbox"; reg = <0x0 0x37020000 0x0 0x1000>, <0x0 0x2000318C 0x0 0x40>; From 6fc655ed4986f88b91e3f7b339222fc1c4ffba08 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 27 Sep 2022 12:19:18 +0100 Subject: [PATCH 022/121] riscv: dts: microchip: icicle: update pci address properties For the v2022.09 reference design the PCI root port's data region has been moved to FIC1 from FIC0. This is a shorter path, allowing for higher clock rates and improved through-put. As a result, the address at which the PCIe's data region appears to the core complex has changed. The config region's address is unchanged. As FIC0 is no longer used, its clock can be removed too. Signed-off-by: Conor Dooley --- .../boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi index 688ef0fc5a64..9ca2ac4ad8e2 100644 --- a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi +++ b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi @@ -38,13 +38,13 @@ fabric_clk1: fabric-clk1 { clock-frequency = <125000000>; }; - pcie: pcie@2000000000 { + pcie: pcie@3000000000 { compatible = "microchip,pcie-host-1.0"; #address-cells = <0x3>; #interrupt-cells = <0x1>; #size-cells = <0x2>; device_type = "pci"; - reg = <0x20 0x0 0x0 0x8000000>, <0x0 0x43000000 0x0 0x10000>; + reg = <0x30 0x0 0x0 0x8000000>, <0x0 0x43000000 0x0 0x10000>; reg-names = "cfg", "apb"; bus-range = <0x0 0x7f>; interrupt-parent = <&plic>; @@ -54,9 +54,9 @@ pcie: pcie@2000000000 { <0 0 0 3 &pcie_intc 2>, <0 0 0 4 &pcie_intc 3>; interrupt-map-mask = <0 0 0 7>; - clocks = <&fabric_clk1>, <&fabric_clk1>, <&fabric_clk3>; - clock-names = "fic0", "fic1", "fic3"; - ranges = <0x3000000 0x0 0x8000000 0x20 0x8000000 0x0 0x80000000>; + clocks = <&fabric_clk1>, <&fabric_clk3>; + clock-names = "fic1", "fic3"; + ranges = <0x3000000 0x0 0x8000000 0x30 0x8000000 0x0 0x80000000>; dma-ranges = <0x02000000 0x0 0x00000000 0x0 0x00000000 0x1 0x00000000>; msi-parent = <&pcie>; msi-controller; From ab291621a8b85269496ae9a964b6d49cd1e030c8 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 27 Sep 2022 12:19:19 +0100 Subject: [PATCH 023/121] riscv: dts: microchip: icicle: re-jig fabric peripheral addresses When users try to add onto the reference design, they find that the current addresses that peripherals connected to Fabric InterConnect (FIC) 3 use are restrictive. For the v2022.09 reference design, the peripherals have been shifted down, leaving more contiguous address space for their custom IP/peripherals. Signed-off-by: Conor Dooley --- arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi index 9ca2ac4ad8e2..35030ea330ee 100644 --- a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi +++ b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi @@ -5,18 +5,18 @@ / { compatible = "microchip,mpfs-icicle-reference-rtlv2210", "microchip,mpfs-icicle-kit", "microchip,mpfs"; - core_pwm0: pwm@41000000 { + core_pwm0: pwm@40000000 { compatible = "microchip,corepwm-rtl-v4"; - reg = <0x0 0x41000000 0x0 0xF0>; + reg = <0x0 0x40000000 0x0 0xF0>; microchip,sync-update-mask = /bits/ 32 <0>; #pwm-cells = <2>; clocks = <&fabric_clk3>; status = "disabled"; }; - i2c2: i2c@44000000 { + i2c2: i2c@40000200 { compatible = "microchip,corei2c-rtl-v7"; - reg = <0x0 0x44000000 0x0 0x1000>; + reg = <0x0 0x40000200 0x0 0x1000>; #address-cells = <1>; #size-cells = <0>; clocks = <&fabric_clk3>; From fa52935abef422d119dda3c10c02787a86e6289d Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 27 Sep 2022 12:19:20 +0100 Subject: [PATCH 024/121] riscv: dts: microchip: reduce the fic3 clock rate For the v2022.09 release of the reference design, the fic3 clock rate been reduced from 62.5 MHz to 50 MHz as it allows timing to be closed significantly more quickly by customers who chose to build the reference design themselves. Signed-off-by: Conor Dooley --- arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi index 35030ea330ee..b6bfe177ccb2 100644 --- a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi +++ b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi @@ -29,7 +29,7 @@ i2c2: i2c@40000200 { fabric_clk3: fabric-clk3 { compatible = "fixed-clock"; #clock-cells = <0>; - clock-frequency = <62500000>; + clock-frequency = <50000000>; }; fabric_clk1: fabric-clk1 { From 978a17d1a688db025275d282665ab3f39407191d Mon Sep 17 00:00:00 2001 From: Vattipalli Praveen Date: Tue, 27 Sep 2022 12:19:21 +0100 Subject: [PATCH 025/121] riscv: dts: microchip: add sevkit device tree Add a basic dts for the Microchip Smart Embedded Vision dev kit. The SEV kit is an upcoming first party board, featuring an MPFS250T and: - Dual Sony Camera Sensors (IMX334) - IEEE 802.11 b/g/n 20MHz (1x1) Wi-Fi - Bluetooth 5 Low Energy - 4 GB DDR4 x64 - 2 GB LPDDR4 x32 - 1 GB SPI Flash - 8 GB eMMC flash & SD card slot (multiplexed) - HDMI2.0 Video Input/Output - MIPI DSI Output - MIPI CSI-2 Input Link: https://onlinedocs.microchip.com/pr/GUID-404D3738-DC76-46BA-8683-6A77E837C2DD-en-US-1/index.html?GUID-065AEBEE-7B2C-4895-8579-B1D73D797F06 Signed-off-by: Vattipalli Praveen Signed-off-by: Conor Dooley --- arch/riscv/boot/dts/microchip/Makefile | 1 + .../dts/microchip/mpfs-sev-kit-fabric.dtsi | 45 ++++++ .../riscv/boot/dts/microchip/mpfs-sev-kit.dts | 145 ++++++++++++++++++ 3 files changed, 191 insertions(+) create mode 100644 arch/riscv/boot/dts/microchip/mpfs-sev-kit-fabric.dtsi create mode 100644 arch/riscv/boot/dts/microchip/mpfs-sev-kit.dts diff --git a/arch/riscv/boot/dts/microchip/Makefile b/arch/riscv/boot/dts/microchip/Makefile index 39aae7b04f1c..f18477b2e86d 100644 --- a/arch/riscv/boot/dts/microchip/Makefile +++ b/arch/riscv/boot/dts/microchip/Makefile @@ -1,4 +1,5 @@ # SPDX-License-Identifier: GPL-2.0 dtb-$(CONFIG_SOC_MICROCHIP_POLARFIRE) += mpfs-icicle-kit.dtb dtb-$(CONFIG_SOC_MICROCHIP_POLARFIRE) += mpfs-polarberry.dtb +dtb-$(CONFIG_SOC_MICROCHIP_POLARFIRE) += mpfs-sev-kit.dtb obj-$(CONFIG_BUILTIN_DTB) += $(addsuffix .o, $(dtb-y)) diff --git a/arch/riscv/boot/dts/microchip/mpfs-sev-kit-fabric.dtsi b/arch/riscv/boot/dts/microchip/mpfs-sev-kit-fabric.dtsi new file mode 100644 index 000000000000..8545baf4d129 --- /dev/null +++ b/arch/riscv/boot/dts/microchip/mpfs-sev-kit-fabric.dtsi @@ -0,0 +1,45 @@ +// SPDX-License-Identifier: (GPL-2.0 OR MIT) +/* Copyright (c) 2022 Microchip Technology Inc */ + +/ { + fabric_clk3: fabric-clk3 { + compatible = "fixed-clock"; + #clock-cells = <0>; + clock-frequency = <0>; + }; + + fabric_clk1: fabric-clk1 { + compatible = "fixed-clock"; + #clock-cells = <0>; + clock-frequency = <125000000>; + }; + + pcie: pcie@2000000000 { + compatible = "microchip,pcie-host-1.0"; + #address-cells = <0x3>; + #interrupt-cells = <0x1>; + #size-cells = <0x2>; + device_type = "pci"; + reg = <0x20 0x0 0x0 0x8000000>, <0x0 0x43000000 0x0 0x10000>; + reg-names = "cfg", "apb"; + bus-range = <0x0 0x7f>; + interrupt-parent = <&plic>; + interrupts = <119>; + interrupt-map = <0 0 0 1 &pcie_intc 0>, + <0 0 0 2 &pcie_intc 1>, + <0 0 0 3 &pcie_intc 2>, + <0 0 0 4 &pcie_intc 3>; + interrupt-map-mask = <0 0 0 7>; + clocks = <&fabric_clk1>, <&fabric_clk1>, <&fabric_clk3>; + clock-names = "fic0", "fic1", "fic3"; + ranges = <0x3000000 0x0 0x8000000 0x20 0x8000000 0x0 0x80000000>; + msi-parent = <&pcie>; + msi-controller; + status = "disabled"; + pcie_intc: interrupt-controller { + #address-cells = <0>; + #interrupt-cells = <1>; + interrupt-controller; + }; + }; +}; diff --git a/arch/riscv/boot/dts/microchip/mpfs-sev-kit.dts b/arch/riscv/boot/dts/microchip/mpfs-sev-kit.dts new file mode 100644 index 000000000000..013cb666c72d --- /dev/null +++ b/arch/riscv/boot/dts/microchip/mpfs-sev-kit.dts @@ -0,0 +1,145 @@ +// SPDX-License-Identifier: (GPL-2.0 OR MIT) +/* Copyright (c) 2022 Microchip Technology Inc */ + +/dts-v1/; + +#include "mpfs.dtsi" +#include "mpfs-sev-kit-fabric.dtsi" + +/* Clock frequency (in Hz) of the rtcclk */ +#define MTIMER_FREQ 1000000 + +/ { + #address-cells = <2>; + #size-cells = <2>; + model = "Microchip PolarFire-SoC SEV Kit"; + compatible = "microchip,mpfs-sev-kit", "microchip,mpfs"; + + aliases { + ethernet0 = &mac1; + serial0 = &mmuart0; + serial1 = &mmuart1; + serial2 = &mmuart2; + serial3 = &mmuart3; + serial4 = &mmuart4; + }; + + chosen { + stdout-path = "serial1:115200n8"; + }; + + cpus { + timebase-frequency = ; + }; + + reserved-memory { + #address-cells = <2>; + #size-cells = <2>; + ranges; + + fabricbuf0ddrc: buffer@80000000 { + compatible = "shared-dma-pool"; + reg = <0x0 0x80000000 0x0 0x2000000>; + }; + + fabricbuf1ddrnc: buffer@c4000000 { + compatible = "shared-dma-pool"; + reg = <0x0 0xc4000000 0x0 0x4000000>; + }; + + fabricbuf2ddrncwcb: buffer@d4000000 { + compatible = "shared-dma-pool"; + reg = <0x0 0xd4000000 0x0 0x4000000>; + }; + }; + + ddrc_cache: memory@1000000000 { + device_type = "memory"; + reg = <0x10 0x0 0x0 0x76000000>; + }; +}; + +&i2c0 { + status = "okay"; +}; + +&gpio2 { + interrupts = <53>, <53>, <53>, <53>, + <53>, <53>, <53>, <53>, + <53>, <53>, <53>, <53>, + <53>, <53>, <53>, <53>, + <53>, <53>, <53>, <53>, + <53>, <53>, <53>, <53>, + <53>, <53>, <53>, <53>, + <53>, <53>, <53>, <53>; + status = "okay"; +}; + +&mac0 { + status = "okay"; + phy-mode = "sgmii"; + phy-handle = <&phy0>; + phy1: ethernet-phy@9 { + reg = <9>; + }; + phy0: ethernet-phy@8 { + reg = <8>; + }; +}; + +&mac1 { + status = "okay"; + phy-mode = "sgmii"; + phy-handle = <&phy1>; +}; + +&mbox { + status = "okay"; +}; + +&mmc { + status = "okay"; + bus-width = <4>; + disable-wp; + cap-sd-highspeed; + cap-mmc-highspeed; + mmc-ddr-1_8v; + mmc-hs200-1_8v; + sd-uhs-sdr12; + sd-uhs-sdr25; + sd-uhs-sdr50; + sd-uhs-sdr104; +}; + +&mmuart1 { + status = "okay"; +}; + +&mmuart2 { + status = "okay"; +}; + +&mmuart3 { + status = "okay"; +}; + +&mmuart4 { + status = "okay"; +}; + +&refclk { + clock-frequency = <125000000>; +}; + +&rtc { + status = "okay"; +}; + +&syscontroller { + status = "okay"; +}; + +&usb { + status = "okay"; + dr_mode = "otg"; +}; From d49166646e44064b694a2e631fcdba4f814746d9 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 27 Sep 2022 12:19:22 +0100 Subject: [PATCH 026/121] riscv: dts: microchip: add a devicetree for aries' m100pfsevp Add device trees for both configs used by the Aries Embedded M100PFSEVP. The M100OFSEVP consists of a MPFS250T on a SOM, featuring: - 2GB DDR4 SDRAM dedicated to the HMS - 512MB DDR4 SDRAM dedicated to the FPGA - 32 MB SPI NOR Flash - 4 GByte eMMC and a carrier board with: - 2x Gigabit Ethernet - USB - 2x UART - 2x CAN - TFT connector - HSMC extension connector - 3x PMOD extension connectors - microSD-card slot Link: https://www.aries-embedded.com/polarfire-soc-fpga-microsemi-m100pfs-som-mpfs025t-pcie-serdes Link: https://www.aries-embedded.com/evaluation-kit/fpga/polarfire-microchip-soc-fpga-m100pfsevp-riscv-hsmc-pmod Link: https://downloads.aries-embedded.de/products/M100PFS/Hardware/M100PFSEVP-Schematics.pdf Co-developed-by: Wolfgang Grandegger Signed-off-by: Wolfgang Grandegger Signed-off-by: Conor Dooley --- arch/riscv/boot/dts/microchip/Makefile | 1 + .../dts/microchip/mpfs-m100pfs-fabric.dtsi | 45 +++++ .../boot/dts/microchip/mpfs-m100pfsevp.dts | 179 ++++++++++++++++++ 3 files changed, 225 insertions(+) create mode 100644 arch/riscv/boot/dts/microchip/mpfs-m100pfs-fabric.dtsi create mode 100644 arch/riscv/boot/dts/microchip/mpfs-m100pfsevp.dts diff --git a/arch/riscv/boot/dts/microchip/Makefile b/arch/riscv/boot/dts/microchip/Makefile index f18477b2e86d..7427a20934f3 100644 --- a/arch/riscv/boot/dts/microchip/Makefile +++ b/arch/riscv/boot/dts/microchip/Makefile @@ -1,5 +1,6 @@ # SPDX-License-Identifier: GPL-2.0 dtb-$(CONFIG_SOC_MICROCHIP_POLARFIRE) += mpfs-icicle-kit.dtb +dtb-$(CONFIG_SOC_MICROCHIP_POLARFIRE) += mpfs-m100pfsevp.dtb dtb-$(CONFIG_SOC_MICROCHIP_POLARFIRE) += mpfs-polarberry.dtb dtb-$(CONFIG_SOC_MICROCHIP_POLARFIRE) += mpfs-sev-kit.dtb obj-$(CONFIG_BUILTIN_DTB) += $(addsuffix .o, $(dtb-y)) diff --git a/arch/riscv/boot/dts/microchip/mpfs-m100pfs-fabric.dtsi b/arch/riscv/boot/dts/microchip/mpfs-m100pfs-fabric.dtsi new file mode 100644 index 000000000000..7b9ee13b6a3a --- /dev/null +++ b/arch/riscv/boot/dts/microchip/mpfs-m100pfs-fabric.dtsi @@ -0,0 +1,45 @@ +// SPDX-License-Identifier: (GPL-2.0 OR MIT) +/* Copyright (c) 2022 Microchip Technology Inc */ + +/ { + fabric_clk3: fabric-clk3 { + compatible = "fixed-clock"; + #clock-cells = <0>; + clock-frequency = <62500000>; + }; + + fabric_clk1: fabric-clk1 { + compatible = "fixed-clock"; + #clock-cells = <0>; + clock-frequency = <125000000>; + }; + + pcie: pcie@2000000000 { + compatible = "microchip,pcie-host-1.0"; + #address-cells = <0x3>; + #interrupt-cells = <0x1>; + #size-cells = <0x2>; + device_type = "pci"; + reg = <0x20 0x0 0x0 0x8000000>, <0x0 0x43000000 0x0 0x10000>; + reg-names = "cfg", "apb"; + bus-range = <0x0 0x7f>; + interrupt-parent = <&plic>; + interrupts = <119>; + interrupt-map = <0 0 0 1 &pcie_intc 0>, + <0 0 0 2 &pcie_intc 1>, + <0 0 0 3 &pcie_intc 2>, + <0 0 0 4 &pcie_intc 3>; + interrupt-map-mask = <0 0 0 7>; + clocks = <&fabric_clk1>, <&fabric_clk1>, <&fabric_clk3>; + clock-names = "fic0", "fic1", "fic3"; + ranges = <0x3000000 0x0 0x8000000 0x20 0x8000000 0x0 0x80000000>; + msi-parent = <&pcie>; + msi-controller; + status = "disabled"; + pcie_intc: interrupt-controller { + #address-cells = <0>; + #interrupt-cells = <1>; + interrupt-controller; + }; + }; +}; diff --git a/arch/riscv/boot/dts/microchip/mpfs-m100pfsevp.dts b/arch/riscv/boot/dts/microchip/mpfs-m100pfsevp.dts new file mode 100644 index 000000000000..184cb36a175e --- /dev/null +++ b/arch/riscv/boot/dts/microchip/mpfs-m100pfsevp.dts @@ -0,0 +1,179 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Original all-in-one devicetree: + * Copyright (C) 2021-2022 - Wolfgang Grandegger + * Rewritten to use includes: + * Copyright (C) 2022 - Conor Dooley + */ +/dts-v1/; + +#include "mpfs.dtsi" +#include "mpfs-m100pfs-fabric.dtsi" + +/* Clock frequency (in Hz) of the rtcclk */ +#define MTIMER_FREQ 1000000 + +/ { + model = "Aries Embedded M100PFEVPS"; + compatible = "aries,m100pfsevp", "microchip,mpfs"; + + aliases { + ethernet0 = &mac0; + ethernet1 = &mac1; + serial0 = &mmuart0; + serial1 = &mmuart1; + serial2 = &mmuart2; + serial3 = &mmuart3; + serial4 = &mmuart4; + gpio0 = &gpio0; + gpio1 = &gpio2; + }; + + chosen { + stdout-path = "serial1:115200n8"; + }; + + cpus { + timebase-frequency = ; + }; + + ddrc_cache_lo: memory@80000000 { + device_type = "memory"; + reg = <0x0 0x80000000 0x0 0x40000000>; + }; + ddrc_cache_hi: memory@1040000000 { + device_type = "memory"; + reg = <0x10 0x40000000 0x0 0x40000000>; + }; +}; + +&can0 { + status = "okay"; +}; + +&i2c0 { + status = "okay"; +}; + +&i2c1 { + status = "okay"; +}; + +&gpio0 { + interrupts = <13>, <14>, <15>, <16>, + <17>, <18>, <19>, <20>, + <21>, <22>, <23>, <24>, + <25>, <26>; + ngpios = <14>; + status = "okay"; + + pmic-irq-hog { + gpio-hog; + gpios = <13 0>; + input; + }; + + /* Set to low for eMMC, high for SD-card */ + mmc-sel-hog { + gpio-hog; + gpios = <12 0>; + output-high; + }; +}; + +&gpio2 { + interrupts = <13>, <14>, <15>, <16>, + <17>, <18>, <19>, <20>, + <21>, <22>, <23>, <24>, + <25>, <26>, <27>, <28>, + <29>, <30>, <31>, <32>, + <33>, <34>, <35>, <36>, + <37>, <38>, <39>, <40>, + <41>, <42>, <43>, <44>; + status = "okay"; +}; + +&mac0 { + status = "okay"; + phy-mode = "gmii"; + phy-handle = <&phy0>; + phy0: ethernet-phy@0 { + reg = <0>; + }; +}; + +&mac1 { + status = "okay"; + phy-mode = "gmii"; + phy-handle = <&phy1>; + phy1: ethernet-phy@0 { + reg = <0>; + }; +}; + +&mbox { + status = "okay"; +}; + +&mmc { + max-frequency = <50000000>; + bus-width = <4>; + cap-mmc-highspeed; + cap-sd-highspeed; + no-1-8-v; + sd-uhs-sdr12; + sd-uhs-sdr25; + sd-uhs-sdr50; + sd-uhs-sdr104; + disable-wp; + status = "okay"; +}; + +&mmuart1 { + status = "okay"; +}; + +&mmuart2 { + status = "okay"; +}; + +&mmuart3 { + status = "okay"; +}; + +&mmuart4 { + status = "okay"; +}; + +&pcie { + status = "okay"; +}; + +&qspi { + status = "okay"; +}; + +&refclk { + clock-frequency = <125000000>; +}; + +&rtc { + status = "okay"; +}; + +&spi0 { + status = "okay"; +}; + +&spi1 { + status = "okay"; +}; + +&syscontroller { + status = "okay"; +}; + +&usb { + status = "okay"; + dr_mode = "host"; +}; From 6c1193301791d3fcc0ad9ff3b861a8216e00773b Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 27 Sep 2022 12:19:23 +0100 Subject: [PATCH 027/121] riscv: dts: microchip: update memory configuration for v2022.10 In the v2022.10 reference design, the seg registers are going to be changed, resulting in a required change to the memory map in Linux. A small 4M reservation is made at the end of 32-bit DDR to provide some memory for the HSS to use, so that it can cache its payload.bin between reboots of a specific context. Signed-off-by: Conor Dooley --- arch/riscv/boot/dts/microchip/mpfs-icicle-kit.dts | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit.dts b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit.dts index 42d350fe6c6b..31f88cb4d5e5 100644 --- a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit.dts +++ b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit.dts @@ -33,15 +33,26 @@ cpus { ddrc_cache_lo: memory@80000000 { device_type = "memory"; - reg = <0x0 0x80000000 0x0 0x2e000000>; + reg = <0x0 0x80000000 0x0 0x40000000>; status = "okay"; }; ddrc_cache_hi: memory@1000000000 { device_type = "memory"; - reg = <0x10 0x0 0x0 0x40000000>; + reg = <0x10 0x40000000 0x0 0x40000000>; status = "okay"; }; + + reserved-memory { + #address-cells = <2>; + #size-cells = <2>; + ranges; + + hss_payload: region@BFC00000 { + reg = <0x0 0xBFC00000 0x0 0x400000>; + no-map; + }; + }; }; &core_pwm0 { From 3baca1a4d490484fcd555413f1fec85b2e071912 Mon Sep 17 00:00:00 2001 From: Anup Patel Date: Wed, 27 Jul 2022 10:08:29 +0530 Subject: [PATCH 028/121] RISC-V: Add mvendorid, marchid, and mimpid to /proc/cpuinfo output Identifying the underlying RISC-V implementation can be important for some of the user space applications. For example, the perf tool uses arch specific CPU implementation id (i.e. CPUID) to select a JSON file describing custom perf events on a CPU. Currently, there is no way to identify RISC-V implementation so we add mvendorid, marchid, and mimpid to /proc/cpuinfo output. Signed-off-by: Anup Patel Reviewed-by: Heinrich Schuchardt Tested-by: Nikita Shubin Reviewed-by: Conor Dooley Link: https://lore.kernel.org/r/20220727043829.151794-1-apatel@ventanamicro.com/ Signed-off-by: Palmer Dabbelt --- arch/riscv/kernel/cpu.c | 51 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 51 insertions(+) diff --git a/arch/riscv/kernel/cpu.c b/arch/riscv/kernel/cpu.c index 0be8a2403212..250220e2cceb 100644 --- a/arch/riscv/kernel/cpu.c +++ b/arch/riscv/kernel/cpu.c @@ -3,10 +3,13 @@ * Copyright (C) 2012 Regents of the University of California */ +#include #include #include #include +#include #include +#include #include #include @@ -68,6 +71,50 @@ int riscv_of_parent_hartid(struct device_node *node, unsigned long *hartid) } #ifdef CONFIG_PROC_FS + +struct riscv_cpuinfo { + unsigned long mvendorid; + unsigned long marchid; + unsigned long mimpid; +}; +static DEFINE_PER_CPU(struct riscv_cpuinfo, riscv_cpuinfo); + +static int riscv_cpuinfo_starting(unsigned int cpu) +{ + struct riscv_cpuinfo *ci = this_cpu_ptr(&riscv_cpuinfo); + +#if IS_ENABLED(CONFIG_RISCV_SBI) + ci->mvendorid = sbi_spec_is_0_1() ? 0 : sbi_get_mvendorid(); + ci->marchid = sbi_spec_is_0_1() ? 0 : sbi_get_marchid(); + ci->mimpid = sbi_spec_is_0_1() ? 0 : sbi_get_mimpid(); +#elif IS_ENABLED(CONFIG_RISCV_M_MODE) + ci->mvendorid = csr_read(CSR_MVENDORID); + ci->marchid = csr_read(CSR_MARCHID); + ci->mimpid = csr_read(CSR_MIMPID); +#else + ci->mvendorid = 0; + ci->marchid = 0; + ci->mimpid = 0; +#endif + + return 0; +} + +static int __init riscv_cpuinfo_init(void) +{ + int ret; + + ret = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "riscv/cpuinfo:starting", + riscv_cpuinfo_starting, NULL); + if (ret < 0) { + pr_err("cpuinfo: failed to register hotplug callbacks.\n"); + return ret; + } + + return 0; +} +device_initcall(riscv_cpuinfo_init); + #define __RISCV_ISA_EXT_DATA(UPROP, EXTID) \ { \ .uprop = #UPROP, \ @@ -185,6 +232,7 @@ static int c_show(struct seq_file *m, void *v) { unsigned long cpu_id = (unsigned long)v - 1; struct device_node *node = of_get_cpu_node(cpu_id, NULL); + struct riscv_cpuinfo *ci = per_cpu_ptr(&riscv_cpuinfo, cpu_id); const char *compat, *isa; seq_printf(m, "processor\t: %lu\n", cpu_id); @@ -195,6 +243,9 @@ static int c_show(struct seq_file *m, void *v) if (!of_property_read_string(node, "compatible", &compat) && strcmp(compat, "riscv")) seq_printf(m, "uarch\t\t: %s\n", compat); + seq_printf(m, "mvendorid\t: 0x%lx\n", ci->mvendorid); + seq_printf(m, "marchid\t\t: 0x%lx\n", ci->marchid); + seq_printf(m, "mimpid\t\t: 0x%lx\n", ci->mimpid); seq_puts(m, "\n"); of_node_put(node); From 417b9c51f59734d852e47252476fadc293ad994a Mon Sep 17 00:00:00 2001 From: Callum Osmotherly Date: Wed, 5 Oct 2022 17:44:16 +1030 Subject: [PATCH 029/121] ALSA: hda/realtek: remove ALC289_FIXUP_DUAL_SPK for Dell 5530 After some feedback from users with Dell Precision 5530 machines, this patch reverts the previous change to add ALC289_FIXUP_DUAL_SPK. While it improved the speaker output quality, it caused the headphone jack to have an audible "pop" sound when power saving was toggled. Fixes: 1885ff13d4c4 ("ALSA: hda/realtek: Enable 4-speaker output Dell Precision 5530 laptop") Signed-off-by: Callum Osmotherly Cc: Link: https://lore.kernel.org/r/Yz0uyN1zwZhnyRD6@piranha Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_realtek.c | 1 - 1 file changed, 1 deletion(-) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index bce82b834cec..d89f95ae0efc 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -9198,7 +9198,6 @@ static const struct snd_pci_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x1028, 0x0871, "Dell Precision 3630", ALC255_FIXUP_DELL_HEADSET_MIC), SND_PCI_QUIRK(0x1028, 0x0872, "Dell Precision 3630", ALC255_FIXUP_DELL_HEADSET_MIC), SND_PCI_QUIRK(0x1028, 0x0873, "Dell Precision 3930", ALC255_FIXUP_DUMMY_LINEOUT_VERB), - SND_PCI_QUIRK(0x1028, 0x087d, "Dell Precision 5530", ALC289_FIXUP_DUAL_SPK), SND_PCI_QUIRK(0x1028, 0x08ad, "Dell WYSE AIO", ALC225_FIXUP_DELL_WYSE_AIO_MIC_NO_PRESENCE), SND_PCI_QUIRK(0x1028, 0x08ae, "Dell WYSE NB", ALC225_FIXUP_DELL1_MIC_NO_PRESENCE), SND_PCI_QUIRK(0x1028, 0x0935, "Dell", ALC274_FIXUP_DELL_AIO_LINEOUT_VERB), From c210b91818e81068ca2573c20684644b8e110a07 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 20 Sep 2022 10:37:35 +0100 Subject: [PATCH 030/121] riscv: dts: microchip: fix fabric i2c reg size The size of the reg should've been changed when the address was changed, but obviously I forgot to do so. Fixes: ab291621a8b8 ("riscv: dts: microchip: icicle: re-jig fabric peripheral addresses") Signed-off-by: Conor Dooley --- arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi index b6bfe177ccb2..24b1cfb9a73e 100644 --- a/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi +++ b/arch/riscv/boot/dts/microchip/mpfs-icicle-kit-fabric.dtsi @@ -16,7 +16,7 @@ core_pwm0: pwm@40000000 { i2c2: i2c@40000200 { compatible = "microchip,corei2c-rtl-v7"; - reg = <0x0 0x40000200 0x0 0x1000>; + reg = <0x0 0x40000200 0x0 0x100>; #address-cells = <1>; #size-cells = <0>; clocks = <&fabric_clk3>; From e1567b4f0eec779af99b372773ddeb5be9b6208b Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Wed, 5 Oct 2022 19:16:42 +0100 Subject: [PATCH 031/121] arm64/sysreg: Fix typo in SCTR_EL1.SPINTMASK SPINTMASK was typoed as SPINMASK, fix it. Signed-off-by: Mark Brown Link: https://lore.kernel.org/r/20221005181642.711734-1-broonie@kernel.org Signed-off-by: Catalin Marinas --- arch/arm64/tools/sysreg | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm64/tools/sysreg b/arch/arm64/tools/sysreg index 7f1fb36f208c..384757a7eda9 100644 --- a/arch/arm64/tools/sysreg +++ b/arch/arm64/tools/sysreg @@ -732,7 +732,7 @@ EndSysreg Sysreg SCTLR_EL1 3 0 1 0 0 Field 63 TIDCP -Field 62 SPINMASK +Field 62 SPINTMASK Field 61 NMI Field 60 EnTP2 Res0 59:58 From 171df58028bf4649460fb146a56a58dcb0c8f75a Mon Sep 17 00:00:00 2001 From: James Morse Date: Fri, 30 Sep 2022 14:19:59 +0100 Subject: [PATCH 032/121] arm64: errata: Add Cortex-A55 to the repeat tlbi list Cortex-A55 is affected by an erratum where in rare circumstances the CPUs may not handle a race between a break-before-make sequence on one CPU, and another CPU accessing the same page. This could allow a store to a page that has been unmapped. Work around this by adding the affected CPUs to the list that needs TLB sequences to be done twice. Signed-off-by: James Morse Cc: Link: https://lore.kernel.org/r/20220930131959.3082594-1-james.morse@arm.com Signed-off-by: Catalin Marinas --- Documentation/arm64/silicon-errata.rst | 2 ++ arch/arm64/Kconfig | 17 +++++++++++++++++ arch/arm64/kernel/cpu_errata.c | 5 +++++ 3 files changed, 24 insertions(+) diff --git a/Documentation/arm64/silicon-errata.rst b/Documentation/arm64/silicon-errata.rst index 17d9fc5d14fb..808ade4cc008 100644 --- a/Documentation/arm64/silicon-errata.rst +++ b/Documentation/arm64/silicon-errata.rst @@ -76,6 +76,8 @@ stable kernels. +----------------+-----------------+-----------------+-----------------------------+ | ARM | Cortex-A55 | #1530923 | ARM64_ERRATUM_1530923 | +----------------+-----------------+-----------------+-----------------------------+ +| ARM | Cortex-A55 | #2441007 | ARM64_ERRATUM_2441007 | ++----------------+-----------------+-----------------+-----------------------------+ | ARM | Cortex-A57 | #832075 | ARM64_ERRATUM_832075 | +----------------+-----------------+-----------------+-----------------------------+ | ARM | Cortex-A57 | #852523 | N/A | diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index 1675310f1791..20d082d54bd8 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -634,6 +634,23 @@ config ARM64_ERRATUM_1530923 config ARM64_WORKAROUND_REPEAT_TLBI bool +config ARM64_ERRATUM_2441007 + bool "Cortex-A55: Completion of affected memory accesses might not be guaranteed by completion of a TLBI" + default y + select ARM64_WORKAROUND_REPEAT_TLBI + help + This option adds a workaround for ARM Cortex-A55 erratum #2441007. + + Under very rare circumstances, affected Cortex-A55 CPUs + may not handle a race between a break-before-make sequence on one + CPU, and another CPU accessing the same page. This could allow a + store to a page that has been unmapped. + + Work around this by adding the affected CPUs to the list that needs + TLB sequences to be done twice. + + If unsure, say Y. + config ARM64_ERRATUM_1286807 bool "Cortex-A76: Modification of the translation table for a virtual address might lead to read-after-read ordering violation" default y diff --git a/arch/arm64/kernel/cpu_errata.c b/arch/arm64/kernel/cpu_errata.c index 58ca4f6b25d6..89ac00084f38 100644 --- a/arch/arm64/kernel/cpu_errata.c +++ b/arch/arm64/kernel/cpu_errata.c @@ -230,6 +230,11 @@ static const struct arm64_cpu_capabilities arm64_repeat_tlbi_list[] = { ERRATA_MIDR_RANGE(MIDR_QCOM_KRYO_4XX_GOLD, 0xc, 0xe, 0xf, 0xe), }, #endif +#ifdef CONFIG_ARM64_ERRATUM_2441007 + { + ERRATA_MIDR_ALL_VERSIONS(MIDR_CORTEX_A55), + }, +#endif #ifdef CONFIG_ARM64_ERRATUM_2441009 { /* Cortex-A510 r0p0 -> r1p1. Fixed in r1p2 */ From ad0112f2d54cafee839e2ee99ec0b5fb9ce5c4b8 Mon Sep 17 00:00:00 2001 From: Sun Ke Date: Sat, 24 Sep 2022 11:21:27 +0800 Subject: [PATCH 033/121] drivers/perf: fix return value check in ali_drw_pmu_probe() In case of error, devm_ioremap_resource() returns ERR_PTR(), and never returns NULL. The NULL test in the return value check should be replaced with IS_ERR(). Fixes: cf7b61073e45 ("drivers/perf: add DDR Sub-System Driveway PMU driver for Yitian 710 SoC") Signed-off-by: Sun Ke Acked-by: Will Deacon Reviewed-by: Shuai Xue Link: https://lore.kernel.org/r/20220924032127.313156-1-sunke32@huawei.com Signed-off-by: Catalin Marinas --- drivers/perf/alibaba_uncore_drw_pmu.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/perf/alibaba_uncore_drw_pmu.c b/drivers/perf/alibaba_uncore_drw_pmu.c index 82729b874f09..a7689fecb49d 100644 --- a/drivers/perf/alibaba_uncore_drw_pmu.c +++ b/drivers/perf/alibaba_uncore_drw_pmu.c @@ -658,8 +658,8 @@ static int ali_drw_pmu_probe(struct platform_device *pdev) res = platform_get_resource(pdev, IORESOURCE_MEM, 0); drw_pmu->cfg_base = devm_ioremap_resource(&pdev->dev, res); - if (!drw_pmu->cfg_base) - return -ENOMEM; + if (IS_ERR(drw_pmu->cfg_base)) + return PTR_ERR(drw_pmu->cfg_base); name = devm_kasprintf(drw_pmu->dev, GFP_KERNEL, "ali_drw_%llx", (u64) (res->start >> ALI_DRW_PMU_PA_SHIFT)); From e08d07dd9f80e997ad36a088eb276509ca484e97 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Tue, 27 Sep 2022 15:37:16 +0200 Subject: [PATCH 034/121] drivers/perf: ALIBABA_UNCORE_DRW_PMU should depend on ACPI The Alibaba T-Head Yitian 710 DDR Sub-system Driveway PMU driver relies solely on ACPI for matching. Hence add a dependency on ACPI, to prevent asking the user about this driver when configuring a kernel without ACPI support. Fixes: cf7b61073e45 ("drivers/perf: add DDR Sub-System Driveway PMU driver for Yitian 710 SoC") Signed-off-by: Geert Uytterhoeven Acked-by: Will Deacon Link: https://lore.kernel.org/r/2a4407bb598285660fa5e604e56823ddb12bb0aa.1664285774.git.geert+renesas@glider.be Signed-off-by: Catalin Marinas --- drivers/perf/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/perf/Kconfig b/drivers/perf/Kconfig index 44c07ea487f4..341010f20b77 100644 --- a/drivers/perf/Kconfig +++ b/drivers/perf/Kconfig @@ -185,7 +185,7 @@ config APPLE_M1_CPU_PMU config ALIBABA_UNCORE_DRW_PMU tristate "Alibaba T-Head Yitian 710 DDR Sub-system Driveway PMU driver" - depends on ARM64 || COMPILE_TEST + depends on (ARM64 && ACPI) || COMPILE_TEST help Support for Driveway PMU events monitoring on Yitian 710 DDR Sub-system. From 5f4853e810943af5e45fcc040cbbcbba07d8fc25 Mon Sep 17 00:00:00 2001 From: Lukas Bulwahn Date: Thu, 29 Sep 2022 14:29:37 +0200 Subject: [PATCH 035/121] MAINTAINERS: rectify file entry in ALIBABA PMU DRIVER Commit cf7b61073e45 ("drivers/perf: add DDR Sub-System Driveway PMU driver for Yitian 710 SoC") adds the DDR Sub-System Driveway PMU driver here: drivers/perf/alibaba_uncore_drw_pmu.c The file entry in MAINTAINERS for the ALIBABA PMU DRIVER, introduced with commit d813a19e7d2e ("MAINTAINERS: add maintainers for Alibaba' T-Head PMU driver"), however refers to: drivers/perf/alibaba_uncore_dwr_pmu.c Note the swapping of characters. Hence, ./scripts/get_maintainer.pl --self-test=patterns complains about a broken file pattern. Repair this file entry in ALIBABA PMU DRIVER. Signed-off-by: Lukas Bulwahn Acked-by: Will Deacon Link: https://lore.kernel.org/r/20220929122937.20132-1-lukas.bulwahn@gmail.com Signed-off-by: Catalin Marinas --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 58d0aefead54..a8e0f73747cf 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -753,7 +753,7 @@ ALIBABA PMU DRIVER M: Shuai Xue S: Supported F: Documentation/admin-guide/perf/alibaba_pmu.rst -F: drivers/perf/alibaba_uncore_dwr_pmu.c +F: drivers/perf/alibaba_uncore_drw_pmu.c ALIENWARE WMI DRIVER L: Dell.Client.Kernel@dell.com From 556a11a082ee208455ed42e7c460849cc3dbd68c Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Amadeusz=20S=C5=82awi=C5=84ski?= Date: Fri, 7 Oct 2022 10:48:56 +0200 Subject: [PATCH 036/121] ALSA: hda: Update register polling macros MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Recent commit d91857059def ("ALSA: hda: Rework snd_hdac_stream_reset() to use macros") missed that on some devices register access needs to be done with unaligned access helper. Change polling macros to use read_poll_timeout_atomic() in order to specify register read function. Fixes: d91857059def ("ALSA: hda: Rework snd_hdac_stream_reset() to use macros") Reported-by: Jon Hunter Link: https://lore.kernel.org/alsa-devel/20220818141517.109280-1-amadeuszx.slawinski@linux.intel.com/T/#m1270737db52b5ef163eff73cb5f862d16a07a428 Reviewed-by: Cezary Rojewski Signed-off-by: Amadeusz Sławiński Link: https://lore.kernel.org/r/20221007084856.1638302-1-amadeuszx.slawinski@linux.intel.com Signed-off-by: Takashi Iwai --- include/sound/hdaudio.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/include/sound/hdaudio.h b/include/sound/hdaudio.h index ddff03e546e9..35778f953a3f 100644 --- a/include/sound/hdaudio.h +++ b/include/sound/hdaudio.h @@ -592,11 +592,11 @@ int snd_hdac_get_stream_stripe_ctl(struct hdac_bus *bus, #define snd_hdac_stream_readb(dev, reg) \ snd_hdac_reg_readb((dev)->bus, (dev)->sd_addr + AZX_REG_ ## reg) #define snd_hdac_stream_readb_poll(dev, reg, val, cond, delay_us, timeout_us) \ - readb_poll_timeout((dev)->sd_addr + AZX_REG_ ## reg, val, cond, \ - delay_us, timeout_us) + read_poll_timeout_atomic(snd_hdac_reg_readb, val, cond, delay_us, timeout_us, \ + false, (dev)->bus, (dev)->sd_addr + AZX_REG_ ## reg) #define snd_hdac_stream_readl_poll(dev, reg, val, cond, delay_us, timeout_us) \ - readl_poll_timeout((dev)->sd_addr + AZX_REG_ ## reg, val, cond, \ - delay_us, timeout_us) + read_poll_timeout_atomic(snd_hdac_reg_readl, val, cond, delay_us, timeout_us, \ + false, (dev)->bus, (dev)->sd_addr + AZX_REG_ ## reg) /* update a register, pass without AZX_REG_ prefix */ #define snd_hdac_stream_updatel(dev, reg, mask, val) \ From 9902b303b5ade208b58f0dd38a09831813582211 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Sun, 9 Oct 2022 12:42:09 +0200 Subject: [PATCH 037/121] ALSA: usb-audio: Avoid unnecessary interface change at EP close We toggle USB interface at PCM prepare and reset at close. When the PCM isn't prepared, resetting again makes little sense. Check the current altset and avoid unnecessary interface reset at EP close. Link: https://lore.kernel.org/r/20221009104212.18877-2-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/usb/endpoint.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/sound/usb/endpoint.c b/sound/usb/endpoint.c index 48a3843a08f1..f21acbc9f4f4 100644 --- a/sound/usb/endpoint.c +++ b/sound/usb/endpoint.c @@ -32,6 +32,7 @@ struct snd_usb_iface_ref { unsigned char iface; bool need_setup; int opened; + int altset; struct list_head list; }; @@ -899,6 +900,9 @@ static int endpoint_set_interface(struct snd_usb_audio *chip, int altset = set ? ep->altsetting : 0; int err; + if (ep->iface_ref->altset == altset) + return 0; + usb_audio_dbg(chip, "Setting usb interface %d:%d for EP 0x%x\n", ep->iface, altset, ep->ep_num); err = usb_set_interface(chip->dev, ep->iface, altset); @@ -910,6 +914,7 @@ static int endpoint_set_interface(struct snd_usb_audio *chip, if (chip->quirk_flags & QUIRK_FLAG_IFACE_DELAY) msleep(50); + ep->iface_ref->altset = altset; return 0; } From a74f8d0aa902ca494676b79226e0b5a1747b81d4 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Sun, 9 Oct 2022 12:42:10 +0200 Subject: [PATCH 038/121] ALSA: usb-audio: Apply mutex around snd_usb_endpoint_set_params() The protection with chip->mutex was lost after splitting snd_usb_endpoint_set_params() and snd_usb_endpoint_prepare(). Apply the same mutex again to the former function. Fixes: 2be79d586454 ("ALSA: usb-audio: Split endpoint setups for hw_params and prepare (take#2)") Link: https://lore.kernel.org/r/20221009104212.18877-3-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/usb/endpoint.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/sound/usb/endpoint.c b/sound/usb/endpoint.c index f21acbc9f4f4..da378e565ef8 100644 --- a/sound/usb/endpoint.c +++ b/sound/usb/endpoint.c @@ -1337,10 +1337,11 @@ int snd_usb_endpoint_set_params(struct snd_usb_audio *chip, const struct audioformat *fmt = ep->cur_audiofmt; int err; + mutex_lock(&chip->mutex); /* release old buffers, if any */ err = release_urbs(ep, false); if (err < 0) - return err; + goto unlock; ep->datainterval = fmt->datainterval; ep->maxpacksize = fmt->maxpacksize; @@ -1378,13 +1379,16 @@ int snd_usb_endpoint_set_params(struct snd_usb_audio *chip, usb_audio_dbg(chip, "Set up %d URBS, ret=%d\n", ep->nurbs, err); if (err < 0) - return err; + goto unlock; /* some unit conversions in runtime */ ep->maxframesize = ep->maxpacksize / ep->cur_frame_bytes; ep->curframesize = ep->curpacksize / ep->cur_frame_bytes; - return update_clock_ref_rate(chip, ep); + err = update_clock_ref_rate(chip, ep); + unlock: + mutex_unlock(&chip->mutex); + return err; } static int init_sample_rate(struct snd_usb_audio *chip, From 9355b60e401d825590d37f04ea873c58efe9b7bf Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Sun, 9 Oct 2022 12:42:11 +0200 Subject: [PATCH 039/121] ALSA: usb-audio: Correct the return code from snd_usb_endpoint_set_params() snd_usb_endpoint_set_params() should return zero for a success, but currently it returns the sample rate. Correct it. Fixes: 2be79d586454 ("ALSA: usb-audio: Split endpoint setups for hw_params and prepare (take#2)") Link: https://lore.kernel.org/r/20221009104212.18877-4-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/usb/endpoint.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/sound/usb/endpoint.c b/sound/usb/endpoint.c index da378e565ef8..44cce6cec9da 100644 --- a/sound/usb/endpoint.c +++ b/sound/usb/endpoint.c @@ -1386,6 +1386,8 @@ int snd_usb_endpoint_set_params(struct snd_usb_audio *chip, ep->curframesize = ep->curpacksize / ep->cur_frame_bytes; err = update_clock_ref_rate(chip, ep); + if (err >= 0) + err = 0; unlock: mutex_unlock(&chip->mutex); return err; From 1045f5f1ff0751423aeb65648e5e1abd7a7a8672 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Sun, 9 Oct 2022 12:42:12 +0200 Subject: [PATCH 040/121] ALSA: usb-audio: Avoid superfluous endpoint setup After splitting to snd_usb_endpoint_set_params() and *_prepare(), the skip of each function should be checked with different flags, while we still use ep->need_setup as the single one. Introduce ep->need_prepare for indicating the need of prepare, and also add the missing check of ep->need_setup at the set_params. Fixes: 2be79d586454 ("ALSA: usb-audio: Split endpoint setups for hw_params and prepare (take#2)") Link: https://lore.kernel.org/r/20221009104212.18877-5-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/usb/card.h | 3 ++- sound/usb/endpoint.c | 17 ++++++++++++----- 2 files changed, 14 insertions(+), 6 deletions(-) diff --git a/sound/usb/card.h b/sound/usb/card.h index ca75f2206170..40061550105a 100644 --- a/sound/usb/card.h +++ b/sound/usb/card.h @@ -129,7 +129,8 @@ struct snd_usb_endpoint { in a stream */ bool implicit_fb_sync; /* syncs with implicit feedback */ bool lowlatency_playback; /* low-latency playback mode */ - bool need_setup; /* (re-)need for configure? */ + bool need_setup; /* (re-)need for hw_params? */ + bool need_prepare; /* (re-)need for prepare? */ /* for hw constraints */ const struct audioformat *cur_audiofmt; diff --git a/sound/usb/endpoint.c b/sound/usb/endpoint.c index 44cce6cec9da..d0b8d61d1d22 100644 --- a/sound/usb/endpoint.c +++ b/sound/usb/endpoint.c @@ -824,6 +824,7 @@ snd_usb_endpoint_open(struct snd_usb_audio *chip, ep->implicit_fb_sync = fp->implicit_fb; ep->need_setup = true; + ep->need_prepare = true; usb_audio_dbg(chip, " channels=%d, rate=%d, format=%s, period_bytes=%d, periods=%d, implicit_fb=%d\n", ep->cur_channels, ep->cur_rate, @@ -952,7 +953,7 @@ void snd_usb_endpoint_close(struct snd_usb_audio *chip, /* Prepare for suspening EP, called from the main suspend handler */ void snd_usb_endpoint_suspend(struct snd_usb_endpoint *ep) { - ep->need_setup = true; + ep->need_prepare = true; if (ep->iface_ref) ep->iface_ref->need_setup = true; if (ep->clock_ref) @@ -1335,9 +1336,12 @@ int snd_usb_endpoint_set_params(struct snd_usb_audio *chip, struct snd_usb_endpoint *ep) { const struct audioformat *fmt = ep->cur_audiofmt; - int err; + int err = 0; mutex_lock(&chip->mutex); + if (!ep->need_setup) + goto unlock; + /* release old buffers, if any */ err = release_urbs(ep, false); if (err < 0) @@ -1386,8 +1390,11 @@ int snd_usb_endpoint_set_params(struct snd_usb_audio *chip, ep->curframesize = ep->curpacksize / ep->cur_frame_bytes; err = update_clock_ref_rate(chip, ep); - if (err >= 0) + if (err >= 0) { + ep->need_setup = false; err = 0; + } + unlock: mutex_unlock(&chip->mutex); return err; @@ -1437,7 +1444,7 @@ int snd_usb_endpoint_prepare(struct snd_usb_audio *chip, mutex_lock(&chip->mutex); if (WARN_ON(!ep->iface_ref)) goto unlock; - if (!ep->need_setup) + if (!ep->need_prepare) goto unlock; /* If the interface has been already set up, just set EP parameters */ @@ -1491,7 +1498,7 @@ int snd_usb_endpoint_prepare(struct snd_usb_audio *chip, ep->iface_ref->need_setup = false; done: - ep->need_setup = false; + ep->need_prepare = false; err = 1; unlock: From f3e59ff348c077a6afd4edb23d7e69e9cba62fdc Mon Sep 17 00:00:00 2001 From: Peter Oberparleiter Date: Wed, 28 Sep 2022 15:34:33 +0200 Subject: [PATCH 041/121] s390/vmur: remove unnecessary BUG statement An existing BUG statement in vmur's interrupt handler triggers if: 1. An online vmur device is removed (e.g. due to driver unload, manual unbind or channel-report words indicating hypervisor-side device removal) 2. Device deactivation fails due to firmware/hypervisor error, leaving subchannel enabled for interrupts + drvdata=NULL 3. Interrupt occurs This situation is highly unlikely and not a clear indication of a general system error that would warrant stopping the full Linux system. Also it can be prevented completely by clearing the interrupt handler when unsetting a vmur device's drvdata. Replace the BUG statement in vmur's interrupt handler by clearing the interrupt handler callback during device removal. Also move the initial setting of the interrupt handler callback under lock for consistency reasons. Reviewed-by: Sven Schnelle Reviewed-by: Heiko Carstens Signed-off-by: Peter Oberparleiter Signed-off-by: Vasily Gorbik --- drivers/s390/char/vmur.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/s390/char/vmur.c b/drivers/s390/char/vmur.c index 68f49e2e964c..471f07ca5066 100644 --- a/drivers/s390/char/vmur.c +++ b/drivers/s390/char/vmur.c @@ -293,7 +293,6 @@ static void ur_int_handler(struct ccw_device *cdev, unsigned long intparm, return; } urd = dev_get_drvdata(&cdev->dev); - BUG_ON(!urd); /* On special conditions irb is an error pointer */ if (IS_ERR(irb)) urd->io_request_rc = PTR_ERR(irb); @@ -809,7 +808,6 @@ static int ur_probe(struct ccw_device *cdev) rc = -ENOMEM; goto fail_urdev_put; } - cdev->handler = ur_int_handler; /* validate virtual unit record device */ urd->class = get_urd_class(urd); @@ -823,6 +821,7 @@ static int ur_probe(struct ccw_device *cdev) } spin_lock_irq(get_ccwdev_lock(cdev)); dev_set_drvdata(&cdev->dev, urd); + cdev->handler = ur_int_handler; spin_unlock_irq(get_ccwdev_lock(cdev)); mutex_unlock(&vmur_mutex); @@ -963,6 +962,7 @@ static void ur_remove(struct ccw_device *cdev) spin_lock_irqsave(get_ccwdev_lock(cdev), flags); urdev_put(dev_get_drvdata(&cdev->dev)); dev_set_drvdata(&cdev->dev, NULL); + cdev->handler = NULL; spin_unlock_irqrestore(get_ccwdev_lock(cdev), flags); mutex_unlock(&vmur_mutex); From bf18140d30541c2c1e5c0f57879634f3d0d04912 Mon Sep 17 00:00:00 2001 From: Peter Oberparleiter Date: Mon, 26 Sep 2022 16:56:04 +0200 Subject: [PATCH 042/121] s390/vmur: generate uevent on unsolicited device end When a traditional channel-attached device transitions from not-ready to ready state, an unsolicited DEVICE END I/O interrupt is raised. This happens for example when a new file arrives in the z/VM virtual reader device. Change the Linux kernel to generate a change uevent when such an interrupt occurs for any online unit record devices supported by the vmur driver. This can be useful to automatically trigger processing of files as they arrive in the reader device. A sample udev rule for running a program when this event occurs looks as follows: ENV{DRIVER}=="vmur", ACTION=="change", ENV{EVENT}=="unsol_de", \ RUN{program}="/path/to/program" The rule can be tested using the following steps: 1. Set reader device online (assuming default reader device number 000c) $ chzdev -ea 0.0.000c 2. Force a ready-state transition using z/VM's READY CP command $ vmcp ready 000c Suggested-by: Alan Altmark Reviewed-by: Heiko Carstens Reviewed-by: Sven Schnelle Signed-off-by: Peter Oberparleiter Signed-off-by: Vasily Gorbik --- drivers/s390/char/vmur.c | 33 ++++++++++++++++++++++++++++++++- drivers/s390/char/vmur.h | 2 ++ 2 files changed, 34 insertions(+), 1 deletion(-) diff --git a/drivers/s390/char/vmur.c b/drivers/s390/char/vmur.c index 471f07ca5066..131293f7f152 100644 --- a/drivers/s390/char/vmur.c +++ b/drivers/s390/char/vmur.c @@ -15,12 +15,14 @@ #include #include #include +#include #include #include #include #include #include +#include #include "vmur.h" @@ -78,6 +80,8 @@ static struct ccw_driver ur_driver = { static DEFINE_MUTEX(vmur_mutex); +static void ur_uevent(struct work_struct *ws); + /* * Allocation, freeing, getting and putting of urdev structures * @@ -108,6 +112,7 @@ static struct urdev *urdev_alloc(struct ccw_device *cdev) ccw_device_get_id(cdev, &urd->dev_id); mutex_init(&urd->io_mutex); init_waitqueue_head(&urd->wait); + INIT_WORK(&urd->uevent_work, ur_uevent); spin_lock_init(&urd->open_lock); refcount_set(&urd->ref_count, 1); urd->cdev = cdev; @@ -275,6 +280,18 @@ static int do_ur_io(struct urdev *urd, struct ccw1 *cpa) return rc; } +static void ur_uevent(struct work_struct *ws) +{ + struct urdev *urd = container_of(ws, struct urdev, uevent_work); + char *envp[] = { + "EVENT=unsol_de", /* Unsolicited device-end interrupt */ + NULL + }; + + kobject_uevent_env(&urd->cdev->dev.kobj, KOBJ_CHANGE, envp); + urdev_put(urd); +} + /* * ur interrupt handler, called from the ccw_device layer */ @@ -288,11 +305,21 @@ static void ur_int_handler(struct ccw_device *cdev, unsigned long intparm, intparm, irb->scsw.cmd.cstat, irb->scsw.cmd.dstat, irb->scsw.cmd.count); } + urd = dev_get_drvdata(&cdev->dev); if (!intparm) { TRACE("ur_int_handler: unsolicited interrupt\n"); + + if (scsw_dstat(&irb->scsw) & DEV_STAT_DEV_END) { + /* + * Userspace might be interested in a transition to + * device-ready state. + */ + urdev_get(urd); + schedule_work(&urd->uevent_work); + } + return; } - urd = dev_get_drvdata(&cdev->dev); /* On special conditions irb is an error pointer */ if (IS_ERR(irb)) urd->io_request_rc = PTR_ERR(irb); @@ -927,6 +954,10 @@ static int ur_set_offline_force(struct ccw_device *cdev, int force) rc = -EBUSY; goto fail_urdev_put; } + if (cancel_work_sync(&urd->uevent_work)) { + /* Work not run yet - need to release reference here */ + urdev_put(urd); + } device_destroy(vmur_class, urd->char_device->dev); cdev_del(urd->char_device); urd->char_device = NULL; diff --git a/drivers/s390/char/vmur.h b/drivers/s390/char/vmur.h index 608b0719ce17..92d17d7cb47b 100644 --- a/drivers/s390/char/vmur.h +++ b/drivers/s390/char/vmur.h @@ -13,6 +13,7 @@ #define _VMUR_H_ #include +#include #define DEV_CLASS_UR_I 0x20 /* diag210 unit record input device class */ #define DEV_CLASS_UR_O 0x10 /* diag210 unit record output device class */ @@ -76,6 +77,7 @@ struct urdev { wait_queue_head_t wait; /* wait queue to serialize open */ int open_flag; /* "urdev is open" flag */ spinlock_t open_lock; /* serialize critical sections */ + struct work_struct uevent_work; /* work to send uevent */ }; /* From 9afd20a5a1b3558950b5e12f3ed057ef93c64a05 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Mon, 10 Oct 2022 09:52:37 +0530 Subject: [PATCH 043/121] clk: spear: Move prototype to accessible header Fixes the following W=1 kernel build warning(s): drivers/clk/spear/spear6xx_clock.c:116:13: warning: no previous prototype for function 'spear6xx_clk_init' [-Wmissing-prototypes] Reported-by: kernel test robot Signed-off-by: Viresh Kumar Signed-off-by: Arnd Bergmann --- arch/arm/mach-spear/generic.h | 3 --- arch/arm/mach-spear/spear3xx.c | 1 + arch/arm/mach-spear/spear6xx.c | 1 + drivers/clk/spear/spear3xx_clock.c | 1 + drivers/clk/spear/spear6xx_clock.c | 1 + include/linux/clk/spear.h | 14 ++++++++++++++ 6 files changed, 18 insertions(+), 3 deletions(-) diff --git a/arch/arm/mach-spear/generic.h b/arch/arm/mach-spear/generic.h index 43b7996ab754..9e36920d4cfd 100644 --- a/arch/arm/mach-spear/generic.h +++ b/arch/arm/mach-spear/generic.h @@ -25,11 +25,8 @@ extern struct pl022_ssp_controller pl022_plat_data; extern struct pl08x_platform_data pl080_plat_data; void __init spear_setup_of_timer(void); -void __init spear3xx_clk_init(void __iomem *misc_base, - void __iomem *soc_config_base); void __init spear3xx_map_io(void); void __init spear3xx_dt_init_irq(void); -void __init spear6xx_clk_init(void __iomem *misc_base); void __init spear13xx_map_io(void); void __init spear13xx_l2x0_init(void); diff --git a/arch/arm/mach-spear/spear3xx.c b/arch/arm/mach-spear/spear3xx.c index 2ba406e92c41..7ef9670d3029 100644 --- a/arch/arm/mach-spear/spear3xx.c +++ b/arch/arm/mach-spear/spear3xx.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include #include "pl080.h" diff --git a/arch/arm/mach-spear/spear6xx.c b/arch/arm/mach-spear/spear6xx.c index 58183493e06d..98d9f2ad6b74 100644 --- a/arch/arm/mach-spear/spear6xx.c +++ b/arch/arm/mach-spear/spear6xx.c @@ -12,6 +12,7 @@ #include #include +#include #include #include #include diff --git a/drivers/clk/spear/spear3xx_clock.c b/drivers/clk/spear/spear3xx_clock.c index 41717ff707f6..ba8791303156 100644 --- a/drivers/clk/spear/spear3xx_clock.c +++ b/drivers/clk/spear/spear3xx_clock.c @@ -8,6 +8,7 @@ #include #include +#include #include #include #include diff --git a/drivers/clk/spear/spear6xx_clock.c b/drivers/clk/spear/spear6xx_clock.c index 490701ac9e93..c192a9141b86 100644 --- a/drivers/clk/spear/spear6xx_clock.c +++ b/drivers/clk/spear/spear6xx_clock.c @@ -7,6 +7,7 @@ */ #include +#include #include #include #include "clk.h" diff --git a/include/linux/clk/spear.h b/include/linux/clk/spear.h index a64d034ceddd..eaf95ca656f8 100644 --- a/include/linux/clk/spear.h +++ b/include/linux/clk/spear.h @@ -8,6 +8,20 @@ #ifndef __LINUX_CLK_SPEAR_H #define __LINUX_CLK_SPEAR_H +#ifdef CONFIG_ARCH_SPEAR3XX +void __init spear3xx_clk_init(void __iomem *misc_base, + void __iomem *soc_config_base); +#else +static inline void __init spear3xx_clk_init(void __iomem *misc_base, + void __iomem *soc_config_base) {} +#endif + +#ifdef CONFIG_ARCH_SPEAR6XX +void __init spear6xx_clk_init(void __iomem *misc_base); +#else +static inline void __init spear6xx_clk_init(void __iomem *misc_base) {} +#endif + #ifdef CONFIG_MACH_SPEAR1310 void __init spear1310_clk_init(void __iomem *misc_base, void __iomem *ras_base); #else From 390ca5bca7cdb95042dcb445375db0d9eba7aa4a Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Mon, 10 Oct 2022 09:41:07 +0530 Subject: [PATCH 044/121] ARM: spear6xx: Staticize few definitions Fix warnings with clang like: arch/arm/mach-spear/spear6xx.c:365:13: warning: no previous prototype for function 'spear6xx_map_io' [-Wmissing-prototypes] by making few definitions static. Reported-by: kernel test robot Signed-off-by: Viresh Kumar Signed-off-by: Arnd Bergmann --- arch/arm/mach-spear/spear6xx.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/arm/mach-spear/spear6xx.c b/arch/arm/mach-spear/spear6xx.c index 98d9f2ad6b74..f0a1e704cceb 100644 --- a/arch/arm/mach-spear/spear6xx.c +++ b/arch/arm/mach-spear/spear6xx.c @@ -340,7 +340,7 @@ static struct pl08x_platform_data spear6xx_pl080_plat_data = { * 0xD0000000 0xFD000000 * 0xFC000000 0xFC000000 */ -struct map_desc spear6xx_io_desc[] __initdata = { +static struct map_desc spear6xx_io_desc[] __initdata = { { .virtual = (unsigned long)VA_SPEAR6XX_ML_CPU_BASE, .pfn = __phys_to_pfn(SPEAR_ICM3_ML1_2_BASE), @@ -360,12 +360,12 @@ struct map_desc spear6xx_io_desc[] __initdata = { }; /* This will create static memory mapping for selected devices */ -void __init spear6xx_map_io(void) +static void __init spear6xx_map_io(void) { iotable_init(spear6xx_io_desc, ARRAY_SIZE(spear6xx_io_desc)); } -void __init spear6xx_timer_init(void) +static void __init spear6xx_timer_init(void) { char pclk_name[] = "pll3_clk"; struct clk *gpt_clk, *pclk; @@ -395,7 +395,7 @@ void __init spear6xx_timer_init(void) } /* Add auxdata to pass platform data */ -struct of_dev_auxdata spear6xx_auxdata_lookup[] __initdata = { +static struct of_dev_auxdata spear6xx_auxdata_lookup[] __initdata = { OF_DEV_AUXDATA("arm,pl080", SPEAR_ICM3_DMA_BASE, NULL, &spear6xx_pl080_plat_data), {} From bd60aafce5e1943fd395b8bf726e9824fa621eca Mon Sep 17 00:00:00 2001 From: Chen Lifu Date: Wed, 17 Aug 2022 16:14:20 +0800 Subject: [PATCH 045/121] ARM: mmp: Make some symbols static These symbols pxa168_usb_phy_resources, pxa168_u2o_resources, pxa168_u2oehci_resources and pxa168_u2ootg_resources are not used outside of arch/arm/mach-mmp/devices.c, so mark them static. Fixes the following sparse warning: arch/arm/mach-mmp/devices.c:241:17: warning: symbol 'pxa168_usb_phy_resources' was not declared. Should it be static? arch/arm/mach-mmp/devices.c:262:17: warning: symbol 'pxa168_u2o_resources' was not declared. Should it be static? arch/arm/mach-mmp/devices.c:297:17: warning: symbol 'pxa168_u2oehci_resources' was not declared. Should it be static? arch/arm/mach-mmp/devices.c:324:17: warning: symbol 'pxa168_u2ootg_resources' was not declared. Should it be static? Reported-by: kernel test robot Signed-off-by: Chen Lifu Signed-off-by: Arnd Bergmann --- arch/arm/mach-mmp/devices.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/arm/mach-mmp/devices.c b/arch/arm/mach-mmp/devices.c index 79f4a2aa5475..9968239d8041 100644 --- a/arch/arm/mach-mmp/devices.c +++ b/arch/arm/mach-mmp/devices.c @@ -238,7 +238,7 @@ void pxa_usb_phy_deinit(void __iomem *phy_reg) static u64 __maybe_unused usb_dma_mask = ~(u32)0; #if IS_ENABLED(CONFIG_PHY_PXA_USB) -struct resource pxa168_usb_phy_resources[] = { +static struct resource pxa168_usb_phy_resources[] = { [0] = { .start = PXA168_U2O_PHYBASE, .end = PXA168_U2O_PHYBASE + USB_PHY_RANGE, @@ -259,7 +259,7 @@ struct platform_device pxa168_device_usb_phy = { #endif /* CONFIG_PHY_PXA_USB */ #if IS_ENABLED(CONFIG_USB_MV_UDC) -struct resource pxa168_u2o_resources[] = { +static struct resource pxa168_u2o_resources[] = { /* regbase */ [0] = { .start = PXA168_U2O_REGBASE + U2x_CAPREGS_OFFSET, @@ -294,7 +294,7 @@ struct platform_device pxa168_device_u2o = { #endif /* CONFIG_USB_MV_UDC */ #if IS_ENABLED(CONFIG_USB_EHCI_MV_U2O) -struct resource pxa168_u2oehci_resources[] = { +static struct resource pxa168_u2oehci_resources[] = { [0] = { .start = PXA168_U2O_REGBASE, .end = PXA168_U2O_REGBASE + USB_REG_RANGE, @@ -321,7 +321,7 @@ struct platform_device pxa168_device_u2oehci = { #endif #if IS_ENABLED(CONFIG_USB_MV_OTG) -struct resource pxa168_u2ootg_resources[] = { +static struct resource pxa168_u2ootg_resources[] = { /* regbase */ [0] = { .start = PXA168_U2O_REGBASE + U2x_CAPREGS_OFFSET, From 66ba7c88507344dee68ad1acbdb630473ab36114 Mon Sep 17 00:00:00 2001 From: "Luke D. Jones" Date: Mon, 10 Oct 2022 19:57:02 +1300 Subject: [PATCH 046/121] ALSA: hda/realtek: Correct pin configs for ASUS G533Z The initial fix for ASUS G533Z was based on faulty information. This fixes the pincfg to values that have been verified with no existing module options or other hacks enabled. Enables headphone jack, and 5.1 surround. [ corrected the indent level by tiwai ] Fixes: bc2c23549ccd ("ALSA: hda/realtek: Add pincfg for ASUS G533Z HP jack") Signed-off-by: Luke D. Jones Cc: Link: https://lore.kernel.org/r/20221010065702.35190-1-luke@ljones.dev Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_realtek.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index d89f95ae0efc..77a308a71cd4 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -8449,11 +8449,13 @@ static const struct hda_fixup alc269_fixups[] = { [ALC285_FIXUP_ASUS_G533Z_PINS] = { .type = HDA_FIXUP_PINS, .v.pins = (const struct hda_pintbl[]) { - { 0x14, 0x90170120 }, + { 0x14, 0x90170152 }, /* Speaker Surround Playback Switch */ + { 0x19, 0x03a19020 }, /* Mic Boost Volume */ + { 0x1a, 0x03a11c30 }, /* Mic Boost Volume */ + { 0x1e, 0x90170151 }, /* Rear jack, IN OUT EAPD Detect */ + { 0x21, 0x03211420 }, { } }, - .chained = true, - .chain_id = ALC294_FIXUP_ASUS_G513_PINS, }, [ALC294_FIXUP_ASUS_COEF_1B] = { .type = HDA_FIXUP_VERBS, From 2ea8e1297801f7b0220ebf6ae61a5b74ca83981e Mon Sep 17 00:00:00 2001 From: "Luke D. Jones" Date: Mon, 10 Oct 2022 20:03:47 +1300 Subject: [PATCH 047/121] ALSA: hda/realtek: Add quirk for ASUS GV601R laptop The ASUS ROG X16 (GV601R) series laptop has the same node-to-DAC pairs as early models and the G14, this includes bass speakers which are by default mapped incorrectly to the 0x06 node. Add a quirk to use the same DAC pairs as the G14. Signed-off-by: Luke D. Jones Cc: Link: https://lore.kernel.org/r/20221010070347.36883-1-luke@ljones.dev Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_realtek.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 77a308a71cd4..54a0f6b4ffc7 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -9423,6 +9423,7 @@ static const struct snd_pci_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x1043, 0x1e8e, "ASUS Zephyrus G15", ALC289_FIXUP_ASUS_GA401), SND_PCI_QUIRK(0x1043, 0x1c52, "ASUS Zephyrus G15 2022", ALC289_FIXUP_ASUS_GA401), SND_PCI_QUIRK(0x1043, 0x1f11, "ASUS Zephyrus G14", ALC289_FIXUP_ASUS_GA401), + SND_PCI_QUIRK(0x1043, 0x1f92, "ASUS ROG Flow X16", ALC289_FIXUP_ASUS_GA401), SND_PCI_QUIRK(0x1043, 0x3030, "ASUS ZN270IE", ALC256_FIXUP_ASUS_AIO_GPIO2), SND_PCI_QUIRK(0x1043, 0x831a, "ASUS P901", ALC269_FIXUP_STEREO_DMIC), SND_PCI_QUIRK(0x1043, 0x834a, "ASUS S101", ALC269_FIXUP_STEREO_DMIC), From 4f2e56a59b9947b3e698d3cabcb858765c12b1e8 Mon Sep 17 00:00:00 2001 From: Saranya Gopal Date: Tue, 11 Oct 2022 10:19:16 +0530 Subject: [PATCH 048/121] ALSA: hda/realtek: Add Intel Reference SSID to support headset keys This patch fixes the issue with 3.5mm headset keys on RPL-P platform. [ Rearranged the entry in SSID order by tiwai ] Signed-off-by: Saranya Gopal Signed-off-by: Ninad Naik Cc: Link: https://lore.kernel.org/r/20221011044916.2278867-1-saranya.gopal@intel.com Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_realtek.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 54a0f6b4ffc7..4b076912bbf4 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -9445,6 +9445,7 @@ static const struct snd_pci_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x10ec, 0x10f2, "Intel Reference board", ALC700_FIXUP_INTEL_REFERENCE), SND_PCI_QUIRK(0x10ec, 0x118c, "Medion EE4254 MD62100", ALC256_FIXUP_MEDION_HEADSET_NO_PRESENCE), SND_PCI_QUIRK(0x10ec, 0x1230, "Intel Reference board", ALC295_FIXUP_CHROME_BOOK), + SND_PCI_QUIRK(0x10ec, 0x124c, "Intel Reference board", ALC295_FIXUP_CHROME_BOOK), SND_PCI_QUIRK(0x10ec, 0x1252, "Intel Reference board", ALC295_FIXUP_CHROME_BOOK), SND_PCI_QUIRK(0x10ec, 0x1254, "Intel Reference board", ALC295_FIXUP_CHROME_BOOK), SND_PCI_QUIRK(0x10f7, 0x8338, "Panasonic CF-SZ6", ALC269_FIXUP_HEADSET_MODE), From a70aef7982b012e86dfd39fbb235e76a21ae778a Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 11 Oct 2022 09:01:46 +0200 Subject: [PATCH 049/121] ALSA: rawmidi: Drop register_mutex in snd_rawmidi_free() The register_mutex taken around the dev_unregister callback call in snd_rawmidi_free() may potentially lead to a mutex deadlock, when OSS emulation and a hot unplug are involved. Since the mutex doesn't protect the actual race (as the registration itself is already protected by another means), let's drop it. Link: https://lore.kernel.org/r/CAB7eexJP7w1B0mVgDF0dQ+gWor7UdkiwPczmL7pn91xx8xpzOA@mail.gmail.com Cc: Link: https://lore.kernel.org/r/20221011070147.7611-1-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/core/rawmidi.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/sound/core/rawmidi.c b/sound/core/rawmidi.c index 6963d5a487b3..d8edb6055072 100644 --- a/sound/core/rawmidi.c +++ b/sound/core/rawmidi.c @@ -1899,10 +1899,8 @@ static int snd_rawmidi_free(struct snd_rawmidi *rmidi) snd_info_free_entry(rmidi->proc_entry); rmidi->proc_entry = NULL; - mutex_lock(®ister_mutex); if (rmidi->ops && rmidi->ops->dev_unregister) rmidi->ops->dev_unregister(rmidi); - mutex_unlock(®ister_mutex); snd_rawmidi_free_substreams(&rmidi->streams[SNDRV_RAWMIDI_STREAM_INPUT]); snd_rawmidi_free_substreams(&rmidi->streams[SNDRV_RAWMIDI_STREAM_OUTPUT]); From 97d917879d7f92df09c3f21fd54609a8bcd654b2 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 11 Oct 2022 09:01:47 +0200 Subject: [PATCH 050/121] ALSA: oss: Fix potential deadlock at unregistration We took sound_oss_mutex around the calls of unregister_sound_special() at unregistering OSS devices. This may, however, lead to a deadlock, because we manage the card release via the card's device object, and the release may happen at unregister_sound_special() call -- which will take sound_oss_mutex again in turn. Although the deadlock might be fixed by relaxing the rawmidi mutex in the previous commit, it's safer to move unregister_sound_special() calls themselves out of the sound_oss_mutex, too. The call is race-safe as the function has a spinlock protection by itself. Link: https://lore.kernel.org/r/CAB7eexJP7w1B0mVgDF0dQ+gWor7UdkiwPczmL7pn91xx8xpzOA@mail.gmail.com Cc: Link: https://lore.kernel.org/r/20221011070147.7611-2-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/core/sound_oss.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/sound/core/sound_oss.c b/sound/core/sound_oss.c index 7ed0a2a91035..2751bf2ff61b 100644 --- a/sound/core/sound_oss.c +++ b/sound/core/sound_oss.c @@ -162,7 +162,6 @@ int snd_unregister_oss_device(int type, struct snd_card *card, int dev) mutex_unlock(&sound_oss_mutex); return -ENOENT; } - unregister_sound_special(minor); switch (SNDRV_MINOR_OSS_DEVICE(minor)) { case SNDRV_MINOR_OSS_PCM: track2 = SNDRV_MINOR_OSS(cidx, SNDRV_MINOR_OSS_AUDIO); @@ -174,12 +173,18 @@ int snd_unregister_oss_device(int type, struct snd_card *card, int dev) track2 = SNDRV_MINOR_OSS(cidx, SNDRV_MINOR_OSS_DMMIDI1); break; } - if (track2 >= 0) { - unregister_sound_special(track2); + if (track2 >= 0) snd_oss_minors[track2] = NULL; - } snd_oss_minors[minor] = NULL; mutex_unlock(&sound_oss_mutex); + + /* call unregister_sound_special() outside sound_oss_mutex; + * otherwise may deadlock, as it can trigger the release of a card + */ + unregister_sound_special(minor); + if (track2 >= 0) + unregister_sound_special(track2); + kfree(mptr); return 0; } From b148766e2b8b7b61c9aef53aefedae33f637a1e7 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Wed, 28 Sep 2022 23:31:20 +0200 Subject: [PATCH 051/121] parisc: Reduce kernel size by packing alternative tables The values stored in the length and condition fields of the alternative tables fit into 16 bits, so we can save 4 bytes per alternative table entry. Since a typical 32-bit kernel has more than 3000 entries this saves > 12k of storage on disc. bloat-o-meter shows a reduction of -0.01% by this change: Total: Before=10196505, After=10195529, chg -0.01% $ ls -la vmlinux vmlinux.before -rwxr-xr-x 14437324 vmlinux -rwxr-xr-x 14449512 vmlinux.before Signed-off-by: Helge Deller --- arch/parisc/include/asm/alternative.h | 21 ++++++++++++--------- arch/parisc/kernel/alternative.c | 7 ++++--- 2 files changed, 16 insertions(+), 12 deletions(-) diff --git a/arch/parisc/include/asm/alternative.h b/arch/parisc/include/asm/alternative.h index 0ec54f43d6d2..1ed45fd085d3 100644 --- a/arch/parisc/include/asm/alternative.h +++ b/arch/parisc/include/asm/alternative.h @@ -22,10 +22,10 @@ struct alt_instr { s32 orig_offset; /* offset to original instructions */ - s32 len; /* end of original instructions */ - u32 cond; /* see ALT_COND_XXX */ + s16 len; /* end of original instructions */ + u16 cond; /* see ALT_COND_XXX */ u32 replacement; /* replacement instruction or code */ -}; +} __packed; void set_kernel_text_rw(int enable_read_write); void apply_alternatives_all(void); @@ -35,8 +35,9 @@ void apply_alternatives(struct alt_instr *start, struct alt_instr *end, /* Alternative SMP implementation. */ #define ALTERNATIVE(cond, replacement) "!0:" \ ".section .altinstructions, \"aw\" !" \ - ".word (0b-4-.), 1, " __stringify(cond) "," \ - __stringify(replacement) " !" \ + ".word (0b-4-.) !" \ + ".hword 1, " __stringify(cond) " !" \ + ".word " __stringify(replacement) " !" \ ".previous" #else @@ -44,15 +45,17 @@ void apply_alternatives(struct alt_instr *start, struct alt_instr *end, /* to replace one single instructions by a new instruction */ #define ALTERNATIVE(from, to, cond, replacement)\ .section .altinstructions, "aw" ! \ - .word (from - .), (to - from)/4 ! \ - .word cond, replacement ! \ + .word (from - .) ! \ + .hword (to - from)/4, cond ! \ + .word replacement ! \ .previous /* to replace multiple instructions by new code */ #define ALTERNATIVE_CODE(from, num_instructions, cond, new_instr_ptr)\ .section .altinstructions, "aw" ! \ - .word (from - .), -num_instructions ! \ - .word cond, (new_instr_ptr - .) ! \ + .word (from - .) ! \ + .hword -num_instructions, cond ! \ + .word (new_instr_ptr - .) ! \ .previous #endif /* __ASSEMBLY__ */ diff --git a/arch/parisc/kernel/alternative.c b/arch/parisc/kernel/alternative.c index daa1e9047275..66f5672c70bd 100644 --- a/arch/parisc/kernel/alternative.c +++ b/arch/parisc/kernel/alternative.c @@ -26,7 +26,7 @@ void __init_or_module apply_alternatives(struct alt_instr *start, struct alt_instr *entry; int index = 0, applied = 0; int num_cpus = num_online_cpus(); - u32 cond_check; + u16 cond_check; cond_check = ALT_COND_ALWAYS | ((num_cpus == 1) ? ALT_COND_NO_SMP : 0) | @@ -45,8 +45,9 @@ void __init_or_module apply_alternatives(struct alt_instr *start, for (entry = start; entry < end; entry++, index++) { - u32 *from, cond, replacement; - s32 len; + u32 *from, replacement; + u16 cond; + s16 len; from = (u32 *)((ulong)&entry->orig_offset + entry->orig_offset); len = entry->len; From 027c3d345e2a1ea61d6e4506a250eb392e6e7b18 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Sat, 1 Oct 2022 00:32:07 +0200 Subject: [PATCH 052/121] parisc: Convert PDC console to an early console Rewrite the PDC console to become an early console. Beside the fact that now boot information is visible until another (text- or graphics) console takes over, this benefits as well machines with a yet-unsupported STI console and kgdb. Signed-off-by: Helge Deller --- arch/parisc/include/asm/pdc.h | 3 - arch/parisc/kernel/pdc_cons.c | 240 ++++------------------------------ arch/parisc/kernel/setup.c | 6 +- arch/parisc/kernel/traps.c | 15 +-- drivers/tty/serial/Kconfig | 15 --- lib/Kconfig.kgdb | 2 +- 6 files changed, 32 insertions(+), 249 deletions(-) diff --git a/arch/parisc/include/asm/pdc.h b/arch/parisc/include/asm/pdc.h index b643092d4b98..fcbcf9a96c11 100644 --- a/arch/parisc/include/asm/pdc.h +++ b/arch/parisc/include/asm/pdc.h @@ -19,9 +19,6 @@ extern unsigned long parisc_pat_pdc_cap; /* PDC capabilities (PAT) */ #define PDC_TYPE_SYSTEM_MAP 1 /* 32-bit, but supports PDC_SYSTEM_MAP */ #define PDC_TYPE_SNAKE 2 /* Doesn't support SYSTEM_MAP */ -void pdc_console_init(void); /* in pdc_console.c */ -void pdc_console_restart(void); - void setup_pdc(void); /* in inventory.c */ /* wrapper-functions from pdc.c */ diff --git a/arch/parisc/kernel/pdc_cons.c b/arch/parisc/kernel/pdc_cons.c index 2661cdd256ae..7d0989f523d0 100644 --- a/arch/parisc/kernel/pdc_cons.c +++ b/arch/parisc/kernel/pdc_cons.c @@ -1,46 +1,18 @@ // SPDX-License-Identifier: GPL-2.0-or-later /* - * PDC Console support - ie use firmware to dump text via boot console + * PDC early console support - use PDC firmware to dump text via boot console * - * Copyright (C) 1999-2003 Matthew Wilcox - * Copyright (C) 2000 Martin K Petersen - * Copyright (C) 2000 John Marvin - * Copyright (C) 2000-2003 Paul Bame - * Copyright (C) 2000 Philipp Rumpf - * Copyright (C) 2000 Michael Ang - * Copyright (C) 2000 Grant Grundler - * Copyright (C) 2001-2002 Ryan Bradetich - * Copyright (C) 2001 Helge Deller - * Copyright (C) 2001 Thomas Bogendoerfer - * Copyright (C) 2002 Randolph Chung - * Copyright (C) 2010 Guy Martin + * Copyright (C) 2001-2022 Helge Deller */ -/* - * The PDC console is a simple console, which can be used for debugging - * boot related problems on HP PA-RISC machines. It is also useful when no - * other console works. - * - * This code uses the ROM (=PDC) based functions to read and write characters - * from and to PDC's boot path. - */ - -/* Define EARLY_BOOTUP_DEBUG to debug kernel related boot problems. - * On production kernels EARLY_BOOTUP_DEBUG should be undefined. */ -#define EARLY_BOOTUP_DEBUG - - -#include #include -#include #include -#include -#include +#include +#include #include /* for PAGE0 */ #include /* for iodc_call() proto and friends */ static DEFINE_SPINLOCK(pdc_console_lock); -static struct console pdc_cons; static void pdc_console_write(struct console *co, const char *s, unsigned count) { @@ -54,7 +26,8 @@ static void pdc_console_write(struct console *co, const char *s, unsigned count) spin_unlock_irqrestore(&pdc_console_lock, flags); } -int pdc_console_poll_key(struct console *co) +#ifdef CONFIG_KGDB +static int kgdb_pdc_read_char(void) { int c; unsigned long flags; @@ -63,201 +36,40 @@ int pdc_console_poll_key(struct console *co) c = pdc_iodc_getc(); spin_unlock_irqrestore(&pdc_console_lock, flags); - return c; + return (c <= 0) ? NO_POLL_CHAR : c; } -static int pdc_console_setup(struct console *co, char *options) +static void kgdb_pdc_write_char(u8 chr) { - return 0; + if (PAGE0->mem_cons.cl_class != CL_DUPLEX) + pdc_console_write(NULL, &chr, 1); } -#if defined(CONFIG_PDC_CONSOLE) -#include -#include - -#define PDC_CONS_POLL_DELAY (30 * HZ / 1000) - -static void pdc_console_poll(struct timer_list *unused); -static DEFINE_TIMER(pdc_console_timer, pdc_console_poll); -static struct tty_port tty_port; - -static int pdc_console_tty_open(struct tty_struct *tty, struct file *filp) -{ - tty_port_tty_set(&tty_port, tty); - mod_timer(&pdc_console_timer, jiffies + PDC_CONS_POLL_DELAY); - - return 0; -} - -static void pdc_console_tty_close(struct tty_struct *tty, struct file *filp) -{ - if (tty->count == 1) { - del_timer_sync(&pdc_console_timer); - tty_port_tty_set(&tty_port, NULL); - } -} - -static int pdc_console_tty_write(struct tty_struct *tty, const unsigned char *buf, int count) -{ - pdc_console_write(NULL, buf, count); - return count; -} - -static unsigned int pdc_console_tty_write_room(struct tty_struct *tty) -{ - return 32768; /* no limit, no buffer used */ -} - -static const struct tty_operations pdc_console_tty_ops = { - .open = pdc_console_tty_open, - .close = pdc_console_tty_close, - .write = pdc_console_tty_write, - .write_room = pdc_console_tty_write_room, +static struct kgdb_io kgdb_pdc_io_ops = { + .name = "kgdb_pdc", + .read_char = kgdb_pdc_read_char, + .write_char = kgdb_pdc_write_char, }; - -static void pdc_console_poll(struct timer_list *unused) -{ - int data, count = 0; - - while (1) { - data = pdc_console_poll_key(NULL); - if (data == -1) - break; - tty_insert_flip_char(&tty_port, data & 0xFF, TTY_NORMAL); - count ++; - } - - if (count) - tty_flip_buffer_push(&tty_port); - - if (pdc_cons.flags & CON_ENABLED) - mod_timer(&pdc_console_timer, jiffies + PDC_CONS_POLL_DELAY); -} - -static struct tty_driver *pdc_console_tty_driver; - -static int __init pdc_console_tty_driver_init(void) -{ - struct tty_driver *driver; - int err; - - /* Check if the console driver is still registered. - * It is unregistered if the pdc console was not selected as the - * primary console. */ - - struct console *tmp; - - console_lock(); - for_each_console(tmp) - if (tmp == &pdc_cons) - break; - console_unlock(); - - if (!tmp) { - printk(KERN_INFO "PDC console driver not registered anymore, not creating %s\n", pdc_cons.name); - return -ENODEV; - } - - printk(KERN_INFO "The PDC console driver is still registered, removing CON_BOOT flag\n"); - pdc_cons.flags &= ~CON_BOOT; - - driver = tty_alloc_driver(1, TTY_DRIVER_REAL_RAW | - TTY_DRIVER_RESET_TERMIOS); - if (IS_ERR(driver)) - return PTR_ERR(driver); - - tty_port_init(&tty_port); - - driver->driver_name = "pdc_cons"; - driver->name = "ttyB"; - driver->major = MUX_MAJOR; - driver->minor_start = 0; - driver->type = TTY_DRIVER_TYPE_SYSTEM; - driver->init_termios = tty_std_termios; - tty_set_operations(driver, &pdc_console_tty_ops); - tty_port_link_device(&tty_port, driver, 0); - - err = tty_register_driver(driver); - if (err) { - printk(KERN_ERR "Unable to register the PDC console TTY driver\n"); - tty_port_destroy(&tty_port); - tty_driver_kref_put(driver); - return err; - } - - pdc_console_tty_driver = driver; - - return 0; -} -device_initcall(pdc_console_tty_driver_init); - -static struct tty_driver * pdc_console_device (struct console *c, int *index) -{ - *index = c->index; - return pdc_console_tty_driver; -} -#else -#define pdc_console_device NULL #endif -static struct console pdc_cons = { - .name = "ttyB", - .write = pdc_console_write, - .device = pdc_console_device, - .setup = pdc_console_setup, - .flags = CON_BOOT | CON_PRINTBUFFER, - .index = -1, -}; - -static int pdc_console_initialized; - -static void pdc_console_init_force(void) +static int __init pdc_earlycon_setup(struct earlycon_device *device, + const char *opt) { - if (pdc_console_initialized) - return; - ++pdc_console_initialized; - + struct console *earlycon_console; + /* If the console is duplex then copy the COUT parameters to CIN. */ if (PAGE0->mem_cons.cl_class == CL_DUPLEX) memcpy(&PAGE0->mem_kbd, &PAGE0->mem_cons, sizeof(PAGE0->mem_cons)); - /* register the pdc console */ - register_console(&pdc_cons); -} + earlycon_console = device->con; + earlycon_console->write = pdc_console_write; + device->port.iotype = UPIO_MEM32BE; -void __init pdc_console_init(void) -{ -#if defined(EARLY_BOOTUP_DEBUG) || defined(CONFIG_PDC_CONSOLE) - pdc_console_init_force(); -#endif -#ifdef EARLY_BOOTUP_DEBUG - printk(KERN_INFO "Initialized PDC Console for debugging.\n"); +#ifdef CONFIG_KGDB + kgdb_register_io_module(&kgdb_pdc_io_ops); #endif + + return 0; } - -/* - * Used for emergencies. Currently only used if an HPMC occurs. If an - * HPMC occurs, it is possible that the current console may not be - * properly initialised after the PDC IO reset. This routine unregisters - * all of the current consoles, reinitializes the pdc console and - * registers it. - */ - -void pdc_console_restart(void) -{ - struct console *console; - - if (pdc_console_initialized) - return; - - /* If we've already seen the output, don't bother to print it again */ - if (console_drivers != NULL) - pdc_cons.flags &= ~CON_PRINTBUFFER; - - while ((console = console_drivers) != NULL) - unregister_console(console_drivers); - - /* force registering the pdc console */ - pdc_console_init_force(); -} +EARLYCON_DECLARE(pdc, pdc_earlycon_setup); diff --git a/arch/parisc/kernel/setup.c b/arch/parisc/kernel/setup.c index f005ddedb50e..375f38d6e1a4 100644 --- a/arch/parisc/kernel/setup.c +++ b/arch/parisc/kernel/setup.c @@ -70,6 +70,10 @@ void __init setup_cmdline(char **cmdline_p) strlcat(p, "tty0", COMMAND_LINE_SIZE); } + /* default to use early console */ + if (!strstr(p, "earlycon")) + strlcat(p, " earlycon=pdc", COMMAND_LINE_SIZE); + #ifdef CONFIG_BLK_DEV_INITRD if (boot_args[2] != 0) /* did palo pass us a ramdisk? */ { @@ -139,8 +143,6 @@ void __init setup_arch(char **cmdline_p) if (__pa((unsigned long) &_end) >= KERNEL_INITIAL_SIZE) panic("KERNEL_INITIAL_ORDER too small!"); - pdc_console_init(); - #ifdef CONFIG_64BIT if(parisc_narrow_firmware) { printk(KERN_INFO "Kernel is using PDC in 32-bit mode.\n"); diff --git a/arch/parisc/kernel/traps.c b/arch/parisc/kernel/traps.c index b78f1b9d45c1..f9696fbf646c 100644 --- a/arch/parisc/kernel/traps.c +++ b/arch/parisc/kernel/traps.c @@ -239,13 +239,6 @@ void die_if_kernel(char *str, struct pt_regs *regs, long err) /* unlock the pdc lock if necessary */ pdc_emergency_unlock(); - /* maybe the kernel hasn't booted very far yet and hasn't been able - * to initialize the serial or STI console. In that case we should - * re-enable the pdc console, so that the user will be able to - * identify the problem. */ - if (!console_drivers) - pdc_console_restart(); - if (err) printk(KERN_CRIT "%s (pid %d): %s (code %ld)\n", current->comm, task_pid_nr(current), str, err); @@ -429,10 +422,6 @@ void parisc_terminate(char *msg, struct pt_regs *regs, int code, unsigned long o /* unlock the pdc lock if necessary */ pdc_emergency_unlock(); - /* restart pdc console if necessary */ - if (!console_drivers) - pdc_console_restart(); - /* Not all paths will gutter the processor... */ switch(code){ @@ -482,9 +471,7 @@ void notrace handle_interruption(int code, struct pt_regs *regs) unsigned long fault_space = 0; int si_code; - if (code == 1) - pdc_console_restart(); /* switch back to pdc if HPMC */ - else if (!irqs_disabled_flags(regs->gr[0])) + if (!irqs_disabled_flags(regs->gr[0])) local_irq_enable(); /* Security check: diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig index 877173907c53..898728ab2c18 100644 --- a/drivers/tty/serial/Kconfig +++ b/drivers/tty/serial/Kconfig @@ -602,21 +602,6 @@ config SERIAL_MUX_CONSOLE select SERIAL_CORE_CONSOLE default y -config PDC_CONSOLE - bool "PDC software console support" - depends on PARISC && !SERIAL_MUX && VT - help - Saying Y here will enable the software based PDC console to be - used as the system console. This is useful for machines in - which the hardware based console has not been written yet. The - following steps must be completed to use the PDC console: - - 1. create the device entry (mknod /dev/ttyB0 c 11 0) - 2. Edit the /etc/inittab to start a getty listening on /dev/ttyB0 - 3. Add device ttyB0 to /etc/securetty (if you want to log on as - root on this console.) - 4. Change the kernel command console parameter to: console=ttyB0 - config SERIAL_SUNSAB tristate "Sun Siemens SAB82532 serial support" depends on SPARC && PCI diff --git a/lib/Kconfig.kgdb b/lib/Kconfig.kgdb index 05dae05b6cc9..3b9a44008433 100644 --- a/lib/Kconfig.kgdb +++ b/lib/Kconfig.kgdb @@ -121,7 +121,7 @@ config KDB_DEFAULT_ENABLE config KDB_KEYBOARD bool "KGDB_KDB: keyboard as input device" - depends on VT && KGDB_KDB + depends on VT && KGDB_KDB && !PARISC default n help KDB can use a PS/2 type keyboard for an input device From 71c8517e004b950a148581ea2d2abe10aa46e02d Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Mon, 10 Oct 2022 23:17:04 +0100 Subject: [PATCH 053/121] MAINTAINERS: update polarfire soc clock binding The clock binding has been renamed and a new binding added for the clock controllers in the FPGA fabric. Generalise the pattern to cover both. Signed-off-by: Conor Dooley Link: https://lore.kernel.org/r/20221010221704.2161221-2-conor@kernel.org/ Signed-off-by: Palmer Dabbelt --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 0dc4a769216b..94cf47ea5f80 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -17645,7 +17645,7 @@ M: Conor Dooley M: Daire McNamara L: linux-riscv@lists.infradead.org S: Supported -F: Documentation/devicetree/bindings/clock/microchip,mpfs.yaml +F: Documentation/devicetree/bindings/clock/microchip,mpfs*.yaml F: Documentation/devicetree/bindings/gpio/microchip,mpfs-gpio.yaml F: Documentation/devicetree/bindings/i2c/microchip,corei2c.yaml F: Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.yaml From abbb388d335f8c400d1baecb15d360fa0062de77 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Mon, 10 Oct 2022 23:17:05 +0100 Subject: [PATCH 054/121] dt-bindings: riscv: update microchip.yaml's maintainership Daire and I are the platform maintainers for Microchip's RISC-V FPGAs. Update the maintainers in microchip.yaml to reflect this and explicitly add the binding to the SoC's MAINTAINERS entry. Acked-by: Krzysztof Kozlowski Signed-off-by: Conor Dooley Link: https://lore.kernel.org/r/20221010221704.2161221-3-conor@kernel.org/ Signed-off-by: Palmer Dabbelt --- Documentation/devicetree/bindings/riscv/microchip.yaml | 4 ++-- MAINTAINERS | 1 + 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/Documentation/devicetree/bindings/riscv/microchip.yaml b/Documentation/devicetree/bindings/riscv/microchip.yaml index 1aa7336a9672..9faf8447332b 100644 --- a/Documentation/devicetree/bindings/riscv/microchip.yaml +++ b/Documentation/devicetree/bindings/riscv/microchip.yaml @@ -7,8 +7,8 @@ $schema: http://devicetree.org/meta-schemas/core.yaml# title: Microchip PolarFire SoC-based boards device tree bindings maintainers: - - Cyril Jean - - Lewis Hanly + - Conor Dooley + - Daire McNamara description: Microchip PolarFire SoC-based boards diff --git a/MAINTAINERS b/MAINTAINERS index 94cf47ea5f80..b3415857a812 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -17651,6 +17651,7 @@ F: Documentation/devicetree/bindings/i2c/microchip,corei2c.yaml F: Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.yaml F: Documentation/devicetree/bindings/net/can/microchip,mpfs-can.yaml F: Documentation/devicetree/bindings/pwm/microchip,corepwm.yaml +F: Documentation/devicetree/bindings/riscv/microchip.yaml F: Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-sys-controller.yaml F: Documentation/devicetree/bindings/spi/microchip,mpfs-spi.yaml F: Documentation/devicetree/bindings/usb/microchip,mpfs-musb.yaml From 10f6913c548b32ecb73801a16b120e761c6957ea Mon Sep 17 00:00:00 2001 From: Wenting Zhang Date: Fri, 8 Jul 2022 16:38:22 -0400 Subject: [PATCH 055/121] riscv: always honor the CONFIG_CMDLINE_FORCE when parsing dtb MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When CONFIG_CMDLINE_FORCE is enabled, cmdline provided by CONFIG_CMDLINE are always used. This allows CONFIG_CMDLINE to be used regardless of the result of device tree scanning. This especially fixes the case where a device tree without the chosen node is supplied to the kernel. In such cases, early_init_dt_scan would return true. But inside early_init_dt_scan_chosen, the cmdline won't be updated as there is no chosen node in the device tree. As a result, CONFIG_CMDLINE is not copied into boot_command_line even if CONFIG_CMDLINE_FORCE is enabled. This commit allows properly update boot_command_line in this situation. Fixes: 8fd6e05c7463 ("arch: riscv: support kernel command line forcing when no DTB passed") Signed-off-by: Wenting Zhang Reviewed-by: Björn Töpel Reviewed-by: Conor Dooley Link: https://lore.kernel.org/r/PSBPR04MB399135DFC54928AB958D0638B1829@PSBPR04MB3991.apcprd04.prod.outlook.com Cc: stable@vger.kernel.org Signed-off-by: Palmer Dabbelt --- arch/riscv/kernel/setup.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/riscv/kernel/setup.c b/arch/riscv/kernel/setup.c index 2dfc463b86bb..ad76bb59b059 100644 --- a/arch/riscv/kernel/setup.c +++ b/arch/riscv/kernel/setup.c @@ -252,10 +252,10 @@ static void __init parse_dtb(void) pr_info("Machine model: %s\n", name); dump_stack_set_arch_desc("%s (DT)", name); } - return; + } else { + pr_err("No DTB passed to the kernel\n"); } - pr_err("No DTB passed to the kernel\n"); #ifdef CONFIG_CMDLINE_FORCE strscpy(boot_command_line, CONFIG_CMDLINE, COMMAND_LINE_SIZE); pr_info("Forcing kernel command line to: %s\n", boot_command_line); From 49b0dea1eb5e0fd5e498a2c2ce50d2e036494072 Mon Sep 17 00:00:00 2001 From: Stefan Binding Date: Tue, 11 Oct 2022 15:35:48 +0100 Subject: [PATCH 056/121] ALSA: hda: hda_cs_dsp_ctl: Minor clean and redundant code removal The cs_dsp core will return an error if passed a NULL cs_dsp struct so there is no need for the hda_cs_dsp_write|read_ctl functions to manually check that. The cs_dsp core will also check the data is within bounds of the control so the additional bounds check is redundant too. Simplify things a bit by removing said code. Signed-off-by: Stefan Binding Link: https://lore.kernel.org/r/20221011143552.621792-2-sbinding@opensource.cirrus.com Signed-off-by: Takashi Iwai --- sound/pci/hda/hda_cs_dsp_ctl.c | 17 ++++------------- 1 file changed, 4 insertions(+), 13 deletions(-) diff --git a/sound/pci/hda/hda_cs_dsp_ctl.c b/sound/pci/hda/hda_cs_dsp_ctl.c index 89ee549cb7d5..41d3e8fd289d 100644 --- a/sound/pci/hda/hda_cs_dsp_ctl.c +++ b/sound/pci/hda/hda_cs_dsp_ctl.c @@ -199,16 +199,10 @@ EXPORT_SYMBOL_NS_GPL(hda_cs_dsp_control_remove, SND_HDA_CS_DSP_CONTROLS); int hda_cs_dsp_write_ctl(struct cs_dsp *dsp, const char *name, int type, unsigned int alg, const void *buf, size_t len) { - struct cs_dsp_coeff_ctl *cs_ctl; + struct cs_dsp_coeff_ctl *cs_ctl = cs_dsp_get_ctl(dsp, name, type, alg); struct hda_cs_dsp_coeff_ctl *ctl; int ret; - cs_ctl = cs_dsp_get_ctl(dsp, name, type, alg); - if (!cs_ctl) - return -EINVAL; - - ctl = cs_ctl->priv; - ret = cs_dsp_coeff_write_ctrl(cs_ctl, 0, buf, len); if (ret) return ret; @@ -216,6 +210,8 @@ int hda_cs_dsp_write_ctl(struct cs_dsp *dsp, const char *name, int type, if (cs_ctl->flags & WMFW_CTL_FLAG_SYS) return 0; + ctl = cs_ctl->priv; + snd_ctl_notify(ctl->card, SNDRV_CTL_EVENT_MASK_VALUE, &ctl->kctl->id); return 0; @@ -225,13 +221,8 @@ EXPORT_SYMBOL_NS_GPL(hda_cs_dsp_write_ctl, SND_HDA_CS_DSP_CONTROLS); int hda_cs_dsp_read_ctl(struct cs_dsp *dsp, const char *name, int type, unsigned int alg, void *buf, size_t len) { - struct cs_dsp_coeff_ctl *cs_ctl; + return cs_dsp_coeff_read_ctrl(cs_dsp_get_ctl(dsp, name, type, alg), 0, buf, len); - cs_ctl = cs_dsp_get_ctl(dsp, name, type, alg); - if (!cs_ctl) - return -EINVAL; - - return cs_dsp_coeff_read_ctrl(cs_ctl, 0, buf, len); } EXPORT_SYMBOL_NS_GPL(hda_cs_dsp_read_ctl, SND_HDA_CS_DSP_CONTROLS); From 06f3a0a758c4246dc644e22fb33f85c6e5f92af6 Mon Sep 17 00:00:00 2001 From: Stefan Binding Date: Tue, 11 Oct 2022 15:35:49 +0100 Subject: [PATCH 057/121] ALSA: hda: hda_cs_dsp_ctl: Ensure pwr_lock is held before reading/writing controls These apis require the pwr_lock to be held. Signed-off-by: Stefan Binding Link: https://lore.kernel.org/r/20221011143552.621792-3-sbinding@opensource.cirrus.com Signed-off-by: Takashi Iwai --- sound/pci/hda/hda_cs_dsp_ctl.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/sound/pci/hda/hda_cs_dsp_ctl.c b/sound/pci/hda/hda_cs_dsp_ctl.c index 41d3e8fd289d..75fb69185817 100644 --- a/sound/pci/hda/hda_cs_dsp_ctl.c +++ b/sound/pci/hda/hda_cs_dsp_ctl.c @@ -199,11 +199,14 @@ EXPORT_SYMBOL_NS_GPL(hda_cs_dsp_control_remove, SND_HDA_CS_DSP_CONTROLS); int hda_cs_dsp_write_ctl(struct cs_dsp *dsp, const char *name, int type, unsigned int alg, const void *buf, size_t len) { - struct cs_dsp_coeff_ctl *cs_ctl = cs_dsp_get_ctl(dsp, name, type, alg); + struct cs_dsp_coeff_ctl *cs_ctl; struct hda_cs_dsp_coeff_ctl *ctl; int ret; + mutex_lock(&dsp->pwr_lock); + cs_ctl = cs_dsp_get_ctl(dsp, name, type, alg); ret = cs_dsp_coeff_write_ctrl(cs_ctl, 0, buf, len); + mutex_unlock(&dsp->pwr_lock); if (ret) return ret; @@ -221,7 +224,13 @@ EXPORT_SYMBOL_NS_GPL(hda_cs_dsp_write_ctl, SND_HDA_CS_DSP_CONTROLS); int hda_cs_dsp_read_ctl(struct cs_dsp *dsp, const char *name, int type, unsigned int alg, void *buf, size_t len) { - return cs_dsp_coeff_read_ctrl(cs_dsp_get_ctl(dsp, name, type, alg), 0, buf, len); + int ret; + + mutex_lock(&dsp->pwr_lock); + ret = cs_dsp_coeff_read_ctrl(cs_dsp_get_ctl(dsp, name, type, alg), 0, buf, len); + mutex_unlock(&dsp->pwr_lock); + + return ret; } EXPORT_SYMBOL_NS_GPL(hda_cs_dsp_read_ctl, SND_HDA_CS_DSP_CONTROLS); From 2176c6b599dba55a640cffec0182c0b6bab680d1 Mon Sep 17 00:00:00 2001 From: Richard Fitzgerald Date: Tue, 11 Oct 2022 15:35:50 +0100 Subject: [PATCH 058/121] ALSA: hda/cs_dsp_ctl: Fix mutex inversion when creating controls Redesign the creation of ALSA controls so that the cs_dsp pwr_lock is not held when calling snd_ctl_add(). Instead of creating the ALSA control from the cs_dsp control_add callback, do it after cs_dsp_power_up() has completed. The existing functions are changed to return void instead of passing errors back - this duplicates the original behaviour, as cs_dsp does not abort firmware load if creation of a control fails. It is safe to walk the control list without taking any mutex provided that the caller is not trying to load a new firmware or remove the driver in parallel. There is no other situation that the list can change. So the caller can trigger creation of ALSA controls after cs_dsp_power_up() has returned. A cs_dsp control will have a non-NULL priv pointer if we have created an ALSA control. With the previous code the ALSA controls were created from the cs_dsp control_add callback. But this is called with pwr_lock held (as it is part of the DSP power-up sequence). The kernel lock checking will show a mutex inversion between this and the control creation path: control_add pwr_lock held, takes controls_rwsem (in snd_ctl_add) get/put controls_rwsem held, takes pwr_lock to call cs_dsp. This is not completely theoretical. Although the time window is very small, it is possible for these to run in parallel and deadlock the old implementation. Signed-off-by: Richard Fitzgerald Signed-off-by: Stefan Binding Link: https://lore.kernel.org/r/20221011143552.621792-4-sbinding@opensource.cirrus.com Signed-off-by: Takashi Iwai --- sound/pci/hda/cs35l41_hda.c | 8 ++--- sound/pci/hda/hda_cs_dsp_ctl.c | 61 ++++++++++++++++++++-------------- sound/pci/hda/hda_cs_dsp_ctl.h | 2 +- 3 files changed, 41 insertions(+), 30 deletions(-) diff --git a/sound/pci/hda/cs35l41_hda.c b/sound/pci/hda/cs35l41_hda.c index 3952f2853703..102ac4a94a9d 100644 --- a/sound/pci/hda/cs35l41_hda.c +++ b/sound/pci/hda/cs35l41_hda.c @@ -91,20 +91,18 @@ static const struct reg_sequence cs35l41_hda_mute[] = { { CS35L41_AMP_DIG_VOL_CTRL, 0x0000A678 }, // AMP_VOL_PCM Mute }; -static int cs35l41_control_add(struct cs_dsp_coeff_ctl *cs_ctl) +static void cs35l41_add_controls(struct cs35l41_hda *cs35l41) { - struct cs35l41_hda *cs35l41 = container_of(cs_ctl->dsp, struct cs35l41_hda, cs_dsp); struct hda_cs_dsp_ctl_info info; info.device_name = cs35l41->amp_name; info.fw_type = cs35l41->firmware_type; info.card = cs35l41->codec->card; - return hda_cs_dsp_control_add(cs_ctl, &info); + hda_cs_dsp_add_controls(&cs35l41->cs_dsp, &info); } static const struct cs_dsp_client_ops client_ops = { - .control_add = cs35l41_control_add, .control_remove = hda_cs_dsp_control_remove, }; @@ -435,6 +433,8 @@ static int cs35l41_init_dsp(struct cs35l41_hda *cs35l41) if (ret) goto err_release; + cs35l41_add_controls(cs35l41); + ret = cs35l41_save_calibration(cs35l41); err_release: diff --git a/sound/pci/hda/hda_cs_dsp_ctl.c b/sound/pci/hda/hda_cs_dsp_ctl.c index 75fb69185817..1622a22f96f6 100644 --- a/sound/pci/hda/hda_cs_dsp_ctl.c +++ b/sound/pci/hda/hda_cs_dsp_ctl.c @@ -97,7 +97,7 @@ static unsigned int wmfw_convert_flags(unsigned int in) return out; } -static int hda_cs_dsp_add_kcontrol(struct hda_cs_dsp_coeff_ctl *ctl, const char *name) +static void hda_cs_dsp_add_kcontrol(struct hda_cs_dsp_coeff_ctl *ctl, const char *name) { struct cs_dsp_coeff_ctl *cs_ctl = ctl->cs_ctl; struct snd_kcontrol_new kcontrol = {0}; @@ -107,7 +107,7 @@ static int hda_cs_dsp_add_kcontrol(struct hda_cs_dsp_coeff_ctl *ctl, const char if (cs_ctl->len > ADSP_MAX_STD_CTRL_SIZE) { dev_err(cs_ctl->dsp->dev, "KControl %s: length %zu exceeds maximum %d\n", name, cs_ctl->len, ADSP_MAX_STD_CTRL_SIZE); - return -EINVAL; + return; } kcontrol.name = name; @@ -120,24 +120,21 @@ static int hda_cs_dsp_add_kcontrol(struct hda_cs_dsp_coeff_ctl *ctl, const char /* Save ctl inside private_data, ctl is owned by cs_dsp, * and will be freed when cs_dsp removes the control */ kctl = snd_ctl_new1(&kcontrol, (void *)ctl); - if (!kctl) { - ret = -ENOMEM; - return ret; - } + if (!kctl) + return; ret = snd_ctl_add(ctl->card, kctl); if (ret) { dev_err(cs_ctl->dsp->dev, "Failed to add KControl %s = %d\n", kcontrol.name, ret); - return ret; + return; } dev_dbg(cs_ctl->dsp->dev, "Added KControl: %s\n", kcontrol.name); ctl->kctl = kctl; - - return 0; } -int hda_cs_dsp_control_add(struct cs_dsp_coeff_ctl *cs_ctl, struct hda_cs_dsp_ctl_info *info) +static void hda_cs_dsp_control_add(struct cs_dsp_coeff_ctl *cs_ctl, + const struct hda_cs_dsp_ctl_info *info) { struct cs_dsp *cs_dsp = cs_ctl->dsp; char name[SNDRV_CTL_ELEM_ID_NAME_MAXLEN]; @@ -145,13 +142,10 @@ int hda_cs_dsp_control_add(struct cs_dsp_coeff_ctl *cs_ctl, struct hda_cs_dsp_ct const char *region_name; int ret; - if (cs_ctl->flags & WMFW_CTL_FLAG_SYS) - return 0; - region_name = cs_dsp_mem_region_name(cs_ctl->alg_region.type); if (!region_name) { - dev_err(cs_dsp->dev, "Unknown region type: %d\n", cs_ctl->alg_region.type); - return -EINVAL; + dev_warn(cs_dsp->dev, "Unknown region type: %d\n", cs_ctl->alg_region.type); + return; } ret = scnprintf(name, SNDRV_CTL_ELEM_ID_NAME_MAXLEN, "%s %s %.12s %x", info->device_name, @@ -171,22 +165,39 @@ int hda_cs_dsp_control_add(struct cs_dsp_coeff_ctl *cs_ctl, struct hda_cs_dsp_ct ctl = kzalloc(sizeof(*ctl), GFP_KERNEL); if (!ctl) - return -ENOMEM; + return; ctl->cs_ctl = cs_ctl; ctl->card = info->card; cs_ctl->priv = ctl; - ret = hda_cs_dsp_add_kcontrol(ctl, name); - if (ret) { - dev_err(cs_dsp->dev, "Error (%d) adding control %s\n", ret, name); - kfree(ctl); - return ret; - } - - return 0; + hda_cs_dsp_add_kcontrol(ctl, name); } -EXPORT_SYMBOL_NS_GPL(hda_cs_dsp_control_add, SND_HDA_CS_DSP_CONTROLS); + +void hda_cs_dsp_add_controls(struct cs_dsp *dsp, const struct hda_cs_dsp_ctl_info *info) +{ + struct cs_dsp_coeff_ctl *cs_ctl; + + /* + * pwr_lock would cause mutex inversion with ALSA control lock compared + * to the get/put functions. + * It is safe to walk the list without holding a mutex because entries + * are persistent and only cs_dsp_power_up() or cs_dsp_remove() can + * change the list. + */ + lockdep_assert_not_held(&dsp->pwr_lock); + + list_for_each_entry(cs_ctl, &dsp->ctl_list, list) { + if (cs_ctl->flags & WMFW_CTL_FLAG_SYS) + continue; + + if (cs_ctl->priv) + continue; + + hda_cs_dsp_control_add(cs_ctl, info); + } +} +EXPORT_SYMBOL_NS_GPL(hda_cs_dsp_add_controls, SND_HDA_CS_DSP_CONTROLS); void hda_cs_dsp_control_remove(struct cs_dsp_coeff_ctl *cs_ctl) { diff --git a/sound/pci/hda/hda_cs_dsp_ctl.h b/sound/pci/hda/hda_cs_dsp_ctl.h index 4babc69cf2f0..2cf93359c4f2 100644 --- a/sound/pci/hda/hda_cs_dsp_ctl.h +++ b/sound/pci/hda/hda_cs_dsp_ctl.h @@ -29,7 +29,7 @@ struct hda_cs_dsp_ctl_info { extern const char * const hda_cs_dsp_fw_ids[HDA_CS_DSP_NUM_FW]; -int hda_cs_dsp_control_add(struct cs_dsp_coeff_ctl *cs_ctl, struct hda_cs_dsp_ctl_info *info); +void hda_cs_dsp_add_controls(struct cs_dsp *dsp, const struct hda_cs_dsp_ctl_info *info); void hda_cs_dsp_control_remove(struct cs_dsp_coeff_ctl *cs_ctl); int hda_cs_dsp_write_ctl(struct cs_dsp *dsp, const char *name, int type, unsigned int alg, const void *buf, size_t len); From 23904f7b2518e9b6bbfe2ac7bbe9e284bcdda18e Mon Sep 17 00:00:00 2001 From: Stefan Binding Date: Tue, 11 Oct 2022 15:35:51 +0100 Subject: [PATCH 059/121] ALSA: hda: cs35l41: Remove suspend/resume hda hooks The current code uses calls from the HDA Codec driver to determine when to suspend/resume by calling hooks via the hda_component binding. However, this means the cs35l41 driver relies on the HDA Codec driver to tell it when to suspend or resume, creating an additional external dependency, and potentially creating race conditions in the future. It is better for the cs35l41 hda driver to decide for itself when the part should be suspended or resumed. This makes supporting system suspend easier. Signed-off-by: Stefan Binding Link: https://lore.kernel.org/r/20221011143552.621792-5-sbinding@opensource.cirrus.com Signed-off-by: Takashi Iwai --- sound/pci/hda/cs35l41_hda.c | 31 ++++++++++++------------------- sound/pci/hda/hda_component.h | 2 -- sound/pci/hda/patch_realtek.c | 19 +------------------ 3 files changed, 13 insertions(+), 39 deletions(-) diff --git a/sound/pci/hda/cs35l41_hda.c b/sound/pci/hda/cs35l41_hda.c index 102ac4a94a9d..89f6b4a28d3d 100644 --- a/sound/pci/hda/cs35l41_hda.c +++ b/sound/pci/hda/cs35l41_hda.c @@ -487,10 +487,10 @@ static void cs35l41_hda_playback_hook(struct device *dev, int action) struct regmap *reg = cs35l41->regmap; int ret = 0; - mutex_lock(&cs35l41->fw_mutex); - switch (action) { case HDA_GEN_PCM_ACT_OPEN: + pm_runtime_get_sync(dev); + mutex_lock(&cs35l41->fw_mutex); cs35l41->playback_started = true; if (cs35l41->firmware_running) { regmap_multi_reg_write(reg, cs35l41_hda_config_dsp, @@ -508,15 +508,21 @@ static void cs35l41_hda_playback_hook(struct device *dev, int action) CS35L41_AMP_EN_MASK, 1 << CS35L41_AMP_EN_SHIFT); if (cs35l41->hw_cfg.bst_type == CS35L41_EXT_BOOST) regmap_write(reg, CS35L41_GPIO1_CTRL1, 0x00008001); + mutex_unlock(&cs35l41->fw_mutex); break; case HDA_GEN_PCM_ACT_PREPARE: + mutex_lock(&cs35l41->fw_mutex); ret = cs35l41_global_enable(reg, cs35l41->hw_cfg.bst_type, 1); + mutex_unlock(&cs35l41->fw_mutex); break; case HDA_GEN_PCM_ACT_CLEANUP: + mutex_lock(&cs35l41->fw_mutex); regmap_multi_reg_write(reg, cs35l41_hda_mute, ARRAY_SIZE(cs35l41_hda_mute)); ret = cs35l41_global_enable(reg, cs35l41->hw_cfg.bst_type, 0); + mutex_unlock(&cs35l41->fw_mutex); break; case HDA_GEN_PCM_ACT_CLOSE: + mutex_lock(&cs35l41->fw_mutex); ret = regmap_update_bits(reg, CS35L41_PWR_CTRL2, CS35L41_AMP_EN_MASK, 0 << CS35L41_AMP_EN_SHIFT); if (cs35l41->hw_cfg.bst_type == CS35L41_EXT_BOOST) @@ -530,14 +536,16 @@ static void cs35l41_hda_playback_hook(struct device *dev, int action) } cs35l41_irq_release(cs35l41); cs35l41->playback_started = false; + mutex_unlock(&cs35l41->fw_mutex); + + pm_runtime_mark_last_busy(dev); + pm_runtime_put_autosuspend(dev); break; default: dev_warn(cs35l41->dev, "Playback action not supported: %d\n", action); break; } - mutex_unlock(&cs35l41->fw_mutex); - if (ret) dev_err(cs35l41->dev, "Regmap access fail: %d\n", ret); } @@ -618,19 +626,6 @@ static int cs35l41_runtime_resume(struct device *dev) return 0; } -static int cs35l41_hda_suspend_hook(struct device *dev) -{ - dev_dbg(dev, "Request Suspend\n"); - pm_runtime_mark_last_busy(dev); - return pm_runtime_put_autosuspend(dev); -} - -static int cs35l41_hda_resume_hook(struct device *dev) -{ - dev_dbg(dev, "Request Resume\n"); - return pm_runtime_get_sync(dev); -} - static int cs35l41_smart_amp(struct cs35l41_hda *cs35l41) { int halo_sts; @@ -863,8 +858,6 @@ static int cs35l41_hda_bind(struct device *dev, struct device *master, void *mas ret = cs35l41_create_controls(cs35l41); comps->playback_hook = cs35l41_hda_playback_hook; - comps->suspend_hook = cs35l41_hda_suspend_hook; - comps->resume_hook = cs35l41_hda_resume_hook; pm_runtime_mark_last_busy(dev); pm_runtime_put_autosuspend(dev); diff --git a/sound/pci/hda/hda_component.h b/sound/pci/hda/hda_component.h index 1223621bd62c..534e845b9cd1 100644 --- a/sound/pci/hda/hda_component.h +++ b/sound/pci/hda/hda_component.h @@ -16,6 +16,4 @@ struct hda_component { char name[HDA_MAX_NAME_SIZE]; struct hda_codec *codec; void (*playback_hook)(struct device *dev, int action); - int (*suspend_hook)(struct device *dev); - int (*resume_hook)(struct device *dev); }; diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 4b076912bbf4..e6c4bb5fa041 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -4022,22 +4022,16 @@ static void alc5505_dsp_init(struct hda_codec *codec) static int alc269_suspend(struct hda_codec *codec) { struct alc_spec *spec = codec->spec; - int i; if (spec->has_alc5505_dsp) alc5505_dsp_suspend(codec); - for (i = 0; i < HDA_MAX_COMPONENTS; i++) - if (spec->comps[i].suspend_hook) - spec->comps[i].suspend_hook(spec->comps[i].dev); - return alc_suspend(codec); } static int alc269_resume(struct hda_codec *codec) { struct alc_spec *spec = codec->spec; - int i; if (spec->codec_variant == ALC269_TYPE_ALC269VB) alc269vb_toggle_power_output(codec, 0); @@ -4068,10 +4062,6 @@ static int alc269_resume(struct hda_codec *codec) if (spec->has_alc5505_dsp) alc5505_dsp_resume(codec); - for (i = 0; i < HDA_MAX_COMPONENTS; i++) - if (spec->comps[i].resume_hook) - spec->comps[i].resume_hook(spec->comps[i].dev); - return 0; } #endif /* CONFIG_PM */ @@ -6664,19 +6654,12 @@ static int comp_bind(struct device *dev) { struct hda_codec *cdc = dev_to_hda_codec(dev); struct alc_spec *spec = cdc->spec; - int ret, i; + int ret; ret = component_bind_all(dev, spec->comps); if (ret) return ret; - if (snd_hdac_is_power_on(&cdc->core)) { - codec_dbg(cdc, "Resuming after bind.\n"); - for (i = 0; i < HDA_MAX_COMPONENTS; i++) - if (spec->comps[i].resume_hook) - spec->comps[i].resume_hook(spec->comps[i].dev); - } - return 0; } From 88672826e2a465d2f4c0a50fb5ced2956f4ffcbc Mon Sep 17 00:00:00 2001 From: Stefan Binding Date: Tue, 11 Oct 2022 15:35:52 +0100 Subject: [PATCH 060/121] ALSA: hda: cs35l41: Support System Suspend Add support for system suspend into the CS35L41 HDA Driver. Since S4 suspend may power off the system, it is required that the driver ensure the part is safe to be shutdown before system suspend, as well as ensuring that the firmware is unloaded before shutdown. The part must then be restored on system resume, including re-downloading the firmware. Signed-off-by: Stefan Binding Link: https://lore.kernel.org/r/20221011143552.621792-6-sbinding@opensource.cirrus.com Signed-off-by: Takashi Iwai --- sound/pci/hda/cs35l41_hda.c | 160 ++++++++++++++++++++++++++++++------ 1 file changed, 136 insertions(+), 24 deletions(-) diff --git a/sound/pci/hda/cs35l41_hda.c b/sound/pci/hda/cs35l41_hda.c index 89f6b4a28d3d..e5f0549bf06d 100644 --- a/sound/pci/hda/cs35l41_hda.c +++ b/sound/pci/hda/cs35l41_hda.c @@ -461,9 +461,12 @@ static void cs35l41_remove_dsp(struct cs35l41_hda *cs35l41) struct cs_dsp *dsp = &cs35l41->cs_dsp; cancel_work_sync(&cs35l41->fw_load_work); + + mutex_lock(&cs35l41->fw_mutex); cs35l41_shutdown_dsp(cs35l41); cs_dsp_remove(dsp); cs35l41->halo_initialized = false; + mutex_unlock(&cs35l41->fw_mutex); } /* Protection release cycle to get the speaker out of Safe-Mode */ @@ -570,45 +573,148 @@ static int cs35l41_hda_channel_map(struct device *dev, unsigned int tx_num, unsi rx_slot); } +static void cs35l41_ready_for_reset(struct cs35l41_hda *cs35l41) +{ + mutex_lock(&cs35l41->fw_mutex); + if (cs35l41->firmware_running) { + + regcache_cache_only(cs35l41->regmap, false); + + cs35l41_exit_hibernate(cs35l41->dev, cs35l41->regmap); + cs35l41_shutdown_dsp(cs35l41); + cs35l41_safe_reset(cs35l41->regmap, cs35l41->hw_cfg.bst_type); + + regcache_cache_only(cs35l41->regmap, true); + regcache_mark_dirty(cs35l41->regmap); + } + mutex_unlock(&cs35l41->fw_mutex); +} + +static int cs35l41_system_suspend(struct device *dev) +{ + struct cs35l41_hda *cs35l41 = dev_get_drvdata(dev); + int ret; + + dev_dbg(cs35l41->dev, "System Suspend\n"); + + if (cs35l41->hw_cfg.bst_type == CS35L41_EXT_BOOST_NO_VSPK_SWITCH) { + dev_err(cs35l41->dev, "System Suspend not supported\n"); + return -EINVAL; + } + + ret = pm_runtime_force_suspend(dev); + if (ret) + return ret; + + /* Shutdown DSP before system suspend */ + cs35l41_ready_for_reset(cs35l41); + + /* + * Reset GPIO may be shared, so cannot reset here. + * However beyond this point, amps may be powered down. + */ + return 0; +} + +static int cs35l41_system_resume(struct device *dev) +{ + struct cs35l41_hda *cs35l41 = dev_get_drvdata(dev); + int ret; + + dev_dbg(cs35l41->dev, "System Resume\n"); + + if (cs35l41->hw_cfg.bst_type == CS35L41_EXT_BOOST_NO_VSPK_SWITCH) { + dev_err(cs35l41->dev, "System Resume not supported\n"); + return -EINVAL; + } + + if (cs35l41->reset_gpio) { + usleep_range(2000, 2100); + gpiod_set_value_cansleep(cs35l41->reset_gpio, 1); + } + + usleep_range(2000, 2100); + + ret = pm_runtime_force_resume(dev); + + mutex_lock(&cs35l41->fw_mutex); + if (!ret && cs35l41->request_fw_load && !cs35l41->fw_request_ongoing) { + cs35l41->fw_request_ongoing = true; + schedule_work(&cs35l41->fw_load_work); + } + mutex_unlock(&cs35l41->fw_mutex); + + return ret; +} + static int cs35l41_runtime_suspend(struct device *dev) { struct cs35l41_hda *cs35l41 = dev_get_drvdata(dev); + int ret = 0; - dev_dbg(cs35l41->dev, "Suspend\n"); + dev_dbg(cs35l41->dev, "Runtime Suspend\n"); - if (!cs35l41->firmware_running) + if (cs35l41->hw_cfg.bst_type == CS35L41_EXT_BOOST_NO_VSPK_SWITCH) { + dev_dbg(cs35l41->dev, "Runtime Suspend not supported\n"); return 0; + } - if (cs35l41_enter_hibernate(cs35l41->dev, cs35l41->regmap, cs35l41->hw_cfg.bst_type) < 0) - return 0; + mutex_lock(&cs35l41->fw_mutex); + + if (cs35l41->playback_started) { + regmap_multi_reg_write(cs35l41->regmap, cs35l41_hda_mute, + ARRAY_SIZE(cs35l41_hda_mute)); + cs35l41_global_enable(cs35l41->regmap, cs35l41->hw_cfg.bst_type, 0); + regmap_update_bits(cs35l41->regmap, CS35L41_PWR_CTRL2, + CS35L41_AMP_EN_MASK, 0 << CS35L41_AMP_EN_SHIFT); + if (cs35l41->hw_cfg.bst_type == CS35L41_EXT_BOOST) + regmap_write(cs35l41->regmap, CS35L41_GPIO1_CTRL1, 0x00000001); + regmap_update_bits(cs35l41->regmap, CS35L41_PWR_CTRL2, + CS35L41_VMON_EN_MASK | CS35L41_IMON_EN_MASK, + 0 << CS35L41_VMON_EN_SHIFT | 0 << CS35L41_IMON_EN_SHIFT); + cs35l41->playback_started = false; + } + + if (cs35l41->firmware_running) { + ret = cs35l41_enter_hibernate(cs35l41->dev, cs35l41->regmap, + cs35l41->hw_cfg.bst_type); + if (ret) + goto err; + } else { + cs35l41_safe_reset(cs35l41->regmap, cs35l41->hw_cfg.bst_type); + } regcache_cache_only(cs35l41->regmap, true); regcache_mark_dirty(cs35l41->regmap); - return 0; +err: + mutex_unlock(&cs35l41->fw_mutex); + + return ret; } static int cs35l41_runtime_resume(struct device *dev) { struct cs35l41_hda *cs35l41 = dev_get_drvdata(dev); - int ret; + int ret = 0; - dev_dbg(cs35l41->dev, "Resume.\n"); + dev_dbg(cs35l41->dev, "Runtime Resume\n"); if (cs35l41->hw_cfg.bst_type == CS35L41_EXT_BOOST_NO_VSPK_SWITCH) { - dev_dbg(cs35l41->dev, "System does not support Resume\n"); + dev_dbg(cs35l41->dev, "Runtime Resume not supported\n"); return 0; } - if (!cs35l41->firmware_running) - return 0; + mutex_lock(&cs35l41->fw_mutex); regcache_cache_only(cs35l41->regmap, false); - ret = cs35l41_exit_hibernate(cs35l41->dev, cs35l41->regmap); - if (ret) { - regcache_cache_only(cs35l41->regmap, true); - return ret; + if (cs35l41->firmware_running) { + ret = cs35l41_exit_hibernate(cs35l41->dev, cs35l41->regmap); + if (ret) { + dev_warn(cs35l41->dev, "Unable to exit Hibernate."); + goto err; + } } /* Test key needs to be unlocked to allow the OTP settings to re-apply */ @@ -617,13 +723,16 @@ static int cs35l41_runtime_resume(struct device *dev) cs35l41_test_key_lock(cs35l41->dev, cs35l41->regmap); if (ret) { dev_err(cs35l41->dev, "Failed to restore register cache: %d\n", ret); - return ret; + goto err; } if (cs35l41->hw_cfg.bst_type == CS35L41_EXT_BOOST) cs35l41_init_boost(cs35l41->dev, cs35l41->regmap, &cs35l41->hw_cfg); - return 0; +err: + mutex_unlock(&cs35l41->fw_mutex); + + return ret; } static int cs35l41_smart_amp(struct cs35l41_hda *cs35l41) @@ -673,8 +782,6 @@ static int cs35l41_smart_amp(struct cs35l41_hda *cs35l41) static void cs35l41_load_firmware(struct cs35l41_hda *cs35l41, bool load) { - pm_runtime_get_sync(cs35l41->dev); - if (cs35l41->firmware_running && !load) { dev_dbg(cs35l41->dev, "Unloading Firmware\n"); cs35l41_shutdown_dsp(cs35l41); @@ -684,9 +791,6 @@ static void cs35l41_load_firmware(struct cs35l41_hda *cs35l41, bool load) } else { dev_dbg(cs35l41->dev, "Unable to Load firmware.\n"); } - - pm_runtime_mark_last_busy(cs35l41->dev); - pm_runtime_put_autosuspend(cs35l41->dev); } static int cs35l41_fw_load_ctl_get(struct snd_kcontrol *kcontrol, @@ -702,16 +806,21 @@ static void cs35l41_fw_load_work(struct work_struct *work) { struct cs35l41_hda *cs35l41 = container_of(work, struct cs35l41_hda, fw_load_work); + pm_runtime_get_sync(cs35l41->dev); + mutex_lock(&cs35l41->fw_mutex); /* Recheck if playback is ongoing, mutex will block playback during firmware loading */ if (cs35l41->playback_started) - dev_err(cs35l41->dev, "Cannot Load/Unload firmware during Playback\n"); + dev_err(cs35l41->dev, "Cannot Load/Unload firmware during Playback. Retrying...\n"); else cs35l41_load_firmware(cs35l41, cs35l41->request_fw_load); cs35l41->fw_request_ongoing = false; mutex_unlock(&cs35l41->fw_mutex); + + pm_runtime_mark_last_busy(cs35l41->dev); + pm_runtime_put_autosuspend(cs35l41->dev); } static int cs35l41_fw_load_ctl_put(struct snd_kcontrol *kcontrol, @@ -835,6 +944,8 @@ static int cs35l41_hda_bind(struct device *dev, struct device *master, void *mas pm_runtime_get_sync(dev); + mutex_lock(&cs35l41->fw_mutex); + comps->dev = dev; if (!cs35l41->acpi_subsystem_id) cs35l41->acpi_subsystem_id = kasprintf(GFP_KERNEL, "%.8x", @@ -847,10 +958,8 @@ static int cs35l41_hda_bind(struct device *dev, struct device *master, void *mas if (firmware_autostart) { dev_dbg(cs35l41->dev, "Firmware Autostart.\n"); cs35l41->request_fw_load = true; - mutex_lock(&cs35l41->fw_mutex); if (cs35l41_smart_amp(cs35l41) < 0) dev_warn(cs35l41->dev, "Cannot Run Firmware, reverting to dsp bypass...\n"); - mutex_unlock(&cs35l41->fw_mutex); } else { dev_dbg(cs35l41->dev, "Firmware Autostart is disabled.\n"); } @@ -859,6 +968,8 @@ static int cs35l41_hda_bind(struct device *dev, struct device *master, void *mas comps->playback_hook = cs35l41_hda_playback_hook; + mutex_unlock(&cs35l41->fw_mutex); + pm_runtime_mark_last_busy(dev); pm_runtime_put_autosuspend(dev); @@ -1426,6 +1537,7 @@ EXPORT_SYMBOL_NS_GPL(cs35l41_hda_remove, SND_HDA_SCODEC_CS35L41); const struct dev_pm_ops cs35l41_hda_pm_ops = { RUNTIME_PM_OPS(cs35l41_runtime_suspend, cs35l41_runtime_resume, NULL) + SYSTEM_SLEEP_PM_OPS(cs35l41_system_suspend, cs35l41_system_resume) }; EXPORT_SYMBOL_NS_GPL(cs35l41_hda_pm_ops, SND_HDA_SCODEC_CS35L41); From a8e5e5146ad08d794c58252bab00b261045ef16d Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Thu, 6 Oct 2022 17:33:54 +0100 Subject: [PATCH 061/121] arm64: mte: Avoid setting PG_mte_tagged if no tags cleared or restored Prior to commit 69e3b846d8a7 ("arm64: mte: Sync tags for pages where PTE is untagged"), mte_sync_tags() was only called for pte_tagged() entries (those mapped with PROT_MTE). Therefore mte_sync_tags() could safely use test_and_set_bit(PG_mte_tagged, &page->flags) without inadvertently setting PG_mte_tagged on an untagged page. The above commit was required as guests may enable MTE without any control at the stage 2 mapping, nor a PROT_MTE mapping in the VMM. However, the side-effect was that any page with a PTE that looked like swap (or migration) was getting PG_mte_tagged set automatically. A subsequent page copy (e.g. migration) copied the tags to the destination page even if the tags were owned by KASAN. This issue was masked by the page_kasan_tag_reset() call introduced in commit e5b8d9218951 ("arm64: mte: reset the page tag in page->flags"). When this commit was reverted (20794545c146), KASAN started reporting access faults because the overriding tags in a page did not match the original page->flags (with CONFIG_KASAN_HW_TAGS=y): BUG: KASAN: invalid-access in copy_page+0x10/0xd0 arch/arm64/lib/copy_page.S:26 Read at addr f5ff000017f2e000 by task syz-executor.1/2218 Pointer tag: [f5], memory tag: [f2] Move the PG_mte_tagged bit setting from mte_sync_tags() to the actual place where tags are cleared (mte_sync_page_tags()) or restored (mte_restore_tags()). Signed-off-by: Catalin Marinas Reported-by: syzbot+c2c79c6d6eddc5262b77@syzkaller.appspotmail.com Fixes: 69e3b846d8a7 ("arm64: mte: Sync tags for pages where PTE is untagged") Cc: # 5.14.x Cc: Steven Price Cc: Andrey Konovalov Cc: Vincenzo Frascino Cc: Will Deacon Link: https://lore.kernel.org/r/0000000000004387dc05e5888ae5@google.com/ Reviewed-by: Steven Price Link: https://lore.kernel.org/r/20221006163354.3194102-1-catalin.marinas@arm.com Signed-off-by: Catalin Marinas --- arch/arm64/kernel/mte.c | 9 +++++++-- arch/arm64/mm/mteswap.c | 7 ++++++- 2 files changed, 13 insertions(+), 3 deletions(-) diff --git a/arch/arm64/kernel/mte.c b/arch/arm64/kernel/mte.c index aca88470fb69..7467217c1eaf 100644 --- a/arch/arm64/kernel/mte.c +++ b/arch/arm64/kernel/mte.c @@ -48,7 +48,12 @@ static void mte_sync_page_tags(struct page *page, pte_t old_pte, if (!pte_is_tagged) return; - mte_clear_page_tags(page_address(page)); + /* + * Test PG_mte_tagged again in case it was racing with another + * set_pte_at(). + */ + if (!test_and_set_bit(PG_mte_tagged, &page->flags)) + mte_clear_page_tags(page_address(page)); } void mte_sync_tags(pte_t old_pte, pte_t pte) @@ -64,7 +69,7 @@ void mte_sync_tags(pte_t old_pte, pte_t pte) /* if PG_mte_tagged is set, tags have already been initialised */ for (i = 0; i < nr_pages; i++, page++) { - if (!test_and_set_bit(PG_mte_tagged, &page->flags)) + if (!test_bit(PG_mte_tagged, &page->flags)) mte_sync_page_tags(page, old_pte, check_swap, pte_is_tagged); } diff --git a/arch/arm64/mm/mteswap.c b/arch/arm64/mm/mteswap.c index 4334dec93bd4..bed803d8e158 100644 --- a/arch/arm64/mm/mteswap.c +++ b/arch/arm64/mm/mteswap.c @@ -53,7 +53,12 @@ bool mte_restore_tags(swp_entry_t entry, struct page *page) if (!tags) return false; - mte_restore_page_tags(page_address(page), tags); + /* + * Test PG_mte_tagged again in case it was racing with another + * set_pte_at(). + */ + if (!test_and_set_bit(PG_mte_tagged, &page->flags)) + mte_restore_page_tags(page_address(page), tags); return true; } From e237506238352f3bfa9cf3983cdab873e35651eb Mon Sep 17 00:00:00 2001 From: Nicholas Piggin Date: Wed, 12 Oct 2022 13:53:34 +1000 Subject: [PATCH 062/121] powerpc/32: fix syscall wrappers with 64-bit arguments of unaligned register-pairs powerpc 32-bit system call (and function) calling convention for 64-bit arguments requires the next available odd-pair (two sequential registers with the first being odd-numbered) from the standard register argument allocation. The first argument register is r3, so a 64-bit argument that appears at an even position in the argument list must skip a register (unless there were preceding 64-bit arguments, which might throw things off). This requires non-standard compat definitions to deal with the holes in the argument register allocation. With pt_regs syscall wrappers which use a standard mapper to map pt_regs GPRs to function arguments, 32-bit kernels hit the same basic problem, the standard definitions don't cope with the unused argument registers. Fix this by having 32-bit kernels share those syscall definitions with compat. Thanks to Jason for spending a lot of time finding and bisecting this and developing a trivial reproducer. The perfect bug report. Reported-by: Jason A. Donenfeld Signed-off-by: Nicholas Piggin Fixes: 7e92e01b72452 ("powerpc: Provide syscall wrapper") Reviewed-by: Arnd Bergmann Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20221012035335.866440-1-npiggin@gmail.com --- arch/powerpc/include/asm/syscalls.h | 16 ++++++++++ arch/powerpc/kernel/Makefile | 1 + arch/powerpc/kernel/sys_ppc32.c | 38 ++++++++++++++++++------ arch/powerpc/kernel/syscalls/syscall.tbl | 16 ++++++---- 4 files changed, 56 insertions(+), 15 deletions(-) diff --git a/arch/powerpc/include/asm/syscalls.h b/arch/powerpc/include/asm/syscalls.h index 9840d572da55..a1142496cd58 100644 --- a/arch/powerpc/include/asm/syscalls.h +++ b/arch/powerpc/include/asm/syscalls.h @@ -89,6 +89,22 @@ long compat_sys_rt_sigreturn(void); * responsible for combining parameter pairs. */ +#ifdef CONFIG_PPC32 +long sys_ppc_pread64(unsigned int fd, + char __user *ubuf, compat_size_t count, + u32 reg6, u32 pos1, u32 pos2); +long sys_ppc_pwrite64(unsigned int fd, + const char __user *ubuf, compat_size_t count, + u32 reg6, u32 pos1, u32 pos2); +long sys_ppc_readahead(int fd, u32 r4, + u32 offset1, u32 offset2, u32 count); +long sys_ppc_truncate64(const char __user *path, u32 reg4, + unsigned long len1, unsigned long len2); +long sys_ppc_ftruncate64(unsigned int fd, u32 reg4, + unsigned long len1, unsigned long len2); +long sys_ppc32_fadvise64(int fd, u32 unused, u32 offset1, u32 offset2, + size_t len, int advice); +#endif #ifdef CONFIG_COMPAT long compat_sys_mmap2(unsigned long addr, size_t len, unsigned long prot, unsigned long flags, diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile index ee2d76cb3187..9b6146056e48 100644 --- a/arch/powerpc/kernel/Makefile +++ b/arch/powerpc/kernel/Makefile @@ -73,6 +73,7 @@ obj-y := cputable.o syscalls.o \ obj-y += ptrace/ obj-$(CONFIG_PPC64) += setup_64.o irq_64.o\ paca.o nvram_64.o note.o +obj-$(CONFIG_PPC32) += sys_ppc32.o obj-$(CONFIG_COMPAT) += sys_ppc32.o signal_32.o obj-$(CONFIG_VDSO32) += vdso32_wrapper.o obj-$(CONFIG_PPC_WATCHDOG) += watchdog.o diff --git a/arch/powerpc/kernel/sys_ppc32.c b/arch/powerpc/kernel/sys_ppc32.c index dcc3c9fd4cfd..1ab4a4d95aba 100644 --- a/arch/powerpc/kernel/sys_ppc32.c +++ b/arch/powerpc/kernel/sys_ppc32.c @@ -1,13 +1,23 @@ // SPDX-License-Identifier: GPL-2.0-or-later /* - * sys_ppc32.c: Conversion between 32bit and 64bit native syscalls. + * sys_ppc32.c: 32-bit system calls with complex calling conventions. * * Copyright (C) 2001 IBM * Copyright (C) 1997,1998 Jakub Jelinek (jj@sunsite.mff.cuni.cz) * Copyright (C) 1997 David S. Miller (davem@caip.rutgers.edu) * - * These routines maintain argument size conversion between 32bit and 64bit - * environment. + * 32-bit system calls with 64-bit arguments pass those in register pairs. + * This must be specially dealt with on 64-bit kernels. The compat_arg_u64_dual + * in generic compat syscalls is not always usable because the register + * pairing is constrained depending on preceding arguments. + * + * An analogous problem exists on 32-bit kernels with ARCH_HAS_SYSCALL_WRAPPER, + * the defined system call functions take the pt_regs as an argument, and there + * is a mapping macro which maps registers to arguments + * (SC_POWERPC_REGS_TO_ARGS) which also does not deal with these 64-bit + * arguments. + * + * This file contains these system calls. */ #include @@ -47,7 +57,17 @@ #include #include -COMPAT_SYSCALL_DEFINE6(ppc_pread64, +#ifdef CONFIG_PPC32 +#define PPC32_SYSCALL_DEFINE4 SYSCALL_DEFINE4 +#define PPC32_SYSCALL_DEFINE5 SYSCALL_DEFINE5 +#define PPC32_SYSCALL_DEFINE6 SYSCALL_DEFINE6 +#else +#define PPC32_SYSCALL_DEFINE4 COMPAT_SYSCALL_DEFINE4 +#define PPC32_SYSCALL_DEFINE5 COMPAT_SYSCALL_DEFINE5 +#define PPC32_SYSCALL_DEFINE6 COMPAT_SYSCALL_DEFINE6 +#endif + +PPC32_SYSCALL_DEFINE6(ppc_pread64, unsigned int, fd, char __user *, ubuf, compat_size_t, count, u32, reg6, u32, pos1, u32, pos2) @@ -55,7 +75,7 @@ COMPAT_SYSCALL_DEFINE6(ppc_pread64, return ksys_pread64(fd, ubuf, count, merge_64(pos1, pos2)); } -COMPAT_SYSCALL_DEFINE6(ppc_pwrite64, +PPC32_SYSCALL_DEFINE6(ppc_pwrite64, unsigned int, fd, const char __user *, ubuf, compat_size_t, count, u32, reg6, u32, pos1, u32, pos2) @@ -63,28 +83,28 @@ COMPAT_SYSCALL_DEFINE6(ppc_pwrite64, return ksys_pwrite64(fd, ubuf, count, merge_64(pos1, pos2)); } -COMPAT_SYSCALL_DEFINE5(ppc_readahead, +PPC32_SYSCALL_DEFINE5(ppc_readahead, int, fd, u32, r4, u32, offset1, u32, offset2, u32, count) { return ksys_readahead(fd, merge_64(offset1, offset2), count); } -COMPAT_SYSCALL_DEFINE4(ppc_truncate64, +PPC32_SYSCALL_DEFINE4(ppc_truncate64, const char __user *, path, u32, reg4, unsigned long, len1, unsigned long, len2) { return ksys_truncate(path, merge_64(len1, len2)); } -COMPAT_SYSCALL_DEFINE4(ppc_ftruncate64, +PPC32_SYSCALL_DEFINE4(ppc_ftruncate64, unsigned int, fd, u32, reg4, unsigned long, len1, unsigned long, len2) { return ksys_ftruncate(fd, merge_64(len1, len2)); } -COMPAT_SYSCALL_DEFINE6(ppc32_fadvise64, +PPC32_SYSCALL_DEFINE6(ppc32_fadvise64, int, fd, u32, unused, u32, offset1, u32, offset2, size_t, len, int, advice) { diff --git a/arch/powerpc/kernel/syscalls/syscall.tbl b/arch/powerpc/kernel/syscalls/syscall.tbl index 2bca64f96164..e9e0df4f9a61 100644 --- a/arch/powerpc/kernel/syscalls/syscall.tbl +++ b/arch/powerpc/kernel/syscalls/syscall.tbl @@ -228,8 +228,10 @@ 176 64 rt_sigtimedwait sys_rt_sigtimedwait 177 nospu rt_sigqueueinfo sys_rt_sigqueueinfo compat_sys_rt_sigqueueinfo 178 nospu rt_sigsuspend sys_rt_sigsuspend compat_sys_rt_sigsuspend -179 common pread64 sys_pread64 compat_sys_ppc_pread64 -180 common pwrite64 sys_pwrite64 compat_sys_ppc_pwrite64 +179 32 pread64 sys_ppc_pread64 compat_sys_ppc_pread64 +179 64 pread64 sys_pread64 +180 32 pwrite64 sys_ppc_pwrite64 compat_sys_ppc_pwrite64 +180 64 pwrite64 sys_pwrite64 181 common chown sys_chown 182 common getcwd sys_getcwd 183 common capget sys_capget @@ -242,10 +244,11 @@ 188 common putpmsg sys_ni_syscall 189 nospu vfork sys_vfork 190 common ugetrlimit sys_getrlimit compat_sys_getrlimit -191 common readahead sys_readahead compat_sys_ppc_readahead +191 32 readahead sys_ppc_readahead compat_sys_ppc_readahead +191 64 readahead sys_readahead 192 32 mmap2 sys_mmap2 compat_sys_mmap2 -193 32 truncate64 sys_truncate64 compat_sys_ppc_truncate64 -194 32 ftruncate64 sys_ftruncate64 compat_sys_ppc_ftruncate64 +193 32 truncate64 sys_ppc_truncate64 compat_sys_ppc_truncate64 +194 32 ftruncate64 sys_ppc_ftruncate64 compat_sys_ppc_ftruncate64 195 32 stat64 sys_stat64 196 32 lstat64 sys_lstat64 197 32 fstat64 sys_fstat64 @@ -288,7 +291,8 @@ 230 common io_submit sys_io_submit compat_sys_io_submit 231 common io_cancel sys_io_cancel 232 nospu set_tid_address sys_set_tid_address -233 common fadvise64 sys_fadvise64 compat_sys_ppc32_fadvise64 +233 32 fadvise64 sys_ppc32_fadvise64 compat_sys_ppc32_fadvise64 +233 64 fadvise64 sys_fadvise64 234 nospu exit_group sys_exit_group 235 nospu lookup_dcookie sys_lookup_dcookie compat_sys_lookup_dcookie 236 common epoll_create sys_epoll_create From 06267eb2decaa6baac81bbd882265a8e7782dba4 Mon Sep 17 00:00:00 2001 From: Palmer Dabbelt Date: Thu, 28 Jul 2022 14:07:15 -0700 Subject: [PATCH 063/121] doc: RISC-V: Document that misaligned accesses are supported The RISC-V ISA manual used to mandate that misaligned accesses were supported in user mode, but that requirement was removed in 2018 via riscv-isa-manual commit 61cadb9 ("Provide new description of misaligned load/store behavior compatible with privileged architecture."). Since the Linux uABI was already frozen at that point it's just been demoted to part of the uABI, but that was never written down. Link: https://lore.kernel.org/r/20220728210715.17214-1-palmer@rivosinc.com Signed-off-by: Palmer Dabbelt --- Documentation/riscv/index.rst | 1 + Documentation/riscv/uabi.rst | 6 ++++++ 2 files changed, 7 insertions(+) create mode 100644 Documentation/riscv/uabi.rst diff --git a/Documentation/riscv/index.rst b/Documentation/riscv/index.rst index e23b876ad6eb..2e5b18fbb145 100644 --- a/Documentation/riscv/index.rst +++ b/Documentation/riscv/index.rst @@ -8,6 +8,7 @@ RISC-V architecture boot-image-header vm-layout patch-acceptance + uabi features diff --git a/Documentation/riscv/uabi.rst b/Documentation/riscv/uabi.rst new file mode 100644 index 000000000000..21a82cfb6c4d --- /dev/null +++ b/Documentation/riscv/uabi.rst @@ -0,0 +1,6 @@ +.. SPDX-License-Identifier: GPL-2.0 + +RISC-V Linux User ABI +===================== + +Misaligned accesses are supported in userspace, but they may perform poorly. From 0e5d5ae837c8ce04d2ddb874ec5f920118bd9d31 Mon Sep 17 00:00:00 2001 From: D Scott Phillips Date: Mon, 10 Oct 2022 19:21:40 -0700 Subject: [PATCH 064/121] arm64: Add AMPERE1 to the Spectre-BHB affected list Per AmpereOne erratum AC03_CPU_12, "Branch history may allow control of speculative execution across software contexts," the AMPERE1 core needs the bhb clearing loop to mitigate Spectre-BHB, with a loop iteration count of 11. Signed-off-by: D Scott Phillips Link: https://lore.kernel.org/r/20221011022140.432370-1-scott@os.amperecomputing.com Reviewed-by: James Morse Signed-off-by: Catalin Marinas --- arch/arm64/include/asm/cputype.h | 4 ++++ arch/arm64/kernel/proton-pack.c | 6 ++++++ 2 files changed, 10 insertions(+) diff --git a/arch/arm64/include/asm/cputype.h b/arch/arm64/include/asm/cputype.h index 8aa0d276a636..abc418650fec 100644 --- a/arch/arm64/include/asm/cputype.h +++ b/arch/arm64/include/asm/cputype.h @@ -60,6 +60,7 @@ #define ARM_CPU_IMP_FUJITSU 0x46 #define ARM_CPU_IMP_HISI 0x48 #define ARM_CPU_IMP_APPLE 0x61 +#define ARM_CPU_IMP_AMPERE 0xC0 #define ARM_CPU_PART_AEM_V8 0xD0F #define ARM_CPU_PART_FOUNDATION 0xD00 @@ -123,6 +124,8 @@ #define APPLE_CPU_PART_M1_ICESTORM_MAX 0x028 #define APPLE_CPU_PART_M1_FIRESTORM_MAX 0x029 +#define AMPERE_CPU_PART_AMPERE1 0xAC3 + #define MIDR_CORTEX_A53 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A53) #define MIDR_CORTEX_A57 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A57) #define MIDR_CORTEX_A72 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A72) @@ -172,6 +175,7 @@ #define MIDR_APPLE_M1_FIRESTORM_PRO MIDR_CPU_MODEL(ARM_CPU_IMP_APPLE, APPLE_CPU_PART_M1_FIRESTORM_PRO) #define MIDR_APPLE_M1_ICESTORM_MAX MIDR_CPU_MODEL(ARM_CPU_IMP_APPLE, APPLE_CPU_PART_M1_ICESTORM_MAX) #define MIDR_APPLE_M1_FIRESTORM_MAX MIDR_CPU_MODEL(ARM_CPU_IMP_APPLE, APPLE_CPU_PART_M1_FIRESTORM_MAX) +#define MIDR_AMPERE1 MIDR_CPU_MODEL(ARM_CPU_IMP_AMPERE, AMPERE_CPU_PART_AMPERE1) /* Fujitsu Erratum 010001 affects A64FX 1.0 and 1.1, (v0r0 and v1r0) */ #define MIDR_FUJITSU_ERRATUM_010001 MIDR_FUJITSU_A64FX diff --git a/arch/arm64/kernel/proton-pack.c b/arch/arm64/kernel/proton-pack.c index a8ea1637b137..bfce41c2a53b 100644 --- a/arch/arm64/kernel/proton-pack.c +++ b/arch/arm64/kernel/proton-pack.c @@ -868,6 +868,10 @@ u8 spectre_bhb_loop_affected(int scope) MIDR_ALL_VERSIONS(MIDR_NEOVERSE_N1), {}, }; + static const struct midr_range spectre_bhb_k11_list[] = { + MIDR_ALL_VERSIONS(MIDR_AMPERE1), + {}, + }; static const struct midr_range spectre_bhb_k8_list[] = { MIDR_ALL_VERSIONS(MIDR_CORTEX_A72), MIDR_ALL_VERSIONS(MIDR_CORTEX_A57), @@ -878,6 +882,8 @@ u8 spectre_bhb_loop_affected(int scope) k = 32; else if (is_midr_in_range_list(read_cpuid_id(), spectre_bhb_k24_list)) k = 24; + else if (is_midr_in_range_list(read_cpuid_id(), spectre_bhb_k11_list)) + k = 11; else if (is_midr_in_range_list(read_cpuid_id(), spectre_bhb_k8_list)) k = 8; From acfac37851e01b40c30a7afd0d93ad8db8914f25 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Fri, 7 Oct 2022 12:59:20 -0700 Subject: [PATCH 065/121] mm/hugetlb.c: make __hugetlb_vma_unlock_write_put() static Reported-by: kernel test robot Cc: Mike Kravetz Signed-off-by: Andrew Morton --- mm/hugetlb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 0ad53ad98e74..41d3aa077837 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -6804,7 +6804,7 @@ void hugetlb_vma_lock_release(struct kref *kref) kfree(vma_lock); } -void __hugetlb_vma_unlock_write_put(struct hugetlb_vma_lock *vma_lock) +static void __hugetlb_vma_unlock_write_put(struct hugetlb_vma_lock *vma_lock) { struct vm_area_struct *vma = vma_lock->vma; From 7efc3b7261030da79001c00d92bc3392fd6c664c Mon Sep 17 00:00:00 2001 From: Chuyi Zhou Date: Wed, 13 Jul 2022 14:20:09 +0800 Subject: [PATCH 066/121] mm/compaction: fix set skip in fast_find_migrateblock When we successfully find a pageblock in fast_find_migrateblock(), the block will be set skip-flag through set_pageblock_skip(). However, when entering isolate_migratepages_block(), the whole pageblock will be skipped due to the branch 'if (!valid_page && IS_ALIGNED(low_pfn, pageblock_nr_pages))'. Eventually we will goto isolate_abort and isolate nothing. That makes fast_find_migrateblock useless. In this patch, when we find a suitable pageblock in fast_find_migrateblock, we do noting but let isolate_migratepages_block to set skip flag to the pageblock after scan it. Normally, we would isolate some pages from the fast-find block. I use mmtest/thpscale-madvhugepage test it. Here is the result: baseline patch Amean fault-both-1 1331.66 ( 0.00%) 1261.04 * 5.30%* Amean fault-both-3 1383.95 ( 0.00%) 1191.69 * 13.89%* Amean fault-both-5 1568.13 ( 0.00%) 1445.20 * 7.84%* Amean fault-both-7 1819.62 ( 0.00%) 1555.13 * 14.54%* Amean fault-both-12 1106.96 ( 0.00%) 1149.43 * -3.84%* Amean fault-both-18 2196.93 ( 0.00%) 1875.77 * 14.62%* Amean fault-both-24 2642.69 ( 0.00%) 2671.21 * -1.08%* Amean fault-both-30 2901.89 ( 0.00%) 2857.32 * 1.54%* Amean fault-both-32 3747.00 ( 0.00%) 3479.23 * 7.15%* Link: https://lkml.kernel.org/r/20220713062009.597255-1-zhouchuyi@bytedance.com Fixes: 70b44595eafe9 ("mm, compaction: use free lists to quickly locate a migration source") Signed-off-by: zhouchuyi Cc: Mel Gorman Signed-off-by: Andrew Morton --- mm/compaction.c | 1 - 1 file changed, 1 deletion(-) diff --git a/mm/compaction.c b/mm/compaction.c index e2a9615f5fde..c4e4453187a2 100644 --- a/mm/compaction.c +++ b/mm/compaction.c @@ -1851,7 +1851,6 @@ static unsigned long fast_find_migrateblock(struct compact_control *cc) pfn = cc->zone->zone_start_pfn; cc->fast_search_fail = 0; found_block = true; - set_pageblock_skip(freepage); break; } } From 92b7399695a5cc961c44fc6e4624d3bc3c699ee7 Mon Sep 17 00:00:00 2001 From: Liam Howlett Date: Tue, 11 Oct 2022 20:36:51 +0000 Subject: [PATCH 067/121] mmap: fix copy_vma() failure path The anon vma was not unlinked and the file was not closed in the failure path when the machine runs out of memory during the maple tree modification. This caused a memory leak of the anon vma chain and vma since neither would be freed. Link: https://lkml.kernel.org/r/20221011203621.1446507-1-Liam.Howlett@oracle.com Fixes: 524e00b36e8c ("mm: remove rb tree") Signed-off-by: Liam R. Howlett Reported-by: Lukas Bulwahn Tested-by: Lukas Bulwahn Signed-off-by: Andrew Morton --- mm/mmap.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/mm/mmap.c b/mm/mmap.c index 6e447544f07d..fc8581cefef7 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -3240,6 +3240,11 @@ struct vm_area_struct *copy_vma(struct vm_area_struct **vmap, out_vma_link: if (new_vma->vm_ops && new_vma->vm_ops->close) new_vma->vm_ops->close(new_vma); + + if (new_vma->vm_file) + fput(new_vma->vm_file); + + unlink_anon_vmas(new_vma); out_free_mempol: mpol_put(vma_policy(new_vma)); out_free_vma: From 7be1c1a3c7b13fb259bb5159662a7b83622013b8 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Tue, 11 Oct 2022 20:55:31 +0300 Subject: [PATCH 068/121] mm: more vma cache removal Link: https://lkml.kernel.org/r/Y0WuE3Riv4iy5Jx8@localhost.localdomain Fixes: 7964cf8caa4d ("mm: remove vmacache") Signed-off-by: Alexey Dobriyan Acked-by: Liam Howlett Signed-off-by: Andrew Morton --- include/linux/sched.h | 2 -- 1 file changed, 2 deletions(-) diff --git a/include/linux/sched.h b/include/linux/sched.h index 88a043f7235e..e0bb85cf8bdd 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -861,8 +861,6 @@ struct task_struct { struct mm_struct *mm; struct mm_struct *active_mm; - /* Per-thread vma caching: */ - #ifdef SPLIT_RSS_COUNTING struct task_rss_stat rss_stat; #endif From 28c5609fb236807910ca347ad3e26c4567998526 Mon Sep 17 00:00:00 2001 From: Liam Howlett Date: Tue, 11 Oct 2022 16:08:37 +0000 Subject: [PATCH 069/121] mm/mmap: preallocate maple nodes for brk vma expansion If the brk VMA is the last vma in a maple node and meets the rare criteria that it can be expanded, then preallocation is necessary to avoid a potential fs_reclaim circular lock issue on low resources. At the same time use the actual vma start address (unaligned) when calling vma_adjust_trans_huge(). Link: https://lkml.kernel.org/r/20221011160624.1253454-1-Liam.Howlett@oracle.com Fixes: 2e7ce7d354f2 (mm/mmap: change do_brk_flags() to expand existing VMA and add do_brk_munmap()) Signed-off-by: Liam R. Howlett Reported-by: Yu Zhao Signed-off-by: Andrew Morton --- mm/mmap.c | 18 ++++++------------ 1 file changed, 6 insertions(+), 12 deletions(-) diff --git a/mm/mmap.c b/mm/mmap.c index fc8581cefef7..5855f26639f9 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -2942,17 +2942,18 @@ static int do_brk_flags(struct ma_state *mas, struct vm_area_struct *vma, if (vma && (!vma->anon_vma || list_is_singular(&vma->anon_vma_chain)) && ((vma->vm_flags & ~VM_SOFTDIRTY) == flags)) { - mas->index = vma->vm_start; - mas->last = addr + len - 1; - vma_adjust_trans_huge(vma, addr, addr + len, 0); + mas_set_range(mas, vma->vm_start, addr + len - 1); + if (mas_preallocate(mas, vma, GFP_KERNEL)) + return -ENOMEM; + + vma_adjust_trans_huge(vma, vma->vm_start, addr + len, 0); if (vma->anon_vma) { anon_vma_lock_write(vma->anon_vma); anon_vma_interval_tree_pre_update_vma(vma); } vma->vm_end = addr + len; vma->vm_flags |= VM_SOFTDIRTY; - if (mas_store_gfp(mas, vma, GFP_KERNEL)) - goto mas_expand_failed; + mas_store_prealloc(mas, vma); if (vma->anon_vma) { anon_vma_interval_tree_post_update_vma(vma); @@ -2993,13 +2994,6 @@ static int do_brk_flags(struct ma_state *mas, struct vm_area_struct *vma, vma_alloc_fail: vm_unacct_memory(len >> PAGE_SHIFT); return -ENOMEM; - -mas_expand_failed: - if (vma->anon_vma) { - anon_vma_interval_tree_post_update_vma(vma); - anon_vma_unlock_write(vma->anon_vma); - } - return -ENOMEM; } int vm_brk_flags(unsigned long addr, unsigned long request, unsigned long flags) From 515778e2d790652a38a24554fdb7f21420d91efc Mon Sep 17 00:00:00 2001 From: Peter Xu Date: Fri, 30 Sep 2022 20:25:55 -0400 Subject: [PATCH 070/121] mm/uffd: fix warning without PTE_MARKER_UFFD_WP compiled in When PTE_MARKER_UFFD_WP not configured, it's still possible to reach pte marker code and trigger an warning. Add a few CONFIG_PTE_MARKER_UFFD_WP ifdefs to make sure the code won't be reached when not compiled in. Link: https://lkml.kernel.org/r/YzeR+R6b4bwBlBHh@x1n Fixes: b1f9e876862d ("mm/uffd: enable write protection for shmem & hugetlbfs") Signed-off-by: Peter Xu Reported-by: Cc: Axel Rasmussen Cc: Brian Geffon Cc: Edward Liaw Cc: Liu Shixin Cc: Mike Kravetz Cc: Signed-off-by: Andrew Morton --- mm/hugetlb.c | 4 ++++ mm/memory.c | 2 ++ mm/mprotect.c | 2 ++ 3 files changed, 8 insertions(+) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 41d3aa077837..9a910612336d 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -5096,6 +5096,7 @@ static void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct * unmapped and its refcount is dropped, so just clear pte here. */ if (unlikely(!pte_present(pte))) { +#ifdef CONFIG_PTE_MARKER_UFFD_WP /* * If the pte was wr-protected by uffd-wp in any of the * swap forms, meanwhile the caller does not want to @@ -5107,6 +5108,7 @@ static void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct set_huge_pte_at(mm, address, ptep, make_pte_marker(PTE_MARKER_UFFD_WP)); else +#endif huge_pte_clear(mm, address, ptep, sz); spin_unlock(ptl); continue; @@ -5135,11 +5137,13 @@ static void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct tlb_remove_huge_tlb_entry(h, tlb, ptep, address); if (huge_pte_dirty(pte)) set_page_dirty(page); +#ifdef CONFIG_PTE_MARKER_UFFD_WP /* Leave a uffd-wp pte marker if needed */ if (huge_pte_uffd_wp(pte) && !(zap_flags & ZAP_FLAG_DROP_MARKER)) set_huge_pte_at(mm, address, ptep, make_pte_marker(PTE_MARKER_UFFD_WP)); +#endif hugetlb_count_sub(pages_per_huge_page(h), mm); page_remove_rmap(page, vma, true); diff --git a/mm/memory.c b/mm/memory.c index df678fa30cdb..2c7723ea4371 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -1393,10 +1393,12 @@ zap_install_uffd_wp_if_needed(struct vm_area_struct *vma, unsigned long addr, pte_t *pte, struct zap_details *details, pte_t pteval) { +#ifdef CONFIG_PTE_MARKER_UFFD_WP if (zap_drop_file_uffd_wp(details)) return; pte_install_uffd_wp_if_needed(vma, addr, pte, pteval); +#endif } static unsigned long zap_pte_range(struct mmu_gather *tlb, diff --git a/mm/mprotect.c b/mm/mprotect.c index 461dcbd4f21a..668bfaa6ed2a 100644 --- a/mm/mprotect.c +++ b/mm/mprotect.c @@ -267,6 +267,7 @@ static unsigned long change_pte_range(struct mmu_gather *tlb, } else { /* It must be an none page, or what else?.. */ WARN_ON_ONCE(!pte_none(oldpte)); +#ifdef CONFIG_PTE_MARKER_UFFD_WP if (unlikely(uffd_wp && !vma_is_anonymous(vma))) { /* * For file-backed mem, we need to be able to @@ -278,6 +279,7 @@ static unsigned long change_pte_range(struct mmu_gather *tlb, make_pte_marker(PTE_MARKER_UFFD_WP)); pages++; } +#endif } } while (pte++, addr += PAGE_SIZE, addr != end); arch_leave_lazy_mmu_mode(); From 826249942679a110353e71a1d92764fcf43e7cf7 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 23 Aug 2022 19:33:17 +0100 Subject: [PATCH 071/121] dt-bindings: timer: sifive,clint: add legacy riscv compatible While "real" hardware might not use the compatible string "riscv,clint0" it is present in the driver & QEMU uses it for automatically generated virt machine dtbs. To avoid dt-validate problems with QEMU produced dtbs, such as the following, add it to the binding. riscv-virt.dtb: clint@2000000: compatible:0: 'sifive,clint0' is not one of ['sifive,fu540-c000-clint', 'starfive,jh7100-clint', 'canaan,k210-clint'] Reported-by: Rob Herring Link: https://lore.kernel.org/linux-riscv/20220803170552.GA2250266-robh@kernel.org/ Reviewed-by: Rob Herring Signed-off-by: Conor Dooley Reviewed-by: Heiko Stuebner Link: https://lore.kernel.org/r/20220823183319.3314940-2-mail@conchuod.ie Signed-off-by: Palmer Dabbelt --- .../bindings/timer/sifive,clint.yaml | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) diff --git a/Documentation/devicetree/bindings/timer/sifive,clint.yaml b/Documentation/devicetree/bindings/timer/sifive,clint.yaml index e64f46339079..bbad24165837 100644 --- a/Documentation/devicetree/bindings/timer/sifive,clint.yaml +++ b/Documentation/devicetree/bindings/timer/sifive,clint.yaml @@ -22,12 +22,18 @@ description: properties: compatible: - items: - - enum: - - sifive,fu540-c000-clint - - starfive,jh7100-clint - - canaan,k210-clint - - const: sifive,clint0 + oneOf: + - items: + - enum: + - sifive,fu540-c000-clint + - starfive,jh7100-clint + - canaan,k210-clint + - const: sifive,clint0 + - items: + - const: sifive,clint0 + - const: riscv,clint0 + deprecated: true + description: For the QEMU virt machine only description: Should be ",-clint" and "sifive,clint". From 6e965c9bd7388762b302dca5852eb25cbe9cc085 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 23 Aug 2022 19:33:18 +0100 Subject: [PATCH 072/121] dt-bindings: interrupt-controller: sifive,plic: add legacy riscv compatible While "real" hardware might not use the compatible string "riscv,plic0" it is present in the driver & QEMU uses it for automatically generated virt machine dtbs. To avoid dt-validate problems with QEMU produced dtbs, such as the following, add it to the binding. riscv-virt.dtb: plic@c000000: compatible: 'oneOf' conditional failed, one must be fixed: 'sifive,plic-1.0.0' is not one of ['sifive,fu540-c000-plic', 'starfive,jh7100-plic', 'canaan,k210-plic'] 'sifive,plic-1.0.0' is not one of ['allwinner,sun20i-d1-plic'] 'sifive,plic-1.0.0' was expected 'thead,c900-plic' was expected riscv-virt.dtb: plic@c000000: '#address-cells' is a required property Reported-by: Rob Herring Link: https://lore.kernel.org/linux-riscv/20220803170552.GA2250266-robh@kernel.org/ Reviewed-by: Rob Herring Signed-off-by: Conor Dooley Reviewed-by: Heiko Stuebner Link: https://lore.kernel.org/r/20220823183319.3314940-3-mail@conchuod.ie Signed-off-by: Palmer Dabbelt --- .../bindings/interrupt-controller/sifive,plic-1.0.0.yaml | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/Documentation/devicetree/bindings/interrupt-controller/sifive,plic-1.0.0.yaml b/Documentation/devicetree/bindings/interrupt-controller/sifive,plic-1.0.0.yaml index 92e0f8c3eff2..99e01f4d0a69 100644 --- a/Documentation/devicetree/bindings/interrupt-controller/sifive,plic-1.0.0.yaml +++ b/Documentation/devicetree/bindings/interrupt-controller/sifive,plic-1.0.0.yaml @@ -66,6 +66,11 @@ properties: - enum: - allwinner,sun20i-d1-plic - const: thead,c900-plic + - items: + - const: sifive,plic-1.0.0 + - const: riscv,plic0 + deprecated: true + description: For the QEMU virt machine only reg: maxItems: 1 From 299824e68bd0fac60f8352c940fd731fde609de1 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 23 Aug 2022 19:33:19 +0100 Subject: [PATCH 073/121] dt-bindings: riscv: add new riscv,isa strings for emulators The QEMU virt and spike machines currently export a riscv,isa string of "rv64imafdcsuh", While the RISC-V foundation has been ratifying a bunch of extenstions etc, the kernel has remained relatively static with what hardware is supported - but the same is not true of QEMU. Using the virt machine and running dt-validate on the dumped dtb fails, partly due to the unexpected isa string. Rather than enumerate the many many possbilities, change the pattern to a regex, with the following assumptions: - ima are required - the single letter order is fixed & we don't care about things that can't even do "ima" - the standard multi letter extensions are all in a "_z" format where the first letter of is a valid single letter extension - _s & _h are used for supervisor and hyper visor extensions - convention says that after the first two chars, a standard multi letter extension name could be an english word (ifencei anyone?) so it's not worth restricting the charset - as the above is just convention, don't apply any charset restrictions to reduce future churn - vendor ISA extensions begind with _x and have no charset restrictions - we don't care about an e extension from an OS pov - that attempting to validate the contents of the multiletter extensions with dt-validate beyond the formatting is a futile, massively verbose or unwieldy exercise at best The following limitations also apply: - multi letter extension ordering is not enforced. dt-schema does not appear to allow for named match groups, so the resulting regex would be even more of a headache - ditto for the numbered extensions Finally, add me as a maintainer of the binding so that when it breaks in the future, I can be held responsible! Reported-by: Rob Herring Link: https://lore.kernel.org/linux-riscv/20220803170552.GA2250266-robh@kernel.org/ Reviewed-by: Andrew Jones Acked-by: Guo Ren Signed-off-by: Conor Dooley Acked-by: Heiko Stuebner Reviewed-by: Rob Herring Link: https://lore.kernel.org/r/20220823183319.3314940-4-mail@conchuod.ie Signed-off-by: Palmer Dabbelt --- Documentation/devicetree/bindings/riscv/cpus.yaml | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/Documentation/devicetree/bindings/riscv/cpus.yaml b/Documentation/devicetree/bindings/riscv/cpus.yaml index 873dd12f6e89..90a7cabf58fe 100644 --- a/Documentation/devicetree/bindings/riscv/cpus.yaml +++ b/Documentation/devicetree/bindings/riscv/cpus.yaml @@ -9,6 +9,7 @@ title: RISC-V bindings for 'cpus' DT nodes maintainers: - Paul Walmsley - Palmer Dabbelt + - Conor Dooley description: | This document uses some terminology common to the RISC-V community @@ -79,9 +80,7 @@ properties: insensitive, letters in the riscv,isa string must be all lowercase to simplify parsing. $ref: "/schemas/types.yaml#/definitions/string" - enum: - - rv64imac - - rv64imafdc + pattern: ^rv(?:64|32)imaf?d?q?c?b?v?k?h?(?:_[hsxz](?:[a-z])+)*$ # RISC-V requires 'timebase-frequency' in /cpus, so disallow it here timebase-frequency: false From deb0f6562884b5b4beb883d73e66a7d3a1b96d99 Mon Sep 17 00:00:00 2001 From: Carlos Llamas Date: Fri, 30 Sep 2022 00:38:43 +0000 Subject: [PATCH 074/121] mm/mmap: undo ->mmap() when arch_validate_flags() fails Commit c462ac288f2c ("mm: Introduce arch_validate_flags()") added a late check in mmap_region() to let architectures validate vm_flags. The check needs to happen after calling ->mmap() as the flags can potentially be modified during this callback. If arch_validate_flags() check fails we unmap and free the vma. However, the error path fails to undo the ->mmap() call that previously succeeded and depending on the specific ->mmap() implementation this translates to reference increments, memory allocations and other operations what will not be cleaned up. There are several places (mainly device drivers) where this is an issue. However, one specific example is bpf_map_mmap() which keeps count of the mappings in map->writecnt. The count is incremented on ->mmap() and then decremented on vm_ops->close(). When arch_validate_flags() fails this count is off since bpf_map_mmap_close() is never called. One can reproduce this issue in arm64 devices with MTE support. Here the vm_flags are checked to only allow VM_MTE if VM_MTE_ALLOWED has been set previously. From userspace then is enough to pass the PROT_MTE flag to mmap() syscall to trigger the arch_validate_flags() failure. The following program reproduces this issue: #include #include #include #include #include int main(void) { union bpf_attr attr = { .map_type = BPF_MAP_TYPE_ARRAY, .key_size = sizeof(int), .value_size = sizeof(long long), .max_entries = 256, .map_flags = BPF_F_MMAPABLE, }; int fd; fd = syscall(__NR_bpf, BPF_MAP_CREATE, &attr, sizeof(attr)); mmap(NULL, 4096, PROT_WRITE | PROT_MTE, MAP_SHARED, fd, 0); return 0; } By manually adding some log statements to the vm_ops callbacks we can confirm that when passing PROT_MTE to mmap() the map->writecnt is off upon ->release(): With PROT_MTE flag: root@debian:~# ./bpf-test [ 111.263874] bpf_map_write_active_inc: map=9 writecnt=1 [ 111.288763] bpf_map_release: map=9 writecnt=1 Without PROT_MTE flag: root@debian:~# ./bpf-test [ 157.816912] bpf_map_write_active_inc: map=10 writecnt=1 [ 157.830442] bpf_map_write_active_dec: map=10 writecnt=0 [ 157.832396] bpf_map_release: map=10 writecnt=0 This patch fixes the above issue by calling vm_ops->close() when the arch_validate_flags() check fails, after this we can proceed to unmap and free the vma on the error path. Link: https://lkml.kernel.org/r/20220930003844.1210987-1-cmllamas@google.com Fixes: c462ac288f2c ("mm: Introduce arch_validate_flags()") Signed-off-by: Carlos Llamas Reviewed-by: Catalin Marinas Acked-by: Andrii Nakryiko Reviewed-by: Liam Howlett Cc: Christian Brauner (Microsoft) Cc: Michal Hocko Cc: Suren Baghdasaryan Cc: [5.10+] Signed-off-by: Andrew Morton --- mm/mmap.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/mm/mmap.c b/mm/mmap.c index 5855f26639f9..bf2122af94e7 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -2673,7 +2673,7 @@ unsigned long mmap_region(struct file *file, unsigned long addr, if (!arch_validate_flags(vma->vm_flags)) { error = -EINVAL; if (file) - goto unmap_and_free_vma; + goto close_and_free_vma; else goto free_vma; } @@ -2742,6 +2742,9 @@ unsigned long mmap_region(struct file *file, unsigned long addr, validate_mm(mm); return addr; +close_and_free_vma: + if (vma->vm_ops && vma->vm_ops->close) + vma->vm_ops->close(vma); unmap_and_free_vma: fput(vma->vm_file); vma->vm_file = NULL; From 4fa0e3ff217f775cb58d2d6d51820ec519243fb9 Mon Sep 17 00:00:00 2001 From: "Matthew Wilcox (Oracle)" Date: Wed, 12 Oct 2022 20:34:19 +0100 Subject: [PATCH 075/121] ext4,f2fs: fix readahead of verity data The recent change of page_cache_ra_unbounded() arguments was buggy in the two callers, causing us to readahead the wrong pages. Move the definition of ractl down to after the index is set correctly. This affected performance on configurations that use fs-verity. Link: https://lkml.kernel.org/r/20221012193419.1453558-1-willy@infradead.org Fixes: 73bb49da50cd ("mm/readahead: make page_cache_ra_unbounded take a readahead_control") Signed-off-by: Matthew Wilcox (Oracle) Reported-by: Jintao Yin Signed-off-by: Andrew Morton --- fs/ext4/verity.c | 3 ++- fs/f2fs/verity.c | 3 ++- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/fs/ext4/verity.c b/fs/ext4/verity.c index b051d19b5c8a..94442c690ca7 100644 --- a/fs/ext4/verity.c +++ b/fs/ext4/verity.c @@ -365,13 +365,14 @@ static struct page *ext4_read_merkle_tree_page(struct inode *inode, pgoff_t index, unsigned long num_ra_pages) { - DEFINE_READAHEAD(ractl, NULL, NULL, inode->i_mapping, index); struct page *page; index += ext4_verity_metadata_pos(inode) >> PAGE_SHIFT; page = find_get_page_flags(inode->i_mapping, index, FGP_ACCESSED); if (!page || !PageUptodate(page)) { + DEFINE_READAHEAD(ractl, NULL, NULL, inode->i_mapping, index); + if (page) put_page(page); else if (num_ra_pages > 1) diff --git a/fs/f2fs/verity.c b/fs/f2fs/verity.c index 7b8f2b41c29b..c0733f867074 100644 --- a/fs/f2fs/verity.c +++ b/fs/f2fs/verity.c @@ -262,13 +262,14 @@ static struct page *f2fs_read_merkle_tree_page(struct inode *inode, pgoff_t index, unsigned long num_ra_pages) { - DEFINE_READAHEAD(ractl, NULL, NULL, inode->i_mapping, index); struct page *page; index += f2fs_verity_metadata_pos(inode) >> PAGE_SHIFT; page = find_get_page_flags(inode->i_mapping, index, FGP_ACCESSED); if (!page || !PageUptodate(page)) { + DEFINE_READAHEAD(ractl, NULL, NULL, inode->i_mapping, index); + if (page) put_page(page); else if (num_ra_pages > 1) From ac801e7e252c5588325e3c983c7d4167fc68c024 Mon Sep 17 00:00:00 2001 From: Alexander Potapenko Date: Mon, 5 Sep 2022 14:24:27 +0200 Subject: [PATCH 076/121] kmsan: unpoison @tlb in arch_tlb_gather_mmu() This is an optimization to reduce stackdepot pressure. struct mmu_gather contains 7 1-bit fields packed into a 32-bit unsigned int value. The remaining 25 bits remain uninitialized and are never used, but KMSAN updates the origin for them in zap_pXX_range() in mm/memory.c, thus creating very long origin chains. This is technically correct, but consumes too much memory. Unpoisoning the whole structure will prevent creating such chains. Link: https://lkml.kernel.org/r/20220905122452.2258262-20-glider@google.com Signed-off-by: Alexander Potapenko Acked-by: Marco Elver Cc: Alexander Viro Cc: Alexei Starovoitov Cc: Andrey Konovalov Cc: Andrey Konovalov Cc: Andy Lutomirski Cc: Arnd Bergmann Cc: Borislav Petkov Cc: Christoph Hellwig Cc: Christoph Lameter Cc: David Rientjes Cc: Dmitry Vyukov Cc: Eric Biggers Cc: Eric Dumazet Cc: Greg Kroah-Hartman Cc: Herbert Xu Cc: Ilya Leoshkevich Cc: Ingo Molnar Cc: Jens Axboe Cc: Joonsoo Kim Cc: Kees Cook Cc: Liu Shixin Cc: Mark Rutland Cc: Matthew Wilcox Cc: Michael S. Tsirkin Cc: Pekka Enberg Cc: Peter Zijlstra Cc: Petr Mladek Cc: Steven Rostedt Cc: Thomas Gleixner Cc: Vasily Gorbik Cc: Vegard Nossum Cc: Vlastimil Babka Signed-off-by: Andrew Morton --- mm/mmu_gather.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/mm/mmu_gather.c b/mm/mmu_gather.c index a71924bd38c0..add4244e5790 100644 --- a/mm/mmu_gather.c +++ b/mm/mmu_gather.c @@ -1,6 +1,7 @@ #include #include #include +#include #include #include #include @@ -265,6 +266,15 @@ void tlb_flush_mmu(struct mmu_gather *tlb) static void __tlb_gather_mmu(struct mmu_gather *tlb, struct mm_struct *mm, bool fullmm) { + /* + * struct mmu_gather contains 7 1-bit fields packed into a 32-bit + * unsigned int value. The remaining 25 bits remain uninitialized + * and are never used, but KMSAN updates the origin for them in + * zap_pXX_range() in mm/memory.c, thus creating very long origin + * chains. This is technically correct, but consumes too much memory. + * Unpoisoning the whole structure will prevent creating such chains. + */ + kmsan_unpoison_memory(tlb, sizeof(*tlb)); tlb->mm = mm; tlb->fullmm = fullmm; From ea091fa53680030881b56520d731e36d3ff6cdd5 Mon Sep 17 00:00:00 2001 From: Xiaoke Wang Date: Fri, 4 Mar 2022 17:12:15 +0800 Subject: [PATCH 077/121] lib/test_meminit: add checks for the allocation functions alloc_pages(), kmalloc() and vmalloc() are all memory allocation functions which can return NULL when some internal memory failures happen. So it is better to check the return of them to catch the failure in time for better test them. Link: https://lkml.kernel.org/r/tencent_D44A49FFB420EDCCBFB9221C8D14DFE12908@qq.com Signed-off-by: Xiaoke Wang Reviewed-by: Alexander Potapenko Cc: Andrey Konovalov Cc: Andrey Ryabinin Cc: Dmitry Vyukov Cc: Marco Elver Signed-off-by: Andrew Morton --- lib/test_meminit.c | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/lib/test_meminit.c b/lib/test_meminit.c index c95db11a6906..60e1984c060f 100644 --- a/lib/test_meminit.c +++ b/lib/test_meminit.c @@ -67,17 +67,24 @@ static int __init do_alloc_pages_order(int order, int *total_failures) size_t size = PAGE_SIZE << order; page = alloc_pages(GFP_KERNEL, order); + if (!page) + goto err; buf = page_address(page); fill_with_garbage(buf, size); __free_pages(page, order); page = alloc_pages(GFP_KERNEL, order); + if (!page) + goto err; buf = page_address(page); if (count_nonzero_bytes(buf, size)) (*total_failures)++; fill_with_garbage(buf, size); __free_pages(page, order); return 1; +err: + (*total_failures)++; + return 1; } /* Test the page allocator by calling alloc_pages with different orders. */ @@ -100,15 +107,22 @@ static int __init do_kmalloc_size(size_t size, int *total_failures) void *buf; buf = kmalloc(size, GFP_KERNEL); + if (!buf) + goto err; fill_with_garbage(buf, size); kfree(buf); buf = kmalloc(size, GFP_KERNEL); + if (!buf) + goto err; if (count_nonzero_bytes(buf, size)) (*total_failures)++; fill_with_garbage(buf, size); kfree(buf); return 1; +err: + (*total_failures)++; + return 1; } /* Test vmalloc() with given parameters. */ @@ -117,15 +131,22 @@ static int __init do_vmalloc_size(size_t size, int *total_failures) void *buf; buf = vmalloc(size); + if (!buf) + goto err; fill_with_garbage(buf, size); vfree(buf); buf = vmalloc(size); + if (!buf) + goto err; if (count_nonzero_bytes(buf, size)) (*total_failures)++; fill_with_garbage(buf, size); vfree(buf); return 1; +err: + (*total_failures)++; + return 1; } /* Test kmalloc()/vmalloc() by allocating objects of different sizes. */ From 652e04464d3944226052c827bdaaf5113b072870 Mon Sep 17 00:00:00 2001 From: Xin Hao Date: Tue, 27 Sep 2022 08:19:45 +0800 Subject: [PATCH 078/121] mm/damon: move sz_damon_region to damon_sz_region Rename sz_damon_region() to damon_sz_region(), and move it to "include/linux/damon.h", because in many places, we can to use this func. Link: https://lkml.kernel.org/r/20220927001946.85375-1-xhao@linux.alibaba.com Signed-off-by: Xin Hao Suggested-by: SeongJae Park Reviewed-by: SeongJae Park Signed-off-by: Andrew Morton --- include/linux/damon.h | 6 ++++++ mm/damon/core.c | 9 ++------- 2 files changed, 8 insertions(+), 7 deletions(-) diff --git a/include/linux/damon.h b/include/linux/damon.h index ed5470f50bab..620ada094c3b 100644 --- a/include/linux/damon.h +++ b/include/linux/damon.h @@ -484,6 +484,12 @@ static inline struct damon_region *damon_first_region(struct damon_target *t) return list_first_entry(&t->regions_list, struct damon_region, list); } +static inline unsigned long damon_sz_region(struct damon_region *r) +{ + return r->ar.end - r->ar.start; +} + + #define damon_for_each_region(r, t) \ list_for_each_entry(r, &t->regions_list, list) diff --git a/mm/damon/core.c b/mm/damon/core.c index 4de8c7c52979..5b9e0d585aef 100644 --- a/mm/damon/core.c +++ b/mm/damon/core.c @@ -864,18 +864,13 @@ static void kdamond_apply_schemes(struct damon_ctx *c) } } -static inline unsigned long sz_damon_region(struct damon_region *r) -{ - return r->ar.end - r->ar.start; -} - /* * Merge two adjacent regions into one region */ static void damon_merge_two_regions(struct damon_target *t, struct damon_region *l, struct damon_region *r) { - unsigned long sz_l = sz_damon_region(l), sz_r = sz_damon_region(r); + unsigned long sz_l = damon_sz_region(l), sz_r = damon_sz_region(r); l->nr_accesses = (l->nr_accesses * sz_l + r->nr_accesses * sz_r) / (sz_l + sz_r); @@ -904,7 +899,7 @@ static void damon_merge_regions_of(struct damon_target *t, unsigned int thres, if (prev && prev->ar.end == r->ar.start && abs(prev->nr_accesses - r->nr_accesses) <= thres && - sz_damon_region(prev) + sz_damon_region(r) <= sz_limit) + damon_sz_region(prev) + damon_sz_region(r) <= sz_limit) damon_merge_two_regions(t, prev, r); else prev = r; From ab63f63f3885d492e62da55304b0483a2a9e6a7d Mon Sep 17 00:00:00 2001 From: Xin Hao Date: Tue, 27 Sep 2022 08:19:46 +0800 Subject: [PATCH 079/121] mm/damon: use damon_sz_region() in appropriate place In many places we can use damon_sz_region() to instead of "r->ar.end - r->ar.start". Link: https://lkml.kernel.org/r/20220927001946.85375-2-xhao@linux.alibaba.com Signed-off-by: Xin Hao Suggested-by: SeongJae Park Reviewed-by: SeongJae Park Signed-off-by: Andrew Morton --- mm/damon/core.c | 17 ++++++++--------- mm/damon/vaddr.c | 4 ++-- 2 files changed, 10 insertions(+), 11 deletions(-) diff --git a/mm/damon/core.c b/mm/damon/core.c index 5b9e0d585aef..515ac4e52a11 100644 --- a/mm/damon/core.c +++ b/mm/damon/core.c @@ -490,7 +490,7 @@ static unsigned long damon_region_sz_limit(struct damon_ctx *ctx) damon_for_each_target(t, ctx) { damon_for_each_region(r, t) - sz += r->ar.end - r->ar.start; + sz += damon_sz_region(r); } if (ctx->attrs.min_nr_regions) @@ -673,7 +673,7 @@ static bool __damos_valid_target(struct damon_region *r, struct damos *s) { unsigned long sz; - sz = r->ar.end - r->ar.start; + sz = damon_sz_region(r); return s->pattern.min_sz_region <= sz && sz <= s->pattern.max_sz_region && s->pattern.min_nr_accesses <= r->nr_accesses && @@ -701,7 +701,7 @@ static void damon_do_apply_schemes(struct damon_ctx *c, damon_for_each_scheme(s, c) { struct damos_quota *quota = &s->quota; - unsigned long sz = r->ar.end - r->ar.start; + unsigned long sz = damon_sz_region(r); struct timespec64 begin, end; unsigned long sz_applied = 0; @@ -730,14 +730,14 @@ static void damon_do_apply_schemes(struct damon_ctx *c, sz = ALIGN_DOWN(quota->charge_addr_from - r->ar.start, DAMON_MIN_REGION); if (!sz) { - if (r->ar.end - r->ar.start <= - DAMON_MIN_REGION) + if (damon_sz_region(r) <= + DAMON_MIN_REGION) continue; sz = DAMON_MIN_REGION; } damon_split_region_at(t, r, sz); r = damon_next_region(r); - sz = r->ar.end - r->ar.start; + sz = damon_sz_region(r); } quota->charge_target_from = NULL; quota->charge_addr_from = 0; @@ -842,8 +842,7 @@ static void kdamond_apply_schemes(struct damon_ctx *c) continue; score = c->ops.get_scheme_score( c, t, r, s); - quota->histogram[score] += - r->ar.end - r->ar.start; + quota->histogram[score] += damon_sz_region(r); if (score > max_score) max_score = score; } @@ -957,7 +956,7 @@ static void damon_split_regions_of(struct damon_target *t, int nr_subs) int i; damon_for_each_region_safe(r, next, t) { - sz_region = r->ar.end - r->ar.start; + sz_region = damon_sz_region(r); for (i = 0; i < nr_subs - 1 && sz_region > 2 * DAMON_MIN_REGION; i++) { diff --git a/mm/damon/vaddr.c b/mm/damon/vaddr.c index ea94e0b2c311..15f03df66db6 100644 --- a/mm/damon/vaddr.c +++ b/mm/damon/vaddr.c @@ -72,7 +72,7 @@ static int damon_va_evenly_split_region(struct damon_target *t, return -EINVAL; orig_end = r->ar.end; - sz_orig = r->ar.end - r->ar.start; + sz_orig = damon_sz_region(r); sz_piece = ALIGN_DOWN(sz_orig / nr_pieces, DAMON_MIN_REGION); if (!sz_piece) @@ -618,7 +618,7 @@ static unsigned long damos_madvise(struct damon_target *target, { struct mm_struct *mm; unsigned long start = PAGE_ALIGN(r->ar.start); - unsigned long len = PAGE_ALIGN(r->ar.end - r->ar.start); + unsigned long len = PAGE_ALIGN(damon_sz_region(r)); unsigned long applied; mm = damon_get_mm(target); From 16ce101db85db694a91380aa4c89b25530871d33 Mon Sep 17 00:00:00 2001 From: Alistair Popple Date: Wed, 28 Sep 2022 22:01:15 +1000 Subject: [PATCH 080/121] mm/memory.c: fix race when faulting a device private page MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Patch series "Fix several device private page reference counting issues", v2 This series aims to fix a number of page reference counting issues in drivers dealing with device private ZONE_DEVICE pages. These result in use-after-free type bugs, either from accessing a struct page which no longer exists because it has been removed or accessing fields within the struct page which are no longer valid because the page has been freed. During normal usage it is unlikely these will cause any problems. However without these fixes it is possible to crash the kernel from userspace. These crashes can be triggered either by unloading the kernel module or unbinding the device from the driver prior to a userspace task exiting. In modules such as Nouveau it is also possible to trigger some of these issues by explicitly closing the device file-descriptor prior to the task exiting and then accessing device private memory. This involves some minor changes to both PowerPC and AMD GPU code. Unfortunately I lack hardware to test either of those so any help there would be appreciated. The changes mimic what is done in for both Nouveau and hmm-tests though so I doubt they will cause problems. This patch (of 8): When the CPU tries to access a device private page the migrate_to_ram() callback associated with the pgmap for the page is called. However no reference is taken on the faulting page. Therefore a concurrent migration of the device private page can free the page and possibly the underlying pgmap. This results in a race which can crash the kernel due to the migrate_to_ram() function pointer becoming invalid. It also means drivers can't reliably read the zone_device_data field because the page may have been freed with memunmap_pages(). Close the race by getting a reference on the page while holding the ptl to ensure it has not been freed. Unfortunately the elevated reference count will cause the migration required to handle the fault to fail. To avoid this failure pass the faulting page into the migrate_vma functions so that if an elevated reference count is found it can be checked to see if it's expected or not. [mpe@ellerman.id.au: fix build] Link: https://lkml.kernel.org/r/87fsgbf3gh.fsf@mpe.ellerman.id.au Link: https://lkml.kernel.org/r/cover.60659b549d8509ddecafad4f498ee7f03bb23c69.1664366292.git-series.apopple@nvidia.com Link: https://lkml.kernel.org/r/d3e813178a59e565e8d78d9b9a4e2562f6494f90.1664366292.git-series.apopple@nvidia.com Signed-off-by: Alistair Popple Acked-by: Felix Kuehling Cc: Jason Gunthorpe Cc: John Hubbard Cc: Ralph Campbell Cc: Michael Ellerman Cc: Lyude Paul Cc: Alex Deucher Cc: Alex Sierra Cc: Ben Skeggs Cc: Christian König Cc: Dan Williams Cc: David Hildenbrand Cc: "Huang, Ying" Cc: Matthew Wilcox Cc: Yang Shi Cc: Zi Yan Signed-off-by: Andrew Morton --- arch/powerpc/kvm/book3s_hv_uvmem.c | 19 +++++++------ drivers/gpu/drm/amd/amdkfd/kfd_migrate.c | 17 +++++++----- drivers/gpu/drm/amd/amdkfd/kfd_migrate.h | 2 +- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 11 +++++--- include/linux/migrate.h | 8 ++++++ lib/test_hmm.c | 7 ++--- mm/memory.c | 16 ++++++++++- mm/migrate.c | 34 ++++++++++++++---------- mm/migrate_device.c | 18 +++++++++---- 9 files changed, 89 insertions(+), 43 deletions(-) diff --git a/arch/powerpc/kvm/book3s_hv_uvmem.c b/arch/powerpc/kvm/book3s_hv_uvmem.c index 598006301620..965c9e9e500b 100644 --- a/arch/powerpc/kvm/book3s_hv_uvmem.c +++ b/arch/powerpc/kvm/book3s_hv_uvmem.c @@ -508,10 +508,10 @@ unsigned long kvmppc_h_svm_init_start(struct kvm *kvm) static int __kvmppc_svm_page_out(struct vm_area_struct *vma, unsigned long start, unsigned long end, unsigned long page_shift, - struct kvm *kvm, unsigned long gpa) + struct kvm *kvm, unsigned long gpa, struct page *fault_page) { unsigned long src_pfn, dst_pfn = 0; - struct migrate_vma mig; + struct migrate_vma mig = { 0 }; struct page *dpage, *spage; struct kvmppc_uvmem_page_pvt *pvt; unsigned long pfn; @@ -525,6 +525,7 @@ static int __kvmppc_svm_page_out(struct vm_area_struct *vma, mig.dst = &dst_pfn; mig.pgmap_owner = &kvmppc_uvmem_pgmap; mig.flags = MIGRATE_VMA_SELECT_DEVICE_PRIVATE; + mig.fault_page = fault_page; /* The requested page is already paged-out, nothing to do */ if (!kvmppc_gfn_is_uvmem_pfn(gpa >> page_shift, kvm, NULL)) @@ -580,12 +581,14 @@ static int __kvmppc_svm_page_out(struct vm_area_struct *vma, static inline int kvmppc_svm_page_out(struct vm_area_struct *vma, unsigned long start, unsigned long end, unsigned long page_shift, - struct kvm *kvm, unsigned long gpa) + struct kvm *kvm, unsigned long gpa, + struct page *fault_page) { int ret; mutex_lock(&kvm->arch.uvmem_lock); - ret = __kvmppc_svm_page_out(vma, start, end, page_shift, kvm, gpa); + ret = __kvmppc_svm_page_out(vma, start, end, page_shift, kvm, gpa, + fault_page); mutex_unlock(&kvm->arch.uvmem_lock); return ret; @@ -634,7 +637,7 @@ void kvmppc_uvmem_drop_pages(const struct kvm_memory_slot *slot, pvt->remove_gfn = true; if (__kvmppc_svm_page_out(vma, addr, addr + PAGE_SIZE, - PAGE_SHIFT, kvm, pvt->gpa)) + PAGE_SHIFT, kvm, pvt->gpa, NULL)) pr_err("Can't page out gpa:0x%lx addr:0x%lx\n", pvt->gpa, addr); } else { @@ -736,7 +739,7 @@ static int kvmppc_svm_page_in(struct vm_area_struct *vma, bool pagein) { unsigned long src_pfn, dst_pfn = 0; - struct migrate_vma mig; + struct migrate_vma mig = { 0 }; struct page *spage; unsigned long pfn; struct page *dpage; @@ -994,7 +997,7 @@ static vm_fault_t kvmppc_uvmem_migrate_to_ram(struct vm_fault *vmf) if (kvmppc_svm_page_out(vmf->vma, vmf->address, vmf->address + PAGE_SIZE, PAGE_SHIFT, - pvt->kvm, pvt->gpa)) + pvt->kvm, pvt->gpa, vmf->page)) return VM_FAULT_SIGBUS; else return 0; @@ -1065,7 +1068,7 @@ kvmppc_h_svm_page_out(struct kvm *kvm, unsigned long gpa, if (!vma || vma->vm_start > start || vma->vm_end < end) goto out; - if (!kvmppc_svm_page_out(vma, start, end, page_shift, kvm, gpa)) + if (!kvmppc_svm_page_out(vma, start, end, page_shift, kvm, gpa, NULL)) ret = H_SUCCESS; out: mmap_read_unlock(kvm->mm); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c index b059a77b6081..776448bd9fe4 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c @@ -409,7 +409,7 @@ svm_migrate_vma_to_vram(struct amdgpu_device *adev, struct svm_range *prange, uint64_t npages = (end - start) >> PAGE_SHIFT; struct kfd_process_device *pdd; struct dma_fence *mfence = NULL; - struct migrate_vma migrate; + struct migrate_vma migrate = { 0 }; unsigned long cpages = 0; dma_addr_t *scratch; void *buf; @@ -668,7 +668,7 @@ svm_migrate_copy_to_ram(struct amdgpu_device *adev, struct svm_range *prange, static long svm_migrate_vma_to_ram(struct amdgpu_device *adev, struct svm_range *prange, struct vm_area_struct *vma, uint64_t start, uint64_t end, - uint32_t trigger) + uint32_t trigger, struct page *fault_page) { struct kfd_process *p = container_of(prange->svms, struct kfd_process, svms); uint64_t npages = (end - start) >> PAGE_SHIFT; @@ -676,7 +676,7 @@ svm_migrate_vma_to_ram(struct amdgpu_device *adev, struct svm_range *prange, unsigned long cpages = 0; struct kfd_process_device *pdd; struct dma_fence *mfence = NULL; - struct migrate_vma migrate; + struct migrate_vma migrate = { 0 }; dma_addr_t *scratch; void *buf; int r = -ENOMEM; @@ -699,6 +699,7 @@ svm_migrate_vma_to_ram(struct amdgpu_device *adev, struct svm_range *prange, migrate.src = buf; migrate.dst = migrate.src + npages; + migrate.fault_page = fault_page; scratch = (dma_addr_t *)(migrate.dst + npages); kfd_smi_event_migration_start(adev->kfd.dev, p->lead_thread->pid, @@ -766,7 +767,7 @@ svm_migrate_vma_to_ram(struct amdgpu_device *adev, struct svm_range *prange, * 0 - OK, otherwise error code */ int svm_migrate_vram_to_ram(struct svm_range *prange, struct mm_struct *mm, - uint32_t trigger) + uint32_t trigger, struct page *fault_page) { struct amdgpu_device *adev; struct vm_area_struct *vma; @@ -807,7 +808,8 @@ int svm_migrate_vram_to_ram(struct svm_range *prange, struct mm_struct *mm, } next = min(vma->vm_end, end); - r = svm_migrate_vma_to_ram(adev, prange, vma, addr, next, trigger); + r = svm_migrate_vma_to_ram(adev, prange, vma, addr, next, trigger, + fault_page); if (r < 0) { pr_debug("failed %ld to migrate prange %p\n", r, prange); break; @@ -851,7 +853,7 @@ svm_migrate_vram_to_vram(struct svm_range *prange, uint32_t best_loc, pr_debug("from gpu 0x%x to gpu 0x%x\n", prange->actual_loc, best_loc); do { - r = svm_migrate_vram_to_ram(prange, mm, trigger); + r = svm_migrate_vram_to_ram(prange, mm, trigger, NULL); if (r) return r; } while (prange->actual_loc && --retries); @@ -938,7 +940,8 @@ static vm_fault_t svm_migrate_to_ram(struct vm_fault *vmf) goto out_unlock_prange; } - r = svm_migrate_vram_to_ram(prange, mm, KFD_MIGRATE_TRIGGER_PAGEFAULT_CPU); + r = svm_migrate_vram_to_ram(prange, mm, KFD_MIGRATE_TRIGGER_PAGEFAULT_CPU, + vmf->page); if (r) pr_debug("failed %d migrate 0x%p [0x%lx 0x%lx] to ram\n", r, prange, prange->start, prange->last); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.h b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.h index b3f0754b32fa..a5d7e6d22264 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.h @@ -43,7 +43,7 @@ enum MIGRATION_COPY_DIR { int svm_migrate_to_vram(struct svm_range *prange, uint32_t best_loc, struct mm_struct *mm, uint32_t trigger); int svm_migrate_vram_to_ram(struct svm_range *prange, struct mm_struct *mm, - uint32_t trigger); + uint32_t trigger, struct page *fault_page); unsigned long svm_migrate_addr_to_pfn(struct amdgpu_device *adev, unsigned long addr); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index 11074cc8c333..9139e5a0b2a0 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -2913,13 +2913,15 @@ svm_range_restore_pages(struct amdgpu_device *adev, unsigned int pasid, */ if (prange->actual_loc) r = svm_migrate_vram_to_ram(prange, mm, - KFD_MIGRATE_TRIGGER_PAGEFAULT_GPU); + KFD_MIGRATE_TRIGGER_PAGEFAULT_GPU, + NULL); else r = 0; } } else { r = svm_migrate_vram_to_ram(prange, mm, - KFD_MIGRATE_TRIGGER_PAGEFAULT_GPU); + KFD_MIGRATE_TRIGGER_PAGEFAULT_GPU, + NULL); } if (r) { pr_debug("failed %d to migrate svms %p [0x%lx 0x%lx]\n", @@ -3242,7 +3244,8 @@ svm_range_trigger_migration(struct mm_struct *mm, struct svm_range *prange, return 0; if (!best_loc) { - r = svm_migrate_vram_to_ram(prange, mm, KFD_MIGRATE_TRIGGER_PREFETCH); + r = svm_migrate_vram_to_ram(prange, mm, + KFD_MIGRATE_TRIGGER_PREFETCH, NULL); *migrated = !r; return r; } @@ -3303,7 +3306,7 @@ static void svm_range_evict_svm_bo_worker(struct work_struct *work) mutex_lock(&prange->migrate_mutex); do { r = svm_migrate_vram_to_ram(prange, mm, - KFD_MIGRATE_TRIGGER_TTM_EVICTION); + KFD_MIGRATE_TRIGGER_TTM_EVICTION, NULL); } while (!r && prange->actual_loc && --retries); if (!r && prange->actual_loc) diff --git a/include/linux/migrate.h b/include/linux/migrate.h index 704a04f5a074..52090d1f9230 100644 --- a/include/linux/migrate.h +++ b/include/linux/migrate.h @@ -62,6 +62,8 @@ extern const char *migrate_reason_names[MR_TYPES]; #ifdef CONFIG_MIGRATION extern void putback_movable_pages(struct list_head *l); +int migrate_folio_extra(struct address_space *mapping, struct folio *dst, + struct folio *src, enum migrate_mode mode, int extra_count); int migrate_folio(struct address_space *mapping, struct folio *dst, struct folio *src, enum migrate_mode mode); extern int migrate_pages(struct list_head *l, new_page_t new, free_page_t free, @@ -197,6 +199,12 @@ struct migrate_vma { */ void *pgmap_owner; unsigned long flags; + + /* + * Set to vmf->page if this is being called to migrate a page as part of + * a migrate_to_ram() callback. + */ + struct page *fault_page; }; int migrate_vma_setup(struct migrate_vma *args); diff --git a/lib/test_hmm.c b/lib/test_hmm.c index 6a33f6b1b465..e566166b5571 100644 --- a/lib/test_hmm.c +++ b/lib/test_hmm.c @@ -907,7 +907,7 @@ static int dmirror_migrate_to_system(struct dmirror *dmirror, struct vm_area_struct *vma; unsigned long src_pfns[64] = { 0 }; unsigned long dst_pfns[64] = { 0 }; - struct migrate_vma args; + struct migrate_vma args = { 0 }; unsigned long next; int ret; @@ -968,7 +968,7 @@ static int dmirror_migrate_to_device(struct dmirror *dmirror, unsigned long src_pfns[64] = { 0 }; unsigned long dst_pfns[64] = { 0 }; struct dmirror_bounce bounce; - struct migrate_vma args; + struct migrate_vma args = { 0 }; unsigned long next; int ret; @@ -1334,7 +1334,7 @@ static void dmirror_devmem_free(struct page *page) static vm_fault_t dmirror_devmem_fault(struct vm_fault *vmf) { - struct migrate_vma args; + struct migrate_vma args = { 0 }; unsigned long src_pfns = 0; unsigned long dst_pfns = 0; struct page *rpage; @@ -1357,6 +1357,7 @@ static vm_fault_t dmirror_devmem_fault(struct vm_fault *vmf) args.dst = &dst_pfns; args.pgmap_owner = dmirror->mdevice; args.flags = dmirror_select_device(dmirror); + args.fault_page = vmf->page; if (migrate_vma_setup(&args)) return VM_FAULT_SIGBUS; diff --git a/mm/memory.c b/mm/memory.c index 2c7723ea4371..4ad6077164cd 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -3750,7 +3750,21 @@ vm_fault_t do_swap_page(struct vm_fault *vmf) ret = remove_device_exclusive_entry(vmf); } else if (is_device_private_entry(entry)) { vmf->page = pfn_swap_entry_to_page(entry); - ret = vmf->page->pgmap->ops->migrate_to_ram(vmf); + vmf->pte = pte_offset_map_lock(vma->vm_mm, vmf->pmd, + vmf->address, &vmf->ptl); + if (unlikely(!pte_same(*vmf->pte, vmf->orig_pte))) { + spin_unlock(vmf->ptl); + goto out; + } + + /* + * Get a page reference while we know the page can't be + * freed. + */ + get_page(vmf->page); + pte_unmap_unlock(vmf->pte, vmf->ptl); + vmf->page->pgmap->ops->migrate_to_ram(vmf); + put_page(vmf->page); } else if (is_hwpoison_entry(entry)) { ret = VM_FAULT_HWPOISON; } else if (is_swapin_error_entry(entry)) { diff --git a/mm/migrate.c b/mm/migrate.c index c228afba0963..1379e1912772 100644 --- a/mm/migrate.c +++ b/mm/migrate.c @@ -625,6 +625,25 @@ EXPORT_SYMBOL(folio_migrate_copy); * Migration functions ***********************************************************/ +int migrate_folio_extra(struct address_space *mapping, struct folio *dst, + struct folio *src, enum migrate_mode mode, int extra_count) +{ + int rc; + + BUG_ON(folio_test_writeback(src)); /* Writeback must be complete */ + + rc = folio_migrate_mapping(mapping, dst, src, extra_count); + + if (rc != MIGRATEPAGE_SUCCESS) + return rc; + + if (mode != MIGRATE_SYNC_NO_COPY) + folio_migrate_copy(dst, src); + else + folio_migrate_flags(dst, src); + return MIGRATEPAGE_SUCCESS; +} + /** * migrate_folio() - Simple folio migration. * @mapping: The address_space containing the folio. @@ -640,20 +659,7 @@ EXPORT_SYMBOL(folio_migrate_copy); int migrate_folio(struct address_space *mapping, struct folio *dst, struct folio *src, enum migrate_mode mode) { - int rc; - - BUG_ON(folio_test_writeback(src)); /* Writeback must be complete */ - - rc = folio_migrate_mapping(mapping, dst, src, 0); - - if (rc != MIGRATEPAGE_SUCCESS) - return rc; - - if (mode != MIGRATE_SYNC_NO_COPY) - folio_migrate_copy(dst, src); - else - folio_migrate_flags(dst, src); - return MIGRATEPAGE_SUCCESS; + return migrate_folio_extra(mapping, dst, src, mode, 0); } EXPORT_SYMBOL(migrate_folio); diff --git a/mm/migrate_device.c b/mm/migrate_device.c index 5ab6ab9d2ed8..8dee38ffcda2 100644 --- a/mm/migrate_device.c +++ b/mm/migrate_device.c @@ -325,14 +325,14 @@ static void migrate_vma_collect(struct migrate_vma *migrate) * folio_migrate_mapping(), except that here we allow migration of a * ZONE_DEVICE page. */ -static bool migrate_vma_check_page(struct page *page) +static bool migrate_vma_check_page(struct page *page, struct page *fault_page) { /* * One extra ref because caller holds an extra reference, either from * isolate_lru_page() for a regular page, or migrate_vma_collect() for * a device page. */ - int extra = 1; + int extra = 1 + (page == fault_page); /* * FIXME support THP (transparent huge page), it is bit more complex to @@ -405,7 +405,8 @@ static void migrate_vma_unmap(struct migrate_vma *migrate) if (folio_mapped(folio)) try_to_migrate(folio, 0); - if (page_mapped(page) || !migrate_vma_check_page(page)) { + if (page_mapped(page) || + !migrate_vma_check_page(page, migrate->fault_page)) { if (!is_zone_device_page(page)) { get_page(page); putback_lru_page(page); @@ -517,6 +518,8 @@ int migrate_vma_setup(struct migrate_vma *args) return -EINVAL; if (!args->src || !args->dst) return -EINVAL; + if (args->fault_page && !is_device_private_page(args->fault_page)) + return -EINVAL; memset(args->src, 0, sizeof(*args->src) * nr_pages); args->cpages = 0; @@ -747,8 +750,13 @@ void migrate_vma_pages(struct migrate_vma *migrate) continue; } - r = migrate_folio(mapping, page_folio(newpage), - page_folio(page), MIGRATE_SYNC_NO_COPY); + if (migrate->fault_page == page) + r = migrate_folio_extra(mapping, page_folio(newpage), + page_folio(page), + MIGRATE_SYNC_NO_COPY, 1); + else + r = migrate_folio(mapping, page_folio(newpage), + page_folio(page), MIGRATE_SYNC_NO_COPY); if (r != MIGRATEPAGE_SUCCESS) migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; } From ef233450898f8893dafa193a9f3211fa077a3d05 Mon Sep 17 00:00:00 2001 From: Alistair Popple Date: Wed, 28 Sep 2022 22:01:16 +1000 Subject: [PATCH 081/121] mm: free device private pages have zero refcount MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Since 27674ef6c73f ("mm: remove the extra ZONE_DEVICE struct page refcount") device private pages have no longer had an extra reference count when the page is in use. However before handing them back to the owning device driver we add an extra reference count such that free pages have a reference count of one. This makes it difficult to tell if a page is free or not because both free and in use pages will have a non-zero refcount. Instead we should return pages to the drivers page allocator with a zero reference count. Kernel code can then safely use kernel functions such as get_page_unless_zero(). Link: https://lkml.kernel.org/r/cf70cf6f8c0bdb8aaebdbfb0d790aea4c683c3c6.1664366292.git-series.apopple@nvidia.com Signed-off-by: Alistair Popple Acked-by: Felix Kuehling Cc: Jason Gunthorpe Cc: Michael Ellerman Cc: Alex Deucher Cc: Christian König Cc: Ben Skeggs Cc: Lyude Paul Cc: Ralph Campbell Cc: Alex Sierra Cc: John Hubbard Cc: Dan Williams Cc: David Hildenbrand Cc: "Huang, Ying" Cc: Matthew Wilcox Cc: Yang Shi Cc: Zi Yan Signed-off-by: Andrew Morton --- arch/powerpc/kvm/book3s_hv_uvmem.c | 2 +- drivers/gpu/drm/amd/amdkfd/kfd_migrate.c | 2 +- drivers/gpu/drm/nouveau/nouveau_dmem.c | 2 +- include/linux/memremap.h | 1 + lib/test_hmm.c | 2 +- mm/memremap.c | 11 ++++++++++- mm/page_alloc.c | 8 ++++++++ 7 files changed, 23 insertions(+), 5 deletions(-) diff --git a/arch/powerpc/kvm/book3s_hv_uvmem.c b/arch/powerpc/kvm/book3s_hv_uvmem.c index 965c9e9e500b..e2f11f9c3f2a 100644 --- a/arch/powerpc/kvm/book3s_hv_uvmem.c +++ b/arch/powerpc/kvm/book3s_hv_uvmem.c @@ -718,7 +718,7 @@ static struct page *kvmppc_uvmem_get_page(unsigned long gpa, struct kvm *kvm) dpage = pfn_to_page(uvmem_pfn); dpage->zone_device_data = pvt; - lock_page(dpage); + zone_device_page_init(dpage); return dpage; out_clear: spin_lock(&kvmppc_uvmem_bitmap_lock); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c index 776448bd9fe4..97a684568ae0 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c @@ -223,7 +223,7 @@ svm_migrate_get_vram_page(struct svm_range *prange, unsigned long pfn) page = pfn_to_page(pfn); svm_range_bo_ref(prange->svm_bo); page->zone_device_data = prange->svm_bo; - lock_page(page); + zone_device_page_init(page); } static void diff --git a/drivers/gpu/drm/nouveau/nouveau_dmem.c b/drivers/gpu/drm/nouveau/nouveau_dmem.c index 16356611b5b9..b092988266a6 100644 --- a/drivers/gpu/drm/nouveau/nouveau_dmem.c +++ b/drivers/gpu/drm/nouveau/nouveau_dmem.c @@ -326,7 +326,7 @@ nouveau_dmem_page_alloc_locked(struct nouveau_drm *drm) return NULL; } - lock_page(page); + zone_device_page_init(page); return page; } diff --git a/include/linux/memremap.h b/include/linux/memremap.h index c3b4cc84877b..7fcaf3180a5b 100644 --- a/include/linux/memremap.h +++ b/include/linux/memremap.h @@ -187,6 +187,7 @@ static inline bool folio_is_device_coherent(const struct folio *folio) } #ifdef CONFIG_ZONE_DEVICE +void zone_device_page_init(struct page *page); void *memremap_pages(struct dev_pagemap *pgmap, int nid); void memunmap_pages(struct dev_pagemap *pgmap); void *devm_memremap_pages(struct device *dev, struct dev_pagemap *pgmap); diff --git a/lib/test_hmm.c b/lib/test_hmm.c index e566166b5571..bc2b94991165 100644 --- a/lib/test_hmm.c +++ b/lib/test_hmm.c @@ -627,8 +627,8 @@ static struct page *dmirror_devmem_alloc_page(struct dmirror_device *mdevice) goto error; } + zone_device_page_init(dpage); dpage->zone_device_data = rpage; - lock_page(dpage); return dpage; error: diff --git a/mm/memremap.c b/mm/memremap.c index 25029a474d30..1c2c038f3410 100644 --- a/mm/memremap.c +++ b/mm/memremap.c @@ -505,9 +505,18 @@ void free_zone_device_page(struct page *page) /* * Reset the page count to 1 to prepare for handing out the page again. */ - set_page_count(page, 1); + if (page->pgmap->type != MEMORY_DEVICE_PRIVATE && + page->pgmap->type != MEMORY_DEVICE_COHERENT) + set_page_count(page, 1); } +void zone_device_page_init(struct page *page) +{ + set_page_count(page, 1); + lock_page(page); +} +EXPORT_SYMBOL_GPL(zone_device_page_init); + #ifdef CONFIG_FS_DAX bool __put_devmap_managed_page_refs(struct page *page, int refs) { diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 12b6184cbbed..059f6946832f 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -6819,6 +6819,14 @@ static void __ref __init_zone_device_page(struct page *page, unsigned long pfn, set_pageblock_migratetype(page, MIGRATE_MOVABLE); cond_resched(); } + + /* + * ZONE_DEVICE pages are released directly to the driver page allocator + * which will set the page count to 1 when allocating the page. + */ + if (pgmap->type == MEMORY_DEVICE_PRIVATE || + pgmap->type == MEMORY_DEVICE_COHERENT) + set_page_count(page, 0); } /* From 0dc45ca1ce18900572282c4f054bbe78351cb6a7 Mon Sep 17 00:00:00 2001 From: Alistair Popple Date: Wed, 28 Sep 2022 22:01:17 +1000 Subject: [PATCH 082/121] mm/memremap.c: take a pgmap reference on page allocation MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit ZONE_DEVICE pages have a struct dev_pagemap which is allocated by a driver. When the struct page is first allocated by the kernel in memremap_pages() a reference is taken on the associated pagemap to ensure it is not freed prior to the pages being freed. Prior to 27674ef6c73f ("mm: remove the extra ZONE_DEVICE struct page refcount") pages were considered free and returned to the driver when the reference count dropped to one. However the pagemap reference was not dropped until the page reference count hit zero. This would occur as part of the final put_page() in memunmap_pages() which would wait for all pages to be freed prior to returning. When the extra refcount was removed the pagemap reference was no longer being dropped in put_page(). Instead memunmap_pages() was changed to explicitly drop the pagemap references. This means that memunmap_pages() can complete even though pages are still mapped by the kernel which can lead to kernel crashes, particularly if a driver frees the pagemap. To fix this drivers should take a pagemap reference when allocating the page. This reference can then be returned when the page is freed. Link: https://lkml.kernel.org/r/12d155ec727935ebfbb4d639a03ab374917ea51b.1664366292.git-series.apopple@nvidia.com Signed-off-by: Alistair Popple Fixes: 27674ef6c73f ("mm: remove the extra ZONE_DEVICE struct page refcount") Cc: Jason Gunthorpe Cc: Felix Kuehling Cc: Alex Deucher Cc: Christian König Cc: Ben Skeggs Cc: Lyude Paul Cc: Ralph Campbell Cc: Alex Sierra Cc: John Hubbard Cc: Dan Williams Cc: David Hildenbrand Cc: "Huang, Ying" Cc: Matthew Wilcox Cc: Michael Ellerman Cc: Yang Shi Cc: Zi Yan Signed-off-by: Andrew Morton --- mm/memremap.c | 25 +++++++++++++++++++------ 1 file changed, 19 insertions(+), 6 deletions(-) diff --git a/mm/memremap.c b/mm/memremap.c index 1c2c038f3410..421bec3a29ee 100644 --- a/mm/memremap.c +++ b/mm/memremap.c @@ -138,8 +138,11 @@ void memunmap_pages(struct dev_pagemap *pgmap) int i; percpu_ref_kill(&pgmap->ref); - for (i = 0; i < pgmap->nr_range; i++) - percpu_ref_put_many(&pgmap->ref, pfn_len(pgmap, i)); + if (pgmap->type != MEMORY_DEVICE_PRIVATE && + pgmap->type != MEMORY_DEVICE_COHERENT) + for (i = 0; i < pgmap->nr_range; i++) + percpu_ref_put_many(&pgmap->ref, pfn_len(pgmap, i)); + wait_for_completion(&pgmap->done); for (i = 0; i < pgmap->nr_range; i++) @@ -264,7 +267,9 @@ static int pagemap_range(struct dev_pagemap *pgmap, struct mhp_params *params, memmap_init_zone_device(&NODE_DATA(nid)->node_zones[ZONE_DEVICE], PHYS_PFN(range->start), PHYS_PFN(range_len(range)), pgmap); - percpu_ref_get_many(&pgmap->ref, pfn_len(pgmap, range_id)); + if (pgmap->type != MEMORY_DEVICE_PRIVATE && + pgmap->type != MEMORY_DEVICE_COHERENT) + percpu_ref_get_many(&pgmap->ref, pfn_len(pgmap, range_id)); return 0; err_add_memory: @@ -502,16 +507,24 @@ void free_zone_device_page(struct page *page) page->mapping = NULL; page->pgmap->ops->page_free(page); - /* - * Reset the page count to 1 to prepare for handing out the page again. - */ if (page->pgmap->type != MEMORY_DEVICE_PRIVATE && page->pgmap->type != MEMORY_DEVICE_COHERENT) + /* + * Reset the page count to 1 to prepare for handing out the page + * again. + */ set_page_count(page, 1); + else + put_dev_pagemap(page->pgmap); } void zone_device_page_init(struct page *page) { + /* + * Drivers shouldn't be allocating pages after calling + * memunmap_pages(). + */ + WARN_ON_ONCE(!percpu_ref_tryget_live(&page->pgmap->ref)); set_page_count(page, 1); lock_page(page); } From 241f68859656836ae3e85179cc224cc4c5e4e6a7 Mon Sep 17 00:00:00 2001 From: Alistair Popple Date: Wed, 28 Sep 2022 22:01:18 +1000 Subject: [PATCH 083/121] mm/migrate_device.c: refactor migrate_vma and migrate_deivce_coherent_page() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit migrate_device_coherent_page() reuses the existing migrate_vma family of functions to migrate a specific page without providing a valid mapping or vma. This looks a bit odd because it means we are calling migrate_vma_*() without setting a valid vma, however it was considered acceptable at the time because the details were internal to migrate_device.c and there was only a single user. One of the reasons the details could be kept internal was that this was strictly for migrating device coherent memory. Such memory can be copied directly by the CPU without intervention from a driver. However this isn't true for device private memory, and a future change requires similar functionality for device private memory. So refactor the code into something more sensible for migrating device memory without a vma. Link: https://lkml.kernel.org/r/c7b2ff84e9b33d022cf4a40f87d051f281a16d8f.1664366292.git-series.apopple@nvidia.com Signed-off-by: Alistair Popple Cc: "Huang, Ying" Cc: Zi Yan Cc: Matthew Wilcox Cc: Yang Shi Cc: David Hildenbrand Cc: Ralph Campbell Cc: John Hubbard Cc: Alex Deucher Cc: Alex Sierra Cc: Ben Skeggs Cc: Christian König Cc: Dan Williams Cc: Felix Kuehling Cc: Jason Gunthorpe Cc: Lyude Paul Cc: Michael Ellerman Signed-off-by: Andrew Morton --- mm/migrate_device.c | 150 +++++++++++++++++++++++++------------------- 1 file changed, 85 insertions(+), 65 deletions(-) diff --git a/mm/migrate_device.c b/mm/migrate_device.c index 8dee38ffcda2..7707c1d898f5 100644 --- a/mm/migrate_device.c +++ b/mm/migrate_device.c @@ -357,26 +357,20 @@ static bool migrate_vma_check_page(struct page *page, struct page *fault_page) } /* - * migrate_vma_unmap() - replace page mapping with special migration pte entry - * @migrate: migrate struct containing all migration information - * - * Isolate pages from the LRU and replace mappings (CPU page table pte) with a - * special migration pte entry and check if it has been pinned. Pinned pages are - * restored because we cannot migrate them. - * - * This is the last step before we call the device driver callback to allocate - * destination memory and copy contents of original page over to new page. + * Unmaps pages for migration. Returns number of unmapped pages. */ -static void migrate_vma_unmap(struct migrate_vma *migrate) +static unsigned long migrate_device_unmap(unsigned long *src_pfns, + unsigned long npages, + struct page *fault_page) { - const unsigned long npages = migrate->npages; unsigned long i, restore = 0; bool allow_drain = true; + unsigned long unmapped = 0; lru_add_drain(); for (i = 0; i < npages; i++) { - struct page *page = migrate_pfn_to_page(migrate->src[i]); + struct page *page = migrate_pfn_to_page(src_pfns[i]); struct folio *folio; if (!page) @@ -391,8 +385,7 @@ static void migrate_vma_unmap(struct migrate_vma *migrate) } if (isolate_lru_page(page)) { - migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; - migrate->cpages--; + src_pfns[i] &= ~MIGRATE_PFN_MIGRATE; restore++; continue; } @@ -406,34 +399,54 @@ static void migrate_vma_unmap(struct migrate_vma *migrate) try_to_migrate(folio, 0); if (page_mapped(page) || - !migrate_vma_check_page(page, migrate->fault_page)) { + !migrate_vma_check_page(page, fault_page)) { if (!is_zone_device_page(page)) { get_page(page); putback_lru_page(page); } - migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; - migrate->cpages--; + src_pfns[i] &= ~MIGRATE_PFN_MIGRATE; restore++; continue; } + + unmapped++; } for (i = 0; i < npages && restore; i++) { - struct page *page = migrate_pfn_to_page(migrate->src[i]); + struct page *page = migrate_pfn_to_page(src_pfns[i]); struct folio *folio; - if (!page || (migrate->src[i] & MIGRATE_PFN_MIGRATE)) + if (!page || (src_pfns[i] & MIGRATE_PFN_MIGRATE)) continue; folio = page_folio(page); remove_migration_ptes(folio, folio, false); - migrate->src[i] = 0; + src_pfns[i] = 0; folio_unlock(folio); folio_put(folio); restore--; } + + return unmapped; +} + +/* + * migrate_vma_unmap() - replace page mapping with special migration pte entry + * @migrate: migrate struct containing all migration information + * + * Isolate pages from the LRU and replace mappings (CPU page table pte) with a + * special migration pte entry and check if it has been pinned. Pinned pages are + * restored because we cannot migrate them. + * + * This is the last step before we call the device driver callback to allocate + * destination memory and copy contents of original page over to new page. + */ +static void migrate_vma_unmap(struct migrate_vma *migrate) +{ + migrate->cpages = migrate_device_unmap(migrate->src, migrate->npages, + migrate->fault_page); } /** @@ -680,41 +693,36 @@ static void migrate_vma_insert_page(struct migrate_vma *migrate, *src &= ~MIGRATE_PFN_MIGRATE; } -/** - * migrate_vma_pages() - migrate meta-data from src page to dst page - * @migrate: migrate struct containing all migration information - * - * This migrates struct page meta-data from source struct page to destination - * struct page. This effectively finishes the migration from source page to the - * destination page. - */ -void migrate_vma_pages(struct migrate_vma *migrate) +static void migrate_device_pages(unsigned long *src_pfns, + unsigned long *dst_pfns, unsigned long npages, + struct migrate_vma *migrate) { - const unsigned long npages = migrate->npages; - const unsigned long start = migrate->start; struct mmu_notifier_range range; - unsigned long addr, i; + unsigned long i; bool notified = false; - for (i = 0, addr = start; i < npages; addr += PAGE_SIZE, i++) { - struct page *newpage = migrate_pfn_to_page(migrate->dst[i]); - struct page *page = migrate_pfn_to_page(migrate->src[i]); + for (i = 0; i < npages; i++) { + struct page *newpage = migrate_pfn_to_page(dst_pfns[i]); + struct page *page = migrate_pfn_to_page(src_pfns[i]); struct address_space *mapping; int r; if (!newpage) { - migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; + src_pfns[i] &= ~MIGRATE_PFN_MIGRATE; continue; } if (!page) { + unsigned long addr; + /* * The only time there is no vma is when called from * migrate_device_coherent_page(). However this isn't * called if the page could not be unmapped. */ - VM_BUG_ON(!migrate->vma); - if (!(migrate->src[i] & MIGRATE_PFN_MIGRATE)) + VM_BUG_ON(!migrate); + addr = migrate->start + i*PAGE_SIZE; + if (!(src_pfns[i] & MIGRATE_PFN_MIGRATE)) continue; if (!notified) { notified = true; @@ -726,7 +734,7 @@ void migrate_vma_pages(struct migrate_vma *migrate) mmu_notifier_invalidate_range_start(&range); } migrate_vma_insert_page(migrate, addr, newpage, - &migrate->src[i]); + &src_pfns[i]); continue; } @@ -739,18 +747,18 @@ void migrate_vma_pages(struct migrate_vma *migrate) * device private or coherent memory. */ if (mapping) { - migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; + src_pfns[i] &= ~MIGRATE_PFN_MIGRATE; continue; } } else if (is_zone_device_page(newpage)) { /* * Other types of ZONE_DEVICE page are not supported. */ - migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; + src_pfns[i] &= ~MIGRATE_PFN_MIGRATE; continue; } - if (migrate->fault_page == page) + if (migrate && migrate->fault_page == page) r = migrate_folio_extra(mapping, page_folio(newpage), page_folio(page), MIGRATE_SYNC_NO_COPY, 1); @@ -758,7 +766,7 @@ void migrate_vma_pages(struct migrate_vma *migrate) r = migrate_folio(mapping, page_folio(newpage), page_folio(page), MIGRATE_SYNC_NO_COPY); if (r != MIGRATEPAGE_SUCCESS) - migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; + src_pfns[i] &= ~MIGRATE_PFN_MIGRATE; } /* @@ -769,28 +777,30 @@ void migrate_vma_pages(struct migrate_vma *migrate) if (notified) mmu_notifier_invalidate_range_only_end(&range); } -EXPORT_SYMBOL(migrate_vma_pages); /** - * migrate_vma_finalize() - restore CPU page table entry + * migrate_vma_pages() - migrate meta-data from src page to dst page * @migrate: migrate struct containing all migration information * - * This replaces the special migration pte entry with either a mapping to the - * new page if migration was successful for that page, or to the original page - * otherwise. - * - * This also unlocks the pages and puts them back on the lru, or drops the extra - * refcount, for device pages. + * This migrates struct page meta-data from source struct page to destination + * struct page. This effectively finishes the migration from source page to the + * destination page. */ -void migrate_vma_finalize(struct migrate_vma *migrate) +void migrate_vma_pages(struct migrate_vma *migrate) +{ + migrate_device_pages(migrate->src, migrate->dst, migrate->npages, migrate); +} +EXPORT_SYMBOL(migrate_vma_pages); + +static void migrate_device_finalize(unsigned long *src_pfns, + unsigned long *dst_pfns, unsigned long npages) { - const unsigned long npages = migrate->npages; unsigned long i; for (i = 0; i < npages; i++) { struct folio *dst, *src; - struct page *newpage = migrate_pfn_to_page(migrate->dst[i]); - struct page *page = migrate_pfn_to_page(migrate->src[i]); + struct page *newpage = migrate_pfn_to_page(dst_pfns[i]); + struct page *page = migrate_pfn_to_page(src_pfns[i]); if (!page) { if (newpage) { @@ -800,7 +810,7 @@ void migrate_vma_finalize(struct migrate_vma *migrate) continue; } - if (!(migrate->src[i] & MIGRATE_PFN_MIGRATE) || !newpage) { + if (!(src_pfns[i] & MIGRATE_PFN_MIGRATE) || !newpage) { if (newpage) { unlock_page(newpage); put_page(newpage); @@ -827,6 +837,22 @@ void migrate_vma_finalize(struct migrate_vma *migrate) } } } + +/** + * migrate_vma_finalize() - restore CPU page table entry + * @migrate: migrate struct containing all migration information + * + * This replaces the special migration pte entry with either a mapping to the + * new page if migration was successful for that page, or to the original page + * otherwise. + * + * This also unlocks the pages and puts them back on the lru, or drops the extra + * refcount, for device pages. + */ +void migrate_vma_finalize(struct migrate_vma *migrate) +{ + migrate_device_finalize(migrate->src, migrate->dst, migrate->npages); +} EXPORT_SYMBOL(migrate_vma_finalize); /* @@ -837,25 +863,19 @@ EXPORT_SYMBOL(migrate_vma_finalize); int migrate_device_coherent_page(struct page *page) { unsigned long src_pfn, dst_pfn = 0; - struct migrate_vma args; struct page *dpage; WARN_ON_ONCE(PageCompound(page)); lock_page(page); src_pfn = migrate_pfn(page_to_pfn(page)) | MIGRATE_PFN_MIGRATE; - args.src = &src_pfn; - args.dst = &dst_pfn; - args.cpages = 1; - args.npages = 1; - args.vma = NULL; /* * We don't have a VMA and don't need to walk the page tables to find * the source page. So call migrate_vma_unmap() directly to unmap the * page as migrate_vma_setup() will fail if args.vma == NULL. */ - migrate_vma_unmap(&args); + migrate_device_unmap(&src_pfn, 1, NULL); if (!(src_pfn & MIGRATE_PFN_MIGRATE)) return -EBUSY; @@ -865,10 +885,10 @@ int migrate_device_coherent_page(struct page *page) dst_pfn = migrate_pfn(page_to_pfn(dpage)); } - migrate_vma_pages(&args); + migrate_device_pages(&src_pfn, &dst_pfn, 1, NULL); if (src_pfn & MIGRATE_PFN_MIGRATE) copy_highpage(dpage, page); - migrate_vma_finalize(&args); + migrate_device_finalize(&src_pfn, &dst_pfn, 1); if (src_pfn & MIGRATE_PFN_MIGRATE) return 0; From e778406b40dbb1342a1888cd751ca9d2982a12e2 Mon Sep 17 00:00:00 2001 From: Alistair Popple Date: Wed, 28 Sep 2022 22:01:19 +1000 Subject: [PATCH 084/121] mm/migrate_device.c: add migrate_device_range() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Device drivers can use the migrate_vma family of functions to migrate existing private anonymous mappings to device private pages. These pages are backed by memory on the device with drivers being responsible for copying data to and from device memory. Device private pages are freed via the pgmap->page_free() callback when they are unmapped and their refcount drops to zero. Alternatively they may be freed indirectly via migration back to CPU memory in response to a pgmap->migrate_to_ram() callback called whenever the CPU accesses an address mapped to a device private page. In other words drivers cannot control the lifetime of data allocated on the devices and must wait until these pages are freed from userspace. This causes issues when memory needs to reclaimed on the device, either because the device is going away due to a ->release() callback or because another user needs to use the memory. Drivers could use the existing migrate_vma functions to migrate data off the device. However this would require them to track the mappings of each page which is both complicated and not always possible. Instead drivers need to be able to migrate device pages directly so they can free up device memory. To allow that this patch introduces the migrate_device family of functions which are functionally similar to migrate_vma but which skips the initial lookup based on mapping. Link: https://lkml.kernel.org/r/868116aab70b0c8ee467d62498bb2cf0ef907295.1664366292.git-series.apopple@nvidia.com Signed-off-by: Alistair Popple Cc: "Huang, Ying" Cc: Zi Yan Cc: Matthew Wilcox Cc: Yang Shi Cc: David Hildenbrand Cc: Ralph Campbell Cc: John Hubbard Cc: Alex Deucher Cc: Alex Sierra Cc: Ben Skeggs Cc: Christian König Cc: Dan Williams Cc: Felix Kuehling Cc: Jason Gunthorpe Cc: Lyude Paul Cc: Michael Ellerman Signed-off-by: Andrew Morton --- include/linux/migrate.h | 7 ++++ mm/migrate_device.c | 89 +++++++++++++++++++++++++++++++++++++---- 2 files changed, 89 insertions(+), 7 deletions(-) diff --git a/include/linux/migrate.h b/include/linux/migrate.h index 52090d1f9230..3ef77f52a4f0 100644 --- a/include/linux/migrate.h +++ b/include/linux/migrate.h @@ -210,6 +210,13 @@ struct migrate_vma { int migrate_vma_setup(struct migrate_vma *args); void migrate_vma_pages(struct migrate_vma *migrate); void migrate_vma_finalize(struct migrate_vma *migrate); +int migrate_device_range(unsigned long *src_pfns, unsigned long start, + unsigned long npages); +void migrate_device_pages(unsigned long *src_pfns, unsigned long *dst_pfns, + unsigned long npages); +void migrate_device_finalize(unsigned long *src_pfns, + unsigned long *dst_pfns, unsigned long npages); + #endif /* CONFIG_MIGRATION */ #endif /* _LINUX_MIGRATE_H */ diff --git a/mm/migrate_device.c b/mm/migrate_device.c index 7707c1d898f5..6fa682eef7a0 100644 --- a/mm/migrate_device.c +++ b/mm/migrate_device.c @@ -693,7 +693,7 @@ static void migrate_vma_insert_page(struct migrate_vma *migrate, *src &= ~MIGRATE_PFN_MIGRATE; } -static void migrate_device_pages(unsigned long *src_pfns, +static void __migrate_device_pages(unsigned long *src_pfns, unsigned long *dst_pfns, unsigned long npages, struct migrate_vma *migrate) { @@ -715,6 +715,9 @@ static void migrate_device_pages(unsigned long *src_pfns, if (!page) { unsigned long addr; + if (!(src_pfns[i] & MIGRATE_PFN_MIGRATE)) + continue; + /* * The only time there is no vma is when called from * migrate_device_coherent_page(). However this isn't @@ -722,8 +725,6 @@ static void migrate_device_pages(unsigned long *src_pfns, */ VM_BUG_ON(!migrate); addr = migrate->start + i*PAGE_SIZE; - if (!(src_pfns[i] & MIGRATE_PFN_MIGRATE)) - continue; if (!notified) { notified = true; @@ -778,6 +779,22 @@ static void migrate_device_pages(unsigned long *src_pfns, mmu_notifier_invalidate_range_only_end(&range); } +/** + * migrate_device_pages() - migrate meta-data from src page to dst page + * @src_pfns: src_pfns returned from migrate_device_range() + * @dst_pfns: array of pfns allocated by the driver to migrate memory to + * @npages: number of pages in the range + * + * Equivalent to migrate_vma_pages(). This is called to migrate struct page + * meta-data from source struct page to destination. + */ +void migrate_device_pages(unsigned long *src_pfns, unsigned long *dst_pfns, + unsigned long npages) +{ + __migrate_device_pages(src_pfns, dst_pfns, npages, NULL); +} +EXPORT_SYMBOL(migrate_device_pages); + /** * migrate_vma_pages() - migrate meta-data from src page to dst page * @migrate: migrate struct containing all migration information @@ -788,12 +805,22 @@ static void migrate_device_pages(unsigned long *src_pfns, */ void migrate_vma_pages(struct migrate_vma *migrate) { - migrate_device_pages(migrate->src, migrate->dst, migrate->npages, migrate); + __migrate_device_pages(migrate->src, migrate->dst, migrate->npages, migrate); } EXPORT_SYMBOL(migrate_vma_pages); -static void migrate_device_finalize(unsigned long *src_pfns, - unsigned long *dst_pfns, unsigned long npages) +/* + * migrate_device_finalize() - complete page migration + * @src_pfns: src_pfns returned from migrate_device_range() + * @dst_pfns: array of pfns allocated by the driver to migrate memory to + * @npages: number of pages in the range + * + * Completes migration of the page by removing special migration entries. + * Drivers must ensure copying of page data is complete and visible to the CPU + * before calling this. + */ +void migrate_device_finalize(unsigned long *src_pfns, + unsigned long *dst_pfns, unsigned long npages) { unsigned long i; @@ -837,6 +864,7 @@ static void migrate_device_finalize(unsigned long *src_pfns, } } } +EXPORT_SYMBOL(migrate_device_finalize); /** * migrate_vma_finalize() - restore CPU page table entry @@ -855,6 +883,53 @@ void migrate_vma_finalize(struct migrate_vma *migrate) } EXPORT_SYMBOL(migrate_vma_finalize); +/** + * migrate_device_range() - migrate device private pfns to normal memory. + * @src_pfns: array large enough to hold migrating source device private pfns. + * @start: starting pfn in the range to migrate. + * @npages: number of pages to migrate. + * + * migrate_vma_setup() is similar in concept to migrate_vma_setup() except that + * instead of looking up pages based on virtual address mappings a range of + * device pfns that should be migrated to system memory is used instead. + * + * This is useful when a driver needs to free device memory but doesn't know the + * virtual mappings of every page that may be in device memory. For example this + * is often the case when a driver is being unloaded or unbound from a device. + * + * Like migrate_vma_setup() this function will take a reference and lock any + * migrating pages that aren't free before unmapping them. Drivers may then + * allocate destination pages and start copying data from the device to CPU + * memory before calling migrate_device_pages(). + */ +int migrate_device_range(unsigned long *src_pfns, unsigned long start, + unsigned long npages) +{ + unsigned long i, pfn; + + for (pfn = start, i = 0; i < npages; pfn++, i++) { + struct page *page = pfn_to_page(pfn); + + if (!get_page_unless_zero(page)) { + src_pfns[i] = 0; + continue; + } + + if (!trylock_page(page)) { + src_pfns[i] = 0; + put_page(page); + continue; + } + + src_pfns[i] = migrate_pfn(pfn) | MIGRATE_PFN_MIGRATE; + } + + migrate_device_unmap(src_pfns, npages, NULL); + + return 0; +} +EXPORT_SYMBOL(migrate_device_range); + /* * Migrate a device coherent page back to normal memory. The caller should have * a reference on page which will be copied to the new page if migration is @@ -885,7 +960,7 @@ int migrate_device_coherent_page(struct page *page) dst_pfn = migrate_pfn(page_to_pfn(dpage)); } - migrate_device_pages(&src_pfn, &dst_pfn, 1, NULL); + migrate_device_pages(&src_pfn, &dst_pfn, 1); if (src_pfn & MIGRATE_PFN_MIGRATE) copy_highpage(dpage, page); migrate_device_finalize(&src_pfn, &dst_pfn, 1); From d9b719394a1147614351961ac454589111c76e76 Mon Sep 17 00:00:00 2001 From: Alistair Popple Date: Wed, 28 Sep 2022 22:01:20 +1000 Subject: [PATCH 085/121] nouveau/dmem: refactor nouveau_dmem_fault_copy_one() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit nouveau_dmem_fault_copy_one() is used during handling of CPU faults via the migrate_to_ram() callback and is used to copy data from GPU to CPU memory. It is currently specific to fault handling, however a future patch implementing eviction of data during teardown needs similar functionality. Refactor out the core functionality so that it is not specific to fault handling. Link: https://lkml.kernel.org/r/20573d7b4e641a78fde9935f948e64e71c9e709e.1664366292.git-series.apopple@nvidia.com Signed-off-by: Alistair Popple Reviewed-by: Lyude Paul Cc: Ben Skeggs Cc: Ralph Campbell Cc: John Hubbard Cc: Alex Deucher Cc: Alex Sierra Cc: Christian König Cc: Dan Williams Cc: David Hildenbrand Cc: Felix Kuehling Cc: "Huang, Ying" Cc: Jason Gunthorpe Cc: Matthew Wilcox Cc: Michael Ellerman Cc: Yang Shi Cc: Zi Yan Signed-off-by: Andrew Morton --- drivers/gpu/drm/nouveau/nouveau_dmem.c | 58 +++++++++++++------------- 1 file changed, 28 insertions(+), 30 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nouveau_dmem.c b/drivers/gpu/drm/nouveau/nouveau_dmem.c index b092988266a6..65f51fb6a70c 100644 --- a/drivers/gpu/drm/nouveau/nouveau_dmem.c +++ b/drivers/gpu/drm/nouveau/nouveau_dmem.c @@ -139,44 +139,24 @@ static void nouveau_dmem_fence_done(struct nouveau_fence **fence) } } -static vm_fault_t nouveau_dmem_fault_copy_one(struct nouveau_drm *drm, - struct vm_fault *vmf, struct migrate_vma *args, - dma_addr_t *dma_addr) +static int nouveau_dmem_copy_one(struct nouveau_drm *drm, struct page *spage, + struct page *dpage, dma_addr_t *dma_addr) { struct device *dev = drm->dev->dev; - struct page *dpage, *spage; - struct nouveau_svmm *svmm; - spage = migrate_pfn_to_page(args->src[0]); - if (!spage || !(args->src[0] & MIGRATE_PFN_MIGRATE)) - return 0; - - dpage = alloc_page_vma(GFP_HIGHUSER, vmf->vma, vmf->address); - if (!dpage) - return VM_FAULT_SIGBUS; lock_page(dpage); *dma_addr = dma_map_page(dev, dpage, 0, PAGE_SIZE, DMA_BIDIRECTIONAL); if (dma_mapping_error(dev, *dma_addr)) - goto error_free_page; + return -EIO; - svmm = spage->zone_device_data; - mutex_lock(&svmm->mutex); - nouveau_svmm_invalidate(svmm, args->start, args->end); if (drm->dmem->migrate.copy_func(drm, 1, NOUVEAU_APER_HOST, *dma_addr, - NOUVEAU_APER_VRAM, nouveau_dmem_page_addr(spage))) - goto error_dma_unmap; - mutex_unlock(&svmm->mutex); + NOUVEAU_APER_VRAM, nouveau_dmem_page_addr(spage))) { + dma_unmap_page(dev, *dma_addr, PAGE_SIZE, DMA_BIDIRECTIONAL); + return -EIO; + } - args->dst[0] = migrate_pfn(page_to_pfn(dpage)); return 0; - -error_dma_unmap: - mutex_unlock(&svmm->mutex); - dma_unmap_page(dev, *dma_addr, PAGE_SIZE, DMA_BIDIRECTIONAL); -error_free_page: - __free_page(dpage); - return VM_FAULT_SIGBUS; } static vm_fault_t nouveau_dmem_migrate_to_ram(struct vm_fault *vmf) @@ -184,9 +164,11 @@ static vm_fault_t nouveau_dmem_migrate_to_ram(struct vm_fault *vmf) struct nouveau_drm *drm = page_to_drm(vmf->page); struct nouveau_dmem *dmem = drm->dmem; struct nouveau_fence *fence; + struct nouveau_svmm *svmm; + struct page *spage, *dpage; unsigned long src = 0, dst = 0; dma_addr_t dma_addr = 0; - vm_fault_t ret; + vm_fault_t ret = 0; struct migrate_vma args = { .vma = vmf->vma, .start = vmf->address, @@ -207,10 +189,26 @@ static vm_fault_t nouveau_dmem_migrate_to_ram(struct vm_fault *vmf) if (!args.cpages) return 0; - ret = nouveau_dmem_fault_copy_one(drm, vmf, &args, &dma_addr); - if (ret || dst == 0) + spage = migrate_pfn_to_page(src); + if (!spage || !(src & MIGRATE_PFN_MIGRATE)) goto done; + dpage = alloc_page_vma(GFP_HIGHUSER, vmf->vma, vmf->address); + if (!dpage) + goto done; + + dst = migrate_pfn(page_to_pfn(dpage)); + + svmm = spage->zone_device_data; + mutex_lock(&svmm->mutex); + nouveau_svmm_invalidate(svmm, args.start, args.end); + ret = nouveau_dmem_copy_one(drm, spage, dpage, &dma_addr); + mutex_unlock(&svmm->mutex); + if (ret) { + ret = VM_FAULT_SIGBUS; + goto done; + } + nouveau_fence_new(dmem->migrate.chan, false, &fence); migrate_vma_pages(&args); nouveau_dmem_fence_done(&fence); From 249881232e1471d28b68f9a3829acc14d150cf5d Mon Sep 17 00:00:00 2001 From: Alistair Popple Date: Wed, 28 Sep 2022 22:01:21 +1000 Subject: [PATCH 086/121] nouveau/dmem: evict device private memory during release MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When the module is unloaded or a GPU is unbound from the module it is possible for device private pages to still be mapped in currently running processes. This can lead to a hangs and RCU stall warnings when unbinding the device as memunmap_pages() will wait in an uninterruptible state until all device pages have been freed which may never happen. Fix this by migrating device mappings back to normal CPU memory prior to freeing the GPU memory chunks and associated device private pages. Link: https://lkml.kernel.org/r/66277601fb8fda9af408b33da9887192bf895bda.1664366292.git-series.apopple@nvidia.com Signed-off-by: Alistair Popple Cc: Lyude Paul Cc: Ben Skeggs Cc: Ralph Campbell Cc: John Hubbard Cc: Alex Deucher Cc: Alex Sierra Cc: Christian König Cc: Dan Williams Cc: David Hildenbrand Cc: Felix Kuehling Cc: "Huang, Ying" Cc: Jason Gunthorpe Cc: Matthew Wilcox Cc: Michael Ellerman Cc: Yang Shi Cc: Zi Yan Signed-off-by: Andrew Morton --- drivers/gpu/drm/nouveau/nouveau_dmem.c | 48 ++++++++++++++++++++++++++ 1 file changed, 48 insertions(+) diff --git a/drivers/gpu/drm/nouveau/nouveau_dmem.c b/drivers/gpu/drm/nouveau/nouveau_dmem.c index 65f51fb6a70c..5fe209107246 100644 --- a/drivers/gpu/drm/nouveau/nouveau_dmem.c +++ b/drivers/gpu/drm/nouveau/nouveau_dmem.c @@ -367,6 +367,52 @@ nouveau_dmem_suspend(struct nouveau_drm *drm) mutex_unlock(&drm->dmem->mutex); } +/* + * Evict all pages mapping a chunk. + */ +static void +nouveau_dmem_evict_chunk(struct nouveau_dmem_chunk *chunk) +{ + unsigned long i, npages = range_len(&chunk->pagemap.range) >> PAGE_SHIFT; + unsigned long *src_pfns, *dst_pfns; + dma_addr_t *dma_addrs; + struct nouveau_fence *fence; + + src_pfns = kcalloc(npages, sizeof(*src_pfns), GFP_KERNEL); + dst_pfns = kcalloc(npages, sizeof(*dst_pfns), GFP_KERNEL); + dma_addrs = kcalloc(npages, sizeof(*dma_addrs), GFP_KERNEL); + + migrate_device_range(src_pfns, chunk->pagemap.range.start >> PAGE_SHIFT, + npages); + + for (i = 0; i < npages; i++) { + if (src_pfns[i] & MIGRATE_PFN_MIGRATE) { + struct page *dpage; + + /* + * _GFP_NOFAIL because the GPU is going away and there + * is nothing sensible we can do if we can't copy the + * data back. + */ + dpage = alloc_page(GFP_HIGHUSER | __GFP_NOFAIL); + dst_pfns[i] = migrate_pfn(page_to_pfn(dpage)); + nouveau_dmem_copy_one(chunk->drm, + migrate_pfn_to_page(src_pfns[i]), dpage, + &dma_addrs[i]); + } + } + + nouveau_fence_new(chunk->drm->dmem->migrate.chan, false, &fence); + migrate_device_pages(src_pfns, dst_pfns, npages); + nouveau_dmem_fence_done(&fence); + migrate_device_finalize(src_pfns, dst_pfns, npages); + kfree(src_pfns); + kfree(dst_pfns); + for (i = 0; i < npages; i++) + dma_unmap_page(chunk->drm->dev->dev, dma_addrs[i], PAGE_SIZE, DMA_BIDIRECTIONAL); + kfree(dma_addrs); +} + void nouveau_dmem_fini(struct nouveau_drm *drm) { @@ -378,8 +424,10 @@ nouveau_dmem_fini(struct nouveau_drm *drm) mutex_lock(&drm->dmem->mutex); list_for_each_entry_safe(chunk, tmp, &drm->dmem->chunks, list) { + nouveau_dmem_evict_chunk(chunk); nouveau_bo_unpin(chunk->bo); nouveau_bo_ref(NULL, &chunk->bo); + WARN_ON(chunk->callocated); list_del(&chunk->list); memunmap_pages(&chunk->pagemap); release_mem_region(chunk->pagemap.range.start, From ad4c365221b0f92f9c24a203119f2bade30c970e Mon Sep 17 00:00:00 2001 From: Alistair Popple Date: Wed, 28 Sep 2022 22:01:22 +1000 Subject: [PATCH 087/121] hmm-tests: add test for migrate_device_range() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Link: https://lkml.kernel.org/r/a73cf109de0224cfd118d22be58ddebac3ae2897.1664366292.git-series.apopple@nvidia.com Signed-off-by: Alistair Popple Cc: Jason Gunthorpe Cc: Ralph Campbell Cc: John Hubbard Cc: Alex Sierra Cc: Felix Kuehling Cc: Alex Deucher Cc: Ben Skeggs Cc: Christian König Cc: Dan Williams Cc: David Hildenbrand Cc: "Huang, Ying" Cc: Lyude Paul Cc: Matthew Wilcox Cc: Michael Ellerman Cc: Yang Shi Cc: Zi Yan Signed-off-by: Andrew Morton --- lib/test_hmm.c | 120 ++++++++++++++++++++----- lib/test_hmm_uapi.h | 1 + tools/testing/selftests/vm/hmm-tests.c | 49 ++++++++++ 3 files changed, 149 insertions(+), 21 deletions(-) diff --git a/lib/test_hmm.c b/lib/test_hmm.c index bc2b94991165..67e6f83fe0f8 100644 --- a/lib/test_hmm.c +++ b/lib/test_hmm.c @@ -100,6 +100,7 @@ struct dmirror { struct dmirror_chunk { struct dev_pagemap pagemap; struct dmirror_device *mdevice; + bool remove; }; /* @@ -192,11 +193,15 @@ static int dmirror_fops_release(struct inode *inode, struct file *filp) return 0; } +static struct dmirror_chunk *dmirror_page_to_chunk(struct page *page) +{ + return container_of(page->pgmap, struct dmirror_chunk, pagemap); +} + static struct dmirror_device *dmirror_page_to_device(struct page *page) { - return container_of(page->pgmap, struct dmirror_chunk, - pagemap)->mdevice; + return dmirror_page_to_chunk(page)->mdevice; } static int dmirror_do_fault(struct dmirror *dmirror, struct hmm_range *range) @@ -1218,6 +1223,85 @@ static int dmirror_snapshot(struct dmirror *dmirror, return ret; } +static void dmirror_device_evict_chunk(struct dmirror_chunk *chunk) +{ + unsigned long start_pfn = chunk->pagemap.range.start >> PAGE_SHIFT; + unsigned long end_pfn = chunk->pagemap.range.end >> PAGE_SHIFT; + unsigned long npages = end_pfn - start_pfn + 1; + unsigned long i; + unsigned long *src_pfns; + unsigned long *dst_pfns; + + src_pfns = kcalloc(npages, sizeof(*src_pfns), GFP_KERNEL); + dst_pfns = kcalloc(npages, sizeof(*dst_pfns), GFP_KERNEL); + + migrate_device_range(src_pfns, start_pfn, npages); + for (i = 0; i < npages; i++) { + struct page *dpage, *spage; + + spage = migrate_pfn_to_page(src_pfns[i]); + if (!spage || !(src_pfns[i] & MIGRATE_PFN_MIGRATE)) + continue; + + if (WARN_ON(!is_device_private_page(spage) && + !is_device_coherent_page(spage))) + continue; + spage = BACKING_PAGE(spage); + dpage = alloc_page(GFP_HIGHUSER_MOVABLE | __GFP_NOFAIL); + lock_page(dpage); + copy_highpage(dpage, spage); + dst_pfns[i] = migrate_pfn(page_to_pfn(dpage)); + if (src_pfns[i] & MIGRATE_PFN_WRITE) + dst_pfns[i] |= MIGRATE_PFN_WRITE; + } + migrate_device_pages(src_pfns, dst_pfns, npages); + migrate_device_finalize(src_pfns, dst_pfns, npages); + kfree(src_pfns); + kfree(dst_pfns); +} + +/* Removes free pages from the free list so they can't be re-allocated */ +static void dmirror_remove_free_pages(struct dmirror_chunk *devmem) +{ + struct dmirror_device *mdevice = devmem->mdevice; + struct page *page; + + for (page = mdevice->free_pages; page; page = page->zone_device_data) + if (dmirror_page_to_chunk(page) == devmem) + mdevice->free_pages = page->zone_device_data; +} + +static void dmirror_device_remove_chunks(struct dmirror_device *mdevice) +{ + unsigned int i; + + mutex_lock(&mdevice->devmem_lock); + if (mdevice->devmem_chunks) { + for (i = 0; i < mdevice->devmem_count; i++) { + struct dmirror_chunk *devmem = + mdevice->devmem_chunks[i]; + + spin_lock(&mdevice->lock); + devmem->remove = true; + dmirror_remove_free_pages(devmem); + spin_unlock(&mdevice->lock); + + dmirror_device_evict_chunk(devmem); + memunmap_pages(&devmem->pagemap); + if (devmem->pagemap.type == MEMORY_DEVICE_PRIVATE) + release_mem_region(devmem->pagemap.range.start, + range_len(&devmem->pagemap.range)); + kfree(devmem); + } + mdevice->devmem_count = 0; + mdevice->devmem_capacity = 0; + mdevice->free_pages = NULL; + kfree(mdevice->devmem_chunks); + mdevice->devmem_chunks = NULL; + } + mutex_unlock(&mdevice->devmem_lock); +} + static long dmirror_fops_unlocked_ioctl(struct file *filp, unsigned int command, unsigned long arg) @@ -1272,6 +1356,11 @@ static long dmirror_fops_unlocked_ioctl(struct file *filp, ret = dmirror_snapshot(dmirror, &cmd); break; + case HMM_DMIRROR_RELEASE: + dmirror_device_remove_chunks(dmirror->mdevice); + ret = 0; + break; + default: return -EINVAL; } @@ -1326,9 +1415,13 @@ static void dmirror_devmem_free(struct page *page) mdevice = dmirror_page_to_device(page); spin_lock(&mdevice->lock); - mdevice->cfree++; - page->zone_device_data = mdevice->free_pages; - mdevice->free_pages = page; + + /* Return page to our allocator if not freeing the chunk */ + if (!dmirror_page_to_chunk(page)->remove) { + mdevice->cfree++; + page->zone_device_data = mdevice->free_pages; + mdevice->free_pages = page; + } spin_unlock(&mdevice->lock); } @@ -1408,22 +1501,7 @@ static int dmirror_device_init(struct dmirror_device *mdevice, int id) static void dmirror_device_remove(struct dmirror_device *mdevice) { - unsigned int i; - - if (mdevice->devmem_chunks) { - for (i = 0; i < mdevice->devmem_count; i++) { - struct dmirror_chunk *devmem = - mdevice->devmem_chunks[i]; - - memunmap_pages(&devmem->pagemap); - if (devmem->pagemap.type == MEMORY_DEVICE_PRIVATE) - release_mem_region(devmem->pagemap.range.start, - range_len(&devmem->pagemap.range)); - kfree(devmem); - } - kfree(mdevice->devmem_chunks); - } - + dmirror_device_remove_chunks(mdevice); cdev_device_del(&mdevice->cdevice, &mdevice->device); } diff --git a/lib/test_hmm_uapi.h b/lib/test_hmm_uapi.h index e31d58c9034a..8c818a2cf4f6 100644 --- a/lib/test_hmm_uapi.h +++ b/lib/test_hmm_uapi.h @@ -36,6 +36,7 @@ struct hmm_dmirror_cmd { #define HMM_DMIRROR_SNAPSHOT _IOWR('H', 0x04, struct hmm_dmirror_cmd) #define HMM_DMIRROR_EXCLUSIVE _IOWR('H', 0x05, struct hmm_dmirror_cmd) #define HMM_DMIRROR_CHECK_EXCLUSIVE _IOWR('H', 0x06, struct hmm_dmirror_cmd) +#define HMM_DMIRROR_RELEASE _IOWR('H', 0x07, struct hmm_dmirror_cmd) /* * Values returned in hmm_dmirror_cmd.ptr for HMM_DMIRROR_SNAPSHOT. diff --git a/tools/testing/selftests/vm/hmm-tests.c b/tools/testing/selftests/vm/hmm-tests.c index f2c2c970eeb2..28232adec883 100644 --- a/tools/testing/selftests/vm/hmm-tests.c +++ b/tools/testing/selftests/vm/hmm-tests.c @@ -1054,6 +1054,55 @@ TEST_F(hmm, migrate_fault) hmm_buffer_free(buffer); } +TEST_F(hmm, migrate_release) +{ + struct hmm_buffer *buffer; + unsigned long npages; + unsigned long size; + unsigned long i; + int *ptr; + int ret; + + npages = ALIGN(HMM_BUFFER_SIZE, self->page_size) >> self->page_shift; + ASSERT_NE(npages, 0); + size = npages << self->page_shift; + + buffer = malloc(sizeof(*buffer)); + ASSERT_NE(buffer, NULL); + + buffer->fd = -1; + buffer->size = size; + buffer->mirror = malloc(size); + ASSERT_NE(buffer->mirror, NULL); + + buffer->ptr = mmap(NULL, size, PROT_READ | PROT_WRITE, + MAP_PRIVATE | MAP_ANONYMOUS, buffer->fd, 0); + ASSERT_NE(buffer->ptr, MAP_FAILED); + + /* Initialize buffer in system memory. */ + for (i = 0, ptr = buffer->ptr; i < size / sizeof(*ptr); ++i) + ptr[i] = i; + + /* Migrate memory to device. */ + ret = hmm_migrate_sys_to_dev(self->fd, buffer, npages); + ASSERT_EQ(ret, 0); + ASSERT_EQ(buffer->cpages, npages); + + /* Check what the device read. */ + for (i = 0, ptr = buffer->mirror; i < size / sizeof(*ptr); ++i) + ASSERT_EQ(ptr[i], i); + + /* Release device memory. */ + ret = hmm_dmirror_cmd(self->fd, HMM_DMIRROR_RELEASE, buffer, npages); + ASSERT_EQ(ret, 0); + + /* Fault pages back to system memory and check them. */ + for (i = 0, ptr = buffer->ptr; i < size / (2 * sizeof(*ptr)); ++i) + ASSERT_EQ(ptr[i], i); + + hmm_buffer_free(buffer); +} + /* * Migrate anonymous shared memory to device private memory. */ From d6e5040bd8e53371fafd7e0c7c63b090b3a675db Mon Sep 17 00:00:00 2001 From: Andrey Konovalov Date: Mon, 26 Sep 2022 20:08:47 +0200 Subject: [PATCH 088/121] kasan: fix array-bounds warnings in tests GCC's -Warray-bounds option detects out-of-bounds accesses to statically-sized allocations in krealloc out-of-bounds tests. Use OPTIMIZER_HIDE_VAR to suppress the warning. Also change kmalloc_memmove_invalid_size to use OPTIMIZER_HIDE_VAR instead of a volatile variable. Link: https://lkml.kernel.org/r/e94399242d32e00bba6fd0d9ec4c897f188128e8.1664215688.git.andreyknvl@google.com Signed-off-by: Andrey Konovalov Reported-by: kernel test robot Reviewed-by: Kees Cook Cc: Alexander Potapenko Cc: Andrey Ryabinin Cc: Dmitry Vyukov Cc: Marco Elver Signed-off-by: Andrew Morton --- mm/kasan/kasan_test.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/mm/kasan/kasan_test.c b/mm/kasan/kasan_test.c index f25692def781..57e4c72aa8bd 100644 --- a/mm/kasan/kasan_test.c +++ b/mm/kasan/kasan_test.c @@ -295,6 +295,9 @@ static void krealloc_more_oob_helper(struct kunit *test, ptr2 = krealloc(ptr1, size2, GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr2); + /* Suppress -Warray-bounds warnings. */ + OPTIMIZER_HIDE_VAR(ptr2); + /* All offsets up to size2 must be accessible. */ ptr2[size1 - 1] = 'x'; ptr2[size1] = 'x'; @@ -327,6 +330,9 @@ static void krealloc_less_oob_helper(struct kunit *test, ptr2 = krealloc(ptr1, size2, GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr2); + /* Suppress -Warray-bounds warnings. */ + OPTIMIZER_HIDE_VAR(ptr2); + /* Must be accessible for all modes. */ ptr2[size2 - 1] = 'x'; @@ -540,13 +546,14 @@ static void kmalloc_memmove_invalid_size(struct kunit *test) { char *ptr; size_t size = 64; - volatile size_t invalid_size = size; + size_t invalid_size = size; ptr = kmalloc(size, GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); memset((char *)ptr, 0, 64); OPTIMIZER_HIDE_VAR(ptr); + OPTIMIZER_HIDE_VAR(invalid_size); KUNIT_EXPECT_KASAN_FAIL(test, memmove((char *)ptr, (char *)ptr + 4, invalid_size)); kfree(ptr); From bce8cb3c04dc01d21b6b17baf1cb6c277e7e6848 Mon Sep 17 00:00:00 2001 From: Qi Zheng Date: Thu, 29 Sep 2022 19:23:17 +0800 Subject: [PATCH 089/121] mm: use update_mmu_tlb() on the second thread As message in commit 7df676974359 ("mm/memory.c: Update local TLB if PTE entry exists") said, we should update local TLB only on the second thread. So in the do_anonymous_page() here, we should use update_mmu_tlb() instead of update_mmu_cache() on the second thread. As David pointed out, this is a performance improvement, not a correctness fix. Link: https://lkml.kernel.org/r/20220929112318.32393-2-zhengqi.arch@bytedance.com Signed-off-by: Qi Zheng Reviewed-by: Muchun Song Acked-by: David Hildenbrand Cc: Bibo Mao Cc: Chris Zankel Cc: Huacai Chen Cc: Max Filippov Signed-off-by: Andrew Morton --- mm/memory.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/memory.c b/mm/memory.c index 4ad6077164cd..f88c351aecd4 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -4134,7 +4134,7 @@ static vm_fault_t do_anonymous_page(struct vm_fault *vmf) vmf->pte = pte_offset_map_lock(vma->vm_mm, vmf->pmd, vmf->address, &vmf->ptl); if (!pte_none(*vmf->pte)) { - update_mmu_cache(vma, vmf->address, vmf->pte); + update_mmu_tlb(vma, vmf->address, vmf->pte); goto release; } From 14c2ac36811b82479b1138383b2c9ff1ab6ba47d Mon Sep 17 00:00:00 2001 From: Qi Zheng Date: Thu, 29 Sep 2022 19:23:18 +0800 Subject: [PATCH 090/121] LoongArch: update local TLB if PTE entry exists Currently, the implementation of update_mmu_tlb() is empty if __HAVE_ARCH_UPDATE_MMU_TLB is not defined. Then if two threads concurrently fault at the same page, the second thread that did not win the race will give up and do nothing. In the LoongArch architecture, this second thread will trigger another fault, and only updates its local TLB. Instead of triggering another fault, it's better to implement update_mmu_tlb() to directly update the local TLB of the second thread. Just do it. Link: https://lkml.kernel.org/r/20220929112318.32393-3-zhengqi.arch@bytedance.com Signed-off-by: Qi Zheng Suggested-by: Bibo Mao Acked-by: Huacai Chen Cc: Chris Zankel Cc: David Hildenbrand Cc: Max Filippov Cc: Muchun Song Signed-off-by: Andrew Morton --- arch/loongarch/include/asm/pgtable.h | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/loongarch/include/asm/pgtable.h b/arch/loongarch/include/asm/pgtable.h index 8ea57e2f0e04..946704bee599 100644 --- a/arch/loongarch/include/asm/pgtable.h +++ b/arch/loongarch/include/asm/pgtable.h @@ -412,6 +412,9 @@ static inline void update_mmu_cache(struct vm_area_struct *vma, __update_tlb(vma, address, ptep); } +#define __HAVE_ARCH_UPDATE_MMU_TLB +#define update_mmu_tlb update_mmu_cache + static inline void update_mmu_cache_pmd(struct vm_area_struct *vma, unsigned long address, pmd_t *pmdp) { From 94541bc3fbde45bbd40e7989995246b22732679a Mon Sep 17 00:00:00 2001 From: Brian Geffon Date: Mon, 3 Oct 2022 10:48:32 -0400 Subject: [PATCH 091/121] zram: always expose rw_page Currently zram will adjust its fops to a version which does not contain rw_page when a backing device has been assigned. This is done to prevent upper layers from assuming a synchronous operation when a page may have been written back. This forces every operation through bio which has overhead associated with bio_alloc/frees. The code can be simplified to always expose an rw_page method and only in the rare event that a page is written back we instead will return -EOPNOTSUPP forcing the upper layer to fallback to bio. Link: https://lkml.kernel.org/r/20221003144832.2906610-1-bgeffon@google.com Signed-off-by: Brian Geffon Reviewed-by: Sergey Senozhatsky Cc: Minchan Kim Cc: Nitin Gupta Cc: Rom Lemarchand Cc: Suleiman Souhlal Signed-off-by: Andrew Morton --- drivers/block/zram/zram_drv.c | 26 +++----------------------- 1 file changed, 3 insertions(+), 23 deletions(-) diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c index 43eeef2b9fbe..2ba5c98319e5 100644 --- a/drivers/block/zram/zram_drv.c +++ b/drivers/block/zram/zram_drv.c @@ -52,9 +52,6 @@ static unsigned int num_devices = 1; static size_t huge_class_size; static const struct block_device_operations zram_devops; -#ifdef CONFIG_ZRAM_WRITEBACK -static const struct block_device_operations zram_wb_devops; -#endif static void zram_free_page(struct zram *zram, size_t index); static int zram_bvec_read(struct zram *zram, struct bio_vec *bvec, @@ -546,17 +543,6 @@ static ssize_t backing_dev_store(struct device *dev, zram->backing_dev = backing_dev; zram->bitmap = bitmap; zram->nr_pages = nr_pages; - /* - * With writeback feature, zram does asynchronous IO so it's no longer - * synchronous device so let's remove synchronous io flag. Othewise, - * upper layer(e.g., swap) could wait IO completion rather than - * (submit and return), which will cause system sluggish. - * Furthermore, when the IO function returns(e.g., swap_readpage), - * upper layer expects IO was done so it could deallocate the page - * freely but in fact, IO is going on so finally could cause - * use-after-free when the IO is really done. - */ - zram->disk->fops = &zram_wb_devops; up_write(&zram->init_lock); pr_info("setup backing device %s\n", file_name); @@ -1270,6 +1256,9 @@ static int __zram_bvec_read(struct zram *zram, struct page *page, u32 index, struct bio_vec bvec; zram_slot_unlock(zram, index); + /* A null bio means rw_page was used, we must fallback to bio */ + if (!bio) + return -EOPNOTSUPP; bvec.bv_page = page; bvec.bv_len = PAGE_SIZE; @@ -1856,15 +1845,6 @@ static const struct block_device_operations zram_devops = { .owner = THIS_MODULE }; -#ifdef CONFIG_ZRAM_WRITEBACK -static const struct block_device_operations zram_wb_devops = { - .open = zram_open, - .submit_bio = zram_submit_bio, - .swap_slot_free_notify = zram_slot_free_notify, - .owner = THIS_MODULE -}; -#endif - static DEVICE_ATTR_WO(compact); static DEVICE_ATTR_RW(disksize); static DEVICE_ATTR_RO(initstate); From 2ea7ff1e39cbe3753d3c649beb70f2cf861dca75 Mon Sep 17 00:00:00 2001 From: Peter Xu Date: Tue, 4 Oct 2022 15:33:58 -0400 Subject: [PATCH 092/121] mm/hugetlb: fix race condition of uffd missing/minor handling Patch series "mm/hugetlb: Fix selftest failures with write check", v3. Currently akpm mm-unstable fails with uffd hugetlb private mapping test randomly on a write check. The initial bisection of that points to the recent pmd unshare series, but it turns out there's no direction relationship with the series but only some timing change caused the race to start trigger. The race should be fixed in patch 1. Patch 2 is a trivial cleanup on the similar race with hugetlb migrations, patch 3 comment on the write check so when anyone read it again it'll be clear why it's there. This patch (of 3): After the recent rework patchset of hugetlb locking on pmd sharing, kselftest for userfaultfd sometimes fails on hugetlb private tests with unexpected write fault checks. It turns out there's nothing wrong within the locking series regarding this matter, but it could have changed the timing of threads so it can trigger an old bug. The real bug is when we call hugetlb_no_page() we're not with the pgtable lock. It means we're reading the pte values lockless. It's perfectly fine in most cases because before we do normal page allocations we'll take the lock and check pte_same() again. However before that, there are actually two paths on userfaultfd missing/minor handling that may directly move on with the fault process without checking the pte values. It means for these two paths we may be generating an uffd message based on an unstable pte, while an unstable pte can legally be anything as long as the modifier holds the pgtable lock. One example, which is also what happened in the failing kselftest and caused the test failure, is that for private mappings wr-protection changes can happen on one page. While hugetlb_change_protection() generally requires pte being cleared before being changed, then there can be a race condition like: thread 1 thread 2 -------- -------- UFFDIO_WRITEPROTECT hugetlb_fault hugetlb_change_protection pgtable_lock() huge_ptep_modify_prot_start pte==NULL hugetlb_no_page generate uffd missing event even if page existed!! huge_ptep_modify_prot_commit pgtable_unlock() Fix this by rechecking the pte after pgtable lock for both userfaultfd missing & minor fault paths. This bug should have been around starting from uffd hugetlb introduced, so attaching a Fixes to the commit. Also attach another Fixes to the minor support commit for easier tracking. Note that userfaultfd is actually fine with false positives (e.g. caused by pte changed), but not wrong logical events (e.g. caused by reading a pte during changing). The latter can confuse the userspace, so the strictness is very much preferred. E.g., MISSING event should never happen on the page after UFFDIO_COPY has correctly installed the page and returned. Link: https://lkml.kernel.org/r/20221004193400.110155-1-peterx@redhat.com Link: https://lkml.kernel.org/r/20221004193400.110155-2-peterx@redhat.com Fixes: 1a1aad8a9b7b ("userfaultfd: hugetlbfs: add userfaultfd hugetlb hook") Fixes: 7677f7fd8be7 ("userfaultfd: add minor fault registration mode") Signed-off-by: Peter Xu Co-developed-by: Mike Kravetz Reviewed-by: Mike Kravetz Cc: Andrea Arcangeli Cc: Axel Rasmussen Cc: Nadav Amit Cc: David Hildenbrand Cc: Mike Rapoport Signed-off-by: Andrew Morton --- mm/hugetlb.c | 59 +++++++++++++++++++++++++++++++++++++++++++++------- 1 file changed, 52 insertions(+), 7 deletions(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 9a910612336d..bf9d8d04bf4f 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -5535,6 +5535,23 @@ static inline vm_fault_t hugetlb_handle_userfault(struct vm_area_struct *vma, return handle_userfault(&vmf, reason); } +/* + * Recheck pte with pgtable lock. Returns true if pte didn't change, or + * false if pte changed or is changing. + */ +static bool hugetlb_pte_stable(struct hstate *h, struct mm_struct *mm, + pte_t *ptep, pte_t old_pte) +{ + spinlock_t *ptl; + bool same; + + ptl = huge_pte_lock(h, mm, ptep); + same = pte_same(huge_ptep_get(ptep), old_pte); + spin_unlock(ptl); + + return same; +} + static vm_fault_t hugetlb_no_page(struct mm_struct *mm, struct vm_area_struct *vma, struct address_space *mapping, pgoff_t idx, @@ -5575,10 +5592,33 @@ static vm_fault_t hugetlb_no_page(struct mm_struct *mm, if (idx >= size) goto out; /* Check for page in userfault range */ - if (userfaultfd_missing(vma)) - return hugetlb_handle_userfault(vma, mapping, idx, - flags, haddr, address, - VM_UFFD_MISSING); + if (userfaultfd_missing(vma)) { + /* + * Since hugetlb_no_page() was examining pte + * without pgtable lock, we need to re-test under + * lock because the pte may not be stable and could + * have changed from under us. Try to detect + * either changed or during-changing ptes and retry + * properly when needed. + * + * Note that userfaultfd is actually fine with + * false positives (e.g. caused by pte changed), + * but not wrong logical events (e.g. caused by + * reading a pte during changing). The latter can + * confuse the userspace, so the strictness is very + * much preferred. E.g., MISSING event should + * never happen on the page after UFFDIO_COPY has + * correctly installed the page and returned. + */ + if (!hugetlb_pte_stable(h, mm, ptep, old_pte)) { + ret = 0; + goto out; + } + + return hugetlb_handle_userfault(vma, mapping, idx, flags, + haddr, address, + VM_UFFD_MISSING); + } page = alloc_huge_page(vma, haddr, 0); if (IS_ERR(page)) { @@ -5644,9 +5684,14 @@ static vm_fault_t hugetlb_no_page(struct mm_struct *mm, if (userfaultfd_minor(vma)) { unlock_page(page); put_page(page); - return hugetlb_handle_userfault(vma, mapping, idx, - flags, haddr, address, - VM_UFFD_MINOR); + /* See comment in userfaultfd_missing() block above */ + if (!hugetlb_pte_stable(h, mm, ptep, old_pte)) { + ret = 0; + goto out; + } + return hugetlb_handle_userfault(vma, mapping, idx, flags, + haddr, address, + VM_UFFD_MINOR); } } From f9bf6c03eca1077cae8de0e6d86427656fa42a9b Mon Sep 17 00:00:00 2001 From: Peter Xu Date: Tue, 4 Oct 2022 15:33:59 -0400 Subject: [PATCH 093/121] mm/hugetlb: use hugetlb_pte_stable in migration race check After hugetlb_pte_stable() introduced, we can also rewrite the migration race condition against page allocation to use the new helper too. Link: https://lkml.kernel.org/r/20221004193400.110155-3-peterx@redhat.com Signed-off-by: Peter Xu Reviewed-by: Mike Kravetz Reviewed-by: David Hildenbrand Cc: Andrea Arcangeli Cc: Axel Rasmussen Cc: Mike Rapoport Cc: Nadav Amit Signed-off-by: Andrew Morton --- mm/hugetlb.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index bf9d8d04bf4f..9b26055f3119 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -5634,11 +5634,10 @@ static vm_fault_t hugetlb_no_page(struct mm_struct *mm, * here. Before returning error, get ptl and make * sure there really is no pte entry. */ - ptl = huge_pte_lock(h, mm, ptep); - ret = 0; - if (huge_pte_none(huge_ptep_get(ptep))) + if (hugetlb_pte_stable(h, mm, ptep, old_pte)) ret = vmf_error(PTR_ERR(page)); - spin_unlock(ptl); + else + ret = 0; goto out; } clear_huge_page(page, address, pages_per_huge_page(h)); From 26c92d37d3dc484157bdb4eb7d29991c017b168b Mon Sep 17 00:00:00 2001 From: Peter Xu Date: Tue, 4 Oct 2022 15:34:00 -0400 Subject: [PATCH 094/121] mm/selftest: uffd: explain the write missing fault check It's not obvious why we had a write check for each of the missing messages, especially when it should be a locking op. Add a rich comment for that, and also try to explain its good side and limitations, so that if someone hit it again for either a bug or a different glibc impl there'll be some clue to start with. Link: https://lkml.kernel.org/r/20221004193400.110155-4-peterx@redhat.com Signed-off-by: Peter Xu Reviewed-by: Mike Kravetz Reviewed-by: David Hildenbrand Cc: Andrea Arcangeli Cc: Axel Rasmussen Cc: Mike Rapoport Cc: Nadav Amit Signed-off-by: Andrew Morton --- tools/testing/selftests/vm/userfaultfd.c | 22 +++++++++++++++++++++- 1 file changed, 21 insertions(+), 1 deletion(-) diff --git a/tools/testing/selftests/vm/userfaultfd.c b/tools/testing/selftests/vm/userfaultfd.c index 74babdbc02e5..297f250c1d95 100644 --- a/tools/testing/selftests/vm/userfaultfd.c +++ b/tools/testing/selftests/vm/userfaultfd.c @@ -774,7 +774,27 @@ static void uffd_handle_page_fault(struct uffd_msg *msg, continue_range(uffd, msg->arg.pagefault.address, page_size); stats->minor_faults++; } else { - /* Missing page faults */ + /* + * Missing page faults. + * + * Here we force a write check for each of the missing mode + * faults. It's guaranteed because the only threads that + * will trigger uffd faults are the locking threads, and + * their first instruction to touch the missing page will + * always be pthread_mutex_lock(). + * + * Note that here we relied on an NPTL glibc impl detail to + * always read the lock type at the entry of the lock op + * (pthread_mutex_t.__data.__type, offset 0x10) before + * doing any locking operations to guarantee that. It's + * actually not good to rely on this impl detail because + * logically a pthread-compatible lib can implement the + * locks without types and we can fail when linking with + * them. However since we used to find bugs with this + * strict check we still keep it around. Hopefully this + * could be a good hint when it fails again. If one day + * it'll break on some other impl of glibc we'll revisit. + */ if (msg->arg.pagefault.flags & UFFD_PAGEFAULT_FLAG_WRITE) err("unexpected write fault"); From 15cd90049d595e592d8860ee15a3f23491d54d17 Mon Sep 17 00:00:00 2001 From: Yafang Shao Date: Thu, 6 Oct 2022 10:15:40 +0000 Subject: [PATCH 095/121] mm/page_alloc: fix incorrect PGFREE and PGALLOC for high-order page PGFREE and PGALLOC represent the number of freed and allocated pages. So the page order must be considered. Link: https://lkml.kernel.org/r/20221006101540.40686-1-laoar.shao@gmail.com Fixes: 44042b449872 ("mm/page_alloc: allow high-order pages to be stored on the per-cpu lists") Signed-off-by: Yafang Shao Acked-by: Mel Gorman Reviewed-by: Miaohe Lin Signed-off-by: Andrew Morton --- mm/page_alloc.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 059f6946832f..8e9b7f08a32c 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3446,7 +3446,7 @@ static void free_unref_page_commit(struct zone *zone, struct per_cpu_pages *pcp, int pindex; bool free_high; - __count_vm_event(PGFREE); + __count_vm_events(PGFREE, 1 << order); pindex = order_to_pindex(migratetype, order); list_add(&page->pcp_list, &pcp->lists[pindex]); pcp->count += 1 << order; @@ -3803,7 +3803,7 @@ static struct page *rmqueue_pcplist(struct zone *preferred_zone, pcp_spin_unlock_irqrestore(pcp, flags); pcp_trylock_finish(UP_flags); if (page) { - __count_zid_vm_events(PGALLOC, page_zonenum(page), 1); + __count_zid_vm_events(PGALLOC, page_zonenum(page), 1 << order); zone_statistics(preferred_zone, zone, 1); } return page; From ef6e06b2ef87077104d1145a0fd452ff8dbbc4b7 Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 5 Oct 2022 21:05:55 -0700 Subject: [PATCH 096/121] highmem: fix kmap_to_page() for kmap_local_page() addresses kmap_to_page() is used to get the page for a virtual address which may be kmap'ed. Unfortunately, kmap_local_page() stores mappings in a thread local array separate from kmap(). These mappings were not checked by the call. Check the kmap_local_page() mappings and return the page if found. Because it is intended to remove kmap_to_page() add a warn on once to the kmap checks to flag potential issues early. NOTE Due to 32bit x86 use of kmap local in iomap atmoic, KMAP_LOCAL does not require HIGHMEM to be set. Therefore the support calls required a new KMAP_LOCAL section to fix 0day build errors. [akpm@linux-foundation.org: fix warning] Link: https://lkml.kernel.org/r/20221006040555.1502679-1-ira.weiny@intel.com Signed-off-by: Ira Weiny Reported-by: Al Viro Reported-by: kernel test robot Cc: "Fabio M. De Francesco" Cc: Thomas Gleixner Cc: Christoph Hellwig Signed-off-by: Andrew Morton --- mm/highmem.c | 43 +++++++++++++++++++++++++++++++------------ 1 file changed, 31 insertions(+), 12 deletions(-) diff --git a/mm/highmem.c b/mm/highmem.c index c707d7202d5f..db251e77f98f 100644 --- a/mm/highmem.c +++ b/mm/highmem.c @@ -30,6 +30,17 @@ #include #include +#ifdef CONFIG_KMAP_LOCAL +static inline int kmap_local_calc_idx(int idx) +{ + return idx + KM_MAX_IDX * smp_processor_id(); +} + +#ifndef arch_kmap_local_map_idx +#define arch_kmap_local_map_idx(idx, pfn) kmap_local_calc_idx(idx) +#endif +#endif /* CONFIG_KMAP_LOCAL */ + /* * Virtual_count is not a pure "count". * 0 means that it is not mapped, and has not been mapped @@ -142,12 +153,29 @@ pte_t *pkmap_page_table; struct page *__kmap_to_page(void *vaddr) { + unsigned long base = (unsigned long) vaddr & PAGE_MASK; + struct kmap_ctrl *kctrl = ¤t->kmap_ctrl; unsigned long addr = (unsigned long)vaddr; + int i; - if (addr >= PKMAP_ADDR(0) && addr < PKMAP_ADDR(LAST_PKMAP)) { - int i = PKMAP_NR(addr); + /* kmap() mappings */ + if (WARN_ON_ONCE(addr >= PKMAP_ADDR(0) && + addr < PKMAP_ADDR(LAST_PKMAP))) + return pte_page(pkmap_page_table[PKMAP_NR(addr)]); - return pte_page(pkmap_page_table[i]); + /* kmap_local_page() mappings */ + if (WARN_ON_ONCE(base >= __fix_to_virt(FIX_KMAP_END) && + base < __fix_to_virt(FIX_KMAP_BEGIN))) { + for (i = 0; i < kctrl->idx; i++) { + unsigned long base_addr; + int idx; + + idx = arch_kmap_local_map_idx(i, pte_pfn(pteval)); + base_addr = __fix_to_virt(FIX_KMAP_BEGIN + idx); + + if (base_addr == base) + return pte_page(kctrl->pteval[i]); + } } return virt_to_page(vaddr); @@ -462,10 +490,6 @@ static inline void kmap_local_idx_pop(void) # define arch_kmap_local_post_unmap(vaddr) do { } while (0) #endif -#ifndef arch_kmap_local_map_idx -#define arch_kmap_local_map_idx(idx, pfn) kmap_local_calc_idx(idx) -#endif - #ifndef arch_kmap_local_unmap_idx #define arch_kmap_local_unmap_idx(idx, vaddr) kmap_local_calc_idx(idx) #endif @@ -494,11 +518,6 @@ static inline bool kmap_high_unmap_local(unsigned long vaddr) return false; } -static inline int kmap_local_calc_idx(int idx) -{ - return idx + KM_MAX_IDX * smp_processor_id(); -} - static pte_t *__kmap_pte; static pte_t *kmap_get_pte(unsigned long vaddr, int idx) From f2913d006fcdb61719635e093d1b5dd0dafecac7 Mon Sep 17 00:00:00 2001 From: Palmer Dabbelt Date: Tue, 20 Sep 2022 13:00:37 -0700 Subject: [PATCH 097/121] RISC-V: Avoid dereferening NULL regs in die() I don't think we can actually die() without a regs pointer, but the compiler was warning about a NULL check after a dereference. It seems prudent to just avoid the possibly-NULL dereference, given that when die()ing the system is already toast so who knows how we got there. Reported-by: kernel test robot Reported-by: Dan Carpenter Reviewed-by: Conor Dooley Link: https://lore.kernel.org/r/20220920200037.6727-1-palmer@rivosinc.com Signed-off-by: Palmer Dabbelt --- arch/riscv/kernel/traps.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/arch/riscv/kernel/traps.c b/arch/riscv/kernel/traps.c index 635e6ec26938..f3e96d60a2ff 100644 --- a/arch/riscv/kernel/traps.c +++ b/arch/riscv/kernel/traps.c @@ -33,6 +33,7 @@ void die(struct pt_regs *regs, const char *str) { static int die_counter; int ret; + long cause; oops_enter(); @@ -42,11 +43,13 @@ void die(struct pt_regs *regs, const char *str) pr_emerg("%s [#%d]\n", str, ++die_counter); print_modules(); - show_regs(regs); + if (regs) + show_regs(regs); - ret = notify_die(DIE_OOPS, str, regs, 0, regs->cause, SIGSEGV); + cause = regs ? regs->cause : -1; + ret = notify_die(DIE_OOPS, str, regs, 0, cause, SIGSEGV); - if (regs && kexec_should_crash(current)) + if (kexec_should_crash(current)) crash_kexec(regs); bust_spinlocks(0); From a4cb3651a174366cc85a677da9e3681fbe97fdae Mon Sep 17 00:00:00 2001 From: Nicholas Piggin Date: Thu, 13 Oct 2022 16:44:18 +1000 Subject: [PATCH 098/121] powerpc/64s/interrupt: Fix lost interrupts when returning to soft-masked context It's possible for an interrupt returning to an irqs-disabled context to lose a pending soft-masked irq because it branches to part of the exit code for irqs-enabled contexts, which is meant to clear only the PACA_IRQS_HARD_DIS flag from PACAIRQHAPPENED by zeroing the byte. This just looks like a simple thinko from a recent commit (if there was no hard mask pending, there would be no reason to clear it anyway). This also adds comment to the code that actually does need to clear the flag. Fixes: e485f6c751e0a ("powerpc/64/interrupt: Fix return to masked context after hard-mask irq becomes pending") Reported-by: Sachin Sant Reported-by: Guenter Roeck Signed-off-by: Nicholas Piggin Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20221013064418.1311104-1-npiggin@gmail.com --- arch/powerpc/kernel/interrupt_64.S | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/kernel/interrupt_64.S b/arch/powerpc/kernel/interrupt_64.S index 904a5608cbe3..978a173eb339 100644 --- a/arch/powerpc/kernel/interrupt_64.S +++ b/arch/powerpc/kernel/interrupt_64.S @@ -538,7 +538,7 @@ _ASM_NOKPROBE_SYMBOL(interrupt_return_\srr\()_kernel) beq .Lfast_kernel_interrupt_return_\srr\() // EE already disabled lbz r11,PACAIRQHAPPENED(r13) andi. r10,r11,PACA_IRQ_MUST_HARD_MASK - beq 1f // No HARD_MASK pending + beq .Lfast_kernel_interrupt_return_\srr\() // No HARD_MASK pending /* Must clear MSR_EE from _MSR */ #ifdef CONFIG_PPC_BOOK3S @@ -555,12 +555,23 @@ _ASM_NOKPROBE_SYMBOL(interrupt_return_\srr\()_kernel) b .Lfast_kernel_interrupt_return_\srr\() .Linterrupt_return_\srr\()_soft_enabled: + /* + * In the soft-enabled case, need to double-check that we have no + * pending interrupts that might have come in before we reached the + * restart section of code, and restart the exit so those can be + * handled. + * + * If there are none, it is be possible that the interrupt still + * has PACA_IRQ_HARD_DIS set, which needs to be cleared for the + * interrupted context. This clear will not clobber a new pending + * interrupt coming in, because we're in the restart section, so + * such would return to the restart location. + */ #ifdef CONFIG_PPC_BOOK3S lbz r11,PACAIRQHAPPENED(r13) andi. r11,r11,(~PACA_IRQ_HARD_DIS)@l bne- interrupt_return_\srr\()_kernel_restart #endif -1: li r11,0 stb r11,PACAIRQHAPPENED(r13) // clear the possible HARD_DIS From 90d5ce82e143b42b2fdfb95401a89f86b71cedb7 Mon Sep 17 00:00:00 2001 From: Nicholas Piggin Date: Thu, 13 Oct 2022 17:31:31 +1000 Subject: [PATCH 099/121] powerpc/pseries: Fix CONFIG_DTL=n build The recently moved dtl code must be compiled-in if CONFIG_VIRT_CPU_ACCOUNTING_NATIVE=y even if CONFIG_DTL=n. Fixes: 6ba5aa541aaa0 ("powerpc/pseries: Move dtl scanning and steal time accounting to pseries platform") Reported-by: Guenter Roeck Signed-off-by: Nicholas Piggin Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20221013073131.1485742-1-npiggin@gmail.com --- arch/powerpc/platforms/pseries/Makefile | 3 +- arch/powerpc/platforms/pseries/dtl.c | 151 +++++++++++++----------- 2 files changed, 80 insertions(+), 74 deletions(-) diff --git a/arch/powerpc/platforms/pseries/Makefile b/arch/powerpc/platforms/pseries/Makefile index 14e143b946a3..92310202bdd7 100644 --- a/arch/powerpc/platforms/pseries/Makefile +++ b/arch/powerpc/platforms/pseries/Makefile @@ -7,7 +7,7 @@ obj-y := lpar.o hvCall.o nvram.o reconfig.o \ setup.o iommu.o event_sources.o ras.o \ firmware.o power.o dlpar.o mobility.o rng.o \ pci.o pci_dlpar.o eeh_pseries.o msi.o \ - papr_platform_attributes.o + papr_platform_attributes.o dtl.o obj-$(CONFIG_SMP) += smp.o obj-$(CONFIG_KEXEC_CORE) += kexec.o obj-$(CONFIG_PSERIES_ENERGY) += pseries_energy.o @@ -19,7 +19,6 @@ obj-$(CONFIG_HVC_CONSOLE) += hvconsole.o obj-$(CONFIG_HVCS) += hvcserver.o obj-$(CONFIG_HCALL_STATS) += hvCall_inst.o obj-$(CONFIG_CMM) += cmm.o -obj-$(CONFIG_DTL) += dtl.o obj-$(CONFIG_IO_EVENT_IRQ) += io_event_irq.o obj-$(CONFIG_LPARCFG) += lparcfg.o obj-$(CONFIG_IBMVIO) += vio.o diff --git a/arch/powerpc/platforms/pseries/dtl.c b/arch/powerpc/platforms/pseries/dtl.c index 1b1977bc78e7..3f1cdccebc9c 100644 --- a/arch/powerpc/platforms/pseries/dtl.c +++ b/arch/powerpc/platforms/pseries/dtl.c @@ -18,6 +18,7 @@ #include #include +#ifdef CONFIG_DTL struct dtl { struct dtl_entry *buf; int cpu; @@ -57,78 +58,6 @@ static DEFINE_PER_CPU(struct dtl_ring, dtl_rings); static atomic_t dtl_count; -/* - * Scan the dispatch trace log and count up the stolen time. - * Should be called with interrupts disabled. - */ -static notrace u64 scan_dispatch_log(u64 stop_tb) -{ - u64 i = local_paca->dtl_ridx; - struct dtl_entry *dtl = local_paca->dtl_curr; - struct dtl_entry *dtl_end = local_paca->dispatch_log_end; - struct lppaca *vpa = local_paca->lppaca_ptr; - u64 tb_delta; - u64 stolen = 0; - u64 dtb; - - if (!dtl) - return 0; - - if (i == be64_to_cpu(vpa->dtl_idx)) - return 0; - while (i < be64_to_cpu(vpa->dtl_idx)) { - dtb = be64_to_cpu(dtl->timebase); - tb_delta = be32_to_cpu(dtl->enqueue_to_dispatch_time) + - be32_to_cpu(dtl->ready_to_enqueue_time); - barrier(); - if (i + N_DISPATCH_LOG < be64_to_cpu(vpa->dtl_idx)) { - /* buffer has overflowed */ - i = be64_to_cpu(vpa->dtl_idx) - N_DISPATCH_LOG; - dtl = local_paca->dispatch_log + (i % N_DISPATCH_LOG); - continue; - } - if (dtb > stop_tb) - break; - if (dtl_consumer) - dtl_consumer(dtl, i); - stolen += tb_delta; - ++i; - ++dtl; - if (dtl == dtl_end) - dtl = local_paca->dispatch_log; - } - local_paca->dtl_ridx = i; - local_paca->dtl_curr = dtl; - return stolen; -} - -/* - * Accumulate stolen time by scanning the dispatch trace log. - * Called on entry from user mode. - */ -void notrace pseries_accumulate_stolen_time(void) -{ - u64 sst, ust; - struct cpu_accounting_data *acct = &local_paca->accounting; - - sst = scan_dispatch_log(acct->starttime_user); - ust = scan_dispatch_log(acct->starttime); - acct->stime -= sst; - acct->utime -= ust; - acct->steal_time += ust + sst; -} - -u64 pseries_calculate_stolen_time(u64 stop_tb) -{ - if (!firmware_has_feature(FW_FEATURE_SPLPAR)) - return 0; - - if (get_paca()->dtl_ridx != be64_to_cpu(get_lppaca()->dtl_idx)) - return scan_dispatch_log(stop_tb); - - return 0; -} - /* * The cpu accounting code controls the DTL ring buffer, and we get * given entries as they are processed. @@ -436,3 +365,81 @@ static int dtl_init(void) return 0; } machine_arch_initcall(pseries, dtl_init); +#endif /* CONFIG_DTL */ + +#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE +/* + * Scan the dispatch trace log and count up the stolen time. + * Should be called with interrupts disabled. + */ +static notrace u64 scan_dispatch_log(u64 stop_tb) +{ + u64 i = local_paca->dtl_ridx; + struct dtl_entry *dtl = local_paca->dtl_curr; + struct dtl_entry *dtl_end = local_paca->dispatch_log_end; + struct lppaca *vpa = local_paca->lppaca_ptr; + u64 tb_delta; + u64 stolen = 0; + u64 dtb; + + if (!dtl) + return 0; + + if (i == be64_to_cpu(vpa->dtl_idx)) + return 0; + while (i < be64_to_cpu(vpa->dtl_idx)) { + dtb = be64_to_cpu(dtl->timebase); + tb_delta = be32_to_cpu(dtl->enqueue_to_dispatch_time) + + be32_to_cpu(dtl->ready_to_enqueue_time); + barrier(); + if (i + N_DISPATCH_LOG < be64_to_cpu(vpa->dtl_idx)) { + /* buffer has overflowed */ + i = be64_to_cpu(vpa->dtl_idx) - N_DISPATCH_LOG; + dtl = local_paca->dispatch_log + (i % N_DISPATCH_LOG); + continue; + } + if (dtb > stop_tb) + break; +#ifdef CONFIG_DTL + if (dtl_consumer) + dtl_consumer(dtl, i); +#endif + stolen += tb_delta; + ++i; + ++dtl; + if (dtl == dtl_end) + dtl = local_paca->dispatch_log; + } + local_paca->dtl_ridx = i; + local_paca->dtl_curr = dtl; + return stolen; +} + +/* + * Accumulate stolen time by scanning the dispatch trace log. + * Called on entry from user mode. + */ +void notrace pseries_accumulate_stolen_time(void) +{ + u64 sst, ust; + struct cpu_accounting_data *acct = &local_paca->accounting; + + sst = scan_dispatch_log(acct->starttime_user); + ust = scan_dispatch_log(acct->starttime); + acct->stime -= sst; + acct->utime -= ust; + acct->steal_time += ust + sst; +} + +u64 pseries_calculate_stolen_time(u64 stop_tb) +{ + if (!firmware_has_feature(FW_FEATURE_SPLPAR)) + return 0; + + if (get_paca()->dtl_ridx != be64_to_cpu(get_lppaca()->dtl_idx)) + return scan_dispatch_log(stop_tb); + + return 0; +} + +#endif From 3cebf80e9a0d3adcb174053be32c88a640b3344b Mon Sep 17 00:00:00 2001 From: Fangrui Song Date: Sun, 18 Sep 2022 02:29:34 -0700 Subject: [PATCH 100/121] riscv: Pass -mno-relax only on lld < 15.0.0 lld since llvm:6611d58f5bbc ("[ELF] Relax R_RISCV_ALIGN"), which will be included in the 15.0.0 release, has implemented some RISC-V linker relaxation. -mno-relax is no longer needed in KBUILD_CFLAGS/KBUILD_AFLAGS to suppress R_RISCV_ALIGN which older lld can not handle: ld.lld: error: capability.c:(.fixup+0x0): relocation R_RISCV_ALIGN requires unimplemented linker relaxation; recompile with -mno-relax but the .o is already compiled with -mno-relax Signed-off-by: Fangrui Song Link: https://lore.kernel.org/r/20220710071117.446112-1-maskray@google.com/ Link: https://lore.kernel.org/r/20220918092933.19943-1-palmer@rivosinc.com Reviewed-by: Nick Desaulniers Tested-by: Nick Desaulniers Tested-by: Nathan Chancellor Tested-by: Conor Dooley Cc: stable@vger.kernel.org Signed-off-by: Palmer Dabbelt --- arch/riscv/Makefile | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/riscv/Makefile b/arch/riscv/Makefile index d63295e21373..76364cf67a72 100644 --- a/arch/riscv/Makefile +++ b/arch/riscv/Makefile @@ -37,6 +37,7 @@ else endif ifeq ($(CONFIG_LD_IS_LLD),y) +ifeq ($(shell test $(CONFIG_LLD_VERSION) -lt 150000; echo $$?),0) KBUILD_CFLAGS += -mno-relax KBUILD_AFLAGS += -mno-relax ifndef CONFIG_AS_IS_LLVM @@ -44,6 +45,7 @@ ifndef CONFIG_AS_IS_LLVM KBUILD_AFLAGS += -Wa,-mno-relax endif endif +endif # ISA string setting riscv-march-$(CONFIG_ARCH_RV32I) := rv32ima From e47bddcb2ec531022a915f896f13586470b593d0 Mon Sep 17 00:00:00 2001 From: Heiko Stuebner Date: Mon, 5 Sep 2022 13:10:23 +0200 Subject: [PATCH 101/121] riscv: cleanup svpbmt cpufeature probing For better readability (and compile time coverage) use IS_ENABLED instead of ifdef and drop the new unneeded switch statement. Signed-off-by: Heiko Stuebner Reviewed-by: Guo Ren Reviewed-by: Conor Dooley Reviewed-by: Andrew Jones Link: https://lore.kernel.org/r/20220905111027.2463297-2-heiko@sntech.de Signed-off-by: Palmer Dabbelt --- arch/riscv/kernel/cpufeature.c | 13 +++++-------- 1 file changed, 5 insertions(+), 8 deletions(-) diff --git a/arch/riscv/kernel/cpufeature.c b/arch/riscv/kernel/cpufeature.c index 553d755483ed..764ea220161f 100644 --- a/arch/riscv/kernel/cpufeature.c +++ b/arch/riscv/kernel/cpufeature.c @@ -253,16 +253,13 @@ void __init riscv_fill_hwcap(void) #ifdef CONFIG_RISCV_ALTERNATIVE static bool __init_or_module cpufeature_probe_svpbmt(unsigned int stage) { -#ifdef CONFIG_RISCV_ISA_SVPBMT - switch (stage) { - case RISCV_ALTERNATIVES_EARLY_BOOT: + if (!IS_ENABLED(CONFIG_RISCV_ISA_SVPBMT)) return false; - default: - return riscv_isa_extension_available(NULL, SVPBMT); - } -#endif - return false; + if (stage == RISCV_ALTERNATIVES_EARLY_BOOT) + return false; + + return riscv_isa_extension_available(NULL, SVPBMT); } static bool __init_or_module cpufeature_probe_zicbom(unsigned int stage) From f055268e3946555deb9bb80b2c8c9798c64dbc47 Mon Sep 17 00:00:00 2001 From: Heiko Stuebner Date: Mon, 5 Sep 2022 13:10:24 +0200 Subject: [PATCH 102/121] riscv: drop some idefs from CMO initialization Wrapping things in #ifdefs makes the code harder to read while we also have IS_ENABLED() macros to do this in regular code and the extension detection is not _that_ runtime critical. So define a stub for riscv_noncoherent_supported() in the non-CONFIG_RISCV_DMA_NONCOHERENT case and move the code to us IS_ENABLED. Suggested-by: Conor Dooley Signed-off-by: Heiko Stuebner Reviewed-by: Guo Ren Reviewed-by: Conor Dooley Reviewed-by: Andrew Jones Link: https://lore.kernel.org/r/20220905111027.2463297-3-heiko@sntech.de Signed-off-by: Palmer Dabbelt --- arch/riscv/errata/thead/errata.c | 7 +++---- arch/riscv/include/asm/cacheflush.h | 2 ++ arch/riscv/kernel/cpufeature.c | 22 +++++++++------------- 3 files changed, 14 insertions(+), 17 deletions(-) diff --git a/arch/riscv/errata/thead/errata.c b/arch/riscv/errata/thead/errata.c index 202c83f677b2..bffa711aaf64 100644 --- a/arch/riscv/errata/thead/errata.c +++ b/arch/riscv/errata/thead/errata.c @@ -30,7 +30,9 @@ static bool errata_probe_pbmt(unsigned int stage, static bool errata_probe_cmo(unsigned int stage, unsigned long arch_id, unsigned long impid) { -#ifdef CONFIG_ERRATA_THEAD_CMO + if (!IS_ENABLED(CONFIG_ERRATA_THEAD_CMO)) + return false; + if (arch_id != 0 || impid != 0) return false; @@ -39,9 +41,6 @@ static bool errata_probe_cmo(unsigned int stage, riscv_noncoherent_supported(); return true; -#else - return false; -#endif } static u32 thead_errata_probe(unsigned int stage, diff --git a/arch/riscv/include/asm/cacheflush.h b/arch/riscv/include/asm/cacheflush.h index a60acaecfeda..4363d0beb38a 100644 --- a/arch/riscv/include/asm/cacheflush.h +++ b/arch/riscv/include/asm/cacheflush.h @@ -50,6 +50,8 @@ static inline void riscv_init_cbom_blocksize(void) { } #ifdef CONFIG_RISCV_DMA_NONCOHERENT void riscv_noncoherent_supported(void); +#else +static inline void riscv_noncoherent_supported(void) {} #endif /* diff --git a/arch/riscv/kernel/cpufeature.c b/arch/riscv/kernel/cpufeature.c index 764ea220161f..729f7a218093 100644 --- a/arch/riscv/kernel/cpufeature.c +++ b/arch/riscv/kernel/cpufeature.c @@ -264,21 +264,17 @@ static bool __init_or_module cpufeature_probe_svpbmt(unsigned int stage) static bool __init_or_module cpufeature_probe_zicbom(unsigned int stage) { -#ifdef CONFIG_RISCV_ISA_ZICBOM - switch (stage) { - case RISCV_ALTERNATIVES_EARLY_BOOT: + if (!IS_ENABLED(CONFIG_RISCV_ISA_ZICBOM)) return false; - default: - if (riscv_isa_extension_available(NULL, ZICBOM)) { - riscv_noncoherent_supported(); - return true; - } else { - return false; - } - } -#endif - return false; + if (stage == RISCV_ALTERNATIVES_EARLY_BOOT) + return false; + + if (!riscv_isa_extension_available(NULL, ZICBOM)) + return false; + + riscv_noncoherent_supported(); + return true; } /* From 499590c084f13b6aca225e5766edeebd48437ee8 Mon Sep 17 00:00:00 2001 From: Heiko Stuebner Date: Mon, 5 Sep 2022 13:10:25 +0200 Subject: [PATCH 103/121] riscv: use BIT() macros in t-head errata init Using the appropriate BIT macro makes the code better readable. Suggested-by: Conor Dooley Signed-off-by: Heiko Stuebner Reviewed-by: Guo Ren Reviewed-by: Conor Dooley Reviewed-by: Andrew Jones Link: https://lore.kernel.org/r/20220905111027.2463297-4-heiko@sntech.de Signed-off-by: Palmer Dabbelt --- arch/riscv/errata/thead/errata.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/riscv/errata/thead/errata.c b/arch/riscv/errata/thead/errata.c index bffa711aaf64..a6f4bd8ccf3f 100644 --- a/arch/riscv/errata/thead/errata.c +++ b/arch/riscv/errata/thead/errata.c @@ -49,10 +49,10 @@ static u32 thead_errata_probe(unsigned int stage, u32 cpu_req_errata = 0; if (errata_probe_pbmt(stage, archid, impid)) - cpu_req_errata |= (1U << ERRATA_THEAD_PBMT); + cpu_req_errata |= BIT(ERRATA_THEAD_PBMT); if (errata_probe_cmo(stage, archid, impid)) - cpu_req_errata |= (1U << ERRATA_THEAD_CMO); + cpu_req_errata |= BIT(ERRATA_THEAD_CMO); return cpu_req_errata; } From e283187c034cd80c1dd98ad732c73ce930a5efa4 Mon Sep 17 00:00:00 2001 From: Heiko Stuebner Date: Mon, 5 Sep 2022 13:10:26 +0200 Subject: [PATCH 104/121] riscv: use BIT() marco for cpufeature probing Using the appropriate BIT macro makes the code better readable. Suggested-by: Conor Dooley Signed-off-by: Heiko Stuebner Reviewed-by: Conor Dooley Link: https://lore.kernel.org/r/20220905111027.2463297-5-heiko@sntech.de Signed-off-by: Palmer Dabbelt --- arch/riscv/kernel/cpufeature.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/riscv/kernel/cpufeature.c b/arch/riscv/kernel/cpufeature.c index 729f7a218093..08f7445985dc 100644 --- a/arch/riscv/kernel/cpufeature.c +++ b/arch/riscv/kernel/cpufeature.c @@ -289,10 +289,10 @@ static u32 __init_or_module cpufeature_probe(unsigned int stage) u32 cpu_req_feature = 0; if (cpufeature_probe_svpbmt(stage)) - cpu_req_feature |= (1U << CPUFEATURE_SVPBMT); + cpu_req_feature |= BIT(CPUFEATURE_SVPBMT); if (cpufeature_probe_zicbom(stage)) - cpu_req_feature |= (1U << CPUFEATURE_ZICBOM); + cpu_req_feature |= BIT(CPUFEATURE_ZICBOM); return cpu_req_feature; } From 14057733109dcc83c35a6730f3b7112aac4d2b82 Mon Sep 17 00:00:00 2001 From: Heiko Stuebner Date: Mon, 5 Sep 2022 13:10:27 +0200 Subject: [PATCH 105/121] riscv: check for kernel config option in t-head memory types errata The t-head variant of page-based memory types should also check first for the enabled kernel config option. Fixes: a35707c3d850 ("riscv: add memory-type errata for T-Head") Signed-off-by: Heiko Stuebner Reviewed-by: Conor Dooley Reviewed-by: Andrew Jones Reviewed-by: Guo Ren Link: https://lore.kernel.org/r/20220905111027.2463297-6-heiko@sntech.de Signed-off-by: Palmer Dabbelt --- arch/riscv/errata/thead/errata.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/riscv/errata/thead/errata.c b/arch/riscv/errata/thead/errata.c index a6f4bd8ccf3f..902e12452821 100644 --- a/arch/riscv/errata/thead/errata.c +++ b/arch/riscv/errata/thead/errata.c @@ -17,6 +17,9 @@ static bool errata_probe_pbmt(unsigned int stage, unsigned long arch_id, unsigned long impid) { + if (!IS_ENABLED(CONFIG_ERRATA_THEAD_PBMT)) + return false; + if (arch_id != 0 || impid != 0) return false; From 44dce4b084f83f41922ed8c2a2c7d148254848bb Mon Sep 17 00:00:00 2001 From: Zong Li Date: Tue, 13 Sep 2022 06:18:11 +0000 Subject: [PATCH 106/121] dt-bindings: sifive-ccache: change Sifive L2 cache to Composable cache Since composable cache may be L3 cache if private L2 cache exists, we should use its original name Composable cache to prevent confusion. Signed-off-by: Zong Li Suggested-by: Conor Dooley Suggested-by: Ben Dooks Reviewed-by: Conor Dooley Reviewed-by: Rob Herring Link: https://lore.kernel.org/r/20220913061817.22564-2-zong.li@sifive.com Signed-off-by: Palmer Dabbelt --- ...five-l2-cache.yaml => sifive,ccache0.yaml} | 28 +++++++++++++++---- 1 file changed, 23 insertions(+), 5 deletions(-) rename Documentation/devicetree/bindings/riscv/{sifive-l2-cache.yaml => sifive,ccache0.yaml} (83%) diff --git a/Documentation/devicetree/bindings/riscv/sifive-l2-cache.yaml b/Documentation/devicetree/bindings/riscv/sifive,ccache0.yaml similarity index 83% rename from Documentation/devicetree/bindings/riscv/sifive-l2-cache.yaml rename to Documentation/devicetree/bindings/riscv/sifive,ccache0.yaml index ca3b9be58058..bf3f07421f7e 100644 --- a/Documentation/devicetree/bindings/riscv/sifive-l2-cache.yaml +++ b/Documentation/devicetree/bindings/riscv/sifive,ccache0.yaml @@ -2,18 +2,18 @@ # Copyright (C) 2020 SiFive, Inc. %YAML 1.2 --- -$id: http://devicetree.org/schemas/riscv/sifive-l2-cache.yaml# +$id: http://devicetree.org/schemas/riscv/sifive,ccache0.yaml# $schema: http://devicetree.org/meta-schemas/core.yaml# -title: SiFive L2 Cache Controller +title: SiFive Composable Cache Controller maintainers: - Sagar Kadam - Paul Walmsley description: - The SiFive Level 2 Cache Controller is used to provide access to fast copies - of memory for masters in a Core Complex. The Level 2 Cache Controller also + The SiFive Composable Cache Controller is used to provide access to fast copies + of memory for masters in a Core Complex. The Composable Cache Controller also acts as directory-based coherency manager. All the properties in ePAPR/DeviceTree specification applies for this platform. @@ -22,6 +22,7 @@ select: compatible: contains: enum: + - sifive,ccache0 - sifive,fu540-c000-ccache - sifive,fu740-c000-ccache @@ -33,6 +34,7 @@ properties: oneOf: - items: - enum: + - sifive,ccache0 - sifive,fu540-c000-ccache - sifive,fu740-c000-ccache - const: cache @@ -45,7 +47,7 @@ properties: const: 64 cache-level: - const: 2 + enum: [2, 3] cache-sets: enum: [1024, 2048] @@ -115,6 +117,22 @@ allOf: cache-sets: const: 1024 + - if: + properties: + compatible: + contains: + const: sifive,ccache0 + + then: + properties: + cache-level: + enum: [2, 3] + + else: + properties: + cache-level: + const: 2 + additionalProperties: false required: From ca120a79cf5a3323172c82e77efd70ae10d120ef Mon Sep 17 00:00:00 2001 From: Greentime Hu Date: Tue, 13 Sep 2022 06:18:12 +0000 Subject: [PATCH 107/121] soc: sifive: ccache: Rename SiFive L2 cache to Composable cache. Since composable cache may be L3 cache if there is a L2 cache, we should use its original name composable cache to prevent confusion. There are some new lines were generated due to adding the compatible "sifive,ccache0" into ID table and indent requirement. The sifive L2 has been renamed to sifive CCACHE, EDAC driver needs to apply the change as well. Signed-off-by: Greentime Hu Signed-off-by: Zong Li Co-developed-by: Zong Li Reviewed-by: Conor Dooley Link: https://lore.kernel.org/r/20220913061817.22564-3-zong.li@sifive.com Signed-off-by: Palmer Dabbelt --- drivers/edac/Kconfig | 2 +- drivers/edac/sifive_edac.c | 12 +- drivers/soc/sifive/Kconfig | 6 +- drivers/soc/sifive/Makefile | 2 +- drivers/soc/sifive/sifive_ccache.c | 245 +++++++++++++++++++++++++++ drivers/soc/sifive/sifive_l2_cache.c | 237 -------------------------- include/soc/sifive/sifive_ccache.h | 16 ++ include/soc/sifive/sifive_l2_cache.h | 16 -- 8 files changed, 272 insertions(+), 264 deletions(-) create mode 100644 drivers/soc/sifive/sifive_ccache.c delete mode 100644 drivers/soc/sifive/sifive_l2_cache.c create mode 100644 include/soc/sifive/sifive_ccache.h delete mode 100644 include/soc/sifive/sifive_l2_cache.h diff --git a/drivers/edac/Kconfig b/drivers/edac/Kconfig index 17562cf1fe97..456602d373b7 100644 --- a/drivers/edac/Kconfig +++ b/drivers/edac/Kconfig @@ -473,7 +473,7 @@ config EDAC_ALTERA_SDMMC config EDAC_SIFIVE bool "Sifive platform EDAC driver" - depends on EDAC=y && SIFIVE_L2 + depends on EDAC=y && SIFIVE_CCACHE help Support for error detection and correction on the SiFive SoCs. diff --git a/drivers/edac/sifive_edac.c b/drivers/edac/sifive_edac.c index ee800aec7d47..b844e2626fd5 100644 --- a/drivers/edac/sifive_edac.c +++ b/drivers/edac/sifive_edac.c @@ -2,7 +2,7 @@ /* * SiFive Platform EDAC Driver * - * Copyright (C) 2018-2019 SiFive, Inc. + * Copyright (C) 2018-2022 SiFive, Inc. * * This driver is partially based on octeon_edac-pc.c * @@ -10,7 +10,7 @@ #include #include #include "edac_module.h" -#include +#include #define DRVNAME "sifive_edac" @@ -32,9 +32,9 @@ int ecc_err_event(struct notifier_block *this, unsigned long event, void *ptr) p = container_of(this, struct sifive_edac_priv, notifier); - if (event == SIFIVE_L2_ERR_TYPE_UE) + if (event == SIFIVE_CCACHE_ERR_TYPE_UE) edac_device_handle_ue(p->dci, 0, 0, msg); - else if (event == SIFIVE_L2_ERR_TYPE_CE) + else if (event == SIFIVE_CCACHE_ERR_TYPE_CE) edac_device_handle_ce(p->dci, 0, 0, msg); return NOTIFY_OK; @@ -67,7 +67,7 @@ static int ecc_register(struct platform_device *pdev) goto err; } - register_sifive_l2_error_notifier(&p->notifier); + register_sifive_ccache_error_notifier(&p->notifier); return 0; @@ -81,7 +81,7 @@ static int ecc_unregister(struct platform_device *pdev) { struct sifive_edac_priv *p = platform_get_drvdata(pdev); - unregister_sifive_l2_error_notifier(&p->notifier); + unregister_sifive_ccache_error_notifier(&p->notifier); edac_device_del_device(&pdev->dev); edac_device_free_ctl_info(p->dci); diff --git a/drivers/soc/sifive/Kconfig b/drivers/soc/sifive/Kconfig index 58cf8c40d08d..ed4c571f8771 100644 --- a/drivers/soc/sifive/Kconfig +++ b/drivers/soc/sifive/Kconfig @@ -2,9 +2,9 @@ if SOC_SIFIVE -config SIFIVE_L2 - bool "Sifive L2 Cache controller" +config SIFIVE_CCACHE + bool "Sifive Composable Cache controller" help - Support for the L2 cache controller on SiFive platforms. + Support for the composable cache controller on SiFive platforms. endif diff --git a/drivers/soc/sifive/Makefile b/drivers/soc/sifive/Makefile index b5caff77938f..1f5dc339bf82 100644 --- a/drivers/soc/sifive/Makefile +++ b/drivers/soc/sifive/Makefile @@ -1,3 +1,3 @@ # SPDX-License-Identifier: GPL-2.0 -obj-$(CONFIG_SIFIVE_L2) += sifive_l2_cache.o +obj-$(CONFIG_SIFIVE_CCACHE) += sifive_ccache.o diff --git a/drivers/soc/sifive/sifive_ccache.c b/drivers/soc/sifive/sifive_ccache.c new file mode 100644 index 000000000000..949b824e89ad --- /dev/null +++ b/drivers/soc/sifive/sifive_ccache.c @@ -0,0 +1,245 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * SiFive composable cache controller Driver + * + * Copyright (C) 2018-2022 SiFive, Inc. + * + */ +#include +#include +#include +#include +#include +#include +#include + +#define SIFIVE_CCACHE_DIRECCFIX_LOW 0x100 +#define SIFIVE_CCACHE_DIRECCFIX_HIGH 0x104 +#define SIFIVE_CCACHE_DIRECCFIX_COUNT 0x108 + +#define SIFIVE_CCACHE_DIRECCFAIL_LOW 0x120 +#define SIFIVE_CCACHE_DIRECCFAIL_HIGH 0x124 +#define SIFIVE_CCACHE_DIRECCFAIL_COUNT 0x128 + +#define SIFIVE_CCACHE_DATECCFIX_LOW 0x140 +#define SIFIVE_CCACHE_DATECCFIX_HIGH 0x144 +#define SIFIVE_CCACHE_DATECCFIX_COUNT 0x148 + +#define SIFIVE_CCACHE_DATECCFAIL_LOW 0x160 +#define SIFIVE_CCACHE_DATECCFAIL_HIGH 0x164 +#define SIFIVE_CCACHE_DATECCFAIL_COUNT 0x168 + +#define SIFIVE_CCACHE_CONFIG 0x00 +#define SIFIVE_CCACHE_WAYENABLE 0x08 +#define SIFIVE_CCACHE_ECCINJECTERR 0x40 + +#define SIFIVE_CCACHE_MAX_ECCINTR 4 + +static void __iomem *ccache_base; +static int g_irq[SIFIVE_CCACHE_MAX_ECCINTR]; +static struct riscv_cacheinfo_ops ccache_cache_ops; + +enum { + DIR_CORR = 0, + DATA_CORR, + DATA_UNCORR, + DIR_UNCORR, +}; + +#ifdef CONFIG_DEBUG_FS +static struct dentry *sifive_test; + +static ssize_t ccache_write(struct file *file, const char __user *data, + size_t count, loff_t *ppos) +{ + unsigned int val; + + if (kstrtouint_from_user(data, count, 0, &val)) + return -EINVAL; + if ((val < 0xFF) || (val >= 0x10000 && val < 0x100FF)) + writel(val, ccache_base + SIFIVE_CCACHE_ECCINJECTERR); + else + return -EINVAL; + return count; +} + +static const struct file_operations ccache_fops = { + .owner = THIS_MODULE, + .open = simple_open, + .write = ccache_write +}; + +static void setup_sifive_debug(void) +{ + sifive_test = debugfs_create_dir("sifive_ccache_cache", NULL); + + debugfs_create_file("sifive_debug_inject_error", 0200, + sifive_test, NULL, &ccache_fops); +} +#endif + +static void ccache_config_read(void) +{ + u32 regval, val; + + regval = readl(ccache_base + SIFIVE_CCACHE_CONFIG); + val = regval & 0xFF; + pr_info("CCACHE: No. of Banks in the cache: %d\n", val); + val = (regval & 0xFF00) >> 8; + pr_info("CCACHE: No. of ways per bank: %d\n", val); + val = (regval & 0xFF0000) >> 16; + pr_info("CCACHE: Sets per bank: %llu\n", (uint64_t)1 << val); + val = (regval & 0xFF000000) >> 24; + pr_info("CCACHE: Bytes per cache block: %llu\n", (uint64_t)1 << val); + + regval = readl(ccache_base + SIFIVE_CCACHE_WAYENABLE); + pr_info("CCACHE: Index of the largest way enabled: %d\n", regval); +} + +static const struct of_device_id sifive_ccache_ids[] = { + { .compatible = "sifive,fu540-c000-ccache" }, + { .compatible = "sifive,fu740-c000-ccache" }, + { .compatible = "sifive,ccache0" }, + { /* end of table */ } +}; + +static ATOMIC_NOTIFIER_HEAD(ccache_err_chain); + +int register_sifive_ccache_error_notifier(struct notifier_block *nb) +{ + return atomic_notifier_chain_register(&ccache_err_chain, nb); +} +EXPORT_SYMBOL_GPL(register_sifive_ccache_error_notifier); + +int unregister_sifive_ccache_error_notifier(struct notifier_block *nb) +{ + return atomic_notifier_chain_unregister(&ccache_err_chain, nb); +} +EXPORT_SYMBOL_GPL(unregister_sifive_ccache_error_notifier); + +static int ccache_largest_wayenabled(void) +{ + return readl(ccache_base + SIFIVE_CCACHE_WAYENABLE) & 0xFF; +} + +static ssize_t number_of_ways_enabled_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + return sprintf(buf, "%u\n", ccache_largest_wayenabled()); +} + +static DEVICE_ATTR_RO(number_of_ways_enabled); + +static struct attribute *priv_attrs[] = { + &dev_attr_number_of_ways_enabled.attr, + NULL, +}; + +static const struct attribute_group priv_attr_group = { + .attrs = priv_attrs, +}; + +static const struct attribute_group *ccache_get_priv_group(struct cacheinfo + *this_leaf) +{ + /* We want to use private group for composable cache only */ + if (this_leaf->level == 2) + return &priv_attr_group; + else + return NULL; +} + +static irqreturn_t ccache_int_handler(int irq, void *device) +{ + unsigned int add_h, add_l; + + if (irq == g_irq[DIR_CORR]) { + add_h = readl(ccache_base + SIFIVE_CCACHE_DIRECCFIX_HIGH); + add_l = readl(ccache_base + SIFIVE_CCACHE_DIRECCFIX_LOW); + pr_err("CCACHE: DirError @ 0x%08X.%08X\n", add_h, add_l); + /* Reading this register clears the DirError interrupt sig */ + readl(ccache_base + SIFIVE_CCACHE_DIRECCFIX_COUNT); + atomic_notifier_call_chain(&ccache_err_chain, + SIFIVE_CCACHE_ERR_TYPE_CE, + "DirECCFix"); + } + if (irq == g_irq[DIR_UNCORR]) { + add_h = readl(ccache_base + SIFIVE_CCACHE_DIRECCFAIL_HIGH); + add_l = readl(ccache_base + SIFIVE_CCACHE_DIRECCFAIL_LOW); + /* Reading this register clears the DirFail interrupt sig */ + readl(ccache_base + SIFIVE_CCACHE_DIRECCFAIL_COUNT); + atomic_notifier_call_chain(&ccache_err_chain, + SIFIVE_CCACHE_ERR_TYPE_UE, + "DirECCFail"); + panic("CCACHE: DirFail @ 0x%08X.%08X\n", add_h, add_l); + } + if (irq == g_irq[DATA_CORR]) { + add_h = readl(ccache_base + SIFIVE_CCACHE_DATECCFIX_HIGH); + add_l = readl(ccache_base + SIFIVE_CCACHE_DATECCFIX_LOW); + pr_err("CCACHE: DataError @ 0x%08X.%08X\n", add_h, add_l); + /* Reading this register clears the DataError interrupt sig */ + readl(ccache_base + SIFIVE_CCACHE_DATECCFIX_COUNT); + atomic_notifier_call_chain(&ccache_err_chain, + SIFIVE_CCACHE_ERR_TYPE_CE, + "DatECCFix"); + } + if (irq == g_irq[DATA_UNCORR]) { + add_h = readl(ccache_base + SIFIVE_CCACHE_DATECCFAIL_HIGH); + add_l = readl(ccache_base + SIFIVE_CCACHE_DATECCFAIL_LOW); + pr_err("CCACHE: DataFail @ 0x%08X.%08X\n", add_h, add_l); + /* Reading this register clears the DataFail interrupt sig */ + readl(ccache_base + SIFIVE_CCACHE_DATECCFAIL_COUNT); + atomic_notifier_call_chain(&ccache_err_chain, + SIFIVE_CCACHE_ERR_TYPE_UE, + "DatECCFail"); + } + + return IRQ_HANDLED; +} + +static int __init sifive_ccache_init(void) +{ + struct device_node *np; + struct resource res; + int i, rc, intr_num; + + np = of_find_matching_node(NULL, sifive_ccache_ids); + if (!np) + return -ENODEV; + + if (of_address_to_resource(np, 0, &res)) + return -ENODEV; + + ccache_base = ioremap(res.start, resource_size(&res)); + if (!ccache_base) + return -ENOMEM; + + intr_num = of_property_count_u32_elems(np, "interrupts"); + if (!intr_num) { + pr_err("CCACHE: no interrupts property\n"); + return -ENODEV; + } + + for (i = 0; i < intr_num; i++) { + g_irq[i] = irq_of_parse_and_map(np, i); + rc = request_irq(g_irq[i], ccache_int_handler, 0, "ccache_ecc", + NULL); + if (rc) { + pr_err("CCACHE: Could not request IRQ %d\n", g_irq[i]); + return rc; + } + } + + ccache_config_read(); + + ccache_cache_ops.get_priv_group = ccache_get_priv_group; + riscv_set_cacheinfo_ops(&ccache_cache_ops); + +#ifdef CONFIG_DEBUG_FS + setup_sifive_debug(); +#endif + return 0; +} + +device_initcall(sifive_ccache_init); diff --git a/drivers/soc/sifive/sifive_l2_cache.c b/drivers/soc/sifive/sifive_l2_cache.c deleted file mode 100644 index 59640a1d0b28..000000000000 --- a/drivers/soc/sifive/sifive_l2_cache.c +++ /dev/null @@ -1,237 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -/* - * SiFive L2 cache controller Driver - * - * Copyright (C) 2018-2019 SiFive, Inc. - * - */ -#include -#include -#include -#include -#include -#include -#include - -#define SIFIVE_L2_DIRECCFIX_LOW 0x100 -#define SIFIVE_L2_DIRECCFIX_HIGH 0x104 -#define SIFIVE_L2_DIRECCFIX_COUNT 0x108 - -#define SIFIVE_L2_DIRECCFAIL_LOW 0x120 -#define SIFIVE_L2_DIRECCFAIL_HIGH 0x124 -#define SIFIVE_L2_DIRECCFAIL_COUNT 0x128 - -#define SIFIVE_L2_DATECCFIX_LOW 0x140 -#define SIFIVE_L2_DATECCFIX_HIGH 0x144 -#define SIFIVE_L2_DATECCFIX_COUNT 0x148 - -#define SIFIVE_L2_DATECCFAIL_LOW 0x160 -#define SIFIVE_L2_DATECCFAIL_HIGH 0x164 -#define SIFIVE_L2_DATECCFAIL_COUNT 0x168 - -#define SIFIVE_L2_CONFIG 0x00 -#define SIFIVE_L2_WAYENABLE 0x08 -#define SIFIVE_L2_ECCINJECTERR 0x40 - -#define SIFIVE_L2_MAX_ECCINTR 4 - -static void __iomem *l2_base; -static int g_irq[SIFIVE_L2_MAX_ECCINTR]; -static struct riscv_cacheinfo_ops l2_cache_ops; - -enum { - DIR_CORR = 0, - DATA_CORR, - DATA_UNCORR, - DIR_UNCORR, -}; - -#ifdef CONFIG_DEBUG_FS -static struct dentry *sifive_test; - -static ssize_t l2_write(struct file *file, const char __user *data, - size_t count, loff_t *ppos) -{ - unsigned int val; - - if (kstrtouint_from_user(data, count, 0, &val)) - return -EINVAL; - if ((val < 0xFF) || (val >= 0x10000 && val < 0x100FF)) - writel(val, l2_base + SIFIVE_L2_ECCINJECTERR); - else - return -EINVAL; - return count; -} - -static const struct file_operations l2_fops = { - .owner = THIS_MODULE, - .open = simple_open, - .write = l2_write -}; - -static void setup_sifive_debug(void) -{ - sifive_test = debugfs_create_dir("sifive_l2_cache", NULL); - - debugfs_create_file("sifive_debug_inject_error", 0200, - sifive_test, NULL, &l2_fops); -} -#endif - -static void l2_config_read(void) -{ - u32 regval, val; - - regval = readl(l2_base + SIFIVE_L2_CONFIG); - val = regval & 0xFF; - pr_info("L2CACHE: No. of Banks in the cache: %d\n", val); - val = (regval & 0xFF00) >> 8; - pr_info("L2CACHE: No. of ways per bank: %d\n", val); - val = (regval & 0xFF0000) >> 16; - pr_info("L2CACHE: Sets per bank: %llu\n", (uint64_t)1 << val); - val = (regval & 0xFF000000) >> 24; - pr_info("L2CACHE: Bytes per cache block: %llu\n", (uint64_t)1 << val); - - regval = readl(l2_base + SIFIVE_L2_WAYENABLE); - pr_info("L2CACHE: Index of the largest way enabled: %d\n", regval); -} - -static const struct of_device_id sifive_l2_ids[] = { - { .compatible = "sifive,fu540-c000-ccache" }, - { .compatible = "sifive,fu740-c000-ccache" }, - { /* end of table */ }, -}; - -static ATOMIC_NOTIFIER_HEAD(l2_err_chain); - -int register_sifive_l2_error_notifier(struct notifier_block *nb) -{ - return atomic_notifier_chain_register(&l2_err_chain, nb); -} -EXPORT_SYMBOL_GPL(register_sifive_l2_error_notifier); - -int unregister_sifive_l2_error_notifier(struct notifier_block *nb) -{ - return atomic_notifier_chain_unregister(&l2_err_chain, nb); -} -EXPORT_SYMBOL_GPL(unregister_sifive_l2_error_notifier); - -static int l2_largest_wayenabled(void) -{ - return readl(l2_base + SIFIVE_L2_WAYENABLE) & 0xFF; -} - -static ssize_t number_of_ways_enabled_show(struct device *dev, - struct device_attribute *attr, - char *buf) -{ - return sprintf(buf, "%u\n", l2_largest_wayenabled()); -} - -static DEVICE_ATTR_RO(number_of_ways_enabled); - -static struct attribute *priv_attrs[] = { - &dev_attr_number_of_ways_enabled.attr, - NULL, -}; - -static const struct attribute_group priv_attr_group = { - .attrs = priv_attrs, -}; - -static const struct attribute_group *l2_get_priv_group(struct cacheinfo *this_leaf) -{ - /* We want to use private group for L2 cache only */ - if (this_leaf->level == 2) - return &priv_attr_group; - else - return NULL; -} - -static irqreturn_t l2_int_handler(int irq, void *device) -{ - unsigned int add_h, add_l; - - if (irq == g_irq[DIR_CORR]) { - add_h = readl(l2_base + SIFIVE_L2_DIRECCFIX_HIGH); - add_l = readl(l2_base + SIFIVE_L2_DIRECCFIX_LOW); - pr_err("L2CACHE: DirError @ 0x%08X.%08X\n", add_h, add_l); - /* Reading this register clears the DirError interrupt sig */ - readl(l2_base + SIFIVE_L2_DIRECCFIX_COUNT); - atomic_notifier_call_chain(&l2_err_chain, SIFIVE_L2_ERR_TYPE_CE, - "DirECCFix"); - } - if (irq == g_irq[DIR_UNCORR]) { - add_h = readl(l2_base + SIFIVE_L2_DIRECCFAIL_HIGH); - add_l = readl(l2_base + SIFIVE_L2_DIRECCFAIL_LOW); - /* Reading this register clears the DirFail interrupt sig */ - readl(l2_base + SIFIVE_L2_DIRECCFAIL_COUNT); - atomic_notifier_call_chain(&l2_err_chain, SIFIVE_L2_ERR_TYPE_UE, - "DirECCFail"); - panic("L2CACHE: DirFail @ 0x%08X.%08X\n", add_h, add_l); - } - if (irq == g_irq[DATA_CORR]) { - add_h = readl(l2_base + SIFIVE_L2_DATECCFIX_HIGH); - add_l = readl(l2_base + SIFIVE_L2_DATECCFIX_LOW); - pr_err("L2CACHE: DataError @ 0x%08X.%08X\n", add_h, add_l); - /* Reading this register clears the DataError interrupt sig */ - readl(l2_base + SIFIVE_L2_DATECCFIX_COUNT); - atomic_notifier_call_chain(&l2_err_chain, SIFIVE_L2_ERR_TYPE_CE, - "DatECCFix"); - } - if (irq == g_irq[DATA_UNCORR]) { - add_h = readl(l2_base + SIFIVE_L2_DATECCFAIL_HIGH); - add_l = readl(l2_base + SIFIVE_L2_DATECCFAIL_LOW); - pr_err("L2CACHE: DataFail @ 0x%08X.%08X\n", add_h, add_l); - /* Reading this register clears the DataFail interrupt sig */ - readl(l2_base + SIFIVE_L2_DATECCFAIL_COUNT); - atomic_notifier_call_chain(&l2_err_chain, SIFIVE_L2_ERR_TYPE_UE, - "DatECCFail"); - } - - return IRQ_HANDLED; -} - -static int __init sifive_l2_init(void) -{ - struct device_node *np; - struct resource res; - int i, rc, intr_num; - - np = of_find_matching_node(NULL, sifive_l2_ids); - if (!np) - return -ENODEV; - - if (of_address_to_resource(np, 0, &res)) - return -ENODEV; - - l2_base = ioremap(res.start, resource_size(&res)); - if (!l2_base) - return -ENOMEM; - - intr_num = of_property_count_u32_elems(np, "interrupts"); - if (!intr_num) { - pr_err("L2CACHE: no interrupts property\n"); - return -ENODEV; - } - - for (i = 0; i < intr_num; i++) { - g_irq[i] = irq_of_parse_and_map(np, i); - rc = request_irq(g_irq[i], l2_int_handler, 0, "l2_ecc", NULL); - if (rc) { - pr_err("L2CACHE: Could not request IRQ %d\n", g_irq[i]); - return rc; - } - } - - l2_config_read(); - - l2_cache_ops.get_priv_group = l2_get_priv_group; - riscv_set_cacheinfo_ops(&l2_cache_ops); - -#ifdef CONFIG_DEBUG_FS - setup_sifive_debug(); -#endif - return 0; -} -device_initcall(sifive_l2_init); diff --git a/include/soc/sifive/sifive_ccache.h b/include/soc/sifive/sifive_ccache.h new file mode 100644 index 000000000000..4d4ed49388a0 --- /dev/null +++ b/include/soc/sifive/sifive_ccache.h @@ -0,0 +1,16 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * SiFive Composable Cache Controller header file + * + */ + +#ifndef __SOC_SIFIVE_CCACHE_H +#define __SOC_SIFIVE_CCACHE_H + +extern int register_sifive_ccache_error_notifier(struct notifier_block *nb); +extern int unregister_sifive_ccache_error_notifier(struct notifier_block *nb); + +#define SIFIVE_CCACHE_ERR_TYPE_CE 0 +#define SIFIVE_CCACHE_ERR_TYPE_UE 1 + +#endif /* __SOC_SIFIVE_CCACHE_H */ diff --git a/include/soc/sifive/sifive_l2_cache.h b/include/soc/sifive/sifive_l2_cache.h deleted file mode 100644 index 92ade10ed67e..000000000000 --- a/include/soc/sifive/sifive_l2_cache.h +++ /dev/null @@ -1,16 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -/* - * SiFive L2 Cache Controller header file - * - */ - -#ifndef __SOC_SIFIVE_L2_CACHE_H -#define __SOC_SIFIVE_L2_CACHE_H - -extern int register_sifive_l2_error_notifier(struct notifier_block *nb); -extern int unregister_sifive_l2_error_notifier(struct notifier_block *nb); - -#define SIFIVE_L2_ERR_TYPE_CE 0 -#define SIFIVE_L2_ERR_TYPE_UE 1 - -#endif /* __SOC_SIFIVE_L2_CACHE_H */ From 95f196f3212bbc258611c22865aef12b98304e1d Mon Sep 17 00:00:00 2001 From: Zong Li Date: Tue, 13 Sep 2022 06:18:13 +0000 Subject: [PATCH 108/121] soc: sifive: ccache: determine the cache level from dts Composable cache could be L2 or L3 cache, use 'cache-level' property of device node to determine the level. Signed-off-by: Zong Li Reviewed-by: Conor Dooley Link: https://lore.kernel.org/r/20220913061817.22564-4-zong.li@sifive.com Signed-off-by: Palmer Dabbelt --- drivers/soc/sifive/sifive_ccache.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/soc/sifive/sifive_ccache.c b/drivers/soc/sifive/sifive_ccache.c index 949b824e89ad..b361b661ea09 100644 --- a/drivers/soc/sifive/sifive_ccache.c +++ b/drivers/soc/sifive/sifive_ccache.c @@ -38,6 +38,7 @@ static void __iomem *ccache_base; static int g_irq[SIFIVE_CCACHE_MAX_ECCINTR]; static struct riscv_cacheinfo_ops ccache_cache_ops; +static int level; enum { DIR_CORR = 0, @@ -144,7 +145,7 @@ static const struct attribute_group *ccache_get_priv_group(struct cacheinfo *this_leaf) { /* We want to use private group for composable cache only */ - if (this_leaf->level == 2) + if (this_leaf->level == level) return &priv_attr_group; else return NULL; @@ -215,6 +216,9 @@ static int __init sifive_ccache_init(void) if (!ccache_base) return -ENOMEM; + if (of_property_read_u32(np, "cache-level", &level)) + return -ENOENT; + intr_num = of_property_count_u32_elems(np, "interrupts"); if (!intr_num) { pr_err("CCACHE: no interrupts property\n"); From 3fb787e5bad50687a65ded7f3bb805cab70dff59 Mon Sep 17 00:00:00 2001 From: Ben Dooks Date: Tue, 13 Sep 2022 06:18:14 +0000 Subject: [PATCH 109/121] soc: sifive: ccache: reduce printing on init The driver prints out 6 lines on startup, which can easily be redcued to two lines without losing any information. Note, to make the types work better, uint64_t has been replaced with ULL to make the unsigned long long match the format in the print statement. Signed-off-by: Ben Dooks Signed-off-by: Zong Li Reviewed-by: Conor Dooley Link: https://lore.kernel.org/r/20220913061817.22564-5-zong.li@sifive.com Signed-off-by: Palmer Dabbelt --- drivers/soc/sifive/sifive_ccache.c | 21 +++++++++------------ 1 file changed, 9 insertions(+), 12 deletions(-) diff --git a/drivers/soc/sifive/sifive_ccache.c b/drivers/soc/sifive/sifive_ccache.c index b361b661ea09..17080af7dfa0 100644 --- a/drivers/soc/sifive/sifive_ccache.c +++ b/drivers/soc/sifive/sifive_ccache.c @@ -81,20 +81,17 @@ static void setup_sifive_debug(void) static void ccache_config_read(void) { - u32 regval, val; + u32 cfg; - regval = readl(ccache_base + SIFIVE_CCACHE_CONFIG); - val = regval & 0xFF; - pr_info("CCACHE: No. of Banks in the cache: %d\n", val); - val = (regval & 0xFF00) >> 8; - pr_info("CCACHE: No. of ways per bank: %d\n", val); - val = (regval & 0xFF0000) >> 16; - pr_info("CCACHE: Sets per bank: %llu\n", (uint64_t)1 << val); - val = (regval & 0xFF000000) >> 24; - pr_info("CCACHE: Bytes per cache block: %llu\n", (uint64_t)1 << val); + cfg = readl(ccache_base + SIFIVE_CCACHE_CONFIG); - regval = readl(ccache_base + SIFIVE_CCACHE_WAYENABLE); - pr_info("CCACHE: Index of the largest way enabled: %d\n", regval); + pr_info("CCACHE: %u banks, %u ways, sets/bank=%llu, bytes/block=%llu\n", + (cfg & 0xff), (cfg >> 8) & 0xff, + BIT_ULL((cfg >> 16) & 0xff), + BIT_ULL((cfg >> 24) & 0xff)); + + cfg = readl(ccache_base + SIFIVE_CCACHE_WAYENABLE); + pr_info("CCACHE: Index of the largest way enabled: %u\n", cfg); } static const struct of_device_id sifive_ccache_ids[] = { From 696ab9bda22a770d079dc3a23bac9aaa553d98f4 Mon Sep 17 00:00:00 2001 From: Ben Dooks Date: Tue, 13 Sep 2022 06:18:15 +0000 Subject: [PATCH 110/121] soc: sifive: ccache: use pr_fmt() to remove CCACHE: prefixes Use the pr_fmt() macro to prefix all the output with "CCACHE:" to avoid having to write it out each time, or make a large diff when the next change comes along. Signed-off-by: Ben Dooks Signed-off-by: Zong Li Reviewed-by: Conor Dooley Link: https://lore.kernel.org/r/20220913061817.22564-6-zong.li@sifive.com Signed-off-by: Palmer Dabbelt --- drivers/soc/sifive/sifive_ccache.c | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-) diff --git a/drivers/soc/sifive/sifive_ccache.c b/drivers/soc/sifive/sifive_ccache.c index 17080af7dfa0..91f0c2b32ea2 100644 --- a/drivers/soc/sifive/sifive_ccache.c +++ b/drivers/soc/sifive/sifive_ccache.c @@ -5,6 +5,9 @@ * Copyright (C) 2018-2022 SiFive, Inc. * */ + +#define pr_fmt(fmt) "CCACHE: " fmt + #include #include #include @@ -85,13 +88,13 @@ static void ccache_config_read(void) cfg = readl(ccache_base + SIFIVE_CCACHE_CONFIG); - pr_info("CCACHE: %u banks, %u ways, sets/bank=%llu, bytes/block=%llu\n", + pr_info("%u banks, %u ways, sets/bank=%llu, bytes/block=%llu\n", (cfg & 0xff), (cfg >> 8) & 0xff, BIT_ULL((cfg >> 16) & 0xff), BIT_ULL((cfg >> 24) & 0xff)); cfg = readl(ccache_base + SIFIVE_CCACHE_WAYENABLE); - pr_info("CCACHE: Index of the largest way enabled: %u\n", cfg); + pr_info("Index of the largest way enabled: %u\n", cfg); } static const struct of_device_id sifive_ccache_ids[] = { @@ -155,7 +158,7 @@ static irqreturn_t ccache_int_handler(int irq, void *device) if (irq == g_irq[DIR_CORR]) { add_h = readl(ccache_base + SIFIVE_CCACHE_DIRECCFIX_HIGH); add_l = readl(ccache_base + SIFIVE_CCACHE_DIRECCFIX_LOW); - pr_err("CCACHE: DirError @ 0x%08X.%08X\n", add_h, add_l); + pr_err("DirError @ 0x%08X.%08X\n", add_h, add_l); /* Reading this register clears the DirError interrupt sig */ readl(ccache_base + SIFIVE_CCACHE_DIRECCFIX_COUNT); atomic_notifier_call_chain(&ccache_err_chain, @@ -175,7 +178,7 @@ static irqreturn_t ccache_int_handler(int irq, void *device) if (irq == g_irq[DATA_CORR]) { add_h = readl(ccache_base + SIFIVE_CCACHE_DATECCFIX_HIGH); add_l = readl(ccache_base + SIFIVE_CCACHE_DATECCFIX_LOW); - pr_err("CCACHE: DataError @ 0x%08X.%08X\n", add_h, add_l); + pr_err("DataError @ 0x%08X.%08X\n", add_h, add_l); /* Reading this register clears the DataError interrupt sig */ readl(ccache_base + SIFIVE_CCACHE_DATECCFIX_COUNT); atomic_notifier_call_chain(&ccache_err_chain, @@ -185,7 +188,7 @@ static irqreturn_t ccache_int_handler(int irq, void *device) if (irq == g_irq[DATA_UNCORR]) { add_h = readl(ccache_base + SIFIVE_CCACHE_DATECCFAIL_HIGH); add_l = readl(ccache_base + SIFIVE_CCACHE_DATECCFAIL_LOW); - pr_err("CCACHE: DataFail @ 0x%08X.%08X\n", add_h, add_l); + pr_err("DataFail @ 0x%08X.%08X\n", add_h, add_l); /* Reading this register clears the DataFail interrupt sig */ readl(ccache_base + SIFIVE_CCACHE_DATECCFAIL_COUNT); atomic_notifier_call_chain(&ccache_err_chain, @@ -218,7 +221,7 @@ static int __init sifive_ccache_init(void) intr_num = of_property_count_u32_elems(np, "interrupts"); if (!intr_num) { - pr_err("CCACHE: no interrupts property\n"); + pr_err("No interrupts property\n"); return -ENODEV; } @@ -227,7 +230,7 @@ static int __init sifive_ccache_init(void) rc = request_irq(g_irq[i], ccache_int_handler, 0, "ccache_ecc", NULL); if (rc) { - pr_err("CCACHE: Could not request IRQ %d\n", g_irq[i]); + pr_err("Could not request IRQ %d\n", g_irq[i]); return rc; } } From afc7a5834f0de13aee46df62f09e479c1bbf7b9d Mon Sep 17 00:00:00 2001 From: Zong Li Date: Tue, 13 Sep 2022 06:18:16 +0000 Subject: [PATCH 111/121] soc: sifive: ccache: define the macro for the register shifts Define the macro for the register shifts, it could make the code be more readable Signed-off-by: Zong Li Reviewed-by: Conor Dooley Link: https://lore.kernel.org/r/20220913061817.22564-7-zong.li@sifive.com Signed-off-by: Palmer Dabbelt --- drivers/soc/sifive/sifive_ccache.c | 16 +++++++++++----- 1 file changed, 11 insertions(+), 5 deletions(-) diff --git a/drivers/soc/sifive/sifive_ccache.c b/drivers/soc/sifive/sifive_ccache.c index 91f0c2b32ea2..1c171150e878 100644 --- a/drivers/soc/sifive/sifive_ccache.c +++ b/drivers/soc/sifive/sifive_ccache.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include @@ -33,6 +34,11 @@ #define SIFIVE_CCACHE_DATECCFAIL_COUNT 0x168 #define SIFIVE_CCACHE_CONFIG 0x00 +#define SIFIVE_CCACHE_CONFIG_BANK_MASK GENMASK_ULL(7, 0) +#define SIFIVE_CCACHE_CONFIG_WAYS_MASK GENMASK_ULL(15, 8) +#define SIFIVE_CCACHE_CONFIG_SETS_MASK GENMASK_ULL(23, 16) +#define SIFIVE_CCACHE_CONFIG_BLKS_MASK GENMASK_ULL(31, 24) + #define SIFIVE_CCACHE_WAYENABLE 0x08 #define SIFIVE_CCACHE_ECCINJECTERR 0x40 @@ -87,11 +93,11 @@ static void ccache_config_read(void) u32 cfg; cfg = readl(ccache_base + SIFIVE_CCACHE_CONFIG); - - pr_info("%u banks, %u ways, sets/bank=%llu, bytes/block=%llu\n", - (cfg & 0xff), (cfg >> 8) & 0xff, - BIT_ULL((cfg >> 16) & 0xff), - BIT_ULL((cfg >> 24) & 0xff)); + pr_info("%llu banks, %llu ways, sets/bank=%llu, bytes/block=%llu\n", + FIELD_GET(SIFIVE_CCACHE_CONFIG_BANK_MASK, cfg), + FIELD_GET(SIFIVE_CCACHE_CONFIG_WAYS_MASK, cfg), + BIT_ULL(FIELD_GET(SIFIVE_CCACHE_CONFIG_SETS_MASK, cfg)), + BIT_ULL(FIELD_GET(SIFIVE_CCACHE_CONFIG_BLKS_MASK, cfg))); cfg = readl(ccache_base + SIFIVE_CCACHE_WAYENABLE); pr_info("Index of the largest way enabled: %u\n", cfg); From da29dbcda49d60f34055df19bd4783b889fc7dfc Mon Sep 17 00:00:00 2001 From: Greentime Hu Date: Tue, 13 Sep 2022 06:18:17 +0000 Subject: [PATCH 112/121] riscv: Add cache information in AUX vector There are no standard CSR registers to provide cache information, the way for RISC-V is to get this information from DT. sysconf syscall could use them to get information of cache through AUX vector. The result of 'getconf -a|grep -i cache' as follows: LEVEL1_ICACHE_SIZE 32768 LEVEL1_ICACHE_ASSOC 2 LEVEL1_ICACHE_LINESIZE 64 LEVEL1_DCACHE_SIZE 32768 LEVEL1_DCACHE_ASSOC 4 LEVEL1_DCACHE_LINESIZE 64 LEVEL2_CACHE_SIZE 524288 LEVEL2_CACHE_ASSOC 8 LEVEL2_CACHE_LINESIZE 64 LEVEL3_CACHE_SIZE 4194304 LEVEL3_CACHE_ASSOC 16 LEVEL3_CACHE_LINESIZE 64 LEVEL4_CACHE_SIZE 0 LEVEL4_CACHE_ASSOC 0 LEVEL4_CACHE_LINESIZE 0 Signed-off-by: Greentime Hu Signed-off-by: Zong Li Suggested-by: Zong Li Reviewed-by: Conor Dooley Link: https://lore.kernel.org/r/20220913061817.22564-8-zong.li@sifive.com Signed-off-by: Palmer Dabbelt --- arch/riscv/include/asm/elf.h | 4 ++++ arch/riscv/include/uapi/asm/auxvec.h | 4 +++- 2 files changed, 7 insertions(+), 1 deletion(-) diff --git a/arch/riscv/include/asm/elf.h b/arch/riscv/include/asm/elf.h index 14fc7342490b..e7acffdf21d2 100644 --- a/arch/riscv/include/asm/elf.h +++ b/arch/riscv/include/asm/elf.h @@ -99,6 +99,10 @@ do { \ get_cache_size(2, CACHE_TYPE_UNIFIED)); \ NEW_AUX_ENT(AT_L2_CACHEGEOMETRY, \ get_cache_geometry(2, CACHE_TYPE_UNIFIED)); \ + NEW_AUX_ENT(AT_L3_CACHESIZE, \ + get_cache_size(3, CACHE_TYPE_UNIFIED)); \ + NEW_AUX_ENT(AT_L3_CACHEGEOMETRY, \ + get_cache_geometry(3, CACHE_TYPE_UNIFIED)); \ } while (0) #define ARCH_HAS_SETUP_ADDITIONAL_PAGES struct linux_binprm; diff --git a/arch/riscv/include/uapi/asm/auxvec.h b/arch/riscv/include/uapi/asm/auxvec.h index 32c73ba1d531..fb187a33ce58 100644 --- a/arch/riscv/include/uapi/asm/auxvec.h +++ b/arch/riscv/include/uapi/asm/auxvec.h @@ -30,8 +30,10 @@ #define AT_L1D_CACHEGEOMETRY 43 #define AT_L2_CACHESIZE 44 #define AT_L2_CACHEGEOMETRY 45 +#define AT_L3_CACHESIZE 46 +#define AT_L3_CACHEGEOMETRY 47 /* entries in ARCH_DLINFO */ -#define AT_VECTOR_SIZE_ARCH 7 +#define AT_VECTOR_SIZE_ARCH 9 #endif /* _UAPI_ASM_RISCV_AUXVEC_H */ From a8616d2dc193b6becc36b5f3cfeaa9ac7a5762f9 Mon Sep 17 00:00:00 2001 From: Jisheng Zhang Date: Sat, 24 Sep 2022 15:07:37 +0800 Subject: [PATCH 113/121] riscv: vdso: fix NULL deference in vdso_join_timens() when vfork Testing tools/testing/selftests/timens/vfork_exec.c got below kernel log: [ 6.838454] Unable to handle kernel access to user memory without uaccess routines at virtual address 0000000000000020 [ 6.842255] Oops [#1] [ 6.842871] Modules linked in: [ 6.844249] CPU: 1 PID: 64 Comm: vfork_exec Not tainted 6.0.0-rc3-rt15+ #8 [ 6.845861] Hardware name: riscv-virtio,qemu (DT) [ 6.848009] epc : vdso_join_timens+0xd2/0x110 [ 6.850097] ra : vdso_join_timens+0xd2/0x110 [ 6.851164] epc : ffffffff8000635c ra : ffffffff8000635c sp : ff6000000181fbf0 [ 6.852562] gp : ffffffff80cff648 tp : ff60000000fdb700 t0 : 3030303030303030 [ 6.853852] t1 : 0000000000000030 t2 : 3030303030303030 s0 : ff6000000181fc40 [ 6.854984] s1 : ff60000001e6c000 a0 : 0000000000000010 a1 : ffffffff8005654c [ 6.856221] a2 : 00000000ffffefff a3 : 0000000000000000 a4 : 0000000000000000 [ 6.858114] a5 : 0000000000000000 a6 : 0000000000000008 a7 : 0000000000000038 [ 6.859484] s2 : ff60000001e6c068 s3 : ff6000000108abb0 s4 : 0000000000000000 [ 6.860751] s5 : 0000000000001000 s6 : ffffffff8089dc40 s7 : ffffffff8089dc38 [ 6.862029] s8 : ffffffff8089dc30 s9 : ff60000000fdbe38 s10: 000000000000005e [ 6.863304] s11: ffffffff80cc3510 t3 : ffffffff80d1112f t4 : ffffffff80d1112f [ 6.864565] t5 : ffffffff80d11130 t6 : ff6000000181fa00 [ 6.865561] status: 0000000000000120 badaddr: 0000000000000020 cause: 000000000000000d [ 6.868046] [] timens_commit+0x38/0x11a [ 6.869089] [] timens_on_fork+0x72/0xb4 [ 6.870055] [] begin_new_exec+0x3c6/0x9f0 [ 6.871231] [] load_elf_binary+0x628/0x1214 [ 6.872304] [] bprm_execve+0x1f2/0x4e4 [ 6.873243] [] do_execveat_common+0x16e/0x1ee [ 6.874258] [] sys_execve+0x3c/0x48 [ 6.875162] [] ret_from_syscall+0x0/0x2 [ 6.877484] ---[ end trace 0000000000000000 ]--- This is because the mm->context.vdso_info is NULL in vfork case. From another side, mm->context.vdso_info either points to vdso info for RV64 or vdso info for compat, there's no need to bloat riscv's mm_context_t, we can handle the difference when setup the additional page for vdso. Signed-off-by: Jisheng Zhang Suggested-by: Palmer Dabbelt Fixes: 3092eb456375 ("riscv: compat: vdso: Add setup additional pages implementation") Link: https://lore.kernel.org/r/20220924070737.3048-1-jszhang@kernel.org Cc: stable@vger.kernel.org Signed-off-by: Palmer Dabbelt --- arch/riscv/include/asm/mmu.h | 1 - arch/riscv/kernel/vdso.c | 13 ++++++++++--- 2 files changed, 10 insertions(+), 4 deletions(-) diff --git a/arch/riscv/include/asm/mmu.h b/arch/riscv/include/asm/mmu.h index cedcf8ea3c76..0099dc116168 100644 --- a/arch/riscv/include/asm/mmu.h +++ b/arch/riscv/include/asm/mmu.h @@ -16,7 +16,6 @@ typedef struct { atomic_long_t id; #endif void *vdso; - void *vdso_info; #ifdef CONFIG_SMP /* A local icache flush is needed before user execution can resume. */ cpumask_t icache_stale_mask; diff --git a/arch/riscv/kernel/vdso.c b/arch/riscv/kernel/vdso.c index 69b05b6c181b..4abc9aebdfae 100644 --- a/arch/riscv/kernel/vdso.c +++ b/arch/riscv/kernel/vdso.c @@ -60,6 +60,11 @@ struct __vdso_info { struct vm_special_mapping *cm; }; +static struct __vdso_info vdso_info; +#ifdef CONFIG_COMPAT +static struct __vdso_info compat_vdso_info; +#endif + static int vdso_mremap(const struct vm_special_mapping *sm, struct vm_area_struct *new_vma) { @@ -114,15 +119,18 @@ int vdso_join_timens(struct task_struct *task, struct time_namespace *ns) { struct mm_struct *mm = task->mm; struct vm_area_struct *vma; - struct __vdso_info *vdso_info = mm->context.vdso_info; mmap_read_lock(mm); for (vma = mm->mmap; vma; vma = vma->vm_next) { unsigned long size = vma->vm_end - vma->vm_start; - if (vma_is_special_mapping(vma, vdso_info->dm)) + if (vma_is_special_mapping(vma, vdso_info.dm)) zap_page_range(vma, vma->vm_start, size); +#ifdef CONFIG_COMPAT + if (vma_is_special_mapping(vma, compat_vdso_info.dm)) + zap_page_range(vma, vma->vm_start, size); +#endif } mmap_read_unlock(mm); @@ -264,7 +272,6 @@ static int __setup_additional_pages(struct mm_struct *mm, vdso_base += VVAR_SIZE; mm->context.vdso = (void *)vdso_base; - mm->context.vdso_info = (void *)vdso_info; ret = _install_special_mapping(mm, vdso_base, vdso_text_len, From 5a5294fbe0200d1327f0e089135dad77b45aa2ee Mon Sep 17 00:00:00 2001 From: Palmer Dabbelt Date: Wed, 28 Sep 2022 06:18:07 -0700 Subject: [PATCH 114/121] RISC-V: Re-enable counter access from userspace These counters were part of the ISA when we froze the uABI, removing them breaks userspace. Link: https://lore.kernel.org/all/YxEhC%2FmDW1lFt36J@aurel32.net/ Fixes: e9991434596f ("RISC-V: Add perf platform driver based on SBI PMU extension") Tested-by: Conor Dooley Reviewed-by: Conor Dooley Link: https://lore.kernel.org/r/20220928131807.30386-1-palmer@rivosinc.com Cc: stable@vger.kernel.org Signed-off-by: Palmer Dabbelt --- drivers/perf/riscv_pmu_sbi.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/perf/riscv_pmu_sbi.c b/drivers/perf/riscv_pmu_sbi.c index 15e5a47be7d5..3852c18362f5 100644 --- a/drivers/perf/riscv_pmu_sbi.c +++ b/drivers/perf/riscv_pmu_sbi.c @@ -652,8 +652,11 @@ static int pmu_sbi_starting_cpu(unsigned int cpu, struct hlist_node *node) struct riscv_pmu *pmu = hlist_entry_safe(node, struct riscv_pmu, node); struct cpu_hw_events *cpu_hw_evt = this_cpu_ptr(pmu->hw_events); - /* Enable the access for TIME csr only from the user mode now */ - csr_write(CSR_SCOUNTEREN, 0x2); + /* + * Enable the access for CYCLE, TIME, and INSTRET CSRs from userspace, + * as is necessary to maintain uABI compatibility. + */ + csr_write(CSR_SCOUNTEREN, 0x7); /* Stop all the counters so that they can be enabled from perf */ pmu_sbi_stop_all(pmu); From c45fc916c2b2cc2a0587659c18d6ceef9b7299be Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Fri, 29 Jul 2022 12:11:17 +0100 Subject: [PATCH 115/121] riscv: enable software resend of irqs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The PLIC specification does not describe the interrupt pendings bits as read-write, only that they "can be read". To allow for retriggering of interrupts (and the use of the irq debugfs interface) enable HARDIRQS_SW_RESEND for RISC-V. Link: https://github.com/riscv/riscv-plic-spec/blob/master/riscv-plic.adoc#interrupt-pending-bits Signed-off-by: Conor Dooley Acked-by: Marc Zyngier Acked-by: Palmer Dabbelt Tested-by: Palmer Dabbelt # on QEMU Reviewed-by: Björn Töpel Link: https://lore.kernel.org/r/20220729111116.259146-1-conor.dooley@microchip.com Signed-off-by: Palmer Dabbelt --- arch/riscv/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig index e84f2742b6bb..c56bc70158ac 100644 --- a/arch/riscv/Kconfig +++ b/arch/riscv/Kconfig @@ -70,6 +70,7 @@ config RISCV select GENERIC_SMP_IDLE_THREAD select GENERIC_TIME_VSYSCALL if MMU && 64BIT select GENERIC_VDSO_TIME_NS if HAVE_GENERIC_VDSO + select HARDIRQS_SW_RESEND select HAVE_ARCH_AUDITSYSCALL select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL select HAVE_ARCH_JUMP_LABEL_RELATIVE if !XIP_KERNEL From 9cc205e3c17d5716da7ebb7fa0c985555e95d009 Mon Sep 17 00:00:00 2001 From: "Maciej W. Rozycki" Date: Thu, 22 Sep 2022 22:56:06 +0100 Subject: [PATCH 116/121] RISC-V: Make port I/O string accessors actually work Fix port I/O string accessors such as `insb', `outsb', etc. which use the physical PCI port I/O address rather than the corresponding memory mapping to get at the requested location, which in turn breaks at least accesses made by our parport driver to a PCIe parallel port such as: PCI parallel port detected: 1415:c118, I/O at 0x1000(0x1008), IRQ 20 parport0: PC-style at 0x1000 (0x1008), irq 20, using FIFO [PCSPP,TRISTATE,COMPAT,EPP,ECP] causing a memory access fault: Unable to handle kernel access to user memory without uaccess routines at virtual address 0000000000001008 Oops [#1] Modules linked in: CPU: 1 PID: 350 Comm: cat Not tainted 6.0.0-rc2-00283-g10d4879f9ef0-dirty #23 Hardware name: SiFive HiFive Unmatched A00 (DT) epc : parport_pc_fifo_write_block_pio+0x266/0x416 ra : parport_pc_fifo_write_block_pio+0xb4/0x416 epc : ffffffff80542c3e ra : ffffffff80542a8c sp : ffffffd88899fc60 gp : ffffffff80fa2700 tp : ffffffd882b1e900 t0 : ffffffd883d0b000 t1 : ffffffffff000002 t2 : 4646393043330a38 s0 : ffffffd88899fcf0 s1 : 0000000000001000 a0 : 0000000000000010 a1 : 0000000000000000 a2 : ffffffd883d0a010 a3 : 0000000000000023 a4 : 00000000ffff8fbb a5 : ffffffd883d0a001 a6 : 0000000100000000 a7 : ffffffc800000000 s2 : ffffffffff000002 s3 : ffffffff80d28880 s4 : ffffffff80fa1f50 s5 : 0000000000001008 s6 : 0000000000000008 s7 : ffffffd883d0a000 s8 : 0004000000000000 s9 : ffffffff80dc1d80 s10: ffffffd8807e4000 s11: 0000000000000000 t3 : 00000000000000ff t4 : 393044410a303930 t5 : 0000000000001000 t6 : 0000000000040000 status: 0000000200000120 badaddr: 0000000000001008 cause: 000000000000000f [] parport_pc_compat_write_block_pio+0xfe/0x200 [] parport_write+0x46/0xf8 [] lp_write+0x158/0x2d2 [] vfs_write+0x8e/0x2c2 [] ksys_write+0x52/0xc2 [] sys_write+0xe/0x16 [] ret_from_syscall+0x0/0x2 ---[ end trace 0000000000000000 ]--- For simplicity address the problem by adding PCI_IOBASE to the physical address requested in the respective wrapper macros only, observing that the raw accessors such as `__insb', `__outsb', etc. are not supposed to be used other than by said macros. Remove the cast to `long' that is no longer needed on `addr' now that it is used as an offset from PCI_IOBASE and add parentheses around `addr' needed for predictable evaluation in macro expansion. No need to make said adjustments in separate changes given that current code is gravely broken and does not ever work. Signed-off-by: Maciej W. Rozycki Fixes: fab957c11efe2 ("RISC-V: Atomic and Locking Code") Cc: stable@vger.kernel.org # v4.15+ Reviewed-by: Arnd Bergmann Link: https://lore.kernel.org/r/alpine.DEB.2.21.2209220223080.29493@angie.orcam.me.uk Signed-off-by: Palmer Dabbelt --- arch/riscv/include/asm/io.h | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/arch/riscv/include/asm/io.h b/arch/riscv/include/asm/io.h index 69605a474270..92080a227937 100644 --- a/arch/riscv/include/asm/io.h +++ b/arch/riscv/include/asm/io.h @@ -101,9 +101,9 @@ __io_reads_ins(reads, u32, l, __io_br(), __io_ar(addr)) __io_reads_ins(ins, u8, b, __io_pbr(), __io_par(addr)) __io_reads_ins(ins, u16, w, __io_pbr(), __io_par(addr)) __io_reads_ins(ins, u32, l, __io_pbr(), __io_par(addr)) -#define insb(addr, buffer, count) __insb((void __iomem *)(long)addr, buffer, count) -#define insw(addr, buffer, count) __insw((void __iomem *)(long)addr, buffer, count) -#define insl(addr, buffer, count) __insl((void __iomem *)(long)addr, buffer, count) +#define insb(addr, buffer, count) __insb(PCI_IOBASE + (addr), buffer, count) +#define insw(addr, buffer, count) __insw(PCI_IOBASE + (addr), buffer, count) +#define insl(addr, buffer, count) __insl(PCI_IOBASE + (addr), buffer, count) __io_writes_outs(writes, u8, b, __io_bw(), __io_aw()) __io_writes_outs(writes, u16, w, __io_bw(), __io_aw()) @@ -115,22 +115,22 @@ __io_writes_outs(writes, u32, l, __io_bw(), __io_aw()) __io_writes_outs(outs, u8, b, __io_pbw(), __io_paw()) __io_writes_outs(outs, u16, w, __io_pbw(), __io_paw()) __io_writes_outs(outs, u32, l, __io_pbw(), __io_paw()) -#define outsb(addr, buffer, count) __outsb((void __iomem *)(long)addr, buffer, count) -#define outsw(addr, buffer, count) __outsw((void __iomem *)(long)addr, buffer, count) -#define outsl(addr, buffer, count) __outsl((void __iomem *)(long)addr, buffer, count) +#define outsb(addr, buffer, count) __outsb(PCI_IOBASE + (addr), buffer, count) +#define outsw(addr, buffer, count) __outsw(PCI_IOBASE + (addr), buffer, count) +#define outsl(addr, buffer, count) __outsl(PCI_IOBASE + (addr), buffer, count) #ifdef CONFIG_64BIT __io_reads_ins(reads, u64, q, __io_br(), __io_ar(addr)) #define readsq(addr, buffer, count) __readsq(addr, buffer, count) __io_reads_ins(ins, u64, q, __io_pbr(), __io_par(addr)) -#define insq(addr, buffer, count) __insq((void __iomem *)addr, buffer, count) +#define insq(addr, buffer, count) __insq(PCI_IOBASE + (addr), buffer, count) __io_writes_outs(writes, u64, q, __io_bw(), __io_aw()) #define writesq(addr, buffer, count) __writesq(addr, buffer, count) __io_writes_outs(outs, u64, q, __io_pbr(), __io_paw()) -#define outsq(addr, buffer, count) __outsq((void __iomem *)addr, buffer, count) +#define outsq(addr, buffer, count) __outsq(PCI_IOBASE + (addr), buffer, count) #endif #include From ab0c23b535f3f9d8345d8ad4c18c0a8594459d55 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Tue, 11 Oct 2022 17:07:45 +0100 Subject: [PATCH 117/121] MAINTAINERS: add RISC-V's patchwork The RISC-V patchwork instance on kernel.org has had some necromancy performed on it & will be used going forward. The statuses that are intended to be used are: - New: No action has been taken yet - Under Review: The maintainer is waiting for review comments from others - Changes Requested: Either the maintainer or a reviewer requested changes in the patch. The patch author is expected to submit a new version - Superseded: There's a new version of the patch available - Not Applicable: The patch is not intended for the RISC-V tree - Accepted: The patch has been applied - Rejected: The patch has been rejected, with reasons stated in an email Signed-off-by: Conor Dooley Link: https://lore.kernel.org/r/20221011160744.2167025-1-conor@kernel.org/ Signed-off-by: Palmer Dabbelt --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index b3415857a812..557da4a32717 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -17634,6 +17634,7 @@ M: Palmer Dabbelt M: Albert Ou L: linux-riscv@lists.infradead.org S: Supported +Q: https://patchwork.kernel.org/project/linux-riscv/list/ P: Documentation/riscv/patch-acceptance.rst T: git git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux.git F: arch/riscv/ From aca7c13d3bee81a968337a5515411409ae9d095d Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Fri, 14 Oct 2022 10:13:55 +0200 Subject: [PATCH 118/121] parisc: fbdev/stifb: Align graphics memory size to 4MB Independend of the current graphics resolution, adjust the reported graphics card memory size to the next 4MB boundary. This fixes the fbtest program which expects a naturally aligned size. Signed-off-by: Helge Deller Cc: --- drivers/video/fbdev/stifb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/video/fbdev/stifb.c b/drivers/video/fbdev/stifb.c index 38a861e22c33..7753e586e65a 100644 --- a/drivers/video/fbdev/stifb.c +++ b/drivers/video/fbdev/stifb.c @@ -1298,7 +1298,7 @@ static int __init stifb_init_fb(struct sti_struct *sti, int bpp_pref) /* limit fbsize to max visible screen size */ if (fix->smem_len > yres*fix->line_length) - fix->smem_len = yres*fix->line_length; + fix->smem_len = ALIGN(yres*fix->line_length, 4*1024*1024); fix->accel = FB_ACCEL_NONE; From 70be49f2f6223ddd2fcddb0089a40864c37e1494 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Fri, 14 Oct 2022 10:18:53 +0200 Subject: [PATCH 119/121] parisc: Fix userspace graphics card breakage due to pgtable special bit Commit df24e1783e6e ("parisc: Add vDSO support") introduced the vDSO support, for which a _PAGE_SPECIAL page table flag was needed. Since we wanted to keep every page table entry in 32-bits, this patch re-used the existing - but yet unused - _PAGE_DMB flag (which triggers a hardware break if a page is accessed) to store the special bit. But when graphics card memory is mmapped into userspace, the kernel uses vm_iomap_memory() which sets the the special flag. So, with the DMB bit set, every access to the graphics memory now triggered a hardware exception and segfaulted the userspace program. Fix this breakage by dropping the DMB bit when writing the page protection bits to the CPU TLB. In addition this patch adds a small optimization: if huge pages aren't configured (which is at least the case for 32-bit kernels), then the special bit is stored in the hpage (HUGE PAGE) bit instead. That way we can skip to reset the DMB bit. Fixes: df24e1783e6e ("parisc: Add vDSO support") Cc: # 5.18+ Signed-off-by: Helge Deller --- arch/parisc/include/asm/pgtable.h | 7 ++++++- arch/parisc/kernel/entry.S | 8 ++++++++ 2 files changed, 14 insertions(+), 1 deletion(-) diff --git a/arch/parisc/include/asm/pgtable.h b/arch/parisc/include/asm/pgtable.h index df7b931865d2..ecd028854469 100644 --- a/arch/parisc/include/asm/pgtable.h +++ b/arch/parisc/include/asm/pgtable.h @@ -192,6 +192,11 @@ extern void __update_cache(pte_t pte); #define _PAGE_PRESENT_BIT 22 /* (0x200) Software: translation valid */ #define _PAGE_HPAGE_BIT 21 /* (0x400) Software: Huge Page */ #define _PAGE_USER_BIT 20 /* (0x800) Software: User accessible page */ +#ifdef CONFIG_HUGETLB_PAGE +#define _PAGE_SPECIAL_BIT _PAGE_DMB_BIT /* DMB feature is currently unused */ +#else +#define _PAGE_SPECIAL_BIT _PAGE_HPAGE_BIT /* use unused HUGE PAGE bit */ +#endif /* N.B. The bits are defined in terms of a 32 bit word above, so the */ /* following macro is ok for both 32 and 64 bit. */ @@ -219,7 +224,7 @@ extern void __update_cache(pte_t pte); #define _PAGE_PRESENT (1 << xlate_pabit(_PAGE_PRESENT_BIT)) #define _PAGE_HUGE (1 << xlate_pabit(_PAGE_HPAGE_BIT)) #define _PAGE_USER (1 << xlate_pabit(_PAGE_USER_BIT)) -#define _PAGE_SPECIAL (_PAGE_DMB) +#define _PAGE_SPECIAL (1 << xlate_pabit(_PAGE_SPECIAL_BIT)) #define _PAGE_TABLE (_PAGE_PRESENT | _PAGE_READ | _PAGE_WRITE | _PAGE_DIRTY | _PAGE_ACCESSED) #define _PAGE_CHG_MASK (PAGE_MASK | _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_SPECIAL) diff --git a/arch/parisc/kernel/entry.S b/arch/parisc/kernel/entry.S index df8102fb435f..0e5ebfe8d9d2 100644 --- a/arch/parisc/kernel/entry.S +++ b/arch/parisc/kernel/entry.S @@ -499,6 +499,10 @@ * Finally, _PAGE_READ goes in the top bit of PL1 (so we * trigger an access rights trap in user space if the user * tries to read an unreadable page */ +#if _PAGE_SPECIAL_BIT == _PAGE_DMB_BIT + /* need to drop DMB bit, as it's used as SPECIAL flag */ + depi 0,_PAGE_SPECIAL_BIT,1,\pte +#endif depd \pte,8,7,\prot /* PAGE_USER indicates the page can be read with user privileges, @@ -529,6 +533,10 @@ * makes the tlb entry for the differently formatted pa11 * insertion instructions */ .macro make_insert_tlb_11 spc,pte,prot +#if _PAGE_SPECIAL_BIT == _PAGE_DMB_BIT + /* need to drop DMB bit, as it's used as SPECIAL flag */ + depi 0,_PAGE_SPECIAL_BIT,1,\pte +#endif zdep \spc,30,15,\prot dep \pte,8,7,\prot extru,= \pte,_PAGE_NO_CACHE_BIT,1,%r0 From 34314cd615af5036e582fad14f2bb13e4383bfe1 Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Thu, 13 Oct 2022 23:19:15 +0100 Subject: [PATCH 120/121] parisc: Fix spelling mistake "mis-match" -> "mismatch" in eisa driver There are several spelling mistakes in kernel error messages. Fix them. Signed-off-by: Colin Ian King Signed-off-by: Helge Deller --- drivers/parisc/eisa_enumerator.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/parisc/eisa_enumerator.c b/drivers/parisc/eisa_enumerator.c index f54a6f450391..f0cb31198a8f 100644 --- a/drivers/parisc/eisa_enumerator.c +++ b/drivers/parisc/eisa_enumerator.c @@ -393,7 +393,7 @@ static int parse_slot_config(int slot, } if (p0 + function_len < pos) { - printk(KERN_ERR "eisa_enumerator: function %d length mis-match " + printk(KERN_ERR "eisa_enumerator: function %d length mismatch " "got %d, expected %d\n", num_func, pos-p0, function_len); res=-1; @@ -407,13 +407,13 @@ static int parse_slot_config(int slot, } if (pos != es->config_data_length) { - printk(KERN_ERR "eisa_enumerator: config data length mis-match got %d, expected %d\n", + printk(KERN_ERR "eisa_enumerator: config data length mismatch got %d, expected %d\n", pos, es->config_data_length); res=-1; } if (num_func != es->num_functions) { - printk(KERN_ERR "eisa_enumerator: number of functions mis-match got %d, expected %d\n", + printk(KERN_ERR "eisa_enumerator: number of functions mismatch got %d, expected %d\n", num_func, es->num_functions); res=-2; } @@ -451,7 +451,7 @@ static int init_slot(int slot, struct eeprom_eisa_slot_info *es) } if (es->eisa_slot_id != id) { print_eisa_id(id_string, id); - printk(KERN_ERR "EISA slot %d id mis-match: got %s", + printk(KERN_ERR "EISA slot %d id mismatch: got %s", slot, id_string); print_eisa_id(id_string, es->eisa_slot_id); From 2130b87b2273389cafe6765bf09ef564cda01407 Mon Sep 17 00:00:00 2001 From: Nathan Chancellor Date: Fri, 14 Oct 2022 08:21:03 -0700 Subject: [PATCH 121/121] drm/amd/display: Fix build breakage with CONFIG_DEBUG_FS=n MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit After commit 8799c0be89eb ("drm/amd/display: Fix vblank refcount in vrr transition"), a build with CONFIG_DEBUG_FS=n is broken due to a misplaced brace, along the lines of: In file included from drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_trace.h:39, from drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm.c:41: drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm.c: At top level: ./include/drm/drm_atomic.h:864:9: error: expected identifier or ‘(’ before ‘for’ 864 | for ((__i) = 0; \ | ^~~ drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm.c:8317:9: note: in expansion of macro ‘for_each_new_crtc_in_state’ 8317 | for_each_new_crtc_in_state(state, crtc, new_crtc_state, j) | ^~~~~~~~~~~~~~~~~~~~~~~~~~ Move the brace within the #ifdef so that the file can be built with or without CONFIG_DEBUG_FS. Fixes: 8799c0be89eb ("drm/amd/display: Fix vblank refcount in vrr transition") Signed-off-by: Nathan Chancellor Signed-off-by: Linus Torvalds --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index f6a9e8fdd87d..c053cb79cd06 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -8310,8 +8310,8 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) crtc, dm_new_crtc_state, cur_crc_src)) DRM_DEBUG_DRIVER("Failed to configure crc source"); } -#endif } +#endif } for_each_new_crtc_in_state(state, crtc, new_crtc_state, j)