With a recent change to the output directory handling from
build_with_bazel.py, some dist directory paths incorrectly used
hyphens instead of underscores.
Fix this issue by mirroring the logic in msm_common.bzl which handles
these paths properly.
Also, run buildifier to fix format of Bazel build files.
Change-Id: Ia82073f059c54f4a7f523c5bd91afdc755c70df1
Signed-off-by: John Moon <quic_johmoo@quicinc.com>
This change changes the qmp command which sends to AOP.
Change-Id: I8b4377a0219c062c4814ae63f2ee9f07526287d6
Signed-off-by: yonliu <quic_yonliu@quicinc.com>
Currently, there is an issue where the request to execute a QOS
request is not being serviced, as requests will only be honored
if a CPU is offline.
Fix this by ensuring QOS requests are made on online CPUs.
Change-Id: I454a5ddf50e373dc6d4fe5ac06674288aa529f1e
Signed-off-by: Shaleen Agrawal <quic_shalagra@quicinc.com>
Enable thermal zone config debug driver to support reading the
entire thermal zone configurations from userspace via debugfs
interface.
Change-Id: Ide9bd09bbd07ff866ad9f8fbec1efa596ad77560
Signed-off-by: Rashid Zafar <quic_rzafar@quicinc.com>
Thermal config debugfs snapshot taken from msm-5.15 branch
'commit 8ff12fd3b6f5 ("drivers: thermal: Add support to read thermal
config rules with debugfs")'.
Updates:
- Change GPL v2 license to GPL
- Account for tz trips name change to num_trips as
changed in thermal.h
Change-Id: Id8212bad1ae374e34c0f4fb4efba20d03905720c
Signed-off-by: Rashid Zafar <quic_rzafar@quicinc.com>
Currently, there's no convenient way to discover which build options
were passed to the kernel build other than examining the output
.config file. That method is somewhat fragile as the mapping between
the CONFIG options we're parsing and the options being set by a
particular build flag could change.
Instead, let's create a new build artifact which writes the build
flags to a file which can be easily consumed later. This will allow
external module (DDK) builds to guarantee a match between the kernel
build and their build.
The output format looks like this (one opt per line):
--kasan
--lto=none
--config=stamp
--user_kmi_symbol_lists=//msm-kernel:android/abi_gki_aarch64_qcom
--ignore_missing_projects
The flags can be pulled into a separate build command line like this:
./tools/bazel build $(xargs -a build_opts.txt) //my_package:my_target
Change-Id: I162c950ff01c99e082945190b50d13b4a139f944
Signed-off-by: John Moon <quic_johmoo@quicinc.com>
During reboot/shutdown we send POWER_OFF_LONG notification.
As per spec the host should issue a power off notification
(POWER_OFF_LONG, POWER_OFF_SHORT ) if it intends to turn off both VCC
and VCCQ power supplies.
Currently we don't power_off VCCQ as it is marked as always_on.
Hence mark the is_always_on as false when power off notification is sent
and turn off both VCC and VCCQ. To check if power off notification
happened we have only one available flag EXT_CSD_NO_POWER_NOTIFICATION
hence use that flag to turn off.
Change-Id: Ifa974be837c841d78b2422bb98dff6ecbd6ec75f
Signed-off-by: Sarthak Garg <quic_sartgarg@quicinc.com>
Signed-off-by: Sachin Gupta <quic_sachgupt@quicinc.com>
Restoring CMDQ_SEND_STATUS_TRIGGER field after XO or hardware reset can
lead to CMD13 being send as soon as CQE is turned on which can lead to
below unnecessary warnings followed by register dumps.
[ 82.554113] mmc: sdhci_cmd_irq: Got command interrupt 0x00000001 even
though no command operation was in progress.
Unset that bit before restoring CMDQ_VENDOR_SPECIFIC_CFG register.
Change-Id: Id5961ecdd68b0d7cbc1c48c116f9e862a1dfc573
Signed-off-by: Sarthak Garg <sartgarg@codeaurora.org>
Signed-off-by: Sachin Gupta <quic_sachgupt@quicinc.com>
This is a squash of below two minor fixes:
8fec754f57a8 mmc: cmdq: Trigger get queue status after dcmd
327fd3225911 mmc: cqhci: CQ Register changes for v5.0 onwards.
Change-Id: I1bb6cae5a647c4ea189a992d5ace7b6c1e3a1c87
Signed-off-by: Sarthak Garg <sartgarg@codeaurora.org>
Signed-off-by: kamasali <quic_kamasali@quicinc.com>
Signed-off-by: Sachin Gupta <quic_sachgupt@quicinc.com>
The MSM SDHCI driver always set the "actual_clock" field to 0. It had
a comment about it not being needed because we weren't using the
standard SDHCI divider mechanism and we'd just fallback to
"host->clock". However, it's still better to provide the actual
clock. Why?
1. It will make timeout calculations slightly better. On one system I
have, the eMMC requets 200 MHz (for HS400-ES) but actually gets 192
MHz. These are close, but why not get the more accurate one.
2. If things are seriously off in the clock driver and it's missing
rates or picking the wrong rate (maybe it's rounding up instead of
down), this will make it much more obvious what's going on.
NOTE: we have to be a little careful here because the "actual_clock"
field shouldn't include the multiplier that sdhci-msm needs
internally.
Change-Id: I16f0ad021209e7a0c2797c695e972888b542246f
Git-Repo: https://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc.git/
Git-Commit: f16c8fd444
Suggested-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Reviewed-by: Veerabhadrarao Badiganti <vbadigan@codeaurora.org>
Signed-off-by: Sarthak Garg <sartgarg@codeaurora.org>
Signed-off-by: kamasali <quic_kamasali@quicinc.com>
Signed-off-by: Sachin Gupta <quic_sachgupt@quicinc.com>
On MDM targets we have only one path for bus voting (sdhc-ddr).
If we provide only one path and not the other one then probe will
itself fail as per the current logic.
Don't return error when bus voting paths not found.
Change-Id: Idf5401266793f4d4003391816f7511f8c0754696
Signed-off-by: Sarthak Garg <sartgarg@codeaurora.org>
Signed-off-by: Kishor Krishna Bhat <quic_kishkris@quicinc.com>
Signed-off-by: Sachin Gupta <quic_sachgupt@quicinc.com>
Support up to 4 pairs of TCSR for context doorbell queue.
Change-Id: I0dd5612fd8ecc6a5f1f9ddcaa9d4aa970b39c546
Signed-off-by: Jeyaprabu J <quic_jeyaprab@quicinc.com>
Add support to submit packet through door bell context queue.
Change-Id: Ie997b30fcfe586e063c75ace288f33d6cdc55b2c
Signed-off-by: Jeyaprabu J <quic_jeyaprab@quicinc.com>
In oom notifier, add the memory block only if Movable zone is exhausted.
This makes sure that we expand movable zone only if oom resulted from
memory pressure in movable zone.
Also fix the get_zone_free_pages to match with zoneinfo.
Change-Id: I24e4e3058922107152dee2244a9c31e476e16a69
Signed-off-by: Sudarshan Rajagopalan <quic_sudaraja@quicinc.com>
