Store the correct values that we happen to have for some A7xx SKUs in
the GPU info struct and fill out the missing information for A6xx GPUs
based on downstream kernel information.
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/611094/
[add missing entry to a615 catalog to resolve conflict]
Signed-off-by: Rob Clark <robdclark@chromium.org>
There is another cause for soft lock-up of GPU in empty ring-buffer:
race between GPU executing last commands and CPU checking ring for
emptiness. On GPU side IRQ for retire is triggered by CACHE_FLUSH_TS
event and RPTR shadow (which is used to check ring emptiness) is updated
a bit later from CP_CONTEXT_SWITCH_YIELD. Thus if GPU is executing its
last commands slow enough or we check that ring too fast we will miss a
chance to trigger switch to lower priority ring because current ring isn't
empty just yet. This can escalate to lock-up situation described in
previous patch.
To work-around this issue we keep track of last submit sequence number
for each ring and compare it with one written to memptrs from GPU during
execution of CACHE_FLUSH_TS event.
Fixes: b1fc2839d2 ("drm/msm: Implement preemption for A5XX targets")
Signed-off-by: Vladimir Lypak <vladimir.lypak@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/612047/
Signed-off-by: Rob Clark <robdclark@chromium.org>
On A5XX GPUs when preemption is used it's invietable to enter a soft
lock-up state in which GPU is stuck at empty ring-buffer doing nothing.
This appears as full UI lockup and not detected as GPU hang (because
it's not). This happens due to not triggering preemption when it was
needed. Sometimes this state can be recovered by some new submit but
generally it won't happen because applications are waiting for old
submits to retire.
One of the reasons why this happens is a race between a5xx_submit and
a5xx_preempt_trigger called from IRQ during submit retire. Former thread
updates ring->cur of previously empty and not current ring right after
latter checks it for emptiness. Then both threads can just exit because
for first one preempt_state wasn't NONE yet and for second one all rings
appeared to be empty.
To prevent such situations from happening we need to establish guarantee
for preempt_trigger to make decision after each submit or retire. To
implement this we serialize preemption initiation using spinlock. If
switch is already in progress we need to re-trigger preemption when it
finishes.
Fixes: b1fc2839d2 ("drm/msm: Implement preemption for A5XX targets")
Signed-off-by: Vladimir Lypak <vladimir.lypak@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/612045/
Signed-off-by: Rob Clark <robdclark@chromium.org>
Two fields of preempt_record which are used by CP aren't reset on
resume: "data" and "info". This is the reason behind faults which happen
when we try to switch to the ring that was active last before suspend.
In addition those faults can't be recovered from because we use suspend
and resume to do so (keeping values of those fields again).
Fixes: b1fc2839d2 ("drm/msm: Implement preemption for A5XX targets")
Signed-off-by: Vladimir Lypak <vladimir.lypak@gmail.com>
Reviewed-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/612043/
Signed-off-by: Rob Clark <robdclark@chromium.org>
Fine grain preemption (switching from/to points within submits)
requires extra handling in command stream of those submits, especially
when rendering with tiling (using GMEM). However this handling is
missing at this point in mesa (and always was). For this reason we get
random GPU faults and hangs if more than one priority level is used
because local preemption is enabled prior to executing command stream
from submit.
With that said it was ahead of time to enable local preemption by
default considering the fact that even on downstream kernel it is only
enabled if requested via UAPI.
Fixes: a7a4c19c36 ("drm/msm/a5xx: fix setting of the CP_PREEMPT_ENABLE_LOCAL register")
Signed-off-by: Vladimir Lypak <vladimir.lypak@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/612041/
Signed-off-by: Rob Clark <robdclark@chromium.org>
There are some cases, such as the one uncovered by Commit 46d4efcccc
("drm/msm/a6xx: Avoid a nullptr dereference when speedbin setting fails")
where
msm_gpu_cleanup() : platform_set_drvdata(gpu->pdev, NULL);
is called on gpu->pdev == NULL, as the GPU device has not been fully
initialized yet.
Turns out that there's more than just the aforementioned path that
causes this to happen (e.g. the case when there's speedbin data in the
catalog, but opp-supported-hw is missing in DT).
Assigning msm_gpu->pdev earlier seems like the least painful solution
to this, therefore do so.
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/602742/
Signed-off-by: Rob Clark <robdclark@chromium.org>
In adreno_request_fw() when debugging information is printed to the log
after firmware load, an incorrect filename is printed. 'newname' is used
instead of 'fwname', so prefix "qcom/" is being added to filename.
Looks like "copy-paste" mistake.
Fix this mistake by replacing 'newname' with 'fwname'.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Fixes: 2c41ef1b6f ("drm/msm/adreno: deal with linux-firmware fw paths")
Signed-off-by: Aleksandr Mishin <amishin@t-argos.ru>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/602382/
Signed-off-by: Rob Clark <robdclark@chromium.org>
Make it match the MDSS settings for sc8180x and downstream.
Note that without the previous commit that exposes the value of
macrotile_mode to mesa, this will break mesa which expects the legacy
default value of 0. Therefore we do *not* want to backport it.
Signed-off-by: Connor Abbott <cwabbott0@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/607398/
Signed-off-by: Rob Clark <robdclark@chromium.org>
According to downstream we should be setting RBBM_NC_MODE_CNTL to a
non-default value on a663 and a680, we don't support a663 and on a680
we're leaving it at the wrong (suboptimal) value. Just set it on all
GPUs. Similarly, plumb through level2_swizzling_dis which will be
necessary on a663.
ubwc_mode is expanded and renamed to ubwc_swizzle to match the name on
the display side. Similarly macrotile_mode should match the display
side.
Signed-off-by: Connor Abbott <cwabbott0@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/607397/
Signed-off-by: Rob Clark <robdclark@chromium.org>
With a7xx, we need to import a new header for each new generation and
switch to a different list of registers, instead of making
backwards-compatible changes. Using the helpers inadvertently made a750
use the a740 list of registers, instead use the family directly to fix
this.
Fixes: f3f8207d8a ("drm/msm: Add devcoredump support for a750")
Signed-off-by: Connor Abbott <cwabbott0@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/607392/
Signed-off-by: Rob Clark <robdclark@chromium.org>
-ENODEV is used to signify that there is no zap shader for the platform,
and the CPU can directly take the GPU out of secure mode. We want to
use this return code when there is no zap-shader node. But not when
there is, but without a firmware-name property. This case we want to
treat as-if the needed fw is not found.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/604564/
Disable the call to qcom_scm_gpu_init_regs() for a730 and a740
after init failures on the HDK8550 and HDK8450 platforms:
msm_dpu ae01000.display-controller: failed to load adreno gpu
msm_dpu ae01000.display-controller: failed to bind 3d00000.gpu (ops a3xx_ops [msm]): -5
msm_dpu ae01000.display-controller: adev bind failed: -5
While debugging, it happens the call to:
qcom_scm_gpu_init_regs(QCOM_SCM_GPU_ALWAYS_EN_REQ)
returns -5 and makes the gpu fail to initialize.
Remove the scm call since it's not done downstream either and
works fine without.
Fixes: 14b27d5df3 ("drm/msm/a7xx: Initialize a750 "software fuse"")
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/600972/
Signed-off-by: Rob Clark <robdclark@chromium.org>
The gpulist has twice the a6xx gpulist, replace the second one
with the a7xx gpulist.
Solves:
msm_dpu ae01000.display-controller: Unknown GPU revision: 7.3.0.1
msm_dpu ae01000.display-controller: Unknown GPU revision: 67.5.10.1
msm_dpu ae01000.display-controller: Unknown GPU revision: 67.5.20.1
on SM8450, SM8550 & SM8560.
Fixes: 8ed322f632 ("drm/msm/adreno: Split up giant device table")
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/600939/
Signed-off-by: Rob Clark <robdclark@chromium.org>
Commit f6ebff4fe810 ("drm/msm/adreno: De-spaghettify the use of memory
barriers") made some fixups relating to write arrival, ensuring that
the GPU's memory interface has *really really really* been told to come
out of reset. That in turn rendered the hacky commit being reverted no
longer necessary.
Get rid of it.
This reverts commit b77532803d ("drm/msm/a6xx: Poll for GBIF unhalt
status in hw_init")
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/600870/
Reviewed-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Signed-off-by: Rob Clark <robdclark@chromium.org>
Memory barriers help ensure instruction ordering, NOT time and order
of actual write arrival at other observers (e.g. memory-mapped IP).
On architectures employing weak memory ordering, the latter can be a
giant pain point, and it has been as part of this driver.
Moreover, the gpu_/gmu_ accessors already use non-relaxed versions of
readl/writel, which include r/w (respectively) barriers.
Replace the barriers with a readback (or drop altogether where possible)
that ensures the previous writes have exited the write buffer (as the CPU
must flush the write to the register it's trying to read back).
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/600869/
Reviewed-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Signed-off-by: Rob Clark <robdclark@chromium.org>
On all Qualcomm platforms with a7xx GPUs, qcom_scm provides a method to
initialize cx_mem. Copy this from downstream (minus BCL which we
currently don't support). On a750, this includes a new "fuse" register
which can be used by qcom_scm to fuse off certain features like
raytracing in software. The fuse is default off, and is initialized by
calling the method. Afterwards we have to read it to find out which
features were enabled.
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Connor Abbott <cwabbott0@gmail.com>
Reviewed-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/592042/
Signed-off-by: Rob Clark <robdclark@chromium.org>
GCC diagnostic pragma method throws below warnings in some of the versions
drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c:16:9: warning: unknown
option after '#pragma GCC diagnostic' kind [-Wpragmas]
#pragma GCC diagnostic ignored "-Wunused-const-variable"
^
In file included from drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c:18:0:
drivers/gpu/drm/msm/adreno/adreno_gen7_0_0_snapshot.h:924:19: warning:
'gen7_0_0_external_core_regs' defined but not used [-Wunused-variable]
static const u32 *gen7_0_0_external_core_regs[] = {
^
In file included from drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c:19:0:
drivers/gpu/drm/msm/adreno/adreno_gen7_2_0_snapshot.h:748:19: warning:
'gen7_2_0_external_core_regs' defined but not used [-Wunused-variable]
static const u32 *gen7_2_0_external_core_regs[] = {
^
In file included from drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c:20:0:
drivers/gpu/drm/msm/adreno/adreno_gen7_9_0_snapshot.h:1188:43: warning:
'gen7_9_0_sptp_clusters' defined but not used [-Wunused-variable]
static struct gen7_sptp_cluster_registers gen7_9_0_sptp_clusters[] = {
^
drivers/gpu/drm/msm/adreno/adreno_gen7_9_0_snapshot.h:1438:19: warning:
'gen7_9_0_external_core_regs' defined but not used [-Wunused-variable]
static const u32 *gen7_9_0_external_core_regs[] = {
Remove GCC version dependency by using __unused__ for the unused gen7_* includes.
Changes in v2:
- Fix the warnings in the commit text
- Use __attribute((__unused__)) instead of local assignment
changes in v3:
- drop the Link from the auto add
changes in v4:
- replace __attribute((__unused__)) with __always_unused
Fixes: 64d6255650 ("drm/msm: More fully implement devcoredump for a7xx")
Suggested-by: Rob Clark <robdclark@chromium.org>
Signed-off-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/597265/
[Add gen7_9_0_cx_debugbus_blocks as well]
Signed-off-by: Rob Clark <robdclark@chromium.org>
These tables were made non-const in commit 3cba4a2cdf ("drm/msm/a6xx:
Update ROQ size in coredump") in order to avoid powering up the GPU when
reading back a devcoredump. Instead let's just stash the count that is
potentially read from hw in struct a6xx_gpu_state_obj, and make the
tables const again.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/592699/
Calling a6xx_destroy() before adreno_gpu_init() leads to a null pointer
dereference on:
msm_gpu_cleanup() : platform_set_drvdata(gpu->pdev, NULL);
as gpu->pdev is only assigned in:
a6xx_gpu_init()
|_ adreno_gpu_init
|_ msm_gpu_init()
Instead of relying on handwavy null checks down the cleanup chain,
explicitly de-allocate the LLC data and free a6xx_gpu instead.
Fixes: 76efc2453d ("drm/msm/gpu: Fix crash during system suspend after unbind")
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/588919/
Signed-off-by: Rob Clark <robdclark@chromium.org>
