- 07 Oct, 2018 4 commits
-
-
Jordan Crouse authored
The CP performance counter selects were accidentally marked as protected so they couldn't be written from PM4 streams. Remove the protection because user space does have an interest in setting up their own counters. Signed-off-by:
Jordan Crouse <jcrouse@codeaurora.org> Signed-off-by:
Rob Clark <robdclark@gmail.com>
-
Sean Paul authored
We are currently leaking a drm_crtc_commit struct for every atomic commit containing plane state. The dpu plane destroy function cleans up the fb reference manually, but fails to release the commit ref. As a result, we just keep allocating drm_crtc_commits without ever freeing them. Fortunately there's a helper function which will clean up all of our mess at once, so use that. Thanks to Doug Anderson for reporting the memory leak (and leaving breadcrumbs from kmemleak!). Reported-by:
Doug Anderson <dianders@chromium.org> Signed-off-by:
Sean Paul <seanpaul@chromium.org> Signed-off-by:
-
Sean Paul authored
This patch uses the proper do_div() macro to perform u64 division and guards against overflow if the result is too large for the unsigned long return type Fixes: de0a3d09 drm/msm: re-factor devfreq code Cc: Sharat Masetty <smasetty@codeaurora.org> Signed-off-by:
-
Rob Clark authored
Signed-off-by:
-
- 04 Oct, 2018 36 commits
-
-
Sharat Masetty authored
Implement routines to estimate GPU busy time and fetching the current frequency for the polling interval. This is required by the devfreq framework which recommends a frequency change if needed. The driver code then tries to set this new frequency on the GPU by sending an Out Of Band(OOB) request to the GMU. Signed-off-by:
Sharat Masetty <smasetty@codeaurora.org> Signed-off-by:
-
Sharat Masetty authored
The devfreq framework requires the drivers to provide busy time estimations. The GPU driver relies on the hardware performance counteres for the busy time estimations, but different hardware revisions have counters which can be sourced from different clocks. So the busy time estimation will be target dependent. Additionally on targets where the clocks are completely controlled by the on chip microcontroller, fetching and setting the current GPU frequency will be different. This patch aims to embrace these differences by re-factoring the devfreq code a bit. Signed-off-by:
-
Sharat Masetty authored
Add a simple function to read 64 registers in the GMU domain Signed-off-by:
-
Sharat Masetty authored
Devfreq turns on and starts recommending power level as soon as it is initialized. The GPU is still not powered on by the time the devfreq init happens and this leads to problems on GPU's where register access is needed to get/set power levels. So we start suspended and only restart devfreq when GPU is powered on. Signed-off-by:
-
Jordan Crouse authored
The target definition for a630 didn't set a reasonable value for inactive_period so it defaulted to zero and we were essentially powering down after every submission. Set it back to the default value to keep the GPU from bouncing too much during regular workloads. Signed-off-by:
-
Thomas Zimmermann authored
This patch unifies the naming of DRM functions for reference counting of struct drm_device. The resulting code is more aligned with the rest of the Linux kernel interfaces. Signed-off-by:
Thomas Zimmermann <tzimmermann@suse.de> Signed-off-by:
-
Thomas Zimmermann authored
This patch unifies the naming of DRM functions for reference counting of struct drm_gem_object. The resulting code is more aligned with the rest of the Linux kernel interfaces. Signed-off-by:
-
Thomas Zimmermann authored
This patch unifies the naming of DRM functions for reference counting of struct drm_framebuffer. The resulting code is more aligned with the rest of the Linux kernel interfaces. Signed-off-by:
-
Jordan Crouse authored
The only HFI communication with the GMU on sdm845 happens during initialization and all commands are synchronous. A fancy interrupt tasklet and associated infrastructure is entirely not eeded and puts us at the mercy of the scheduler. Instead poll for the message signal and handle the response immediately and go on our way. Signed-off-by:
-
Jordan Crouse authored
The point of the 'force_dma' parameter for of_dma_configure is to force the device to be set up even if DMA capability is not described by the firmware which is exactly the use case we have for GMU - we need SMMU to get set up but we have no other dma capabilities since memory is managed by the GPU driver. Currently we pass false so of_dma_configure() fails and subsequently GMU and GPU probe does as well. Fixes: 4b565ca5 ("drm/msm: Add A6XX device support") Signed-off-by:
Sibi Sankar <sibis@codeaurora.org> Signed-off-by:
-
Sharat Masetty authored
In the case where preemption is not enabled, this patch simply skips preemption related initialization in hardware init sequence. Signed-off-by:
-
Colin Ian King authored
The return from the call to _mixer_stages can be a negative error code however this is being assigned to an unsigned variable 'stages' hence the check is always false. Fix this by making 'stages' an int. Detected by Coccinelle ("Unsigned expression compared with zero: stages < 0") Fixes: 25fdd593 ("drm/msm: Add SDM845 DPU support") Signed-off-by:Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
Jordan Crouse authored
The current design greedily takes a big chunk of the PDC register space instead of just the GPU specific sections which conflicts with other drivers and generally makes a mess of things. Furthermore we only need to map the GPU PDC sections just once during init so map the memory inside the function that uses it and adjust the pointers and register offsets accordingly. Signed-off-by:
-
Jordan Crouse authored
>From the review for the DT bindings for the GPU/GMU it was suggested that the phandle for the GMU be 'qcom,gmu' instead of just 'gmu'. Signed-off-by:
-
Anders Roxell authored
When CONFIG_DEV_COREDUMP isn't defined msm_gpu_crashstate_capture doesn't pass the correct parameters. drivers/gpu/drm/msm/msm_gpu.c: In function ‘recover_worker’: drivers/gpu/drm/msm/msm_gpu.c:479:34: error: passing argument 2 of ‘msm_gpu_crashstate_capture’ from incompatible pointer type [-Werror=incompatible-pointer-types] msm_gpu_crashstate_capture(gpu, submit, comm, cmd); ^~~~~~ drivers/gpu/drm/msm/msm_gpu.c:388:13: note: expected ‘char *’ but argument is of type ‘struct msm_gem_submit *’ static void msm_gpu_crashstate_capture(struct msm_gpu *gpu, char *comm, ^~~~~~~~~~~~~~~~~~~~~~~~~~ drivers/gpu/drm/msm/msm_gpu.c:479:2: error: too many arguments to function ‘msm_gpu_crashstate_capture’ msm_gpu_crashstate_capture(gpu, submit, comm, cmd); ^~~~~~~~~~~~~~~~~~~~~~~~~~ drivers/gpu/drm/msm/msm_gpu.c:388:13: note: declared here static void msm_gpu_crashstate_capture(struct msm_gpu *gpu, char *comm, In current code the function msm_gpu_crashstate_capture parameters. Fixes: cdb95931 ("drm/msm/gpu: Add the buffer objects from the submit to the crash dump") Signed-off-by:Anders Roxell <anders.roxell@linaro.org> Reviewed-By:
-
Sharat Masetty authored
The index of the perf table was being set in the wrong bit position in the register. With this fix, the GPU clock can be seen running at desired frequency. Signed-off-by:
-
Rob Clark authored
The [v]snprintf() functions return the size that *would have* been written into the buffer, rather than the size *actually* written. Which results in us trying to memcpy() past the end of the stack. What we really want is [v]scnprintf(). Signed-off-by: -
Sean Paul authored
cur_master in dpu_encoder is assigned at modeset and cleared on .disable(). Unfortunately dpms (or enable/disable) does not guarantee a modeset, so cur_master is NULL when we try to re-enable it. This patch moves the NULL assignment to setup_display where it will be re-assigned later in the function. Tested-by:
Bruce Wang <bzwang@chromium.org> Reviewed-by:
Jeykumar Sankaran <jsanka@codeaurora.org> Signed-off-by:
-
Bruce Wang authored
Remove unneeded checks from _dpu_plane_get_aspace. v3: change _dpu_plane_get_aspace to return a struct *msm_gem_address_space instead passing in a pointer of the same type to edit. Remove uneeded arguments. Reviewed-by:
-
Bruce Wang authored
Removed all impossible checks from the function, which eliminates the need for a return value. This function is also never used outside of dpu_plane.c, so the function is made static. v3: Using helper function _dpu_plane_get_kms() instead of doing it locally. Signed-off-by:
-
Bruce Wang authored
Removes redundant tests for _dpu_crtc_vblank_enable_no_lock. Function return type is now void and all function calls have been changed accordingly. Signed-off-by:
-
Bruce Wang authored
All checks for _dpu_crtc_power_enable are not true, so the function can never return an error code. This removes the need for the function as pm_runtime functions can be used instead. v3: Separated _dpu_crtc_power_enable into _dpu_crtc_power_enable and _dpu_crtc_power_disable for clarity. v4: Removed both _dpu_crtc_power_enable and _dpu_crtc_power_disable and called pm_runtime_get_sync and pm_runtime_put_sync from all call points Reviewed-by:
-
Bruce Wang authored
Removes impossible checks in dpu_crtc.c. Variable assignments are moved up to be initializations where possible. Some variables are no longer used, these are removed. v3: reverted back to original patch Reviewed-by:
-
Bruce Wang authored
Removes unnecessary checks from dpu_plane_atomic_disable, old_state argument for both dpu_plane_atomic_disable and dpu_plane_sspp_atomic_update is removed as it is no longer used. Reviewed-by:
-
Bruce Wang authored
Removes some checks from dpu_plane.c that will never result in an error. Subsequent variable assignments become part of the initialization wherever possible. Unused variables are removed. v3: removed additional impossible checks and called helper function _dpu_plane_get_kms() where possible. Reviewed-by:
-
Sean Paul authored
TP_printk is not synchronous, so storing pointers and then later dereferencing them is a Bad Idea. This patch stores everything locally to avoid display stomped memory. Reviewed-by:
Abhinav Kumar <abhinavk@codeaurora.org> [seanpaul fixed up commit msg typo on apply] Signed-off-by:
-
Sean Paul authored
It's useful to know which bits of the flush come from extra_flush_bits Reviewed-by:
-
Sean Paul authored
We're printing the frame_busy_mask in a trace, but after it's been cleared. This, as it turns out, is pretty pointless. Reviewed-by:
-
Sean Paul authored
There's no benefit in falling out of the if, just return directly. Changes in v2: - None Reviewed-by:
-
Sean Paul authored
All of the checks in dpu_plane_sspp_atomic_update() are impossible, so remove them and make the function void. This removes the need to error check in dpu_plane_atomic_update(). Additionally, remove impossible checks in dpu_plane_atomic_update(). Changes in v2: - None Reviewed-by:
-
Sean Paul authored
plane->state->visible encompasses all of these checks and more, so we can just check visible. Changes in v2: - None Reviewed-by:
-
Sean Paul authored
It's doing the same thing dpu_plane_enabled() is. Changes in v2: - None Reviewed-by:
-
Sean Paul authored
dpu_plane_atomic_check() is a very thin wrapper around dpu_plane_sspp_atomic_check(). All it does is a NULL-check of state->fb, which is already done by drm_atomic_helper_check_plane_state(). Further, the helper sets state->visible = false when this is true. So remove dpu_plane_atomic_check() and just use dpu_plane_sspp_atomic_check() directly. Changes in v2: - Fix spelling mistake in Subject (Jeykumar) Reviewed-by:
-
Sean Paul authored
src/dst rects are checked in both atomic_check and atomic_update, with the more comprehensive check occurring in atomic_update, which is backwards. So consolodate the checks in atomic_check. Changes in v2: - Use the correct crtc state (Jeykumar) Cc: Jeykumar Sankaran <jsanka@codeaurora.org> Reviewed-by:
-
Sean Paul authored
This patch removes some checks which are impossible to hit. As a result, we can move some of the local var assignments into the declarations. Changes in v2: - None Reviewed-by:
-
Jordan Crouse authored
enum dpu_ad isn't used and can be safely removed. Reviewed-by:
-
