- 23 Jun, 2021 40 commits
-
-
Yangtao Li authored
Use resource-managed OPP API to simplify code. Signed-off-by:
Yangtao Li <tiny.windzz@gmail.com> Signed-off-by:
Dmitry Osipenko <digetx@gmail.com> Link: https://lore.kernel.org/r/20210314163408.22292-12-digetx@gmail.comReviewed-by:
Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by:
Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Signed-off-by:
Rob Clark <robdclark@chromium.org>
-
Bhaskar Chowdhury authored
s/modueles/modules/ ....two different places Signed-off-by:
Bhaskar Chowdhury <unixbhaskar@gmail.com> Link: https://lore.kernel.org/r/20210318062650.19886-1-unixbhaskar@gmail.comAcked-by:
Randy Dunlap <rdunlap@infradead.org> Reviewed-by:
Stephen Boyd <swboyd@chromium.org> Signed-off-by:
-
Bhaskar Chowdhury authored
s/poiner/pointer/ Signed-off-by:
-
Bhaskar Chowdhury authored
s/struture/structure/ Signed-off-by:
-
Zhen Lei authored
The error code returned by platform_get_irq() is stored in 'irq', it's forgotten to be copied to 'ret' before being returned. As a result, the value 0 of 'ret' is returned incorrectly. After the above fix is completed, initializing the local variable 'ret' to 0 is no longer needed, remove it. In addition, when dpu_mdss_init() is successfully returned, the value of 'ret' is always 0. Therefore, replace "return ret" with "return 0" to make the code clearer. Fixes: 070e64dc ("drm/msm/dpu: Convert to a chained irq chip") Reported-by:
Hulk Robot <hulkci@huawei.com> Signed-off-by:
Zhen Lei <thunder.leizhen@huawei.com> Link: https://lore.kernel.org/r/20210510063805.3262-2-thunder.leizhen@huawei.comReviewed-by:
-
Zhen Lei authored
Fix to return a negative error code from the error handling case instead of 0, as done elsewhere in this function. Fixes: 7f9743ab ("drm/msm: validate display and event threads") Reported-by:
-
Rob Clark authored
Syncing up with -rc6 fixes to avoid conflicts with a660 patches. Signed-off-by: -
Krishna Manikandan authored
Crtc perf update from frame event work can result in wrong bandwidth and clock update from dpu if the work is scheduled after the swap state has happened. Avoid such issues by moving perf update to complete commit once the frame is accepted by the hardware. Fixes: a29c8c02 ("drm/msm/disp/dpu1: fix display underruns during modeset") Signed-off-by:
Krishna Manikandan <mkrishn@codeaurora.org> Tested-by:
Douglas Anderson <dianders@chromium.org> Link: https://lore.kernel.org/r/1622092076-5100-1-git-send-email-mkrishn@codeaurora.orgSigned-off-by:
-
Stephen Boyd authored
These prints flood the logs with drm debugging set to enable kms and driver logging (DRM_UT_KMS and DRM_UT_DRIVER). Let's move these prints to the atomic bucket (DRM_UT_ATOMIC) as they're related to the atomic paths. Cc: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Cc: Abhinav Kumar <abhinavk@codeaurora.org> Cc: Kuogee Hsieh <khsieh@codeaurora.org> Cc: aravindh@codeaurora.org Cc: Sean Paul <sean@poorly.run> Signed-off-by:
Abhinav Kumar <abhinavk@codeaurora.org> Signed-off-by:
-
Stephen Boyd authored
Use the DPU_DEBUG_PLANE() helper to print the plane number instead of open coding it. Cc: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Cc: Abhinav Kumar <abhinavk@codeaurora.org> Cc: Kuogee Hsieh <khsieh@codeaurora.org> Cc: aravindh@codeaurora.org Cc: Sean Paul <sean@poorly.run> Signed-off-by:
-
Stephen Boyd authored
These are verbose prints that tell us about the framebuffer state. Let's move them to drm_dbg_state() so that they're only printed if we're interested in verbose state logging while drm debugging. Cc: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Cc: Abhinav Kumar <abhinavk@codeaurora.org> Cc: Kuogee Hsieh <khsieh@codeaurora.org> Cc: aravindh@codeaurora.org Cc: Sean Paul <sean@poorly.run> Signed-off-by:
-
Stephen Boyd authored
This print is missing a newline, and doesn't really provide any value. Drop it. Cc: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Cc: Abhinav Kumar <abhinavk@codeaurora.org> Cc: Kuogee Hsieh <khsieh@codeaurora.org> Cc: aravindh@codeaurora.org Cc: Sean Paul <sean@poorly.run> Signed-off-by:
-
Stephen Boyd authored
Put these debug prints in the vblank code into the appropriate vblank category via drm_dbg_vbl(). Cc: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Cc: Abhinav Kumar <abhinavk@codeaurora.org> Cc: Kuogee Hsieh <khsieh@codeaurora.org> Cc: aravindh@codeaurora.org Cc: Sean Paul <sean@poorly.run> Signed-off-by:
-
Dmitry Baryshkov authored
Merge dpu_core_irq_enable() into dpu_core_irq_register_callback() and dpu_core_irq_disable() into dpu_core_irq_unregister_callback(), because they are called in pairs. There is no need to have separate enable/disable pair, we can enable hardware IRQ when first callback is registered and when the last callback is unregistered. Signed-off-by:
-
Dmitry Baryshkov authored
There is no more need for the dpu_intr_type types, dpu_irq_map table, individual intr defines and obsolete_irq field. Drop all of them now. Signed-off-by:
-
Dmitry Baryshkov authored
The IRQ table in the dpu_hw_interrupts.h is big, ugly, and hard to maintain. There are only few interrupts used from that table. Newer generations use different IRQ locations. Move this data to hw catalog. Signed-off-by:
-
Dmitry Baryshkov authored
In order to make mdss_irqs readable (and error-prone) define names for interrupt register indices. Signed-off-by:
-
Dmitry Baryshkov authored
Always call dpu_hw_intr_clear_intr_status_nolock() from the dpu_hw_intr_dispatch_irqs(). This simplifies the callback function (which call clears the interrupts anyway) and enforces clearing the hw interrupt status. Signed-off-by:
Bjorn Andersson <bjorn.andersson@linaro.org> Reviewed-by:
-
Dmitry Baryshkov authored
There is little sense in reading interrupt statuses and right after that going after the array of statuses to dispatch them. Merge both loops into single function doing read and dispatch. Signed-off-by:
-
Stephen Boyd authored
Let's look at the irq status bits after a transfer and see if we got a nack or a defer or a timeout, instead of telling drm layers that everything was fine, while still printing an error message. I wasn't sure about NACK+DEFER so I lumped all those various errors along with a nack so that the drm core can figure out that things are just not going well. The important thing is that we're now returning -ETIMEDOUT when the message times out and nacks for bad addresses. Cc: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Cc: Abhinav Kumar <abhinavk@codeaurora.org> Cc: Kuogee Hsieh <khsieh@codeaurora.org> Cc: aravindh@codeaurora.org Cc: Sean Paul <sean@poorly.run> Signed-off-by:
Kuogee Hsieh <khsieh@codeaurora.org> Link: https://lore.kernel.org/r/20210507212505.1224111-4-swboyd@chromium.orgSigned-off-by:
-
Stephen Boyd authored
We don't need to hold the lock to inspect the message we're going to transfer, and we don't need to clear the busy flag either. Take the lock later and bail out earlier if conditions aren't met. Cc: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Cc: Abhinav Kumar <abhinavk@codeaurora.org> Cc: Kuogee Hsieh <khsieh@codeaurora.org> Cc: aravindh@codeaurora.org Cc: Sean Paul <sean@poorly.run> Signed-off-by:
-
Stephen Boyd authored
We don't need to stash away 'isr' in the aux structure to pass to two functions. Let's use a local variable instead. And we can complete the completion variable in one place instead of two to simplify the code. Cc: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Cc: Abhinav Kumar <abhinavk@codeaurora.org> Cc: Kuogee Hsieh <khsieh@codeaurora.org> Cc: aravindh@codeaurora.org Cc: Sean Paul <sean@poorly.run> Signed-off-by:
-
Dmitry Baryshkov authored
Fix following warnings generated when either DP or DSI support is disabled: drivers/gpu/drm/msm/disp/msm_disp_snapshot_util.c:141:3: error: implicit declaration of function 'msm_dp_snapshot'; did you mean 'msm_dsi_snapshot'? [-Werror=implicit-function-declaration] drivers/gpu/drm/msm/msm_kms.h:127:26: warning: 'struct msm_disp_state' declared inside parameter list will not be visible outside of this definition or declaration drivers/gpu/drm/msm/disp/dpu1/dpu_kms.c:867:21: error: initialization of 'void (*)(struct msm_disp_state *, struct msm_kms *)' from incompatible pointer type 'void (*)(struct msm_disp_state *, struct msm_kms *)' [-Werror=incompatible-pointer-types] drivers/gpu/drm/msm/dsi/dsi.h:94:30: warning: 'struct msm_disp_state' declared inside parameter list will not be visible outside of this definition or declaration Reported-by:
kernel test robot <lkp@intel.com> Cc: Abhinav Kumar <abhinavk@codeaurora.org> Fixes: 1c3b7ac1a71d ("drm/msm: pass dump state as a function argument") Signed-off-by:
-
Jonathan Marek authored
There shouldn't be any reason to ever use uncached over writecombine, so just use writecombine for MSM_BO_UNCACHED. Note: userspace never used MSM_BO_UNCACHED anyway Signed-off-by:
Jonathan Marek <jonathan@marek.ca> Acked-by:
Jordan Crouse <jordan@cosmicpenguin.net> Link: https://lore.kernel.org/r/20210423190833.25319-6-jonathan@marek.caSigned-off-by:
-
Jonathan Marek authored
Add a new cache mode for creating coherent host-cached BOs. Signed-off-by:
Jordan Crouse <jcrouse@codeaurora.org> Link: https://lore.kernel.org/r/20210423190833.25319-5-jonathan@marek.caSigned-off-by:
-
Jonathan Marek authored
Use the same logic as the userspace mapping. This fixes msm_rd with cached BOs. Signed-off-by:
-
Jonathan Marek authored
msm_gem_get_vaddr() currently always maps as writecombine, so use the right flag instead of relying on broken behavior (things don't actually work if they are mapped as uncached). Signed-off-by:
-
Jonathan Marek authored
No one knows what this is for anymore, so just remove it. Signed-off-by:
-
Dmitry Baryshkov authored
Add DSI PHY registers to the msm state snapshots to be able to check their contents. Signed-off-by:
-
Dmitry Baryshkov authored
Instead of looping throught the resources each time to get the DSI CTRL area size, get it at the ioremap time. Signed-off-by:
-
Dmitry Baryshkov authored
Instead of allocating snapshotting structure at the driver probe time and later handling concurrent access, actual state, etc, make msm_disp_state transient struct. Allocate one when snapshotting happens and free it after coredump data is read by userspace. Signed-off-by:
-
Dmitry Baryshkov authored
Instead of always getting the disp_state from drm device, pass it as an argument. Signed-off-by:
-
Abhinav Kumar authored
Add snapshot points across dpu driver to trigger dumps when critical errors are hit. changes in v5: - change the callers to use the snapshot function directly Signed-off-by:
-
Abhinav Kumar authored
Add support to take the register snapshot of dsi, dp and dpu modules. Signed-off-by:
-
Abhinav Kumar authored
Add an API to take a snapshot of DPU controller registers. This API will be used by the msm_disp_snapshot module to capture the DPU snapshot. Signed-off-by:
-
Abhinav Kumar authored
Add an API to take a snapshot of DP controller registers. This API will be used by the msm_disp_snapshot module to capture the DP snapshot. Signed-off-by:
-
Abhinav Kumar authored
Add an API to take a snapshot of DSI controller registers. This API will be used by the msm_disp_snapshot module to capture the DSI snapshot. Signed-off-by:
-
Abhinav Kumar authored
Add the msm_disp_snapshot module which adds supports to dump dpu registers and capture the drm atomic state which can be used in case of error conditions. changes in v5: - start storing disp_state in msm_kms instead of dpu_kms - get rid of MSM_DISP_SNAPSHOT_IN_* enum by simplifying the functions - move snprintf inside the snapshot core by using varargs - get rid of some stale code comments - allow snapshot module for non-DPU targets Reported-by:
-
Abhinav Kumar authored
Currently drm_atomic_print_state() internally allocates and uses a drm_info printer. Allow it to accept any drm_printer type so that the API can be leveraged even for taking drm snapshot. Rename the drm_atomic_print_state() to drm_atomic_print_new_state() so that it reflects its functionality better. changes in v5: - none Reported-by:
Daniel Vetter <daniel.vetter@ffwll.ch> Reviewed-by:
-
Kuogee Hsieh authored
irq_hpd interrupt should be handled after dongle plugged in and before dongle unplugged. Hence irq_hpd interrupt is enabled at the end of the plugin handle and disabled at the beginning of unplugged handle. Current irq_hpd with sink_count = 0 is wrongly handled same as the dongle unplugged which tears down the mainlink and disables the phy. This patch fixes this problem by only tearing down the mainlink but keeping phy enabled at irq_hpd with sink_count = 0 handle so that next irq_hpd with sink_count =1 can be handled by setup mainlink only. This patch also set dongle into D3 (power off) state at end of handling irq_hpd with sink_count = 0. Changes in v2: -- add ctrl->phy_Power_count Changes in v3: -- del ctrl->phy_Power_count -- add phy_power_off to dp_ctrl_off_link_stream() Changes in v4: -- return immediately if clock disable failed at dp_ctrl_off_link_stream() Changes in v5: -- set dongle to D3 (power off) state at dp_ctrl_off_link_stream() Changes in v6: -- add Fixes tag Fixes: ea9f337c ("drm/msm/dp: reset dp controller only at boot up and pm_resume") Signed-off-by:
-
