In V3D, the conclusion of a job is indicated by a IRQ. When a job
finishes, then we update the local and the global GPU stats of that
queue. But, while the GPU stats are being updated, a user might be
reading the stats from sysfs or fdinfo.
For example, on `gpu_stats_show()`, we could think about a scenario where
`v3d->queue[queue].start_ns != 0`, then an interrupt happens, we update
the value of `v3d->queue[queue].start_ns` to 0, we come back to
`gpu_stats_show()` to calculate `active_runtime` and now,
`active_runtime = timestamp`.
In this simple example, the user would see a spike in the queue usage,
that didn't match reality.
In order to address this issue properly, use a seqcount to protect read
and write sections of the code.
Fixes: 09a93cc4f7 ("drm/v3d: Implement show_fdinfo() callback for GPU usage stats")
Reported-by: Tvrtko Ursulin <tursulin@igalia.com>
Signed-off-by: Maíra Canal <mcanal@igalia.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240420213632.339941-7-mcanal@igalia.com
This will make it easier to instantiate the GPU stats variables and it
will create a structure where we can store all the variables that refer
to GPU stats.
Note that, when we created the struct `v3d_stats`, we renamed
`jobs_sent` to `jobs_completed`. This better express the semantics of
the variable, as we are only accounting jobs that have been completed.
Signed-off-by: Maíra Canal <mcanal@igalia.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Reviewed-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240420213632.339941-4-mcanal@igalia.com
Currently, we manually perform all operations to update the GPU stats
variables. Apart from the code repetition, this is very prone to errors,
as we can see on commit 35f4f8c9fc ("drm/v3d: Don't increment
`enabled_ns` twice").
Therefore, create two functions to manage updating all GPU stats
variables. Now, the jobs only need to call for `v3d_job_update_stats()`
when the job is done and `v3d_job_start_stats()` when starting the job.
Co-developed-by: Tvrtko Ursulin <tursulin@igalia.com>
Signed-off-by: Tvrtko Ursulin <tursulin@igalia.com>
Signed-off-by: Maíra Canal <mcanal@igalia.com>
Reviewed-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240420213632.339941-3-mcanal@igalia.com
The tc358775 bridge is pin compatible with earlier tc358765 according to
the tc358774xbg_datasheet_en_20190118.pdf documentation. Compared to the
tc358765, the tc358775 supports a STBY GPIO and higher data rates.
The tc358765 has a register bit for video event mode vs video pulse mode.
We must set it to video event mode for the LCD output to work, and on the
tc358775, this bit no longer exists.
Looks like the registers seem to match otherwise based on a quick glance
comparing the defines to the earlier Android kernel tc358765 driver.
Reviewed-by: Michael Walle <mwalle@kernel.org>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Robert Foss <rfoss@kernel.org>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240225062008.33191-10-tony@atomide.com
The current code assumes the data-lanes property is configured on the
DSI host side instead of the bridge side, and assumes DSI host endpoint 1.
Let's standardize on what the other bridge drivers are doing and parse the
data-lanes property for the bridge. Only if data-lanes property is not found,
let's be nice and also check the DSI host for old dtb in use and warn.
And as Dmitry pointed out, the lanes for the host and the bridge may be
different because the lanes may be swapped on the host side.
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Michael Walle <mwalle@kernel.org>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Robert Foss <rfoss@kernel.org>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240225062008.33191-7-tony@atomide.com
The bridge always uses 24bpp internally. Therefore, for jeida-18
mapping we need to discard the lowest two bits for each channel and thus
starting with LV_[RGB]2. jeida-24 has the same mapping but uses four
lanes instead of three, with the forth pair transmitting the lowest two
bits of each channel. Thus, the mapping between jeida-18 and jeida-24
is actually the same, except that one channel is turned off (by
selecting the RGB666 format in VPCTRL).
While at it, remove the bogus comment about the hardware default because
the default is overwritten in any case.
Tested with a jeida-18 display (Evervision VGG644804).
Fixes: b26975593b ("display/drm/bridge: TC358775 DSI/LVDS driver")
Signed-off-by: Michael Walle <mwalle@kernel.org>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Robert Foss <rfoss@kernel.org>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240225062008.33191-5-tony@atomide.com
The device uses a clock lane, and 1 to 4 DSI data lanes. Let's add the
data-lanes property starting at 1 similar to what the other bridge
bindings are doing.
Let's also drop the data-lanes properties in the example for the DSI host
controller to avoid confusion. The configuration of the DSI host depends
on the controller used and is unrelated to the bridge binding.
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240225062008.33191-3-tony@atomide.com
For a normal operation, the stby GPIO is not needed.
The reset pin is required because once the PPI (PHY protocol interface)
is started, it can only be stopped by asserting the reset pin.
Signed-off-by: Michael Walle <mwalle@kernel.org>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
[tony@atomide.com: dropped regulator related changes]
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240225062008.33191-2-tony@atomide.com
Add clear page support in vram memory region.
v1(Christian):
- Dont handle clear page as TTM flag since when moving the BO back
in from GTT again we don't need that.
- Make a specialized version of amdgpu_fill_buffer() which only
clears the VRAM areas which are not already cleared
- Drop the TTM_PL_FLAG_WIPE_ON_RELEASE check in
amdgpu_object.c
v2:
- Modify the function name amdgpu_ttm_* (Alex)
- Drop the delayed parameter (Christian)
- handle amdgpu_res_cleared(&cursor) just above the size
calculation (Christian)
- Use AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE for clearing the buffers
in the free path to properly wait for fences etc.. (Christian)
v3(Christian):
- Remove buffer clear code in VRAM manager instead change the
AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE handling to set
the DRM_BUDDY_CLEARED flag.
- Remove ! from amdgpu_res_cleared(&cursor) check.
v4(Christian):
- vres flag setting move to vram manager file
- use dma_fence_get_stub in amdgpu_ttm_clear_buffer function
- make fence a mandatory parameter and drop the if and the get/put dance
Signed-off-by: Arunpravin Paneer Selvam <Arunpravin.PaneerSelvam@amd.com>
Suggested-by: Christian König <christian.koenig@amd.com>
Acked-by: Felix Kuehling <felix.kuehling@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240419063538.11957-2-Arunpravin.PaneerSelvam@amd.com
Signed-off-by: Christian König <christian.koenig@amd.com>
- Add tracking clear page feature.
- Driver should enable the DRM_BUDDY_CLEARED flag if it
successfully clears the blocks in the free path. On the otherhand,
DRM buddy marks each block as cleared.
- Track the available cleared pages size
- If driver requests cleared memory we prefer cleared memory
but fallback to uncleared if we can't find the cleared blocks.
when driver requests uncleared memory we try to use uncleared but
fallback to cleared memory if necessary.
- When a block gets freed we clear it and mark the freed block as cleared,
when there are buddies which are cleared as well we can merge them.
Otherwise, we prefer to keep the blocks as separated.
- Add a function to support defragmentation.
v1:
- Depends on the flag check DRM_BUDDY_CLEARED, enable the block as
cleared. Else, reset the clear flag for each block in the list(Christian)
- For merging the 2 cleared blocks compare as below,
drm_buddy_is_clear(block) != drm_buddy_is_clear(buddy)(Christian)
- Defragment the memory beginning from min_order
till the required memory space is available.
v2: (Matthew)
- Add a wrapper drm_buddy_free_list_internal for the freeing of blocks
operation within drm buddy.
- Write a macro block_incompatible() to allocate the required blocks.
- Update the xe driver for the drm_buddy_free_list change in arguments.
- add a warning if the two blocks are incompatible on
defragmentation
- call full defragmentation in the fini() function
- place a condition to test if min_order is equal to 0
- replace the list with safe_reverse() variant as we might
remove the block from the list.
v3:
- fix Gitlab user reported lockup issue.
- Keep DRM_BUDDY_HEADER_CLEAR define sorted(Matthew)
- modify to pass the root order instead max_order in fini()
function(Matthew)
- change bool 1 to true(Matthew)
- add check if min_block_size is power of 2(Matthew)
- modify the min_block_size datatype to u64(Matthew)
v4:
- rename the function drm_buddy_defrag with __force_merge.
- Include __force_merge directly in drm buddy file and remove
the defrag use in amdgpu driver.
- Remove list_empty() check(Matthew)
- Remove unnecessary space, headers and placement of new variables(Matthew)
- Add a unit test case(Matthew)
v5:
- remove force merge support to actual range allocation and not to bail
out when contains && split(Matthew)
- add range support to force merge function.
v6:
- modify the alloc_range() function clear page non merged blocks
allocation(Matthew)
- correct the list_insert function name(Matthew).
Signed-off-by: Arunpravin Paneer Selvam <Arunpravin.PaneerSelvam@amd.com>
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Suggested-by: Christian König <christian.koenig@amd.com>
Suggested-by: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240419063538.11957-1-Arunpravin.PaneerSelvam@amd.com
Signed-off-by: Christian König <christian.koenig@amd.com>
DisplayID spec v1.3 revision history notes do claim that
the toplogy block was added in v1.3 so requiring structure
v1.2 would seem correct, but there is at least one EDID in
edid.tv with a topology block and structure v1.0. And
there are also EDIDs with DisplayID structure v1.3 which
seems to be totally incorrect as DisplayID spec v1.3 lists
structure v1.2 as the only legal value.
Unfortunately I couldn't find copies of DisplayID spec
v1.0-v1.2 anywhere (even on vesa.org), so I'll have to
go on empirical evidence alone.
We used to parse the topology block on all v1.x
structures until the check for structure v2.0 was added.
Let's go back to doing that as the evidence does suggest
that there are DisplayIDs in the wild that would miss
out on the topology stuff otherwise.
Also toss out DISPLAY_ID_STRUCTURE_VER_12 entirely as
it doesn't appear we can really use it for anything.
I *think* we could technically skip all the structure
version checks as the block tags shouldn't conflict
between v2.0 and v1.x. But no harm in having a bit of
extra sanity checks I guess.
So far I'm not aware of any user reported regressions
from overly strict check, but I do know that it broke
igt/kms_tiled_display's fake DisplayID as that one
gets generated with structure v1.0.
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Fixes: c5a486af9d ("drm/edid: parse Tiled Display Topology Data Block for DisplayID 2.0")
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240410180139.21352-1-ville.syrjala@linux.intel.com
Acked-by: Jani Nikula <jani.nikula@intel.com>
Linux 6.9-rc5
I've had a persistent msm failure on clang, and the fix is in fixes
so just pull it back to fix that.
Signed-off-by: Dave Airlie <airlied@redhat.com>
