linux/drivers/gpu/drm/i915/display/intel_atomic_plane.c

875 lines
26 KiB
C
Raw Normal View History

drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
/*
* Copyright © 2014 Intel Corporation
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice (including the next
* paragraph) shall be included in all copies or substantial portions of the
* Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
* DEALINGS IN THE SOFTWARE.
*/
/**
* DOC: atomic plane helpers
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
*
* The functions here are used by the atomic plane helper functions to
* implement legacy plane updates (i.e., drm_plane->update_plane() and
* drm_plane->disable_plane()). This allows plane updates to use the
* atomic state infrastructure and perform plane updates as separate
* prepare/check/commit/cleanup steps.
*/
#include <drm/drm_atomic_helper.h>
#include <drm/drm_fourcc.h>
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
#include <drm/drm_plane_helper.h>
#include "gt/intel_rps.h"
#include "intel_atomic_plane.h"
#include "intel_cdclk.h"
#include "intel_display_trace.h"
#include "intel_display_types.h"
#include "intel_fb.h"
#include "intel_fb_pin.h"
#include "intel_pm.h"
#include "intel_sprite.h"
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
static void intel_plane_state_reset(struct intel_plane_state *plane_state,
struct intel_plane *plane)
{
memset(plane_state, 0, sizeof(*plane_state));
__drm_atomic_helper_plane_state_reset(&plane_state->uapi, &plane->base);
plane_state->scaler_id = -1;
}
struct intel_plane *intel_plane_alloc(void)
{
struct intel_plane_state *plane_state;
struct intel_plane *plane;
plane = kzalloc(sizeof(*plane), GFP_KERNEL);
if (!plane)
return ERR_PTR(-ENOMEM);
plane_state = kzalloc(sizeof(*plane_state), GFP_KERNEL);
if (!plane_state) {
kfree(plane);
return ERR_PTR(-ENOMEM);
}
intel_plane_state_reset(plane_state, plane);
plane->base.state = &plane_state->uapi;
return plane;
}
void intel_plane_free(struct intel_plane *plane)
{
intel_plane_destroy_state(&plane->base, plane->base.state);
kfree(plane);
}
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
/**
* intel_plane_duplicate_state - duplicate plane state
* @plane: drm plane
*
* Allocates and returns a copy of the plane state (both common and
* Intel-specific) for the specified plane.
*
* Returns: The newly allocated plane state, or NULL on failure.
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
*/
struct drm_plane_state *
intel_plane_duplicate_state(struct drm_plane *plane)
{
struct intel_plane_state *intel_state;
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
intel_state = to_intel_plane_state(plane->state);
intel_state = kmemdup(intel_state, sizeof(*intel_state), GFP_KERNEL);
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
if (!intel_state)
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
return NULL;
__drm_atomic_helper_plane_duplicate_state(plane, &intel_state->uapi);
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
intel_state->ggtt_vma = NULL;
intel_state->dpt_vma = NULL;
drm/i915: Move the policy for placement of the GGTT vma into the caller Currently we make the unilateral decision inside i915_gem_object_pin_to_display() where the VMA should resided (inside the fence and mappable region or above?). This is not our decision to make as it impacts on how the display engine can use the resulting scanout object, and it would rather instruct us where to place the VMA so that it can enable the features it wants. As such, make the pin flags an argument to i915_gem_object_pin_to_display() and control them from intel_pin_and_fence_fb_obj() Whilst taking control of the mapping for ourselves, start tracking how we use it to avoid trying to free a fence we never claimed: <3>[ 227.151869] GEM_BUG_ON(vma->fence->pin_count <= 0) <4>[ 227.152064] ------------[ cut here ]------------ <2>[ 227.152068] kernel BUG at drivers/gpu/drm/i915/i915_vma.h:391! <4>[ 227.152084] invalid opcode: 0000 [#1] PREEMPT SMP KASAN PTI <0>[ 227.152092] Dumping ftrace buffer: <0>[ 227.152099] (ftrace buffer empty) <4>[ 227.152102] Modules linked in: i915 snd_hda_codec_analog snd_hda_codec_generic coretemp snd_hda_intel snd_hda_codec snd_hwdep snd_hda_core snd_pcm lpc_ich e1000e mei_me mei prime_numbers <4>[ 227.152131] CPU: 1 PID: 1587 Comm: kworker/u16:49 Tainted: G U 4.16.0-rc1-gbab67b2f6177-kasan_7+ #1 <4>[ 227.152134] Hardware name: Dell Inc. OptiPlex 755 /0PU052, BIOS A08 02/19/2008 <4>[ 227.152236] Workqueue: events_unbound intel_atomic_commit_work [i915] <4>[ 227.152292] RIP: 0010:intel_unpin_fb_vma+0x23a/0x2a0 [i915] <4>[ 227.152295] RSP: 0018:ffff88005aad7b68 EFLAGS: 00010286 <4>[ 227.152300] RAX: 0000000000000026 RBX: ffff88005c359580 RCX: 0000000000000000 <4>[ 227.152304] RDX: 0000000000000026 RSI: ffffffff8707d840 RDI: ffffed000b55af63 <4>[ 227.152307] RBP: ffff880056817e58 R08: 0000000000000001 R09: 0000000000000000 <4>[ 227.152311] R10: ffff88005aad7b88 R11: 0000000000000000 R12: ffff8800568184d0 <4>[ 227.152314] R13: ffff880065b5ab08 R14: 0000000000000000 R15: dffffc0000000000 <4>[ 227.152318] FS: 0000000000000000(0000) GS:ffff88006ac40000(0000) knlGS:0000000000000000 <4>[ 227.152322] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 <4>[ 227.152325] CR2: 00007f5fb25550a8 CR3: 0000000068c78000 CR4: 00000000000006e0 <4>[ 227.152328] Call Trace: <4>[ 227.152385] intel_cleanup_plane_fb+0x6b/0xd0 [i915] <4>[ 227.152395] drm_atomic_helper_cleanup_planes+0x166/0x280 <4>[ 227.152452] intel_atomic_commit_tail+0x159d/0x3380 [i915] <4>[ 227.152463] ? process_one_work+0x66e/0x1460 <4>[ 227.152516] ? skl_update_crtcs+0x9c0/0x9c0 [i915] <4>[ 227.152523] ? lock_acquire+0x13d/0x390 <4>[ 227.152527] ? lock_acquire+0x13d/0x390 <4>[ 227.152534] process_one_work+0x71a/0x1460 <4>[ 227.152540] ? __schedule+0x815/0x1e20 <4>[ 227.152547] ? pwq_dec_nr_in_flight+0x2b0/0x2b0 <4>[ 227.152553] ? _raw_spin_lock_irq+0xa/0x40 <4>[ 227.152559] worker_thread+0xdf/0xf60 <4>[ 227.152569] ? process_one_work+0x1460/0x1460 <4>[ 227.152573] kthread+0x2cf/0x3c0 <4>[ 227.152578] ? _kthread_create_on_node+0xa0/0xa0 <4>[ 227.152583] ret_from_fork+0x3a/0x50 <4>[ 227.152591] Code: c6 00 11 86 c0 48 c7 c7 e0 bd 85 c0 e8 60 e7 a9 c4 0f ff e9 1f fe ff ff 48 c7 c6 40 10 86 c0 48 c7 c7 e0 ca 85 c0 e8 2b 95 bd c4 <0f> 0b 48 89 ef e8 4c 44 e8 c4 e9 ef fd ff ff e8 42 44 e8 c4 e9 <1>[ 227.152720] RIP: intel_unpin_fb_vma+0x23a/0x2a0 [i915] RSP: ffff88005aad7b68 v2: i915_vma_pin_fence() is a no-op if a fence isn't required, so check vma->fence as well. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180220134208.24988-2-chris@chris-wilson.co.uk
2018-02-20 13:42:06 +00:00
intel_state->flags = 0;
intel_state->do_async_flip = false;
/* add reference to fb */
if (intel_state->hw.fb)
drm_framebuffer_get(intel_state->hw.fb);
return &intel_state->uapi;
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
}
/**
* intel_plane_destroy_state - destroy plane state
* @plane: drm plane
* @state: state object to destroy
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
*
* Destroys the plane state (both common and Intel-specific) for the
* specified plane.
*/
void
intel_plane_destroy_state(struct drm_plane *plane,
struct drm_plane_state *state)
{
struct intel_plane_state *plane_state = to_intel_plane_state(state);
drm_WARN_ON(plane->dev, plane_state->ggtt_vma);
drm_WARN_ON(plane->dev, plane_state->dpt_vma);
__drm_atomic_helper_plane_destroy_state(&plane_state->uapi);
if (plane_state->hw.fb)
drm_framebuffer_put(plane_state->hw.fb);
kfree(plane_state);
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
}
unsigned int intel_adjusted_rate(const struct drm_rect *src,
const struct drm_rect *dst,
unsigned int rate)
drm/i915: Use intel_plane_data_rate for min_cdclk calculation There seems to be a bit of confusing redundancy in a way, how plane data rate/min cdclk are calculated. In fact both min cdclk, pixel rate and plane data rate are all part of the same formula as per BSpec. However currently we have intel_plane_data_rate, which is used to calculate plane data rate and which is also used in bandwidth calculations. However for calculating min_cdclk we have another piece of code, doing almost same calculation, but a bit differently and in a different place. However as both are actually part of same formula, probably would be wise to use plane data rate calculations as a basis anyway, thus avoiding code duplication and possible bugs related to this. Another thing is that I've noticed that during min_cdclk calculations we account for plane scaling, while for plane data rate, we don't. crtc->pixel_rate seems to account only for pipe ratio, however it is clearly stated in BSpec that plane data rate also need to account plane ratio as well. So what this commit does is: - Adds a plane ratio calculation to intel_plane_data_rate - Removes redundant calculations from skl_plane_min_cdclk which is used for gen9+ and now uses intel_plane_data_rate as a basis from there as well. v2: - Don't use 64 division if not needed(Ville Syrjälä) - Now use intel_plane_pixel_rate as a basis for calculations both at intel_plane_data_rate and skl_plane_min_cdclk(Ville Syrjälä) v3: - Again fix the division macro - Fix plane_pixel_rate to pixel_rate at intel_plane_pixel_rate callsites v4: - Renamed skl_plane_ratio function back(Ville Syrjälä) v5: - Don't precalculate plane pixel rate for invisible plane, check for visibility first, as in invisible case it will have dst_w and dst_h equal to zero, causing divide error. v6: - Removed useless warn in intel_plane_pixel_rate(Ville Syrjälä) - Fixed alignment in intel_plane_data_rate(Ville Syrjälä) - Changed pixel_rate type to be unsigned int in skl_plane_min_cdclk(Ville Syrjälä) Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200227150935.2107-1-stanislav.lisovskiy@intel.com
2020-02-27 17:09:35 +02:00
{
unsigned int src_w, src_h, dst_w, dst_h;
src_w = drm_rect_width(src) >> 16;
src_h = drm_rect_height(src) >> 16;
dst_w = drm_rect_width(dst);
dst_h = drm_rect_height(dst);
drm/i915: Use intel_plane_data_rate for min_cdclk calculation There seems to be a bit of confusing redundancy in a way, how plane data rate/min cdclk are calculated. In fact both min cdclk, pixel rate and plane data rate are all part of the same formula as per BSpec. However currently we have intel_plane_data_rate, which is used to calculate plane data rate and which is also used in bandwidth calculations. However for calculating min_cdclk we have another piece of code, doing almost same calculation, but a bit differently and in a different place. However as both are actually part of same formula, probably would be wise to use plane data rate calculations as a basis anyway, thus avoiding code duplication and possible bugs related to this. Another thing is that I've noticed that during min_cdclk calculations we account for plane scaling, while for plane data rate, we don't. crtc->pixel_rate seems to account only for pipe ratio, however it is clearly stated in BSpec that plane data rate also need to account plane ratio as well. So what this commit does is: - Adds a plane ratio calculation to intel_plane_data_rate - Removes redundant calculations from skl_plane_min_cdclk which is used for gen9+ and now uses intel_plane_data_rate as a basis from there as well. v2: - Don't use 64 division if not needed(Ville Syrjälä) - Now use intel_plane_pixel_rate as a basis for calculations both at intel_plane_data_rate and skl_plane_min_cdclk(Ville Syrjälä) v3: - Again fix the division macro - Fix plane_pixel_rate to pixel_rate at intel_plane_pixel_rate callsites v4: - Renamed skl_plane_ratio function back(Ville Syrjälä) v5: - Don't precalculate plane pixel rate for invisible plane, check for visibility first, as in invisible case it will have dst_w and dst_h equal to zero, causing divide error. v6: - Removed useless warn in intel_plane_pixel_rate(Ville Syrjälä) - Fixed alignment in intel_plane_data_rate(Ville Syrjälä) - Changed pixel_rate type to be unsigned int in skl_plane_min_cdclk(Ville Syrjälä) Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200227150935.2107-1-stanislav.lisovskiy@intel.com
2020-02-27 17:09:35 +02:00
/* Downscaling limits the maximum pixel rate */
dst_w = min(src_w, dst_w);
dst_h = min(src_h, dst_h);
return DIV_ROUND_UP_ULL(mul_u32_u32(rate, src_w * src_h),
drm/i915: Use intel_plane_data_rate for min_cdclk calculation There seems to be a bit of confusing redundancy in a way, how plane data rate/min cdclk are calculated. In fact both min cdclk, pixel rate and plane data rate are all part of the same formula as per BSpec. However currently we have intel_plane_data_rate, which is used to calculate plane data rate and which is also used in bandwidth calculations. However for calculating min_cdclk we have another piece of code, doing almost same calculation, but a bit differently and in a different place. However as both are actually part of same formula, probably would be wise to use plane data rate calculations as a basis anyway, thus avoiding code duplication and possible bugs related to this. Another thing is that I've noticed that during min_cdclk calculations we account for plane scaling, while for plane data rate, we don't. crtc->pixel_rate seems to account only for pipe ratio, however it is clearly stated in BSpec that plane data rate also need to account plane ratio as well. So what this commit does is: - Adds a plane ratio calculation to intel_plane_data_rate - Removes redundant calculations from skl_plane_min_cdclk which is used for gen9+ and now uses intel_plane_data_rate as a basis from there as well. v2: - Don't use 64 division if not needed(Ville Syrjälä) - Now use intel_plane_pixel_rate as a basis for calculations both at intel_plane_data_rate and skl_plane_min_cdclk(Ville Syrjälä) v3: - Again fix the division macro - Fix plane_pixel_rate to pixel_rate at intel_plane_pixel_rate callsites v4: - Renamed skl_plane_ratio function back(Ville Syrjälä) v5: - Don't precalculate plane pixel rate for invisible plane, check for visibility first, as in invisible case it will have dst_w and dst_h equal to zero, causing divide error. v6: - Removed useless warn in intel_plane_pixel_rate(Ville Syrjälä) - Fixed alignment in intel_plane_data_rate(Ville Syrjälä) - Changed pixel_rate type to be unsigned int in skl_plane_min_cdclk(Ville Syrjälä) Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200227150935.2107-1-stanislav.lisovskiy@intel.com
2020-02-27 17:09:35 +02:00
dst_w * dst_h);
}
unsigned int intel_plane_pixel_rate(const struct intel_crtc_state *crtc_state,
const struct intel_plane_state *plane_state)
{
/*
* Note we don't check for plane visibility here as
* we want to use this when calculating the cursor
* watermarks even if the cursor is fully offscreen.
* That depends on the src/dst rectangles being
* correctly populated whenever the watermark code
* considers the cursor to be visible, whether or not
* it is actually visible.
*
* See: intel_wm_plane_visible() and intel_check_cursor()
*/
return intel_adjusted_rate(&plane_state->uapi.src,
&plane_state->uapi.dst,
crtc_state->pixel_rate);
}
drm/i915: Make sure we have enough memory bandwidth on ICL ICL has so many planes that it can easily exceed the maximum effective memory bandwidth of the system. We must therefore check that we don't exceed that limit. The algorithm is very magic number heavy and lacks sufficient explanation for now. We also have no sane way to query the memory clock and timings, so we must rely on a combination of raw readout from the memory controller and hardcoded assumptions. The memory controller values obviously change as the system jumps between the different SAGV points, so we try to stabilize it first by disabling SAGV for the duration of the readout. The utilized bandwidth is tracked via a device wide atomic private object. That is actually not robust because we can't afford to enforce strict global ordering between the pipes. Thus I think I'll need to change this to simply chop up the available bandwidth between all the active pipes. Each pipe can then do whatever it wants as long as it doesn't exceed its budget. That scheme will also require that we assume that any number of planes could be active at any time. TODO: make it robust and deal with all the open questions v2: Sleep longer after disabling SAGV v3: Poll for the dclk to get raised (seen it take 250ms!) If the system has 2133MT/s memory then we pointlessly wait one full second :( v4: Use the new pcode interface to get the qgv points rather that using hardcoded numbers v5: Move the pcode stuff into intel_bw.c (Matt) s/intel_sagv_info/intel_qgv_info/ Do the NV12/P010 as per spec for now (Matt) s/IS_ICELAKE/IS_GEN11/ v6: Ignore bandwidth limits if the pcode query fails Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Acked-by: Clint Taylor <Clinton.A.Taylor@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190524153614.32410-1-ville.syrjala@linux.intel.com
2019-05-24 18:36:14 +03:00
unsigned int intel_plane_data_rate(const struct intel_crtc_state *crtc_state,
const struct intel_plane_state *plane_state)
{
const struct drm_framebuffer *fb = plane_state->hw.fb;
drm/i915: Make sure we have enough memory bandwidth on ICL ICL has so many planes that it can easily exceed the maximum effective memory bandwidth of the system. We must therefore check that we don't exceed that limit. The algorithm is very magic number heavy and lacks sufficient explanation for now. We also have no sane way to query the memory clock and timings, so we must rely on a combination of raw readout from the memory controller and hardcoded assumptions. The memory controller values obviously change as the system jumps between the different SAGV points, so we try to stabilize it first by disabling SAGV for the duration of the readout. The utilized bandwidth is tracked via a device wide atomic private object. That is actually not robust because we can't afford to enforce strict global ordering between the pipes. Thus I think I'll need to change this to simply chop up the available bandwidth between all the active pipes. Each pipe can then do whatever it wants as long as it doesn't exceed its budget. That scheme will also require that we assume that any number of planes could be active at any time. TODO: make it robust and deal with all the open questions v2: Sleep longer after disabling SAGV v3: Poll for the dclk to get raised (seen it take 250ms!) If the system has 2133MT/s memory then we pointlessly wait one full second :( v4: Use the new pcode interface to get the qgv points rather that using hardcoded numbers v5: Move the pcode stuff into intel_bw.c (Matt) s/intel_sagv_info/intel_qgv_info/ Do the NV12/P010 as per spec for now (Matt) s/IS_ICELAKE/IS_GEN11/ v6: Ignore bandwidth limits if the pcode query fails Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Acked-by: Clint Taylor <Clinton.A.Taylor@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190524153614.32410-1-ville.syrjala@linux.intel.com
2019-05-24 18:36:14 +03:00
unsigned int cpp;
drm/i915: Use intel_plane_data_rate for min_cdclk calculation There seems to be a bit of confusing redundancy in a way, how plane data rate/min cdclk are calculated. In fact both min cdclk, pixel rate and plane data rate are all part of the same formula as per BSpec. However currently we have intel_plane_data_rate, which is used to calculate plane data rate and which is also used in bandwidth calculations. However for calculating min_cdclk we have another piece of code, doing almost same calculation, but a bit differently and in a different place. However as both are actually part of same formula, probably would be wise to use plane data rate calculations as a basis anyway, thus avoiding code duplication and possible bugs related to this. Another thing is that I've noticed that during min_cdclk calculations we account for plane scaling, while for plane data rate, we don't. crtc->pixel_rate seems to account only for pipe ratio, however it is clearly stated in BSpec that plane data rate also need to account plane ratio as well. So what this commit does is: - Adds a plane ratio calculation to intel_plane_data_rate - Removes redundant calculations from skl_plane_min_cdclk which is used for gen9+ and now uses intel_plane_data_rate as a basis from there as well. v2: - Don't use 64 division if not needed(Ville Syrjälä) - Now use intel_plane_pixel_rate as a basis for calculations both at intel_plane_data_rate and skl_plane_min_cdclk(Ville Syrjälä) v3: - Again fix the division macro - Fix plane_pixel_rate to pixel_rate at intel_plane_pixel_rate callsites v4: - Renamed skl_plane_ratio function back(Ville Syrjälä) v5: - Don't precalculate plane pixel rate for invisible plane, check for visibility first, as in invisible case it will have dst_w and dst_h equal to zero, causing divide error. v6: - Removed useless warn in intel_plane_pixel_rate(Ville Syrjälä) - Fixed alignment in intel_plane_data_rate(Ville Syrjälä) - Changed pixel_rate type to be unsigned int in skl_plane_min_cdclk(Ville Syrjälä) Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200227150935.2107-1-stanislav.lisovskiy@intel.com
2020-02-27 17:09:35 +02:00
unsigned int pixel_rate;
drm/i915: Make sure we have enough memory bandwidth on ICL ICL has so many planes that it can easily exceed the maximum effective memory bandwidth of the system. We must therefore check that we don't exceed that limit. The algorithm is very magic number heavy and lacks sufficient explanation for now. We also have no sane way to query the memory clock and timings, so we must rely on a combination of raw readout from the memory controller and hardcoded assumptions. The memory controller values obviously change as the system jumps between the different SAGV points, so we try to stabilize it first by disabling SAGV for the duration of the readout. The utilized bandwidth is tracked via a device wide atomic private object. That is actually not robust because we can't afford to enforce strict global ordering between the pipes. Thus I think I'll need to change this to simply chop up the available bandwidth between all the active pipes. Each pipe can then do whatever it wants as long as it doesn't exceed its budget. That scheme will also require that we assume that any number of planes could be active at any time. TODO: make it robust and deal with all the open questions v2: Sleep longer after disabling SAGV v3: Poll for the dclk to get raised (seen it take 250ms!) If the system has 2133MT/s memory then we pointlessly wait one full second :( v4: Use the new pcode interface to get the qgv points rather that using hardcoded numbers v5: Move the pcode stuff into intel_bw.c (Matt) s/intel_sagv_info/intel_qgv_info/ Do the NV12/P010 as per spec for now (Matt) s/IS_ICELAKE/IS_GEN11/ v6: Ignore bandwidth limits if the pcode query fails Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Acked-by: Clint Taylor <Clinton.A.Taylor@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190524153614.32410-1-ville.syrjala@linux.intel.com
2019-05-24 18:36:14 +03:00
if (!plane_state->uapi.visible)
drm/i915: Make sure we have enough memory bandwidth on ICL ICL has so many planes that it can easily exceed the maximum effective memory bandwidth of the system. We must therefore check that we don't exceed that limit. The algorithm is very magic number heavy and lacks sufficient explanation for now. We also have no sane way to query the memory clock and timings, so we must rely on a combination of raw readout from the memory controller and hardcoded assumptions. The memory controller values obviously change as the system jumps between the different SAGV points, so we try to stabilize it first by disabling SAGV for the duration of the readout. The utilized bandwidth is tracked via a device wide atomic private object. That is actually not robust because we can't afford to enforce strict global ordering between the pipes. Thus I think I'll need to change this to simply chop up the available bandwidth between all the active pipes. Each pipe can then do whatever it wants as long as it doesn't exceed its budget. That scheme will also require that we assume that any number of planes could be active at any time. TODO: make it robust and deal with all the open questions v2: Sleep longer after disabling SAGV v3: Poll for the dclk to get raised (seen it take 250ms!) If the system has 2133MT/s memory then we pointlessly wait one full second :( v4: Use the new pcode interface to get the qgv points rather that using hardcoded numbers v5: Move the pcode stuff into intel_bw.c (Matt) s/intel_sagv_info/intel_qgv_info/ Do the NV12/P010 as per spec for now (Matt) s/IS_ICELAKE/IS_GEN11/ v6: Ignore bandwidth limits if the pcode query fails Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Acked-by: Clint Taylor <Clinton.A.Taylor@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190524153614.32410-1-ville.syrjala@linux.intel.com
2019-05-24 18:36:14 +03:00
return 0;
drm/i915: Use intel_plane_data_rate for min_cdclk calculation There seems to be a bit of confusing redundancy in a way, how plane data rate/min cdclk are calculated. In fact both min cdclk, pixel rate and plane data rate are all part of the same formula as per BSpec. However currently we have intel_plane_data_rate, which is used to calculate plane data rate and which is also used in bandwidth calculations. However for calculating min_cdclk we have another piece of code, doing almost same calculation, but a bit differently and in a different place. However as both are actually part of same formula, probably would be wise to use plane data rate calculations as a basis anyway, thus avoiding code duplication and possible bugs related to this. Another thing is that I've noticed that during min_cdclk calculations we account for plane scaling, while for plane data rate, we don't. crtc->pixel_rate seems to account only for pipe ratio, however it is clearly stated in BSpec that plane data rate also need to account plane ratio as well. So what this commit does is: - Adds a plane ratio calculation to intel_plane_data_rate - Removes redundant calculations from skl_plane_min_cdclk which is used for gen9+ and now uses intel_plane_data_rate as a basis from there as well. v2: - Don't use 64 division if not needed(Ville Syrjälä) - Now use intel_plane_pixel_rate as a basis for calculations both at intel_plane_data_rate and skl_plane_min_cdclk(Ville Syrjälä) v3: - Again fix the division macro - Fix plane_pixel_rate to pixel_rate at intel_plane_pixel_rate callsites v4: - Renamed skl_plane_ratio function back(Ville Syrjälä) v5: - Don't precalculate plane pixel rate for invisible plane, check for visibility first, as in invisible case it will have dst_w and dst_h equal to zero, causing divide error. v6: - Removed useless warn in intel_plane_pixel_rate(Ville Syrjälä) - Fixed alignment in intel_plane_data_rate(Ville Syrjälä) - Changed pixel_rate type to be unsigned int in skl_plane_min_cdclk(Ville Syrjälä) Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200227150935.2107-1-stanislav.lisovskiy@intel.com
2020-02-27 17:09:35 +02:00
pixel_rate = intel_plane_pixel_rate(crtc_state, plane_state);
drm/i915: Make sure we have enough memory bandwidth on ICL ICL has so many planes that it can easily exceed the maximum effective memory bandwidth of the system. We must therefore check that we don't exceed that limit. The algorithm is very magic number heavy and lacks sufficient explanation for now. We also have no sane way to query the memory clock and timings, so we must rely on a combination of raw readout from the memory controller and hardcoded assumptions. The memory controller values obviously change as the system jumps between the different SAGV points, so we try to stabilize it first by disabling SAGV for the duration of the readout. The utilized bandwidth is tracked via a device wide atomic private object. That is actually not robust because we can't afford to enforce strict global ordering between the pipes. Thus I think I'll need to change this to simply chop up the available bandwidth between all the active pipes. Each pipe can then do whatever it wants as long as it doesn't exceed its budget. That scheme will also require that we assume that any number of planes could be active at any time. TODO: make it robust and deal with all the open questions v2: Sleep longer after disabling SAGV v3: Poll for the dclk to get raised (seen it take 250ms!) If the system has 2133MT/s memory then we pointlessly wait one full second :( v4: Use the new pcode interface to get the qgv points rather that using hardcoded numbers v5: Move the pcode stuff into intel_bw.c (Matt) s/intel_sagv_info/intel_qgv_info/ Do the NV12/P010 as per spec for now (Matt) s/IS_ICELAKE/IS_GEN11/ v6: Ignore bandwidth limits if the pcode query fails Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Acked-by: Clint Taylor <Clinton.A.Taylor@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190524153614.32410-1-ville.syrjala@linux.intel.com
2019-05-24 18:36:14 +03:00
cpp = fb->format->cpp[0];
/*
* Based on HSD#:1408715493
* NV12 cpp == 4, P010 cpp == 8
*
* FIXME what is the logic behind this?
*/
if (fb->format->is_yuv && fb->format->num_planes > 1)
cpp *= 4;
drm/i915: Use intel_plane_data_rate for min_cdclk calculation There seems to be a bit of confusing redundancy in a way, how plane data rate/min cdclk are calculated. In fact both min cdclk, pixel rate and plane data rate are all part of the same formula as per BSpec. However currently we have intel_plane_data_rate, which is used to calculate plane data rate and which is also used in bandwidth calculations. However for calculating min_cdclk we have another piece of code, doing almost same calculation, but a bit differently and in a different place. However as both are actually part of same formula, probably would be wise to use plane data rate calculations as a basis anyway, thus avoiding code duplication and possible bugs related to this. Another thing is that I've noticed that during min_cdclk calculations we account for plane scaling, while for plane data rate, we don't. crtc->pixel_rate seems to account only for pipe ratio, however it is clearly stated in BSpec that plane data rate also need to account plane ratio as well. So what this commit does is: - Adds a plane ratio calculation to intel_plane_data_rate - Removes redundant calculations from skl_plane_min_cdclk which is used for gen9+ and now uses intel_plane_data_rate as a basis from there as well. v2: - Don't use 64 division if not needed(Ville Syrjälä) - Now use intel_plane_pixel_rate as a basis for calculations both at intel_plane_data_rate and skl_plane_min_cdclk(Ville Syrjälä) v3: - Again fix the division macro - Fix plane_pixel_rate to pixel_rate at intel_plane_pixel_rate callsites v4: - Renamed skl_plane_ratio function back(Ville Syrjälä) v5: - Don't precalculate plane pixel rate for invisible plane, check for visibility first, as in invisible case it will have dst_w and dst_h equal to zero, causing divide error. v6: - Removed useless warn in intel_plane_pixel_rate(Ville Syrjälä) - Fixed alignment in intel_plane_data_rate(Ville Syrjälä) - Changed pixel_rate type to be unsigned int in skl_plane_min_cdclk(Ville Syrjälä) Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200227150935.2107-1-stanislav.lisovskiy@intel.com
2020-02-27 17:09:35 +02:00
return pixel_rate * cpp;
drm/i915: Make sure we have enough memory bandwidth on ICL ICL has so many planes that it can easily exceed the maximum effective memory bandwidth of the system. We must therefore check that we don't exceed that limit. The algorithm is very magic number heavy and lacks sufficient explanation for now. We also have no sane way to query the memory clock and timings, so we must rely on a combination of raw readout from the memory controller and hardcoded assumptions. The memory controller values obviously change as the system jumps between the different SAGV points, so we try to stabilize it first by disabling SAGV for the duration of the readout. The utilized bandwidth is tracked via a device wide atomic private object. That is actually not robust because we can't afford to enforce strict global ordering between the pipes. Thus I think I'll need to change this to simply chop up the available bandwidth between all the active pipes. Each pipe can then do whatever it wants as long as it doesn't exceed its budget. That scheme will also require that we assume that any number of planes could be active at any time. TODO: make it robust and deal with all the open questions v2: Sleep longer after disabling SAGV v3: Poll for the dclk to get raised (seen it take 250ms!) If the system has 2133MT/s memory then we pointlessly wait one full second :( v4: Use the new pcode interface to get the qgv points rather that using hardcoded numbers v5: Move the pcode stuff into intel_bw.c (Matt) s/intel_sagv_info/intel_qgv_info/ Do the NV12/P010 as per spec for now (Matt) s/IS_ICELAKE/IS_GEN11/ v6: Ignore bandwidth limits if the pcode query fails Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Acked-by: Clint Taylor <Clinton.A.Taylor@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190524153614.32410-1-ville.syrjala@linux.intel.com
2019-05-24 18:36:14 +03:00
}
int intel_plane_calc_min_cdclk(struct intel_atomic_state *state,
struct intel_plane *plane,
bool *need_cdclk_calc)
2019-10-15 22:30:26 +03:00
{
struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
const struct intel_plane_state *plane_state =
intel_atomic_get_new_plane_state(state, plane);
struct intel_crtc *crtc = to_intel_crtc(plane_state->hw.crtc);
const struct intel_cdclk_state *cdclk_state;
drm/i915/display: Fix NULL-crtc deref in calc_min_cdclk() [ 23.419442] BUG: KASAN: null-ptr-deref in intel_plane_calc_min_cdclk+0x82/0x440 [i915] [ 23.419527] Read of size 4 at addr 00000000000000f8 by task insmod/735 [ 23.419578] [ 23.419644] CPU: 2 PID: 735 Comm: insmod Not tainted 5.5.0+ #114 [ 23.419716] Hardware name: ��������������������������������� ���������������������������������/���������������������������������, BIOS RYBDWi35.86A.0246.2 [ 23.419793] Call Trace: [ 23.419864] dump_stack+0xef/0x16e [ 23.419927] __kasan_report.cold+0x60/0x90 [ 23.420157] ? intel_plane_calc_min_cdclk+0x82/0x440 [i915] [ 23.420397] intel_plane_calc_min_cdclk+0x82/0x440 [i915] [ 23.420630] intel_atomic_check+0x455f/0x65a0 [i915] [ 23.420708] ? mark_held_locks+0x90/0x90 [ 23.420929] ? intel_crtc_duplicate_state+0x2e/0x1b0 [i915] [ 23.421172] ? intel_plane_duplicate_state+0x2d/0xc0 [i915] [ 23.421239] ? __drm_dbg+0xa4/0x120 [ 23.421303] ? __kasan_kmalloc.constprop.0+0xc2/0xd0 [ 23.421355] ? __kmalloc_track_caller+0x23a/0x320 [ 23.421602] ? intel_calc_active_pipes+0x1c0/0x1c0 [i915] [ 23.421852] sanitize_watermarks+0x220/0x510 [i915] [ 23.422092] ? intel_atomic_check+0x65a0/0x65a0 [i915] [ 23.422164] ? drm_modeset_unlock_all+0x88/0x130 [ 23.422402] intel_modeset_init+0x1b76/0x3c90 [i915] [ 23.422647] ? intel_finish_reset+0x2d0/0x2d0 [i915] [ 23.422851] ? intel_irq_install+0x12c/0x210 [i915] [ 23.423076] i915_driver_probe+0x13e7/0x2930 [i915] v2: No crtc is implied by an invisible plane, so the extra !crtc check is redundant. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200203133824.198872-1-chris@chris-wilson.co.uk
2020-02-03 13:38:24 +00:00
const struct intel_crtc_state *old_crtc_state;
struct intel_crtc_state *new_crtc_state;
2019-10-15 22:30:26 +03:00
if (!plane_state->uapi.visible || !plane->min_cdclk)
return 0;
2019-10-15 22:30:26 +03:00
drm/i915/display: Fix NULL-crtc deref in calc_min_cdclk() [ 23.419442] BUG: KASAN: null-ptr-deref in intel_plane_calc_min_cdclk+0x82/0x440 [i915] [ 23.419527] Read of size 4 at addr 00000000000000f8 by task insmod/735 [ 23.419578] [ 23.419644] CPU: 2 PID: 735 Comm: insmod Not tainted 5.5.0+ #114 [ 23.419716] Hardware name: ��������������������������������� ���������������������������������/���������������������������������, BIOS RYBDWi35.86A.0246.2 [ 23.419793] Call Trace: [ 23.419864] dump_stack+0xef/0x16e [ 23.419927] __kasan_report.cold+0x60/0x90 [ 23.420157] ? intel_plane_calc_min_cdclk+0x82/0x440 [i915] [ 23.420397] intel_plane_calc_min_cdclk+0x82/0x440 [i915] [ 23.420630] intel_atomic_check+0x455f/0x65a0 [i915] [ 23.420708] ? mark_held_locks+0x90/0x90 [ 23.420929] ? intel_crtc_duplicate_state+0x2e/0x1b0 [i915] [ 23.421172] ? intel_plane_duplicate_state+0x2d/0xc0 [i915] [ 23.421239] ? __drm_dbg+0xa4/0x120 [ 23.421303] ? __kasan_kmalloc.constprop.0+0xc2/0xd0 [ 23.421355] ? __kmalloc_track_caller+0x23a/0x320 [ 23.421602] ? intel_calc_active_pipes+0x1c0/0x1c0 [i915] [ 23.421852] sanitize_watermarks+0x220/0x510 [i915] [ 23.422092] ? intel_atomic_check+0x65a0/0x65a0 [i915] [ 23.422164] ? drm_modeset_unlock_all+0x88/0x130 [ 23.422402] intel_modeset_init+0x1b76/0x3c90 [i915] [ 23.422647] ? intel_finish_reset+0x2d0/0x2d0 [i915] [ 23.422851] ? intel_irq_install+0x12c/0x210 [i915] [ 23.423076] i915_driver_probe+0x13e7/0x2930 [i915] v2: No crtc is implied by an invisible plane, so the extra !crtc check is redundant. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200203133824.198872-1-chris@chris-wilson.co.uk
2020-02-03 13:38:24 +00:00
old_crtc_state = intel_atomic_get_old_crtc_state(state, crtc);
new_crtc_state = intel_atomic_get_new_crtc_state(state, crtc);
new_crtc_state->min_cdclk[plane->id] =
plane->min_cdclk(new_crtc_state, plane_state);
2019-10-15 22:30:26 +03:00
/*
* No need to check against the cdclk state if
* the min cdclk for the plane doesn't increase.
*
* Ie. we only ever increase the cdclk due to plane
* requirements. This can reduce back and forth
* display blinking due to constant cdclk changes.
*/
if (new_crtc_state->min_cdclk[plane->id] <=
old_crtc_state->min_cdclk[plane->id])
return 0;
cdclk_state = intel_atomic_get_cdclk_state(state);
if (IS_ERR(cdclk_state))
return PTR_ERR(cdclk_state);
2019-10-15 22:30:26 +03:00
/*
* No need to recalculate the cdclk state if
* the min cdclk for the pipe doesn't increase.
2019-10-15 22:30:26 +03:00
*
* Ie. we only ever increase the cdclk due to plane
* requirements. This can reduce back and forth
* display blinking due to constant cdclk changes.
2019-10-15 22:30:26 +03:00
*/
if (new_crtc_state->min_cdclk[plane->id] <=
cdclk_state->min_cdclk[crtc->pipe])
return 0;
drm_dbg_kms(&dev_priv->drm,
"[PLANE:%d:%s] min cdclk (%d kHz) > [CRTC:%d:%s] min cdclk (%d kHz)\n",
plane->base.base.id, plane->base.name,
new_crtc_state->min_cdclk[plane->id],
crtc->base.base.id, crtc->base.name,
cdclk_state->min_cdclk[crtc->pipe]);
*need_cdclk_calc = true;
2019-10-15 22:30:26 +03:00
return 0;
2019-10-15 22:30:26 +03:00
}
static void intel_plane_clear_hw_state(struct intel_plane_state *plane_state)
{
if (plane_state->hw.fb)
drm_framebuffer_put(plane_state->hw.fb);
memset(&plane_state->hw, 0, sizeof(plane_state->hw));
}
void intel_plane_copy_uapi_to_hw_state(struct intel_plane_state *plane_state,
const struct intel_plane_state *from_plane_state,
struct intel_crtc *crtc)
{
intel_plane_clear_hw_state(plane_state);
/*
* For the bigjoiner slave uapi.crtc will point at
* the master crtc. So we explicitly assign the right
* slave crtc to hw.crtc. uapi.crtc!=NULL simply indicates
* the plane is logically enabled on the uapi level.
*/
plane_state->hw.crtc = from_plane_state->uapi.crtc ? &crtc->base : NULL;
plane_state->hw.fb = from_plane_state->uapi.fb;
if (plane_state->hw.fb)
drm_framebuffer_get(plane_state->hw.fb);
plane_state->hw.alpha = from_plane_state->uapi.alpha;
plane_state->hw.pixel_blend_mode =
from_plane_state->uapi.pixel_blend_mode;
plane_state->hw.rotation = from_plane_state->uapi.rotation;
plane_state->hw.color_encoding = from_plane_state->uapi.color_encoding;
plane_state->hw.color_range = from_plane_state->uapi.color_range;
plane_state->hw.scaling_filter = from_plane_state->uapi.scaling_filter;
plane_state->uapi.src = drm_plane_state_src(&from_plane_state->uapi);
plane_state->uapi.dst = drm_plane_state_dest(&from_plane_state->uapi);
}
void intel_plane_copy_hw_state(struct intel_plane_state *plane_state,
const struct intel_plane_state *from_plane_state)
{
intel_plane_clear_hw_state(plane_state);
memcpy(&plane_state->hw, &from_plane_state->hw,
sizeof(plane_state->hw));
if (plane_state->hw.fb)
drm_framebuffer_get(plane_state->hw.fb);
}
drm/i915: Fix crtc nv12 etc. plane bitmasks for DPMS off We only consider crtc_state->enable when initially calculating plane visibility. Later on we try to override the plane's state to invisible if the crtc is in DPMS off state (crtc_state->active==false). Unfortunately the code doing that only updates the plane_state.visible flag and the crtc_state.active_planes bimask, but forgets to update some of the other plane bitmasks stored in the crtc_state. Namely crtc_state.nv12_planes is left set up based on the original visibility check which makes icl_check_nv12_planes() pick a slave plane for the flagged plane in the bitmask. Later on we hit the watermark code which sees a plane with a slave assigned and it then makes the logical assumption that the master plane must itself be visible. Since the master's plane_state.visible flag was already cleared we get a WARN. Fix the problem by clearing all the plane bitmasks for DPMS off. This is more or less the wrong approach and instead we should calculate all the plane related state purely based crtc_state->enable (to guarantee that the subsequent DPMS on can't fail). However in the past we definitely had some roadblocks to making that happen. Not sure how many are left these days, but let's stick to the current approach since it's a much simpler fix to the immediate problem (the WARN). v2: Keep the visible=false, it's important (Rodrigo) Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200318174515.31637-1-ville.syrjala@linux.intel.com Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
2020-03-18 19:45:15 +02:00
void intel_plane_set_invisible(struct intel_crtc_state *crtc_state,
struct intel_plane_state *plane_state)
{
struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
crtc_state->active_planes &= ~BIT(plane->id);
crtc_state->nv12_planes &= ~BIT(plane->id);
crtc_state->c8_planes &= ~BIT(plane->id);
crtc_state->data_rate[plane->id] = 0;
crtc_state->min_cdclk[plane->id] = 0;
plane_state->uapi.visible = false;
}
int intel_plane_atomic_check_with_state(const struct intel_crtc_state *old_crtc_state,
struct intel_crtc_state *new_crtc_state,
const struct intel_plane_state *old_plane_state,
struct intel_plane_state *new_plane_state)
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
{
struct intel_plane *plane = to_intel_plane(new_plane_state->uapi.plane);
const struct drm_framebuffer *fb = new_plane_state->hw.fb;
int ret;
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
drm/i915: Fix crtc nv12 etc. plane bitmasks for DPMS off We only consider crtc_state->enable when initially calculating plane visibility. Later on we try to override the plane's state to invisible if the crtc is in DPMS off state (crtc_state->active==false). Unfortunately the code doing that only updates the plane_state.visible flag and the crtc_state.active_planes bimask, but forgets to update some of the other plane bitmasks stored in the crtc_state. Namely crtc_state.nv12_planes is left set up based on the original visibility check which makes icl_check_nv12_planes() pick a slave plane for the flagged plane in the bitmask. Later on we hit the watermark code which sees a plane with a slave assigned and it then makes the logical assumption that the master plane must itself be visible. Since the master's plane_state.visible flag was already cleared we get a WARN. Fix the problem by clearing all the plane bitmasks for DPMS off. This is more or less the wrong approach and instead we should calculate all the plane related state purely based crtc_state->enable (to guarantee that the subsequent DPMS on can't fail). However in the past we definitely had some roadblocks to making that happen. Not sure how many are left these days, but let's stick to the current approach since it's a much simpler fix to the immediate problem (the WARN). v2: Keep the visible=false, it's important (Rodrigo) Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200318174515.31637-1-ville.syrjala@linux.intel.com Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
2020-03-18 19:45:15 +02:00
intel_plane_set_invisible(new_crtc_state, new_plane_state);
new_crtc_state->enabled_planes &= ~BIT(plane->id);
if (!new_plane_state->hw.crtc && !old_plane_state->hw.crtc)
return 0;
ret = plane->check_plane(new_crtc_state, new_plane_state);
if (ret)
return ret;
if (fb)
new_crtc_state->enabled_planes |= BIT(plane->id);
/* FIXME pre-g4x don't work like this */
if (new_plane_state->uapi.visible)
new_crtc_state->active_planes |= BIT(plane->id);
if (new_plane_state->uapi.visible &&
intel_format_info_is_yuv_semiplanar(fb->format, fb->modifier))
new_crtc_state->nv12_planes |= BIT(plane->id);
if (new_plane_state->uapi.visible &&
fb->format->format == DRM_FORMAT_C8)
new_crtc_state->c8_planes |= BIT(plane->id);
if (new_plane_state->uapi.visible || old_plane_state->uapi.visible)
new_crtc_state->update_planes |= BIT(plane->id);
drm/i915: Make sure we have enough memory bandwidth on ICL ICL has so many planes that it can easily exceed the maximum effective memory bandwidth of the system. We must therefore check that we don't exceed that limit. The algorithm is very magic number heavy and lacks sufficient explanation for now. We also have no sane way to query the memory clock and timings, so we must rely on a combination of raw readout from the memory controller and hardcoded assumptions. The memory controller values obviously change as the system jumps between the different SAGV points, so we try to stabilize it first by disabling SAGV for the duration of the readout. The utilized bandwidth is tracked via a device wide atomic private object. That is actually not robust because we can't afford to enforce strict global ordering between the pipes. Thus I think I'll need to change this to simply chop up the available bandwidth between all the active pipes. Each pipe can then do whatever it wants as long as it doesn't exceed its budget. That scheme will also require that we assume that any number of planes could be active at any time. TODO: make it robust and deal with all the open questions v2: Sleep longer after disabling SAGV v3: Poll for the dclk to get raised (seen it take 250ms!) If the system has 2133MT/s memory then we pointlessly wait one full second :( v4: Use the new pcode interface to get the qgv points rather that using hardcoded numbers v5: Move the pcode stuff into intel_bw.c (Matt) s/intel_sagv_info/intel_qgv_info/ Do the NV12/P010 as per spec for now (Matt) s/IS_ICELAKE/IS_GEN11/ v6: Ignore bandwidth limits if the pcode query fails Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Acked-by: Clint Taylor <Clinton.A.Taylor@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190524153614.32410-1-ville.syrjala@linux.intel.com
2019-05-24 18:36:14 +03:00
new_crtc_state->data_rate[plane->id] =
intel_plane_data_rate(new_crtc_state, new_plane_state);
return intel_plane_atomic_calc_changes(old_crtc_state, new_crtc_state,
old_plane_state, new_plane_state);
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
}
static struct intel_plane *
intel_crtc_get_plane(struct intel_crtc *crtc, enum plane_id plane_id)
{
struct drm_i915_private *i915 = to_i915(crtc->base.dev);
struct intel_plane *plane;
for_each_intel_plane_on_crtc(&i915->drm, crtc, plane) {
if (plane->id == plane_id)
return plane;
}
return NULL;
}
int intel_plane_atomic_check(struct intel_atomic_state *state,
struct intel_plane *plane)
{
struct drm_i915_private *i915 = to_i915(state->base.dev);
struct intel_plane_state *new_plane_state =
intel_atomic_get_new_plane_state(state, plane);
const struct intel_plane_state *old_plane_state =
intel_atomic_get_old_plane_state(state, plane);
const struct intel_plane_state *new_master_plane_state;
struct intel_crtc *crtc = intel_crtc_for_pipe(i915, plane->pipe);
const struct intel_crtc_state *old_crtc_state =
intel_atomic_get_old_crtc_state(state, crtc);
struct intel_crtc_state *new_crtc_state =
intel_atomic_get_new_crtc_state(state, crtc);
if (new_crtc_state && new_crtc_state->bigjoiner_slave) {
struct intel_plane *master_plane =
intel_crtc_get_plane(new_crtc_state->bigjoiner_linked_crtc,
plane->id);
new_master_plane_state =
intel_atomic_get_new_plane_state(state, master_plane);
} else {
new_master_plane_state = new_plane_state;
}
intel_plane_copy_uapi_to_hw_state(new_plane_state,
new_master_plane_state,
crtc);
new_plane_state->uapi.visible = false;
if (!new_crtc_state)
return 0;
return intel_plane_atomic_check_with_state(old_crtc_state,
new_crtc_state,
old_plane_state,
new_plane_state);
}
static struct intel_plane *
skl_next_plane_to_commit(struct intel_atomic_state *state,
struct intel_crtc *crtc,
struct skl_ddb_entry entries_y[I915_MAX_PLANES],
struct skl_ddb_entry entries_uv[I915_MAX_PLANES],
unsigned int *update_mask)
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
{
struct intel_crtc_state *crtc_state =
intel_atomic_get_new_crtc_state(state, crtc);
struct intel_plane_state *plane_state;
struct intel_plane *plane;
int i;
if (*update_mask == 0)
return NULL;
for_each_new_intel_plane_in_state(state, plane, plane_state, i) {
enum plane_id plane_id = plane->id;
if (crtc->pipe != plane->pipe ||
!(*update_mask & BIT(plane_id)))
continue;
if (skl_ddb_allocation_overlaps(&crtc_state->wm.skl.plane_ddb_y[plane_id],
entries_y,
I915_MAX_PLANES, plane_id) ||
skl_ddb_allocation_overlaps(&crtc_state->wm.skl.plane_ddb_uv[plane_id],
entries_uv,
I915_MAX_PLANES, plane_id))
continue;
*update_mask &= ~BIT(plane_id);
entries_y[plane_id] = crtc_state->wm.skl.plane_ddb_y[plane_id];
entries_uv[plane_id] = crtc_state->wm.skl.plane_ddb_uv[plane_id];
return plane;
}
/* should never happen */
drm_WARN_ON(state->base.dev, 1);
return NULL;
}
drm/i915: Split update_plane() into update_noarm() + update_arm() The amount of plane registers we have to write has been steadily increasing, putting more pressure on the vblank evasion mechanism and forcing us to increase its time budget. Let's try to take some of the pressure off by splitting plane updates into two parts: 1) write all non-self arming plane registers, ie. the registers where the write actually does nothing until a separate arming register is also written which will cause the hardware to latch the new register values at the next start of vblank 2) write all self arming plane registers, ie. registers which always just latch at the next start of vblank, and registers which also arm other registers to do so Here we just provide the mechanism, but don't actually implement the split on any platform yet. so everything stays now in the _arm() hooks. Subsequently we can move a whole bunch of stuff into the _noarm() part, especially in more modern platforms where the number of registers we have to write is also the greatest. On older platforms this is less beneficial probably, but no real reason to deviate from a common behaviour. And let's sprinkle some TODOs around the areas that will need adapting. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211018115030.3547-5-ville.syrjala@linux.intel.com Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
2021-10-18 14:50:25 +03:00
void intel_plane_update_noarm(struct intel_plane *plane,
const struct intel_crtc_state *crtc_state,
const struct intel_plane_state *plane_state)
{
struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
drm/i915: Split update_plane() into update_noarm() + update_arm() The amount of plane registers we have to write has been steadily increasing, putting more pressure on the vblank evasion mechanism and forcing us to increase its time budget. Let's try to take some of the pressure off by splitting plane updates into two parts: 1) write all non-self arming plane registers, ie. the registers where the write actually does nothing until a separate arming register is also written which will cause the hardware to latch the new register values at the next start of vblank 2) write all self arming plane registers, ie. registers which always just latch at the next start of vblank, and registers which also arm other registers to do so Here we just provide the mechanism, but don't actually implement the split on any platform yet. so everything stays now in the _arm() hooks. Subsequently we can move a whole bunch of stuff into the _noarm() part, especially in more modern platforms where the number of registers we have to write is also the greatest. On older platforms this is less beneficial probably, but no real reason to deviate from a common behaviour. And let's sprinkle some TODOs around the areas that will need adapting. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211018115030.3547-5-ville.syrjala@linux.intel.com Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
2021-10-18 14:50:25 +03:00
trace_intel_plane_update_noarm(&plane->base, crtc);
if (plane->update_noarm)
plane->update_noarm(plane, crtc_state, plane_state);
}
void intel_plane_update_arm(struct intel_plane *plane,
const struct intel_crtc_state *crtc_state,
const struct intel_plane_state *plane_state)
{
struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
trace_intel_plane_update_arm(&plane->base, crtc);
if (plane_state->do_async_flip)
plane->async_flip(plane, crtc_state, plane_state, true);
else
drm/i915: Split update_plane() into update_noarm() + update_arm() The amount of plane registers we have to write has been steadily increasing, putting more pressure on the vblank evasion mechanism and forcing us to increase its time budget. Let's try to take some of the pressure off by splitting plane updates into two parts: 1) write all non-self arming plane registers, ie. the registers where the write actually does nothing until a separate arming register is also written which will cause the hardware to latch the new register values at the next start of vblank 2) write all self arming plane registers, ie. registers which always just latch at the next start of vblank, and registers which also arm other registers to do so Here we just provide the mechanism, but don't actually implement the split on any platform yet. so everything stays now in the _arm() hooks. Subsequently we can move a whole bunch of stuff into the _noarm() part, especially in more modern platforms where the number of registers we have to write is also the greatest. On older platforms this is less beneficial probably, but no real reason to deviate from a common behaviour. And let's sprinkle some TODOs around the areas that will need adapting. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211018115030.3547-5-ville.syrjala@linux.intel.com Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
2021-10-18 14:50:25 +03:00
plane->update_arm(plane, crtc_state, plane_state);
}
drm/i915: Split update_plane() into update_noarm() + update_arm() The amount of plane registers we have to write has been steadily increasing, putting more pressure on the vblank evasion mechanism and forcing us to increase its time budget. Let's try to take some of the pressure off by splitting plane updates into two parts: 1) write all non-self arming plane registers, ie. the registers where the write actually does nothing until a separate arming register is also written which will cause the hardware to latch the new register values at the next start of vblank 2) write all self arming plane registers, ie. registers which always just latch at the next start of vblank, and registers which also arm other registers to do so Here we just provide the mechanism, but don't actually implement the split on any platform yet. so everything stays now in the _arm() hooks. Subsequently we can move a whole bunch of stuff into the _noarm() part, especially in more modern platforms where the number of registers we have to write is also the greatest. On older platforms this is less beneficial probably, but no real reason to deviate from a common behaviour. And let's sprinkle some TODOs around the areas that will need adapting. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211018115030.3547-5-ville.syrjala@linux.intel.com Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
2021-10-18 14:50:25 +03:00
void intel_plane_disable_arm(struct intel_plane *plane,
const struct intel_crtc_state *crtc_state)
{
struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
drm/i915: Split update_plane() into update_noarm() + update_arm() The amount of plane registers we have to write has been steadily increasing, putting more pressure on the vblank evasion mechanism and forcing us to increase its time budget. Let's try to take some of the pressure off by splitting plane updates into two parts: 1) write all non-self arming plane registers, ie. the registers where the write actually does nothing until a separate arming register is also written which will cause the hardware to latch the new register values at the next start of vblank 2) write all self arming plane registers, ie. registers which always just latch at the next start of vblank, and registers which also arm other registers to do so Here we just provide the mechanism, but don't actually implement the split on any platform yet. so everything stays now in the _arm() hooks. Subsequently we can move a whole bunch of stuff into the _noarm() part, especially in more modern platforms where the number of registers we have to write is also the greatest. On older platforms this is less beneficial probably, but no real reason to deviate from a common behaviour. And let's sprinkle some TODOs around the areas that will need adapting. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211018115030.3547-5-ville.syrjala@linux.intel.com Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
2021-10-18 14:50:25 +03:00
trace_intel_plane_disable_arm(&plane->base, crtc);
plane->disable_arm(plane, crtc_state);
}
drm/i915: Split update_plane() into update_noarm() + update_arm() The amount of plane registers we have to write has been steadily increasing, putting more pressure on the vblank evasion mechanism and forcing us to increase its time budget. Let's try to take some of the pressure off by splitting plane updates into two parts: 1) write all non-self arming plane registers, ie. the registers where the write actually does nothing until a separate arming register is also written which will cause the hardware to latch the new register values at the next start of vblank 2) write all self arming plane registers, ie. registers which always just latch at the next start of vblank, and registers which also arm other registers to do so Here we just provide the mechanism, but don't actually implement the split on any platform yet. so everything stays now in the _arm() hooks. Subsequently we can move a whole bunch of stuff into the _noarm() part, especially in more modern platforms where the number of registers we have to write is also the greatest. On older platforms this is less beneficial probably, but no real reason to deviate from a common behaviour. And let's sprinkle some TODOs around the areas that will need adapting. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211018115030.3547-5-ville.syrjala@linux.intel.com Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
2021-10-18 14:50:25 +03:00
void intel_update_planes_on_crtc(struct intel_atomic_state *state,
struct intel_crtc *crtc)
{
struct intel_crtc_state *new_crtc_state =
intel_atomic_get_new_crtc_state(state, crtc);
u32 update_mask = new_crtc_state->update_planes;
struct intel_plane_state *new_plane_state;
struct intel_plane *plane;
int i;
if (new_crtc_state->uapi.async_flip)
return;
/*
* Since we only write non-arming registers here,
* the order does not matter even for skl+.
*/
for_each_new_intel_plane_in_state(state, plane, new_plane_state, i) {
if (crtc->pipe != plane->pipe ||
!(update_mask & BIT(plane->id)))
continue;
/* TODO: for mailbox updates this should be skipped */
if (new_plane_state->uapi.visible ||
new_plane_state->planar_slave)
intel_plane_update_noarm(plane, new_crtc_state, new_plane_state);
}
}
void skl_arm_planes_on_crtc(struct intel_atomic_state *state,
struct intel_crtc *crtc)
{
struct intel_crtc_state *old_crtc_state =
intel_atomic_get_old_crtc_state(state, crtc);
struct intel_crtc_state *new_crtc_state =
intel_atomic_get_new_crtc_state(state, crtc);
struct skl_ddb_entry entries_y[I915_MAX_PLANES];
struct skl_ddb_entry entries_uv[I915_MAX_PLANES];
u32 update_mask = new_crtc_state->update_planes;
struct intel_plane *plane;
memcpy(entries_y, old_crtc_state->wm.skl.plane_ddb_y,
sizeof(old_crtc_state->wm.skl.plane_ddb_y));
memcpy(entries_uv, old_crtc_state->wm.skl.plane_ddb_uv,
sizeof(old_crtc_state->wm.skl.plane_ddb_uv));
while ((plane = skl_next_plane_to_commit(state, crtc,
entries_y, entries_uv,
&update_mask))) {
struct intel_plane_state *new_plane_state =
intel_atomic_get_new_plane_state(state, plane);
drm/i915: Split update_plane() into update_noarm() + update_arm() The amount of plane registers we have to write has been steadily increasing, putting more pressure on the vblank evasion mechanism and forcing us to increase its time budget. Let's try to take some of the pressure off by splitting plane updates into two parts: 1) write all non-self arming plane registers, ie. the registers where the write actually does nothing until a separate arming register is also written which will cause the hardware to latch the new register values at the next start of vblank 2) write all self arming plane registers, ie. registers which always just latch at the next start of vblank, and registers which also arm other registers to do so Here we just provide the mechanism, but don't actually implement the split on any platform yet. so everything stays now in the _arm() hooks. Subsequently we can move a whole bunch of stuff into the _noarm() part, especially in more modern platforms where the number of registers we have to write is also the greatest. On older platforms this is less beneficial probably, but no real reason to deviate from a common behaviour. And let's sprinkle some TODOs around the areas that will need adapting. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211018115030.3547-5-ville.syrjala@linux.intel.com Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
2021-10-18 14:50:25 +03:00
/*
* TODO: for mailbox updates intel_plane_update_noarm()
* would have to be called here as well.
*/
if (new_plane_state->uapi.visible ||
drm/i915: Split update_plane() into update_noarm() + update_arm() The amount of plane registers we have to write has been steadily increasing, putting more pressure on the vblank evasion mechanism and forcing us to increase its time budget. Let's try to take some of the pressure off by splitting plane updates into two parts: 1) write all non-self arming plane registers, ie. the registers where the write actually does nothing until a separate arming register is also written which will cause the hardware to latch the new register values at the next start of vblank 2) write all self arming plane registers, ie. registers which always just latch at the next start of vblank, and registers which also arm other registers to do so Here we just provide the mechanism, but don't actually implement the split on any platform yet. so everything stays now in the _arm() hooks. Subsequently we can move a whole bunch of stuff into the _noarm() part, especially in more modern platforms where the number of registers we have to write is also the greatest. On older platforms this is less beneficial probably, but no real reason to deviate from a common behaviour. And let's sprinkle some TODOs around the areas that will need adapting. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211018115030.3547-5-ville.syrjala@linux.intel.com Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
2021-10-18 14:50:25 +03:00
new_plane_state->planar_slave)
intel_plane_update_arm(plane, new_crtc_state, new_plane_state);
else
intel_plane_disable_arm(plane, new_crtc_state);
}
}
drm/i915: Split update_plane() into update_noarm() + update_arm() The amount of plane registers we have to write has been steadily increasing, putting more pressure on the vblank evasion mechanism and forcing us to increase its time budget. Let's try to take some of the pressure off by splitting plane updates into two parts: 1) write all non-self arming plane registers, ie. the registers where the write actually does nothing until a separate arming register is also written which will cause the hardware to latch the new register values at the next start of vblank 2) write all self arming plane registers, ie. registers which always just latch at the next start of vblank, and registers which also arm other registers to do so Here we just provide the mechanism, but don't actually implement the split on any platform yet. so everything stays now in the _arm() hooks. Subsequently we can move a whole bunch of stuff into the _noarm() part, especially in more modern platforms where the number of registers we have to write is also the greatest. On older platforms this is less beneficial probably, but no real reason to deviate from a common behaviour. And let's sprinkle some TODOs around the areas that will need adapting. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211018115030.3547-5-ville.syrjala@linux.intel.com Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
2021-10-18 14:50:25 +03:00
void i9xx_arm_planes_on_crtc(struct intel_atomic_state *state,
struct intel_crtc *crtc)
{
struct intel_crtc_state *new_crtc_state =
intel_atomic_get_new_crtc_state(state, crtc);
u32 update_mask = new_crtc_state->update_planes;
struct intel_plane_state *new_plane_state;
struct intel_plane *plane;
int i;
for_each_new_intel_plane_in_state(state, plane, new_plane_state, i) {
if (crtc->pipe != plane->pipe ||
!(update_mask & BIT(plane->id)))
continue;
drm/i915: Split update_plane() into update_noarm() + update_arm() The amount of plane registers we have to write has been steadily increasing, putting more pressure on the vblank evasion mechanism and forcing us to increase its time budget. Let's try to take some of the pressure off by splitting plane updates into two parts: 1) write all non-self arming plane registers, ie. the registers where the write actually does nothing until a separate arming register is also written which will cause the hardware to latch the new register values at the next start of vblank 2) write all self arming plane registers, ie. registers which always just latch at the next start of vblank, and registers which also arm other registers to do so Here we just provide the mechanism, but don't actually implement the split on any platform yet. so everything stays now in the _arm() hooks. Subsequently we can move a whole bunch of stuff into the _noarm() part, especially in more modern platforms where the number of registers we have to write is also the greatest. On older platforms this is less beneficial probably, but no real reason to deviate from a common behaviour. And let's sprinkle some TODOs around the areas that will need adapting. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211018115030.3547-5-ville.syrjala@linux.intel.com Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
2021-10-18 14:50:25 +03:00
/*
* TODO: for mailbox updates intel_plane_update_noarm()
* would have to be called here as well.
*/
if (new_plane_state->uapi.visible)
drm/i915: Split update_plane() into update_noarm() + update_arm() The amount of plane registers we have to write has been steadily increasing, putting more pressure on the vblank evasion mechanism and forcing us to increase its time budget. Let's try to take some of the pressure off by splitting plane updates into two parts: 1) write all non-self arming plane registers, ie. the registers where the write actually does nothing until a separate arming register is also written which will cause the hardware to latch the new register values at the next start of vblank 2) write all self arming plane registers, ie. registers which always just latch at the next start of vblank, and registers which also arm other registers to do so Here we just provide the mechanism, but don't actually implement the split on any platform yet. so everything stays now in the _arm() hooks. Subsequently we can move a whole bunch of stuff into the _noarm() part, especially in more modern platforms where the number of registers we have to write is also the greatest. On older platforms this is less beneficial probably, but no real reason to deviate from a common behaviour. And let's sprinkle some TODOs around the areas that will need adapting. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211018115030.3547-5-ville.syrjala@linux.intel.com Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
2021-10-18 14:50:25 +03:00
intel_plane_update_arm(plane, new_crtc_state, new_plane_state);
else
drm/i915: Split update_plane() into update_noarm() + update_arm() The amount of plane registers we have to write has been steadily increasing, putting more pressure on the vblank evasion mechanism and forcing us to increase its time budget. Let's try to take some of the pressure off by splitting plane updates into two parts: 1) write all non-self arming plane registers, ie. the registers where the write actually does nothing until a separate arming register is also written which will cause the hardware to latch the new register values at the next start of vblank 2) write all self arming plane registers, ie. registers which always just latch at the next start of vblank, and registers which also arm other registers to do so Here we just provide the mechanism, but don't actually implement the split on any platform yet. so everything stays now in the _arm() hooks. Subsequently we can move a whole bunch of stuff into the _noarm() part, especially in more modern platforms where the number of registers we have to write is also the greatest. On older platforms this is less beneficial probably, but no real reason to deviate from a common behaviour. And let's sprinkle some TODOs around the areas that will need adapting. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211018115030.3547-5-ville.syrjala@linux.intel.com Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
2021-10-18 14:50:25 +03:00
intel_plane_disable_arm(plane, new_crtc_state);
}
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
}
int intel_atomic_plane_check_clipping(struct intel_plane_state *plane_state,
struct intel_crtc_state *crtc_state,
int min_scale, int max_scale,
bool can_position)
{
struct drm_i915_private *i915 = to_i915(plane_state->uapi.plane->dev);
struct drm_framebuffer *fb = plane_state->hw.fb;
struct drm_rect *src = &plane_state->uapi.src;
struct drm_rect *dst = &plane_state->uapi.dst;
unsigned int rotation = plane_state->hw.rotation;
struct drm_rect clip = {};
int hscale, vscale;
if (!fb) {
plane_state->uapi.visible = false;
return 0;
}
drm_rect_rotate(src, fb->width << 16, fb->height << 16, rotation);
/* Check scaling */
hscale = drm_rect_calc_hscale(src, dst, min_scale, max_scale);
vscale = drm_rect_calc_vscale(src, dst, min_scale, max_scale);
if (hscale < 0 || vscale < 0) {
drm_dbg_kms(&i915->drm, "Invalid scaling of plane\n");
drm_rect_debug_print("src: ", src, true);
drm_rect_debug_print("dst: ", dst, false);
return -ERANGE;
}
if (crtc_state->hw.enable) {
clip.x2 = crtc_state->pipe_src_w;
clip.y2 = crtc_state->pipe_src_h;
}
/* right side of the image is on the slave crtc, adjust dst to match */
if (crtc_state->bigjoiner_slave)
drm_rect_translate(dst, -crtc_state->pipe_src_w, 0);
/*
* FIXME: This might need further adjustment for seamless scaling
* with phase information, for the 2p2 and 2p1 scenarios.
*/
plane_state->uapi.visible = drm_rect_clip_scaled(src, dst, &clip);
drm_rect_rotate_inv(src, fb->width << 16, fb->height << 16, rotation);
if (!can_position && plane_state->uapi.visible &&
!drm_rect_equals(dst, &clip)) {
drm_dbg_kms(&i915->drm, "Plane must cover entire CRTC\n");
drm_rect_debug_print("dst: ", dst, false);
drm_rect_debug_print("clip: ", &clip, false);
return -EINVAL;
}
return 0;
}
struct wait_rps_boost {
struct wait_queue_entry wait;
struct drm_crtc *crtc;
struct i915_request *request;
};
static int do_rps_boost(struct wait_queue_entry *_wait,
unsigned mode, int sync, void *key)
{
struct wait_rps_boost *wait = container_of(_wait, typeof(*wait), wait);
struct i915_request *rq = wait->request;
/*
* If we missed the vblank, but the request is already running it
* is reasonable to assume that it will complete before the next
* vblank without our intervention, so leave RPS alone.
*/
if (!i915_request_started(rq))
intel_rps_boost(rq);
i915_request_put(rq);
drm_crtc_vblank_put(wait->crtc);
list_del(&wait->wait.entry);
kfree(wait);
return 1;
}
static void add_rps_boost_after_vblank(struct drm_crtc *crtc,
struct dma_fence *fence)
{
struct wait_rps_boost *wait;
if (!dma_fence_is_i915(fence))
return;
if (DISPLAY_VER(to_i915(crtc->dev)) < 6)
return;
if (drm_crtc_vblank_get(crtc))
return;
wait = kmalloc(sizeof(*wait), GFP_KERNEL);
if (!wait) {
drm_crtc_vblank_put(crtc);
return;
}
wait->request = to_request(dma_fence_get(fence));
wait->crtc = crtc;
wait->wait.func = do_rps_boost;
wait->wait.flags = 0;
add_wait_queue(drm_crtc_vblank_waitqueue(crtc), &wait->wait);
}
/**
* intel_prepare_plane_fb - Prepare fb for usage on plane
* @_plane: drm plane to prepare for
* @_new_plane_state: the plane state being prepared
*
* Prepares a framebuffer for usage on a display plane. Generally this
* involves pinning the underlying object and updating the frontbuffer tracking
* bits. Some older platforms need special physical address handling for
* cursor planes.
*
* Returns 0 on success, negative error code on failure.
*/
static int
intel_prepare_plane_fb(struct drm_plane *_plane,
struct drm_plane_state *_new_plane_state)
{
struct i915_sched_attr attr = { .priority = I915_PRIORITY_DISPLAY };
struct intel_plane *plane = to_intel_plane(_plane);
struct intel_plane_state *new_plane_state =
to_intel_plane_state(_new_plane_state);
struct intel_atomic_state *state =
to_intel_atomic_state(new_plane_state->uapi.state);
struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
const struct intel_plane_state *old_plane_state =
intel_atomic_get_old_plane_state(state, plane);
struct drm_i915_gem_object *obj = intel_fb_obj(new_plane_state->hw.fb);
struct drm_i915_gem_object *old_obj = intel_fb_obj(old_plane_state->hw.fb);
int ret;
if (old_obj) {
const struct intel_crtc_state *crtc_state =
intel_atomic_get_new_crtc_state(state,
to_intel_crtc(old_plane_state->hw.crtc));
/* Big Hammer, we also need to ensure that any pending
* MI_WAIT_FOR_EVENT inside a user batch buffer on the
* current scanout is retired before unpinning the old
* framebuffer. Note that we rely on userspace rendering
* into the buffer attached to the pipe they are waiting
* on. If not, userspace generates a GPU hang with IPEHR
* point to the MI_WAIT_FOR_EVENT.
*
* This should only fail upon a hung GPU, in which case we
* can safely continue.
*/
if (intel_crtc_needs_modeset(crtc_state)) {
ret = i915_sw_fence_await_reservation(&state->commit_ready,
old_obj->base.resv, NULL,
false, 0,
GFP_KERNEL);
if (ret < 0)
return ret;
}
}
if (new_plane_state->uapi.fence) { /* explicit fencing */
i915_gem_fence_wait_priority(new_plane_state->uapi.fence,
&attr);
ret = i915_sw_fence_await_dma_fence(&state->commit_ready,
new_plane_state->uapi.fence,
i915_fence_timeout(dev_priv),
GFP_KERNEL);
if (ret < 0)
return ret;
}
if (!obj)
return 0;
ret = intel_plane_pin_fb(new_plane_state);
if (ret)
return ret;
i915_gem_object_wait_priority(obj, 0, &attr);
if (!new_plane_state->uapi.fence) { /* implicit fencing */
struct dma_resv_iter cursor;
struct dma_fence *fence;
ret = i915_sw_fence_await_reservation(&state->commit_ready,
obj->base.resv, NULL,
false,
i915_fence_timeout(dev_priv),
GFP_KERNEL);
if (ret < 0)
goto unpin_fb;
dma_resv_iter_begin(&cursor, obj->base.resv, false);
dma_resv_for_each_fence_unlocked(&cursor, fence) {
add_rps_boost_after_vblank(new_plane_state->hw.crtc,
fence);
}
dma_resv_iter_end(&cursor);
} else {
add_rps_boost_after_vblank(new_plane_state->hw.crtc,
new_plane_state->uapi.fence);
}
/*
* We declare pageflips to be interactive and so merit a small bias
* towards upclocking to deliver the frame on time. By only changing
* the RPS thresholds to sample more regularly and aim for higher
* clocks we can hopefully deliver low power workloads (like kodi)
* that are not quite steady state without resorting to forcing
* maximum clocks following a vblank miss (see do_rps_boost()).
*/
if (!state->rps_interactive) {
intel_rps_mark_interactive(&to_gt(dev_priv)->rps, true);
state->rps_interactive = true;
}
return 0;
unpin_fb:
intel_plane_unpin_fb(new_plane_state);
return ret;
}
/**
* intel_cleanup_plane_fb - Cleans up an fb after plane use
* @plane: drm plane to clean up for
* @_old_plane_state: the state from the previous modeset
*
* Cleans up a framebuffer that has just been removed from a plane.
*/
static void
intel_cleanup_plane_fb(struct drm_plane *plane,
struct drm_plane_state *_old_plane_state)
{
struct intel_plane_state *old_plane_state =
to_intel_plane_state(_old_plane_state);
struct intel_atomic_state *state =
to_intel_atomic_state(old_plane_state->uapi.state);
struct drm_i915_private *dev_priv = to_i915(plane->dev);
struct drm_i915_gem_object *obj = intel_fb_obj(old_plane_state->hw.fb);
if (!obj)
return;
if (state->rps_interactive) {
intel_rps_mark_interactive(&to_gt(dev_priv)->rps, false);
state->rps_interactive = false;
}
/* Should only be called after a successful intel_prepare_plane_fb()! */
intel_plane_unpin_fb(old_plane_state);
}
static const struct drm_plane_helper_funcs intel_plane_helper_funcs = {
drm/i915: Move to atomic plane helpers (v9) Switch plane handling to use the atomic plane helpers. This means that rather than provide our own implementations of .update_plane() and .disable_plane(), we expose the lower-level check/prepare/commit/cleanup entrypoints and let the DRM core implement update/disable for us using those entrypoints. The other main change that falls out of this patch is that our drm_plane's will now always have a valid plane->state that contains the relevant plane state (initial state is allocated at plane creation). The base drm_plane_state pointed to holds the requested source/dest coordinates, and the subclassed intel_plane_state holds the adjusted values that our driver actually uses. v2: - Renamed file from intel_atomic.c to intel_atomic_plane.c (Daniel) - Fix a copy/paste comment mistake (Bob) v3: - Use prepare/cleanup functions that we've already factored out - Use newly refactored pre_commit/commit/post_commit to avoid sleeping during vblank evasion v4: - Rebase to latest di-nightly requires adding an 'old_state' parameter to atomic_update; v5: - Must have botched a rebase somewhere and lost some work. Restore state 'dirty' flag to let begin/end code know which planes to run the pre_commit/post_commit hooks for. This would have actually shown up as broken in the next commit rather than this one. v6: - Squash kerneldoc patch into this one. - Previous patches have now already taken care of most of the infrastructure that used to be in this patch. All we're adding here now is some thin wrappers. v7: - Check return of intel_plane_duplicate_state() for allocation failures. v8: - Drop unused drm_plane_state -> intel_plane_state cast. (Ander) - Squash in actual transition to plane helpers. Significant refactoring earlier in the patchset has made the combined prep+transition much easier to swallow than it was in earlier iterations. (Ander) v9: - s/track_fbs/disabled_planes/ in the atomic crtc flags. The only fb's we need to update frontbuffer tracking for are those on a plane about to be disabled (since the atomic helpers never call prepare_fb() when disabling a plane), so the new name more accurately describes what we're actually tracking. Testcase: igt/kms_plane Testcase: igt/kms_universal_plane Testcase: igt/kms_cursor_crc Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-12-23 10:41:52 -08:00
.prepare_fb = intel_prepare_plane_fb,
.cleanup_fb = intel_cleanup_plane_fb,
};
void intel_plane_helper_add(struct intel_plane *plane)
{
drm_plane_helper_add(&plane->base, &intel_plane_helper_funcs);
}