leandrof/linux
1
0
Fork 0
forked from Minki/linux
Code Issues Pull Requests Packages Projects Releases Wiki Activity
4e0cd68115
linux/drivers/gpu/drm/drm_atomic_helper.c

1967 lines
51 KiB
C
Raw Normal View History

drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
/*
* Copyright (C) 2014 Red Hat
* Copyright (C) 2014 Intel Corp.
*
* 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 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 COPYRIGHT HOLDER(S) OR AUTHOR(S) 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.
*
* Authors:
* Rob Clark <robdclark@gmail.com>
* Daniel Vetter <daniel.vetter@ffwll.ch>
*/
#include <drm/drmP.h>
#include <drm/drm_atomic.h>
#include <drm/drm_plane_helper.h>
#include <drm/drm_crtc_helper.h>
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
#include <drm/drm_atomic_helper.h>
drm/atomic: Integrate fence support This patch is for enabling async commits. It replaces an earlier approach which added an async boolean paramter to the ->prepare_fb callbacks. The idea is that prepare_fb picks up the right fence to synchronize against, which is then used by the synchronous commit helper. For async commits drivers can either register a callback to the fence or simply do the synchronous wait in their async work queue. v2: Remove unused variable. v3: Only wait for fences after the point of no return in the part of the commit function which can be run asynchronously. This is after the atomic state has been swapped in, hence now check plane->state->fence. Also add a WARN_ON to make sure we don't try to wait on a fence when there's no fb, just as a sanity check. Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-10-29 10:34:56 +00:00
#include <linux/fence.h>
drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
drm: Docbook integration and over sections for all the new helpers In all cases the text requires that new drivers are converted to the atomic interfaces. v2: Add overview for state handling. v3: Review from Sean: Some spelling fixes and drop the misguided hunk to remove rgba8888 from the plane helpers compat list. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-06 19:53:29 +00:00
/**
* DOC: overview
*
* This helper library provides implementations of check and commit functions on
* top of the CRTC modeset helper callbacks and the plane helper callbacks. It
* also provides convenience implementations for the atomic state handling
* callbacks for drivers which don't need to subclass the drm core structures to
* add their own additional internal state.
*
* This library also provides default implementations for the check callback in
* drm_atomic_helper_check and for the commit callback with
* drm_atomic_helper_commit. But the individual stages and callbacks are expose
* to allow drivers to mix and match and e.g. use the plane helpers only
* together with a driver private modeset implementation.
*
* This library also provides implementations for all the legacy driver
* interfaces on top of the atomic interface. See drm_atomic_helper_set_config,
* drm_atomic_helper_disable_plane, drm_atomic_helper_disable_plane and the
* various functions to implement set_property callbacks. New drivers must not
* implement these functions themselves but must use the provided helpers.
*/
drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
static void
drm_atomic_helper_plane_changed(struct drm_atomic_state *state,
struct drm_plane_state *plane_state,
struct drm_plane *plane)
{
struct drm_crtc_state *crtc_state;
if (plane->state->crtc) {
crtc_state = state->crtc_states[drm_crtc_index(plane->crtc)];
if (WARN_ON(!crtc_state))
return;
crtc_state->planes_changed = true;
}
if (plane_state->crtc) {
crtc_state =
state->crtc_states[drm_crtc_index(plane_state->crtc)];
if (WARN_ON(!crtc_state))
return;
crtc_state->planes_changed = true;
}
}
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
static struct drm_crtc *
get_current_crtc_for_encoder(struct drm_device *dev,
struct drm_encoder *encoder)
{
struct drm_mode_config *config = &dev->mode_config;
struct drm_connector *connector;
WARN_ON(!drm_modeset_is_locked(&config->connection_mutex));
list_for_each_entry(connector, &config->connector_list, head) {
if (connector->state->best_encoder != encoder)
continue;
return connector->state->crtc;
}
return NULL;
}
static int
steal_encoder(struct drm_atomic_state *state,
struct drm_encoder *encoder,
struct drm_crtc *encoder_crtc)
{
struct drm_mode_config *config = &state->dev->mode_config;
struct drm_crtc_state *crtc_state;
struct drm_connector *connector;
struct drm_connector_state *connector_state;
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
int ret;
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
/*
* We can only steal an encoder coming from a connector, which means we
* must already hold the connection_mutex.
*/
WARN_ON(!drm_modeset_is_locked(&config->connection_mutex));
DRM_DEBUG_KMS("[ENCODER:%d:%s] in use on [CRTC:%d], stealing it\n",
encoder->base.id, encoder->name,
encoder_crtc->base.id);
crtc_state = drm_atomic_get_crtc_state(state, encoder_crtc);
if (IS_ERR(crtc_state))
return PTR_ERR(crtc_state);
crtc_state->mode_changed = true;
list_for_each_entry(connector, &config->connector_list, head) {
if (connector->state->best_encoder != encoder)
continue;
DRM_DEBUG_KMS("Stealing encoder from [CONNECTOR:%d:%s]\n",
connector->base.id,
connector->name);
connector_state = drm_atomic_get_connector_state(state,
connector);
if (IS_ERR(connector_state))
return PTR_ERR(connector_state);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
ret = drm_atomic_set_crtc_for_connector(connector_state, NULL);
if (ret)
return ret;
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
connector_state->best_encoder = NULL;
}
return 0;
}
static int
update_connector_routing(struct drm_atomic_state *state, int conn_idx)
{
struct drm_connector_helper_funcs *funcs;
struct drm_encoder *new_encoder;
struct drm_crtc *encoder_crtc;
struct drm_connector *connector;
struct drm_connector_state *connector_state;
struct drm_crtc_state *crtc_state;
int idx, ret;
connector = state->connectors[conn_idx];
connector_state = state->connector_states[conn_idx];
if (!connector)
return 0;
DRM_DEBUG_KMS("Updating routing for [CONNECTOR:%d:%s]\n",
connector->base.id,
connector->name);
if (connector->state->crtc != connector_state->crtc) {
if (connector->state->crtc) {
idx = drm_crtc_index(connector->state->crtc);
crtc_state = state->crtc_states[idx];
crtc_state->mode_changed = true;
}
if (connector_state->crtc) {
idx = drm_crtc_index(connector_state->crtc);
crtc_state = state->crtc_states[idx];
crtc_state->mode_changed = true;
}
}
if (!connector_state->crtc) {
DRM_DEBUG_KMS("Disabling [CONNECTOR:%d:%s]\n",
connector->base.id,
connector->name);
connector_state->best_encoder = NULL;
return 0;
}
funcs = connector->helper_private;
new_encoder = funcs->best_encoder(connector);
if (!new_encoder) {
DRM_DEBUG_KMS("No suitable encoder found for [CONNECTOR:%d:%s]\n",
connector->base.id,
connector->name);
return -EINVAL;
}
if (new_encoder == connector_state->best_encoder) {
DRM_DEBUG_KMS("[CONNECTOR:%d:%s] keeps [ENCODER:%d:%s], now on [CRTC:%d]\n",
connector->base.id,
connector->name,
new_encoder->base.id,
new_encoder->name,
connector_state->crtc->base.id);
return 0;
}
encoder_crtc = get_current_crtc_for_encoder(state->dev,
new_encoder);
if (encoder_crtc) {
ret = steal_encoder(state, new_encoder, encoder_crtc);
if (ret) {
DRM_DEBUG_KMS("Encoder stealing failed for [CONNECTOR:%d:%s]\n",
connector->base.id,
connector->name);
return ret;
}
}
connector_state->best_encoder = new_encoder;
idx = drm_crtc_index(connector_state->crtc);
crtc_state = state->crtc_states[idx];
crtc_state->mode_changed = true;
DRM_DEBUG_KMS("[CONNECTOR:%d:%s] using [ENCODER:%d:%s] on [CRTC:%d]\n",
connector->base.id,
connector->name,
new_encoder->base.id,
new_encoder->name,
connector_state->crtc->base.id);
return 0;
}
static int
mode_fixup(struct drm_atomic_state *state)
{
int ncrtcs = state->dev->mode_config.num_crtc;
struct drm_crtc_state *crtc_state;
struct drm_connector_state *conn_state;
int i;
bool ret;
for (i = 0; i < ncrtcs; i++) {
crtc_state = state->crtc_states[i];
if (!crtc_state || !crtc_state->mode_changed)
continue;
drm_mode_copy(&crtc_state->adjusted_mode, &crtc_state->mode);
}
drm/atomic: Don't overrun the connector array when hotplugging Yet another fallout from not considering DP MST hotplug. With the previous patches we have stable indices, but it might still happen that a connector gets added between when we allocate the array and when we actually add a connector. Especially when we back off due to ww mutex contention or similar issues. So store the sizes of the arrays in struct drm_atomic_state and double check them. We don't really care about races except that we want to use a consistent value, so ACCESS_ONCE is all we need. And if we indeed notice that we'd overrun the array then just give up and restart the entire ioctl. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:08 +00:00
for (i = 0; i < state->num_connector; i++) {
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
struct drm_encoder_helper_funcs *funcs;
struct drm_encoder *encoder;
conn_state = state->connector_states[i];
if (!conn_state)
continue;
WARN_ON(!!conn_state->best_encoder != !!conn_state->crtc);
if (!conn_state->crtc || !conn_state->best_encoder)
continue;
crtc_state =
state->crtc_states[drm_crtc_index(conn_state->crtc)];
/*
* Each encoder has at most one connector (since we always steal
* it away), so we won't call ->mode_fixup twice.
*/
encoder = conn_state->best_encoder;
funcs = encoder->helper_private;
if (encoder->bridge && encoder->bridge->funcs->mode_fixup) {
ret = encoder->bridge->funcs->mode_fixup(
encoder->bridge, &crtc_state->mode,
&crtc_state->adjusted_mode);
if (!ret) {
DRM_DEBUG_KMS("Bridge fixup failed\n");
return -EINVAL;
}
}
ret = funcs->mode_fixup(encoder, &crtc_state->mode,
&crtc_state->adjusted_mode);
if (!ret) {
DRM_DEBUG_KMS("[ENCODER:%d:%s] fixup failed\n",
encoder->base.id, encoder->name);
return -EINVAL;
}
}
for (i = 0; i < ncrtcs; i++) {
struct drm_crtc_helper_funcs *funcs;
struct drm_crtc *crtc;
crtc_state = state->crtc_states[i];
crtc = state->crtcs[i];
if (!crtc_state || !crtc_state->mode_changed)
continue;
funcs = crtc->helper_private;
ret = funcs->mode_fixup(crtc, &crtc_state->mode,
&crtc_state->adjusted_mode);
if (!ret) {
DRM_DEBUG_KMS("[CRTC:%d] fixup failed\n",
crtc->base.id);
return -EINVAL;
}
}
return 0;
}
static int
drm/atomic: check mode_changed *after* atomic_check The intention is that drivers can set crtc_state->mode_changed in their atomic_check() fxns if they encounter a scenario that requires full modeset. Signed-off-by: Rob Clark <robdclark@gmail.com> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-19 21:41:33 +00:00
drm_atomic_helper_check_modeset(struct drm_device *dev,
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
struct drm_atomic_state *state)
{
int ncrtcs = dev->mode_config.num_crtc;
struct drm_crtc *crtc;
struct drm_crtc_state *crtc_state;
int i, ret;
for (i = 0; i < ncrtcs; i++) {
crtc = state->crtcs[i];
crtc_state = state->crtc_states[i];
if (!crtc)
continue;
if (!drm_mode_equal(&crtc->state->mode, &crtc_state->mode)) {
DRM_DEBUG_KMS("[CRTC:%d] mode changed\n",
crtc->base.id);
crtc_state->mode_changed = true;
}
if (crtc->state->enable != crtc_state->enable) {
DRM_DEBUG_KMS("[CRTC:%d] enable changed\n",
crtc->base.id);
crtc_state->mode_changed = true;
}
}
drm/atomic: Don't overrun the connector array when hotplugging Yet another fallout from not considering DP MST hotplug. With the previous patches we have stable indices, but it might still happen that a connector gets added between when we allocate the array and when we actually add a connector. Especially when we back off due to ww mutex contention or similar issues. So store the sizes of the arrays in struct drm_atomic_state and double check them. We don't really care about races except that we want to use a consistent value, so ACCESS_ONCE is all we need. And if we indeed notice that we'd overrun the array then just give up and restart the entire ioctl. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:08 +00:00
for (i = 0; i < state->num_connector; i++) {
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
/*
* This only sets crtc->mode_changed for routing changes,
* drivers must set crtc->mode_changed themselves when connector
* properties need to be updated.
*/
ret = update_connector_routing(state, i);
if (ret)
return ret;
}
/*
* After all the routing has been prepared we need to add in any
* connector which is itself unchanged, but who's crtc changes it's
* configuration. This must be done before calling mode_fixup in case a
* crtc only changed its mode but has the same set of connectors.
*/
for (i = 0; i < ncrtcs; i++) {
int num_connectors;
crtc = state->crtcs[i];
crtc_state = state->crtc_states[i];
if (!crtc || !crtc_state->mode_changed)
continue;
DRM_DEBUG_KMS("[CRTC:%d] needs full modeset, enable: %c\n",
crtc->base.id,
crtc_state->enable ? 'y' : 'n');
ret = drm_atomic_add_affected_connectors(state, crtc);
if (ret != 0)
return ret;
num_connectors = drm_atomic_connectors_for_crtc(state,
crtc);
if (crtc_state->enable != !!num_connectors) {
DRM_DEBUG_KMS("[CRTC:%d] enabled/connectors mismatch\n",
crtc->base.id);
return -EINVAL;
}
}
return mode_fixup(state);
}
drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
/**
* drm_atomic_helper_check - validate state object
* @dev: DRM device
* @state: the driver state object
*
* Check the state object to see if the requested state is physically possible.
* Only crtcs and planes have check callbacks, so for any additional (global)
* checking that a driver needs it can simply wrap that around this function.
* Drivers without such needs can directly use this as their ->atomic_check()
* callback.
*
* RETURNS
* Zero for success or -errno
*/
int drm_atomic_helper_check(struct drm_device *dev,
struct drm_atomic_state *state)
{
int nplanes = dev->mode_config.num_total_plane;
int ncrtcs = dev->mode_config.num_crtc;
int i, ret = 0;
for (i = 0; i < nplanes; i++) {
struct drm_plane_helper_funcs *funcs;
struct drm_plane *plane = state->planes[i];
struct drm_plane_state *plane_state = state->plane_states[i];
if (!plane)
continue;
funcs = plane->helper_private;
drm_atomic_helper_plane_changed(state, plane_state, plane);
if (!funcs || !funcs->atomic_check)
continue;
ret = funcs->atomic_check(plane, plane_state);
if (ret) {
DRM_DEBUG_KMS("[PLANE:%d] atomic check failed\n",
plane->base.id);
return ret;
}
}
for (i = 0; i < ncrtcs; i++) {
struct drm_crtc_helper_funcs *funcs;
struct drm_crtc *crtc = state->crtcs[i];
if (!crtc)
continue;
funcs = crtc->helper_private;
if (!funcs || !funcs->atomic_check)
continue;
ret = funcs->atomic_check(crtc, state->crtc_states[i]);
if (ret) {
DRM_DEBUG_KMS("[CRTC:%d] atomic check failed\n",
crtc->base.id);
return ret;
}
}
drm/atomic: check mode_changed *after* atomic_check The intention is that drivers can set crtc_state->mode_changed in their atomic_check() fxns if they encounter a scenario that requires full modeset. Signed-off-by: Rob Clark <robdclark@gmail.com> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-19 21:41:33 +00:00
ret = drm_atomic_helper_check_modeset(dev, state);
if (ret)
return ret;
drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
return ret;
}
EXPORT_SYMBOL(drm_atomic_helper_check);
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
static void
disable_outputs(struct drm_device *dev, struct drm_atomic_state *old_state)
{
int ncrtcs = old_state->dev->mode_config.num_crtc;
int i;
drm/atomic: Don't overrun the connector array when hotplugging Yet another fallout from not considering DP MST hotplug. With the previous patches we have stable indices, but it might still happen that a connector gets added between when we allocate the array and when we actually add a connector. Especially when we back off due to ww mutex contention or similar issues. So store the sizes of the arrays in struct drm_atomic_state and double check them. We don't really care about races except that we want to use a consistent value, so ACCESS_ONCE is all we need. And if we indeed notice that we'd overrun the array then just give up and restart the entire ioctl. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:08 +00:00
for (i = 0; i < old_state->num_connector; i++) {
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
struct drm_connector_state *old_conn_state;
struct drm_connector *connector;
struct drm_encoder_helper_funcs *funcs;
struct drm_encoder *encoder;
old_conn_state = old_state->connector_states[i];
connector = old_state->connectors[i];
/* Shut down everything that's in the changeset and currently
* still on. So need to check the old, saved state. */
if (!old_conn_state || !old_conn_state->crtc)
continue;
drm/atomic: shutdown *current* encoder In disable_outputs() we need to shut down the outgoing encoder, not the incoming one (we have already swapped-state at this point). Without this, we end up telling the driver to crtc->dpms(OFF) without first encoder->dpms(OFF), and that makes some hw quite unhappy. v2: missing WARN_ON() hunk and comment Reviewed-by: Daniel Vetter <daniel.vetter@intel.com> Signed-off-by: Rob Clark <robdclark@gmail.com>
2014-11-20 20:40:36 +00:00
encoder = old_conn_state->best_encoder;
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
drm/atomic: shutdown *current* encoder In disable_outputs() we need to shut down the outgoing encoder, not the incoming one (we have already swapped-state at this point). Without this, we end up telling the driver to crtc->dpms(OFF) without first encoder->dpms(OFF), and that makes some hw quite unhappy. v2: missing WARN_ON() hunk and comment Reviewed-by: Daniel Vetter <daniel.vetter@intel.com> Signed-off-by: Rob Clark <robdclark@gmail.com>
2014-11-20 20:40:36 +00:00
/* We shouldn't get this far if we didn't previously have
* an encoder.. but WARN_ON() rather than explode.
*/
if (WARN_ON(!encoder))
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
continue;
funcs = encoder->helper_private;
/*
* Each encoder has at most one connector (since we always steal
* it away), so we won't call call disable hooks twice.
*/
if (encoder->bridge)
encoder->bridge->funcs->disable(encoder->bridge);
/* Right function depends upon target state. */
if (connector->state->crtc)
funcs->prepare(encoder);
else if (funcs->disable)
funcs->disable(encoder);
else
funcs->dpms(encoder, DRM_MODE_DPMS_OFF);
if (encoder->bridge)
encoder->bridge->funcs->post_disable(encoder->bridge);
}
for (i = 0; i < ncrtcs; i++) {
struct drm_crtc_helper_funcs *funcs;
struct drm_crtc *crtc;
crtc = old_state->crtcs[i];
/* Shut down everything that needs a full modeset. */
if (!crtc || !crtc->state->mode_changed)
continue;
funcs = crtc->helper_private;
/* Right function depends upon target state. */
if (crtc->state->enable)
funcs->prepare(crtc);
else if (funcs->disable)
funcs->disable(crtc);
else
funcs->dpms(crtc, DRM_MODE_DPMS_OFF);
}
}
static void
set_routing_links(struct drm_device *dev, struct drm_atomic_state *old_state)
{
int ncrtcs = old_state->dev->mode_config.num_crtc;
int i;
/* clear out existing links */
drm/atomic: Don't overrun the connector array when hotplugging Yet another fallout from not considering DP MST hotplug. With the previous patches we have stable indices, but it might still happen that a connector gets added between when we allocate the array and when we actually add a connector. Especially when we back off due to ww mutex contention or similar issues. So store the sizes of the arrays in struct drm_atomic_state and double check them. We don't really care about races except that we want to use a consistent value, so ACCESS_ONCE is all we need. And if we indeed notice that we'd overrun the array then just give up and restart the entire ioctl. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:08 +00:00
for (i = 0; i < old_state->num_connector; i++) {
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
struct drm_connector *connector;
connector = old_state->connectors[i];
if (!connector || !connector->encoder)
continue;
WARN_ON(!connector->encoder->crtc);
connector->encoder->crtc = NULL;
connector->encoder = NULL;
}
/* set new links */
drm/atomic: Don't overrun the connector array when hotplugging Yet another fallout from not considering DP MST hotplug. With the previous patches we have stable indices, but it might still happen that a connector gets added between when we allocate the array and when we actually add a connector. Especially when we back off due to ww mutex contention or similar issues. So store the sizes of the arrays in struct drm_atomic_state and double check them. We don't really care about races except that we want to use a consistent value, so ACCESS_ONCE is all we need. And if we indeed notice that we'd overrun the array then just give up and restart the entire ioctl. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:08 +00:00
for (i = 0; i < old_state->num_connector; i++) {
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
struct drm_connector *connector;
connector = old_state->connectors[i];
if (!connector || !connector->state->crtc)
continue;
if (WARN_ON(!connector->state->best_encoder))
continue;
connector->encoder = connector->state->best_encoder;
connector->encoder->crtc = connector->state->crtc;
}
/* set legacy state in the crtc structure */
for (i = 0; i < ncrtcs; i++) {
struct drm_crtc *crtc;
crtc = old_state->crtcs[i];
if (!crtc)
continue;
crtc->mode = crtc->state->mode;
crtc->enabled = crtc->state->enable;
crtc->x = crtc->primary->state->src_x >> 16;
crtc->y = crtc->primary->state->src_y >> 16;
}
}
static void
crtc_set_mode(struct drm_device *dev, struct drm_atomic_state *old_state)
{
int ncrtcs = old_state->dev->mode_config.num_crtc;
int i;
for (i = 0; i < ncrtcs; i++) {
struct drm_crtc_helper_funcs *funcs;
struct drm_crtc *crtc;
crtc = old_state->crtcs[i];
if (!crtc || !crtc->state->mode_changed)
continue;
funcs = crtc->helper_private;
if (crtc->state->enable)
funcs->mode_set_nofb(crtc);
}
drm/atomic: Don't overrun the connector array when hotplugging Yet another fallout from not considering DP MST hotplug. With the previous patches we have stable indices, but it might still happen that a connector gets added between when we allocate the array and when we actually add a connector. Especially when we back off due to ww mutex contention or similar issues. So store the sizes of the arrays in struct drm_atomic_state and double check them. We don't really care about races except that we want to use a consistent value, so ACCESS_ONCE is all we need. And if we indeed notice that we'd overrun the array then just give up and restart the entire ioctl. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:08 +00:00
for (i = 0; i < old_state->num_connector; i++) {
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
struct drm_connector *connector;
struct drm_crtc_state *new_crtc_state;
struct drm_encoder_helper_funcs *funcs;
struct drm_encoder *encoder;
struct drm_display_mode *mode, *adjusted_mode;
connector = old_state->connectors[i];
if (!connector || !connector->state->best_encoder)
continue;
encoder = connector->state->best_encoder;
funcs = encoder->helper_private;
new_crtc_state = connector->state->crtc->state;
mode = &new_crtc_state->mode;
adjusted_mode = &new_crtc_state->adjusted_mode;
/*
* Each encoder has at most one connector (since we always steal
* it away), so we won't call call mode_set hooks twice.
*/
funcs->mode_set(encoder, mode, adjusted_mode);
if (encoder->bridge && encoder->bridge->funcs->mode_set)
encoder->bridge->funcs->mode_set(encoder->bridge,
mode, adjusted_mode);
}
}
/**
* drm_atomic_helper_commit_pre_planes - modeset commit before plane updates
* @dev: DRM device
* @state: atomic state
*
* This function commits the modeset changes that need to be committed before
* updating planes. It shuts down all the outputs that need to be shut down and
* prepares them (if required) with the new mode.
*/
void drm_atomic_helper_commit_pre_planes(struct drm_device *dev,
struct drm_atomic_state *state)
{
disable_outputs(dev, state);
set_routing_links(dev, state);
crtc_set_mode(dev, state);
}
EXPORT_SYMBOL(drm_atomic_helper_commit_pre_planes);
/**
* drm_atomic_helper_commit_post_planes - modeset commit after plane updates
* @dev: DRM device
* @old_state: atomic state object with old state structures
*
* This function commits the modeset changes that need to be committed after
* updating planes: It enables all the outputs with the new configuration which
* had to be turned off for the update.
*/
void drm_atomic_helper_commit_post_planes(struct drm_device *dev,
struct drm_atomic_state *old_state)
{
int ncrtcs = old_state->dev->mode_config.num_crtc;
int i;
for (i = 0; i < ncrtcs; i++) {
struct drm_crtc_helper_funcs *funcs;
struct drm_crtc *crtc;
crtc = old_state->crtcs[i];
/* Need to filter out CRTCs where only planes change. */
if (!crtc || !crtc->state->mode_changed)
continue;
funcs = crtc->helper_private;
if (crtc->state->enable)
funcs->commit(crtc);
}
drm/atomic: Don't overrun the connector array when hotplugging Yet another fallout from not considering DP MST hotplug. With the previous patches we have stable indices, but it might still happen that a connector gets added between when we allocate the array and when we actually add a connector. Especially when we back off due to ww mutex contention or similar issues. So store the sizes of the arrays in struct drm_atomic_state and double check them. We don't really care about races except that we want to use a consistent value, so ACCESS_ONCE is all we need. And if we indeed notice that we'd overrun the array then just give up and restart the entire ioctl. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:08 +00:00
for (i = 0; i < old_state->num_connector; i++) {
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
struct drm_connector *connector;
struct drm_encoder_helper_funcs *funcs;
struct drm_encoder *encoder;
connector = old_state->connectors[i];
if (!connector || !connector->state->best_encoder)
continue;
encoder = connector->state->best_encoder;
funcs = encoder->helper_private;
/*
* Each encoder has at most one connector (since we always steal
* it away), so we won't call call enable hooks twice.
*/
if (encoder->bridge)
encoder->bridge->funcs->pre_enable(encoder->bridge);
funcs->commit(encoder);
if (encoder->bridge)
encoder->bridge->funcs->enable(encoder->bridge);
}
}
EXPORT_SYMBOL(drm_atomic_helper_commit_post_planes);
drm/atomic: Integrate fence support This patch is for enabling async commits. It replaces an earlier approach which added an async boolean paramter to the ->prepare_fb callbacks. The idea is that prepare_fb picks up the right fence to synchronize against, which is then used by the synchronous commit helper. For async commits drivers can either register a callback to the fence or simply do the synchronous wait in their async work queue. v2: Remove unused variable. v3: Only wait for fences after the point of no return in the part of the commit function which can be run asynchronously. This is after the atomic state has been swapped in, hence now check plane->state->fence. Also add a WARN_ON to make sure we don't try to wait on a fence when there's no fb, just as a sanity check. Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-10-29 10:34:56 +00:00
static void wait_for_fences(struct drm_device *dev,
struct drm_atomic_state *state)
{
int nplanes = dev->mode_config.num_total_plane;
int i;
for (i = 0; i < nplanes; i++) {
struct drm_plane *plane = state->planes[i];
if (!plane || !plane->state->fence)
continue;
WARN_ON(!plane->state->fb);
fence_wait(plane->state->fence, false);
fence_put(plane->state->fence);
plane->state->fence = NULL;
}
}
drm/atomic-helper: Skip vblank waits for unchanged fbs Especially with legacy cursor ioctls existing userspace assumes that you can pile up lots of updates in one go. The super-proper way to support this would be a special commit mode which overwrites the last update. But getting there will be quite a bit of work. Meanwhile do what pretty much all the drivers have done for the plane update functions: Simply skip the vblank wait for the buffer cleanup if the buffer is the same. Since the universal cursor plane code will not recreate framebuffers needlessly this allows us to not slow down legacy pageflip events while someone moves the cursor around. v2: Drop the async plane update hunk from a previous attempt at this issue. v3: Fix up kerneldoc. v4: Don't oops so badly. Reported by Jasper. Cc: Rob Clark <robdclark@gmail.com> Cc: "Jasper St. Pierre" <jstpierre@mecheye.net> Reviewed-by: Rob Clark <robdclark@gmail.com> Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> Tested-by: Jasper St. Pierre <jstpierre@mecheye.net> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-11-24 19:42:42 +00:00
static bool framebuffer_changed(struct drm_device *dev,
struct drm_atomic_state *old_state,
struct drm_crtc *crtc)
{
struct drm_plane *plane;
struct drm_plane_state *old_plane_state;
int nplanes = old_state->dev->mode_config.num_total_plane;
int i;
for (i = 0; i < nplanes; i++) {
plane = old_state->planes[i];
old_plane_state = old_state->plane_states[i];
if (!plane)
continue;
if (plane->state->crtc != crtc &&
old_plane_state->crtc != crtc)
continue;
if (plane->state->fb != old_plane_state->fb)
return true;
}
return false;
}
drm: export atomic wait_for_vblanks helper (v2) v1: original v2: danvet's kerneldoc nitpicks Signed-off-by: Rob Clark <robdclark@gmail.com> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-12 00:38:59 +00:00
/**
* drm_atomic_helper_wait_for_vblanks - wait for vblank on crtcs
* @dev: DRM device
* @old_state: atomic state object with old state structures
*
* Helper to, after atomic commit, wait for vblanks on all effected
* crtcs (ie. before cleaning up old framebuffers using
drm/atomic-helper: Skip vblank waits for unchanged fbs Especially with legacy cursor ioctls existing userspace assumes that you can pile up lots of updates in one go. The super-proper way to support this would be a special commit mode which overwrites the last update. But getting there will be quite a bit of work. Meanwhile do what pretty much all the drivers have done for the plane update functions: Simply skip the vblank wait for the buffer cleanup if the buffer is the same. Since the universal cursor plane code will not recreate framebuffers needlessly this allows us to not slow down legacy pageflip events while someone moves the cursor around. v2: Drop the async plane update hunk from a previous attempt at this issue. v3: Fix up kerneldoc. v4: Don't oops so badly. Reported by Jasper. Cc: Rob Clark <robdclark@gmail.com> Cc: "Jasper St. Pierre" <jstpierre@mecheye.net> Reviewed-by: Rob Clark <robdclark@gmail.com> Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> Tested-by: Jasper St. Pierre <jstpierre@mecheye.net> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-11-24 19:42:42 +00:00
* drm_atomic_helper_cleanup_planes()). It will only wait on crtcs where the
* framebuffers have actually changed to optimize for the legacy cursor and
* plane update use-case.
drm: export atomic wait_for_vblanks helper (v2) v1: original v2: danvet's kerneldoc nitpicks Signed-off-by: Rob Clark <robdclark@gmail.com> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-12 00:38:59 +00:00
*/
void
drm_atomic_helper_wait_for_vblanks(struct drm_device *dev,
struct drm_atomic_state *old_state)
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
{
struct drm_crtc *crtc;
struct drm_crtc_state *old_crtc_state;
int ncrtcs = old_state->dev->mode_config.num_crtc;
int i, ret;
for (i = 0; i < ncrtcs; i++) {
crtc = old_state->crtcs[i];
old_crtc_state = old_state->crtc_states[i];
if (!crtc)
continue;
/* No one cares about the old state, so abuse it for tracking
* and store whether we hold a vblank reference (and should do a
* vblank wait) in the ->enable boolean. */
old_crtc_state->enable = false;
if (!crtc->state->enable)
continue;
drm/atomic-helper: Skip vblank waits for unchanged fbs Especially with legacy cursor ioctls existing userspace assumes that you can pile up lots of updates in one go. The super-proper way to support this would be a special commit mode which overwrites the last update. But getting there will be quite a bit of work. Meanwhile do what pretty much all the drivers have done for the plane update functions: Simply skip the vblank wait for the buffer cleanup if the buffer is the same. Since the universal cursor plane code will not recreate framebuffers needlessly this allows us to not slow down legacy pageflip events while someone moves the cursor around. v2: Drop the async plane update hunk from a previous attempt at this issue. v3: Fix up kerneldoc. v4: Don't oops so badly. Reported by Jasper. Cc: Rob Clark <robdclark@gmail.com> Cc: "Jasper St. Pierre" <jstpierre@mecheye.net> Reviewed-by: Rob Clark <robdclark@gmail.com> Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> Tested-by: Jasper St. Pierre <jstpierre@mecheye.net> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-11-24 19:42:42 +00:00
if (!framebuffer_changed(dev, old_state, crtc))
continue;
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
ret = drm_crtc_vblank_get(crtc);
if (ret != 0)
continue;
old_crtc_state->enable = true;
old_crtc_state->last_vblank_count = drm_vblank_count(dev, i);
}
for (i = 0; i < ncrtcs; i++) {
crtc = old_state->crtcs[i];
old_crtc_state = old_state->crtc_states[i];
if (!crtc || !old_crtc_state->enable)
continue;
ret = wait_event_timeout(dev->vblank[i].queue,
old_crtc_state->last_vblank_count !=
drm_vblank_count(dev, i),
msecs_to_jiffies(50));
drm_crtc_vblank_put(crtc);
}
}
drm: export atomic wait_for_vblanks helper (v2) v1: original v2: danvet's kerneldoc nitpicks Signed-off-by: Rob Clark <robdclark@gmail.com> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-12 00:38:59 +00:00
EXPORT_SYMBOL(drm_atomic_helper_wait_for_vblanks);
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
/**
* drm_atomic_helper_commit - commit validated state object
* @dev: DRM device
* @state: the driver state object
* @async: asynchronous commit
*
* This function commits a with drm_atomic_helper_check() pre-validated state
* object. This can still fail when e.g. the framebuffer reservation fails. For
* now this doesn't implement asynchronous commits.
*
* RETURNS
* Zero for success or -errno.
*/
int drm_atomic_helper_commit(struct drm_device *dev,
struct drm_atomic_state *state,
bool async)
{
int ret;
if (async)
return -EBUSY;
ret = drm_atomic_helper_prepare_planes(dev, state);
if (ret)
return ret;
/*
* This is the point of no return - everything below never fails except
* when the hw goes bonghits. Which means we can commit the new state on
* the software side now.
*/
drm_atomic_helper_swap_state(dev, state);
/*
* Everything below can be run asynchronously without the need to grab
* any modeset locks at all under one conditions: It must be guaranteed
* that the asynchronous work has either been cancelled (if the driver
* supports it, which at least requires that the framebuffers get
* cleaned up with drm_atomic_helper_cleanup_planes()) or completed
* before the new state gets committed on the software side with
* drm_atomic_helper_swap_state().
*
* This scheme allows new atomic state updates to be prepared and
* checked in parallel to the asynchronous completion of the previous
* update. Which is important since compositors need to figure out the
* composition of the next frame right after having submitted the
* current layout.
*/
drm/atomic: Integrate fence support This patch is for enabling async commits. It replaces an earlier approach which added an async boolean paramter to the ->prepare_fb callbacks. The idea is that prepare_fb picks up the right fence to synchronize against, which is then used by the synchronous commit helper. For async commits drivers can either register a callback to the fence or simply do the synchronous wait in their async work queue. v2: Remove unused variable. v3: Only wait for fences after the point of no return in the part of the commit function which can be run asynchronously. This is after the atomic state has been swapped in, hence now check plane->state->fence. Also add a WARN_ON to make sure we don't try to wait on a fence when there's no fb, just as a sanity check. Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-10-29 10:34:56 +00:00
wait_for_fences(dev, state);
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
drm_atomic_helper_commit_pre_planes(dev, state);
drm_atomic_helper_commit_planes(dev, state);
drm_atomic_helper_commit_post_planes(dev, state);
drm: export atomic wait_for_vblanks helper (v2) v1: original v2: danvet's kerneldoc nitpicks Signed-off-by: Rob Clark <robdclark@gmail.com> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-12 00:38:59 +00:00
drm_atomic_helper_wait_for_vblanks(dev, state);
drm: Atomic crtc/connector updates using crtc/plane helper interfaces So this is finally the integration of the crtc and plane helper interfaces into the atomic helper functions. In the check function we now have a few steps: - First we update the output routing and figure out which crtcs need a full mode set. Suitable encoders are selected using ->best_encoder, with the same semantics as the crtc helpers of implicitly disabling all connectors currently using the encoder. - Then we pull all other connectors into the state update which feed from a crtc which changes. This must be done do catch mode changes and similar updates - atomic updates are differences on top of the current state. - Then we call all the various ->mode_fixup to compute the adjusted mode. Note that here we have a slight semantic difference compared to the crtc helpers: We have not yet updated the encoder->crtc link when calling the encoder's ->mode_fixup function. But that's a requirement when converting to atomic since we want to prepare the entire state completely contained with the over drm_atomic_state structure. So this must be carefully checked when converting drivers over to atomic helpers. - Finally we do call the atomic_check functions on planes and crtcs. The commit function is also quite a beast: - The only step that can fail is done first, namely pinning the framebuffers. After that we cross the point of no return, an async commit would push all that into the worker thread. - The disabling of encoders and connectors is a bit tricky, since depending upon the final state we need to select different crtc helper functions. - Software tracking is a bit clarified compared to the crtc helpers: We commit the software state before starting to touch the hardware, like crtc helpers. But since we just swap them we still have the old state (i.e. the current hw state) around, which is really handy to write simple disable functions. So no more drm_crtc_helper_disable_all_unused_functions kind of fun because we're leaving unused crtcs/encoders behind. Everything gets shut down in-order now, which is one of the key differences of the i915 helpers compared to crtc helpers and a really nice additional guarantee. - Like with the plane helpers the atomic commit function waits for one vblank to pass before calling the framebuffer cleanup function. Compared to Rob's helper approach there's a bunch of upsides: - All the interfaces which can fail are called in the ->check hook (i.e. ->best_match and the various ->mode_fixup hooks). This means that drivers can just reuse those functions and don't need to move everything into ->atomic_check callbacks. If drivers have no need for additional constraint checking beyong their existing crtc helper callbacks they don't need to do anything. - The actual commit operation is properly stage: First we prepare framebuffers, which can potentially still fail (due to memory exhausting). This is important for the async case, where this must be done synchronously to correctly return errors. - The output configuration changes (done with crtc helper functions) and the plane update (using atomic plane helpers) are correctly interleaved: First we shut down any crtcs that need changing, then we update planes and finally we enable everything again. Hardware without GO bits must be more careful with ordering, which this sequence enables. - Also for hardware with shared output resources (like display PLLs) we first must shut down the old configuration before we can enable the new one. Otherwise we can hit an impossible intermediate state where there's not enough PLLs (which is the point behind atomic updates). v2: - Ensure that users of ->check update crtc_state->enable correctly. - Update the legacy state in crtc/plane structures. Eventually we want to remove that, but for now the drm core still expects this (especially the plane->fb pointer). v3: A few changes for better async handling: - Reorder the software side state commit so that it happens all before we touch the hardware. This way async support becomes very easy since we can punt all the actual hw touching to a worker thread. And as long as we synchronize with that thread (flushing or cancelling, depending upon what the driver can handle) before we commit the next software state there's no need for any locking in the worker thread at all. Which greatly simplifies things. And as long as we synchronize with all relevant threads we can have a lot of them (e.g. per-crtc for per-crtc updates) running in parallel. - Expose pre/post plane commit steps separately. We need to expose the actual hw commit step anyway for drivers to be able to implement asynchronous commit workers. But if we expose pre/post and plane commit steps individually we allow drivers to selectively use atomic helpers. - I've forgotten to call encoder/bridge ->mode_set functions, fix this. v4: Add debug output and fix a mixup between current and new state that resulted in crtcs not getting updated correctly. And in an Oops ... v5: - Be kind to driver writers in the vblank wait functions.. if thing aren't working yet, and vblank irq will never come, then let's not block forever.. especially under console-lock. - Correctly clear connector_state->best_encoder when disabling. Spotted while trying to understand a report from Rob Clark. - Only steal encoder if it actually changed, otherwise hilarity ensues if we steal from the current connector and so set the ->crtc pointer unexpectedly to NULL. Reported by Rob Clark. - Bail out in disable_outputs if an output currently doesn't have a best_encoder - this means it's already disabled. v6: Fixupe kerneldoc as reported by Paulo. And also fix up kerneldoc in drm_crtc.h. v7: Take ownership of the atomic state and clean it up with drm_atomic_state_free(). v8 Various improvements all over: - Polish code comments and kerneldoc. - Improve debug output to make sure all failure cases are logged. - Treat enabled crtc with no connectors as invalid input from userspace. - Don't ignore the return value from mode_fixup(). v9: - Improve debug output for crtc_state->mode_changed. v10: - Fixup the vblank waiting code to properly balance the vblank_get/put calls. - Better comments when checking/computing crtc->mode_changed v11: Fixup the encoder stealing logic: We can't look at encoder->crtc since that's not in the atomic state structures and might be updated asynchronously in and async commit. Instead we need to inspect all the connector states and check whether the encoder is currently in used and if so, on which crtc. v12: Review from Sean: - A few spelling fixes. - Flatten control flow indent by converting if blocks to early continue/return in 2 places. - Capture connectors_for_crtc return value in int num_connectors instead of bool has_connectors and do an explicit int->bool conversion with !!. I think the helper is more useful for drivers if it returns the number of connectors (e.g. to detect cloning configurations), so decided to keep that return value. Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-09-16 15:50:47 +00:00
drm_atomic_helper_cleanup_planes(dev, state);
drm_atomic_state_free(state);
return 0;
}
EXPORT_SYMBOL(drm_atomic_helper_commit);
drm/atomic-helpers: document how to implement async commit No helper function to do it all yet provided since no driver has support for driver core fences yet. Which we'd need to make the implementation really generic. v2: Clarify async howto a bit per the discussion With Rob Clark. Cc: Rob Clark <robdclark@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 16:30:19 +00:00
/**
* DOC: implementing async commit
*
* For now the atomic helpers don't support async commit directly. If there is
* real need it could be added though, using the dma-buf fence infrastructure
* for generic synchronization with outstanding rendering.
*
* For now drivers have to implement async commit themselves, with the following
* sequence being the recommended one:
*
* 1. Run drm_atomic_helper_prepare_planes() first. This is the only function
* which commit needs to call which can fail, so we want to run it first and
* synchronously.
*
* 2. Synchronize with any outstanding asynchronous commit worker threads which
* might be affected the new state update. This can be done by either cancelling
* or flushing the work items, depending upon whether the driver can deal with
* cancelled updates. Note that it is important to ensure that the framebuffer
* cleanup is still done when cancelling.
*
* For sufficient parallelism it is recommended to have a work item per crtc
* (for updates which don't touch global state) and a global one. Then we only
* need to synchronize with the crtc work items for changed crtcs and the global
* work item, which allows nice concurrent updates on disjoint sets of crtcs.
*
* 3. The software state is updated synchronously with
* drm_atomic_helper_swap_state. Doing this under the protection of all modeset
* locks means concurrent callers never see inconsistent state. And doing this
* while it's guaranteed that no relevant async worker runs means that async
* workers do not need grab any locks. Actually they must not grab locks, for
* otherwise the work flushing will deadlock.
*
* 4. Schedule a work item to do all subsequent steps, using the split-out
* commit helpers: a) pre-plane commit b) plane commit c) post-plane commit and
* then cleaning up the framebuffers after the old framebuffer is no longer
* being displayed.
*/
drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
/**
* drm_atomic_helper_prepare_planes - prepare plane resources after commit
* @dev: DRM device
* @state: atomic state object with old state structures
*
* This function prepares plane state, specifically framebuffers, for the new
* configuration. If any failure is encountered this function will call
* ->cleanup_fb on any already successfully prepared framebuffer.
*
* Returns:
* 0 on success, negative error code on failure.
*/
int drm_atomic_helper_prepare_planes(struct drm_device *dev,
struct drm_atomic_state *state)
{
int nplanes = dev->mode_config.num_total_plane;
int ret, i;
for (i = 0; i < nplanes; i++) {
struct drm_plane_helper_funcs *funcs;
struct drm_plane *plane = state->planes[i];
struct drm_framebuffer *fb;
if (!plane)
continue;
funcs = plane->helper_private;
fb = state->plane_states[i]->fb;
if (fb && funcs->prepare_fb) {
ret = funcs->prepare_fb(plane, fb);
if (ret)
goto fail;
}
}
return 0;
fail:
for (i--; i >= 0; i--) {
struct drm_plane_helper_funcs *funcs;
struct drm_plane *plane = state->planes[i];
struct drm_framebuffer *fb;
if (!plane)
continue;
funcs = plane->helper_private;
fb = state->plane_states[i]->fb;
if (fb && funcs->cleanup_fb)
funcs->cleanup_fb(plane, fb);
}
return ret;
}
EXPORT_SYMBOL(drm_atomic_helper_prepare_planes);
/**
* drm_atomic_helper_commit_planes - commit plane state
* @dev: DRM device
drm/atomic_helper: Make it clear that commit_planes gets the old state Oversight from my kerneldoc cleanup when doing the original atomic helper series - I've only applied this clarification to the modeset related helpers, and not the plane update code. Remedy this asap. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:11 +00:00
* @old_state: atomic state object with old state structures
drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
*
* This function commits the new plane state using the plane and atomic helper
* functions for planes and crtcs. It assumes that the atomic state has already
* been pushed into the relevant object state pointers, since this step can no
* longer fail.
*
drm/atomic_helper: Make it clear that commit_planes gets the old state Oversight from my kerneldoc cleanup when doing the original atomic helper series - I've only applied this clarification to the modeset related helpers, and not the plane update code. Remedy this asap. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:11 +00:00
* It still requires the global state object @old_state to know which planes and
drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
* crtcs need to be updated though.
*/
void drm_atomic_helper_commit_planes(struct drm_device *dev,
drm/atomic_helper: Make it clear that commit_planes gets the old state Oversight from my kerneldoc cleanup when doing the original atomic helper series - I've only applied this clarification to the modeset related helpers, and not the plane update code. Remedy this asap. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:11 +00:00
struct drm_atomic_state *old_state)
drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
{
int nplanes = dev->mode_config.num_total_plane;
int ncrtcs = dev->mode_config.num_crtc;
int i;
for (i = 0; i < ncrtcs; i++) {
struct drm_crtc_helper_funcs *funcs;
drm/atomic_helper: Make it clear that commit_planes gets the old state Oversight from my kerneldoc cleanup when doing the original atomic helper series - I've only applied this clarification to the modeset related helpers, and not the plane update code. Remedy this asap. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:11 +00:00
struct drm_crtc *crtc = old_state->crtcs[i];
drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
if (!crtc)
continue;
funcs = crtc->helper_private;
if (!funcs || !funcs->atomic_begin)
continue;
funcs->atomic_begin(crtc);
}
for (i = 0; i < nplanes; i++) {
struct drm_plane_helper_funcs *funcs;
drm/atomic_helper: Make it clear that commit_planes gets the old state Oversight from my kerneldoc cleanup when doing the original atomic helper series - I've only applied this clarification to the modeset related helpers, and not the plane update code. Remedy this asap. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:11 +00:00
struct drm_plane *plane = old_state->planes[i];
drm/plane: Pass old state to ->atomic_update() In most situations it will be useful to have the old state passed to the ->atomic_update() callback. For example if a plane is being disabled the new state's .crtc field will be NULL, but some drivers may rely on this field to program the CRTCs registers. v2: rename variable to old_plane_state and remove redundant comment as suggested by Daniel Vetter, remove an Exynos hunk that doesn't apply to drm-next and add a hunk for pending MSM mdp5 changes Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Thierry Reding <treding@nvidia.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-25 11:09:44 +00:00
struct drm_plane_state *old_plane_state;
drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
if (!plane)
continue;
funcs = plane->helper_private;
if (!funcs || !funcs->atomic_update)
continue;
drm/plane: Pass old state to ->atomic_update() In most situations it will be useful to have the old state passed to the ->atomic_update() callback. For example if a plane is being disabled the new state's .crtc field will be NULL, but some drivers may rely on this field to program the CRTCs registers. v2: rename variable to old_plane_state and remove redundant comment as suggested by Daniel Vetter, remove an Exynos hunk that doesn't apply to drm-next and add a hunk for pending MSM mdp5 changes Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Thierry Reding <treding@nvidia.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-25 11:09:44 +00:00
old_plane_state = old_state->plane_states[i];
funcs->atomic_update(plane, old_plane_state);
drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
}
for (i = 0; i < ncrtcs; i++) {
struct drm_crtc_helper_funcs *funcs;
drm/atomic_helper: Make it clear that commit_planes gets the old state Oversight from my kerneldoc cleanup when doing the original atomic helper series - I've only applied this clarification to the modeset related helpers, and not the plane update code. Remedy this asap. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:11 +00:00
struct drm_crtc *crtc = old_state->crtcs[i];
drm: Add atomic/plane helpers This is the first cut of atomic helper code. As-is it's only useful to implement a pure atomic interface for plane updates. Later patches will integrate this with the crtc helpers so that full atomic updates are possible. We also need a pile of helpers to aid drivers in transitioning from the legacy world to the shiny new atomic age. Finally we need helpers to implement legacy ioctls on top of the atomic interface. The design of the overall helpers<->driver interaction is fairly simple, but has an unfortunate large interface: - We have ->atomic_check callbacks for crtcs and planes. The idea is that connectors don't need any checking, and if they do they can adjust the relevant crtc driver-private state. So no connector hooks should be needed. Also the crtc helpers integration will do the ->best_encoder checks, so no need for that. - Framebuffer pinning needs to be done before we can commit to the hw state. This is especially important for async updates where we must pin all buffers before returning to userspace, so that really only hw failures can happen in the asynchronous worker. Hence we add ->prepare_fb and ->cleanup_fb hooks for this resources management. - The actual atomic plane commit can't fail (except hw woes), so has void return type. It has three stages: 1. Prepare all affected crtcs with crtc->atomic_begin. Drivers can use this to unset the GO bit or similar latches to prevent plane updates. 2. Update plane state by looping over all changed planes and calling plane->atomic_update. Presuming the hardware is sane and has GO bits drivers can simply bash the state into the hardware in this function. Other drivers might use this to precompute hw state for the final step. 3. Finally latch the update for the next vblank with crtc->atomic_flush. Note that this function doesn't need to wait for the vblank to happen even for the synchronous case. v2: Clear drm_<obj>_state->state to NULL when swapping in state. v3: Add TODO that we don't short-circuit plane updates for now. Likely no one will care. v4: Squash in a bit of polish that somehow landed in the wrong (later) patche. v5: Integrate atomic functions into the drm docbook and fixup the kerneldoc. v6: Fixup fixup patch squashing fumble. v7: Don't touch the legacy plane state plane->fb and plane->crtc. This is only used by the legacy ioctl code in the drm core, and that code already takes care of updating the pointers in all relevant cases. This is in stark contrast to connector->encoder->crtc links on the modeset side, which we still need to set since the core doesn't touch them. Also some more kerneldoc polish. v8: Drop outdated comment. v9: Handle the state->state pointer correctly: Only clearing the ->state pointer when assigning the state to the kms object isn't good enough. We also need to re-link the swapped out state into the drm_atomic_state structure. v10: Shuffle the misplaced docbook template hunk around that Sean spotted. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-04 23:14:14 +00:00
if (!crtc)
continue;
funcs = crtc->helper_private;
if (!funcs || !funcs->atomic_flush)
continue;
funcs->atomic_flush(crtc);
}
}
EXPORT_SYMBOL(drm_atomic_helper_commit_planes);
/**
* drm_atomic_helper_cleanup_planes - cleanup plane resources after commit
* @dev: DRM device
* @old_state: atomic state object with old state structures
*
* This function cleans up plane state, specifically framebuffers, from the old
* configuration. Hence the old configuration must be perserved in @old_state to
* be able to call this function.
*
* This function must also be called on the new state when the atomic update
* fails at any point after calling drm_atomic_helper_prepare_planes().
*/
void drm_atomic_helper_cleanup_planes(struct drm_device *dev,
struct drm_atomic_state *old_state)
{
int nplanes = dev->mode_config.num_total_plane;
int i;
for (i = 0; i < nplanes; i++) {
struct drm_plane_helper_funcs *funcs;
struct drm_plane *plane = old_state->planes[i];
struct drm_framebuffer *old_fb;
if (!plane)
continue;
funcs = plane->helper_private;
old_fb = old_state->plane_states[i]->fb;
if (old_fb && funcs->cleanup_fb)
funcs->cleanup_fb(plane, old_fb);
}
}
EXPORT_SYMBOL(drm_atomic_helper_cleanup_planes);
/**
* drm_atomic_helper_swap_state - store atomic state into current sw state
* @dev: DRM device
* @state: atomic state
*
* This function stores the atomic state into the current state pointers in all
* driver objects. It should be called after all failing steps have been done
* and succeeded, but before the actual hardware state is committed.
*
* For cleanup and error recovery the current state for all changed objects will
* be swaped into @state.
*
* With that sequence it fits perfectly into the plane prepare/cleanup sequence:
*
* 1. Call drm_atomic_helper_prepare_planes() with the staged atomic state.
*
* 2. Do any other steps that might fail.
*
* 3. Put the staged state into the current state pointers with this function.
*
* 4. Actually commit the hardware state.
*
* 5. Call drm_atomic_helper_cleanup_planes with @state, which since step 3
* contains the old state. Also do any other cleanup required with that state.
*/
void drm_atomic_helper_swap_state(struct drm_device *dev,
struct drm_atomic_state *state)
{
int i;
for (i = 0; i < dev->mode_config.num_connector; i++) {
struct drm_connector *connector = state->connectors[i];
if (!connector)
continue;
connector->state->state = state;
swap(state->connector_states[i], connector->state);
connector->state->state = NULL;
}
for (i = 0; i < dev->mode_config.num_crtc; i++) {
struct drm_crtc *crtc = state->crtcs[i];
if (!crtc)
continue;
crtc->state->state = state;
swap(state->crtc_states[i], crtc->state);
crtc->state->state = NULL;
}
for (i = 0; i < dev->mode_config.num_total_plane; i++) {
struct drm_plane *plane = state->planes[i];
if (!plane)
continue;
plane->state->state = state;
swap(state->plane_states[i], plane->state);
plane->state->state = NULL;
}
}
EXPORT_SYMBOL(drm_atomic_helper_swap_state);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
/**
* drm_atomic_helper_update_plane - Helper for primary plane update using atomic
* @plane: plane object to update
* @crtc: owning CRTC of owning plane
* @fb: framebuffer to flip onto plane
* @crtc_x: x offset of primary plane on crtc
* @crtc_y: y offset of primary plane on crtc
* @crtc_w: width of primary plane rectangle on crtc
* @crtc_h: height of primary plane rectangle on crtc
* @src_x: x offset of @fb for panning
* @src_y: y offset of @fb for panning
* @src_w: width of source rectangle in @fb
* @src_h: height of source rectangle in @fb
*
* Provides a default plane update handler using the atomic driver interface.
*
* RETURNS:
* Zero on success, error code on failure
*/
int drm_atomic_helper_update_plane(struct drm_plane *plane,
struct drm_crtc *crtc,
struct drm_framebuffer *fb,
int crtc_x, int crtc_y,
unsigned int crtc_w, unsigned int crtc_h,
uint32_t src_x, uint32_t src_y,
uint32_t src_w, uint32_t src_h)
{
struct drm_atomic_state *state;
struct drm_plane_state *plane_state;
int ret = 0;
state = drm_atomic_state_alloc(plane->dev);
if (!state)
return -ENOMEM;
state->acquire_ctx = drm_modeset_legacy_acquire_ctx(crtc);
retry:
plane_state = drm_atomic_get_plane_state(state, plane);
if (IS_ERR(plane_state)) {
ret = PTR_ERR(plane_state);
goto fail;
}
drm/atomic: track bitmask of planes attached to crtc Chasing plane->state->crtc of planes that are *not* part of the same atomic update is racy, making it incredibly awkward (or impossible) to do something simple like iterate over all planes and figure out which ones are attached to a crtc. Solve this by adding a bitmask of currently attached planes in the crtc-state. Note that the transitional helpers do not maintain the plane_mask. But they only support the legacy ioctls, which have sufficient brute-force locking around plane updates that they can continue to loop over all planes to see what is attached to a crtc the old way. Signed-off-by: Rob Clark <robdclark@gmail.com> [danvet: - Drop comments about locking in set_crtc_for_plane since they're a bit misleading - we already should hold lock for the current crtc. - Also WARN_ON if get_state on the old crtc fails since that should have been done already. - Squash in fixup to check get_plane_state return value, reported by Dan Carpenter and acked by Rob Clark.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-21 20:28:31 +00:00
ret = drm_atomic_set_crtc_for_plane(state, plane, crtc);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
if (ret != 0)
goto fail;
drm/atomic: Refcounting for plane_state->fb So my original plan was that the drm core refcounts framebuffers like with the legacy ioctls. But that doesn't work for a bunch of reasons: - State objects might live longer than until the next fb change happens for a plane. For example delayed cleanup work only happens _after_ the pageflip ioctl has completed. So this definitely doesn't work without the plane state holding its own references. - The other issue is transition from legacy to atomic implementations, where the driver works under a mix of both worlds. Which means legacy paths might not properly update the ->fb pointer under plane->state->fb. Which is a bit a problem when then someone comes around and _does_ try to clean it up when it's long gone. The second issue is just a bit a transition bug, since drivers should update plane->state->fb in all the paths that aren't converted yet. But a bit more robustness for the transition can't hurt - we pull similar tricks with cleaning up the old fb in the transitional helpers already. The pattern for drivers that transition is if (plane->state) drm_atomic_set_fb_for_plane(plane->state, plane->fb); inserted after the fb update has logically completed at the end of ->set_config (or ->set_base/mode_set if using the crtc helpers), ->page_flip, ->update_plane or any other entry point which updates plane->fb. v2: Update kerneldoc - copypasta fail. v3: Fix spelling in the commit message (Sean). Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-11-04 21:57:27 +00:00
drm_atomic_set_fb_for_plane(plane_state, fb);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
plane_state->crtc_x = crtc_x;
plane_state->crtc_y = crtc_y;
plane_state->crtc_h = crtc_h;
plane_state->crtc_w = crtc_w;
plane_state->src_x = src_x;
plane_state->src_y = src_y;
plane_state->src_h = src_h;
plane_state->src_w = src_w;
ret = drm_atomic_commit(state);
if (ret != 0)
goto fail;
/* Driver takes ownership of state on successful commit. */
return 0;
fail:
if (ret == -EDEADLK)
goto backoff;
drm_atomic_state_free(state);
return ret;
backoff:
drm_atomic_state_clear(state);
drm/atomic: Only destroy connector states with connection mutex held Otherwise the connector might have been unplugged and destroyed while we didn't look. Yet another fallout from DP MST hotplugging that I didn't consider. To make sure we get this right add an appropriate WARN_ON to drm_atomic_state_clear (obviously only when we actually have a state to clear up). And reorder all the state_clear and backoff calls to make it work out properly. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:07 +00:00
drm_atomic_legacy_backoff(state);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
/*
* Someone might have exchanged the framebuffer while we dropped locks
* in the backoff code. We need to fix up the fb refcount tracking the
* core does for us.
*/
plane->old_fb = plane->fb;
goto retry;
}
EXPORT_SYMBOL(drm_atomic_helper_update_plane);
/**
* drm_atomic_helper_disable_plane - Helper for primary plane disable using * atomic
* @plane: plane to disable
*
* Provides a default plane disable handler using the atomic driver interface.
*
* RETURNS:
* Zero on success, error code on failure
*/
int drm_atomic_helper_disable_plane(struct drm_plane *plane)
{
struct drm_atomic_state *state;
struct drm_plane_state *plane_state;
int ret = 0;
drm/atomic_helper: Cope with plane->crtc == NULL in disable helper The drm core can call the plane disable hook multiple times, which means it can get called when plane->crtc is already NULL. That in turn means we can't get at the implicit acquire ctx we use in the atomic helpers for legacy entries points. We could try to pass drm_modeset_legacy_acquire_ctx a drm_device pointer so that it can cope with a NULL crtc. But that still doesn't work since the cursor ioctls (remapped with the universal cursor plane support code) only grabs the crtc locks. So the global acquire context isn't set eitehr. The real solution here would be to bite the bullet and wire up explicit acquire context parameters to all relevant functions. We need to do that anyway (to be able to get rid of some small allocations which we can't cope with failing). But that's a lot of work and better done once atomic has settled a bit. So meanwhile just catch this case in the helper and bail out. Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> Reviewed-by: Rob Clark <robdclark@gmail.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> [danvet: Completely rewrite commit message and comment but keep Jasper's logic and author credits since his patch is the only short-term solution that works.] Tested-by: Thierry Reding <treding@nvidia.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-21 03:59:15 +00:00
/*
* FIXME: Without plane->crtc set we can't get at the implicit legacy
* acquire context. The real fix will be to wire the acquire ctx through
* everywhere we need it, but meanwhile prevent chaos by just skipping
* this noop. The critical case is the cursor ioctls which a) only grab
* crtc/cursor-plane locks (so we need the crtc to get at the right
* acquire context) and b) can try to disable the plane multiple times.
*/
if (!plane->crtc)
return 0;
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
state = drm_atomic_state_alloc(plane->dev);
if (!state)
return -ENOMEM;
state->acquire_ctx = drm_modeset_legacy_acquire_ctx(plane->crtc);
retry:
plane_state = drm_atomic_get_plane_state(state, plane);
if (IS_ERR(plane_state)) {
ret = PTR_ERR(plane_state);
goto fail;
}
drm/atomic: track bitmask of planes attached to crtc Chasing plane->state->crtc of planes that are *not* part of the same atomic update is racy, making it incredibly awkward (or impossible) to do something simple like iterate over all planes and figure out which ones are attached to a crtc. Solve this by adding a bitmask of currently attached planes in the crtc-state. Note that the transitional helpers do not maintain the plane_mask. But they only support the legacy ioctls, which have sufficient brute-force locking around plane updates that they can continue to loop over all planes to see what is attached to a crtc the old way. Signed-off-by: Rob Clark <robdclark@gmail.com> [danvet: - Drop comments about locking in set_crtc_for_plane since they're a bit misleading - we already should hold lock for the current crtc. - Also WARN_ON if get_state on the old crtc fails since that should have been done already. - Squash in fixup to check get_plane_state return value, reported by Dan Carpenter and acked by Rob Clark.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-21 20:28:31 +00:00
ret = drm_atomic_set_crtc_for_plane(state, plane, NULL);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
if (ret != 0)
goto fail;
drm/atomic: Refcounting for plane_state->fb So my original plan was that the drm core refcounts framebuffers like with the legacy ioctls. But that doesn't work for a bunch of reasons: - State objects might live longer than until the next fb change happens for a plane. For example delayed cleanup work only happens _after_ the pageflip ioctl has completed. So this definitely doesn't work without the plane state holding its own references. - The other issue is transition from legacy to atomic implementations, where the driver works under a mix of both worlds. Which means legacy paths might not properly update the ->fb pointer under plane->state->fb. Which is a bit a problem when then someone comes around and _does_ try to clean it up when it's long gone. The second issue is just a bit a transition bug, since drivers should update plane->state->fb in all the paths that aren't converted yet. But a bit more robustness for the transition can't hurt - we pull similar tricks with cleaning up the old fb in the transitional helpers already. The pattern for drivers that transition is if (plane->state) drm_atomic_set_fb_for_plane(plane->state, plane->fb); inserted after the fb update has logically completed at the end of ->set_config (or ->set_base/mode_set if using the crtc helpers), ->page_flip, ->update_plane or any other entry point which updates plane->fb. v2: Update kerneldoc - copypasta fail. v3: Fix spelling in the commit message (Sean). Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-11-04 21:57:27 +00:00
drm_atomic_set_fb_for_plane(plane_state, NULL);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
plane_state->crtc_x = 0;
plane_state->crtc_y = 0;
plane_state->crtc_h = 0;
plane_state->crtc_w = 0;
plane_state->src_x = 0;
plane_state->src_y = 0;
plane_state->src_h = 0;
plane_state->src_w = 0;
ret = drm_atomic_commit(state);
if (ret != 0)
goto fail;
/* Driver takes ownership of state on successful commit. */
return 0;
fail:
if (ret == -EDEADLK)
goto backoff;
drm_atomic_state_free(state);
return ret;
backoff:
drm_atomic_state_clear(state);
drm/atomic: Only destroy connector states with connection mutex held Otherwise the connector might have been unplugged and destroyed while we didn't look. Yet another fallout from DP MST hotplugging that I didn't consider. To make sure we get this right add an appropriate WARN_ON to drm_atomic_state_clear (obviously only when we actually have a state to clear up). And reorder all the state_clear and backoff calls to make it work out properly. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:07 +00:00
drm_atomic_legacy_backoff(state);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
/*
* Someone might have exchanged the framebuffer while we dropped locks
* in the backoff code. We need to fix up the fb refcount tracking the
* core does for us.
*/
plane->old_fb = plane->fb;
goto retry;
}
EXPORT_SYMBOL(drm_atomic_helper_disable_plane);
static int update_output_state(struct drm_atomic_state *state,
struct drm_mode_set *set)
{
struct drm_device *dev = set->crtc->dev;
struct drm_connector_state *conn_state;
int ncrtcs = state->dev->mode_config.num_crtc;
int ret, i, j;
ret = drm_modeset_lock(&dev->mode_config.connection_mutex,
state->acquire_ctx);
if (ret)
return ret;
/* First grab all affected connector/crtc states. */
for (i = 0; i < set->num_connectors; i++) {
conn_state = drm_atomic_get_connector_state(state,
set->connectors[i]);
if (IS_ERR(conn_state))
return PTR_ERR(conn_state);
}
for (i = 0; i < ncrtcs; i++) {
struct drm_crtc *crtc = state->crtcs[i];
if (!crtc)
continue;
ret = drm_atomic_add_affected_connectors(state, crtc);
if (ret)
return ret;
}
/* Then recompute connector->crtc links and crtc enabling state. */
drm/atomic: Don't overrun the connector array when hotplugging Yet another fallout from not considering DP MST hotplug. With the previous patches we have stable indices, but it might still happen that a connector gets added between when we allocate the array and when we actually add a connector. Especially when we back off due to ww mutex contention or similar issues. So store the sizes of the arrays in struct drm_atomic_state and double check them. We don't really care about races except that we want to use a consistent value, so ACCESS_ONCE is all we need. And if we indeed notice that we'd overrun the array then just give up and restart the entire ioctl. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:08 +00:00
for (i = 0; i < state->num_connector; i++) {
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
struct drm_connector *connector;
connector = state->connectors[i];
conn_state = state->connector_states[i];
if (!connector)
continue;
if (conn_state->crtc == set->crtc) {
ret = drm_atomic_set_crtc_for_connector(conn_state,
NULL);
if (ret)
return ret;
}
for (j = 0; j < set->num_connectors; j++) {
if (set->connectors[j] == connector) {
ret = drm_atomic_set_crtc_for_connector(conn_state,
set->crtc);
if (ret)
return ret;
break;
}
}
}
for (i = 0; i < ncrtcs; i++) {
struct drm_crtc *crtc = state->crtcs[i];
struct drm_crtc_state *crtc_state = state->crtc_states[i];
if (!crtc)
continue;
/* Don't update ->enable for the CRTC in the set_config request,
* since a mismatch would indicate a bug in the upper layers.
* The actual modeset code later on will catch any
* inconsistencies here. */
if (crtc == set->crtc)
continue;
crtc_state->enable =
drm_atomic_connectors_for_crtc(state, crtc);
}
return 0;
}
/**
* drm_atomic_helper_set_config - set a new config from userspace
* @set: mode set configuration
*
* Provides a default crtc set_config handler using the atomic driver interface.
*
* Returns:
* Returns 0 on success, negative errno numbers on failure.
*/
int drm_atomic_helper_set_config(struct drm_mode_set *set)
{
struct drm_atomic_state *state;
struct drm_crtc *crtc = set->crtc;
struct drm_crtc_state *crtc_state;
struct drm_plane_state *primary_state;
int ret = 0;
state = drm_atomic_state_alloc(crtc->dev);
if (!state)
return -ENOMEM;
state->acquire_ctx = drm_modeset_legacy_acquire_ctx(crtc);
retry:
crtc_state = drm_atomic_get_crtc_state(state, crtc);
if (IS_ERR(crtc_state)) {
ret = PTR_ERR(crtc_state);
goto fail;
}
drm/atomic: clear plane's CRTC and FB when shutting down Otherwise we'd still end up w/ the plane attached to the CRTC, and seemingly active, but without an FB. Which ends up going *boom* in the drivers. Slightly modified version of Daniel's irc suggestion. Note that the big problem isn't drivers going *boom* here (since we already have the situation of planes being left enabled when the crtc goes down). The real issue is that the core assumes the primary plane always goes down when calling ->set_config with a NULL mode. Ignoring that assumption leads to the legacy state pointers plane->fb/crtc getting out of sync with atomic, and that then leads to the subsequent *boom* all over the place. CC: Daniel Vetter <daniel@ffwll.ch> Signed-off-by: Rob Clark <robdclark@gmail.com> [danvet: Drop my opinion of what's going sidewides here into the commit message as a note.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-26 23:58:04 +00:00
primary_state = drm_atomic_get_plane_state(state, crtc->primary);
if (IS_ERR(primary_state)) {
ret = PTR_ERR(primary_state);
goto fail;
}
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
if (!set->mode) {
WARN_ON(set->fb);
WARN_ON(set->num_connectors);
crtc_state->enable = false;
drm/atomic: clear plane's CRTC and FB when shutting down Otherwise we'd still end up w/ the plane attached to the CRTC, and seemingly active, but without an FB. Which ends up going *boom* in the drivers. Slightly modified version of Daniel's irc suggestion. Note that the big problem isn't drivers going *boom* here (since we already have the situation of planes being left enabled when the crtc goes down). The real issue is that the core assumes the primary plane always goes down when calling ->set_config with a NULL mode. Ignoring that assumption leads to the legacy state pointers plane->fb/crtc getting out of sync with atomic, and that then leads to the subsequent *boom* all over the place. CC: Daniel Vetter <daniel@ffwll.ch> Signed-off-by: Rob Clark <robdclark@gmail.com> [danvet: Drop my opinion of what's going sidewides here into the commit message as a note.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-26 23:58:04 +00:00
ret = drm_atomic_set_crtc_for_plane(state, crtc->primary, NULL);
if (ret != 0)
goto fail;
drm_atomic_set_fb_for_plane(primary_state, NULL);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
goto commit;
}
WARN_ON(!set->fb);
WARN_ON(!set->num_connectors);
crtc_state->enable = true;
drm_mode_copy(&crtc_state->mode, set->mode);
drm/atomic: track bitmask of planes attached to crtc Chasing plane->state->crtc of planes that are *not* part of the same atomic update is racy, making it incredibly awkward (or impossible) to do something simple like iterate over all planes and figure out which ones are attached to a crtc. Solve this by adding a bitmask of currently attached planes in the crtc-state. Note that the transitional helpers do not maintain the plane_mask. But they only support the legacy ioctls, which have sufficient brute-force locking around plane updates that they can continue to loop over all planes to see what is attached to a crtc the old way. Signed-off-by: Rob Clark <robdclark@gmail.com> [danvet: - Drop comments about locking in set_crtc_for_plane since they're a bit misleading - we already should hold lock for the current crtc. - Also WARN_ON if get_state on the old crtc fails since that should have been done already. - Squash in fixup to check get_plane_state return value, reported by Dan Carpenter and acked by Rob Clark.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-21 20:28:31 +00:00
ret = drm_atomic_set_crtc_for_plane(state, crtc->primary, crtc);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
if (ret != 0)
goto fail;
drm/atomic: Refcounting for plane_state->fb So my original plan was that the drm core refcounts framebuffers like with the legacy ioctls. But that doesn't work for a bunch of reasons: - State objects might live longer than until the next fb change happens for a plane. For example delayed cleanup work only happens _after_ the pageflip ioctl has completed. So this definitely doesn't work without the plane state holding its own references. - The other issue is transition from legacy to atomic implementations, where the driver works under a mix of both worlds. Which means legacy paths might not properly update the ->fb pointer under plane->state->fb. Which is a bit a problem when then someone comes around and _does_ try to clean it up when it's long gone. The second issue is just a bit a transition bug, since drivers should update plane->state->fb in all the paths that aren't converted yet. But a bit more robustness for the transition can't hurt - we pull similar tricks with cleaning up the old fb in the transitional helpers already. The pattern for drivers that transition is if (plane->state) drm_atomic_set_fb_for_plane(plane->state, plane->fb); inserted after the fb update has logically completed at the end of ->set_config (or ->set_base/mode_set if using the crtc helpers), ->page_flip, ->update_plane or any other entry point which updates plane->fb. v2: Update kerneldoc - copypasta fail. v3: Fix spelling in the commit message (Sean). Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-11-04 21:57:27 +00:00
drm_atomic_set_fb_for_plane(primary_state, set->fb);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
primary_state->crtc_x = 0;
primary_state->crtc_y = 0;
primary_state->crtc_h = set->mode->vdisplay;
primary_state->crtc_w = set->mode->hdisplay;
primary_state->src_x = set->x << 16;
primary_state->src_y = set->y << 16;
primary_state->src_h = set->mode->vdisplay << 16;
primary_state->src_w = set->mode->hdisplay << 16;
commit:
ret = update_output_state(state, set);
if (ret)
goto fail;
ret = drm_atomic_commit(state);
if (ret != 0)
goto fail;
/* Driver takes ownership of state on successful commit. */
return 0;
fail:
if (ret == -EDEADLK)
goto backoff;
drm_atomic_state_free(state);
return ret;
backoff:
drm_atomic_state_clear(state);
drm/atomic: Only destroy connector states with connection mutex held Otherwise the connector might have been unplugged and destroyed while we didn't look. Yet another fallout from DP MST hotplugging that I didn't consider. To make sure we get this right add an appropriate WARN_ON to drm_atomic_state_clear (obviously only when we actually have a state to clear up). And reorder all the state_clear and backoff calls to make it work out properly. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:07 +00:00
drm_atomic_legacy_backoff(state);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
/*
* Someone might have exchanged the framebuffer while we dropped locks
* in the backoff code. We need to fix up the fb refcount tracking the
* core does for us.
*/
crtc->primary->old_fb = crtc->primary->fb;
goto retry;
}
EXPORT_SYMBOL(drm_atomic_helper_set_config);
/**
* drm_atomic_helper_crtc_set_property - helper for crtc prorties
* @crtc: DRM crtc
* @property: DRM property
* @val: value of property
*
* Provides a default plane disablle handler using the atomic driver interface.
*
* RETURNS:
* Zero on success, error code on failure
*/
int
drm_atomic_helper_crtc_set_property(struct drm_crtc *crtc,
struct drm_property *property,
uint64_t val)
{
struct drm_atomic_state *state;
struct drm_crtc_state *crtc_state;
int ret = 0;
state = drm_atomic_state_alloc(crtc->dev);
if (!state)
return -ENOMEM;
/* ->set_property is always called with all locks held. */
state->acquire_ctx = crtc->dev->mode_config.acquire_ctx;
retry:
crtc_state = drm_atomic_get_crtc_state(state, crtc);
if (IS_ERR(crtc_state)) {
ret = PTR_ERR(crtc_state);
goto fail;
}
ret = crtc->funcs->atomic_set_property(crtc, crtc_state,
property, val);
if (ret)
goto fail;
ret = drm_atomic_commit(state);
if (ret != 0)
goto fail;
/* Driver takes ownership of state on successful commit. */
return 0;
fail:
if (ret == -EDEADLK)
goto backoff;
drm_atomic_state_free(state);
return ret;
backoff:
drm_atomic_state_clear(state);
drm/atomic: Only destroy connector states with connection mutex held Otherwise the connector might have been unplugged and destroyed while we didn't look. Yet another fallout from DP MST hotplugging that I didn't consider. To make sure we get this right add an appropriate WARN_ON to drm_atomic_state_clear (obviously only when we actually have a state to clear up). And reorder all the state_clear and backoff calls to make it work out properly. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:07 +00:00
drm_atomic_legacy_backoff(state);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
goto retry;
}
EXPORT_SYMBOL(drm_atomic_helper_crtc_set_property);
/**
* drm_atomic_helper_plane_set_property - helper for plane prorties
* @plane: DRM plane
* @property: DRM property
* @val: value of property
*
* Provides a default plane disable handler using the atomic driver interface.
*
* RETURNS:
* Zero on success, error code on failure
*/
int
drm_atomic_helper_plane_set_property(struct drm_plane *plane,
struct drm_property *property,
uint64_t val)
{
struct drm_atomic_state *state;
struct drm_plane_state *plane_state;
int ret = 0;
state = drm_atomic_state_alloc(plane->dev);
if (!state)
return -ENOMEM;
/* ->set_property is always called with all locks held. */
state->acquire_ctx = plane->dev->mode_config.acquire_ctx;
retry:
plane_state = drm_atomic_get_plane_state(state, plane);
if (IS_ERR(plane_state)) {
ret = PTR_ERR(plane_state);
goto fail;
}
ret = plane->funcs->atomic_set_property(plane, plane_state,
property, val);
if (ret)
goto fail;
ret = drm_atomic_commit(state);
if (ret != 0)
goto fail;
/* Driver takes ownership of state on successful commit. */
return 0;
fail:
if (ret == -EDEADLK)
goto backoff;
drm_atomic_state_free(state);
return ret;
backoff:
drm_atomic_state_clear(state);
drm/atomic: Only destroy connector states with connection mutex held Otherwise the connector might have been unplugged and destroyed while we didn't look. Yet another fallout from DP MST hotplugging that I didn't consider. To make sure we get this right add an appropriate WARN_ON to drm_atomic_state_clear (obviously only when we actually have a state to clear up). And reorder all the state_clear and backoff calls to make it work out properly. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:07 +00:00
drm_atomic_legacy_backoff(state);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
goto retry;
}
EXPORT_SYMBOL(drm_atomic_helper_plane_set_property);
/**
* drm_atomic_helper_connector_set_property - helper for connector prorties
* @connector: DRM connector
* @property: DRM property
* @val: value of property
*
* Provides a default plane disablle handler using the atomic driver interface.
*
* RETURNS:
* Zero on success, error code on failure
*/
int
drm_atomic_helper_connector_set_property(struct drm_connector *connector,
struct drm_property *property,
uint64_t val)
{
struct drm_atomic_state *state;
struct drm_connector_state *connector_state;
int ret = 0;
state = drm_atomic_state_alloc(connector->dev);
if (!state)
return -ENOMEM;
/* ->set_property is always called with all locks held. */
state->acquire_ctx = connector->dev->mode_config.acquire_ctx;
retry:
connector_state = drm_atomic_get_connector_state(state, connector);
if (IS_ERR(connector_state)) {
ret = PTR_ERR(connector_state);
goto fail;
}
ret = connector->funcs->atomic_set_property(connector, connector_state,
property, val);
if (ret)
goto fail;
ret = drm_atomic_commit(state);
if (ret != 0)
goto fail;
/* Driver takes ownership of state on successful commit. */
return 0;
fail:
if (ret == -EDEADLK)
goto backoff;
drm_atomic_state_free(state);
return ret;
backoff:
drm_atomic_state_clear(state);
drm/atomic: Only destroy connector states with connection mutex held Otherwise the connector might have been unplugged and destroyed while we didn't look. Yet another fallout from DP MST hotplugging that I didn't consider. To make sure we get this right add an appropriate WARN_ON to drm_atomic_state_clear (obviously only when we actually have a state to clear up). And reorder all the state_clear and backoff calls to make it work out properly. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:07 +00:00
drm_atomic_legacy_backoff(state);
drm/atomic-helper: implementatations for legacy interfaces Well, except page_flip since that requires async commit, which isn't there yet. For the functions which changes planes there's a bit of trickery involved to keep the fb refcounting working. But otherwise fairly straight-forward atomic updates. The property setting functions are still a bit incomplete. Once we have generic properties (e.g. rotation, but also all the properties needed by the atomic ioctl) we need to filter those out and parse them in the helper. Preferrably with the same function as used by the real atomic ioctl implementation. v2: Fixup kerneldoc, reported by Paulo. v3: Add missing EXPORT_SYMBOL. v4: We need to look at the crtc of the modeset, not some random leftover one from a previous loop when udpating the connector->crtc routing. Also push some local variables into inner loops to avoid these kinds of bugs. v5: Adjust semantics - drivers now own the atomic state upon successfully synchronous commit. v6: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v7: - Improve comments. - Filter out the crtc of the ->set_config call when recomputing crtc_state->enabled: We should compute the same state, but not doing so will give us a good chance to catch bugs and inconsistencies - the atomic helper's atomic_check function re-validates this again. - Fix the set_config implementation logic when disabling the crtc: We still need to update the output routing to disable all the connectors properly in the state. Caught by the atomic_check functions, so at least that part worked ;-) Also add some WARN_ONs to ensure ->set_config preconditions all apply. v8: Fixup an embarrassing h/vdisplay mixup. v9: Shuffled bad squash to the right patch, spotted by Daniel v10: Use set_crtc_for_connector as suggested by Sean. v11: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v12: Review and discussion with Sean: - One spelling fix. - Correctly skip the crtc from the set_config set when recomputing ->enable state. That should allow us to catch any bugs in higher levels in computing that state (which is supplied to the ->set_config implementation). I've screwed this up and Sean spotted that the current code is pointless. Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 11:46:52 +00:00
goto retry;
}
EXPORT_SYMBOL(drm_atomic_helper_connector_set_property);
drm/atomic-helper: implement ->page_flip Currently there is no way to implement async flips using atomic, that essentially requires us to be able to cancel pending requests mid-flight. To be able to do that (and I guess we want this since vblank synced updates which opportunistically cancel still pending updates seem to be wanted) we'd need to add a mandatory cancellation mode. Depending upon the exact semantics we decide upon that could mean that userspace will not get completion events, or will get them all stacked up. So reject async updates for now. Also async updates usually means not vblank synced at all, and I guess for drivers which want to support this they should simply add a special pageflip handler (since usually you need a special flip cmd to achieve this). That kind of async flip is pretty much exclusively just used for games and benchmarks where dropping just one frame means you'll get a headshot or something bad like that ... And so slight amounts of tearing is acceptable. v2: Fixup kerneldoc, reported by Paulo. v3: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v4: Update crtc->primary->fb since ->page_flip is the only driver callback where the core won't do this itself. We might want to fix this inconsistency eventually. v5: Use set_crtc_for_connector as suggested by Sean. v6: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v7: Fix spelling mistake in the commit message (Sean). Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 16:42:37 +00:00
/**
* drm_atomic_helper_page_flip - execute a legacy page flip
* @crtc: DRM crtc
* @fb: DRM framebuffer
* @event: optional DRM event to signal upon completion
* @flags: flip flags for non-vblank sync'ed updates
*
* Provides a default page flip implementation using the atomic driver interface.
*
* Note that for now so called async page flips (i.e. updates which are not
* synchronized to vblank) are not supported, since the atomic interfaces have
* no provisions for this yet.
*
* Returns:
* Returns 0 on success, negative errno numbers on failure.
*/
int drm_atomic_helper_page_flip(struct drm_crtc *crtc,
struct drm_framebuffer *fb,
struct drm_pending_vblank_event *event,
uint32_t flags)
{
struct drm_plane *plane = crtc->primary;
struct drm_atomic_state *state;
struct drm_plane_state *plane_state;
struct drm_crtc_state *crtc_state;
int ret = 0;
if (flags & DRM_MODE_PAGE_FLIP_ASYNC)
return -EINVAL;
state = drm_atomic_state_alloc(plane->dev);
if (!state)
return -ENOMEM;
state->acquire_ctx = drm_modeset_legacy_acquire_ctx(crtc);
retry:
crtc_state = drm_atomic_get_crtc_state(state, crtc);
if (IS_ERR(crtc_state)) {
ret = PTR_ERR(crtc_state);
goto fail;
}
crtc_state->event = event;
plane_state = drm_atomic_get_plane_state(state, plane);
if (IS_ERR(plane_state)) {
ret = PTR_ERR(plane_state);
goto fail;
}
drm/atomic: track bitmask of planes attached to crtc Chasing plane->state->crtc of planes that are *not* part of the same atomic update is racy, making it incredibly awkward (or impossible) to do something simple like iterate over all planes and figure out which ones are attached to a crtc. Solve this by adding a bitmask of currently attached planes in the crtc-state. Note that the transitional helpers do not maintain the plane_mask. But they only support the legacy ioctls, which have sufficient brute-force locking around plane updates that they can continue to loop over all planes to see what is attached to a crtc the old way. Signed-off-by: Rob Clark <robdclark@gmail.com> [danvet: - Drop comments about locking in set_crtc_for_plane since they're a bit misleading - we already should hold lock for the current crtc. - Also WARN_ON if get_state on the old crtc fails since that should have been done already. - Squash in fixup to check get_plane_state return value, reported by Dan Carpenter and acked by Rob Clark.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-21 20:28:31 +00:00
ret = drm_atomic_set_crtc_for_plane(state, plane, crtc);
drm/atomic-helper: implement ->page_flip Currently there is no way to implement async flips using atomic, that essentially requires us to be able to cancel pending requests mid-flight. To be able to do that (and I guess we want this since vblank synced updates which opportunistically cancel still pending updates seem to be wanted) we'd need to add a mandatory cancellation mode. Depending upon the exact semantics we decide upon that could mean that userspace will not get completion events, or will get them all stacked up. So reject async updates for now. Also async updates usually means not vblank synced at all, and I guess for drivers which want to support this they should simply add a special pageflip handler (since usually you need a special flip cmd to achieve this). That kind of async flip is pretty much exclusively just used for games and benchmarks where dropping just one frame means you'll get a headshot or something bad like that ... And so slight amounts of tearing is acceptable. v2: Fixup kerneldoc, reported by Paulo. v3: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v4: Update crtc->primary->fb since ->page_flip is the only driver callback where the core won't do this itself. We might want to fix this inconsistency eventually. v5: Use set_crtc_for_connector as suggested by Sean. v6: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v7: Fix spelling mistake in the commit message (Sean). Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 16:42:37 +00:00
if (ret != 0)
goto fail;
drm/atomic: Refcounting for plane_state->fb So my original plan was that the drm core refcounts framebuffers like with the legacy ioctls. But that doesn't work for a bunch of reasons: - State objects might live longer than until the next fb change happens for a plane. For example delayed cleanup work only happens _after_ the pageflip ioctl has completed. So this definitely doesn't work without the plane state holding its own references. - The other issue is transition from legacy to atomic implementations, where the driver works under a mix of both worlds. Which means legacy paths might not properly update the ->fb pointer under plane->state->fb. Which is a bit a problem when then someone comes around and _does_ try to clean it up when it's long gone. The second issue is just a bit a transition bug, since drivers should update plane->state->fb in all the paths that aren't converted yet. But a bit more robustness for the transition can't hurt - we pull similar tricks with cleaning up the old fb in the transitional helpers already. The pattern for drivers that transition is if (plane->state) drm_atomic_set_fb_for_plane(plane->state, plane->fb); inserted after the fb update has logically completed at the end of ->set_config (or ->set_base/mode_set if using the crtc helpers), ->page_flip, ->update_plane or any other entry point which updates plane->fb. v2: Update kerneldoc - copypasta fail. v3: Fix spelling in the commit message (Sean). Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-11-04 21:57:27 +00:00
drm_atomic_set_fb_for_plane(plane_state, fb);
drm/atomic-helper: implement ->page_flip Currently there is no way to implement async flips using atomic, that essentially requires us to be able to cancel pending requests mid-flight. To be able to do that (and I guess we want this since vblank synced updates which opportunistically cancel still pending updates seem to be wanted) we'd need to add a mandatory cancellation mode. Depending upon the exact semantics we decide upon that could mean that userspace will not get completion events, or will get them all stacked up. So reject async updates for now. Also async updates usually means not vblank synced at all, and I guess for drivers which want to support this they should simply add a special pageflip handler (since usually you need a special flip cmd to achieve this). That kind of async flip is pretty much exclusively just used for games and benchmarks where dropping just one frame means you'll get a headshot or something bad like that ... And so slight amounts of tearing is acceptable. v2: Fixup kerneldoc, reported by Paulo. v3: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v4: Update crtc->primary->fb since ->page_flip is the only driver callback where the core won't do this itself. We might want to fix this inconsistency eventually. v5: Use set_crtc_for_connector as suggested by Sean. v6: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v7: Fix spelling mistake in the commit message (Sean). Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 16:42:37 +00:00
ret = drm_atomic_async_commit(state);
if (ret != 0)
goto fail;
/* TODO: ->page_flip is the only driver callback where the core
* doesn't update plane->fb. For now patch it up here. */
plane->fb = plane->state->fb;
/* Driver takes ownership of state on successful async commit. */
return 0;
fail:
if (ret == -EDEADLK)
goto backoff;
drm_atomic_state_free(state);
return ret;
backoff:
drm_atomic_state_clear(state);
drm/atomic: Only destroy connector states with connection mutex held Otherwise the connector might have been unplugged and destroyed while we didn't look. Yet another fallout from DP MST hotplugging that I didn't consider. To make sure we get this right add an appropriate WARN_ON to drm_atomic_state_clear (obviously only when we actually have a state to clear up). And reorder all the state_clear and backoff calls to make it work out properly. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-11-19 17:38:07 +00:00
drm_atomic_legacy_backoff(state);
drm/atomic-helper: implement ->page_flip Currently there is no way to implement async flips using atomic, that essentially requires us to be able to cancel pending requests mid-flight. To be able to do that (and I guess we want this since vblank synced updates which opportunistically cancel still pending updates seem to be wanted) we'd need to add a mandatory cancellation mode. Depending upon the exact semantics we decide upon that could mean that userspace will not get completion events, or will get them all stacked up. So reject async updates for now. Also async updates usually means not vblank synced at all, and I guess for drivers which want to support this they should simply add a special pageflip handler (since usually you need a special flip cmd to achieve this). That kind of async flip is pretty much exclusively just used for games and benchmarks where dropping just one frame means you'll get a headshot or something bad like that ... And so slight amounts of tearing is acceptable. v2: Fixup kerneldoc, reported by Paulo. v3: Use the set_crtc_for_plane function to assign the crtc, since otherwise the book-keeping is off. v4: Update crtc->primary->fb since ->page_flip is the only driver callback where the core won't do this itself. We might want to fix this inconsistency eventually. v5: Use set_crtc_for_connector as suggested by Sean. v6: Daniel Thompson noticed that my error handling is inconsistent and that in a few cases I didn't handle fatal errors (i.e. not -EDEADLK). Fix this by consolidate the ww mutex backoff handling into one check in the fail: block and flatten the error control flow everywhere else. v7: Fix spelling mistake in the commit message (Sean). Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Cc: Paulo Zanoni <przanoni@gmail.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-27 16:42:37 +00:00
/*
* Someone might have exchanged the framebuffer while we dropped locks
* in the backoff code. We need to fix up the fb refcount tracking the
* core does for us.
*/
plane->old_fb = plane->fb;
goto retry;
}
EXPORT_SYMBOL(drm_atomic_helper_page_flip);
drm/atomic-helpers: functions for state duplicate/destroy/reset The atomic users and helpers assume that there is always a obj->state structure around. Which means drivers need to somehow create that at driver load time. Also it should obviously reset hardware state, so needs to be reset upon resume. Finally the destroy/duplicate_state functions are an awful lot of boilerplate if the driver doesn't need anything beyond the default state objects. So add helper functions for all of this. v2: Somehow the plane/connector versions got lost in the first version. v3: Add kerneldoc. v4: Make duplicate_state functions a bit more robust, which is useful for debugging state tracking issues when transitioning to atomic. v5: Clear temporary variables in the crtc state when duplicating it, like ->mode_changed or ->planes_changed. If we don't do this stale values for these might pollute the next atomic modeset. v6: Also clear crtc_state->event in case the driver didn't (yet) clear this out. v7: Split out wrong squashed commit. Also improve the kerneldoc to mention that obj->state can be NULL and when. Both suggested by Daniel Thompson. Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-03 14:56:43 +00:00
drm: Docbook integration and over sections for all the new helpers In all cases the text requires that new drivers are converted to the atomic interfaces. v2: Add overview for state handling. v3: Review from Sean: Some spelling fixes and drop the misguided hunk to remove rgba8888 from the plane helpers compat list. Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-06 19:53:29 +00:00
/**
* DOC: atomic state reset and initialization
*
* Both the drm core and the atomic helpers assume that there is always the full
* and correct atomic software state for all connectors, CRTCs and planes
* available. Which is a bit a problem on driver load and also after system
* suspend. One way to solve this is to have a hardware state read-out
* infrastructure which reconstructs the full software state (e.g. the i915
* driver).
*
* The simpler solution is to just reset the software state to everything off,
* which is easiest to do by calling drm_mode_config_reset(). To facilitate this
* the atomic helpers provide default reset implementations for all hooks.
*/
drm/atomic-helpers: functions for state duplicate/destroy/reset The atomic users and helpers assume that there is always a obj->state structure around. Which means drivers need to somehow create that at driver load time. Also it should obviously reset hardware state, so needs to be reset upon resume. Finally the destroy/duplicate_state functions are an awful lot of boilerplate if the driver doesn't need anything beyond the default state objects. So add helper functions for all of this. v2: Somehow the plane/connector versions got lost in the first version. v3: Add kerneldoc. v4: Make duplicate_state functions a bit more robust, which is useful for debugging state tracking issues when transitioning to atomic. v5: Clear temporary variables in the crtc state when duplicating it, like ->mode_changed or ->planes_changed. If we don't do this stale values for these might pollute the next atomic modeset. v6: Also clear crtc_state->event in case the driver didn't (yet) clear this out. v7: Split out wrong squashed commit. Also improve the kerneldoc to mention that obj->state can be NULL and when. Both suggested by Daniel Thompson. Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-03 14:56:43 +00:00
/**
* drm_atomic_helper_crtc_reset - default ->reset hook for CRTCs
* @crtc: drm CRTC
*
* Resets the atomic state for @crtc by freeing the state pointer (which might
* be NULL, e.g. at driver load time) and allocating a new empty state object.
*/
void drm_atomic_helper_crtc_reset(struct drm_crtc *crtc)
{
kfree(crtc->state);
crtc->state = kzalloc(sizeof(*crtc->state), GFP_KERNEL);
}
EXPORT_SYMBOL(drm_atomic_helper_crtc_reset);
/**
* drm_atomic_helper_crtc_duplicate_state - default state duplicate hook
* @crtc: drm CRTC
*
* Default CRTC state duplicate hook for drivers which don't have their own
* subclassed CRTC state structure.
*/
struct drm_crtc_state *
drm_atomic_helper_crtc_duplicate_state(struct drm_crtc *crtc)
{
struct drm_crtc_state *state;
if (WARN_ON(!crtc->state))
return NULL;
state = kmemdup(crtc->state, sizeof(*crtc->state), GFP_KERNEL);
if (state) {
state->mode_changed = false;
state->planes_changed = false;
state->event = NULL;
}
return state;
}
EXPORT_SYMBOL(drm_atomic_helper_crtc_duplicate_state);
/**
* drm_atomic_helper_crtc_destroy_state - default state destroy hook
* @crtc: drm CRTC
* @state: CRTC state object to release
*
* Default CRTC state destroy hook for drivers which don't have their own
* subclassed CRTC state structure.
*/
void drm_atomic_helper_crtc_destroy_state(struct drm_crtc *crtc,
struct drm_crtc_state *state)
{
kfree(state);
}
EXPORT_SYMBOL(drm_atomic_helper_crtc_destroy_state);
/**
* drm_atomic_helper_plane_reset - default ->reset hook for planes
* @plane: drm plane
*
* Resets the atomic state for @plane by freeing the state pointer (which might
* be NULL, e.g. at driver load time) and allocating a new empty state object.
*/
void drm_atomic_helper_plane_reset(struct drm_plane *plane)
{
drm/atomic: Refcounting for plane_state->fb So my original plan was that the drm core refcounts framebuffers like with the legacy ioctls. But that doesn't work for a bunch of reasons: - State objects might live longer than until the next fb change happens for a plane. For example delayed cleanup work only happens _after_ the pageflip ioctl has completed. So this definitely doesn't work without the plane state holding its own references. - The other issue is transition from legacy to atomic implementations, where the driver works under a mix of both worlds. Which means legacy paths might not properly update the ->fb pointer under plane->state->fb. Which is a bit a problem when then someone comes around and _does_ try to clean it up when it's long gone. The second issue is just a bit a transition bug, since drivers should update plane->state->fb in all the paths that aren't converted yet. But a bit more robustness for the transition can't hurt - we pull similar tricks with cleaning up the old fb in the transitional helpers already. The pattern for drivers that transition is if (plane->state) drm_atomic_set_fb_for_plane(plane->state, plane->fb); inserted after the fb update has logically completed at the end of ->set_config (or ->set_base/mode_set if using the crtc helpers), ->page_flip, ->update_plane or any other entry point which updates plane->fb. v2: Update kerneldoc - copypasta fail. v3: Fix spelling in the commit message (Sean). Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-11-04 21:57:27 +00:00
if (plane->state && plane->state->fb)
drm_framebuffer_unreference(plane->state->fb);
drm/atomic-helpers: functions for state duplicate/destroy/reset The atomic users and helpers assume that there is always a obj->state structure around. Which means drivers need to somehow create that at driver load time. Also it should obviously reset hardware state, so needs to be reset upon resume. Finally the destroy/duplicate_state functions are an awful lot of boilerplate if the driver doesn't need anything beyond the default state objects. So add helper functions for all of this. v2: Somehow the plane/connector versions got lost in the first version. v3: Add kerneldoc. v4: Make duplicate_state functions a bit more robust, which is useful for debugging state tracking issues when transitioning to atomic. v5: Clear temporary variables in the crtc state when duplicating it, like ->mode_changed or ->planes_changed. If we don't do this stale values for these might pollute the next atomic modeset. v6: Also clear crtc_state->event in case the driver didn't (yet) clear this out. v7: Split out wrong squashed commit. Also improve the kerneldoc to mention that obj->state can be NULL and when. Both suggested by Daniel Thompson. Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-03 14:56:43 +00:00
kfree(plane->state);
plane->state = kzalloc(sizeof(*plane->state), GFP_KERNEL);
}
EXPORT_SYMBOL(drm_atomic_helper_plane_reset);
/**
* drm_atomic_helper_plane_duplicate_state - default state duplicate hook
* @plane: drm plane
*
* Default plane state duplicate hook for drivers which don't have their own
* subclassed plane state structure.
*/
struct drm_plane_state *
drm_atomic_helper_plane_duplicate_state(struct drm_plane *plane)
{
drm/atomic: Refcounting for plane_state->fb So my original plan was that the drm core refcounts framebuffers like with the legacy ioctls. But that doesn't work for a bunch of reasons: - State objects might live longer than until the next fb change happens for a plane. For example delayed cleanup work only happens _after_ the pageflip ioctl has completed. So this definitely doesn't work without the plane state holding its own references. - The other issue is transition from legacy to atomic implementations, where the driver works under a mix of both worlds. Which means legacy paths might not properly update the ->fb pointer under plane->state->fb. Which is a bit a problem when then someone comes around and _does_ try to clean it up when it's long gone. The second issue is just a bit a transition bug, since drivers should update plane->state->fb in all the paths that aren't converted yet. But a bit more robustness for the transition can't hurt - we pull similar tricks with cleaning up the old fb in the transitional helpers already. The pattern for drivers that transition is if (plane->state) drm_atomic_set_fb_for_plane(plane->state, plane->fb); inserted after the fb update has logically completed at the end of ->set_config (or ->set_base/mode_set if using the crtc helpers), ->page_flip, ->update_plane or any other entry point which updates plane->fb. v2: Update kerneldoc - copypasta fail. v3: Fix spelling in the commit message (Sean). Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-11-04 21:57:27 +00:00
struct drm_plane_state *state;
drm/atomic-helpers: functions for state duplicate/destroy/reset The atomic users and helpers assume that there is always a obj->state structure around. Which means drivers need to somehow create that at driver load time. Also it should obviously reset hardware state, so needs to be reset upon resume. Finally the destroy/duplicate_state functions are an awful lot of boilerplate if the driver doesn't need anything beyond the default state objects. So add helper functions for all of this. v2: Somehow the plane/connector versions got lost in the first version. v3: Add kerneldoc. v4: Make duplicate_state functions a bit more robust, which is useful for debugging state tracking issues when transitioning to atomic. v5: Clear temporary variables in the crtc state when duplicating it, like ->mode_changed or ->planes_changed. If we don't do this stale values for these might pollute the next atomic modeset. v6: Also clear crtc_state->event in case the driver didn't (yet) clear this out. v7: Split out wrong squashed commit. Also improve the kerneldoc to mention that obj->state can be NULL and when. Both suggested by Daniel Thompson. Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-03 14:56:43 +00:00
if (WARN_ON(!plane->state))
return NULL;
drm/atomic: Refcounting for plane_state->fb So my original plan was that the drm core refcounts framebuffers like with the legacy ioctls. But that doesn't work for a bunch of reasons: - State objects might live longer than until the next fb change happens for a plane. For example delayed cleanup work only happens _after_ the pageflip ioctl has completed. So this definitely doesn't work without the plane state holding its own references. - The other issue is transition from legacy to atomic implementations, where the driver works under a mix of both worlds. Which means legacy paths might not properly update the ->fb pointer under plane->state->fb. Which is a bit a problem when then someone comes around and _does_ try to clean it up when it's long gone. The second issue is just a bit a transition bug, since drivers should update plane->state->fb in all the paths that aren't converted yet. But a bit more robustness for the transition can't hurt - we pull similar tricks with cleaning up the old fb in the transitional helpers already. The pattern for drivers that transition is if (plane->state) drm_atomic_set_fb_for_plane(plane->state, plane->fb); inserted after the fb update has logically completed at the end of ->set_config (or ->set_base/mode_set if using the crtc helpers), ->page_flip, ->update_plane or any other entry point which updates plane->fb. v2: Update kerneldoc - copypasta fail. v3: Fix spelling in the commit message (Sean). Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-11-04 21:57:27 +00:00
state = kmemdup(plane->state, sizeof(*plane->state), GFP_KERNEL);
if (state && state->fb)
drm_framebuffer_reference(state->fb);
return state;
drm/atomic-helpers: functions for state duplicate/destroy/reset The atomic users and helpers assume that there is always a obj->state structure around. Which means drivers need to somehow create that at driver load time. Also it should obviously reset hardware state, so needs to be reset upon resume. Finally the destroy/duplicate_state functions are an awful lot of boilerplate if the driver doesn't need anything beyond the default state objects. So add helper functions for all of this. v2: Somehow the plane/connector versions got lost in the first version. v3: Add kerneldoc. v4: Make duplicate_state functions a bit more robust, which is useful for debugging state tracking issues when transitioning to atomic. v5: Clear temporary variables in the crtc state when duplicating it, like ->mode_changed or ->planes_changed. If we don't do this stale values for these might pollute the next atomic modeset. v6: Also clear crtc_state->event in case the driver didn't (yet) clear this out. v7: Split out wrong squashed commit. Also improve the kerneldoc to mention that obj->state can be NULL and when. Both suggested by Daniel Thompson. Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-03 14:56:43 +00:00
}
EXPORT_SYMBOL(drm_atomic_helper_plane_duplicate_state);
/**
* drm_atomic_helper_plane_destroy_state - default state destroy hook
* @plane: drm plane
* @state: plane state object to release
*
* Default plane state destroy hook for drivers which don't have their own
* subclassed plane state structure.
*/
void drm_atomic_helper_plane_destroy_state(struct drm_plane *plane,
drm/atomic: Refcounting for plane_state->fb So my original plan was that the drm core refcounts framebuffers like with the legacy ioctls. But that doesn't work for a bunch of reasons: - State objects might live longer than until the next fb change happens for a plane. For example delayed cleanup work only happens _after_ the pageflip ioctl has completed. So this definitely doesn't work without the plane state holding its own references. - The other issue is transition from legacy to atomic implementations, where the driver works under a mix of both worlds. Which means legacy paths might not properly update the ->fb pointer under plane->state->fb. Which is a bit a problem when then someone comes around and _does_ try to clean it up when it's long gone. The second issue is just a bit a transition bug, since drivers should update plane->state->fb in all the paths that aren't converted yet. But a bit more robustness for the transition can't hurt - we pull similar tricks with cleaning up the old fb in the transitional helpers already. The pattern for drivers that transition is if (plane->state) drm_atomic_set_fb_for_plane(plane->state, plane->fb); inserted after the fb update has logically completed at the end of ->set_config (or ->set_base/mode_set if using the crtc helpers), ->page_flip, ->update_plane or any other entry point which updates plane->fb. v2: Update kerneldoc - copypasta fail. v3: Fix spelling in the commit message (Sean). Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-11-04 21:57:27 +00:00
struct drm_plane_state *state)
drm/atomic-helpers: functions for state duplicate/destroy/reset The atomic users and helpers assume that there is always a obj->state structure around. Which means drivers need to somehow create that at driver load time. Also it should obviously reset hardware state, so needs to be reset upon resume. Finally the destroy/duplicate_state functions are an awful lot of boilerplate if the driver doesn't need anything beyond the default state objects. So add helper functions for all of this. v2: Somehow the plane/connector versions got lost in the first version. v3: Add kerneldoc. v4: Make duplicate_state functions a bit more robust, which is useful for debugging state tracking issues when transitioning to atomic. v5: Clear temporary variables in the crtc state when duplicating it, like ->mode_changed or ->planes_changed. If we don't do this stale values for these might pollute the next atomic modeset. v6: Also clear crtc_state->event in case the driver didn't (yet) clear this out. v7: Split out wrong squashed commit. Also improve the kerneldoc to mention that obj->state can be NULL and when. Both suggested by Daniel Thompson. Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-03 14:56:43 +00:00
{
drm/atomic: Refcounting for plane_state->fb So my original plan was that the drm core refcounts framebuffers like with the legacy ioctls. But that doesn't work for a bunch of reasons: - State objects might live longer than until the next fb change happens for a plane. For example delayed cleanup work only happens _after_ the pageflip ioctl has completed. So this definitely doesn't work without the plane state holding its own references. - The other issue is transition from legacy to atomic implementations, where the driver works under a mix of both worlds. Which means legacy paths might not properly update the ->fb pointer under plane->state->fb. Which is a bit a problem when then someone comes around and _does_ try to clean it up when it's long gone. The second issue is just a bit a transition bug, since drivers should update plane->state->fb in all the paths that aren't converted yet. But a bit more robustness for the transition can't hurt - we pull similar tricks with cleaning up the old fb in the transitional helpers already. The pattern for drivers that transition is if (plane->state) drm_atomic_set_fb_for_plane(plane->state, plane->fb); inserted after the fb update has logically completed at the end of ->set_config (or ->set_base/mode_set if using the crtc helpers), ->page_flip, ->update_plane or any other entry point which updates plane->fb. v2: Update kerneldoc - copypasta fail. v3: Fix spelling in the commit message (Sean). Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
2014-11-04 21:57:27 +00:00
if (state->fb)
drm_framebuffer_unreference(state->fb);
drm/atomic-helpers: functions for state duplicate/destroy/reset The atomic users and helpers assume that there is always a obj->state structure around. Which means drivers need to somehow create that at driver load time. Also it should obviously reset hardware state, so needs to be reset upon resume. Finally the destroy/duplicate_state functions are an awful lot of boilerplate if the driver doesn't need anything beyond the default state objects. So add helper functions for all of this. v2: Somehow the plane/connector versions got lost in the first version. v3: Add kerneldoc. v4: Make duplicate_state functions a bit more robust, which is useful for debugging state tracking issues when transitioning to atomic. v5: Clear temporary variables in the crtc state when duplicating it, like ->mode_changed or ->planes_changed. If we don't do this stale values for these might pollute the next atomic modeset. v6: Also clear crtc_state->event in case the driver didn't (yet) clear this out. v7: Split out wrong squashed commit. Also improve the kerneldoc to mention that obj->state can be NULL and when. Both suggested by Daniel Thompson. Cc: Daniel Thompson <daniel.thompson@linaro.org> Cc: Sean Paul <seanpaul@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-11-03 14:56:43 +00:00
kfree(state);
}
EXPORT_SYMBOL(drm_atomic_helper_plane_destroy_state);
/**
* drm_atomic_helper_connector_reset - default ->reset hook for connectors
* @connector: drm connector
*
* Resets the atomic state for @connector by freeing the state pointer (which
* might be NULL, e.g. at driver load time) and allocating a new empty state
* object.
*/
void drm_atomic_helper_connector_reset(struct drm_connector *connector)
{
kfree(connector->state);
connector->state = kzalloc(sizeof(*connector->state), GFP_KERNEL);
}
EXPORT_SYMBOL(drm_atomic_helper_connector_reset);
/**
* drm_atomic_helper_connector_duplicate_state - default state duplicate hook
* @connector: drm connector
*
* Default connector state duplicate hook for drivers which don't have their own
* subclassed connector state structure.
*/
struct drm_connector_state *
drm_atomic_helper_connector_duplicate_state(struct drm_connector *connector)
{
if (WARN_ON(!connector->state))
return NULL;
return kmemdup(connector->state, sizeof(*connector->state), GFP_KERNEL);
}
EXPORT_SYMBOL(drm_atomic_helper_connector_duplicate_state);
/**
* drm_atomic_helper_connector_destroy_state - default state destroy hook
* @connector: drm connector
* @state: connector state object to release
*
* Default connector state destroy hook for drivers which don't have their own
* subclassed connector state structure.
*/
void drm_atomic_helper_connector_destroy_state(struct drm_connector *connector,
struct drm_connector_state *state)
{
kfree(state);
}
EXPORT_SYMBOL(drm_atomic_helper_connector_destroy_state);
Reference in New Issue Copy Permalink