dri-devel April 2021

dri-devel@lists.freedesktop.org

317 participants
715 discussions

[PATCH v4] drm/amd/amdgpu/amdgpu_drv.c: Replace drm_modeset_lock_all with drm_modeset_lock
by Fabio M. De Francesco 04 May '21

04 May '21

drm_modeset_lock_all() is not needed here, so it is replaced with drm_modeset_lock(). The crtc list around which we are looping never changes, therefore the only lock we need is to protect access to crtc->state. Suggested-by: Daniel Vetter <daniel(a)ffwll.ch> Suggested-by: Matthew Wilcox <willy(a)infradead.org> Signed-off-by: Fabio M. De Francesco <fmdefrancesco(a)gmail.com> Reviewed-by: Matthew Wilcox (Oracle) <willy(a)infradead.org> --- Changes from v3: CC'ed more (previously missing) maintainers. Changes from v2: Drop file name from the Subject. Cc'ed all maintainers. Changes from v1: Removed unnecessary braces around single statement block. drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 80130c1c0c68..39204dbc168b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -1595,17 +1595,15 @@ static int amdgpu_pmops_runtime_idle(struct device *dev) if (amdgpu_device_has_dc_support(adev)) { struct drm_crtc *crtc; - drm_modeset_lock_all(drm_dev); - drm_for_each_crtc(crtc, drm_dev) { - if (crtc->state->active) { + drm_modeset_lock(&crtc->mutex, NULL); + if (crtc->state->active) ret = -EBUSY; + drm_modeset_unlock(&crtc->mutex); + if (ret < 0) break; - } } - drm_modeset_unlock_all(drm_dev); - } else { struct drm_connector *list_connector; struct drm_connector_list_iter iter; -- 2.31.1

2 1

[PATCH] drm/amd/pm: initialize variable
by trix＠redhat.com 04 May '21

04 May '21

From: Tom Rix <trix(a)redhat.com> Static analysis reports this problem amdgpu_pm.c:478:16: warning: The right operand of '<' is a garbage value for (i = 0; i < data.nums; i++) { ^ ~~~~~~~~~ In some cases data is not set. Initialize to 0 and flag not setting data as an error with the existing check. Signed-off-by: Tom Rix <trix(a)redhat.com> --- drivers/gpu/drm/amd/pm/amdgpu_pm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c index 4e459ef632ef..9a54066ec0af 100644 --- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c @@ -451,7 +451,7 @@ static ssize_t amdgpu_get_pp_cur_state(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = drm_to_adev(ddev); const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; - struct pp_states_info data; + struct pp_states_info data = {0}; enum amd_pm_state_type pm = 0; int i = 0, ret = 0; -- 2.26.3

2 1

i.MX53 error during GPU use
by Otavio Salvador 04 May '21

04 May '21

Hello all, We found this error when using Freedreno driver on an i.MX53 device with Wayland. Any idea how to fix this? [ 32.414110] [drm:msm_ioctl_gem_submit] *ERROR* invalid cmdstream size: 0 [ 39.177075] [ 39.178617] ====================================================== [ 39.184804] WARNING: possible circular locking dependency detected [ 39.190997] 5.10.31+g7ae1de1d2bd3 #1 Not tainted [ 39.195619] ------------------------------------------------------ [ 39.201805] kworker/u2:1/68 is trying to acquire lock: [ 39.206953] c2b952fc (reservation_ww_class_mutex){+.+.}-{3:3}, at: dma_buf_detach+0xd0/0x1e0 [ 39.215448] [ 39.215448] but task is already holding lock: [ 39.221286] c2b957b0 (&msm_obj->lock){+.+.}-{3:3}, at: msm_gem_free_work+0x118/0x1b0 [ 39.229062] [ 39.229062] which lock already depends on the new lock. [ 39.229062] [ 39.237243] [ 39.237243] the existing dependency chain (in reverse order) is: [ 39.244729] [ 39.244729] -> #1 (&msm_obj->lock){+.+.}-{3:3}: [ 39.250769] mutex_lock_nested+0x1c/0x24 [ 39.255220] msm_gem_get_and_pin_iova_range+0x38/0x170 [ 39.260885] msm_gem_get_and_pin_iova+0x24/0x2c [ 39.265946] msm_ioctl_gem_submit+0x12c0/0x1964 [ 39.271008] drm_ioctl+0x1f0/0x3e0 [ 39.274940] sys_ioctl+0x3bc/0xbe0 [ 39.278873] ret_fast_syscall+0x0/0x28 [ 39.283150] 0xbed4deac [ 39.286122] [ 39.286122] -> #0 (reservation_ww_class_mutex){+.+.}-{3:3}: [ 39.293208] lock_acquire+0x134/0x560 [ 39.297398] __ww_mutex_lock.constprop.0+0xbc/0x1124 [ 39.302889] ww_mutex_lock+0x68/0xb4 [ 39.306993] dma_buf_detach+0xd0/0x1e0 [ 39.311271] drm_prime_gem_destroy+0x2c/0x38 [ 39.316068] msm_gem_free_work+0xa0/0x1b0 [ 39.320611] process_one_work+0x2a8/0x7ec [ 39.325149] worker_thread+0x4c/0x554 [ 39.329339] kthread+0x140/0x154 [ 39.333095] ret_from_fork+0x14/0x20 [ 39.337196] 0x0 [ 39.339560] [ 39.339560] other info that might help us debug this: [ 39.339560] [ 39.347568] Possible unsafe locking scenario: [ 39.347568] [ 39.353491] CPU0 CPU1 [ 39.358025] ---- ---- [ 39.362557] lock(&msm_obj->lock); [ 39.366061] lock(reservation_ww_class_mutex); [ 39.373123] lock(&msm_obj->lock); [ 39.379143] lock(reservation_ww_class_mutex); [ 39.383689] [ 39.383689] *** DEADLOCK *** [ 39.383689] [ 39.389614] 4 locks held by kworker/u2:1/68: [ 39.393888] #0: c24138a8 ((wq_completion)msm){+.+.}-{0:0}, at: process_one_work+0x1f4/0x7ec [ 39.402362] #1: c26a1f20 ((work_completion)(&priv->free_work)){+.+.}-{0:0}, at: process_one_work+0x1f4/0x7ec [ 39.412312] #2: c24188a4 (&dev->struct_mutex){+.+.}-{3:3}, at: msm_gem_free_work+0x68/0x1b0 [ 39.420784] #3: c2b957b0 (&msm_obj->lock){+.+.}-{3:3}, at: msm_gem_free_work+0x118/0x1b0 [ 39.428996] [ 39.428996] stack backtrace: [ 39.433363] CPU: 0 PID: 68 Comm: kworker/u2:1 Not tainted 5.10.31+g7ae1de1d2bd3 #1 [ 39.440937] Hardware name: Freescale i.MX53 (Device Tree Support) [ 39.447044] Workqueue: msm msm_gem_free_work [ 39.451346] [<c0111258>] (unwind_backtrace) from [<c010ba00>] (show_stack+0x10/0x14) [ 39.459107] [<c010ba00>] (show_stack) from [<c0e2b560>] (dump_stack+0xdc/0x104) [ 39.466430] [<c0e2b560>] (dump_stack) from [<c01838d8>] (check_noncircular+0xf0/0x160) [ 39.474358] [<c01838d8>] (check_noncircular) from [<c0187060>] (__lock_acquire+0x1628/0x2e04) [ 39.482893] [<c0187060>] (__lock_acquire) from [<c01892e8>] (lock_acquire+0x134/0x560) [ 39.490821] [<c01892e8>] (lock_acquire) from [<c0e3da60>] (__ww_mutex_lock.constprop.0+0xbc/0x1124) [ 39.499877] [<c0e3da60>] (__ww_mutex_lock.constprop.0) from [<c0e3ebe4>] (ww_mutex_lock+0x68/0xb4) [ 39.508844] [<c0e3ebe4>] (ww_mutex_lock) from [<c080e6a8>] (dma_buf_detach+0xd0/0x1e0) [ 39.516771] [<c080e6a8>] (dma_buf_detach) from [<c0701d04>] (drm_prime_gem_destroy+0x2c/0x38) [ 39.525305] [<c0701d04>] (drm_prime_gem_destroy) from [<c077d334>] (msm_gem_free_work+0xa0/0x1b0) [ 39.534188] [<c077d334>] (msm_gem_free_work) from [<c0147588>] (process_one_work+0x2a8/0x7ec) [ 39.542725] [<c0147588>] (process_one_work) from [<c0147b18>] (worker_thread+0x4c/0x554) [ 39.550825] [<c0147b18>] (worker_thread) from [<c014ebb0>] (kthread+0x140/0x154) [ 39.558229] [<c014ebb0>] (kthread) from [<c0100134>] (ret_from_fork+0x14/0x20) [ 39.565457] Exception stack(0xc26a1fb0 to 0xc26a1ff8) [ 39.570517] 1fa0: 00000000 00000000 00000000 00000000 [ 39.578703] 1fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 [ 39.586890] 1fe0: 00000000 00000000 00000000 00000000 00000013 00000000 This looks like a regression as it used to run fine on 5.4 Linux kernel. Any idea of patch reference to test is welcome. -- Otavio Salvador O.S. Systems http://www.ossystems.com.br http://code.ossystems.com.br Mobile: +55 (53) 9 9981-7854 Mobile: +1 (347) 903-9750

2 2

[PATCH 0/9] drm + usb-type-c: Add support for out-of-band hotplug notification
by Hans de Goede 04 May '21

04 May '21

Hi All, Here is a new attempt to make DP over Type-C work on devices where the Type-C controller does not drive the HPD pin on the GPU, but instead we need to forward HPD events from the Type-C controller to the DRM driver. For anyone interested here are the old (2019!) patches for this: https://patchwork.freedesktop.org/patch/288491/ https://patchwork.freedesktop.org/patch/288493/ https://patchwork.freedesktop.org/patch/288495/ Last time I posted this the biggest change requested was for more info to be included in the event send to the DRM-subsystem, specifically sending the following info was requested: 1. Which DP connector on the GPU the event is for 2. How many lanes are available 3. Connector orientation This series is basically an entirely new approach, which no longer uses the notifier framework at all. Instead the Type-C code looksup a connector based on a fwnode (this was suggested by Heikki Krogerus) and then calls a new oob_hotplug_event drm_connector_func directly on the connector, passing the requested info as argument. This series not only touches drm subsys files but it also touches drivers/usb/typec/altmodes/typec_displayport.c, that file usually does not see a whole lot of changes. So I believe it would be best to just merge the entire series through drm-misc, Assuming we can get an ack from Greg for merging the typec_displayport.c changes this way. Regards, Hans Hans de Goede (8): drm/connector: Make the drm_sysfs connector->kdev device hold a reference to the connector drm/connector: Add a fwnode pointer to drm_connector and register with ACPI drm/connector: Add drm_connector_find_by_fwnode() function drm/connector: Add support for out-of-band hotplug notification drm/i915/dp: Add support for out-of-bound hotplug events usb: typec: altmodes/displayport: Make dp_altmode_notify() more generic usb: typec: altmodes/displayport: Notify drm subsys of hotplug events platform/x86/intel_cht_int33fe: Correct "displayport" fwnode reference Heikki Krogerus (1): drm/i915: Associate ACPI connector nodes with connector entries drivers/gpu/drm/drm_connector.c | 20 +++ drivers/gpu/drm/drm_sysfs.c | 129 ++++++++++++++++-- drivers/gpu/drm/i915/display/intel_acpi.c | 40 ++++++ drivers/gpu/drm/i915/display/intel_acpi.h | 3 + drivers/gpu/drm/i915/display/intel_display.c | 1 + drivers/gpu/drm/i915/display/intel_dp.c | 13 ++ .../platform/x86/intel_cht_int33fe_typec.c | 4 +- drivers/usb/typec/altmodes/displayport.c | 78 ++++++++--- include/drm/drm_connector.h | 36 +++++ 9 files changed, 292 insertions(+), 32 deletions(-) -- 2.31.1

5 26

[PATCH resend 0/2] drm/i915/display: Make vlv_find_free_pps() skip pipes which are in use for non DP purposes
by Hans de Goede 04 May '21

04 May '21

Hi All, Here is a resend of my patch-set to deal with an "transcoder A assertion failure (expected off, current on)" error + WARN (and backtrace) seen on some Bay Trail devices with a DSI panel. I've rebased it on the latest drm-intel-next, so this time around the CI should be able to actually apply and test it, please review. Regards, Hans Hans de Goede (2): drm/i915/display: Add a intel_pipe_is_enabled() helper drm/i915/display: Make vlv_find_free_pps() skip pipes which are in use for non DP purposes drivers/gpu/drm/i915/display/intel_display.c | 20 ++++++++++++------ drivers/gpu/drm/i915/display/intel_display.h | 2 ++ drivers/gpu/drm/i915/display/intel_pps.c | 22 ++++++++++++++++++++ 3 files changed, 38 insertions(+), 6 deletions(-) -- 2.30.1

3 10

[PATCH 0/6] Workaround building improvements
by Tvrtko Ursulin 04 May '21

04 May '21

From: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Some cleanups and improvements to checks being done when building workaround lists. First five patches are small cleanups while the last one contains the actual story of what gets improved. Test-with: 20210429084130.850426-1-tvrtko.ursulin(a)linux.intel.com Tvrtko Ursulin (6): drm/i915: Drop duplicate WaDisable4x2SubspanOptimization:hsw drm/i915/debugfs: Expose read mask in i915_wa_registers drm/i915: Add a separate low-level helper for masked workarounds drm/i915/icl: Use appropriate helper for a masked workaround drm/i915/icl: Stop conflating mask and readback verify drm/i915: Add more checks when building workaround lists drivers/gpu/drm/i915/gt/intel_workarounds.c | 163 +++++++++++++----- .../gpu/drm/i915/gt/intel_workarounds_types.h | 4 + .../gpu/drm/i915/gt/selftest_workarounds.c | 4 +- drivers/gpu/drm/i915/i915_debugfs.c | 4 +- 4 files changed, 124 insertions(+), 51 deletions(-) -- 2.30.2

2 10

[PATCH v8 0/1] drm/drm_mst: Use Extended Base Receiver Capability
by Nikola Cornij 04 May '21

04 May '21

Change history: v8: - Chaged link lanes and rate parameters to u8 v7: - Fixed formatting - Fixed 'unused variable' compile warning - Fixed comment format v6: - Submited from (hopefully) the correct repo to fix build error v5: - Fixed min_t() macro arguments v4: - Fixed drm/radeon/ lane count and rate v3: - Fixed check-patch errors v2: - No changes, this was my mistaken reply to my patch v1: - Initial revision Nikola Cornij (1): drm/drm_mst: Use Extended Base Receiver Capability DPCD space .../display/amdgpu_dm/amdgpu_dm_mst_types.c | 5 +++ .../gpu/drm/amd/display/dc/core/dc_link_dp.c | 18 ++++++++++ drivers/gpu/drm/amd/display/dc/dc_link.h | 2 ++ drivers/gpu/drm/drm_dp_mst_topology.c | 33 ++++++++++++------- drivers/gpu/drm/i915/display/intel_dp_mst.c | 6 +++- drivers/gpu/drm/nouveau/dispnv50/disp.c | 3 +- drivers/gpu/drm/radeon/radeon_dp_mst.c | 7 ++++ include/drm/drm_dp_mst_helper.h | 12 ++++++- 8 files changed, 71 insertions(+), 15 deletions(-) -- 2.25.1

5 7

[PATCH v3 01/11] drm/atomic: Pass the full state to planes async atomic check and update
by Maxime Ripard 04 May '21

04 May '21

The current atomic helpers have either their object state being passed as an argument or the full atomic state. The former is the pattern that was done at first, before switching to the latter for new hooks or when it was needed. Let's start convert all the remaining helpers to provide a consistent interface, starting with the planes atomic_async_check and atomic_async_update. The conversion was done using the coccinelle script below, built tested on all the drivers. @@ identifier plane, plane_state; symbol state; @@ struct drm_plane_helper_funcs { ... int (*atomic_async_check)(struct drm_plane *plane, - struct drm_plane_state *plane_state); + struct drm_atomic_state *state); ... } @@ identifier plane, plane_state; symbol state; @@ struct drm_plane_helper_funcs { ... void (*atomic_async_update)(struct drm_plane *plane, - struct drm_plane_state *plane_state); + struct drm_atomic_state *state); ... } @ plane_atomic_func @ identifier helpers; identifier func; @@ ( static const struct drm_plane_helper_funcs helpers = { ..., .atomic_async_check = func, ..., }; | static const struct drm_plane_helper_funcs helpers = { ..., .atomic_async_update = func, ..., }; ) @@ struct drm_plane_helper_funcs *FUNCS; identifier f; identifier dev; identifier plane, plane_state, state; @@ f(struct drm_device *dev, struct drm_atomic_state *state) { <+... - FUNCS->atomic_async_check(plane, plane_state) + FUNCS->atomic_async_check(plane, state) ...+> } @@ struct drm_plane_helper_funcs *FUNCS; identifier f; identifier dev; identifier plane, plane_state, state; @@ f(struct drm_device *dev, struct drm_atomic_state *state) { <+... - FUNCS->atomic_async_update(plane, plane_state) + FUNCS->atomic_async_update(plane, state) ...+> } @@ identifier mtk_plane_atomic_async_update; identifier plane; symbol new_state, state; expression e; @@ void mtk_plane_atomic_async_update(struct drm_plane *plane, struct drm_plane_state *new_state) { ... - struct mtk_plane_state *state = e; + struct mtk_plane_state *new_plane_state = e; <+... - state + new_plane_state ...+> } @@ identifier plane_atomic_func.func; identifier plane; symbol state; @@ func(struct drm_plane *plane, - struct drm_plane_state *state) + struct drm_plane_state *new_plane_state) { <... - state + new_plane_state ...> } @ ignores_new_state @ identifier plane_atomic_func.func; identifier plane, new_plane_state; @@ func(struct drm_plane *plane, struct drm_plane_state *new_plane_state) { ... when != new_plane_state } @ adds_new_state depends on plane_atomic_func && !ignores_new_state @ identifier plane_atomic_func.func; identifier plane, new_plane_state; @@ func(struct drm_plane *plane, struct drm_plane_state *new_plane_state) { + struct drm_plane_state *new_plane_state = drm_atomic_get_new_plane_state(state, plane); ... } @ depends on plane_atomic_func @ identifier plane_atomic_func.func; identifier plane, plane_state; @@ func(struct drm_plane *plane, - struct drm_plane_state *plane_state + struct drm_atomic_state *state ) { ... } @ include depends on adds_new_state @ @@ #include <drm/drm_atomic.h> @ no_include depends on !include && adds_new_state @ @@ + #include <drm/drm_atomic.h> #include <drm/...> @@ identifier plane_atomic_func.func; identifier plane, state; identifier plane_state; @@ func(struct drm_plane *plane, struct drm_atomic_state *state) { ... struct drm_plane_state *plane_state = drm_atomic_get_new_plane_state(state, plane); <+... - plane_state->state + state ...+> } Acked-by: Thomas Zimmermann <tzimmermann(a)suse.de> Signed-off-by: Maxime Ripard <maxime(a)cerno.tech> --- Changes from v1: - Updated the comment according to Thomas suggestions --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 8 ++- drivers/gpu/drm/drm_atomic_helper.c | 4 +- drivers/gpu/drm/mediatek/mtk_drm_plane.c | 26 +++++---- drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c | 33 ++++++----- drivers/gpu/drm/rockchip/rockchip_drm_vop.c | 16 ++++-- drivers/gpu/drm/vc4/vc4_plane.c | 56 ++++++++++--------- include/drm/drm_modeset_helper_vtables.h | 18 +++--- 7 files changed, 89 insertions(+), 72 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 6ed96633425f..63f839679a0a 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -6468,7 +6468,7 @@ static int dm_plane_atomic_check(struct drm_plane *plane, } static int dm_plane_atomic_async_check(struct drm_plane *plane, - struct drm_plane_state *new_plane_state) + struct drm_atomic_state *state) { /* Only support async updates on cursor planes. */ if (plane->type != DRM_PLANE_TYPE_CURSOR) @@ -6478,10 +6478,12 @@ static int dm_plane_atomic_async_check(struct drm_plane *plane, } static void dm_plane_atomic_async_update(struct drm_plane *plane, - struct drm_plane_state *new_state) + struct drm_atomic_state *state) { + struct drm_plane_state *new_state = drm_atomic_get_new_plane_state(state, + plane); struct drm_plane_state *old_state = - drm_atomic_get_old_plane_state(new_state->state, plane); + drm_atomic_get_old_plane_state(state, plane); trace_amdgpu_dm_atomic_update_cursor(new_state); diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c index 9fa3f97223a1..bd0bb74437c7 100644 --- a/drivers/gpu/drm/drm_atomic_helper.c +++ b/drivers/gpu/drm/drm_atomic_helper.c @@ -1742,7 +1742,7 @@ int drm_atomic_helper_async_check(struct drm_device *dev, return -EBUSY; } - return funcs->atomic_async_check(plane, new_plane_state); + return funcs->atomic_async_check(plane, state); } EXPORT_SYMBOL(drm_atomic_helper_async_check); @@ -1772,7 +1772,7 @@ void drm_atomic_helper_async_commit(struct drm_device *dev, struct drm_framebuffer *old_fb = plane->state->fb; funcs = plane->helper_private; - funcs->atomic_async_update(plane, plane_state); + funcs->atomic_async_update(plane, state); /* * ->atomic_async_update() is supposed to update the diff --git a/drivers/gpu/drm/mediatek/mtk_drm_plane.c b/drivers/gpu/drm/mediatek/mtk_drm_plane.c index 92141a19681b..30ebcfd8832f 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_plane.c +++ b/drivers/gpu/drm/mediatek/mtk_drm_plane.c @@ -77,12 +77,14 @@ static void mtk_drm_plane_destroy_state(struct drm_plane *plane, } static int mtk_plane_atomic_async_check(struct drm_plane *plane, - struct drm_plane_state *state) + struct drm_atomic_state *state) { + struct drm_plane_state *new_plane_state = drm_atomic_get_new_plane_state(state, + plane); struct drm_crtc_state *crtc_state; int ret; - if (plane != state->crtc->cursor) + if (plane != new_plane_state->crtc->cursor) return -EINVAL; if (!plane->state) @@ -91,16 +93,16 @@ static int mtk_plane_atomic_async_check(struct drm_plane *plane, if (!plane->state->fb) return -EINVAL; - ret = mtk_drm_crtc_plane_check(state->crtc, plane, - to_mtk_plane_state(state)); + ret = mtk_drm_crtc_plane_check(new_plane_state->crtc, plane, + to_mtk_plane_state(new_plane_state)); if (ret) return ret; - if (state->state) - crtc_state = drm_atomic_get_existing_crtc_state(state->state, - state->crtc); + if (state) + crtc_state = drm_atomic_get_existing_crtc_state(state, + new_plane_state->crtc); else /* Special case for asynchronous cursor updates. */ - crtc_state = state->crtc->state; + crtc_state = new_plane_state->crtc->state; return drm_atomic_helper_check_plane_state(plane->state, crtc_state, DRM_PLANE_HELPER_NO_SCALING, @@ -109,9 +111,11 @@ static int mtk_plane_atomic_async_check(struct drm_plane *plane, } static void mtk_plane_atomic_async_update(struct drm_plane *plane, - struct drm_plane_state *new_state) + struct drm_atomic_state *state) { - struct mtk_plane_state *state = to_mtk_plane_state(plane->state); + struct drm_plane_state *new_state = drm_atomic_get_new_plane_state(state, + plane); + struct mtk_plane_state *new_plane_state = to_mtk_plane_state(plane->state); plane->state->crtc_x = new_state->crtc_x; plane->state->crtc_y = new_state->crtc_y; @@ -122,7 +126,7 @@ static void mtk_plane_atomic_async_update(struct drm_plane *plane, plane->state->src_h = new_state->src_h; plane->state->src_w = new_state->src_w; swap(plane->state->fb, new_state->fb); - state->pending.async_dirty = true; + new_plane_state->pending.async_dirty = true; mtk_drm_crtc_async_update(new_state->crtc, plane, new_state); } diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c index 83423092de2f..05fa8255caeb 100644 --- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c +++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c @@ -5,6 +5,7 @@ * Author: Rob Clark <robdclark(a)gmail.com> */ +#include <drm/drm_atomic.h> #include <drm/drm_damage_helper.h> #include <drm/drm_fourcc.h> #include <drm/drm_print.h> @@ -438,41 +439,43 @@ static void mdp5_plane_atomic_update(struct drm_plane *plane, } static int mdp5_plane_atomic_async_check(struct drm_plane *plane, - struct drm_plane_state *state) + struct drm_atomic_state *state) { - struct mdp5_plane_state *mdp5_state = to_mdp5_plane_state(state); + struct drm_plane_state *new_plane_state = drm_atomic_get_new_plane_state(state, + plane); + struct mdp5_plane_state *mdp5_state = to_mdp5_plane_state(new_plane_state); struct drm_crtc_state *crtc_state; int min_scale, max_scale; int ret; - crtc_state = drm_atomic_get_existing_crtc_state(state->state, - state->crtc); + crtc_state = drm_atomic_get_existing_crtc_state(state, + new_plane_state->crtc); if (WARN_ON(!crtc_state)) return -EINVAL; if (!crtc_state->active) return -EINVAL; - mdp5_state = to_mdp5_plane_state(state); + mdp5_state = to_mdp5_plane_state(new_plane_state); /* don't use fast path if we don't have a hwpipe allocated yet */ if (!mdp5_state->hwpipe) return -EINVAL; /* only allow changing of position(crtc x/y or src x/y) in fast path */ - if (plane->state->crtc != state->crtc || - plane->state->src_w != state->src_w || - plane->state->src_h != state->src_h || - plane->state->crtc_w != state->crtc_w || - plane->state->crtc_h != state->crtc_h || + if (plane->state->crtc != new_plane_state->crtc || + plane->state->src_w != new_plane_state->src_w || + plane->state->src_h != new_plane_state->src_h || + plane->state->crtc_w != new_plane_state->crtc_w || + plane->state->crtc_h != new_plane_state->crtc_h || !plane->state->fb || - plane->state->fb != state->fb) + plane->state->fb != new_plane_state->fb) return -EINVAL; min_scale = FRAC_16_16(1, 8); max_scale = FRAC_16_16(8, 1); - ret = drm_atomic_helper_check_plane_state(state, crtc_state, + ret = drm_atomic_helper_check_plane_state(new_plane_state, crtc_state, min_scale, max_scale, true, true); if (ret) @@ -485,15 +488,17 @@ static int mdp5_plane_atomic_async_check(struct drm_plane *plane, * also assign/unassign the hwpipe(s) tied to the plane. We avoid * taking the fast path for both these reasons. */ - if (state->visible != plane->state->visible) + if (new_plane_state->visible != plane->state->visible) return -EINVAL; return 0; } static void mdp5_plane_atomic_async_update(struct drm_plane *plane, - struct drm_plane_state *new_state) + struct drm_atomic_state *state) { + struct drm_plane_state *new_state = drm_atomic_get_new_plane_state(state, + plane); struct drm_framebuffer *old_fb = plane->state->fb; plane->state->src_x = new_state->src_x; diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c index 8d15cabdcb02..fefeab73ca27 100644 --- a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c +++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c @@ -1021,8 +1021,10 @@ static void vop_plane_atomic_update(struct drm_plane *plane, } static int vop_plane_atomic_async_check(struct drm_plane *plane, - struct drm_plane_state *state) + struct drm_atomic_state *state) { + struct drm_plane_state *new_plane_state = drm_atomic_get_new_plane_state(state, + plane); struct vop_win *vop_win = to_vop_win(plane); const struct vop_win_data *win = vop_win->data; int min_scale = win->phy->scl ? FRAC_16_16(1, 8) : @@ -1031,7 +1033,7 @@ static int vop_plane_atomic_async_check(struct drm_plane *plane, DRM_PLANE_HELPER_NO_SCALING; struct drm_crtc_state *crtc_state; - if (plane != state->crtc->cursor) + if (plane != new_plane_state->crtc->cursor) return -EINVAL; if (!plane->state) @@ -1040,9 +1042,9 @@ static int vop_plane_atomic_async_check(struct drm_plane *plane, if (!plane->state->fb) return -EINVAL; - if (state->state) - crtc_state = drm_atomic_get_existing_crtc_state(state->state, - state->crtc); + if (state) + crtc_state = drm_atomic_get_existing_crtc_state(state, + new_plane_state->crtc); else /* Special case for asynchronous cursor updates. */ crtc_state = plane->crtc->state; @@ -1052,8 +1054,10 @@ static int vop_plane_atomic_async_check(struct drm_plane *plane, } static void vop_plane_atomic_async_update(struct drm_plane *plane, - struct drm_plane_state *new_state) + struct drm_atomic_state *state) { + struct drm_plane_state *new_state = drm_atomic_get_new_plane_state(state, + plane); struct vop *vop = to_vop(plane->state->crtc); struct drm_framebuffer *old_fb = plane->state->fb; diff --git a/drivers/gpu/drm/vc4/vc4_plane.c b/drivers/gpu/drm/vc4/vc4_plane.c index 6bd8260aa9f2..ff485e58d271 100644 --- a/drivers/gpu/drm/vc4/vc4_plane.c +++ b/drivers/gpu/drm/vc4/vc4_plane.c @@ -1118,32 +1118,34 @@ void vc4_plane_async_set_fb(struct drm_plane *plane, struct drm_framebuffer *fb) } static void vc4_plane_atomic_async_update(struct drm_plane *plane, - struct drm_plane_state *state) + struct drm_atomic_state *state) { + struct drm_plane_state *new_plane_state = drm_atomic_get_new_plane_state(state, + plane); struct vc4_plane_state *vc4_state, *new_vc4_state; - swap(plane->state->fb, state->fb); - plane->state->crtc_x = state->crtc_x; - plane->state->crtc_y = state->crtc_y; - plane->state->crtc_w = state->crtc_w; - plane->state->crtc_h = state->crtc_h; - plane->state->src_x = state->src_x; - plane->state->src_y = state->src_y; - plane->state->src_w = state->src_w; - plane->state->src_h = state->src_h; - plane->state->src_h = state->src_h; - plane->state->alpha = state->alpha; - plane->state->pixel_blend_mode = state->pixel_blend_mode; - plane->state->rotation = state->rotation; - plane->state->zpos = state->zpos; - plane->state->normalized_zpos = state->normalized_zpos; - plane->state->color_encoding = state->color_encoding; - plane->state->color_range = state->color_range; - plane->state->src = state->src; - plane->state->dst = state->dst; - plane->state->visible = state->visible; + swap(plane->state->fb, new_plane_state->fb); + plane->state->crtc_x = new_plane_state->crtc_x; + plane->state->crtc_y = new_plane_state->crtc_y; + plane->state->crtc_w = new_plane_state->crtc_w; + plane->state->crtc_h = new_plane_state->crtc_h; + plane->state->src_x = new_plane_state->src_x; + plane->state->src_y = new_plane_state->src_y; + plane->state->src_w = new_plane_state->src_w; + plane->state->src_h = new_plane_state->src_h; + plane->state->src_h = new_plane_state->src_h; + plane->state->alpha = new_plane_state->alpha; + plane->state->pixel_blend_mode = new_plane_state->pixel_blend_mode; + plane->state->rotation = new_plane_state->rotation; + plane->state->zpos = new_plane_state->zpos; + plane->state->normalized_zpos = new_plane_state->normalized_zpos; + plane->state->color_encoding = new_plane_state->color_encoding; + plane->state->color_range = new_plane_state->color_range; + plane->state->src = new_plane_state->src; + plane->state->dst = new_plane_state->dst; + plane->state->visible = new_plane_state->visible; - new_vc4_state = to_vc4_plane_state(state); + new_vc4_state = to_vc4_plane_state(new_plane_state); vc4_state = to_vc4_plane_state(plane->state); vc4_state->crtc_x = new_vc4_state->crtc_x; @@ -1187,23 +1189,25 @@ static void vc4_plane_atomic_async_update(struct drm_plane *plane, } static int vc4_plane_atomic_async_check(struct drm_plane *plane, - struct drm_plane_state *state) + struct drm_atomic_state *state) { + struct drm_plane_state *new_plane_state = drm_atomic_get_new_plane_state(state, + plane); struct vc4_plane_state *old_vc4_state, *new_vc4_state; int ret; u32 i; - ret = vc4_plane_mode_set(plane, state); + ret = vc4_plane_mode_set(plane, new_plane_state); if (ret) return ret; old_vc4_state = to_vc4_plane_state(plane->state); - new_vc4_state = to_vc4_plane_state(state); + new_vc4_state = to_vc4_plane_state(new_plane_state); if (old_vc4_state->dlist_count != new_vc4_state->dlist_count || old_vc4_state->pos0_offset != new_vc4_state->pos0_offset || old_vc4_state->pos2_offset != new_vc4_state->pos2_offset || old_vc4_state->ptr0_offset != new_vc4_state->ptr0_offset || - vc4_lbm_size(plane->state) != vc4_lbm_size(state)) + vc4_lbm_size(plane->state) != vc4_lbm_size(new_plane_state)) return -EINVAL; /* Only pos0, pos2 and ptr0 DWORDS can be updated in an async update diff --git a/include/drm/drm_modeset_helper_vtables.h b/include/drm/drm_modeset_helper_vtables.h index eb706342861d..2b4d085da450 100644 --- a/include/drm/drm_modeset_helper_vtables.h +++ b/include/drm/drm_modeset_helper_vtables.h @@ -1292,9 +1292,9 @@ struct drm_plane_helper_funcs { /** * @atomic_async_check: * - * Drivers should set this function pointer to check if the plane state - * can be updated in a async fashion. Here async means "not vblank - * synchronized". + * Drivers should set this function pointer to check if the plane's + * atomic state can be updated in a async fashion. Here async means + * "not vblank synchronized". * * This hook is called by drm_atomic_async_check() to establish if a * given update can be committed asynchronously, that is, if it can @@ -1306,7 +1306,7 @@ struct drm_plane_helper_funcs { * can not be applied in asynchronous manner. */ int (*atomic_async_check)(struct drm_plane *plane, - struct drm_plane_state *state); + struct drm_atomic_state *state); /** * @atomic_async_update: @@ -1322,11 +1322,9 @@ struct drm_plane_helper_funcs { * update won't happen if there is an outstanding commit modifying * the same plane. * - * Note that unlike &drm_plane_helper_funcs.atomic_update this hook - * takes the new &drm_plane_state as parameter. When doing async_update - * drivers shouldn't replace the &drm_plane_state but update the - * current one with the new plane configurations in the new - * plane_state. + * When doing async_update drivers shouldn't replace the + * &drm_plane_state but update the current one with the new plane + * configurations in the new plane_state. * * Drivers should also swap the framebuffers between current plane * state (&drm_plane.state) and new_state. @@ -1345,7 +1343,7 @@ struct drm_plane_helper_funcs { * for deferring if needed, until a common solution is created. */ void (*atomic_async_update)(struct drm_plane *plane, - struct drm_plane_state *new_state); + struct drm_atomic_state *state); }; /** -- 2.29.2

5 25

[PATCH 00/21] drm/i915/gem: ioctl clean-ups
by Jason Ekstrand 04 May '21

04 May '21

Overview: --------- This patch series attempts to clean up some of the IOCTL mess we've created over the last few years. The most egregious bit being context mutability. In summary, this series: 1. Drops two never-used context params: RINGSIZE and NO_ZEROMAP 2. Drops the entire CONTEXT_CLONE API 3. Implements SINGLE_TIMELINE with a syncobj instead of actually sharing intel_timeline between engines. 4. Adds a few sanity restrictions to the balancing/bonding API. 5. Implements a proto-ctx mechanism so that the engine set and VM can only be set early on in the lifetime of a context, before anything ever executes on it. This effectively makes the VM and engine set immutable. This series has been tested with IGT as well as the Iris, ANV, and the Intel media driver doing an 8K decode (this uses bonding/balancing). I've also done quite a bit of git archeology to ensure that nothing in here will break anything that's already shipped at some point in history. It's possible I've missed something, but I've dug quite a bit. Details and motivation: ----------------------- In very broad strokes, there's an effort going on right now within Intel to try and clean up and simplify i915 anywhere we can. We obviously don't want to break any shipping userspace but, as can be seen by this series, there's a lot i915 theoretically supports which userspace doesn't actually need. Some of this, like the two context params used here, were simply oversights where we went through the usual API review process and merged the i915 bits but the userspace bits never landed for some reason. Not all are so innocent, however. For instance, there's an entire context cloning API which allows one to create a context with certain parameters "cloned" from some other context. This entire API has never been used by any userspace except IGT and there were never patches to any other userspace to use it. It never should have landed. Also, when we added support for setting explicit engine sets and sharing VMs across contexts, people decided to do so via SET_CONTEXT_PARAM. While this allowed them to re-use existing API, it did so at the cost of making those states mutable which leads to a plethora of potential race conditions. There were even IGT tests merged to cover some of theses: - gem_vm_create@async-destroy and gem_vm_create@destroy-race which test swapping out the VM on a running context. - gem_ctx_persistence@replace* which test whether a client can escape a non-persistent context by submitting a hanging batch and then swapping out the engine set before the hang is detected. - api_intel_bb@bb-with-vm which tests the that intel_bb_assign_vm works properly. This API is never used by any other IGT test. There is also an entire deferred flush and set state framework in i915_gem_cotnext.c which exists for safely swapping out the VM while there is work in-flight on a context. So, clearly people knew that this API was inherently racy and difficult to implement but they landed it anyway. Why? The best explanation I've been given is because it makes the API more "unified" or "symmetric" for this stuff to go through SET_CONTEXT_PARAM. It's not because any userspace actually wants to be able to swap out the VM or the set of engines on a running context. That would be utterly insane. This patch series cleans up this particular mess by introducing the concept of a i915_gem_proto_context data structure which contains context creation information. When you initially call GEM_CONTEXT_CREATE, a proto-context in created instead of an actual context. Then, the first time something is done on the context besides SET_CONTEXT_PARAM, an actual context is created. This allows us to keep the old drivers which use SET_CONTEXT_PARAM to set up the engine set (see also media) while ensuring that, once you have an i915_gem_context, the VM and the engine set are immutable state. Eventually, there are more clean-ups I'd like to do on top of this which should make working with contexts inside i915 simpler and safer: 1. Move the GEM handle -> vma LUT from i915_gem_context into either i915_ppgtt or drm_i915_file_private depending on whether or not the hardware has a full PPGTT. 2. Move the delayed context destruction code into intel_context or a per-engine wrapper struct rather than i915_gem_context. 3. Get rid of the separation between context close and context destroy 4. Get rid of the RCU on i915_gem_context However, these should probably be done as a separate patch series as this one is already starting to get longish, especially if you consider the 89 IGT patches that go along with it. Test-with: 20210423214853.876911-1-jason(a)jlekstrand.net Jason Ekstrand (21): drm/i915: Drop I915_CONTEXT_PARAM_RINGSIZE drm/i915: Drop I915_CONTEXT_PARAM_NO_ZEROMAP drm/i915/gem: Set the watchdog timeout directly in intel_context_set_gem drm/i915/gem: Return void from context_apply_all drm/i915: Drop the CONTEXT_CLONE API drm/i915: Implement SINGLE_TIMELINE with a syncobj (v3) drm/i915: Drop getparam support for I915_CONTEXT_PARAM_ENGINES drm/i915/gem: Disallow bonding of virtual engines drm/i915/gem: Disallow creating contexts with too many engines drm/i915/request: Remove the hook from await_execution drm/i915: Stop manually RCU banging in reset_stats_ioctl drm/i915/gem: Add a separate validate_priority helper drm/i915/gem: Add an intermediate proto_context struct drm/i915/gem: Return an error ptr from context_lookup drm/i915/gt: Drop i915_address_space::file drm/i915/gem: Delay context creation drm/i915/gem: Don't allow changing the VM on running contexts drm/i915/gem: Don't allow changing the engine set on running contexts drm/i915/selftests: Take a VM in kernel_context() i915/gem/selftests: Assign the VM at context creation in igt_shared_ctx_exec drm/i915/gem: Roll all of context creation together drivers/gpu/drm/i915/Makefile | 1 - drivers/gpu/drm/i915/gem/i915_gem_context.c | 2967 +++++++---------- drivers/gpu/drm/i915/gem/i915_gem_context.h | 3 + .../gpu/drm/i915/gem/i915_gem_context_types.h | 68 +- .../gpu/drm/i915/gem/i915_gem_execbuffer.c | 31 +- .../drm/i915/gem/selftests/i915_gem_context.c | 127 +- .../gpu/drm/i915/gem/selftests/mock_context.c | 62 +- .../gpu/drm/i915/gem/selftests/mock_context.h | 4 +- drivers/gpu/drm/i915/gt/intel_context_param.c | 63 - drivers/gpu/drm/i915/gt/intel_context_param.h | 6 +- drivers/gpu/drm/i915/gt/intel_engine_types.h | 7 - .../drm/i915/gt/intel_execlists_submission.c | 100 - .../drm/i915/gt/intel_execlists_submission.h | 4 - drivers/gpu/drm/i915/gt/intel_gtt.h | 10 - drivers/gpu/drm/i915/gt/selftest_execlists.c | 249 +- drivers/gpu/drm/i915/gt/selftest_hangcheck.c | 2 +- drivers/gpu/drm/i915/i915_drv.h | 23 +- drivers/gpu/drm/i915/i915_perf.c | 4 +- drivers/gpu/drm/i915/i915_request.c | 42 +- drivers/gpu/drm/i915/i915_request.h | 4 +- .../drm/i915/selftests/i915_mock_selftests.h | 1 - drivers/gpu/drm/i915/selftests/mock_gtt.c | 1 - include/uapi/drm/i915_drm.h | 40 +- 23 files changed, 1438 insertions(+), 2381 deletions(-) delete mode 100644 drivers/gpu/drm/i915/gt/intel_context_param.c -- 2.31.1

5 109

[led-backlight] default-brightness-level issue
by pgeiem 04 May '21

04 May '21

Dear all, On a custom board I have a simple DPI panel. Panel's backlight is drive with an I2C led driver (PCA9632). led-backlight driver is sued to manage this as a backlight. When using brightness-levels and default-brightness-level the backlight stay turned-off even if manually trying to set a different index value to brightness through sysfs. I traced this issue as follow: When led_bl_update_status() is called the brightness value is returned from backlight_get_brightness() which call backlight_is_blank(). In my case backlight_is_blank() return true due to props.power = FB_BLANK_POWERDOWN which is != FB_BLANK_UNBLANK. I traced why at startup props.power is FB_BLANK_POWERDOWN and found that in led_bl_probe() when a default brightness is set (>0) props.power is set to FB_BLANK_POWERDOWN which seems incorrect to me. I made the small change below and default brightness is correctly used at startup. I am not really sure this is an issue and if my change is correct or if I am doing something incorrect somewhere else. So I first would like to get your opinion on this and if correct will send a patch. Many thanks in advance Patrick diff --git a/drivers/video/backlight/led_bl.c b/drivers/video/backlight/led_bl.c --- a/drivers/video/backlight/led_bl.c +++ b/drivers/video/backlight/led_bl.c @@ -200,7 +200,7 @@ static int led_bl_probe(struct platform_device *pdev) props.type = BACKLIGHT_RAW; props.max_brightness = priv->max_brightness; props.brightness = priv->default_brightness; - props.power = (priv->default_brightness > 0) ? FB_BLANK_POWERDOWN : + props.power = (priv->default_brightness == 0) ? FB_BLANK_POWERDOWN : FB_BLANK_UNBLANK; priv->bl_dev = backlight_device_register(dev_name(&pdev->dev), &pdev->dev, priv, &led_bl_ops, &props);

2 4

← Newer
1
...
16
17
18
19
20
21
22
...
72
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel April 2021