dri-devel December 2021

dri-devel@lists.freedesktop.org

310 participants
720 discussions

[PATCH v2 0/1] drm/vkms: zpos
by José Expósito 26 Dec '21

26 Dec '21

Hi all, This patch adds support for the zpos plane prop in the VKMS driver. It should be applied after the "drm/vkms: add support for multiple overlay planes" series [1] because this new patch takes advantage of the new constant "NUM_OVERLAY_PLANES" to set the maximum overlay plane zpos. Notice that, as explained in the commit message, there is one test failling (plane-immutable-zpos) because of the timeout capturing CRC. This is a known bug [2] that needs to be fixed in a different series. It happens when the primary plane is disabled. The vkms_composer.c vkms_composer_worker() function is not able to find the "primary_composer" variable and the test fails. As a quick fix, commenting the line: if (act_plane->base.base.plane->type == DRM_PLANE_TYPE_PRIMARY) Fixes the issue. However, more work is required to properly clear the background and blend the first active plane. I'll look into it, feedback from someone that already investigated the issue is welcome :) Thanks in advance, José Expósito [1] https://lore.kernel.org/dri-devel/20211226104059.11265-1-jose.exposito89@gm… [2] https://www.kernel.org/doc/html/latest/gpu/vkms.html#igt-better-support José Expósito (1): drm/vkms: add zpos plane property drivers/gpu/drm/vkms/vkms_crtc.c | 3 +-- drivers/gpu/drm/vkms/vkms_drv.c | 1 + drivers/gpu/drm/vkms/vkms_plane.c | 23 +++++++++++++++++++++++ 3 files changed, 25 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH v2 1/3] drm/vkms: refactor overlay plane creation
by José Expósito 26 Dec '21

26 Dec '21

Move the logic to create an overlay plane to its own function. Refactor, no functional changes. Signed-off-by: José Expósito <jose.exposito89(a)gmail.com> --- drivers/gpu/drm/vkms/vkms_output.c | 26 +++++++++++++++++++------- 1 file changed, 19 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/vkms/vkms_output.c b/drivers/gpu/drm/vkms/vkms_output.c index 04406bd3ff02..2e805b2d36ae 100644 --- a/drivers/gpu/drm/vkms/vkms_output.c +++ b/drivers/gpu/drm/vkms/vkms_output.c @@ -32,6 +32,21 @@ static const struct drm_connector_helper_funcs vkms_conn_helper_funcs = { .get_modes = vkms_conn_get_modes, }; +static int vkms_add_overlay_plane(struct vkms_device *vkmsdev, int index, + struct drm_crtc *crtc) +{ + struct vkms_plane *overlay; + + overlay = vkms_plane_init(vkmsdev, DRM_PLANE_TYPE_OVERLAY, index); + if (IS_ERR(overlay)) + return PTR_ERR(overlay); + + if (!overlay->base.possible_crtcs) + overlay->base.possible_crtcs = drm_crtc_mask(crtc); + + return 0; +} + int vkms_output_init(struct vkms_device *vkmsdev, int index) { struct vkms_output *output = &vkmsdev->output; @@ -39,7 +54,7 @@ int vkms_output_init(struct vkms_device *vkmsdev, int index) struct drm_connector *connector = &output->connector; struct drm_encoder *encoder = &output->encoder; struct drm_crtc *crtc = &output->crtc; - struct vkms_plane *primary, *cursor = NULL, *overlay = NULL; + struct vkms_plane *primary, *cursor = NULL; int ret; int writeback; @@ -48,12 +63,9 @@ int vkms_output_init(struct vkms_device *vkmsdev, int index) return PTR_ERR(primary); if (vkmsdev->config->overlay) { - overlay = vkms_plane_init(vkmsdev, DRM_PLANE_TYPE_OVERLAY, index); - if (IS_ERR(overlay)) - return PTR_ERR(overlay); - - if (!overlay->base.possible_crtcs) - overlay->base.possible_crtcs = drm_crtc_mask(crtc); + ret = vkms_add_overlay_plane(vkmsdev, index, crtc); + if (ret) + return ret; } if (vkmsdev->config->cursor) { -- 2.25.1

1 2

[PATCH] drm/panel: JDI LT070ME05000 drop broken link
by David Heidelberg 24 Dec '21

24 Dec '21

Link is no longer functional and web.archive.org doesn't provide PDF with detail information. Some informations can be found from web.archive.org here: https://web.archive.org/web/20170629205602/http://panelone.net/en/7-0-inch/… Signed-off-by: David Heidelberg <david(a)ixit.cz> --- drivers/gpu/drm/panel/panel-jdi-lt070me05000.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c b/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c index 3c86ad262d5e..ea0e7221e706 100644 --- a/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c +++ b/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c @@ -5,10 +5,6 @@ * * Copyright (C) 2016 Linaro Ltd * Author: Sumit Semwal <sumit.semwal(a)linaro.org> - * - * From internet archives, the panel for Nexus 7 2nd Gen, 2013 model is a - * JDI model LT070ME05000, and its data sheet is at: - * http://panelone.net/en/7-0-inch/JDI_LT070ME05000_7.0_inch-datasheet */ #include <linux/backlight.h> -- 2.34.1

1 0

[PATCH] drm/i915/guc: Log engine resets
by Tvrtko Ursulin 24 Dec '21

24 Dec '21

From: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Log engine resets done by the GuC firmware in the similar way it is done by the execlists backend. This way we have notion of where the hangs are before the GuC gains support for proper error capture. Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Cc: Matthew Brost <matthew.brost(a)intel.com> Cc: John Harrison <John.C.Harrison(a)Intel.com> --- drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c index 97311119da6f..51512123dc1a 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c @@ -11,6 +11,7 @@ #include "gt/intel_context.h" #include "gt/intel_engine_pm.h" #include "gt/intel_engine_heartbeat.h" +#include "gt/intel_engine_user.h" #include "gt/intel_gpu_commands.h" #include "gt/intel_gt.h" #include "gt/intel_gt_clock_utils.h" @@ -3934,9 +3935,18 @@ static void capture_error_state(struct intel_guc *guc, { struct intel_gt *gt = guc_to_gt(guc); struct drm_i915_private *i915 = gt->i915; - struct intel_engine_cs *engine = __context_to_physical_engine(ce); + struct intel_engine_cs *engine = ce->engine; intel_wakeref_t wakeref; + if (intel_engine_is_virtual(engine)) { + drm_notice(&i915->drm, "%s class, engines 0x%x; GuC engine reset\n", + intel_engine_class_repr(engine->class), + engine->mask); + engine = guc_virtual_get_sibling(engine, 0); + } else { + drm_notice(&i915->drm, "%s GuC engine reset\n", engine->name); + } + intel_engine_set_hung_context(engine, ce); with_intel_runtime_pm(&i915->runtime_pm, wakeref) i915_capture_error_state(gt, engine->mask); -- 2.32.0

3 12

[PATCH 0/3] drm/vkms: add support for multiple overlay planes
by José Expósito 24 Dec '21

24 Dec '21

Hi all, First of all, let me quickly introduce myself. I'm José Expósito and I'm a hobbyist open source developer. My contributions are usually in the kernel input/HID subsystems and in its userspace library (libinput) as well as other projects. I'm trying to learn more about the GPU and I found VKMS as a good project to get started. So, now that you have been warned about my (lack) of experience in this subsystem, let's get into the patches. The series adds support for multiple overlay planes by adding a new module parameter that allows to configure the number of overlay planes to add. I checked that the planes are properly created using drm_info[1] and also executed the IGT test "kms_plane" to make sure that the planes were listed in the output. In addition, I checked the vkms_composer and -even though I'm not sure how to properly test it- it looks like it is already prepared to compose an arbitrary number of planes. Having said that, I really hope I didn't make any obvious mistakes. I noticed that the docs say: > For all of these, we also want to review the igt test coverage and > make sure all relevant igt testcases work on vkms I ran some plane related tests, but some of them were already failing without my code, so I'd appreciate some help with this bit. Thank you very much in advance for your time. Any suggestions to improve this patchset or about what to work on next are very welcome. Jose [1] https://github.com/ascent12/drm_info José Expósito (3): drm/vkms: refactor overlay plane creation drm/vkms: add support for multiple overlay planes drm/vkms: drop "Multiple overlay planes" TODO Documentation/gpu/vkms.rst | 2 -- drivers/gpu/drm/vkms/vkms_drv.c | 5 +++++ drivers/gpu/drm/vkms/vkms_drv.h | 1 + drivers/gpu/drm/vkms/vkms_output.c | 29 ++++++++++++++++++++++------- 4 files changed, 28 insertions(+), 9 deletions(-) -- 2.25.1

2 6

Empty IN_FORMATS in sun4i-drm
by Emmanuel Gil Peyrot 24 Dec '21

24 Dec '21

Hi, After updating Weston from 9f8561e9 to 07326040 (latest master), it fails to run on my PinePhone saying “format 0x34325258 not supported by output DSI-1” and then exiting. This format is XR24, which would be extremely surprising not to be present, and drm_info[1] says it is present. Looking into Weston’s code, I found that drm_plane_populate_formats()’s docstring says it uses “either the IN_FORMATS blob property (if available), or the plane's format list if not.” Looking back at drm_info, I saw said IN_FORMATS blob being empty of formats (while the format list is fully populated): "IN_FORMATS" (immutable): blob = 32 └───DRM_FORMAT_MOD_LINEAR (0x0) This makes me think the kernel should populate IN_FORMATS with at least the same formats as the format list when supported, or stop advertising this property altogether. Other compositors (such as phoc) still run file, probably because they use the format list exclusively, without consideration for modifiers. Besides fixing this driver, would it make sense to also make Weston ignore an empty IN_FORMATS and fall back to the format list? Thanks, [1] https://github.com/ascent12/drm_info -- Emmanuel Gil Peyrot

2 3

[git pull] drm fixes for 5.16-rc7
by Dave Airlie 23 Dec '21

23 Dec '21

Hi Linus. Happy Xmas, pull request. Nothing major, one mediatek and a couple of i915 locking fixes. There might be a few stragglers over next week or so but I don't expect much before next release. Dave. drm-fixes-2021-12-24: drm fixes for 5.16-rc7 mediatek: - NULL pointer check i915: - guc submission locking fixes The following changes since commit a7904a538933c525096ca2ccde1e60d0ee62c08e: Linux 5.16-rc6 (2021-12-19 14:14:33 -0800) are available in the Git repository at: git://anongit.freedesktop.org/drm/drm tags/drm-fixes-2021-12-24 for you to fetch changes up to dbfba788c7ef839849e013264fef551e4d119d0b: Merge tag 'drm-intel-fixes-2021-12-22' of git://anongit.freedesktop.org/drm/drm-intel into drm-fixes (2021-12-23 11:21:44 +1000) ---------------------------------------------------------------- drm fixes for 5.16-rc7 mediatek: - NULL pointer check i915: - guc submission locking fixes ---------------------------------------------------------------- AngeloGioacchino Del Regno (1): drm/mediatek: hdmi: Perform NULL pointer check for mtk_hdmi_conf Dave Airlie (2): Merge tag 'mediatek-drm-fixes-5.16' of https://git.kernel.org/pub/scm/linux/kernel/git/chunkuang.hu/linux into drm-fixes Merge tag 'drm-intel-fixes-2021-12-22' of git://anongit.freedesktop.org/drm/drm-intel into drm-fixes Matthew Brost (2): drm/i915/guc: Use correct context lock when callig clr_context_registered drm/i915/guc: Only assign guc_id.id when stealing guc_id drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c | 6 +++--- drivers/gpu/drm/mediatek/mtk_hdmi.c | 12 +++++++----- 2 files changed, 10 insertions(+), 8 deletions(-)

2 1

[GIT PULL] etnaviv-next for 5.17
by Lucas Stach 23 Dec '21

23 Dec '21

Hi Dave, hi Daniel, please pull the following etnaviv changes for the next merge window: - make etnaviv work on IOMMU enabled systems - fix mapping of command buffers on systems with more than 4GB RAM - close a DoS vector - fix spurious GPU resets Regards, Lucas The following changes since commit fa55b7dcdc43c1aa1ba12bca9d2dd4318c2a0dbf: Linux 5.16-rc1 (2021-11-14 13:56:52 -0800) are available in the Git repository at: https://git.pengutronix.de/git/lst/linux etnaviv/next for you to fetch changes up to cdd156955f946beaa5f3a00d8ccf90e5a197becc: drm/etnaviv: consider completed fence seqno in hang check (2021-12-23 20:21:33 +0100) ---------------------------------------------------------------- Lucas Stach (2): drm/etnaviv: limit submit sizes drm/etnaviv: consider completed fence seqno in hang check Michael Walle (3): drm/etnaviv: use PLATFORM_DEVID_NONE drm/etnaviv: fix dma configuration of the virtual device drm/etnaviv: use a 32 bit mask as coherent DMA mask Rikard Falkeborn (1): drm/etnaviv: constify static struct cooling_ops drivers/gpu/drm/etnaviv/etnaviv_drv.c | 41 +++++++++++++++++++++++++++++++---------- drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c | 6 ++++++ drivers/gpu/drm/etnaviv/etnaviv_gpu.c | 2 +- drivers/gpu/drm/etnaviv/etnaviv_gpu.h | 1 + drivers/gpu/drm/etnaviv/etnaviv_sched.c | 4 +++- 5 files changed, 42 insertions(+), 12 deletions(-)

1 0

[PATCH] drm/i915/guc: Use lockless list for destroyed contexts
by Matthew Brost 23 Dec '21

23 Dec '21

Use a lockless list structure for destroyed contexts to avoid hammering on global submission spin lock. Suggested-by: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Signed-off-by: Matthew Brost <matthew.brost(a)intel.com> --- drivers/gpu/drm/i915/gt/intel_context.c | 2 - drivers/gpu/drm/i915/gt/intel_context_types.h | 3 +- drivers/gpu/drm/i915/gt/uc/intel_guc.h | 3 +- .../gpu/drm/i915/gt/uc/intel_guc_submission.c | 44 +++++-------------- 4 files changed, 16 insertions(+), 36 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_context.c b/drivers/gpu/drm/i915/gt/intel_context.c index 5d0ec7c49b6a..4aacb4b0418d 100644 --- a/drivers/gpu/drm/i915/gt/intel_context.c +++ b/drivers/gpu/drm/i915/gt/intel_context.c @@ -403,8 +403,6 @@ intel_context_init(struct intel_context *ce, struct intel_engine_cs *engine) ce->guc_id.id = GUC_INVALID_LRC_ID; INIT_LIST_HEAD(&ce->guc_id.link); - INIT_LIST_HEAD(&ce->destroyed_link); - INIT_LIST_HEAD(&ce->parallel.child_list); /* diff --git a/drivers/gpu/drm/i915/gt/intel_context_types.h b/drivers/gpu/drm/i915/gt/intel_context_types.h index 30cd81ad8911..4532d43ec9c0 100644 --- a/drivers/gpu/drm/i915/gt/intel_context_types.h +++ b/drivers/gpu/drm/i915/gt/intel_context_types.h @@ -9,6 +9,7 @@ #include <linux/average.h> #include <linux/kref.h> #include <linux/list.h> +#include <linux/llist.h> #include <linux/mutex.h> #include <linux/types.h> @@ -224,7 +225,7 @@ struct intel_context { * list when context is pending to be destroyed (deregistered with the * GuC), protected by guc->submission_state.lock */ - struct list_head destroyed_link; + struct llist_node destroyed_link; /** @parallel: sub-structure for parallel submission members */ struct { diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc.h b/drivers/gpu/drm/i915/gt/uc/intel_guc.h index f9240d4baa69..705085058411 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc.h +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc.h @@ -8,6 +8,7 @@ #include <linux/xarray.h> #include <linux/delay.h> +#include <linux/llist.h> #include "intel_uncore.h" #include "intel_guc_fw.h" @@ -112,7 +113,7 @@ struct intel_guc { * @destroyed_contexts: list of contexts waiting to be destroyed * (deregistered with the GuC) */ - struct list_head destroyed_contexts; + struct llist_head destroyed_contexts; /** * @destroyed_worker: worker to deregister contexts, need as we * need to take a GT PM reference and can't from destroy diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c index 0a03a30e4c6d..6f7643edc139 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c @@ -1771,7 +1771,7 @@ int intel_guc_submission_init(struct intel_guc *guc) spin_lock_init(&guc->submission_state.lock); INIT_LIST_HEAD(&guc->submission_state.guc_id_list); ida_init(&guc->submission_state.guc_ids); - INIT_LIST_HEAD(&guc->submission_state.destroyed_contexts); + init_llist_head(&guc->submission_state.destroyed_contexts); INIT_WORK(&guc->submission_state.destroyed_worker, destroyed_worker_func); @@ -2696,26 +2696,18 @@ static void __guc_context_destroy(struct intel_context *ce) } } +#define take_destroyed_contexts(guc) \ + llist_del_all(&guc->submission_state.destroyed_contexts) + static void guc_flush_destroyed_contexts(struct intel_guc *guc) { - struct intel_context *ce; - unsigned long flags; + struct intel_context *ce, *cn; GEM_BUG_ON(!submission_disabled(guc) && guc_submission_initialized(guc)); - while (!list_empty(&guc->submission_state.destroyed_contexts)) { - spin_lock_irqsave(&guc->submission_state.lock, flags); - ce = list_first_entry_or_null(&guc->submission_state.destroyed_contexts, - struct intel_context, - destroyed_link); - if (ce) - list_del_init(&ce->destroyed_link); - spin_unlock_irqrestore(&guc->submission_state.lock, flags); - - if (!ce) - break; - + llist_for_each_entry_safe(ce, cn, take_destroyed_contexts(guc), + destroyed_link) { release_guc_id(guc, ce); __guc_context_destroy(ce); } @@ -2723,23 +2715,11 @@ static void guc_flush_destroyed_contexts(struct intel_guc *guc) static void deregister_destroyed_contexts(struct intel_guc *guc) { - struct intel_context *ce; - unsigned long flags; - - while (!list_empty(&guc->submission_state.destroyed_contexts)) { - spin_lock_irqsave(&guc->submission_state.lock, flags); - ce = list_first_entry_or_null(&guc->submission_state.destroyed_contexts, - struct intel_context, - destroyed_link); - if (ce) - list_del_init(&ce->destroyed_link); - spin_unlock_irqrestore(&guc->submission_state.lock, flags); - - if (!ce) - break; + struct intel_context *ce, *cn; + llist_for_each_entry_safe(ce, cn, take_destroyed_contexts(guc), + destroyed_link) guc_lrc_desc_unpin(ce); - } } static void destroyed_worker_func(struct work_struct *w) @@ -2771,8 +2751,8 @@ static void guc_context_destroy(struct kref *kref) if (likely(!destroy)) { if (!list_empty(&ce->guc_id.link)) list_del_init(&ce->guc_id.link); - list_add_tail(&ce->destroyed_link, - &guc->submission_state.destroyed_contexts); + llist_add(&ce->destroyed_link, + &guc->submission_state.destroyed_contexts); } else { __release_guc_id(guc, ce); } -- 2.34.1

3 4

[PATCH v2] drm/i915/guc: Check for wedged before doing stuff
by John.C.Harrison＠Intel.com 23 Dec '21

23 Dec '21

From: John Harrison <John.C.Harrison(a)Intel.com> A fault injection probe test hit a BUG_ON in a GuC error path. It showed that the GuC code could potentially attempt to do many things when the device is actually wedged. So, add a check in to prevent that. v2: Use intel_gt_is_wedged instead of testing bits directly in the GuC submission code (review feedback from Tvrtko). Signed-off-by: John Harrison <John.C.Harrison(a)Intel.com> --- drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c index e7517206af82..756b29d8326b 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c @@ -1349,7 +1349,8 @@ submission_disabled(struct intel_guc *guc) struct i915_sched_engine * const sched_engine = guc->sched_engine; return unlikely(!sched_engine || - !__tasklet_is_enabled(&sched_engine->tasklet)); + !__tasklet_is_enabled(&sched_engine->tasklet) || + intel_gt_is_wedged(guc_to_gt(guc))); } static void disable_submission(struct intel_guc *guc) @@ -1725,7 +1726,7 @@ void intel_guc_submission_reset_finish(struct intel_guc *guc) { /* Reset called during driver load or during wedge? */ if (unlikely(!guc_submission_initialized(guc) || - test_bit(I915_WEDGED, &guc_to_gt(guc)->reset.flags))) { + intel_gt_is_wedged(guc_to_gt(guc)))) { return; } -- 2.25.1

2 1

← Newer
1
...
23
24
25
26
27
28
29
...
72
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel December 2021