dri-devel January 2022

dri-devel@lists.freedesktop.org

373 participants
842 discussions

[PATCH v3 0/5] drm/dp: Move DisplayPort helpers into own module
by Thomas Zimmermann 20 Jan '22

20 Jan '22

Split-off DisplayPort functions from KMS helper library and move them into their own module. Reduces the size of drm_kms_helper.ko by ~50%. This patchset is part of an on-going effort to reduce the minimum binary size of the DRM core and helpers. It's helpful for systems with early-boot DRM graphics, which requires DRM to be linked into the kernel image. v3: * fix include statement in DRM selftests * rebase v2: * move helper library into subdirectory (Jani) * also move DP AUX bus helpers Thomas Zimmermann (5): drm/dp_mst: Remove trailing whitespace. drm/dp: Move DP declarations into separate header file drm/dp: Move DisplayPort helpers into separate helper module drm/dp: Move public DisplayPort headers into dp/ drm/dp: Move DisplayPort AUX bus helpers into dp/ drivers/gpu/drm/Kconfig | 8 +++++ drivers/gpu/drm/Makefile | 12 +++---- .../gpu/drm/amd/amdgpu/amdgpu_connectors.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 4 +-- drivers/gpu/drm/amd/amdgpu/atombios_dp.c | 2 +- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 2 +- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 2 +- .../display/amdgpu_dm/amdgpu_dm_mst_types.c | 4 +-- .../drm/amd/display/dc/core/dc_link_dpcd.c | 2 +- drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c | 2 +- drivers/gpu/drm/amd/display/dc/os_types.h | 2 +- .../gpu/drm/amd/display/include/dpcd_defs.h | 2 +- .../gpu/drm/amd/display/modules/hdcp/hdcp.h | 2 +- drivers/gpu/drm/bridge/Kconfig | 4 +++ drivers/gpu/drm/bridge/analogix/Kconfig | 2 ++ .../drm/bridge/analogix/analogix-anx6345.c | 2 +- .../drm/bridge/analogix/analogix-anx78xx.c | 2 +- .../drm/bridge/analogix/analogix-i2c-dptx.c | 2 +- .../drm/bridge/analogix/analogix_dp_core.h | 2 +- drivers/gpu/drm/bridge/analogix/anx7625.c | 2 +- drivers/gpu/drm/bridge/cadence/Kconfig | 1 + .../drm/bridge/cadence/cdns-mhdp8546-core.c | 2 +- .../drm/bridge/cadence/cdns-mhdp8546-core.h | 2 +- drivers/gpu/drm/bridge/parade-ps8640.c | 4 +-- drivers/gpu/drm/bridge/tc358767.c | 2 +- drivers/gpu/drm/bridge/tc358775.c | 2 +- drivers/gpu/drm/bridge/ti-sn65dsi86.c | 4 +-- drivers/gpu/drm/dp/Makefile | 9 +++++ .../gpu/drm/{drm_dp_helper.c => dp/drm_dp.c} | 6 ++-- drivers/gpu/drm/{ => dp}/drm_dp_aux_bus.c | 4 +-- drivers/gpu/drm/{ => dp}/drm_dp_aux_dev.c | 6 ++-- drivers/gpu/drm/{ => dp}/drm_dp_cec.c | 2 +- .../drm/{ => dp}/drm_dp_dual_mode_helper.c | 2 +- drivers/gpu/drm/dp/drm_dp_helper_internal.h | 33 +++++++++++++++++++ drivers/gpu/drm/dp/drm_dp_helper_mod.c | 22 +++++++++++++ .../gpu/drm/{ => dp}/drm_dp_mst_topology.c | 6 ++-- .../{ => dp}/drm_dp_mst_topology_internal.h | 2 +- drivers/gpu/drm/drm_crtc_helper_internal.h | 27 --------------- drivers/gpu/drm/drm_dsc.c | 2 +- drivers/gpu/drm/drm_kms_helper_common.c | 14 -------- drivers/gpu/drm/gma500/cdv_intel_dp.c | 2 +- drivers/gpu/drm/gma500/intel_bios.c | 2 +- drivers/gpu/drm/i915/Kconfig | 1 + drivers/gpu/drm/i915/display/intel_bios.c | 2 +- drivers/gpu/drm/i915/display/intel_display.c | 2 +- .../drm/i915/display/intel_display_types.h | 4 +-- drivers/gpu/drm/i915/display/intel_dp.c | 2 +- drivers/gpu/drm/i915/display/intel_dp_hdcp.c | 4 +-- .../drm/i915/display/intel_dp_link_training.h | 2 +- drivers/gpu/drm/i915/display/intel_lspcon.c | 2 +- drivers/gpu/drm/msm/Kconfig | 1 + drivers/gpu/drm/msm/dp/dp_audio.c | 2 +- drivers/gpu/drm/msm/dp/dp_aux.h | 2 +- drivers/gpu/drm/msm/dp/dp_catalog.c | 2 +- drivers/gpu/drm/msm/dp/dp_ctrl.c | 2 +- drivers/gpu/drm/nouveau/Kconfig | 1 + drivers/gpu/drm/nouveau/dispnv50/disp.c | 2 +- drivers/gpu/drm/nouveau/nouveau_connector.h | 2 +- drivers/gpu/drm/nouveau/nouveau_dp.c | 2 +- drivers/gpu/drm/nouveau/nouveau_encoder.h | 4 +-- drivers/gpu/drm/panel/panel-edp.c | 4 +-- .../gpu/drm/panel/panel-samsung-atna33xc20.c | 4 +-- drivers/gpu/drm/radeon/atombios_dp.c | 2 +- drivers/gpu/drm/radeon/radeon_connectors.c | 2 +- drivers/gpu/drm/radeon/radeon_dp_mst.c | 2 +- drivers/gpu/drm/radeon/radeon_mode.h | 4 +-- drivers/gpu/drm/rockchip/Kconfig | 1 + .../gpu/drm/rockchip/analogix_dp-rockchip.c | 2 +- drivers/gpu/drm/rockchip/cdn-dp-core.c | 2 +- drivers/gpu/drm/rockchip/cdn-dp-core.h | 2 +- drivers/gpu/drm/rockchip/rockchip_lvds.c | 2 +- drivers/gpu/drm/rockchip/rockchip_rgb.c | 2 +- .../drm/selftests/test-drm_dp_mst_helper.c | 4 +-- drivers/gpu/drm/tegra/Kconfig | 1 + drivers/gpu/drm/tegra/dp.c | 2 +- drivers/gpu/drm/tegra/dpaux.c | 2 +- drivers/gpu/drm/tegra/sor.c | 2 +- drivers/gpu/drm/xlnx/Kconfig | 1 + drivers/gpu/drm/xlnx/zynqmp_dp.c | 2 +- include/drm/{ => dp}/drm_dp_aux_bus.h | 0 .../drm/{ => dp}/drm_dp_dual_mode_helper.h | 0 include/drm/{ => dp}/drm_dp_helper.h | 0 include/drm/{ => dp}/drm_dp_mst_helper.h | 2 +- include/drm/drm_dsc.h | 2 +- 84 files changed, 172 insertions(+), 132 deletions(-) create mode 100644 drivers/gpu/drm/dp/Makefile rename drivers/gpu/drm/{drm_dp_helper.c => dp/drm_dp.c} (99%) rename drivers/gpu/drm/{ => dp}/drm_dp_aux_bus.c (99%) rename drivers/gpu/drm/{ => dp}/drm_dp_aux_dev.c (98%) rename drivers/gpu/drm/{ => dp}/drm_dp_cec.c (99%) rename drivers/gpu/drm/{ => dp}/drm_dp_dual_mode_helper.c (99%) create mode 100644 drivers/gpu/drm/dp/drm_dp_helper_internal.h create mode 100644 drivers/gpu/drm/dp/drm_dp_helper_mod.c rename drivers/gpu/drm/{ => dp}/drm_dp_mst_topology.c (99%) rename drivers/gpu/drm/{ => dp}/drm_dp_mst_topology_internal.h (94%) rename include/drm/{ => dp}/drm_dp_aux_bus.h (100%) rename include/drm/{ => dp}/drm_dp_dual_mode_helper.h (100%) rename include/drm/{ => dp}/drm_dp_helper.h (100%) rename include/drm/{ => dp}/drm_dp_mst_helper.h (99%) base-commit: fbce7b8d8df5af8d404b6aeaf63779f91bdbeb5d prerequisite-patch-id: c2b2f08f0eccc9f5df0c0da49fa1d36267deb11d prerequisite-patch-id: c67e5d886a47b7d0266d81100837557fda34cb24 prerequisite-patch-id: f89fb57d67dafa8cd0adc24bdcccd8faa1d790b4 -- 2.34.1

2 7

linux-next: build warning after merge of the drm-misc tree
by Stephen Rothwell 20 Jan '22

20 Jan '22

Hi all, After merging the drm-misc tree, today's linux-next build (htmldocs) produced this warning: include/drm/drm_modeset_lock.h:74: warning: Function parameter or member 'stack_depot' not described in 'drm_modeset_acquire_ctx' Introduced by commit cd06ab2fd48f ("drm/locking: add backtrace for locking contended locks without backoff") -- Cheers, Stephen Rothwell

2 2

[RFC] amdgpu MST questions, and future plans
by Lyude Paul 20 Jan '22

20 Jan '22

Hi! I'm writing this email because I'm currently finishing up removing pretty much all of the non-atomic MST code in drm_dp_mst_topology_mgr.c as it's really made it difficult to maintain MST over time. As well, once that's complete it's likely I'm going to start on the (extremely overdue) task of moving as much of amdgpu's MST code for DSC out of amdgpu and into the DRM code where it's supposed to live. This brings me to two questions. The first major one being: is anyone capable of testing the MST support in radeon.ko to figure out whether it works or not? I've already talked with hwentlan and ag5df about this and they haven't been able to find anyone to help with testing this. The reason I ask is because radeon isn't an atomic driver, and is basically the only user of any of the non-atomic parts of the MST helpers. If anyone want to prevent this from breaking in the future, I would definitely recommend they step up to try and help with testing it - otherwise I'm probably going to be pushing for us just to drop the code entirely. The second question is: is anyone willing to help me figure out how much of the code in amdgpu related to DSC is definitely not amdgpu specific and can be moved out? I'm honestly having a lot of trouble wrapping my head around how some of this works, and how much of this code is even needed. As well, with the amount of issues I've already found in it (there's numerous spots where we're storing MST state outside of atomic state for instance, lots of duplicates of DP helper functions that should not be here, etc.) it's quite likely I'm going to be rewriting rather large chunks of it. If anyone would like to volunteer please let me know, it'd be super appreciated and likely will make reviewing the patches that will come out of this easier. -- Cheers, Lyude Paul (she/her) Software Engineer at Red Hat

1 1

[Bug 213561] New: [bisected] AMD GPU can no longer idle state after commit 1c0b0efd148d5b24c4932ddb3fa03c8edd6097b3
by bugzilla-daemon＠bugzilla.kernel.org 20 Jan '22

20 Jan '22

https://bugzilla.kernel.org/show_bug.cgi?id=213561 Bug ID: 213561 Summary: [bisected] AMD GPU can no longer idle state after commit 1c0b0efd148d5b24c4932ddb3fa03c8edd6097b3 Product: Drivers Version: 2.5 Kernel Version: 5.13rc7 Hardware: All OS: Linux Tree: Mainline Status: NEW Severity: normal Priority: P1 Component: Video(DRI - non Intel) Assignee: drivers_video-dri(a)kernel-bugs.osdl.org Reporter: untaintableangel(a)hotmail.co.uk Regression: No Nature of the problem: RX 5700 is unable to enter low power state at idle (see below for usual behaviour) Sensors at idle prior to the commit: amdgpu-pci-0f00 Adapter: PCI adapter vddgfx: 775.00 mV fan1: 0 RPM (min = 0 RPM, max = 3200 RPM) edge: +48.0°C (crit = +100.0°C, hyst = -273.1°C) (emerg = +105.0°C) junction: +48.0°C (crit = +110.0°C, hyst = -273.1°C) (emerg = +115.0°C) mem: +52.0°C (crit = +105.0°C, hyst = -273.1°C) (emerg = +110.0°C) power1: 8.00 W (cap = 165.00 W) After the commit, the lowest is: amdgpu-pci-0f00 Adapter: PCI adapter vddgfx: 1.03 V fan1: 0 RPM (min = 0 RPM, max = 3200 RPM) edge: +54.0°C (crit = +100.0°C, hyst = -273.1°C) (emerg = +105.0°C) junction: +56.0°C (crit = +110.0°C, hyst = -273.1°C) (emerg = +115.0°C) mem: +52.0°C (crit = +105.0°C, hyst = -273.1°C) (emerg = +110.0°C) power1: 31.00 W (cap = 165.00 W) This problem wasn't present in rc6 but is present in 5.13rc7 and bisects to: 1c0b0efd148d5b24c4932ddb3fa03c8edd6097b3 is the first bad commit commit 1c0b0efd148d5b24c4932ddb3fa03c8edd6097b3 Author: Yifan Zhang <yifan1.zhang(a)amd.com> Date: Thu Jun 10 10:10:07 2021 +0800 drm/amdgpu/gfx10: enlarge CP_MEC_DOORBELL_RANGE_UPPER to cover full doorbell. If GC has entered CGPG, ringing doorbell > first page doesn't wakeup GC. Enlarge CP_MEC_DOORBELL_RANGE_UPPER to workaround this issue. Signed-off-by: Yifan Zhang <yifan1.zhang(a)amd.com> Reviewed-by: Felix Kuehling <Felix.Kuehling(a)amd.com> Reviewed-by: Alex Deucher <alexander.deucher(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> Cc: stable(a)vger.kernel.org The device is a Sapphire Pulse RX5700 and this problem is seen even with one monitor set at 60Hz. GPU: 0f:00.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Navi 10 [Radeon RX 5600 OEM/5600 XT / 5700/5700 XT] [1002:731f] (rev c4) -- You may reply to this email to add a comment. You are receiving this mail because: You are watching the assignee of the bug.

1 20

[pull] amdgpu, radeon drm-fixes-5.17
by Alex Deucher 20 Jan '22

20 Jan '22

Hi Dave, Daniel, Fixes for 5.17. This does not yet include the proper fix for the patch Linux reverted when he pulled in the original drm merge windows PR. I was waiting for rc1. At that point do you plan to merge rc1 into drm-fixes? If so, I can just drop the fix on top. The following changes since commit 4efdddbce7c1329f00c458e85dcaf105aebdc0ed: Merge tag 'amd-drm-next-5.17-2022-01-12' of https://gitlab.freedesktop.org/agd5f/linux into drm-next (2022-01-14 15:42:28 +0100) are available in the Git repository at: https://gitlab.freedesktop.org/agd5f/linux.git tags/amd-drm-fixes-5.17-2022-01-19 for you to fetch changes up to 4722f463896cc0ef1a6f1c3cb2e171e949831249: drm/radeon: fix error handling in radeon_driver_open_kms (2022-01-18 18:00:58 -0500) ---------------------------------------------------------------- amd-drm-fixes-5.17-2022-01-19: amdgpu: - SR-IOV fix - VCN harvest fix - Suspend/resume fixes - Tahiti fix - Enable GPU recovery on yellow carp radeon: - Fix error handling regression in radeon_driver_open_kms ---------------------------------------------------------------- Alex Deucher (3): drm/amdgpu: invert the logic in amdgpu_device_should_recover_gpu() drm/amdgpu: don't do resets on APUs which don't support it drm/amdgpu: drop flags check for CHIP_IP_DISCOVERY CHANDAN VURDIGERE NATARAJ (1): drm/amdgpu: Enable recovery on yellow carp Christian König (1): drm/radeon: fix error handling in radeon_driver_open_kms Guchun Chen (1): drm/amdgpu: apply vcn harvest quirk Jingwen Chen (1): drm/amd/amdgpu: fixing read wrong pf2vf data in SRIOV Lukas Fink (1): drm/amdgpu: Fix rejecting Tahiti GPUs Mario Limonciello (1): drm/amd/display: Revert W/A for hard hangs on DCN20/DCN21 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 45 +++++++++------------- drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c | 32 ++++++++++++--- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 5 --- drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 20 ++++------ drivers/gpu/drm/amd/amdgpu/cik.c | 4 ++ drivers/gpu/drm/amd/amdgpu/vi.c | 4 ++ .../amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c | 11 +----- .../drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c | 11 +----- .../amd/display/dc/irq/dcn20/irq_service_dcn20.c | 25 ------------ .../amd/display/dc/irq/dcn20/irq_service_dcn20.h | 2 - .../amd/display/dc/irq/dcn21/irq_service_dcn21.c | 25 ------------ .../amd/display/dc/irq/dcn21/irq_service_dcn21.h | 2 - drivers/gpu/drm/amd/display/dc/irq/irq_service.c | 2 +- drivers/gpu/drm/amd/display/dc/irq/irq_service.h | 4 -- drivers/gpu/drm/radeon/radeon_kms.c | 22 ++++++----- 15 files changed, 75 insertions(+), 139 deletions(-)

1 0

[PATCH 1/3] dma-buf: make the flags can be configured during dma fence init
by Huang Rui 20 Jan '22

20 Jan '22

In some user scenarios, the get_timeline_name callback uses the flags to decide which way to return the timeline string name. Once the trace_dma_fence_init event is enabled, it will call get_timeline_name callback to dump the fence structure. However, at this moment, the flags are always 0, and it might trigger some issues in get_timeline_name callback implementation of different gpu driver. So make a member to initialize the flags in dma_fence_init(). Signed-off-by: Huang Rui <ray.huang(a)amd.com> --- drivers/dma-buf/dma-fence.c | 2 +- include/linux/dma-fence.h | 7 +++++++ 2 files changed, 8 insertions(+), 1 deletion(-) diff --git a/drivers/dma-buf/dma-fence.c b/drivers/dma-buf/dma-fence.c index 066400ed8841..3e0622bf385f 100644 --- a/drivers/dma-buf/dma-fence.c +++ b/drivers/dma-buf/dma-fence.c @@ -952,7 +952,7 @@ dma_fence_init(struct dma_fence *fence, const struct dma_fence_ops *ops, fence->lock = lock; fence->context = context; fence->seqno = seqno; - fence->flags = 0UL; + fence->flags = ops->init_flags; fence->error = 0; trace_dma_fence_init(fence); diff --git a/include/linux/dma-fence.h b/include/linux/dma-fence.h index 1ea691753bd3..f9270c5bc07a 100644 --- a/include/linux/dma-fence.h +++ b/include/linux/dma-fence.h @@ -131,6 +131,13 @@ struct dma_fence_ops { */ bool use_64bit_seqno; + /** + * @init_flags: + * + * The initial value of fence flags (A mask of DMA_FENCE_FLAG_* defined). + */ + unsigned long init_flags; + /** * @get_driver_name: * -- 2.25.1

3 11

[PATCH] drm/msm/hdmi: Fix missing put_device() call in msm_hdmi_get_phy
by Miaoqian Lin 19 Jan '22

19 Jan '22

The reference taken by 'of_find_device_by_node()' must be released when not needed anymore. Add the corresponding 'put_device()' in the error handling path. Fixes: e00012b256d4 ("drm/msm/hdmi: Make HDMI core get its PHY") Signed-off-by: Miaoqian Lin <linmq006(a)gmail.com> --- drivers/gpu/drm/msm/hdmi/hdmi.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.c b/drivers/gpu/drm/msm/hdmi/hdmi.c index 75b64e6ae035..a439794a32e8 100644 --- a/drivers/gpu/drm/msm/hdmi/hdmi.c +++ b/drivers/gpu/drm/msm/hdmi/hdmi.c @@ -95,10 +95,15 @@ static int msm_hdmi_get_phy(struct hdmi *hdmi) of_node_put(phy_node); - if (!phy_pdev || !hdmi->phy) { + if (!phy_pdev) { DRM_DEV_ERROR(&pdev->dev, "phy driver is not ready\n"); return -EPROBE_DEFER; } + if (!hdmi->phy) { + DRM_DEV_ERROR(&pdev->dev, "phy driver is not ready\n"); + put_device(&phy_pdev->dev); + return -EPROBE_DEFER; + } hdmi->phy_dev = get_device(&phy_pdev->dev); -- 2.17.1

2 1

[PATCH] drm/msm/dsi: Fix missing put_device() call in dsi_get_phy
by Miaoqian Lin 19 Jan '22

19 Jan '22

If of_find_device_by_node() succeeds, dsi_get_phy() doesn't a corresponding put_device(). Thus add put_device() to fix the exception handling. Fixes: ec31abf ("drm/msm/dsi: Separate PHY to another platform device") Signed-off-by: Miaoqian Lin <linmq006(a)gmail.com> --- drivers/gpu/drm/msm/dsi/dsi.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/msm/dsi/dsi.c b/drivers/gpu/drm/msm/dsi/dsi.c index 75ae3008b68f..35be526e907a 100644 --- a/drivers/gpu/drm/msm/dsi/dsi.c +++ b/drivers/gpu/drm/msm/dsi/dsi.c @@ -40,7 +40,12 @@ static int dsi_get_phy(struct msm_dsi *msm_dsi) of_node_put(phy_node); - if (!phy_pdev || !msm_dsi->phy) { + if (!phy_pdev) { + DRM_DEV_ERROR(&pdev->dev, "%s: phy driver is not ready\n", __func__); + return -EPROBE_DEFER; + } + if (!msm_dsi->phy) { + put_device(&phy_pdev->dev); DRM_DEV_ERROR(&pdev->dev, "%s: phy driver is not ready\n", __func__); return -EPROBE_DEFER; } -- 2.17.1

2 1

[PATCH -next] drm/msm: remove variable set but not used
by Yang Li 19 Jan '22

19 Jan '22

The code that uses variable mdss has been removed, So the declaration and assignment of the variable can be removed. Eliminate the following clang warning: drivers/gpu/drm/msm/msm_drv.c:513:19: warning: variable 'mdss' set but not used [-Wunused-but-set-variable] Reported-by: Abaci Robot <abaci(a)linux.alibaba.com> Fixes: 2027e5b3413d ("drm/msm: Initialize MDSS irq domain at probe time") Signed-off-by: Yang Li <yang.lee(a)linux.alibaba.com> --- drivers/gpu/drm/msm/msm_drv.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c index ad35a5d94053..59e30192cdf6 100644 --- a/drivers/gpu/drm/msm/msm_drv.c +++ b/drivers/gpu/drm/msm/msm_drv.c @@ -510,7 +510,6 @@ static int msm_drm_init(struct device *dev, const struct drm_driver *drv) struct msm_drm_private *priv = dev_get_drvdata(dev); struct drm_device *ddev; struct msm_kms *kms; - struct msm_mdss *mdss; int ret, i; ddev = drm_dev_alloc(drv, dev); @@ -521,8 +520,6 @@ static int msm_drm_init(struct device *dev, const struct drm_driver *drv) ddev->dev_private = priv; priv->dev = ddev; - mdss = priv->mdss; - priv->wq = alloc_ordered_workqueue("msm", 0); priv->hangcheck_period = DRM_MSM_HANGCHECK_DEFAULT_PERIOD; -- 2.20.1.7.g153144c

2 1

[PATCH] drm/i915/guc: Ensure multi-lrc fini breadcrumb math is correct
by Matthew Brost 19 Jan '22

19 Jan '22

Realized that the GuC multi-lrc fini breadcrumb emit code is very delicate as the math this code does relies on functions it calls to emit a certain number of DWs. Add a few GEM_BUG_ONs to assert the math is correct. v2: - Rebase + resend for CI (Checkpatch) - Fix blank line warning Signed-off-by: Matthew Brost <matthew.brost(a)intel.com> Reviewed-by: John Harrison <John.C.Harrison(a)Intel.com> --- .../gpu/drm/i915/gt/uc/intel_guc_submission.c | 33 +++++++++++++++---- 1 file changed, 27 insertions(+), 6 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 3918f1be114f..650efd3d3a48 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c @@ -4429,27 +4429,31 @@ static inline bool skip_handshake(struct i915_request *rq) return test_bit(I915_FENCE_FLAG_SKIP_PARALLEL, &rq->fence.flags); } +#define NON_SKIP_LEN 6 static u32 * emit_fini_breadcrumb_parent_no_preempt_mid_batch(struct i915_request *rq, u32 *cs) { struct intel_context *ce = rq->context; + __maybe_unused u32 *before_fini_breadcrumb_user_interrupt_cs; + __maybe_unused u32 *start_fini_breadcrumb_cs = cs; GEM_BUG_ON(!intel_context_is_parent(ce)); if (unlikely(skip_handshake(rq))) { /* * NOP everything in __emit_fini_breadcrumb_parent_no_preempt_mid_batch, - * the -6 comes from the length of the emits below. + * the NON_SKIP_LEN comes from the length of the emits below. */ memset(cs, 0, sizeof(u32) * - (ce->engine->emit_fini_breadcrumb_dw - 6)); - cs += ce->engine->emit_fini_breadcrumb_dw - 6; + (ce->engine->emit_fini_breadcrumb_dw - NON_SKIP_LEN)); + cs += ce->engine->emit_fini_breadcrumb_dw - NON_SKIP_LEN; } else { cs = __emit_fini_breadcrumb_parent_no_preempt_mid_batch(rq, cs); } /* Emit fini breadcrumb */ + before_fini_breadcrumb_user_interrupt_cs = cs; cs = gen8_emit_ggtt_write(cs, rq->fence.seqno, i915_request_active_timeline(rq)->hwsp_offset, @@ -4459,6 +4463,12 @@ emit_fini_breadcrumb_parent_no_preempt_mid_batch(struct i915_request *rq, *cs++ = MI_USER_INTERRUPT; *cs++ = MI_NOOP; + /* Ensure our math for skip + emit is correct */ + GEM_BUG_ON(before_fini_breadcrumb_user_interrupt_cs + NON_SKIP_LEN != + cs); + GEM_BUG_ON(start_fini_breadcrumb_cs + + ce->engine->emit_fini_breadcrumb_dw != cs); + rq->tail = intel_ring_offset(rq, cs); return cs; @@ -4501,22 +4511,25 @@ emit_fini_breadcrumb_child_no_preempt_mid_batch(struct i915_request *rq, u32 *cs) { struct intel_context *ce = rq->context; + __maybe_unused u32 *before_fini_breadcrumb_user_interrupt_cs; + __maybe_unused u32 *start_fini_breadcrumb_cs = cs; GEM_BUG_ON(!intel_context_is_child(ce)); if (unlikely(skip_handshake(rq))) { /* * NOP everything in __emit_fini_breadcrumb_child_no_preempt_mid_batch, - * the -6 comes from the length of the emits below. + * the NON_SKIP_LEN comes from the length of the emits below. */ memset(cs, 0, sizeof(u32) * - (ce->engine->emit_fini_breadcrumb_dw - 6)); - cs += ce->engine->emit_fini_breadcrumb_dw - 6; + (ce->engine->emit_fini_breadcrumb_dw - NON_SKIP_LEN)); + cs += ce->engine->emit_fini_breadcrumb_dw - NON_SKIP_LEN; } else { cs = __emit_fini_breadcrumb_child_no_preempt_mid_batch(rq, cs); } /* Emit fini breadcrumb */ + before_fini_breadcrumb_user_interrupt_cs = cs; cs = gen8_emit_ggtt_write(cs, rq->fence.seqno, i915_request_active_timeline(rq)->hwsp_offset, @@ -4526,11 +4539,19 @@ emit_fini_breadcrumb_child_no_preempt_mid_batch(struct i915_request *rq, *cs++ = MI_USER_INTERRUPT; *cs++ = MI_NOOP; + /* Ensure our math for skip + emit is correct */ + GEM_BUG_ON(before_fini_breadcrumb_user_interrupt_cs + NON_SKIP_LEN != + cs); + GEM_BUG_ON(start_fini_breadcrumb_cs + + ce->engine->emit_fini_breadcrumb_dw != cs); + rq->tail = intel_ring_offset(rq, cs); return cs; } +#undef NON_SKIP_LEN + static struct intel_context * guc_create_virtual(struct intel_engine_cs **siblings, unsigned int count, unsigned long flags) -- 2.34.1

1 0

← Newer
1
...
41
42
43
44
45
46
47
...
85
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel January 2022