dri-devel March 2021

dri-devel@lists.freedesktop.org

326 participants
867 discussions

[PATCH v1 0/1] Fixes 30e2ae943c26 "drm/bridge: Introduce LT8912B DSI to HDMI
by Adrien Grassein 31 Mar '21

31 Mar '21

Hi, This patch fixes issues found by a static checker. Thanks, Adrien Grassein (1): drm/bridge: lt8912b: Fix issues found during static analysis drivers/gpu/drm/bridge/lontium-lt8912b.c | 20 ++++++++++++-------- 1 file changed, 12 insertions(+), 8 deletions(-) -- 2.25.1

2 3

[PATCH v5 0/2] HDMI2.1 PCON Misc Fixes
by Ankit Nautiyal 31 Mar '21

31 Mar '21

Patch1: Tweaks the drm_helpers for PCON configuration. Patch2: Removes unwanted code not applicable for older platforms. [Merged] Patch3: Fixes condition for starting FRL link training. rev3: Patch-1 from rev2 [Read PCON DSC ENC caps only for DPCD rev >= 1.4] is dropped as it mixes DPCD and DP revisions. rev4: Rebased Ankit Nautiyal (2): drm/dp_helper: Define options for FRL training for HDMI2.1 PCON drm/i915/display: Configure HDMI2.1 Pcon for FRL only if Src-Ctl mode is available drivers/gpu/drm/drm_dp_helper.c | 24 ++++++++++++++---------- drivers/gpu/drm/i915/display/intel_dp.c | 19 +++++++++++-------- include/drm/drm_dp_helper.h | 6 ++++-- 3 files changed, 29 insertions(+), 20 deletions(-) base-commit: 8e469260b87ff08620181d86570f7ec4e41c0ef7 -- 2.29.2

3 5

[Bug 211501] New: Kernel freeze when waking monitor from blanking / DPMS (AMDGPU / DC)
by bugzilla-daemon＠bugzilla.kernel.org 31 Mar '21

31 Mar '21

https://bugzilla.kernel.org/show_bug.cgi?id=211501 Bug ID: 211501 Summary: Kernel freeze when waking monitor from blanking / DPMS (AMDGPU / DC) Product: Drivers Version: 2.5 Kernel Version: 5.10.11 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: borisovjasen(a)protonmail.com Regression: No Created attachment 295021 --> https://bugzilla.kernel.org/attachment.cgi?id=295021&action=edit relevant dmesg output after kernel recovered from freeze Occasional temporary kernel freeze when monitor wakes up. The freeze only happens sometimes (which makes it hard to reproduce). Most of the time, it wakes up just fine. The last time it happened, I decided to wait, rather than resetting the computer, and the system recovered after some time (less than a minute). My graphics card is a Radeon RX Vega 64. CPU is Intel i7-7700K. I have 2 connected monitors: - Gigabyte G27QC connected via DisplayPort, running at its highest mode: 2560x1440@165Hz. - Samsung SyncMaster 2243WM connected via HDMI through a DvI adapter, also at its highest mode: 1680x1050@60Hz I don't know if the issue will occur if only one of the two monitors is connected; as I said, the bug is hard to reproduce. I am using the `sway` wayland compositor. After the system recovered from the freeze, I obtained the relevant `dmesg` output. I am attaching it as `dmesg-post-freeze.txt` (with relative timestamps). I am also attaching `dmesg-boot.txt` with the kernel messages from the system booting, in case you need it for more info about my system. -- 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 5

[bug report] drm/bridge: Introduce LT8912B DSI to HDMI bridge
by Dan Carpenter 31 Mar '21

31 Mar '21

Hello Adrien Grassein, The patch 30e2ae943c26: "drm/bridge: Introduce LT8912B DSI to HDMI bridge" from Mar 26, 2021, leads to the following static checker warning: drivers/gpu/drm/bridge/lontium-lt8912b.c:638 lt8912_parse_dt() warn: 'endpoint' isn't an ERR_PTR drivers/gpu/drm/bridge/lontium-lt8912b.c 620 static int lt8912_parse_dt(struct lt8912 *lt) 621 { 622 struct gpio_desc *gp_reset; 623 struct device *dev = lt->dev; 624 int ret = 0; 625 struct device_node *port_node; 626 struct device_node *endpoint; 627 628 gp_reset = devm_gpiod_get_optional(dev, "reset", GPIOD_OUT_HIGH); 629 if (IS_ERR(gp_reset)) { 630 ret = PTR_ERR(gp_reset); 631 if (ret != -EPROBE_DEFER) 632 dev_err(dev, "Failed to get reset gpio: %d\n", ret); 633 return ret; 634 } 635 lt->gp_reset = gp_reset; 636 637 endpoint = of_graph_get_endpoint_by_regs(dev->of_node, 0, -1); 638 if (IS_ERR(endpoint)) { Endpoint isn't an error pointer. You could just delete this check if you wanted and check of_property_count_u32_elems() for errors. 639 ret = PTR_ERR(endpoint); 640 goto end; goto end is a do nothing goto. Genereally direct returns are more readable have fewer bugs (based on ten years of reviewing static analysis warnings). 641 } 642 643 lt->data_lanes = of_property_count_u32_elems(endpoint, "data-lanes"); Either way, it's probably a good idea to check if of_property_count_u32_elems() fails. 644 of_node_put(endpoint); 645 646 lt->host_node = of_graph_get_remote_node(dev->of_node, 0, -1); 647 if (!lt->host_node) { 648 dev_err(lt->dev, "%s: Failed to get remote port\n", __func__); 649 ret = -ENODEV; 650 goto end; 651 } 652 653 port_node = of_graph_get_remote_node(dev->of_node, 1, -1); 654 if (!port_node) { 655 dev_err(lt->dev, "%s: Failed to get connector port\n", __func__); 656 ret = -ENODEV; 657 goto err_free_host_node; 658 } 659 660 lt->hdmi_port = of_drm_find_bridge(port_node); 661 if (IS_ERR(lt->hdmi_port)) { 662 dev_err(lt->dev, "%s: Failed to get hdmi port\n", __func__); 663 ret = PTR_ERR(lt->hdmi_port); 664 of_node_put(lt->host_node); 665 goto end; This should call of_node_put(port_node); then goto err_free_host_node; 666 } 667 668 if (!of_device_is_compatible(port_node, "hdmi-connector")) { 669 dev_err(lt->dev, "%s: Failed to get hdmi port\n", __func__); 670 ret = -EINVAL; This should call of_node_put(port_node); then goto err_free_host_node; as well. 671 } 672 673 of_node_put(port_node); 674 675 end: 676 return ret; 677 678 err_free_host_node: 679 of_node_put(lt->host_node); 680 return ret; 681 } regards, dan carpenter

2 4

[PATCH] [v2] amd: display: dc: struct dc_state is declared twice
by Wan Jiabing 31 Mar '21

31 Mar '21

struct dc_state is declared twice. One has been declared at 273rd line. Remove the duplicate. Signed-off-by: Wan Jiabing <wanjiabing(a)vivo.com> --- drivers/gpu/drm/amd/display/dc/dc.h | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 18ed0d3f247e..8ccf4bcd76b3 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -581,7 +581,6 @@ struct dc_bounding_box_overrides { int min_dcfclk_mhz; }; -struct dc_state; struct resource_pool; struct dce_hwseq; struct gpu_info_soc_bounding_box_v1_0; -- 2.25.1

1 0

[PATCH 0/2] ensure alignment on CPU page for bo mapping
by Xℹ Ruoyao 31 Mar '21

31 Mar '21

In AMDGPU driver, the bo mapping should always align to CPU page or the page table is corrupted. The first patch is cherry-picked from Loongson community, which sets a suitable dev_info.gart_page_size so Mesa will handle the alignment correctly. The second patch is added to ensure an ioctl with unaligned parameter to be rejected -EINVAL, instead of causing page table corruption. The patches should be applied for drm-next. Huacai Chen (1): drm/amdgpu: Set a suitable dev_info.gart_page_size Xℹ Ruoyao (1): drm/amdgpu: check alignment on CPU page for bo map drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 4 ++-- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 8 ++++---- 2 files changed, 6 insertions(+), 6 deletions(-) base-commit: a0c8b193bfe81cc8e9c7c162bb8d777ba12596f0 -- 2.31.1

3 4

[PATCH] amd: display: dc: struct dc_state is declared twice
by Wan Jiabing 31 Mar '21

31 Mar '21

struct dc_state has been declared at 273rd line. Remove the duplicate. Delete duplicate blank lines. Signed-off-by: Wan Jiabing <wanjiabing(a)vivo.com> --- drivers/gpu/drm/amd/display/dc/dc.h | 10 ---------- 1 file changed, 10 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 18ed0d3f247e..dc667298ab5b 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -234,7 +234,6 @@ struct dc_static_screen_params { unsigned int num_frames; }; - /* Surface update type is used by dc_update_surfaces_and_stream * The update type is determined at the very beginning of the function based * on parameters passed in and decides how much programming (or updating) is @@ -272,7 +271,6 @@ struct dc; struct dc_plane_state; struct dc_state; - struct dc_cap_funcs { bool (*get_dcc_compression_cap)(const struct dc *dc, const struct dc_dcc_surface_param *input, @@ -281,7 +279,6 @@ struct dc_cap_funcs { struct link_training_settings; - /* Structure to hold configuration flags set by dm at dc creation. */ struct dc_config { bool gpu_vm_support; @@ -581,7 +578,6 @@ struct dc_bounding_box_overrides { int min_dcfclk_mhz; }; -struct dc_state; struct resource_pool; struct dce_hwseq; struct gpu_info_soc_bounding_box_v1_0; @@ -757,7 +753,6 @@ enum dc_transfer_func_predefined { TRANSFER_FUNCTION_GAMMA26 }; - struct dc_transfer_func { struct kref refcount; enum dc_transfer_func_type type; @@ -770,7 +765,6 @@ struct dc_transfer_func { }; }; - union dc_3dlut_state { struct { uint32_t initialized:1; /*if 3dlut is went through color module for initialization */ @@ -784,7 +778,6 @@ union dc_3dlut_state { uint32_t raw; }; - struct dc_3dlut { struct kref refcount; struct tetrahedral_params lut_3d; @@ -1014,7 +1007,6 @@ enum dc_status dc_validate_global_state( struct dc_state *new_ctx, bool fast_validate); - void dc_resource_state_construct( const struct dc *dc, struct dc_state *dst_ctx); @@ -1167,7 +1159,6 @@ struct dc_container_id { unsigned short productCode; }; - struct dc_sink_dsc_caps { // 'true' if these are virtual DPCD's DSC caps (immediately upstream of sink in MST topology), // 'false' if they are sink's DSC caps @@ -1229,7 +1220,6 @@ struct dc_cursor { struct dc_cursor_attributes attributes; }; - /******************************************************************************* * Interrupt interfaces ******************************************************************************/ -- 2.25.1

3 3

linux-next: manual merge of the drm tree with Linus' tree
by Stephen Rothwell 30 Mar '21

30 Mar '21

Hi all, Today's linux-next merge of the drm tree got a conflict in: drivers/gpu/drm/amd/amdgpu/amdgpu_device.c between commits: 9adb125dde69 ("drm/amdgpu: re-enable suspend phase 2 for S0ix") 4021229e32bd ("drm/amdgpu/swsmu: skip gfx cgpg on s0ix suspend") 9bb735abcbd8 ("drm/amdgpu: update comments about s0ix suspend/resume") from Linus' tree and commit: e3c1b0712fdb ("drm/amdgpu: Reset the devices in the XGMI hive duirng probe") from the drm tree. I fixed it up (I think - see below) and can carry the fix as necessary. This is now fixed as far as linux-next is concerned, but any non trivial conflicts should be mentioned to your upstream maintainer when your tree is submitted for merging. You may also want to consider cooperating with the maintainer of the conflicting tree to minimise any particularly complex conflicts. -- Cheers, Stephen Rothwell diff --cc drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 8a5a8ff5d362,0f82c5d21237..000000000000 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@@ -2743,16 -2712,15 +2720,25 @@@ static int amdgpu_device_ip_suspend_pha continue; } + /* skip suspend of gfx and psp for S0ix + * gfx is in gfxoff state, so on resume it will exit gfxoff just + * like at runtime. PSP is also part of the always on hardware + * so no need to suspend it. + */ + if (adev->in_s0ix && + (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_PSP || + adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GFX)) + continue; + + /* skip unnecessary suspend if we do not initialize them yet */ + if (adev->gmc.xgmi.pending_reset && + !(adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC || + adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SMC || + adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_COMMON || + adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH)) { + adev->ip_blocks[i].status.hw = false; + continue; + } /* XXX handle errors */ r = adev->ip_blocks[i].version->funcs->suspend(adev); /* XXX handle errors */

3 2

[PATCH 00/30] DMA: Mundane typo fixes
by Bhaskar Chowdhury 30 Mar '21

30 Mar '21

This patch series fixes some trivial and rudimentary spellings in the COMMENT sections. Bhaskar Chowdhury (30): acpi-dma.c: Fix couple of typos altera-msgdma.c: Couple of typos fixed amba-pl08x.c: Fixed couple of typos bcm-sba-raid.c: Few typos fixed bcm2835-dma.c: Fix a typo idma64.c: Fix couple of typos iop-adma.c: Few typos fixed mv_xor.c: Fix a typo mv_xor.h: Fixed a typo mv_xor_v2.c: Fix a typo nbpfaxi.c: Fixed a typo of-dma.c: Fixed a typo s3c24xx-dma.c: Fix a typo Revert "s3c24xx-dma.c: Fix a typo" s3c24xx-dma.c: Few typos fixed st_fdma.h: Fix couple of typos ste_dma40_ll.h: Fix a typo tegra20-apb-dma.c: Fixed a typo xgene-dma.c: Few spello fixes at_hdmac.c: Quite a few spello fixes owl-dma.c: Fix a typo at_hdmac_regs.h: Couple of typo fixes dma-jz4780.c: Fix a typo Kconfig: Change Synopsys to Synopsis ste_dma40.c: Few spello fixes dw-axi-dmac-platform.c: Few typos fixed dpaa2-qdma.c: Fix a typo usb-dmac.c: Fix a typo edma.c: Fix a typo xilinx_dma.c: Fix a typo drivers/dma/Kconfig | 8 ++++---- drivers/dma/acpi-dma.c | 4 ++-- drivers/dma/altera-msgdma.c | 4 ++-- drivers/dma/amba-pl08x.c | 4 ++-- drivers/dma/at_hdmac.c | 14 +++++++------- drivers/dma/at_hdmac_regs.h | 4 ++-- drivers/dma/bcm-sba-raid.c | 8 ++++---- drivers/dma/bcm2835-dma.c | 2 +- drivers/dma/dma-jz4780.c | 2 +- drivers/dma/dw-axi-dmac/dw-axi-dmac-platform.c | 8 ++++---- drivers/dma/idma64.c | 4 ++-- drivers/dma/iop-adma.c | 6 +++--- drivers/dma/mv_xor.c | 2 +- drivers/dma/mv_xor.h | 2 +- drivers/dma/mv_xor_v2.c | 2 +- drivers/dma/nbpfaxi.c | 2 +- drivers/dma/of-dma.c | 2 +- drivers/dma/owl-dma.c | 2 +- drivers/dma/s3c24xx-dma.c | 6 +++--- drivers/dma/sh/shdmac.c | 2 +- drivers/dma/sh/usb-dmac.c | 2 +- drivers/dma/st_fdma.h | 4 ++-- drivers/dma/ste_dma40.c | 10 +++++----- drivers/dma/ste_dma40_ll.h | 2 +- drivers/dma/tegra20-apb-dma.c | 2 +- drivers/dma/ti/edma.c | 2 +- drivers/dma/xgene-dma.c | 6 +++--- drivers/dma/xilinx/xilinx_dma.c | 2 +- 28 files changed, 59 insertions(+), 59 deletions(-) -- 2.26.3

3 34

[PATCH v6 00/10] Fixes and cleanups for Host1x
by Mikko Perttunen 30 Mar '21

30 Mar '21

This is the first part of the Host1x/TegraDRM UAPI series, split out into a separate series that should be easier to merge. It contains a number of Host1x-related cleanups and fixes. In addition to the previous series there are a couple of new fixes. Tested on Jetson TX2. Thanks, Mikko Jon Hunter (1): gpu: host1x: Fix Tegra194 syncpt interrupt threshold Mikko Perttunen (9): gpu: host1x: Use different lock classes for each client gpu: host1x: Allow syncpoints without associated client gpu: host1x: Show number of pending waiters in debugfs gpu: host1x: Remove cancelled waiters immediately gpu: host1x: Use HW-equivalent syncpoint expiration check gpu: host1x: Cleanup and refcounting for syncpoints gpu: host1x: Reset max value when freeing a syncpoint gpu: host1x: Reserve VBLANK syncpoints at initialization gpu: host1x: Assign intr waiter inside lock drivers/gpu/drm/tegra/dc.c | 10 +- drivers/gpu/drm/tegra/drm.c | 14 +- drivers/gpu/drm/tegra/gr2d.c | 4 +- drivers/gpu/drm/tegra/gr3d.c | 4 +- drivers/gpu/drm/tegra/vic.c | 4 +- drivers/gpu/host1x/bus.c | 10 +- drivers/gpu/host1x/cdma.c | 11 +- drivers/gpu/host1x/debug.c | 14 +- drivers/gpu/host1x/dev.c | 6 + drivers/gpu/host1x/dev.h | 13 +- drivers/gpu/host1x/hw/cdma_hw.c | 2 +- drivers/gpu/host1x/hw/channel_hw.c | 10 +- drivers/gpu/host1x/hw/debug_hw.c | 2 +- drivers/gpu/host1x/hw/hw_host1x07_vm.h | 2 +- drivers/gpu/host1x/intr.c | 28 +++- drivers/gpu/host1x/intr.h | 4 +- drivers/gpu/host1x/job.c | 5 +- drivers/gpu/host1x/syncpt.c | 202 +++++++++++++++---------- drivers/gpu/host1x/syncpt.h | 4 +- drivers/staging/media/tegra-video/vi.c | 4 +- include/linux/host1x.h | 23 ++- 21 files changed, 235 insertions(+), 141 deletions(-) -- 2.30.1

3 16

← Newer
1
...
19
20
21
22
23
24
25
...
87
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel March 2021