RE: [PATCH] drm/sched: fix the bug of time out calculation(v3)

[AMD Official Use Only] what is that 'ts' representing for ? it looks to me the jiffies that it get scheduled to the ring, but a job scheduled to the ring doesn't represent it's being processed by hw. Thanks -- Monk Liu | Cloud-GPU Core team --

RE: [PATCH] drm/sched: fix the bug of time out calculation(v3)

[AMD Official Use Only] Yeah, that "kthread_should_park" is also irrelevant looks to me as well and it delays the signaled job's cleanup/free Thanks -- Monk Liu | Cloud-GPU Core team -- -Original Message-

RE: [PATCH] drm/sched: fix the bug of time out calculation(v3)

[AMD Official Use Only] >>> I'm not sure if the work_tdr is initialized when a maximum timeout is >>> specified. Please double check. Even timeout set to max the work_tdr is still initialized: int drm_sched_init(struct drm_gpu_scheduler *sched, const struct drm_sched_backend_o

Re: [PATCH] drm/amd/amdgpu: Add ready_to_reset resp for vega10

Reviewed-by: Jingwen Chen On Fri Aug 27, 2021 at 02:56:51PM +0800, YuBiao Wang wrote: > Send response to host after received the flr notification from host. > Port NV change to vega10. > > Signed-off-by: YuBiao Wang > --- > drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c | 2 ++ > drivers/gpu/drm/amd/amd

[PATCH] drm/amdgpu: adjust get preferred domain policy

current preferred domain policy is static, which makes vram > 256M APU never get a chance to allocate system S/G framebuffer. Change this policy to take vram memory pressure into accout. If system support both vram/system memory, return the preferred domain based on vram usage. Signed-off-by: Yifa

Re: [PATCH] drm/amdgpu: adjust get preferred domain policy

Am 27.08.21 um 10:45 schrieb Yifan Zhang: current preferred domain policy is static, which makes vram > 256M APU never get a chance to allocate system S/G framebuffer. Change this policy to take vram memory pressure into accout. If system support both vram/system memory, return the preferred doma

RE: [PATCH] drm/amdgpu: adjust get preferred domain policy

[Public] Hi Christian, Actually we met some issues when tried to enable multiple 4k displays in some APU platforms. With current code logic, framebuffers are always allocated from VRAM and eventually pin failed after vram is used up. Although system is able to support S/G framebuffer, it never

[PATCH] drm/amd/amdgpu: consolidate PSP TA unload function

Create common PSP TA unload function and replace all common TA unloading sequences. Signed-off-by: Candice Li --- drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 160 ++-- 1 file changed, 40 insertions(+), 120 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/dr

RE: [PATCH] drm/amdgpu: avoid over-handle of fence driver fini in s3 test (v2)

[Public] Hi Andrey and Christian, I just send out a new patch to address this, I am not sure if I understand your point correctly. Please review. The patch is to stop scheduler in fence_hw_fini and start the scheduler in fence_hw_init. Regards, Guchun -Original Message- From: Grodzov

[PATCH] drm/amdgpu: stop scheduler when calling hw_fini

This gurantees no more work on the ring can be submitted to hardware in suspend/resume case, otherwise the ring will not be empty before suspend. Suggested-by: Christian König Signed-off-by: Guchun Chen --- drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | 6 ++ 1 file changed, 6 insertions(+) d

Re: [PATCH] drm/amdgpu: stop scheduler when calling hw_fini

In general that looks good to me, but what could be is that we now try to stop/start the scheduler during reset twice. Andrey what do you think? Christian. Am 27.08.21 um 12:40 schrieb Guchun Chen: This gurantees no more work on the ring can be submitted to hardware in suspend/resume case, ot

[PATCH v2 1/3] drm/amdgpu/vce: set the priority for each ring

VCE has multiple rings. Set the proper priority level for each ring while initializing. Signed-off-by: Satyajit Sahu --- drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c | 14 ++ drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h | 1 + drivers/gpu/drm/amd/amdgpu/vce_v2_0.c | 4 +++- drivers/gpu/drm

[PATCH v2 2/3] drm/amdgpu/vcn: set the priority for each encode ring

VCN has multiple rings. Set the proper priority level for each encode ring while initializing. Signed-off-by: Satyajit Sahu --- drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 14 ++ drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h | 2 ++ drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c | 4 +++- drivers

[PATCH v2 3/3] drm/amdgpu:schedule vce/vcn encode based on priority

Schedule the encode job in VCE/VCN encode ring based on the priority set by UMD. Signed-off-by: Satyajit Sahu --- drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 16 1 file changed, 16 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/am

Re: [PATCH v1 02/14] mm: remove extra ZONE_DEVICE struct page refcount

On 8/25/21 19:49, Ralph Campbell wrote: > > On 8/25/21 4:15 AM, Vlastimil Babka wrote: >> On 8/25/21 05:48, Alex Sierra wrote: >>> From: Ralph Campbell >>> >>> ZONE_DEVICE struct pages have an extra reference count that complicates the >>> code for put_page() and several places in the kernel that

Re: [PATCH v2] drm/amd/display: Fix two cursor duplication when using overlay

Hi Nicholas! On Tuesday, August 24th, 2021 at 16:56, Kazlauskas, Nicholas wrote: > It's easiest to under the hardware cursor as being constrained within > the DRM plane specifications. Each DRM plane maps to 1 (or 2) hardware > pipes and the cursor has to be drawn along with it. The cursor will

Re: [PATCH v2] drm/amd/display: Fix two cursor duplication when using overlay

On Tuesday, August 24th, 2021 at 18:48, Harry Wentland wrote: > To elaborate on this a bit more, each HW plane's scanout engine > has the ability to scan out a cursor, in addition to the plane's > framebuffer. This cursor is drawn onto the plane at the scanout > phase. Any further scaling, color

Re: [PATCH] drm/sched: fix the bug of time out calculation(v3)

The TS  represents the point in time when the job was inserted into the pending list. I don't think it matters when it actually starts to be processed, what matters is when this job was inserted into pending list because right at that point you arm the TO timer (when no other is running already)

Re: [PATCH v2] drm/amd/display: Fix two cursor duplication when using overlay

On 2021-08-27 9:53 a.m., Simon Ser wrote: > On Tuesday, August 24th, 2021 at 18:48, Harry Wentland > wrote: > >> To elaborate on this a bit more, each HW plane's scanout engine >> has the ability to scan out a cursor, in addition to the plane's >> framebuffer. This cursor is drawn onto the pl

Re: [PATCH v2] drm/amd/display: Fix two cursor duplication when using overlay

On Friday, August 27th, 2021 at 16:04, Harry Wentland wrote: > > Is it possible to draw the cursor only on the overlay pipe (not on the > > primary > > pipe), even though the overlay pipe doesn't cover the whole CRTC? > > > > Or will the overlay pipe crop the cursor image? > > > > It will be cr

Re: [PATCH] drm/sched: fix the bug of time out calculation(v3)

So we agree if (kthread_should_park()) return NULL should go away ? Andrey On 2021-08-27 3:46 a.m., Liu, Monk wrote: [AMD Official Use Only] Yeah, that "kthread_should_park" is also irrelevant looks to me as well and it delays the signaled job's cleanup/free Thanks

Re: [PATCH] drm/sched: fix the bug of time out calculation(v3)

I don't think that makes sense. See we don't want to start the time when the job is inserted into the ring buffer, but rather when it starts processing. Starting processing is a bit swampy defined, but just starting the timer when the previous job completes should be fine enough. Christian.

Re: [PATCH v2 1/3] drm/amdgpu/vce: set the priority for each ring

Am 27.08.21 um 14:42 schrieb Satyajit Sahu: VCE has multiple rings. Set the proper priority level for each ring while initializing. Signed-off-by: Satyajit Sahu --- drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c | 14 ++ drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h | 1 + drivers/gpu/drm/a

Re: [PATCH] drm/sched: fix the bug of time out calculation(v3)

Yes, I don't see any good reason for that either. Christian. Am 27.08.21 um 15:45 schrieb Andrey Grodzovsky: So we agree if (kthread_should_park()) return NULL should go away ? Andrey On 2021-08-27 3:46 a.m., Liu, Monk wrote: [AMD Official Use Only] Yeah, that "kthread_should_park" is also

Re: [PATCH v6 01/11] moduleparam: add data member to struct kernel_param

On Wed, Aug 25, 2021 at 11:13 AM Jason Baron wrote: > > > > On 8/22/21 6:19 PM, Jim Cromie wrote: > > Add a const void* data member to the struct, to allow attaching > > private data that will be used soon by a setter method (via kp->data) > > to perform more elaborate actions. > > > > To attach t

Re: [PATCH v2 0/4] Various fixes to pass libdrm hotunplug tests

Ping Andrey On 2021-08-26 1:27 p.m., Andrey Grodzovsky wrote: Bunch of fixes to enable passing hotplug tests i previosly added here[1] with latest code. Once accepted I will enable the tests on libdrm side. [1] - https://gitlab.freedesktop.org/mesa/drm/-/merge_requests/172 v2: Dropping VCE pa

Re: [PATCH v3] drm/amd/pm: And destination bounds checking to struct copy

Applied. Thanks! Alex On Thu, Aug 26, 2021 at 11:16 PM Kees Cook wrote: > > In preparation for FORTIFY_SOURCE performing compile-time and run-time > field bounds checking for memcpy(), memmove(), and memset(), avoid > intentionally writing across neighboring fields. > > The "Board Parameters" m

Re: [PATCH] drm/sched: fix the bug of time out calculation(v3)

As I mentioned to Monk before - what about cases such as in this test - https://gitlab.freedesktop.org/mesa/drm/-/commit/bc21168fa924d3fc4a000492e861f50a1a135b25 Here you don't have serialized sequence where when jobs finishes processing and second starts, they execute together  concurrently - fo

Re: [PATCH] drm/sched: fix the bug of time out calculation(v3)

Yeah, that's what I meant with that the start of processing a job is a bit swampy defined. Jobs overload, but we simply don't have another good indicator that a job started except that the previous one completed. It's still better than starting the timer when pushing the job to the ring buff

Re: [PATCH] drm/sched: fix the bug of time out calculation(v3)

Sure then. Andrey On 2021-08-27 2:30 p.m., Christian König wrote: Yeah, that's what I meant with that the start of processing a job is a bit swampy defined. Jobs overload, but we simply don't have another good indicator that a job started except that the previous one completed. It's still

Re: [PATCH] drm/amdgpu: stop scheduler when calling hw_fini

I don't think it will start/stop twice because amdgpu_fence_driver_hw_fini/inint is not called during reset. I am worried about calling drm_sched_start without calling drm_sched_resubmit_job first since that the place where the jobs are actually restarted. Also calling drm_sched_start with fal

[PATCH 00/19] DC Patches August 23, 2021

This DC patchset brings improvements in multiple areas. In summary, we highlight: * Memory leak fixes and false positive warnings removed caught by coverity * Backlight fix for Carrizo/Stoney laptops * Unblocking ABM when ODM is enabled on DCN31 and up * Fork thread to offload work of hpd_rx_i

[PATCH] drm/amdgpu/swsmu: fix spelling mistake "minimun" -> "minimum"

From: Colin Ian King There are three identical spelling mistakes in dev_err messages. Fix these. Signed-off-by: Colin Ian King --- drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c | 2 +- drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c | 2 +- drivers/gpu/drm/amd/pm/swsmu/smu13/yellow

[pull] amdgpu, amdkfd drm-next-5.15

Hi Dave, Daniel, Fixes for 5.15. The following changes since commit 90a9266269eb9f71af1f323c33e1dca53527bd22: drm/amdgpu: Cancel delayed work when GFXOFF is disabled (2021-08-20 12:09:44 -0400) are available in the Git repository at: https://gitlab.freedesktop.org/agd5f/linux.git tags/am

[PATCH 01/19] drm/amd/display: cleanup idents after a revert

From: Angus Wang [WHY] The change has caused high idle memory clock speed and power consumption at some resolutions and frame rates for Navi10 [HOW] Reverted change "drm/amd/display: Fixed Intermittent blue screen on OLED panel" Reviewed-by: Aric Cyr Acked-by: Mikita Lipski Signed-off-by: An

[PATCH 07/19] drm/amd/display: move bpp range decision in decide dsc bw range function

From: Wenjing Liu [why] Before get dsc bw range is used to compute DSC bw range based on the given fixed bpp min/max input. The new change will merge any specs, signal, timing specific bpp range decision into this function. So the function needs to make a decision with all aspects considered. Re

[PATCH 17/19] drm/amd/display: Get backlight from PWM if DMCU is not initialized

From: Harry Wentland On Carrizo/Stoney systems we set backlight through panel_cntl, i.e. directly via the PWM registers, if DMCU is not initialized. We always read it back through ABM registers which leads to a mismatch and forces atomic_commit to program the backlight each time. Instead make su

[PATCH v3 0/4] Various fixes to pass libdrm hotunplug tests

Bunch of fixes to enable passing hotplug tests i previosly added here[1] with latest code. Once accepted I will enable the tests on libdrm side. [1] - https://gitlab.freedesktop.org/mesa/drm/-/merge_requests/172 v2: Dropping VCE patch since relevant function already fixed in latest code. Moving

[PATCH v3 1/4] drm/ttm: Create pinned list

This list will be used to capture all non VRAM BOs not on LRU so when device is hot unplugged we can iterate the list and unmap DMA mappings before device is removed. v2: Reanme function to ttm_bo_move_to_pinned v3: Move the pinned list to ttm device Signed-off-by: Andrey Grodzovsky Suggested-by

[PATCH v3 4/4] drm/amdgpu: Add a UAPI flag for hot plug/unplug

To support libdrm tests. Signed-off-by: Andrey Grodzovsky --- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 6400259a7c4b..c2fdf67ff551 100644 -

[PATCH v3 2/4] drm/ttm: Clear all DMA mappings on demand

Used by drivers supporting hot unplug to handle all DMA IOMMU group related dependencies before the group is removed during device removal and we try to access it after free when last device pointer from user space is dropped. v3: Switch to ttm_bo_get_unless_zerom Iterate bdev for pinned list Swit

[PATCH v3 3/4] drm/amdgpu: drm/amdgpu: Handle IOMMU enabled case

Handle all DMA IOMMU group related dependencies before the group is removed and we try to access it after free. v2: Move the actul handling function to TTM Signed-off-by: Andrey Grodzovsky --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drive

[PATCH 19/19] drm/amd/display: Add regamma/degamma coefficients and set sRGB when TF is BT709

From: Jaehyun Chung [Why] In YUV case, need to set the input TF to sRGB instead of BT709, even though the input TF type is distributed. SRGB was not being used because pixel format was not being set in the surface update sequence. Also, we were using the same coefficients for degamma and regamma

[PATCH 12/19] drm/amd/display: Fix false BAD_FREE warning from Coverity

From: Anson Jacob This is an attempt to fix false warning raised by Coverity via multiple CID's. Addresses-Coverity-ID: 1487412 ("Free of address-of expression") Cc: Wesley Chalmers Reviewed-by: Wesley Chalmers Acked-by: Mikita Lipski Signed-off-by: Anson Jacob --- drivers/gpu/drm/amd/disp

[PATCH 10/19] drm/amd/display: unblock abm when odm is enabled only on configs that support it

From: Josip Pavic [Why] When ODM is enabled, ABM is blocked on dcn31 but unblocked on dcn30. Since the dcn31 firmware is now able to handle ABM interop with ODM, it is no longer necessary to block ABM when ODM is enabled. Since the dcn30 firmware does not handle ABM interop with ODM, leaving th

[PATCH 03/19] drm/amd/display: Add DPCD writes at key points

From: "Leo (Hanghong) Ma" This reverts "drm/amd/display: Revert "Add DPCD writes at key points"." The following patch will fix the system hang issue. Reviewed-by: Aric Cyr Acked-by: Mikita Lipski Signed-off-by: Leo (Hanghong) Ma --- drivers/gpu/drm/amd/display/dc/core/dc_link.c | 7 +++

[PATCH 00/19] DC Patches August 23, 2021

This DC patchset brings improvements in multiple areas. In summary, we highlight: * Memory leak fixes and false positive warnings removed caught by coverity * Backlight fix for Carrizo/Stoney laptops * Unblocking ABM when ODM is enabled on DCN31 and up * Fork thread to offload work of hpd_rx_irq

[PATCH 09/19] drm/amd/display: Fork thread to offload work of hpd_rx_irq

From: Wayne Lin [Why] Currently, we will try to get dm.dc_lock in handle_hpd_rx_irq() when link lost happened, which is risky and could cause deadlock. e.g. If we are under procedure to enable MST streams and then monitor happens to toggle short hpd to notify link lost, then handle_hpd_rx_irq() w

[PATCH 11/19] drm/amd/display: Add flag to detect dpms force off during HPD

From: Aurabindo Pillai [Why] When a connector is unplugged, dpms is forced off so that some connector allocations are cleared off. This is done outside the commit sequence from the userspace. This causes HUBP blank. Due to the blank hubp, a non blocking commit which queues flip will encounter a t

[PATCH 13/19] drm/amd/display: Fix for null pointer access for ddc pin and aux engine.

From: Meenakshikumar Somasundaram [Why] Need a check for NULL pointer access for ddc pin and aux engine. [How] Adding a check for ddc pin and aux engine accesses. Reviewed-by: Jimmy Kizito Acked-by: Mikita Lipski Signed-off-by: Meenakshikumar Somasundaram --- drivers/gpu/drm/amd/display/dc/

[PATCH 02/19] drm/amd/display: Initialize lt_settings on instantiation

From: Michael Strauss [WHY] lt_settings' pointers remain uninitialized but nonzero if display fails to light up with no DPCD/EDID info populated, leading to a hang on access Reviewed-by: Nicholas Kazlauskas Acked-by: Mikita Lipski Signed-off-by: Michael Strauss --- drivers/gpu/drm/amd/displa

[PATCH 08/19] drm/amd/display: Add option to defer works of hpd_rx_irq

From: Wayne Lin [Why & How] Due to some code flow constraints, we need to defer dc_lock needed works from dc_link_handle_hpd_rx_irq(). Thus, do following changes: * Change allow_hpd_rx_irq() from static to public * Change handle_automated_test() from static to public * Extract link lost handling

[PATCH 14/19] drm/amd/display: [FW Promotion] Release 0.0.81

From: Anthony Koo - Fix PSR1 residency to be captured per panel - Merge VBIOS DP AUX change - Pass panel instance in dirty rect message - Z10 bug fixes - add detection required bit to boot status - Match abm_state to interrupt event specified otg_inst Reviewed-by: Anthony Koo Acked-by: Mikita L

[PATCH 16/19] drm/amd/display: Fix multiple memory leaks reported by coverity

From: Anson Jacob coccinelle patch used: @@ expression enc1,vpg,afmt; @@ - if (!enc1 || !vpg || !afmt) + if (!enc1 || !vpg || !afmt) { + kfree(enc1); + kfree(vpg); + kfree(afmt); return NULL; + } Addresses-Coverity-ID:

[PATCH 05/19] drm/amd/display: Drop unused privacy_mask setters and getters

From: Oliver Logush [Why and How] dwbc_funcs.set/get_privacy_mask isn't being used anymore, drop it Reviewed-by: Charlene Liu Acked-by: Mikita Lipski Signed-off-by: Oliver Logush --- drivers/gpu/drm/amd/display/dc/inc/hw/dwb.h | 6 -- 1 file changed, 6 deletions(-) diff --git a/drivers/

[PATCH 04/19] drm/amd/display: Fix system hang at boot

From: "Leo (Hanghong) Ma" [Why] During DQE's promotion test, system hang issue is found on linux system; [How] 1. Add NULL pointor check for the link in the sequence trace function; 2. Get the right link for the stream encoder before blank DP stream; Reviewed-by: Aric Cyr Acked-by: Mikit

[PATCH 18/19] drm/amd/display: Revert "Directly retrain link from debugfs"

From: Anson Jacob This reverts commit "drm/amd/display: Directly retrain link from debugfs". This patch broke new settings from taking effect. Hotplug is required for new settings to take effect. Reviewed-by: Mikita Lipski Acked-by: Mikita Lipski Signed-off-by: Anson Jacob --- drivers/gpu/d

[PATCH 15/19] drm/amd/display: 3.2.151

From: Aric Cyr - memory leak fixes - fix unitialized lt_settings pointers - add DP trace functions - add null pointer checks for DP sequence trace functions - expose dsc overhead bw parameter to DM - block abm when odm is enabled Reviewed-by: Aric Cyr Acked-by: Mikita Lipski Signed-off-by: Ari

[PATCH 06/19] drm/amd/display: expose dsc overhead bw in dc dsc header

From: Wenjing Liu [why] DM needs to know how much overhead is added to DSC as result of AMD internal DSC limitation. Reviewed-by: George Shen Acked-by: Mikita Lipski Signed-off-by: Wenjing Liu --- drivers/gpu/drm/amd/display/dc/dc_dsc.h | 5 ++ drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.

Re: [PATCH] drm/amd/display: add dcn register DP_MSA_VBID_MISC for dcn1.x and dcn2.x

On 2021-08-26 12:59 p.m., Wu, Hersen wrote: > [AMD Official Use Only] > > > This patch add missing AMD ASIC register for DP programming in up stream. > >   > > From 05768b78865d9b41a1d35e9f8e34901321208f2a Mon Sep 17 00:00:00 2001 > > From: Hersen Wu herse...@amd.com

Re: [Intel-gfx] [PATCH 2/2] drm/amdgpu: Disable PCIE_DPM on Intel RKL Platform

Hi Koba, Thank you for the patch! Perhaps something to improve: [auto build test WARNING on drm-intel/for-linux-next] [also build test WARNING on drm-tip/drm-tip linus/master v5.14-rc7 next-20210827] [If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting