RE: [PATCH] drm/gpu-sched: fix force APP kill hang

> 2)if a fence signaled and try to clear some entity's dependency, > should set this entity guilty to prevent its job really run since the > dependency is fake signaled. Well, that is a clear NAK. It would mean that you mark things like the X server or Wayland queue is marked guilty because

Re: [PATCH] drm/gpu-sched: fix force APP kill hang

On 2018年03月29日 10:45, Deng, Emily wrote: Hi Christian, Thanks for your review, could you please give some advices on how to resolve the issue? How about adding the fence status ESRCH checking when check the fence signal? If so, need to identify the detail behavior if the fence status

Re: [PATCH libdrm] headers: sync up amdgpu_drm.h with drm-next

Acked-by: Chunming Zhou On 2018年03月29日 09:10, Yu, Qiang wrote: Hi guys, Ping. Regards, Qiang From: amd-gfx on behalf of Qiang Yu Sent: Tuesday, March 20, 2018 2:08:09

RE: [PATCH] drm/gpu-sched: fix force APP kill hang

Hi Christian, Thanks for your review, could you please give some advices on how to resolve the issue? How about adding the fence status ESRCH checking when check the fence signal? If so, need to identify the detail behavior if the fence status is ESRCH. Best Wishes, Emily Deng >

Re: [PATCH libdrm] headers: sync up amdgpu_drm.h with drm-next

Hi guys, Ping. Regards, Qiang From: amd-gfx on behalf of Qiang Yu Sent: Tuesday, March 20, 2018 2:08:09 PM To: amd-gfx@lists.freedesktop.org Cc: Zhu, Rex Subject: [PATCH libdrm] headers: sync up

Re: [PATCH 1/1] drm: add parameter explanation for some gem dmabuf_ops

On Mon, Mar 26, 2018 at 3:22 PM, Samuel Li wrote: > To reduce some warnings. > > Signed-off-by: Samuel Li Reviewed-by: Alex Deucher And pushed. Alex > --- > drivers/gpu/drm/drm_prime.c | 13 + > 1 file changed, 13

Re: [PATCH 2/8] PCI: Add pci_find_common_upstream_dev()

On 28/03/18 01:44 PM, Christian König wrote: > Well, isn't that exactly what dma_map_resource() is good for? As far as > I can see it makes sure IOMMU is aware of the access route and > translates a CPU address into a PCI Bus address. > I'm using that with the AMD IOMMU driver and at least

Re: [PATCH 2/8] PCI: Add pci_find_common_upstream_dev()

Am 28.03.2018 um 20:57 schrieb Logan Gunthorpe: On 28/03/18 12:28 PM, Christian König wrote: I'm just using amdgpu as blueprint because I'm the co-maintainer of it and know it mostly inside out. Ah, I see. The resource addresses are translated using dma_map_resource(). As far as I know that

Re: [PATCH 2/8] PCI: Add pci_find_common_upstream_dev()

On 28/03/18 12:28 PM, Christian König wrote: > I'm just using amdgpu as blueprint because I'm the co-maintainer of it > and know it mostly inside out. Ah, I see. > The resource addresses are translated using dma_map_resource(). As far > as I know that should be sufficient to offload all the

Re: [PATCH 2/8] PCI: Add pci_find_common_upstream_dev()

Am 28.03.2018 um 18:25 schrieb Logan Gunthorpe: On 28/03/18 10:02 AM, Christian König wrote: Yeah, that looks very similar to what I picked up from the older patches, going to read up on that after my vacation. Yeah, I was just reading through your patchset and there are a lot of

Re: [PATCH 2/8] PCI: Add pci_find_common_upstream_dev()

On 28/03/18 10:02 AM, Christian König wrote: > Yeah, that looks very similar to what I picked up from the older > patches, going to read up on that after my vacation. Yeah, I was just reading through your patchset and there are a lot of similarities. Though, I'm not sure what you're trying to

[PATCH] amdkfd: always select MMU_NOTIFIER

Building amdkfd without MMU notifiers is broken: In file included from drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c:28: drivers/gpu/drm/amd/amdkfd/kfd_priv.h:584:22: error: field 'mmu_notifier' has incomplete type This adds the missing 'select MMU_NOTIFIER' line to make it build cleanly all

Re: [PATCH 2/8] PCI: Add pci_find_common_upstream_dev()

On 28/03/18 09:07 AM, Christian König wrote: > Am 28.03.2018 um 14:38 schrieb Christoph Hellwig: >> On Sun, Mar 25, 2018 at 12:59:54PM +0200, Christian König wrote: >>> From: "wda...@nvidia.com" >>> >>> Add an interface to find the first device which is upstream of both >>>

Re: [PATCH] amdkfd: always select MMU_NOTIFIER

Am 28.03.2018 um 17:53 schrieb Arnd Bergmann: Building amdkfd without MMU notifiers is broken: In file included from drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c:28: drivers/gpu/drm/amd/amdkfd/kfd_priv.h:584:22: error: field 'mmu_notifier' has incomplete type This adds the missing 'select

Re: [PATCH 2/8] PCI: Add pci_find_common_upstream_dev()

Am 28.03.2018 um 17:47 schrieb Logan Gunthorpe: On 28/03/18 09:07 AM, Christian König wrote: Am 28.03.2018 um 14:38 schrieb Christoph Hellwig: On Sun, Mar 25, 2018 at 12:59:54PM +0200, Christian König wrote: From: "wda...@nvidia.com" Add an interface to find the first

Re: [PATCH 2/8] PCI: Add pci_find_common_upstream_dev()

Am 28.03.2018 um 14:38 schrieb Christoph Hellwig: On Sun, Mar 25, 2018 at 12:59:54PM +0200, Christian König wrote: From: "wda...@nvidia.com" Add an interface to find the first device which is upstream of both devices. Please work with Logan and base this on top of the

[PATCH -next] drm/amdgpu: fix error return code in amdgpu_amdkfd_gpuvm_create_process_vm()

Fix to return error code -ENOMEM from the eviction fence create fail error handling case instead of 0, as done elsewhere in this function. Fixes: a46a2cd103a8 ("drm/amdgpu: Add GPUVM memory management functions for KFD") Signed-off-by: Wei Yongjun ---

Re: [PATCH 1/8] lib/scatterlist: add sg_set_dma_addr() helper

On Sun, Mar 25, 2018 at 12:59:53PM +0200, Christian König wrote: > Use this function to set an sg entry to point to device resources mapped > using dma_map_resource(). The page pointer is set to NULL and only the DMA > address, length and offset values are valid. NAK. Please provide a higher

Re: [PATCH 2/8] PCI: Add pci_find_common_upstream_dev()

On Sun, Mar 25, 2018 at 12:59:54PM +0200, Christian König wrote: > From: "wda...@nvidia.com" > > Add an interface to find the first device which is upstream of both > devices. Please work with Logan and base this on top of the outstanding peer to peer patchset.

Re: [PATCH] drm/amdgpu: Fix KIQ hang on bare metal for device unbind/bind back.

On Wed, Mar 28, 2018 at 8:37 AM, Andrey Grodzovsky wrote: > Problem: When unbind and then bind back the device KIQ hangs on Vega > after mapping KCQs request. > > Fix: Adding deinitialzie code from CAIL during HW fini solves the > hang. > > Signed-off-by: Andrey

[PATCH] drm/amdgpu: Fix KIQ hang on bare metal for device unbind/bind back.

Problem: When unbind and then bind back the device KIQ hangs on Vega after mapping KCQs request. Fix: Adding deinitialzie code from CAIL during HW fini solves the hang. Signed-off-by: Andrey Grodzovsky --- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 52

OpenCL with Mesa on E8860 on big endian system (PowerPC)

Hi, I've managed to configure my device-tree correctly, and am now able to correctly start the radeon driver with acceleration on my T2080 (an e6500 PowerPC chip) with an AMD E8860 GPU (radeon/radeonsi). Thanks for the help with that. I am able to run a small OpenCL program that fills a float

Re: Raven Ridge Ryzen 2500U hang reproduced

On Wed, Mar 28, 2018 at 8:46 AM Mike Lothian wrote: > On Tue, 27 Mar 2018, 00:30 Bráulio Bhavamitra, > wrote: > >> Hi all, >> >> Following the random crashes happenning with many users (e.g. >>

Re: [PATCH] drm/gpu-sched: fix force APP kill hang

Am 28.03.2018 um 10:07 schrieb Emily Deng: issue: there are VMC page fault occured if force APP kill during 3dmark test, the cause is in entity_fini we manually signal all those jobs in entity's queue which confuse the sync/dep mechanism: 1)page fault occured in sdma's clear job which operate

Re: Raven Ridge Ryzen 2500U hang reproduced

On Tue, 27 Mar 2018, 00:30 Bráulio Bhavamitra, wrote: > Hi all, > > Following the random crashes happenning with many users (e.g. > https://www.phoronix.com/scan.php?page=news_item; > px=Raven-Ridge-March-Update), not only on Linux but also Windows, I've > been struggling to

Re: [PATCH] drm/amd/powerplay: Enable ACG SS feature

On Wed, Mar 28, 2018 at 03:51:16PM +0800, Kenneth Feng wrote: > port the new atomfirmware.h change in order to > support ACG SS feature and populate the ACG SS > parameters into SMU > We would better to say "update atomfirmware header". Because the guys who are in community not from AMD will not

RE: [PATCH] drm/amd/powerplay: Enable ACG SS feature

Hi Kenneth, There are several places which fix old coding style(comment inline). Please drop them since this is not a patch to fix coding style. Other than that, the patch is reviewed-by: Evan Quan Regards, Evan > -Original Message- > From: amd-gfx

Re: [PATCH 2/3] drm/amdgpu: refresh per vm bo lru

On 2018年03月27日 21:44, Christian König wrote: How about we update the LRU only when we need to re-validate at least one BO? I tried this just now, performance still isn't stable, sometime drop to 28fps by accident. I also tried to check num_evictions, if eviction happens, then update LRU,

[PATCH] drm/gpu-sched: fix force APP kill hang

issue: there are VMC page fault occured if force APP kill during 3dmark test, the cause is in entity_fini we manually signal all those jobs in entity's queue which confuse the sync/dep mechanism: 1)page fault occured in sdma's clear job which operate on shadow buffer, and shadow buffer's Gart

[PATCH] drm/amd/powerplay: Enable ACG SS feature

port the new atomfirmware.h change in order to support ACG SS feature and populate the ACG SS parameters into SMU Change-Id: I3297b93b166abc6e430d14ccdd362e353771ea36 Signed-off-by: Kenneth Feng --- drivers/gpu/drm/amd/include/atomfirmware.h | 12

Re: [PATCH] drm/amdgpu/sdma: fix mask in emit_pipeline_sync

Am 28.03.2018 um 05:24 schrieb Huang Rui: On Tue, Mar 27, 2018 at 04:33:55PM -0500, Alex Deucher wrote: Needs to be a 32 bit mask. Signed-off-by: Alex Deucher Cc: sta...@vger.kernel.org Nice catch! Acked-by: Huang Rui Indeed good catch! Patch

Re: [PATCH 1/8] drm/amdgpu: add emit_reg_write_reg_wait ring callback

Am 28.03.2018 um 00:32 schrieb Alex Deucher: This callback writes a value to a register and then reads back another register and waits for a value in a single operation. Provide a helper function using two operations for engines that don't support this opertion. Signed-off-by: Alex Deucher

Re: [PATCH v2 1/1] drm/amdgpu: Enable scatter gather display support

Am 28.03.2018 um 00:22 schrieb Samuel Li: It's auto by default. For CZ/ST, auto setting enables sg display when vram size is small; otherwise still uses vram. This patch fixed some potential hang issue introduced by change "allow framebuffer in GART memory as well" due to CZ/ST hardware

Re: [PATCH] drm/amdgpu: fix a kcq hang issue for SRIOV

Am 28.03.2018 um 06:36 schrieb Liu, Monk: The SDMA is not directly connected to the GFXHUB, so even if the SDMA would provide a single command for this the write/wait would still be executed as two operations. I don't understand this point, more details may be ?? For SDMA from v148 ucode,

Re: [PATCH] drm/amd/pp: Not call cgs interface to get display info

On Wed, Mar 28, 2018 at 12:44:55PM +0800, Rex Zhu wrote: > DC/Non DC all will update display configuration > when the display state changed > No need to get display info through cgs interface We would better add "." at the end of these two sentences. With that fixed, Reviewed-by: Huang Rui

Re: [PATCH 1/8] drm/amdgpu: add emit_reg_write_reg_wait ring callback

On Tue, Mar 27, 2018 at 05:32:51PM -0500, Alex Deucher wrote: > This callback writes a value to a register and then reads > back another register and waits for a value in a single > operation. > > Provide a helper function using two operations for engines > that don't support this opertion. > >