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

issue: the vmflush in KCQ could be preempted (not like GFX ring which doesn't allow preemption in ring buffer) and this lead to vm flush fail when there is a world switch during the vm flush procedure (between write invalidate request and query invalidate ack) fix: separate vm flush for gfx and

[PATCH v2 2/2] drm/amdgpu: Set pm_display_cfg in non-dc mode

v2: keep the powerplay support in amdgpu_pm_compute_clocks those parameters are needed by powerplay so we can remove the wrapper cgs interface for get display info Change-Id: Idd332a4dbd34f72767a64d205df2affdb456042b Acked-by: Alex Deucher Signed-off-by: Rex Zhu

[PATCH v2 1/2] drm/amd/dc: Use pp interface to apply display cfg in DC

v2: drop the change for amdgpu_pm_computer_clocks in amdgpu_pm.c Change-Id: I362213550487c10746b058c9162b10df49777a47 Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git

RE: [PATCH 3/5] drm/ttm: remove the backing store if no placement is given

Acked-by: Roger He -Original Message- From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Christian K?nig Sent: Sunday, March 25, 2018 6:58 PM To: linaro-mm-...@lists.linaro.org; linux-me...@vger.kernel.org; dri-de...@lists.freedesktop.org;

RE: [PATCH 2/5] drm/ttm: keep a reference to transfer pipelined BOs

Reviewed-by: Roger He -Original Message- From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Christian K?nig Sent: Friday, March 16, 2018 9:21 PM To: linaro-mm-...@lists.linaro.org; linux-me...@vger.kernel.org; dri-de...@lists.freedesktop.org;

[PATCH 2/2] drm/amdgpu: Set pm_display_cfg in non-dc mode

those parameters are needed by powerplay so we can remove the wrapper cgs interface for get display info Change-Id: Idd332a4dbd34f72767a64d205df2affdb456042b Acked-by: Alex Deucher Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c |

[PATCH 1/2] drm/amd/dc: Use pp interface to apply display cfg in DC

amdgpu_pm_compute_clocks is used for none dc mode. Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 41 ++ .../drm/amd/display/amdgpu_dm/amdgpu_dm_services.c | 3 +- 2 files changed, 20 insertions(+), 24 deletions(-) diff

RE: [PATCH] drm/amdgpu: no job timeout setting on compute queues

Ok, fine by me /Monk From: Quan, Evan Sent: 2018年3月27日 9:34 To: Deucher, Alexander ; Liu, Monk ; Michel Dänzer Cc: amd-gfx@lists.freedesktop.org; Koenig, Christian Subject: RE: [PATCH] drm/amdgpu: no

Re: [PATCH 2/2] drm/amd/pp: Use atombios interface directly in powerplay

On Mon, Mar 26, 2018 at 07:56:02PM +0800, Rex Zhu wrote: > In order to remove the cgs wrapper functions for atombios api. > > Signed-off-by: Rex Zhu Reviewed-by: Huang Rui > --- > drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c | 145 >

Re: [PATCH] drm/amdgpu: Set pm_display_cfg in non-dc mode

On Mon, Mar 26, 2018 at 10:11 AM, Rex Zhu wrote: > those parameters are needed by powerplay > so we can remove the wrapper cgs interface for > get display info > > Change-Id: I503decd322837f016a16d6c45602d3d3c9fd953d > Signed-off-by: Rex Zhu > --- >

RE: [PATCH] drm/amdgpu: no job timeout setting on compute queues

Hi Monk, That’s a fix reached after a long discussion(I believe you were in that mail thread). 50s is for the dgemm test. We are not sure whether it’s enough for other compute use case. If virtualization still needs these messages, I believe we can list that as an exception -

Raven Ridge Ryzen 2500U hang reproduced

Hi all, Following the random crashes happenning with many users (e.g. https://www.phoronix.com/scan.php?page=news_item=Raven-Ridge-March-Update), not only on Linux but also Windows, I've been struggling to reproduce and generate any error log. After discovering that the error only happenned with

Re: [PATCH 1/2] drm/amd/dc: Use atombios api directly in DC

On 2018-03-26 07:56 AM, Rex Zhu wrote: > In order to remove the cgs wrapper functions for atombios api. > > Signed-off-by: Rex Zhu Patch 1 is Reviewed-by: Harry Wentland Harry > --- > .../gpu/drm/amd/display/dc/bios/command_table.c| 22

[PATCH xf86-video-amdgpu 5/5] Modify output_create_resources to use configure & change helper

From: "Leo (Sunpeng) Li" The conditionals within the for-loop do the same thing, since they're essentially copied to the configure helper. Signed-off-by: Leo (Sunpeng) Li --- src/drmmode_display.c | 82 ++-

[PATCH xf86-video-amdgpu 4/5] Enable configure & change helper to handle enums

From: "Leo (Sunpeng) Li" Copy code from output_create_resources, insert into configure & change helper, and rename variables. Some formatting changes as well. Modify doc-string to reflect change. Signed-off-by: Leo (Sunpeng) Li --- src/drmmode_display.c

[PATCH xf86-video-amdgpu 1/5] Add functions for changing CRTC color management properties

From: "Leo (Sunpeng) Li" This change adds a few functions in preparation of enabling CRTC color managment via the randr interface. The driver-private CRTC object now contains a list of properties, mirroring the driver-private output object. The lifecycle of the CRTC

[PATCH xf86-video-amdgpu 3/5] Keep CRTC properties consistent

From: "Leo (Sunpeng) Li" In cases where CRTC properties are updated without going through RRChangeOutputProperty, we don't update the properties in user land. Consider setting legacy gamma. It doesn't go through RRChangeOutputProperty, but modifies the CRTC's color

[PATCH xf86-video-amdgpu 0/5] Implementing non-legacy color management

From: "Leo (Sunpeng) Li" These patches will enable modification of non-legacy color management properties via xrandr. On top of the current legacy gamma, DRM allows the setting of three color management tables: the degamma LUT, the color transform matrix (CTM), and the

[PATCH xf86-video-amdgpu 2/5] Hook up CRTC color management functions

From: "Leo (Sunpeng) Li" The functions insert into the output resource creation, and property change functions. CRTC destroy is also hooked-up for proper cleanup of the CRTC property list. Signed-off-by: Leo (Sunpeng) Li --- src/drmmode_display.c | 19

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

To reduce some warnings. Signed-off-by: Samuel Li --- drivers/gpu/drm/drm_prime.c | 13 + 1 file changed, 13 insertions(+) diff --git a/drivers/gpu/drm/drm_prime.c b/drivers/gpu/drm/drm_prime.c index 7856a9b..caf675e 100644 --- a/drivers/gpu/drm/drm_prime.c +++

[PATCH][next] drm/amd/pp: fix logical or'ing of garbage in result by initializing result to zero

From: Colin Ian King Currently result is not initialized, so it contains a garbage value and this is or'd with data from vega10_program_didt_config_registers() which can end up with a non-zero value when the result should be zero. Fix this by ensuring result is

Re: [PATCH] drm/amdgpu: no job timeout setting on compute queues

That's fine with me too. Or make them DRM_INFO. Alex From: Liu, Monk Sent: Monday, March 26, 2018 8:55:51 AM To: Quan, Evan; Deucher, Alexander; Michel Dänzer Cc: amd-gfx@lists.freedesktop.org Subject: RE: [PATCH] drm/amdgpu: no job timeout setting on compute

Re: [PATCH] dma-buf: use parameter structure for dma_buf_attach

On Mon, Mar 26, 2018 at 12:47:01PM +0200, Christian König wrote: > Am 26.03.2018 um 10:36 schrieb Daniel Vetter: > > On Sun, Mar 25, 2018 at 01:34:51PM +0200, Christian König wrote: > [SNIP] > > > - attach->dev = dev; > > > + attach->dev = info->dev; > > > attach->dmabuf = dmabuf; > > >

Re: [RFC] Per file OOM badness

Hi all, Am Dienstag, den 30.01.2018, 11:28 +0100 schrieb Michal Hocko: > On Tue 30-01-18 10:29:10, Michel Dänzer wrote: > > On 2018-01-24 12:50 PM, Michal Hocko wrote: > > > On Wed 24-01-18 12:23:10, Michel Dänzer wrote: > > > > On 2018-01-24 12:01 PM, Michal Hocko wrote: > > > > > On Wed

[PATCH] drm/amdgpu: Set pm_display_cfg in non-dc mode

those parameters are needed by powerplay so we can remove the wrapper cgs interface for get display info Change-Id: I503decd322837f016a16d6c45602d3d3c9fd953d Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c | 21 +

RE: [PATCH] drm/amdgpu: no job timeout setting on compute queues

Besides, if some compute shader takes time more them 50 seconds, you can just set lockuptime out to 50s Why change the logic in kmd side ? I don’t think it’s a good idea to disable the time out message for compute ring, we have virtualization end-user Still want those message printed out Can

RE: [PATCH] drm/amdgpu: no job timeout setting on compute queues

Please don’t do it for SR-IOV /Monk From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Quan, Evan Sent: 2018年3月19日 9:47 To: Deucher, Alexander ; Michel Dänzer Cc: amd-gfx@lists.freedesktop.org Subject: RE: [PATCH]

[PATCH 2/2] drm/amd/pp: Use atombios interface directly in powerplay

In order to remove the cgs wrapper functions for atombios api. Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c | 145 +++-- drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h | 2 +-

[PATCH 1/2] drm/amd/dc: Use atombios api directly in DC

In order to remove the cgs wrapper functions for atombios api. Signed-off-by: Rex Zhu --- .../gpu/drm/amd/display/dc/bios/command_table.c| 22 +- .../gpu/drm/amd/display/dc/bios/command_table2.c | 26 +- 2 files changed, 28

Re: [PATCH] dma-buf: use parameter structure for dma_buf_attach

Am 26.03.2018 um 10:36 schrieb Daniel Vetter: On Sun, Mar 25, 2018 at 01:34:51PM +0200, Christian König wrote: [SNIP] - attach->dev = dev; + attach->dev = info->dev; attach->dmabuf = dmabuf; + attach->priv = info->priv; The ->priv field is for the exporter, not the

Re: [PATCH] dma-buf: use parameter structure for dma_buf_attach

Am 26.03.2018 um 10:36 schrieb Daniel Vetter: On Sun, Mar 25, 2018 at 01:34:51PM +0200, Christian König wrote: [SNIP] - attach->dev = dev; + attach->dev = info->dev; attach->dmabuf = dmabuf; + attach->priv = info->priv; The ->priv field is for the exporter, not the

Re: [PATCH 1/2] drm/amdgpu: re-validate per VM BOs if required

Am 26.03.2018 um 10:47 schrieb Chunming Zhou: From: Christian König If a per VM BO ends up in a allowed domain it never moves back into the prefered domain. Change-Id: Ifb3e561785d3b464da28c439b271c26825224c5e Signed-off-by: Christian König

Re: [PATCH 3/4] drm/amdgpu: Fix acquiring VM on large-BAR systems

Am 26.03.2018 um 03:31 schrieb Felix Kühling: Am 2018-03-24 um 04:34 AM schrieb Christian König: Am 23.03.2018 um 20:53 schrieb Felix Kuehling: On 2018-03-23 03:35 PM, Christian König wrote: Am 23.03.2018 um 20:30 schrieb Felix Kuehling: On large-BAR systems the VM page tables for compute

RE: [PATCH v2] drm/amdgpu: Parse module param amdgpu_dpm in check_arguments

Sorry, Please ignore this patch. Setting dpm_enabled when check_arguments didn't make sense. We will set dpm_enabled to true when smu hw init successfully. Best Regards Rex -Original Message- From: Quan, Evan Sent: Monday, March 26, 2018 4:36 PM To: Zhu, Rex;

Re: [PATCH] drm/amdgpu: Use dpm_enabled as the dpm state flag

On Mon, Mar 26, 2018 at 04:23:32PM +0800, Rex Zhu wrote: > driver will set dpm_enabled to true only when > module paramter amdgpu_dpm not set to 0 and > smu initialize successfully. > Reviewed-by: Huang Rui > Change-Id: Ic835dc93e0cfdfc2be7d10109d0c63b3cc4c2dff >

[PATCH 2/2] drm/amdgpu: add per vm bo validation order

Change-Id: I59f1da8a765efae8196c0b1599f47a8b8485c4d3 Signed-off-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 14 +- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 6 ++ 2 files changed, 19 insertions(+), 1 deletion(-) diff --git

[PATCH 1/2] drm/amdgpu: re-validate per VM BOs if required

From: Christian König If a per VM BO ends up in a allowed domain it never moves back into the prefered domain. Change-Id: Ifb3e561785d3b464da28c439b271c26825224c5e Signed-off-by: Christian König Reviewed-and-Tested-by: Chunming Zhou

RE: [PATCH] drm/amdgpu: Use dpm_enabled as the dpm state flag

Reviewed-by: Evan Quan > -Original Message- > From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf > Of Rex Zhu > Sent: Monday, March 26, 2018 4:24 PM > To: amd-gfx@lists.freedesktop.org > Cc: Zhu, Rex > Subject: [PATCH] drm/amdgpu:

[PATCH] drm/amdgpu: Use dpm_enabled as the dpm state flag

driver will set dpm_enabled to true only when module paramter amdgpu_dpm not set to 0 and smu initialize successfully. Change-Id: Ic835dc93e0cfdfc2be7d10109d0c63b3cc4c2dff Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 2 +-

Re: [PATCH 2/2] drm/amdgpu: add per vm bo validation order

pls ignore this one, will re-send it. On 2018年03月26日 16:29, Chunming Zhou wrote: Change-Id: I59f1da8a765efae8196c0b1599f47a8b8485c4d3 Signed-off-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 12 +++- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 6

[PATCH 2/2] drm/amdgpu: add per vm bo validation order

Change-Id: I59f1da8a765efae8196c0b1599f47a8b8485c4d3 Signed-off-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 12 +++- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 6 ++ 2 files changed, 17 insertions(+), 1 deletion(-) diff --git

[PATCH 1/2] drm/amdgpu: re-validate per VM BOs if required

From: Christian König If a per VM BO ends up in a allowed domain it never moves back into the prefered domain. Change-Id: Ifb3e561785d3b464da28c439b271c26825224c5e Signed-off-by: Christian König Reviewed-and-Tested-by: Chunming Zhou

[PATCH v2] drm/amdgpu: Parse module param amdgpu_dpm in check_arguments

v2: Patch for sriov-vf case, smu ip block not added to driver on vega, set dpm_enabled false. Change-Id: Ib68ffa14492419df444756b6a75ed7f52ccfb738 Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c

RE: [PATCH] amd/powerplay: enable the ACG SS feature

Hi Kenneth, Per my understanding, this can be splitted into three patches: 1. add new PPSMC_MSG 2. GNLD_ACG renamed as GNLD_ACG_DEBUG related. In fact, this change seem not so necessary. 3. smc_dpm structure change related Regards, Evan > -Original Message- > From: amd-gfx

Re: [Linaro-mm-sig] [PATCH 1/5] dma-buf: add optional invalidate_mappings callback v2

On Thu, Mar 22, 2018 at 10:58:55AM +0100, Christian König wrote: > Am 22.03.2018 um 08:18 schrieb Daniel Vetter: > > On Wed, Mar 21, 2018 at 12:54:20PM +0100, Christian König wrote: > > > Am 21.03.2018 um 09:28 schrieb Daniel Vetter: > > > > On Tue, Mar 20, 2018 at 06:47:57PM +0100, Christian

Re: [PATCH] drm/amdgpu: Parse module param amdgpu_dpm in check_arguments

On Mon, Mar 26, 2018 at 03:27:42PM +0800, Rex Zhu wrote: > Change-Id: If9c30a51942ca6544d1f6fa411622ea8900befc0 > Signed-off-by: Rex Zhu Acked-by: Huang Rui > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 2 +- >

Re: [PATCH] amd/powerplay: enable the ACG SS feature

On Mon, Mar 26, 2018 at 03:39:49PM +0800, Kenneth Feng wrote: > Change-Id: Id4bc19b0df93d96f38e5254246a885af8679c034 It would better to add the description for ACG SS feature such as the benefit after enabled it. > Signed-off-by: Kenneth Feng > --- >

Re: [Linaro-mm-sig] [PATCH 1/5] dma-buf: add optional invalidate_mappings callback v2

On Thu, Mar 22, 2018 at 10:37:55AM +0100, Christian König wrote: > Am 22.03.2018 um 08:14 schrieb Daniel Vetter: > > On Wed, Mar 21, 2018 at 10:34:05AM +0100, Christian König wrote: > > > Am 21.03.2018 um 09:18 schrieb Daniel Vetter: > > > > [SNIP] > > > For correct operation you always need to

[PATCH] drm/amdgpu: Parse module param amdgpu_dpm in check_arguments

Change-Id: If9c30a51942ca6544d1f6fa411622ea8900befc0 Signed-off-by: Rex Zhu --- drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c| 2 ++ drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 2 +- drivers/gpu/drm/amd/amdgpu/ci_dpm.c

[PATCH] amd/powerplay: enable the ACG SS feature

Change-Id: Id4bc19b0df93d96f38e5254246a885af8679c034 Signed-off-by: Kenneth Feng --- drivers/gpu/drm/amd/include/atomfirmware.h | 12 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c | 10 +++--- drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h