RE: [PATCH 2/2] drm/amdgpu: disable vm fault irq during prt accessed

2019-01-03 Thread Zhou, David(ChunMing)
Seems we don't need two page table, we just map every prt range to dummy page, any problem? -David > -Original Message- > From: Zhou, David(ChunMing) > Sent: Thursday, January 03, 2019 5:23 PM > To: Koenig, Christian ; amd- > g...@lists.freedesktop.org > Subject: RE: [PATCH 2/2]

Re: [PATCH 2/2] drm/amdgpu: disable vm fault irq during prt accessed

2019-01-03 Thread Koenig, Christian
On pre-gmc9 PTE prt bit is completely ignored. Check the OpenGL spec for PRTs, writes needs to be ignored and reads must return zero. Not sure about Vulkan and OpenCL. Christian. Am 03.01.19 um 10:56 schrieb Zhou, David(ChunMing): > Doesn't gpu check PTE prt bit first and then access va

Re: [PATCH 2/2] drm/amdgpu: disable vm fault irq during prt accessed

2019-01-03 Thread Christian König
Yes, exactly. Problem is that we then probably need two page tables. One for the CB/TC and one for the SDMA. Christian. Am 03.01.19 um 10:02 schrieb zhoucm1: need dummy page for that? -David On 2019年01月03日 17:01, Christian König wrote: NAK, the problem is not the interrupt. E.g.

Re: [PATCH 2/2] drm/amdgpu: set the executable flag on unused Vega10 PTEs

2019-01-03 Thread Christian König
Oh, good point. Completely forgot about those. Going to fix the patch, Christian. Am 02.01.19 um 21:26 schrieb Kuehling, Felix: Don't you need to update amdgpu_vm_clear_freed as well? Regards,   Felix On 2019-01-02 8:35 a.m., Christian König wrote: Otherwise we run into a non-retry fault

RE: [PATCH 2/2] drm/amdgpu: disable vm fault irq during prt accessed

2019-01-03 Thread Zhou, David(ChunMing)
> -Original Message- > From: Christian König > Sent: Thursday, January 03, 2019 5:05 PM > To: Zhou, David(ChunMing) ; Koenig, Christian > ; Zhou, David(ChunMing) > ; amd-gfx@lists.freedesktop.org > Subject: Re: [PATCH 2/2] drm/amdgpu: disable vm fault irq during prt > accessed > > Yes,

Re: [PATCH 2/2] drm/amdgpu: disable vm fault irq during prt accessed

2019-01-03 Thread Christian König
Writes are then not ignored and garbage the dummy page. Christian. Am 03.01.19 um 10:46 schrieb Zhou, David(ChunMing): Seems we don't need two page table, we just map every prt range to dummy page, any problem? -David -Original Message- From: Zhou, David(ChunMing) Sent: Thursday,

Re: [PATCH] drm/amdkfd: Allow building KFD on ARM64

2019-01-03 Thread Christian König
Am 03.01.19 um 04:42 schrieb Alex Deucher: On Wed, Jan 2, 2019 at 5:26 PM Kuehling, Felix wrote: ifdef x86_64 specific code. Allow enabling CONFIG_HSA_AMD on ARM64. CC: Mark Nutter Signed-off-by: Felix Kuehling Acked-by: Alex Deucher Nice to see that we finally pick up some ARM support.

Re: [PATCH 2/2] drm/amdgpu: disable vm fault irq during prt accessed

2019-01-03 Thread zhoucm1
need dummy page for that? -David On 2019年01月03日 17:01, Christian König wrote: NAK, the problem is not the interrupt. E.g. causing faults by accessing unmapped pages with the SDMA can still crash the MC. The key point is that SDMA can't work with PRT tiles on pre-gmc9 and we need to

Re: [PATCH 2/2] drm/amdgpu: disable vm fault irq during prt accessed

2019-01-03 Thread Christian König
NAK, the problem is not the interrupt. E.g. causing faults by accessing unmapped pages with the SDMA can still crash the MC. The key point is that SDMA can't work with PRT tiles on pre-gmc9 and we need to forbid access on the application side. Regards, Christian. Am 03.01.19 um 09:54

RE: [PATCH 2/2] drm/amdgpu: disable vm fault irq during prt accessed

2019-01-03 Thread Zhou, David(ChunMing)
Doesn't gpu check PTE prt bit first and then access va range? Even wrte to dummy page, seem there still is no problem, we don't care that content at all. -David > -Original Message- > From: Christian König > Sent: Thursday, January 03, 2019 5:54 PM > To: Zhou, David(ChunMing) ;

Re: [PATCH 3/3] drm/amdgpu: replace get_user_pages with HMM address mirror helpers v5

2019-01-03 Thread Christian König
I'm just getting up to speed again after winter vacation and have tons of mails to dig through and a few important personal appointments coming as well. From skimming over it looks like it should work, but please give me at least till Monday to take a closer look. Christian. Am 02.01.19 um

[PATCH 2/2] drm/amdgpu: disable vm fault irq during prt accessed

2019-01-03 Thread Chunming Zhou
For pre-gmc9, UMD can only access unmapped PRT tile from CB/TC without firing VM fault. Kernel would still receive the VM fault interrupt and output the error message if SDMA is the mc_client. GMC9 don't need the same since it handle the PRT in different way. We cannot just skip message for SDMA,

Re: [PATCH v5 1/2] drm/sched: Refactor ring mirror list handling.

2019-01-03 Thread Koenig, Christian
Am 21.12.18 um 21:36 schrieb Grodzovsky, Andrey: > > On 12/21/2018 01:37 PM, Christian König wrote: >> Am 20.12.18 um 20:23 schrieb Andrey Grodzovsky: >>> Decauple sched threads stop and start and ring mirror >>> list handling from the policy of what to do about the >>> guilty jobs. >>> When

[PATCH 1/2] drm/amdgpu: fix prt_warning not balance

2019-01-03 Thread Chunming Zhou
use DRM_WARN instead of dev_warn Change-Id: I8ede212b5f99b1f77d0bd7499da3d85873403c7a Signed-off-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c | 6 +- drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c | 6 +- drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c | 6 +- 3 files changed, 15

Re: [PATCH 2/2] drm/amdgpu: disable vm fault irq during prt accessed

2019-01-03 Thread Koenig, Christian
Am 03.01.19 um 10:23 schrieb Zhou, David(ChunMing): > >> -Original Message- >> From: Christian König >> Sent: Thursday, January 03, 2019 5:05 PM >> To: Zhou, David(ChunMing) ; Koenig, Christian >> ; Zhou, David(ChunMing) >> ; amd-gfx@lists.freedesktop.org >> Subject: Re: [PATCH 2/2]

[PATCH 1/2] drm/amdgpu: set the executable flag on unused Vega10 PTEs v2

2019-01-03 Thread Christian König
Otherwise we run into a non-retry fault on access. It seems to be a hardware bug that the executable bit has higher priority than the valid bit. v2: handle clears as well Signed-off-by: Christian König --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 36 ++ 1 file changed,

[PATCH 2/2] drm/amdgpu: fix unused variable warnings

2019-01-03 Thread Christian König
Just remove some leftovers. Signed-off-by: Christian König --- drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 164ffc91b645..57cb3a51bda7 100644 ---

Re: ATI FirePro 2260 doesn't work after the update 'drm-next-2018-12-14'

2019-01-03 Thread Michel Dänzer
On 2019-01-01 4:03 p.m., Christian Zigotzky wrote: > Hello, > > An end user reportet that his ATI FirePro 2260 doesn't work anymore with > the latest Git kernel. We are a first level Linux support and we usually > compile Git kernels during the merge window time for testing. After the > update

Re: Regression with kernel 4.20 on armhf

2019-01-03 Thread Alex Deucher
Does this patch help by any chance? https://cgit.freedesktop.org/~agd5f/linux/commit/?h=amd-staging-drm-next=5e01c09ce3b7263d88873105f21a82eda904664b Alex On Thu, Jan 3, 2019 at 7:14 AM Luís Mendes wrote: > > Hi Christian, Alex, > > I've set the kernel command line with drm.debug=0xf, and I

Re: Regression with kernel 4.20 on armhf

2019-01-03 Thread Luís Mendes
Hi Alex, I've made a backport of that patch to linux 4.20, but didn't notice any improvements. It looks like, from the logs with drm debug messages, that the issue is happening after the ring tests, while entering X session. Thanks for your suggestion, anyway. I'll try that again once

Re: [PATCH v5 1/2] drm/amd: validate user pitch alignment

2019-01-03 Thread Michel Dänzer
On 2018-12-30 2:00 a.m., Yu Zhao wrote: > Userspace may request pitch alignment that is not supported by GPU. > Some requests 32, but GPU ignores it and uses default 64 when cpp is > 4. If GEM object is allocated based on the smaller alignment, GPU > DMA will go out of bound. > > Cc:

Re: [PATCH v5 1/2] drm/sched: Refactor ring mirror list handling.

2019-01-03 Thread Grodzovsky, Andrey
On 01/03/2019 11:20 AM, Grodzovsky, Andrey wrote: > > On 01/03/2019 03:54 AM, Koenig, Christian wrote: >> Am 21.12.18 um 21:36 schrieb Grodzovsky, Andrey: >>> On 12/21/2018 01:37 PM, Christian König wrote: Am 20.12.18 um 20:23 schrieb Andrey Grodzovsky: > Decauple sched threads stop and

Re: [PATCH 1/2] drm/amdgpu: set the executable flag on unused Vega10 PTEs v2

2019-01-03 Thread Kuehling, Felix
On 2019-01-03 4:59 a.m., Christian König wrote: > Otherwise we run into a non-retry fault on access. > > It seems to be a hardware bug that the executable bit has > higher priority than the valid bit. > > v2: handle clears as well > > Signed-off-by: Christian König > --- >

Re: [PATCH v5 1/2] drm/sched: Refactor ring mirror list handling.

2019-01-03 Thread Grodzovsky, Andrey
On 01/03/2019 03:54 AM, Koenig, Christian wrote: > Am 21.12.18 um 21:36 schrieb Grodzovsky, Andrey: >> On 12/21/2018 01:37 PM, Christian König wrote: >>> Am 20.12.18 um 20:23 schrieb Andrey Grodzovsky: Decauple sched threads stop and start and ring mirror list handling from the policy

Re: [PATCH] drm/amdgpu/sriov: For finishing routine send rel event after init failed

2019-01-03 Thread Deucher, Alexander
Reviewed-by: Alex Deucher From: amd-gfx on behalf of Emily Deng Sent: Wednesday, January 2, 2019 10:58:37 PM To: amd-gfx@lists.freedesktop.org Cc: Deng, Emily Subject: [PATCH] drm/amdgpu/sriov: For finishing routine send rel event after init failed When init

Re: [PATCH 2/2] drm/amdgpu/psp: make get_fw_type and prep_cmd_buf to be common interfaces

2019-01-03 Thread Deucher, Alexander
Series is: Reviewed-by: Alex Deucher From: amd-gfx on behalf of Hawking Zhang Sent: Thursday, January 3, 2019 9:00:44 AM To: amd-gfx@lists.freedesktop.org Cc: Zhang, Hawking Subject: [PATCH 2/2] drm/amdgpu/psp: make get_fw_type and prep_cmd_buf to be common

Re: [PATCH] drm/amdgpu: Cleanup 2 compiler warnings

2019-01-03 Thread Alex Deucher
On Thu, Jan 3, 2019 at 8:37 AM Russell, Kent wrote: > > These 2 variables are unused now, so remove their references. > > Fixes: e4ae0fc drm/amdgpu: implement gfx8 post_soft_reset > Fixes: 5e01c09 drm/amdgpu/gfx_v8_0: Reorder the gfx, kiq and kcq rings > test sequence > > Change-Id:

Re: [PATCH] drm/amd/powerplay: hint and error out when prerequisite not meet

2019-01-03 Thread Deucher, Alexander
Reviewed-by: Alex Deucher From: amd-gfx on behalf of Evan Quan Sent: Thursday, January 3, 2019 1:06:56 AM To: amd-gfx@lists.freedesktop.org Cc: Quan, Evan Subject: [PATCH] drm/amd/powerplay: hint and error out when prerequisite not meet Do not ignore the

Re: [PATCH 2/2] drm/amdgpu/psp: Fix can't detect psp INVOKE command failed

2019-01-03 Thread Christian König
Hi guys, this patch broke loading older versions of PSP firmware on Vega10. What exactly is the background here? E.g. why do we need it? And can we revert it ASAP? Thanks, Christian. Am 18.12.18 um 03:45 schrieb Xu, Feifei: Reviewed-by: Feifei Xu -Original Message- From: amd-gfx

RE: [PATCH 2/2] drm/amdgpu/psp: Fix can't detect psp INVOKE command failed

2019-01-03 Thread Yu, Xiangliang
XGMI command will not load ucode, so it can't find the command failed if check ucode at first. For code logic, should check response status to see if command complete successfully at first. For Vega10, could you explain why it block older version fw? -Original Message- From: Christian

Re: [PATCH 2/2] drm/amdgpu/psp: Fix can't detect psp INVOKE command failed

2019-01-03 Thread Koenig, Christian
> For Vega10, could you explain why it block older version fw? I'm not very familiar with the PSP, so I don't know what exactly is going wrong here. Symptoms are that we get the following in dmesg: > Jan 03 15:55:17 abel kernel: [drm] PSP loading VCE firmware > Jan 03 15:55:17 abel kernel: [drm]

[PATCH] drm/amdgpu: Cleanup 2 compiler warnings

2019-01-03 Thread Russell, Kent
These 2 variables are unused now, so remove their references. Fixes: e4ae0fc drm/amdgpu: implement gfx8 post_soft_reset Fixes: 5e01c09 drm/amdgpu/gfx_v8_0: Reorder the gfx, kiq and kcq rings test sequence Change-Id: Ic3f39c48b8deef2ff81589c31ce1cc24573b47eb Signed-off-by: Kent Russell ---

[PATCH 1/2] drm/amdgpu/psp: update the naming of GFX_FW_TYPE_RLC_RESTORE_LIST_CNTL

2019-01-03 Thread Hawking Zhang
GFX_FW_TYPE_RLC_RESTORE_LIST_CNTL was reanmed to GFX_FW_TYPE_RLC_RESTORE_LIST_SRM_CNTL in latest psp_gfx_if drop Change-Id: Ie9ac5b640a3d5c0faa8f1b7700cb0f85d6a6002a Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h | 2 +- drivers/gpu/drm/amd/amdgpu/psp_v10_0.c | 2 +-

[PATCH 2/2] drm/amdgpu/psp: make get_fw_type and prep_cmd_buf to be common interfaces

2019-01-03 Thread Hawking Zhang
get_fw_type and prep_cmd_buf should be common interface instead of IP specific ones Change-Id: I115a8d3fafdbe143008b3698ad3a5f3bd4b87481 Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 94 - drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h | 3

Re: [PATCH 2/2] drm/amdgpu/psp: Fix can't detect psp INVOKE command failed

2019-01-03 Thread Deucher, Alexander
This bug may also be related: https://bugs.freedesktop.org/show_bug.cgi?id=109206 Alex From: Koenig, Christian Sent: Thursday, January 3, 2019 10:57:50 AM To: Yu, Xiangliang; Xu, Feifei; amd-gfx@lists.freedesktop.org; Deucher, Alexander Subject: Re: [PATCH

Re: ATI FirePro 2260 doesn't work after the update 'drm-next-2018-12-14'

2019-01-03 Thread Christian Zigotzky
It‘s an end user. That means bisecting isn‘t possible. — Christian Sent from my iPhone > On 3. Jan 2019, at 12:03, Michel Dänzer wrote: > >> On 2019-01-01 4:03 p.m., Christian Zigotzky wrote: >> Hello, >> >> An end user reportet that his ATI FirePro 2260 doesn't work anymore with >> the

Re: [PATCH] drm/amd/display: Fix boolean expression in get_surf_rq_param

2019-01-03 Thread Wentland, Harry
On 2019-01-03 2:48 p.m., Gustavo A. R. Silva wrote: > Fix boolean expression by using logical AND operator '&&' > instead of bitwise operator '&'. > > This issue was detected with the help of Coccinelle. > > Fixes: 6d04ee9dc101 ("drm/amd/display: Restructuring and cleaning up DML") > Cc:

[PATCH] drm/amdgpu_vm: fix boolean expressions

2019-01-03 Thread Gustavo A. R. Silva
Fix boolean expressions by using logical AND operator '&&' instead of bitwise operator '&'. This issue was detected with the help of Coccinelle. Fixes: c8c5e569c5b0 ("drm/amdgpu: Consolidate visible vs. real vram check v2.") Cc: sta...@vger.kernel.org Signed-off-by: Gustavo A. R. Silva ---

Re: [PATCH] drm/amdgpu_vm: fix boolean expressions

2019-01-03 Thread Kuehling, Felix
On 2019-01-03 12:34 p.m., Gustavo A. R. Silva wrote: > Fix boolean expressions by using logical AND operator '&&' > instead of bitwise operator '&'. > > This issue was detected with the help of Coccinelle. > > Fixes: c8c5e569c5b0 ("drm/amdgpu: Consolidate visible vs. real vram check > v2.")

Re: [PATCH] drm/amd/display: Add below the range support for FreeSync

2019-01-03 Thread Wentland, Harry
On 2018-12-05 8:40 a.m., Nicholas Kazlauskas wrote: > [Why] > When application flip-rate is below the minimum vrr refresh rate. > > Variable refresh rate monitors extend the front porch duration until > flip or timeout occurs. For cases where the application flip-rate is > > When the flip-rate

[PATCH v2] drm/amdgpu_vm: fix boolean expressions

2019-01-03 Thread Gustavo A. R. Silva
Fix boolean expressions by using logical AND operator '&&' instead of bitwise operator '&'. This issue was detected with the help of Coccinelle. Fixes: 9a4b7d4c769e ("drm/amdgpu: Add vm context module param") Cc: sta...@vger.kernel.org Reviewed-by: Felix Kuehling Signed-off-by: Gustavo A. R.

Re: [PATCH] drm/amdgpu_vm: fix boolean expressions

2019-01-03 Thread Gustavo A. R. Silva
On 1/3/19 1:17 PM, Kuehling, Felix wrote: On 2019-01-03 12:34 p.m., Gustavo A. R. Silva wrote: Fix boolean expressions by using logical AND operator '&&' instead of bitwise operator '&'. This issue was detected with the help of Coccinelle. Fixes: c8c5e569c5b0 ("drm/amdgpu: Consolidate

[PATCH] drm/amd/display: Fix boolean expression in get_surf_rq_param

2019-01-03 Thread Gustavo A. R. Silva
Fix boolean expression by using logical AND operator '&&' instead of bitwise operator '&'. This issue was detected with the help of Coccinelle. Fixes: 6d04ee9dc101 ("drm/amd/display: Restructuring and cleaning up DML") Cc: sta...@vger.kernel.org Signed-off-by: Gustavo A. R. Silva ---