RE: [PATCH 1/3] drm/amdgpu: add amdgpu_mmhub_funcs definition

For mmhub/gfxhub, currently we didn't organize both in callback functions structure, instead, we declared all the functions in its header file respectively (see mmhub_vxx_x.h/gfxhub_vxx_x.h). We'd either centralize all the functions into callback function structure, or continue declare the

RE: [PATCH] drm/amdgpu: add check to avoid array bound issue

Reviewed-by: Tao Zhou > -Original Message- > From: amd-gfx On Behalf Of > Guchun Chen > Sent: 2019年8月8日 14:59 > To: amd-gfx@lists.freedesktop.org; Zhang, Hawking > ; Li, Dennis ; Pan, Xinhui > ; Zhou1, Tao > Cc: Chen, Guchun > Subject: [PATCH] drm/amdgpu: add check to avoid array

RE: [PATCH] drm/amdgpu: add sub block parameter in ras inject command

Reviewed-by: Guchun Chen -Original Message- From: amd-gfx On Behalf Of Tao Zhou Sent: Thursday, August 8, 2019 2:48 PM To: amd-gfx@lists.freedesktop.org; Zhang, Hawking ; Chen, Guchun ; Li, Dennis ; Pan, Xinhui Cc: Zhou1, Tao Subject: [PATCH] drm/amdgpu: add sub block parameter in

Re: [PATCH] nouveau/hmm: map pages after migration

On Wed, Aug 07, 2019 at 08:02:14AM -0700, Ralph Campbell wrote: > When memory is migrated to the GPU it is likely to be accessed by GPU > code soon afterwards. Instead of waiting for a GPU fault, map the > migrated memory into the GPU page tables with the same access permissions > as the source

Re: [PATCH 04/15] mm: remove the pgmap field from struct hmm_vma_walk

On Wed, Aug 07, 2019 at 11:47:22AM -0700, Dan Williams wrote: > > Unrelated to this patch, but what is the point of getting checking > > that the pgmap exists for the page and then immediately releasing it? > > This code has this pattern in several places. > > > > It feels racy > > Agree, not

RE: [PATCH 1/3] drm/amdgpu: add amdgpu_mmhub_funcs definition

Mmhub ras function will be called in amdgpu_ras.c, amdgpu_ras.c should pay no attention to the version of mmhub functions. Centralize all the functions into callback function structure is a good idea, but this is a another story and needs a huge effort. Regards, Tao > -Original

RE: [Regression] "drm/amdgpu: enable gfxoff again on raven series (v2)"

> -Original Message- > From: Kai-Heng Feng > Sent: Thursday, August 08, 2019 1:45 AM > To: Huang, Ray > Cc: Deucher, Alexander ; Koenig, Christian > ; Zhou, David(ChunMing) > ; amd-gfx list ; > dri-de...@lists.freedesktop.org; LKML ; > Anthony Wong > Subject: Re: [Regression]

RE: [PATCH 2/2] drm/amd/powerplay: update smu11_driver_if_navi10.h

Series is Reviewed-by: Hawking Zhang Regards, Hawking -Original Message- From: amd-gfx On Behalf Of Tianci Yin Sent: 2019年8月8日 13:49 To: amd-gfx@lists.freedesktop.org Cc: Xiao, Jack ; Wang, Kevin(Yang) ; Yin, Tianci (Rico) ; Zhang, Hawking ; Quan, Evan ; Yuan, Xiaojie Subject:

Re: [Regression] "drm/amdgpu: enable gfxoff again on raven series (v2)"

at 14:29, Huang, Ray wrote: -Original Message- From: Kai-Heng Feng Sent: Thursday, August 08, 2019 1:45 AM To: Huang, Ray Cc: Deucher, Alexander ; Koenig, Christian ; Zhou, David(ChunMing) ; amd-gfx list ; dri-de...@lists.freedesktop.org; LKML ; Anthony Wong Subject: Re:

[PATCH] drm/amdgpu: add sub block parameter in ras inject command

ras sub block index could be passed from shell command Signed-off-by: Tao Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 17 ++--- 1 file changed, 10 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c

[PATCH 1/3] drm/amdgpu: add amdgpu_mmhub_funcs definition

add amdgpu_mmhub_funcs definition and initialize it, prepare for mmhub ras enablement Signed-off-by: Tao Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 2 ++ drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h | 31 +++ drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 12 +

[PATCH 2/3] drm/amdgpu: support mmhub ras in amdgpu ras

call mmhub ras query/inject in amdgpu ras Signed-off-by: Tao Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 5 + 1 file changed, 5 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index bc766cdbeaae..be59102526a1 100644 ---

[PATCH 3/3] drm/amdgpu: create mmhub ras framework

enable mmhub ras feature and create sysfs/debugfs node for mmhub Signed-off-by: Tao Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h | 3 +- drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 112 +--- 2 files changed, 82 insertions(+), 33 deletions(-) diff --git

[PATCH] drm/amdgpu: add check to avoid array bound issue

Sub_block_index can be passed from user level, so add one check before accessing the array first to prevent array index out of bound problem. Change-Id: I556fc560b44215848ba3a95f757febdf3d0af422 Signed-off-by: Guchun Chen --- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 3 +++ 1 file changed, 3

Re: [PATCH 2/2] drm/amd/powerplay: update smu11_driver_if_navi10.h

Thanks Hawking! From: Zhang, Hawking Sent: Thursday, August 8, 2019 14:44 To: Yin, Tianci (Rico) ; amd-gfx@lists.freedesktop.org Cc: Xiao, Jack ; Wang, Kevin(Yang) ; Yin, Tianci (Rico) ; Quan, Evan ; Yuan, Xiaojie Subject: RE: [PATCH 2/2] drm/amd/powerplay:

Re: [Regression] "drm/amdgpu: enable gfxoff again on raven series (v2)"

On 2019-08-08 8:29 a.m., Huang, Ray wrote: >> From: Kai-Heng Feng >> at 00:03, Huang, Ray wrote: >> >>> May I know the all firmware version in your system? > > Seems to the issue we encountered with IOMMU enabled. Could you please > disable iommu in SBIOS or GRUB? The driver needs to work

Re: The issue with page allocation 5.3 rc1-rc2 (seems drm culprit here)

On 2019-08-08 7:31 a.m., Alex Deucher wrote: > On Wed, Aug 7, 2019 at 11:49 PM Mikhail Gavrilov > wrote: >> >> Unfortunately error "gnome-shell: page allocation failure: order:4, >> mode:0x40cc0(GFP_KERNEL|__GFP_COMP), >> nodemask=(null),cpuset=/,mems_allowed=0" still happens even with >>

RE: [Regression] "drm/amdgpu: enable gfxoff again on raven series (v2)"

> -Original Message- > From: Michel Dänzer > Sent: Thursday, August 08, 2019 4:10 PM > To: Huang, Ray ; Kai-Heng Feng > > Cc: Zhou, David(ChunMing) ; LKML ker...@vger.kernel.org>; dri-de...@lists.freedesktop.org; Anthony Wong > ; amd-gfx list g...@lists.freedesktop.org>; Deucher,

RE: [PATCH] drm/amdgpu: add sub block parameter in ras inject command

Reviewed-by: Hawking Zhang Regards, Hawking -Original Message- From: amd-gfx On Behalf Of Tao Zhou Sent: 2019年8月8日 14:48 To: amd-gfx@lists.freedesktop.org; Zhang, Hawking ; Chen, Guchun ; Li, Dennis ; Pan, Xinhui Cc: Zhou1, Tao Subject: [PATCH] drm/amdgpu: add sub block parameter in

RE: [PATCH 1/3] drm/amdgpu: add amdgpu_mmhub_funcs definition

Okay, we can have mmhub ras function go first and find another timeslot to refine those two huge blocks. The series is Reviewed-by: Hawking Zhang Regards, Hawking -Original Message- From: Zhou1, Tao Sent: 2019年8月8日 15:58 To: Zhang, Hawking ; amd-gfx@lists.freedesktop.org; Chen,

libdrm patch merge request

Hi Alex, Would you mind merging attached 3 patches to libdrm master branch? These changes are implemented for gfx and umc ras inject unit test by amdgpu_test. Thanks a lot. Regards, Guchun 0001-amdgpu-add-gfx-ras-inject-configuration-file.patch Description:

RE: [PATCH] drm/amdgpu: update ras sysfs feature info

Yeah, that should work as well. The extra text is a bit questionable, but I think we can live with that as well. Christian. Am 08.08.2019 17:48 schrieb "Zhang, Hawking" : Not exactly, there will be only one line left feature mask: 0x3ffb Regards, Hawking -Original Message- From:

RE: [PATCH 1/5] drm/amdgpu: Extends amdgpu vm definitions

+@Mirza, Jimshed: do we have the ability to snoop another GPU's cache, in a multiple GPU system? Regards, Oak From: Deucher, Alexander Sent: Thursday, August 8, 2019 12:10 PM To: Zeng, Oak ; Koenig, Christian ; amd-gfx@lists.freedesktop.org Cc: Kuehling, Felix ;

RE: [PATCH] drm/amdgpu: update ras sysfs feature info

Hi Chris, I'm not aware of how ROCM SMI using feature nodes. but not all the sysfs are intended to be used by upper level apps/libs. There are bunches of sysfs entries that have multiple line value. The most complicate one would be pp_power_profile_mode, which looks like. 0 BOOTUP_DEFAULT*:

Re: [PATCH] drm/amdgpu: update ras sysfs feature info

Hi Hawking, a multi line value is not the problem, but here you have multiple values. E.g. in the case of the pp tables that is one big array of power profiles and we actually had a discussion if exposing them like this is ok or not. But in the case of the ras features you got multiple

Re: [PATCH 1/5] drm/amdgpu: Extends amdgpu vm definitions

The snoop bit is for snooping the CPU cache by the GPU when doing system memory mappings. Alex From: amd-gfx on behalf of Zeng, Oak Sent: Thursday, August 8, 2019 12:02 PM To: Koenig, Christian ; amd-gfx@lists.freedesktop.org Cc: Kuehling, Felix ; Keely,

Re: [Regression] "drm/amdgpu: enable gfxoff again on raven series (v2)"

Maybe we can check if iommu is enabled in the driver and do the appropriate thing with respect to gfxoff, assuming this is not platform specific. Alex From: Huang, Ray Sent: Thursday, August 8, 2019 4:14 AM To: Michel Dänzer ; Kai-Heng Feng ; Olsak, Marek Cc:

RE: [PATCH] drm/amdgpu: update ras sysfs feature info

Hi Chris, I thought I've stated as "The feature mask is already good enough for this node". It is just a uint32 value. We don't expect ROCm SMI to read any other information from feature node except for the feature mask. Regards, Hawking -Original Message- From: Koenig, Christian

Re: [PATCH] drm/amdgpu: update ras sysfs feature info

Hi Hawking, yeah, but this isn't sufficient to comply to the upstream rules. See what we need to do is to remove the extra text and the per IP information. In other words something like this: diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index

Re: [PATCH 1/5] drm/amdgpu: Extends amdgpu vm definitions

Yeah, exactly. This is controlling PCIe bus request behavior and routing. That is something we never ever want to expose to userspace. Christian. Am 08.08.19 um 18:09 schrieb Deucher, Alexander: The snoop bit is for snooping the CPU cache by the GPU when doing system memory mappings. Alex

Re: The issue with page allocation 5.3 rc1-rc2 (seems drm culprit here)

On Thu, Aug 8, 2019 at 4:13 AM Michel Dänzer wrote: > > On 2019-08-08 7:31 a.m., Alex Deucher wrote: > > On Wed, Aug 7, 2019 at 11:49 PM Mikhail Gavrilov > > wrote: > >> > >> Unfortunately error "gnome-shell: page allocation failure: order:4, > >> mode:0x40cc0(GFP_KERNEL|__GFP_COMP), > >>

Re: [PATCH] drm/amdgpu: update ras sysfs feature info

Hi Hawking, looks like you skipped my response. Even the current way how sysfs is used in the ras code is a clear NO-GO and should be fixed before this is pushed upstream. A sysfs entry should seriously NOT return a multi line value which needs to be extensively parsed by the application.

RE: [PATCH] drm/amdgpu: update ras sysfs feature info

Hi Hawking, I don't want to divert from this important discussion, but would like clarification... In the rocm-smi library, I previously had this interpretation of the values (I think someone had confirmed this): none, //!< No current errors disabled //!< ECC is disabled parity //!<

RE: [PATCH 1/5] drm/amdgpu: Extends amdgpu vm definitions

Hi Christian, My understanding of the snoop bit (C bit in the PTE definition) is to probe remote gpu's L2 cache after this gpu write remote gpu's vram. Is this correct? I am still checking this point with HW engineer. If this is correct, then the snooping (or probing) is a way to maintain

Re: libdrm patch merge request

Done! Alex On Thu, Aug 8, 2019 at 5:18 AM Chen, Guchun wrote: > > Hi Alex, > > > > Would you mind merging attached 3 patches to libdrm master branch? > > These changes are implemented for gfx and umc ras inject unit test by > amdgpu_test. > > Thanks a lot. > > > > Regards, > > Guchun > > > >

Re: [PATCH 00/34] put_user_pages(): miscellaneous call sites

On 8/8/19 9:25 AM, Weiny, Ira wrote: >> >> On 8/7/19 7:36 PM, Ira Weiny wrote: >>> On Wed, Aug 07, 2019 at 10:46:49AM +0200, Michal Hocko wrote: On Wed 07-08-19 10:37:26, Jan Kara wrote: > On Fri 02-08-19 12:14:09, John Hubbard wrote: >> On 8/2/19 7:52 AM, Jan Kara wrote: >>> On

RE: [PATCH 00/34] put_user_pages(): miscellaneous call sites

> > On 8/7/19 7:36 PM, Ira Weiny wrote: > > On Wed, Aug 07, 2019 at 10:46:49AM +0200, Michal Hocko wrote: > >> On Wed 07-08-19 10:37:26, Jan Kara wrote: > >>> On Fri 02-08-19 12:14:09, John Hubbard wrote: > On 8/2/19 7:52 AM, Jan Kara wrote: > > On Fri 02-08-19 07:24:43, Matthew Wilcox

Re: next/master build: 230 builds: 5 failed, 225 passed, 6 errors, 1344 warnings (next-20190805)

On Mon, Aug 05, 2019 at 12:12:05PM +0100, Mark Brown wrote: > On Mon, Aug 05, 2019 at 02:40:32AM -0700, kernelci.org bot wrote: > > Today's -next fails to build an arm allmodconfig due to: > > > allmodconfig (arm, gcc-8) — FAIL, 2 errors, 16 warnings, 0 section > > mismatches > > > > Errors: >

Re: drm/amdgpu: replace readq/writeq with atomic64 operations

/gpu/drm/amd/amdgpu/amdgpu_device.c: In function ‘amdgpu_mm_wreg64’: drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:298:3: error: implicit declaration of function ‘writeq’ This is with next-20190808. Guenter ___ amd-gfx mailing list amd-gfx@lists.freedeskt

Re: [PATCH v19 00/15] arm64: untag user pointers passed to the kernel

On Wed, Aug 07, 2019 at 07:17:35PM +0200, Andrey Konovalov wrote: > On Tue, Aug 6, 2019 at 7:13 PM Will Deacon wrote: > > > > On Wed, Jul 24, 2019 at 03:20:59PM +0100, Will Deacon wrote: > > > On Wed, Jul 24, 2019 at 04:16:49PM +0200, Andrey Konovalov wrote: > > > > On Wed, Jul 24, 2019 at 4:02

Re: [PATCH] nouveau/hmm: map pages after migration

On 8/8/19 12:07 AM, Christoph Hellwig wrote: On Wed, Aug 07, 2019 at 08:02:14AM -0700, Ralph Campbell wrote: When memory is migrated to the GPU it is likely to be accessed by GPU code soon afterwards. Instead of waiting for a GPU fault, map the migrated memory into the GPU page tables with the

Re: drm/amdgpu: replace readq/writeq with atomic64 operations

reg64’: > drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:298:3: error: implicit declaration > of function ‘writeq’ > > This is with next-20190808. > > Guenter > ___ > amd-gfx mailing list > amd-gfx@lists.freedesktop.org > https

Re: [PATCH v19 00/15] arm64: untag user pointers passed to the kernel

On Thu, 8 Aug 2019 14:12:19 -0700 Kees Cook wrote: > > The ones that are left are the mm ones: 4, 5, 6, 7 and 8. > > > > Andrew, could you take a look and give your Acked-by or pick them up > > directly? > > Given the subsystem Acks, it seems like 3-10 and 12 could all just go > via Andrew? I

Re: next/master build: 230 builds: 5 failed, 225 passed, 6 errors, 1344 warnings (next-20190805)

On Thu, Aug 8, 2019 at 2:53 PM Guenter Roeck wrote: > > On Mon, Aug 05, 2019 at 12:12:05PM +0100, Mark Brown wrote: > > On Mon, Aug 05, 2019 at 02:40:32AM -0700, kernelci.org bot wrote: > > > > Today's -next fails to build an arm allmodconfig due to: > > > > > allmodconfig (arm, gcc-8) — FAIL, 2

Re: next/master build: 230 builds: 5 failed, 225 passed, 6 errors, 1344 warnings (next-20190805)

On 8/8/19 11:53 AM, Alex Deucher wrote: On Thu, Aug 8, 2019 at 2:53 PM Guenter Roeck wrote: On Mon, Aug 05, 2019 at 12:12:05PM +0100, Mark Brown wrote: On Mon, Aug 05, 2019 at 02:40:32AM -0700, kernelci.org bot wrote: Today's -next fails to build an arm allmodconfig due to: allmodconfig

RE: [PATCH] drm/amdgpu: update ras sysfs feature info

Not exactly, there will be only one line left feature mask: 0x3ffb Regards, Hawking -Original Message- From: Freehill, Chris Sent: 2019年8月8日 23:28 To: Koenig, Christian ; Zhang, Hawking ; Russell, Kent ; Zhou1, Tao ; amd-gfx@lists.freedesktop.org; Pan, Xinhui Subject: RE: [PATCH]

Re: The issue with page allocation 5.3 rc1-rc2 (seems drm culprit here)

On Thu, 8 Aug 2019 13:32:06 +0800 Alex Deucher wrote: > > On Wed, Aug 7, 2019 at 11:49 PM Mikhail Gavrilov wrote: > > > > Unfortunately error "gnome-shell: page allocation failure: order:4, > > mode:0x40cc0(GFP_KERNEL|__GFP_COMP), > > nodemask=(null),cpuset=/,mems_allowed=0" still happens even

Re: [PATCH v3 hmm 02/11] mm/mmu_notifiers: do not speculatively allocate a mmu_notifier_mm

On Tue, Aug 06, 2019 at 08:15:39PM -0300, Jason Gunthorpe wrote: > From: Jason Gunthorpe > > A prior commit e0f3c3f78da2 ("mm/mmu_notifier: init notifier if necessary") > made an attempt at doing this, but had to be reverted as calling > the GFP_KERNEL allocator under the i_mmap_mutex causes

[PATCH 4/5] drm/amdgpu: Support snooped PTE flag

Set snooped PTE flag according to mapping flag. Write request to a page with snooped bit set, will send out invalidate probe request to TCC of the remote GPU where the vram page resides. Change-Id: I799f68ec7a5a1abf32075f5ef31051641a0b3736 Signed-off-by: Oak Zeng ---

[PATCH 3/5] drm/amdkfd: Postpone memory mapping flags calculation to mapping time

Some mapping flags are decided by memory mapping destination which is not know at memory object allocation time. So it is reasonable to decide memory mapping flags at mapping time, instead of alloc time. Record memory allocation flags during allocation time and calculate mapping flags during

[PATCH 5/5] amd/amdgpu: Introduce new page mapping scheme for arcturus

The new memory mapping scheme is: For vram: Fine-grain coherency: local CC and remote UC, with snoop. Coarse-grain coherency: local RW and remote UC, with snoop. For host memory (not changed) Fine-grain coherency: UC Coarse-grain coherency: NC Change-Id: I6a071249f953cbed813bfd953b6a2e0826f54f86

[PATCH 2/5] drm/amdgpu: Support new arcturus mtype

Arcturus repurposed mtype WC to RW. Modify gmc functions to support the new mtype Change-Id: Idc338e5386a57020f45262025e2664ab4ba9f291 Signed-off-by: Oak Zeng --- drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 3 +++ drivers/gpu/drm/amd/include/vega10_enum.h | 1 + 2 files changed, 4 insertions(+)

Re: [pull] amdgpu, amdkfd, radeon, ttm drm-next-5.4

On Wed, 7 Aug 2019 at 06:03, Alex Deucher wrote: > > Hi Dave, Daniel, > > The big updates here are support for new asics (navi14, navi12, arcturus). Thanks Alex, but due to the readq/writeq this break my local validation builds which means we need to land a fix for that somehow first. Also this

RE: [PATCH 1/9] drm/amdgpu: add gfx clock gating for Arcturus

Reviewed-by: Kenneth Feng -Original Message- From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Le Ma Sent: Thursday, August 08, 2019 6:22 PM To: amd-gfx@lists.freedesktop.org Cc: Ma, Le Subject: [PATCH 1/9] drm/amdgpu: add gfx clock gating for Arcturus

RE: [PATCH 2/9] drm/amdgpu: enable gfx clock gating for Arcturus

Reviewed-by: Kenneth Feng -Original Message- From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Le Ma Sent: Thursday, August 08, 2019 6:22 PM To: amd-gfx@lists.freedesktop.org Cc: Ma, Le Subject: [PATCH 2/9] drm/amdgpu: enable gfx clock gating for Arcturus

RE: [PATCH 3/9] drm/amdgpu: add hdp clock gating for Arcturus

Reviewed-by: Kenneth Feng -Original Message- From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Le Ma Sent: Thursday, August 08, 2019 6:22 PM To: amd-gfx@lists.freedesktop.org Cc: Ma, Le Subject: [PATCH 3/9] drm/amdgpu: add hdp clock gating for Arcturus

RE: [PATCH 1/1] drm/amdgpu: split athub clock gating from mmhub

Serier is Reviewed-by: Feifei Xu -Original Message- From: amd-gfx On Behalf Of Le Ma Sent: Thursday, August 8, 2019 6:32 PM To: amd-gfx@lists.freedesktop.org Cc: Ma, Le Subject: [PATCH 1/1] drm/amdgpu: split athub clock gating from mmhub Untie the bind of get/set athub CG state from

RE: libdrm patch merge request

Thanks, Alex. Regards, Guchun -Original Message- From: Alex Deucher Sent: Friday, August 9, 2019 1:24 AM To: Chen, Guchun Cc: Deucher, Alexander ; amd-gfx@lists.freedesktop.org; Zhou1, Tao ; Li, Candice ; Li, Dennis ; Zhang, Hawking Subject: Re: libdrm patch merge request Done!

RE: [PATCH 3/5] drm/amdkfd: Postpone memory mapping flags calculation to mapping time

Hi Felix/Sean, See one comment inline [Oak] Regards, Oak -Original Message- From: Kuehling, Felix Sent: Wednesday, August 7, 2019 12:05 AM To: Zeng, Oak ; amd-gfx@lists.freedesktop.org Cc: Koenig, Christian ; Keely, Sean Subject: Re: [PATCH 3/5] drm/amdkfd: Postpone memory mapping

[PATCH 1/2] drm/amdgpu: implement UMC 64 bits REG operations

implement 64 bits operations via 32 bits interface Signed-off-by: Tao Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h | 9 + drivers/gpu/drm/amd/amdgpu/umc_v6_1.c | 10 +- 2 files changed, 14 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h

Re: [PATCH v19 00/15] arm64: untag user pointers passed to the kernel

On Thu, Aug 08, 2019 at 03:33:00PM -0700, Andrew Morton wrote: > On Thu, 8 Aug 2019 14:12:19 -0700 Kees Cook wrote: > > > > The ones that are left are the mm ones: 4, 5, 6, 7 and 8. > > > > > > Andrew, could you take a look and give your Acked-by or pick them up > > > directly? > > > > Given

RE: [PATCH 4/9] drm/amdgpu: enable hdp clock gating for Arcturus

Reviewed-by: Kenneth Feng -Original Message- From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Le Ma Sent: Thursday, August 08, 2019 6:22 PM To: amd-gfx@lists.freedesktop.org Cc: Ma, Le Subject: [PATCH 4/9] drm/amdgpu: enable hdp clock gating for Arcturus

RE: [PATCH 6/9] drm/amdgpu: add sdma clock gating for Arcturus

Reviewed-by: Kenneth Feng -Original Message- From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Le Ma Sent: Thursday, August 08, 2019 6:22 PM To: amd-gfx@lists.freedesktop.org Cc: Ma, Le Subject: [PATCH 6/9] drm/amdgpu: add sdma clock gating for Arcturus

[PATCH 2/2] drm/amdgpu: remove RREG64/WREG64

atomic 64 bits REG operations are useless currently Signed-off-by: Tao Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu.h| 4 --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 33 -- 2 files changed, 37 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h

RE: [PATCH 2/2] drm/amdgpu: remove RREG64/WREG64

Series is Reviewed-by: Hawking Zhang Regards, Hawking -Original Message- From: amd-gfx On Behalf Of Tao Zhou Sent: 2019年8月9日 12:28 To: amd-gfx@lists.freedesktop.org; Deucher, Alexander ; Zhang, Hawking ; Koenig, Christian Cc: Zhou1, Tao Subject: [PATCH 2/2] drm/amdgpu: remove

[PATCH 1/5] drm/amdgpu: Extends amdgpu vm definitions

Add definition of all supported mtypes. The RW mtype is recently introduced for arcturus. Also add definition of a flag to probe and possibly invalidate remote GPU cache, which will be used later in this series. Change-Id: I96fc9bb4b6b1e62bdc10b600d8aaa6a802128d6d Signed-off-by: Oak Zeng ---

RE: [PATCH 7/9] drm/amdgpu: enable sdma clock gating for Arcturus

Reviewed-by: Kenneth Feng -Original Message- From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Le Ma Sent: Thursday, August 08, 2019 6:22 PM To: amd-gfx@lists.freedesktop.org Cc: Ma, Le Subject: [PATCH 7/9] drm/amdgpu: enable sdma clock gating for Arcturus

RE: [PATCH 5/9] drm/amdgpu: support sdma clock gating for more instances

Reviewed-by: Kenneth Feng -Original Message- From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Le Ma Sent: Thursday, August 08, 2019 6:22 PM To: amd-gfx@lists.freedesktop.org Cc: Ma, Le Subject: [PATCH 5/9] drm/amdgpu: support sdma clock gating for more

[PATCH 2/2] drm/amdkfd/gfx10: Calling amdgpu functions to invalidate TLB

Calling amdgpu function to invalidate TLB, instead of using a kfd implementation. Delete the kfd local TLB invalidation implementation. Change-Id: Ia16fcf9091a93b8c0acdaf9981f8a5a1f9a5ca1d Signed-off-by: Oak Zeng --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10.c | 41 ++

[PATCH 1/2] drm/amdgpu: Export function to flush TLB of specific vm hub

This is for kfd to reuse amdgpu TLB invalidation function. On gfx10, kfd only needs to flush TLB on gfx hub but not on mm hub. So export a function for KFD flush TLB only on specific hub. Change-Id: I58ff00969f88438cfd3dc7e9deb7bff0c1bb4133 Signed-off-by: Oak Zeng ---

Re: [PATCH] drm/amdgpu: update ras sysfs feature info

Hi Tao, yeah Kent is absolutely right. A must have requirement for sysfs is a stable interface. But there are a couple of other violations for sysfs rules as well: s = scnprintf(buf, PAGE_SIZE, "feature mask: 0x%x\n", con->features); for (i = 0; i < ras_block_count; i++) { A

Re: [PATCH v3 hmm 04/11] misc/sgi-gru: use mmu_notifier_get/put for struct gru_mm_struct

Looks good, Reviewed-by: Christoph Hellwig

[PATCH 3/9] drm/amdgpu: add hdp clock gating for Arcturus

Add hdp CGLS for Arcturus in set common clockgating function Change-Id: I44e392fa5f7653908b36b0902e721d56eed3eb92 Signed-off-by: Le Ma --- drivers/gpu/drm/amd/amdgpu/soc15.c | 7 ++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c

[PATCH 2/9] drm/amdgpu: enable gfx clock gating for Arcturus

Init gfx MGCG/LS, CGCG/LS, CP_LS flag. Change-Id: I88db76d1b8f2b2cecce10846a4d22eec638eea8a Signed-off-by: Le Ma --- drivers/gpu/drm/amd/amdgpu/soc15.c | 6 +- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c

[PATCH 4/9] drm/amdgpu: enable hdp clock gating for Arcturus

Init hdp MGCG/LS flag as Vega20 Change-Id: Ia33ca064f79ac409c53d3beb6f01b6e814a92041 Signed-off-by: Le Ma --- drivers/gpu/drm/amd/amdgpu/soc15.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index

Re: [PATCH v3 hmm 11/11] mm/mmu_notifiers: remove unregister_no_release

On Tue, Aug 06, 2019 at 08:15:48PM -0300, Jason Gunthorpe wrote: > From: Jason Gunthorpe > > mmu_notifier_unregister_no_release() and mmu_notifier_call_srcu() no > longer have any users, they have all been converted to use > mmu_notifier_put(). > > So delete this difficult to use interface. >

[PATCH 1/9] drm/amdgpu: add gfx clock gating for Arcturus

Add ARCTURUS case in gfx set clockgating function. No 3d clock on Arcturus. Change-Id: I9893a2afea7f0b5d433baa14f48ae55a36516fac Signed-off-by: Le Ma --- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 4 1 file changed, 4 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c

[PATCH 5/9] drm/amdgpu: support sdma clock gating for more instances

Shorten the code with RREG32_SDMA/WREG32_SDMA macro in CG part. Change-Id: Icbf94169bb703877b105a307f14c708609faaae4 Signed-off-by: Le Ma --- drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 105 +++-- 1 file changed, 34 insertions(+), 71 deletions(-) diff --git

[PATCH 6/9] drm/amdgpu: add sdma clock gating for Arcturus

Add ARCTURUS case in sdma set clockgating function Change-Id: I65a3d99a140a8a76949b4d03c20bc6e0195c9854 Signed-off-by: Le Ma --- drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c

[PATCH 7/9] drm/amdgpu: enable sdma clock gating for Arcturus

Init sdma MGCG/LS flag Change-Id: I600b8c67b1dfa74240269f2f028960b2c93a0ec2 Signed-off-by: Le Ma --- drivers/gpu/drm/amd/amdgpu/soc15.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index

Re: [PATCH v3 hmm 05/11] hmm: use mmu_notifier_get/put for 'struct hmm'

Looks good, Reviewed-by: Christoph Hellwig

RE: [PATCH 7/9] drm/amdgpu: enable sdma clock gating for Arcturus

This series only contains 7 patches. Sorry for the confusion. Regards, Ma Le -Original Message- From: amd-gfx On Behalf Of Le Ma Sent: Thursday, August 08, 2019 6:22 PM To: amd-gfx@lists.freedesktop.org Cc: Ma, Le Subject: [PATCH 7/9] drm/amdgpu: enable sdma clock gating for Arcturus

Re: [PATCH v3 hmm 01/11] mm/mmu_notifiers: hoist do_mmu_notifier_register down_write to the caller

On Tue, Aug 06, 2019 at 08:15:38PM -0300, Jason Gunthorpe wrote: > From: Jason Gunthorpe > > This simplifies the code to not have so many one line functions and extra > logic. __mmu_notifier_register() simply becomes the entry point to > register the notifier, and the other one calls it under

[PATCH 1/1] drm/amdgpu: split athub clock gating from mmhub

Untie the bind of get/set athub CG state from mmhub, for cosmetic fix and Asic not using mmhub 1.0. Besides, also fix wrong athub CG state in amdgpu_pm_info. Change-Id: I4ba970cae558ad5163e93fa9bc77f589196a22b1 Signed-off-by: Le Ma --- drivers/gpu/drm/amd/amdgpu/Makefile | 1 +

RE: [PATCH] drm/amdgpu: update ras sysfs feature info

+Chris Freehill While I can understand this change, this broke our SMI interface, which was expecting a specific string format for the ras/features file. This has happened a few times now, where changes to the RAS sysfs files has broke the SMI CLI and/or SMI LIB. Can we please get a stable

RE: [PATCH] drm/amdgpu: update ras sysfs feature info

Understood and agree we should keep stable interfaces. But the information in feature node is not correct and makes people confusing. Basically, each IP blocks can support all the four error types, not just multi-uncorrectable. As a result, any upper level apps/libs that read from this file