RE: [PATCH 10/11] drm/amdgpu: Rework pcie_bif ras sw_init

> -Original Message- > From: Zhang, Hawking > Sent: Monday, March 6, 2023 10:32 AM > To: amd-gfx@lists.freedesktop.org; Zhou1, Tao ; > Yang, Stanley ; Li, Candice ; > Chai, Thomas > Cc: Zhang, Hawking > Subject: [PATCH 10/11] drm/amdgpu: Rework pcie_bif ras sw_init > > pcie_bif ras b

RE: [PATCH 09/11] drm/amdgpu: Rework xgmi_wafl_pcs ras sw_init

> -Original Message- > From: Zhang, Hawking > Sent: Monday, March 6, 2023 10:32 AM > To: amd-gfx@lists.freedesktop.org; Zhou1, Tao ; > Yang, Stanley ; Li, Candice ; > Chai, Thomas > Cc: Zhang, Hawking > Subject: [PATCH 09/11] drm/amdgpu: Rework xgmi_wafl_pcs ras sw_init > > To align

[PATCH] drm/amdgpu: Init MMVM_CONTEXTS_DISABLE in gmc11 golden setting under SRIOV

[Why] If disable the mmhub vm contexts(set MMVM_CONTEXTS_DISABLE to 0x), driver loading failed on vf due to fence fallback timer expired on all rings. FLR cannot reset MMVM_CONTEXTS_DISABLE. So this vf can not be recovered anymore unless trigger a whole gpu reset. [How] Under SRIOV, init MMVM_

RE: [PATCH 08/11] drm/amdgpu: Rework mca ras sw_init

> -Original Message- > From: Zhang, Hawking > Sent: Monday, March 6, 2023 10:32 AM > To: amd-gfx@lists.freedesktop.org; Zhou1, Tao ; > Yang, Stanley ; Li, Candice ; > Chai, Thomas > Cc: Zhang, Hawking > Subject: [PATCH 08/11] drm/amdgpu: Rework mca ras sw_init > > To align with other

[PATCH 1/1] amdkfd: Fix memory availability double accounting of kfd pinned objects

Pinned objects that are not kfd objects reduce the total vram available to kfd, so we subtract the total size of pinned objects from kdf vram availability. However this double counts objects pinned by kfd itself because they are counted both as used and pinned. So track the total size of objects pi

Re: [PATCH] drm/amdgpu: Use uncached ioremap() for LoongArch

Am 06.03.23 um 04:01 schrieb WANG Xuerui: On 2023/3/6 10:49, Huacai Chen wrote: Hi, Christian, On Mon, Mar 6, 2023 at 12:40 AM Christian König wrote: Am 05.03.23 um 06:21 schrieb Huacai Chen: LoongArch maintains cache coherency in hardware, but its WUC attribute (Weak-ordered UnCached, whic

RE: [PATCH 03/11] drm/amdgpu: Init gfx ras block when ras is supported

I have one additional suggestion, the amdgpu_gfx_ras_sw_init is declared twice in amdgpu_gfx.h file, it can be removed one in this patch. Regards, Stanley > -Original Message- > From: Zhang, Hawking > Sent: Monday, March 6, 2023 10:32 AM > To: amd-gfx@lists.freedesktop.org; Zhou1, Tao ;

RE: [PATCH 01/11] drm/amdgpu: Move jpeg ras block init to ras sw_init

> -Original Message- > From: Zhang, Hawking > Sent: Monday, March 6, 2023 10:32 AM > To: amd-gfx@lists.freedesktop.org; Zhou1, Tao ; Yang, > Stanley ; Li, Candice ; Chai, > Thomas > Cc: Zhang, Hawking > Subject: [PATCH 01/11] drm/amdgpu: Move jpeg ras block init to ras sw_init > > In

Re: [PATCH] drm/amdgpu: disable cstate properly for driver reloading scenario

On 3/2/2023 9:05 PM, Lazar, Lijo wrote: On 3/2/2023 8:56 PM, Deucher, Alexander wrote: [AMD Official Use Only - General] -Original Message- From: Quan, Evan Sent: Thursday, March 2, 2023 4:31 AM To: Lazar, Lijo ; amd-gfx@lists.freedesktop.org; Deucher, Alexander Subject: RE: [PA

RE: [PATCH 0/3] enable new capped/uncapped power profiles

[AMD Official Use Only - General] Acked-by: Kenneth Feng -Original Message- From: amd-gfx On Behalf Of Perry Yuan Sent: Friday, February 24, 2023 4:33 PM To: Deucher, Alexander ; Quan, Evan ; Lazar, Lijo ; Li, Candice ; amd-gfx@lists.freedesktop.org Cc: Huang, Shimmer ; Liu, Kun Sub

RE: [PATCH 0/3] enable new capped/uncapped power profiles

[AMD Official Use Only - General] Series is reviewed-by: Evan Quan > -Original Message- > From: Yuan, Perry > Sent: Friday, February 24, 2023 4:33 PM > To: Deucher, Alexander ; Quan, Evan > ; Lazar, Lijo ; Li, Candice > ; amd-gfx@lists.freedesktop.org > Cc: Huang, Shimmer ; Liu, Kun >

Re: [PATCH] drm/amdgpu: Use uncached ioremap() for LoongArch

Hi, Christian, On Mon, Mar 6, 2023 at 12:40 AM Christian König wrote: > > Am 05.03.23 um 06:21 schrieb Huacai Chen: > > LoongArch maintains cache coherency in hardware, but its WUC attribute > > (Weak-ordered UnCached, which is similar to WC) is out of the scope of > > cache coherency machanism.

[PATCH 11/11] drm/amdgpu: drop ras check at asic level for new blocks

amdgpu_ras_register_ras_block should always be invoked by ras_sw_init, where driver needs to check ras caps at ip level, instead of asic level. Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgp

[PATCH 10/11] drm/amdgpu: Rework pcie_bif ras sw_init

pcie_bif ras blocks needs to be initialized as early as possible to handle fatal error detected in hw_init phase. also align the pcie_bif ras sw_init with other ras blocks Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.c | 23 +++ drivers/gpu/drm/amd/

[PATCH 09/11] drm/amdgpu: Rework xgmi_wafl_pcs ras sw_init

To align with other IP blocks. Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c | 9 +--- drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c | 28 +++- drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h | 1 + drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c| 7 ++ 4

[PATCH 08/11] drm/amdgpu: Rework mca ras sw_init

To align with other IP blocks Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c | 21 drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c | 72 + drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h | 9 ++-- drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 15 +++--- dr

[PATCH 07/11] drm/amdgpu: Move hdp ras block init to ras sw_init

Initialize hdp ras block only when mmhub ip block supports ras features. Driver queries ras capabilities after early_init, ras block init needs to be moved to sw_init. Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/Makefile | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c | 9 +

[PATCH 06/11] drm/amdgpu: Move mmhub ras block init to ras sw_init

Initialize mmhub ras block only when mmhub ip block supports ras features. Driver queries ras capabilities after early_init, ras block init needs to be moved to sw_init. Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/Makefile | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c

[PATCH 05/11] drm/amdgpu: Move umc ras block init to gmc ras sw_init

Initialize umc ras block only when umc ip block supports ras. Driver queries ras capabilities after early_init, ras block init needs to be moved to sw_init. Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c | 16 +++-- drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h | 2 +

[PATCH 04/11] drm/amdgpu: Init sdma ras block when ras is supported

Initialize sdma ras block only when sdma ip block supports ras features. Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 9 ++--- drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c | 8 drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c | 9 ++--- 3 files changed, 20 ins

[PATCH 03/11] drm/amdgpu: Init gfx ras block when ras is supported

Initialize gfx ras block only when gfx ip block supports ras features. Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c | 9 ++--- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 9 ++--- 2 files changed, 12 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/

[PATCH 02/11] drm/amdgpu: Move vcn ras block init to ras sw_init

Initialize vcn ras block only when vcn ip block supports ras features. Driver queries ras capabilities after early_init, ras block init needs to be moved to sw_int. Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 29 - drivers/gpu/drm/amd/amdgpu

[PATCH 01/11] drm/amdgpu: Move jpeg ras block init to ras sw_init

Initialize jpeg ras block only when jpeg ip block supports ras features. Driver queries ras capabilities after early_init, ras block init needs to be moved to sw_int. Signed-off-by: Hawking Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.c | 29 drivers/gpu/drm/amd/amdg

[PATCH 00/11] Add ras sw_init

We are moving from soc ras to ip ras to address issues as follows - RAS sw block init is mixed in early_init and sw_init - RAS cap check is mixed with both soc check and ip check. RAS cap check is now only avaialble in amdgpu_ras_init, based on the cap query from bios. RAS sw block init is all mov

Re: [PATCH] drm/amdgpu: Use uncached ioremap() for LoongArch

Am 05.03.23 um 06:21 schrieb Huacai Chen: LoongArch maintains cache coherency in hardware, but its WUC attribute (Weak-ordered UnCached, which is similar to WC) is out of the scope of cache coherency machanism. This means WUC can only used for write-only memory regions. So use uncached ioremap()