Re: [PATCH] amdgpu/sriov Stop data exchange for wholegpu reset
Dear Jack, Thank you for your patch. Please add a colon after amdgpu/sriov in the commit message summary. Am 07.01.21 um 11:46 schrieb Jack Zhang: [Why] When host trigger a whole gpu reset, guest will keep *hosts trigger* or *host triggers* waiting till host finish reset. But there's a work finishes queue in guest exchanging data between vf which need needs to access frame buffer. During whole gpu reset, frame buffer is not accessable, and this causes the call trace. accessible (a spell checker should have caught that) Can you please paste part of the trace, so it’s easily findable by users running into this. [How] After vf get reset notification from pf, stop data exchange. How can this be reproduced and tested? Kind regards, Paul Signed-off-by: Jingwen Chen Signed-off-by: Jack Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 1 + drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c| 1 + drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c| 1 + 3 files changed, 3 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index 83ca5cbffe2c..3e212862cf5d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -571,6 +571,7 @@ void amdgpu_virt_fini_data_exchange(struct amdgpu_device *adev) DRM_INFO("clean up the vf2pf work item\n"); flush_delayed_work(>virt.vf2pf_work); cancel_delayed_work_sync(>virt.vf2pf_work); + adev->virt.vf2pf_update_interval_ms = 0; } } diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c index 7767ccca526b..3ee481557fc9 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c @@ -255,6 +255,7 @@ static void xgpu_ai_mailbox_flr_work(struct work_struct *work) if (!down_read_trylock(>reset_sem)) return; + amdgpu_virt_fini_data_exchange(adev); atomic_set(>in_gpu_reset, 1); do { diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c index dd5c1e6ce009..48e588d3c409 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c @@ -276,6 +276,7 @@ static void xgpu_nv_mailbox_flr_work(struct work_struct *work) if (!down_read_trylock(>reset_sem)) return; + amdgpu_virt_fini_data_exchange(adev); atomic_set(>in_gpu_reset, 1); do { ___ amd-gfx mailing list amd-gfx@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/amd-gfx
RE: [PATCH] amdgpu/sriov Stop data exchange for wholegpu reset
[AMD Official Use Only - Internal Distribution Only] Reviewed-by: Monk Liu Thanks -- Monk Liu | Cloud-GPU Core team -- -Original Message- From: Zhang, Jack (Jian) Sent: Tuesday, January 12, 2021 11:20 AM To: Zhang, Hawking ; amd-gfx@lists.freedesktop.org; Liu, Monk ; Chen, JingWen ; Deucher, Alexander ; Deng, Emily Subject: RE: [PATCH] amdgpu/sriov Stop data exchange for wholegpu reset [AMD Official Use Only - Internal Distribution Only] Ping... -Original Message- From: Zhang, Jack (Jian) Sent: Friday, January 8, 2021 11:07 AM To: amd-gfx@lists.freedesktop.org; Liu, Monk ; Chen, JingWen ; Deucher, Alexander ; Deng, Emily Subject: RE: [PATCH] amdgpu/sriov Stop data exchange for wholegpu reset Ping -Original Message- From: Jack Zhang Sent: Thursday, January 7, 2021 6:47 PM To: amd-gfx@lists.freedesktop.org Cc: Zhang, Jack (Jian) ; Zhang, Jack (Jian) ; Chen, JingWen Subject: [PATCH] amdgpu/sriov Stop data exchange for wholegpu reset [Why] When host trigger a whole gpu reset, guest will keep waiting till host finish reset. But there's a work queue in guest exchanging data between vf which need to access frame buffer. During whole gpu reset, frame buffer is not accessable, and this causes the call trace. [How] After vf get reset notification from pf, stop data exchange. Signed-off-by: Jingwen Chen Signed-off-by: Jack Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 1 + drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c| 1 + drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c| 1 + 3 files changed, 3 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index 83ca5cbffe2c..3e212862cf5d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -571,6 +571,7 @@ void amdgpu_virt_fini_data_exchange(struct amdgpu_device *adev) DRM_INFO("clean up the vf2pf work item\n"); flush_delayed_work(>virt.vf2pf_work); cancel_delayed_work_sync(>virt.vf2pf_work); +adev->virt.vf2pf_update_interval_ms = 0; } } diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c index 7767ccca526b..3ee481557fc9 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c @@ -255,6 +255,7 @@ static void xgpu_ai_mailbox_flr_work(struct work_struct *work) if (!down_read_trylock(>reset_sem)) return; +amdgpu_virt_fini_data_exchange(adev); atomic_set(>in_gpu_reset, 1); do { diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c index dd5c1e6ce009..48e588d3c409 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c @@ -276,6 +276,7 @@ static void xgpu_nv_mailbox_flr_work(struct work_struct *work) if (!down_read_trylock(>reset_sem)) return; +amdgpu_virt_fini_data_exchange(adev); atomic_set(>in_gpu_reset, 1); do { -- 2.25.1 ___ amd-gfx mailing list amd-gfx@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/amd-gfx
RE: [PATCH] amdgpu/sriov Stop data exchange for wholegpu reset
[AMD Official Use Only - Internal Distribution Only] Ping... -Original Message- From: Zhang, Jack (Jian) Sent: Friday, January 8, 2021 11:07 AM To: amd-gfx@lists.freedesktop.org; Liu, Monk ; Chen, JingWen ; Deucher, Alexander ; Deng, Emily Subject: RE: [PATCH] amdgpu/sriov Stop data exchange for wholegpu reset Ping -Original Message- From: Jack Zhang Sent: Thursday, January 7, 2021 6:47 PM To: amd-gfx@lists.freedesktop.org Cc: Zhang, Jack (Jian) ; Zhang, Jack (Jian) ; Chen, JingWen Subject: [PATCH] amdgpu/sriov Stop data exchange for wholegpu reset [Why] When host trigger a whole gpu reset, guest will keep waiting till host finish reset. But there's a work queue in guest exchanging data between vf which need to access frame buffer. During whole gpu reset, frame buffer is not accessable, and this causes the call trace. [How] After vf get reset notification from pf, stop data exchange. Signed-off-by: Jingwen Chen Signed-off-by: Jack Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 1 + drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c| 1 + drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c| 1 + 3 files changed, 3 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index 83ca5cbffe2c..3e212862cf5d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -571,6 +571,7 @@ void amdgpu_virt_fini_data_exchange(struct amdgpu_device *adev) DRM_INFO("clean up the vf2pf work item\n"); flush_delayed_work(>virt.vf2pf_work); cancel_delayed_work_sync(>virt.vf2pf_work); +adev->virt.vf2pf_update_interval_ms = 0; } } diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c index 7767ccca526b..3ee481557fc9 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c @@ -255,6 +255,7 @@ static void xgpu_ai_mailbox_flr_work(struct work_struct *work) if (!down_read_trylock(>reset_sem)) return; +amdgpu_virt_fini_data_exchange(adev); atomic_set(>in_gpu_reset, 1); do { diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c index dd5c1e6ce009..48e588d3c409 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c @@ -276,6 +276,7 @@ static void xgpu_nv_mailbox_flr_work(struct work_struct *work) if (!down_read_trylock(>reset_sem)) return; +amdgpu_virt_fini_data_exchange(adev); atomic_set(>in_gpu_reset, 1); do { -- 2.25.1 ___ amd-gfx mailing list amd-gfx@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/amd-gfx
RE: [PATCH] amdgpu/sriov Stop data exchange for wholegpu reset
Ping -Original Message- From: Jack Zhang Sent: Thursday, January 7, 2021 6:47 PM To: amd-gfx@lists.freedesktop.org Cc: Zhang, Jack (Jian) ; Zhang, Jack (Jian) ; Chen, JingWen Subject: [PATCH] amdgpu/sriov Stop data exchange for wholegpu reset [Why] When host trigger a whole gpu reset, guest will keep waiting till host finish reset. But there's a work queue in guest exchanging data between vf which need to access frame buffer. During whole gpu reset, frame buffer is not accessable, and this causes the call trace. [How] After vf get reset notification from pf, stop data exchange. Signed-off-by: Jingwen Chen Signed-off-by: Jack Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 1 + drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c| 1 + drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c| 1 + 3 files changed, 3 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index 83ca5cbffe2c..3e212862cf5d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -571,6 +571,7 @@ void amdgpu_virt_fini_data_exchange(struct amdgpu_device *adev) DRM_INFO("clean up the vf2pf work item\n"); flush_delayed_work(>virt.vf2pf_work); cancel_delayed_work_sync(>virt.vf2pf_work); + adev->virt.vf2pf_update_interval_ms = 0; } } diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c index 7767ccca526b..3ee481557fc9 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c @@ -255,6 +255,7 @@ static void xgpu_ai_mailbox_flr_work(struct work_struct *work) if (!down_read_trylock(>reset_sem)) return; + amdgpu_virt_fini_data_exchange(adev); atomic_set(>in_gpu_reset, 1); do { diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c index dd5c1e6ce009..48e588d3c409 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c @@ -276,6 +276,7 @@ static void xgpu_nv_mailbox_flr_work(struct work_struct *work) if (!down_read_trylock(>reset_sem)) return; + amdgpu_virt_fini_data_exchange(adev); atomic_set(>in_gpu_reset, 1); do { -- 2.25.1 ___ amd-gfx mailing list amd-gfx@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/amd-gfx