Re: [PATCH] gpu/drm: Replace "%p" with "%pK"
Am 30.06.20 um 14:14 schrieb Tiezhu Yang: On 06/30/2020 04:05 PM, Christian König wrote: Am 30.06.20 um 09:36 schrieb Tiezhu Yang: When I update the latest kernel, I see the following "ptrval" boot messages. Use "%pK" instead of "%p" so that the cpu address can be printed when the kptr_restrict sysctl is set to 1. Both radeon_fence_driver_start_ring() and amdgpu_fence_driver_start_ring() have this similar issue, fix them. [ 1.872600] radeon :01:05.0: fence driver on ring 0 use gpu addr 0x48000c00 and cpu addr 0x(ptrval) [ 1.879095] radeon :01:05.0: fence driver on ring 5 use gpu addr 0x40056038 and cpu addr 0x(ptrval) We can probably just completely drop the CPU address here. OK, maybe the CPU address is not much useful. If nobody has any objections, I will send v2 to remove the debug info about CPU address. The CPU address was useful in the past, but isn't any more. We now have a debugfs file to read the current fence values. Please go ahead and send a patch to just remove printing them. Thanks, Christian. Thanks, Tiezhu Christian. Signed-off-by: Tiezhu Yang --- drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | 5 ++--- drivers/gpu/drm/radeon/radeon_fence.c | 2 +- 2 files changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index d878fe7..d4d1e8c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -422,9 +422,8 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring, ring->fence_drv.irq_type = irq_type; ring->fence_drv.initialized = true; - DRM_DEV_DEBUG(adev->dev, "fence driver on ring %s use gpu addr " - "0x%016llx, cpu addr 0x%p\n", ring->name, - ring->fence_drv.gpu_addr, ring->fence_drv.cpu_addr); + DRM_DEV_DEBUG(adev->dev, "fence driver on ring %s use gpu addr 0x%016llx, cpu addr 0x%pK\n", + ring->name, ring->fence_drv.gpu_addr, ring->fence_drv.cpu_addr); return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c index 43f2f93..c51b094 100644 --- a/drivers/gpu/drm/radeon/radeon_fence.c +++ b/drivers/gpu/drm/radeon/radeon_fence.c @@ -865,7 +865,7 @@ int radeon_fence_driver_start_ring(struct radeon_device *rdev, int ring) } radeon_fence_write(rdev, atomic64_read(>fence_drv[ring].last_seq), ring); rdev->fence_drv[ring].initialized = true; - dev_info(rdev->dev, "fence driver on ring %d use gpu addr 0x%016llx and cpu addr 0x%p\n", + dev_info(rdev->dev, "fence driver on ring %d use gpu addr 0x%016llx and cpu addr 0x%pK\n", ring, rdev->fence_drv[ring].gpu_addr, rdev->fence_drv[ring].cpu_addr); return 0; }
Re: [PATCH] gpu/drm: Replace "%p" with "%pK"
On 06/30/2020 04:05 PM, Christian König wrote: Am 30.06.20 um 09:36 schrieb Tiezhu Yang: When I update the latest kernel, I see the following "ptrval" boot messages. Use "%pK" instead of "%p" so that the cpu address can be printed when the kptr_restrict sysctl is set to 1. Both radeon_fence_driver_start_ring() and amdgpu_fence_driver_start_ring() have this similar issue, fix them. [1.872600] radeon :01:05.0: fence driver on ring 0 use gpu addr 0x48000c00 and cpu addr 0x(ptrval) [1.879095] radeon :01:05.0: fence driver on ring 5 use gpu addr 0x40056038 and cpu addr 0x(ptrval) We can probably just completely drop the CPU address here. OK, maybe the CPU address is not much useful. If nobody has any objections, I will send v2 to remove the debug info about CPU address. Thanks, Tiezhu Christian. Signed-off-by: Tiezhu Yang --- drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | 5 ++--- drivers/gpu/drm/radeon/radeon_fence.c | 2 +- 2 files changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index d878fe7..d4d1e8c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -422,9 +422,8 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring, ring->fence_drv.irq_type = irq_type; ring->fence_drv.initialized = true; -DRM_DEV_DEBUG(adev->dev, "fence driver on ring %s use gpu addr " - "0x%016llx, cpu addr 0x%p\n", ring->name, - ring->fence_drv.gpu_addr, ring->fence_drv.cpu_addr); +DRM_DEV_DEBUG(adev->dev, "fence driver on ring %s use gpu addr 0x%016llx, cpu addr 0x%pK\n", + ring->name, ring->fence_drv.gpu_addr, ring->fence_drv.cpu_addr); return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c index 43f2f93..c51b094 100644 --- a/drivers/gpu/drm/radeon/radeon_fence.c +++ b/drivers/gpu/drm/radeon/radeon_fence.c @@ -865,7 +865,7 @@ int radeon_fence_driver_start_ring(struct radeon_device *rdev, int ring) } radeon_fence_write(rdev, atomic64_read(>fence_drv[ring].last_seq), ring); rdev->fence_drv[ring].initialized = true; -dev_info(rdev->dev, "fence driver on ring %d use gpu addr 0x%016llx and cpu addr 0x%p\n", +dev_info(rdev->dev, "fence driver on ring %d use gpu addr 0x%016llx and cpu addr 0x%pK\n", ring, rdev->fence_drv[ring].gpu_addr, rdev->fence_drv[ring].cpu_addr); return 0; }
Re: [PATCH] gpu/drm: Replace "%p" with "%pK"
Am 30.06.20 um 09:36 schrieb Tiezhu Yang: When I update the latest kernel, I see the following "ptrval" boot messages. Use "%pK" instead of "%p" so that the cpu address can be printed when the kptr_restrict sysctl is set to 1. Both radeon_fence_driver_start_ring() and amdgpu_fence_driver_start_ring() have this similar issue, fix them. [1.872600] radeon :01:05.0: fence driver on ring 0 use gpu addr 0x48000c00 and cpu addr 0x(ptrval) [1.879095] radeon :01:05.0: fence driver on ring 5 use gpu addr 0x40056038 and cpu addr 0x(ptrval) We can probably just completely drop the CPU address here. Christian. Signed-off-by: Tiezhu Yang --- drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | 5 ++--- drivers/gpu/drm/radeon/radeon_fence.c | 2 +- 2 files changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index d878fe7..d4d1e8c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -422,9 +422,8 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring, ring->fence_drv.irq_type = irq_type; ring->fence_drv.initialized = true; - DRM_DEV_DEBUG(adev->dev, "fence driver on ring %s use gpu addr " - "0x%016llx, cpu addr 0x%p\n", ring->name, - ring->fence_drv.gpu_addr, ring->fence_drv.cpu_addr); + DRM_DEV_DEBUG(adev->dev, "fence driver on ring %s use gpu addr 0x%016llx, cpu addr 0x%pK\n", + ring->name, ring->fence_drv.gpu_addr, ring->fence_drv.cpu_addr); return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c index 43f2f93..c51b094 100644 --- a/drivers/gpu/drm/radeon/radeon_fence.c +++ b/drivers/gpu/drm/radeon/radeon_fence.c @@ -865,7 +865,7 @@ int radeon_fence_driver_start_ring(struct radeon_device *rdev, int ring) } radeon_fence_write(rdev, atomic64_read(>fence_drv[ring].last_seq), ring); rdev->fence_drv[ring].initialized = true; - dev_info(rdev->dev, "fence driver on ring %d use gpu addr 0x%016llx and cpu addr 0x%p\n", + dev_info(rdev->dev, "fence driver on ring %d use gpu addr 0x%016llx and cpu addr 0x%pK\n", ring, rdev->fence_drv[ring].gpu_addr, rdev->fence_drv[ring].cpu_addr); return 0; }
[PATCH] gpu/drm: Replace "%p" with "%pK"
When I update the latest kernel, I see the following "ptrval" boot messages. Use "%pK" instead of "%p" so that the cpu address can be printed when the kptr_restrict sysctl is set to 1. Both radeon_fence_driver_start_ring() and amdgpu_fence_driver_start_ring() have this similar issue, fix them. [1.872600] radeon :01:05.0: fence driver on ring 0 use gpu addr 0x48000c00 and cpu addr 0x(ptrval) [1.879095] radeon :01:05.0: fence driver on ring 5 use gpu addr 0x40056038 and cpu addr 0x(ptrval) Signed-off-by: Tiezhu Yang --- drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | 5 ++--- drivers/gpu/drm/radeon/radeon_fence.c | 2 +- 2 files changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index d878fe7..d4d1e8c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -422,9 +422,8 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring, ring->fence_drv.irq_type = irq_type; ring->fence_drv.initialized = true; - DRM_DEV_DEBUG(adev->dev, "fence driver on ring %s use gpu addr " - "0x%016llx, cpu addr 0x%p\n", ring->name, - ring->fence_drv.gpu_addr, ring->fence_drv.cpu_addr); + DRM_DEV_DEBUG(adev->dev, "fence driver on ring %s use gpu addr 0x%016llx, cpu addr 0x%pK\n", + ring->name, ring->fence_drv.gpu_addr, ring->fence_drv.cpu_addr); return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c index 43f2f93..c51b094 100644 --- a/drivers/gpu/drm/radeon/radeon_fence.c +++ b/drivers/gpu/drm/radeon/radeon_fence.c @@ -865,7 +865,7 @@ int radeon_fence_driver_start_ring(struct radeon_device *rdev, int ring) } radeon_fence_write(rdev, atomic64_read(>fence_drv[ring].last_seq), ring); rdev->fence_drv[ring].initialized = true; - dev_info(rdev->dev, "fence driver on ring %d use gpu addr 0x%016llx and cpu addr 0x%p\n", + dev_info(rdev->dev, "fence driver on ring %d use gpu addr 0x%016llx and cpu addr 0x%pK\n", ring, rdev->fence_drv[ring].gpu_addr, rdev->fence_drv[ring].cpu_addr); return 0; } -- 2.1.0
