Commit 12938fad authored by Christian König's avatar Christian König Committed by Alex Deucher
Browse files

drm/amdgpu: cleanup GPU recovery check a bit (v2) 



Check if we should call the function instead of providing the forced
flag.

v2: rebase on KFD changes (Alex)

Signed-off-by: default avatarChristian König <christian.koenig@amd.com>
Acked-by: default avatarAndrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: default avatarHuang Rui <ray.huang@amd.com>
Signed-off-by: default avatarAlex Deucher <alexander.deucher@amd.com>
parent 1849e737

drivers/gpu/drm/amd/amdgpu/amdgpu.h

+2 −1
Original line number Original line Diff line number Diff line
@@ -1158,8 +1158,9 @@ int emu_soc_asic_init(struct amdgpu_device *adev); 
#define amdgpu_asic_need_full_reset(adev) (adev)->asic_funcs->need_full_reset((adev))

#define amdgpu_asic_need_full_reset(adev) (adev)->asic_funcs->need_full_reset((adev))





/* Common functions */

/* Common functions */

bool amdgpu_device_should_recover_gpu(struct amdgpu_device *adev);

int amdgpu_device_gpu_recover(struct amdgpu_device *adev,

int amdgpu_device_gpu_recover(struct amdgpu_device *adev,

			      struct amdgpu_job* job, bool force);

			      struct amdgpu_job* job);

void amdgpu_device_pci_config_reset(struct amdgpu_device *adev);

void amdgpu_device_pci_config_reset(struct amdgpu_device *adev);

bool amdgpu_device_need_post(struct amdgpu_device *adev);

bool amdgpu_device_need_post(struct amdgpu_device *adev);

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c

+2 −1
Original line number Original line Diff line number Diff line
@@ -267,7 +267,8 @@ void amdgpu_amdkfd_gpu_reset(struct kgd_dev *kgd) 
{

{

	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;

	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;





	amdgpu_device_gpu_recover(adev, NULL, false);

	if (amdgpu_device_should_recover_gpu(adev))

		amdgpu_device_gpu_recover(adev, NULL);

}

}





int alloc_gtt_mem(struct kgd_dev *kgd, size_t size,

int alloc_gtt_mem(struct kgd_dev *kgd, size_t size,

drivers/gpu/drm/amd/amdgpu/amdgpu_device.c

+25 −13
Original line number Original line Diff line number Diff line
@@ -3243,32 +3243,44 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev, 
	return r;

	return r;

}

}





/**

 * amdgpu_device_should_recover_gpu - check if we should try GPU recovery

 *

 * @adev: amdgpu device pointer

 *

 * Check amdgpu_gpu_recovery and SRIOV status to see if we should try to recover

 * a hung GPU.

 */

bool amdgpu_device_should_recover_gpu(struct amdgpu_device *adev)

{

	if (!amdgpu_device_ip_check_soft_reset(adev)) {

		DRM_INFO("Timeout, but no hardware hang detected.\n");

		return false;

	}



	if (amdgpu_gpu_recovery == 0 || (amdgpu_gpu_recovery == -1  &&

					 !amdgpu_sriov_vf(adev))) {

		DRM_INFO("GPU recovery disabled.\n");

		return false;

	}



	return true;

}



/**

/**

 * amdgpu_device_gpu_recover - reset the asic and recover scheduler

 * amdgpu_device_gpu_recover - reset the asic and recover scheduler

 *

 *

 * @adev: amdgpu device pointer

 * @adev: amdgpu device pointer

 * @job: which job trigger hang

 * @job: which job trigger hang

 * @force: forces reset regardless of amdgpu_gpu_recovery

 *

 *

 * Attempt to reset the GPU if it has hung (all asics).

 * Attempt to reset the GPU if it has hung (all asics).

 * Returns 0 for success or an error on failure.

 * Returns 0 for success or an error on failure.

 */

 */

int amdgpu_device_gpu_recover(struct amdgpu_device *adev,

int amdgpu_device_gpu_recover(struct amdgpu_device *adev,

			      struct amdgpu_job *job, bool force)

			      struct amdgpu_job *job)

{

{

	int i, r, resched;

	int i, r, resched;





	if (!force && !amdgpu_device_ip_check_soft_reset(adev)) {

		DRM_INFO("No hardware hang detected. Did some blocks stall?\n");

		return 0;

	}



	if (!force && (amdgpu_gpu_recovery == 0 ||

			(amdgpu_gpu_recovery == -1  && !amdgpu_sriov_vf(adev)))) {

		DRM_INFO("GPU recovery disabled.\n");

		return 0;

	}



	dev_info(adev->dev, "GPU reset begin!\n");

	dev_info(adev->dev, "GPU reset begin!\n");





	mutex_lock(&adev->lock_reset);

	mutex_lock(&adev->lock_reset);

drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c

+1 −1
Original line number Original line Diff line number Diff line
@@ -701,7 +701,7 @@ static int amdgpu_debugfs_gpu_recover(struct seq_file *m, void *data) 
	struct amdgpu_device *adev = dev->dev_private;

	struct amdgpu_device *adev = dev->dev_private;





	seq_printf(m, "gpu recover\n");

	seq_printf(m, "gpu recover\n");

	amdgpu_device_gpu_recover(adev, NULL, true);

	amdgpu_device_gpu_recover(adev, NULL);





	return 0;

	return 0;

}

}

drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c

+2 −2
Original line number Original line Diff line number Diff line
@@ -105,8 +105,8 @@ static void amdgpu_irq_reset_work_func(struct work_struct *work) 
	struct amdgpu_device *adev = container_of(work, struct amdgpu_device,

	struct amdgpu_device *adev = container_of(work, struct amdgpu_device,

						  reset_work);

						  reset_work);





	if (!amdgpu_sriov_vf(adev))

	if (!amdgpu_sriov_vf(adev) && amdgpu_device_should_recover_gpu(adev))

		amdgpu_device_gpu_recover(adev, NULL, false);

		amdgpu_device_gpu_recover(adev, NULL);

}

}





/**

/**