[PATCH] drm/amdgpu: Simplify amdgpu_lockup_timeout usage.

[Liu, Monk]

-		timeout = msecs_to_jiffies(amdgpu_lockup_timeout);
-		if (timeout == 0) {
-			/*
-			 * FIXME:
-			 * Delayed workqueue cannot use it directly,
-			 * so the scheduler will not use delayed workqueue if
-			 * MAX_SCHEDULE_TIMEOUT is set.
-			 * Currently keep it simple and silly.
-			 */
-			timeout = MAX_SCHEDULE_TIMEOUT;
-		}
 		r = drm_sched_init(&ring->sched, &amdgpu_sched_ops,
 				   num_hw_submission, amdgpu_job_hang_limit,
-				   timeout, ring->name);
+				   msecs_to_jiffies(amdgpu_lockup_timeout), ring->name);


What's the background for above change ? why use scheduler for KIQ ?

-----Original Message-----
From: Andrey Grodzovsky [mailto:andrey.grodzovsky at amd.com] 
Sent: 2017年12月14日 3:45
To: amd-gfx at lists.freedesktop.org; Koenig, Christian <Christian.Koenig at amd.com>
Cc: maraeo at gmail.com; Liu, Monk <Monk.Liu at amd.com>; Grodzovsky, Andrey <Andrey.Grodzovsky at amd.com>
Subject: [PATCH] drm/amdgpu: Simplify amdgpu_lockup_timeout usage.

With introduction of amdgpu_gpu_recovery we don't need any more to rely on amdgpu_lockup_timeout == 0 for disabling GPU reset.

Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky at amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c |  7 ++++++-
 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c    |  4 ++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c  | 14 +-------------
 drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c      |  2 +-
 drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c      |  2 +-
 5 files changed, 11 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index a074502..98fb9f9d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1229,6 +1229,11 @@ static void amdgpu_check_arguments(struct amdgpu_device *adev)
 			 amdgpu_vram_page_split);
 		amdgpu_vram_page_split = 1024;
 	}
+
+	if (amdgpu_lockup_timeout == 0) {
+		dev_warn(adev->dev, "lockup_timeout msut be > 0, adjusting to 10000\n");
+		amdgpu_lockup_timeout = 10000;
+	}
 }
 
 /**
@@ -2831,7 +2836,7 @@ bool amdgpu_need_backup(struct amdgpu_device *adev)
 	if (adev->flags & AMD_IS_APU)
 		return false;
 
-	return amdgpu_lockup_timeout > 0 ? true : false;
+	return amdgpu_gpu_recovery;
 }
 
 static int amdgpu_recover_vram_from_shadow(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index b734cd6..1fc5499 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -90,7 +90,7 @@ int amdgpu_disp_priority = 0;  int amdgpu_hw_i2c = 0;  int amdgpu_pcie_gen2 = -1;  int amdgpu_msi = -1; -int amdgpu_lockup_timeout = 0;
+int amdgpu_lockup_timeout = 10000;
 int amdgpu_dpm = -1;
 int amdgpu_fw_load_type = -1;
 int amdgpu_aspm = -1;
@@ -166,7 +166,7 @@ module_param_named(pcie_gen2, amdgpu_pcie_gen2, int, 0444);  MODULE_PARM_DESC(msi, "MSI support (1 = enable, 0 = disable, -1 = auto)");  module_param_named(msi, amdgpu_msi, int, 0444);
 
-MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms (default 0 = disable)");
+MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms > 0 (default 
+10000)");
 module_param_named(lockup_timeout, amdgpu_lockup_timeout, int, 0444);
 
 MODULE_PARM_DESC(dpm, "DPM support (1 = enable, 0 = disable, -1 = auto)"); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
index 854baf0..9484aed 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
@@ -410,7 +410,6 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring,  int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring,
 				  unsigned num_hw_submission)
 {
-	long timeout;
 	int r;
 
 	/* Check that num_hw_submission is a power of two */ @@ -434,20 +433,9 @@ int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring,
 
 	/* No need to setup the GPU scheduler for KIQ ring */
 	if (ring->funcs->type != AMDGPU_RING_TYPE_KIQ) {
-		timeout = msecs_to_jiffies(amdgpu_lockup_timeout);
-		if (timeout == 0) {
-			/*
-			 * FIXME:
-			 * Delayed workqueue cannot use it directly,
-			 * so the scheduler will not use delayed workqueue if
-			 * MAX_SCHEDULE_TIMEOUT is set.
-			 * Currently keep it simple and silly.
-			 */
-			timeout = MAX_SCHEDULE_TIMEOUT;
-		}
 		r = drm_sched_init(&ring->sched, &amdgpu_sched_ops,
 				   num_hw_submission, amdgpu_job_hang_limit,
-				   timeout, ring->name);
+				   msecs_to_jiffies(amdgpu_lockup_timeout), ring->name);
 		if (r) {
 			DRM_ERROR("Failed to create scheduler on ring %s.\n",
 				  ring->name);
diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
index 7ade56d..43e74ec 100644
--- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
+++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
@@ -277,7 +277,7 @@ static int xgpu_ai_mailbox_rcv_irq(struct amdgpu_device *adev,
 	int r;
 
 	/* trigger gpu-reset by hypervisor only if TDR disbaled */
-	if (amdgpu_lockup_timeout == 0) {
+	if (!amdgpu_gpu_recovery) {
 		/* see what event we get */
 		r = xgpu_ai_mailbox_rcv_msg(adev, IDH_FLR_NOTIFICATION);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c
index e05823d..da7c261 100644
--- a/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c
@@ -545,7 +545,7 @@ static int xgpu_vi_mailbox_rcv_irq(struct amdgpu_device *adev,
 	int r;
 
 	/* trigger gpu-reset by hypervisor only if TDR disbaled */
-	if (amdgpu_lockup_timeout == 0) {
+	if (!amdgpu_gpu_recovery) {
 		/* see what event we get */
 		r = xgpu_vi_mailbox_rcv_msg(adev, IDH_FLR_NOTIFICATION);
 
--
2.7.4