drm/amdgpu/gfx_v8_0: Reorder the gfx, kiq and kcq rings test sequence

The kiq ring and the very first compute ring may fail occasionally
if they are tested directly following kiq_kcq_enable.

Insert the gfx ring test before kiq ring test to delay the kiq and kcq
ring tests will fix the issue.

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Tiecheng Zhou <Tiecheng.Zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index 381f593b0cda83bdda6f67f2d2b41cd8488583c4..164ffc91b6453ce4c547e3cea83d59600f7ec5ba 100644 (file)
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -4278,9 +4278,8 @@ static int gfx_v8_0_cp_gfx_resume(struct amdgpu_device *adev)
        amdgpu_ring_clear_ring(ring);
        gfx_v8_0_cp_gfx_start(adev);
        ring->sched.ready = true;
-       r = amdgpu_ring_test_helper(ring);
 
-       return r;
+       return 0;
 }
 
 static void gfx_v8_0_cp_compute_enable(struct amdgpu_device *adev, bool enable)
@@ -4369,10 +4368,9 @@ static int gfx_v8_0_kiq_kcq_enable(struct amdgpu_device *adev)
                amdgpu_ring_write(kiq_ring, upper_32_bits(wptr_addr));
        }
 
-       r = amdgpu_ring_test_helper(kiq_ring);
-       if (r)
-               DRM_ERROR("KCQ enable failed\n");
-       return r;
+       amdgpu_ring_commit(kiq_ring);
+
+       return 0;
 }
 
 static int gfx_v8_0_deactivate_hqd(struct amdgpu_device *adev, u32 req)
@@ -4709,16 +4707,32 @@ static int gfx_v8_0_kcq_resume(struct amdgpu_device *adev)
        if (r)
                goto done;
 
-       /* Test KCQs - reversing the order of rings seems to fix ring test failure
-        * after GPU reset
-        */
-       for (i = adev->gfx.num_compute_rings - 1; i >= 0; i--) {
+done:
+       return r;
+}
+
+static int gfx_v8_0_cp_test_all_rings(struct amdgpu_device *adev)
+{
+       int r, i;
+       struct amdgpu_ring *ring;
+
+       /* collect all the ring_tests here, gfx, kiq, compute */
+       ring = &adev->gfx.gfx_ring[0];
+       r = amdgpu_ring_test_helper(ring);
+       if (r)
+               return r;
+
+       ring = &adev->gfx.kiq.ring;
+       r = amdgpu_ring_test_helper(ring);
+       if (r)
+               return r;
+
+       for (i = 0; i < adev->gfx.num_compute_rings; i++) {
                ring = &adev->gfx.compute_ring[i];
-               r = amdgpu_ring_test_helper(ring);
+               amdgpu_ring_test_helper(ring);
        }
 
-done:
-       return r;
+       return 0;
 }
 
 static int gfx_v8_0_cp_resume(struct amdgpu_device *adev)
@@ -4739,6 +4753,11 @@ static int gfx_v8_0_cp_resume(struct amdgpu_device *adev)
        r = gfx_v8_0_kcq_resume(adev);
        if (r)
                return r;
+
+       r = gfx_v8_0_cp_test_all_rings(adev);
+       if (r)
+               return r;
+
        gfx_v8_0_enable_gui_idle_interrupt(adev, true);
 
        return 0;
@@ -5056,6 +5075,7 @@ static int gfx_v8_0_post_soft_reset(void *handle)
 {
        struct amdgpu_device *adev = (struct amdgpu_device *)handle;
        u32 grbm_soft_reset = 0;
+       struct amdgpu_ring *ring;
 
        if ((!adev->gfx.grbm_soft_reset) &&
            (!adev->gfx.srbm_soft_reset))
@@ -5086,6 +5106,8 @@ static int gfx_v8_0_post_soft_reset(void *handle)
            REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_GFX))
                gfx_v8_0_cp_gfx_resume(adev);
 
+       gfx_v8_0_cp_test_all_rings(adev);
+
        adev->gfx.rlc.funcs->start(adev);
 
        return 0;