1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
|
From 5c2be5f5d4d600e258d4d6465ed621952ab67c1f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Marek=20Ol=C5=A1=C3=A1k?= <marek.olsak@amd.com>
Date: Wed, 19 Jun 2019 19:26:59 -0400
Subject: [PATCH 3354/4256] drm/amdgpu: handle
AMDGPU_IB_FLAG_RESET_GDS_MAX_WAVE_ID on gfx10
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
---
drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 21 +++++++++++++++++++--
1 file changed, 19 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
index ba8f66f27d2f..7d3e6a3161dd 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
@@ -4406,6 +4406,22 @@ static void gfx_v10_0_ring_emit_ib_compute(struct amdgpu_ring *ring,
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 control = INDIRECT_BUFFER_VALID | ib->length_dw | (vmid << 24);
+ /* Currently, there is a high possibility to get wave ID mismatch
+ * between ME and GDS, leading to a hw deadlock, because ME generates
+ * different wave IDs than the GDS expects. This situation happens
+ * randomly when at least 5 compute pipes use GDS ordered append.
+ * The wave IDs generated by ME are also wrong after suspend/resume.
+ * Those are probably bugs somewhere else in the kernel driver.
+ *
+ * Writing GDS_COMPUTE_MAX_WAVE_ID resets wave ID counters in ME and
+ * GDS to 0 for this ring (me/pipe).
+ */
+ if (ib->flags & AMDGPU_IB_FLAG_RESET_GDS_MAX_WAVE_ID) {
+ amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
+ amdgpu_ring_write(ring, mmGDS_COMPUTE_MAX_WAVE_ID);
+ amdgpu_ring_write(ring, ring->adev->gds.gds_compute_max_wave_id);
+ }
+
amdgpu_ring_write(ring, PACKET3(PACKET3_INDIRECT_BUFFER, 2));
BUG_ON(ib->gpu_addr & 0x3); /* Dword align */
amdgpu_ring_write(ring,
@@ -5142,7 +5158,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_compute = {
SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 +
2 + /* gfx_v10_0_ring_emit_vm_flush */
8 + 8 + 8, /* gfx_v10_0_ring_emit_fence x3 for user fence, vm fence */
- .emit_ib_size = 4, /* gfx_v10_0_ring_emit_ib_compute */
+ .emit_ib_size = 7, /* gfx_v10_0_ring_emit_ib_compute */
.emit_ib = gfx_v10_0_ring_emit_ib_compute,
.emit_fence = gfx_v10_0_ring_emit_fence,
.emit_pipeline_sync = gfx_v10_0_ring_emit_pipeline_sync,
@@ -5175,7 +5191,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_kiq = {
SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 +
2 + /* gfx_v10_0_ring_emit_vm_flush */
8 + 8 + 8, /* gfx_v10_0_ring_emit_fence_kiq x3 for user fence, vm fence */
- .emit_ib_size = 4, /* gfx_v10_0_ring_emit_ib_compute */
+ .emit_ib_size = 7, /* gfx_v10_0_ring_emit_ib_compute */
.emit_ib = gfx_v10_0_ring_emit_ib_compute,
.emit_fence = gfx_v10_0_ring_emit_fence_kiq,
.test_ring = gfx_v10_0_ring_test_ring,
@@ -5255,6 +5271,7 @@ static void gfx_v10_0_set_gds_init(struct amdgpu_device *adev)
case CHIP_NAVI10:
default:
adev->gds.gds_size = 0x10000;
+ adev->gds.gds_compute_max_wave_id = 0x4ff;
adev->gds.vgt_gs_max_wave_id = 0x3ff;
break;
}
--
2.17.1
|
