sravan/linux-tkg
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

linux 6.0.y: misc-additions: Add 0004-drm-sched-add-DRM_SCHED_FENCE_DONT_PIPELINE-flag and 0005-drm-amdgpu-use-DRM_SCHED_FENCE_DONT_PIPELINE-for-VM-, pending fixes for https://gitlab.freedesktop.org/drm/amd/-/issues/2113 from Christian König

Browse Source
This commit is contained in:
Tk-Glitch
2022-10-15 15:22:52 +02:00
parent d60046d9fb
commit 9e06bbfb46
1 changed files with 96 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

96
linux-tkg-patches/6.0/0012-misc-additions.patch
View File

@@ -426,3 +426,99 @@ index 4bf4ea6cbb5eee..4850dafbaa05fb 100644
} }
/* ---- Socket functions ---- */ /* ---- Socket functions ---- */
From 50e6a66675f6c9835d4f1d4f8c947d1699ce8e24 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Fri, 7 Oct 2022 09:51:13 +0200
Subject: [PATCH 4/5] drm/sched: add DRM_SCHED_FENCE_DONT_PIPELINE flag
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Setting this flag on a scheduler fence prevents pipelining of jobs
depending on this fence. In other words we always insert a full CPU
round trip before dependen jobs are pushed to the pipeline.
Signed-off-by: Christian König <christian.koenig@amd.com>
---
drivers/gpu/drm/scheduler/sched_entity.c | 3 ++-
include/drm/gpu_scheduler.h | 9 +++++++++
2 files changed, 11 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c
index 6b25b2f4f5a3..6137537aaea4 100644
--- a/drivers/gpu/drm/scheduler/sched_entity.c
+++ b/drivers/gpu/drm/scheduler/sched_entity.c
@@ -385,7 +385,8 @@ static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
}
s_fence = to_drm_sched_fence(fence);
- if (s_fence && s_fence->sched == sched) {
+ if (s_fence && s_fence->sched == sched &&
+ !test_bit(DRM_SCHED_FENCE_DONT_PIPELINE, &fence->flags)) {
/*
* Fence is from the same scheduler, only need to wait for
diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h
index addb135eeea6..289a33e80639 100644
--- a/include/drm/gpu_scheduler.h
+++ b/include/drm/gpu_scheduler.h
@@ -32,6 +32,15 @@
#define MAX_WAIT_SCHED_ENTITY_Q_EMPTY msecs_to_jiffies(1000)
+/**
+ * DRM_SCHED_FENCE_DONT_PIPELINE - Prefent dependency pipelining
+ *
+ * Setting this flag on a scheduler fence prevents pipelining of jobs depending
+ * on this fence. In other words we always insert a full CPU round trip before
+ * dependen jobs are pushed to the hw queue.
+ */
+#define DRM_SCHED_FENCE_DONT_PIPELINE DMA_FENCE_FLAG_USER_BITS
+
struct drm_gem_object;
struct drm_gpu_scheduler;
--
2.25.1
From e15e1601fba660124acd7ad41b6f61d46a1c4835 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Fri, 7 Oct 2022 10:59:58 +0200
Subject: [PATCH 5/5] drm/amdgpu: use DRM_SCHED_FENCE_DONT_PIPELINE for VM
updates
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Make sure that we always have a CPU round trip to let the submission
code correctly decide if a TLB flush is necessary or not.
Signed-off-by: Christian König <christian.koenig@amd.com>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c
index 1fd3cbca20a2..c7bf189d50de 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c
@@ -115,8 +115,15 @@ static int amdgpu_vm_sdma_commit(struct amdgpu_vm_update_params *p,
amdgpu_bo_fence(p->vm->root.bo, f, true);
}
- if (fence && !p->immediate)
+ if (fence && !p->immediate) {
+ /*
+ * Most hw generations now have a separate queue for page table
+ * updates, but when the queue is shared with userspace we need
+ * the extra CPU round trip to correctly flush the TLB.
+ */
+ set_bit(DRM_SCHED_FENCE_DONT_PIPELINE, &f->flags);
swap(*fence, f);
+ }
dma_fence_put(f);
return 0;
--
2.25.1