diff options
Diffstat (limited to 'meta-amd-bsp/recipes-kernel/linux/linux-yocto-4.19.8/2839-drm-amdgpu-don-t-invalidate-caches-in-RELEASE_MEM-on.patch')
| -rw-r--r-- | meta-amd-bsp/recipes-kernel/linux/linux-yocto-4.19.8/2839-drm-amdgpu-don-t-invalidate-caches-in-RELEASE_MEM-on.patch | 48 |
1 files changed, 48 insertions, 0 deletions
diff --git a/meta-amd-bsp/recipes-kernel/linux/linux-yocto-4.19.8/2839-drm-amdgpu-don-t-invalidate-caches-in-RELEASE_MEM-on.patch b/meta-amd-bsp/recipes-kernel/linux/linux-yocto-4.19.8/2839-drm-amdgpu-don-t-invalidate-caches-in-RELEASE_MEM-on.patch new file mode 100644 index 00000000..3e99b0a8 --- /dev/null +++ b/meta-amd-bsp/recipes-kernel/linux/linux-yocto-4.19.8/2839-drm-amdgpu-don-t-invalidate-caches-in-RELEASE_MEM-on.patch @@ -0,0 +1,48 @@ +From 02dd67e78d090a43ff341c0d79d9b87262ccd13a Mon Sep 17 00:00:00 2001 +From: =?UTF-8?q?Marek=20Ol=C5=A1=C3=A1k?= <marek.olsak@amd.com> +Date: Fri, 28 Jun 2019 18:31:26 -0400 +Subject: [PATCH 2839/2940] drm/amdgpu: don't invalidate caches in RELEASE_MEM, + only do the writeback +MIME-Version: 1.0 +Content-Type: text/plain; charset=UTF-8 +Content-Transfer-Encoding: 8bit + +This RELEASE_MEM use has the Release semantic, which means we should write +back but not invalidate. Invalidations only make sense with the Acquire +semantic (ACQUIRE_MEM), or when RELEASE_MEM is used to do the combined +Acquire-Release semantic, which is a barrier, not a fence. + +The undesirable side effect of doing invalidations for the Release semantic +is that it invalidates caches while shaders are running, because the Release +can execute in the middle of the next IB. + +UMDs should use ACQUIRE_MEM at the beginning of IBs. Doing cache +invalidations for a fence (like in this case) doesn't do anything +for correctness. + +Signed-off-by: Marek Olšák <marek.olsak@amd.com> +Acked-by: Alex Deucher <alexander.deucher@amd.com> +--- + drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 6 +----- + 1 file changed, 1 insertion(+), 5 deletions(-) + +diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +index e5e3fc88de46..09b1634b97c2 100644 +--- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c ++++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +@@ -4339,11 +4339,7 @@ static void gfx_v10_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, + amdgpu_ring_write(ring, PACKET3(PACKET3_RELEASE_MEM, 6)); + amdgpu_ring_write(ring, (PACKET3_RELEASE_MEM_GCR_SEQ | + PACKET3_RELEASE_MEM_GCR_GL2_WB | +- PACKET3_RELEASE_MEM_GCR_GL2_INV | +- PACKET3_RELEASE_MEM_GCR_GL2_US | +- PACKET3_RELEASE_MEM_GCR_GL1_INV | +- PACKET3_RELEASE_MEM_GCR_GLV_INV | +- PACKET3_RELEASE_MEM_GCR_GLM_INV | ++ PACKET3_RELEASE_MEM_GCR_GLM_INV | /* must be set with GLM_WB */ + PACKET3_RELEASE_MEM_GCR_GLM_WB | + PACKET3_RELEASE_MEM_CACHE_POLICY(3) | + PACKET3_RELEASE_MEM_EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) | +-- +2.17.1 + |