diff options
author | seshendra Gadagottu <sgadagottu@nvidia.com> | 2017-06-19 20:40:45 -0400 |
---|---|---|
committer | mobile promotions <svcmobile_promotions@nvidia.com> | 2017-06-20 13:25:32 -0400 |
commit | 492c7cc198f3315b468fa3d612d894edf2bb5e89 (patch) | |
tree | 8b3acf6b15e1e972b30d25b755e455e8ec54860e /drivers/gpu/nvgpu/gv11b/subctx_gv11b.c | |
parent | 47eed280133ee75327942539c6d07e9b89d237f4 (diff) |
gpu: nvgpu: gv11b: make subcontext mappings non-cacheable
Without non-cacheable, gpu filled subcontext data not visible
to cpu without additional l2 flush. Similarly, there will be issues
where cpu updates to subcontext header will not visible to gpu without
additional l2 flush.
Making subcontext header mapping non-cacheable fixes this issue.
Bug 1937331
Change-Id: I8e25b7cac165e7481eec7c9f1f93bc7992183c46
Signed-off-by: seshendra Gadagottu <sgadagottu@nvidia.com>
Reviewed-on: http://git-master/r/1505283
Reviewed-by: Alex Waterman <alexw@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Diffstat (limited to 'drivers/gpu/nvgpu/gv11b/subctx_gv11b.c')
-rw-r--r-- | drivers/gpu/nvgpu/gv11b/subctx_gv11b.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/drivers/gpu/nvgpu/gv11b/subctx_gv11b.c b/drivers/gpu/nvgpu/gv11b/subctx_gv11b.c index 6aa58337..79ed0d1e 100644 --- a/drivers/gpu/nvgpu/gv11b/subctx_gv11b.c +++ b/drivers/gpu/nvgpu/gv11b/subctx_gv11b.c | |||
@@ -66,7 +66,7 @@ int gv11b_alloc_subctx_header(struct channel_gk20a *c) | |||
66 | ctx->mem.gpu_va = nvgpu_gmmu_map(c->vm, | 66 | ctx->mem.gpu_va = nvgpu_gmmu_map(c->vm, |
67 | &ctx->mem, | 67 | &ctx->mem, |
68 | ctx->mem.size, | 68 | ctx->mem.size, |
69 | NVGPU_MAP_BUFFER_FLAGS_CACHEABLE_TRUE, | 69 | NVGPU_MAP_BUFFER_FLAGS_CACHEABLE_FALSE, |
70 | gk20a_mem_flag_none, true, | 70 | gk20a_mem_flag_none, true, |
71 | ctx->mem.aperture); | 71 | ctx->mem.aperture); |
72 | if (!ctx->mem.gpu_va) { | 72 | if (!ctx->mem.gpu_va) { |