diff options
| author | Vineet Gupta <[email protected]> | 2016-09-28 18:53:17 +0000 |
|---|---|---|
| committer | Vineet Gupta <[email protected]> | 2017-11-21 23:21:08 +0000 |
| commit | 82385732b1c9d6a22942b5fe6e48a99891cb806f (patch) | |
| tree | cf0c5e5c0ca3f7490eaab51f8346368d24875e9c /drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | |
| parent | ARCv2: perf: optimize given that num counters <= 32 (diff) | |
| download | kernel-82385732b1c9d6a22942b5fe6e48a99891cb806f.tar.gz kernel-82385732b1c9d6a22942b5fe6e48a99891cb806f.zip | |
ARC: perf: avoid vmalloc backed mmap
For non-alising Dcache, vmalloc is not needed.
vmalloc triggers additonal D-TLB Misses in the perf interrupt code path
making it slightly inefficient as evident from hackbench runs below.
| [ARCLinux]# perf stat -e dTLB-load-misses --repeat 5 hackbench
| Running with 10*40 (== 400) tasks.
| Time: 35.060
| ...
| Performance counter stats for 'hackbench' (5 runs):
Before: 399235 dTLB-load-misses ( +- 2.08% )
After : 397676 dTLB-load-misses ( +- 2.27% )
Acked-by: Peter Zijlstra (Intel) <[email protected]>
Signed-off-by: Vineet Gupta <[email protected]>
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c')
0 files changed, 0 insertions, 0 deletions