Cleanup atomics and fix deadlock in DP bucket_can_pool() #1151
Conversation
|
Compute Benchmarks run (with params: ): |
bratpiorka
force-pushed
the
rrudnick_fix_dp_atomic2
branch
from
7fe2ca1 to
6d6cb89
Compare
|
Compute Benchmarks run (): Summary(Emphasized values are the best results) Improved 7 (threshold 2.00%)
Regressed 4 (threshold 2.00%)
Performance change in benchmark groupsUMFRelative perf in group multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:500000/threads:4 (5)
Relative perf in group multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:500000/threads:1 (5)
Relative perf in group multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:4096/granularity:8/iterations:500000/threads:4 (5)
Relative perf in group multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:4096/granularity:8/iterations:500000/threads:1 (5)
Relative perf in group multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:128/granularity:8/iterations:500000/threads:4 (5)
Relative perf in group multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:128/granularity:8/iterations:500000/threads:1 (5)
Relative perf in group multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:50000/threads:1 (2)
DetailsBenchmark details - environment, command...multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:500000/threads:4 glibcCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:500000/threads:1 glibcCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:4096/granularity:8/iterations:500000/threads:4 glibcCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:4096/granularity:8/iterations:500000/threads:1 glibcCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:128/granularity:8/iterations:500000/threads:4 glibcCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:128/granularity:8/iterations:500000/threads:1 glibcCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:50000/threads:1 proxy_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:50000/threads:1 os_providerCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:500000/threads:4 disjoint_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:500000/threads:1 disjoint_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:4096/granularity:8/iterations:500000/threads:4 disjoint_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:4096/granularity:8/iterations:500000/threads:1 disjoint_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:128/granularity:8/iterations:500000/threads:4 disjoint_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:128/granularity:8/iterations:500000/threads:1 disjoint_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:500000/threads:4 jemalloc_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:500000/threads:1 jemalloc_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:4096/granularity:8/iterations:500000/threads:4 jemalloc_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:4096/granularity:8/iterations:500000/threads:1 jemalloc_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:128/granularity:8/iterations:500000/threads:4 jemalloc_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:128/granularity:8/iterations:500000/threads:1 jemalloc_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:500000/threads:4 scalable_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:500000/threads:1 scalable_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:4096/granularity:8/iterations:500000/threads:4 scalable_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:4096/granularity:8/iterations:500000/threads:1 scalable_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:128/granularity:8/iterations:500000/threads:4 scalable_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:128/granularity:8/iterations:500000/threads:1 scalable_poolCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:500000/threads:4 umfProxyCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv --benchmark_filter=glibc Environment Variables:LD_PRELOAD=/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/lib/libumf_proxy.so multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/size:4096/iterations:500000/threads:1 umfProxyCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv --benchmark_filter=glibc Environment Variables:LD_PRELOAD=/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/lib/libumf_proxy.so multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:4096/granularity:8/iterations:500000/threads:4 umfProxyCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv --benchmark_filter=glibc Environment Variables:LD_PRELOAD=/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/lib/libumf_proxy.so multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:4096/granularity:8/iterations:500000/threads:1 umfProxyCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv --benchmark_filter=glibc Environment Variables:LD_PRELOAD=/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/lib/libumf_proxy.so multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:128/granularity:8/iterations:500000/threads:4 umfProxyCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv --benchmark_filter=glibc Environment Variables:LD_PRELOAD=/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/lib/libumf_proxy.so multiple_malloc_free/max_allocs:10000/thread_local_allocations:1/min_size:8/max_size:128/granularity:8/iterations:500000/threads:1 umfProxyCommand:/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/benchmark/umf-benchmark --benchmark_format=csv --benchmark_filter=glibc Environment Variables:LD_PRELOAD=/home/test-user/actions-runners/umf-perf-runner/_work/unified-memory-framework/unified-memory-framework/umf-repo/build/lib/libumf_proxy.so |
src/pool/pool_disjoint.c
Outdated
| // If this was an empty slab, it was in the pool. | ||
| // Now it is no longer in the pool, so update count. | ||
| --bucket->chunked_slabs_in_pool; | ||
| bucket_decrement_pool(bucket); | ||
| uint64_t size_to_add = bucket_slab_alloc_size(bucket); | ||
| utils_fetch_and_sub_u64(&bucket->shared_limits->total_size, |
There was a problem hiding this comment.
Why did you remove the assert from here?
bratpiorka
force-pushed
the
rrudnick_fix_dp_atomic2
branch
from
6d6cb89 to
76a9875
Compare
bratpiorka
force-pushed
the
rrudnick_fix_dp_atomic2
branch
from
76a9875 to
9478a27
Compare
| @@ -14,3 +14,40 @@ | |||
| fun:critnib_find | |||
| ... | |||
| } | |||
|
|
|||
| { | |||
| [false-positive] Double check locking pattern in trackingOpenIpcHandle | |||
There was a problem hiding this comment.
Do you add extra false positives in this patch? This looks supspicius
There was a problem hiding this comment.
We already have the same suppression for IPC tests with other providers (for example see drd-umf_test-provider_file_memory_ipc.supp). Looks like the same false positive started to reproduce with other providers on the same IPC tests.
bratpiorka
force-pushed
the
rrudnick_fix_dp_atomic2
branch
from
9478a27 to
ab4a76f
Compare
Cleanup atomics and replace while(true) loop in Disjoint Pool bucket_can_pool() with a pair of atomic add/sub.
fix for #1125 and #1115
This PR is required by #1143