Mikhail Brinskiy
Mikhail Brinskiy
pls check CI failures. Coverity error looks relevant: ``` + '[' 1 -gt 0 ']' + cov-format-errors --dir /home/swx-azure-svc_azpcontainer/82170-20240610.12/build/cov_build_devel --emacs-style /__w/1/s/test/gtest/ucs/test_async.cc:388 Type: Calling risky function (DC.WEAK_CRYPTO) /__w/1/s/test/gtest/ucs/test_async.cc:388: dont_call: "rand" should...
some CI errors are relevant: ``` /__w/1/s/ucx_src_89282/contrib/../test/gtest/common/mem_buffer.cc: In static member function 'static void* mem_buffer::allocate(size_t, ucs_memory_type_t, bool)': /__w/1/s/ucx_src_89282/contrib/../test/gtest/common/mem_buffer.cc:242:23: error: 'cudaMallocAsync' was not declared in this scope CUDA_CALL(cudaMallocAsync(&ptr, size, 0), ": size="
When you select RC, UCX would use UD anyway as an auxiliary transport (to wireup RC connection)
The attached log is truncated. Is it a regression? (i.e. did you try to run the same app with older OMPI/UCX) Is it possible to share application source code? Seems...
CI errors are relevant
/azp run UCX PR
/azp run UCX PR
/azp run UCX PR
is `ulimit -l` set to unlimited? If yes, can you pls try to use `UCX_RCACHE_MAX_REGIONS=2048` instead of `UCX_RCACHE_ENABLE=n`