Bug#949767: clblas: *gemm wrong answers in out-of-order queues
Source: clblas Followup-For: Bug #949767 X-Debbugs-Cc: witold.bary...@gmail.com Hi, any progress on this? Is this something that the upstream should look into maybe? Thanks, Witold -- System Information: Debian Release: bullseye/sid APT prefers unstable-debug APT policy: (500, 'unstable-debug'), (500, 'unstable'), (1, 'experimental') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.7.0-1-amd64 (SMP w/32 CPU threads) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled
Bug#949767: clblas: *gemm wrong answers in out-of-order queues
Control: retitle -1 clblas: *gemm wrong answers in out-of-order queues Control: reassign -1 src:clblas Control: found -1 2.12-1 I think I've found the actual bug, in clblas src/library/blas/xgemm.cc: clblasGemm (with a single command queue) enqueues up to 4 kernels and returns an event that depends on only the last of them, so if the queue is out-of-order, waiting on this event doesn't necessarily wait for all of them to finish. This was previously noticed in https://github.com/clMathLibraries/clBLAS/issues/269#issuecomment-225453543 , but not actually reported as a bug. clblas includes a client/performance tester that creates an out-of-order queue (at src/client/clfunc_common.hpp:306), implying that it intends to allow such queues. (We don't run clblas' own tests, possibly because of https://github.com/clMathLibraries/clBLAS/issues/338.) The real fix would be to return an event that depends on all the kernels' events (e.g. created with clEnqueueMarkerWithWaitList). As a workaround for now, I intend to disable out-of-order queues in libgpuarray. (It appears to be the only reverse dependency of clblas that also uses out-of-order queues.)
