diff options
Diffstat (limited to 'results/classifier/105/instruction/1111')
| -rw-r--r-- | results/classifier/105/instruction/1111 | 31 |
1 files changed, 31 insertions, 0 deletions
diff --git a/results/classifier/105/instruction/1111 b/results/classifier/105/instruction/1111 new file mode 100644 index 00000000..35e76d27 --- /dev/null +++ b/results/classifier/105/instruction/1111 @@ -0,0 +1,31 @@ +instruction: 0.926 +graphic: 0.901 +device: 0.845 +semantic: 0.744 +network: 0.672 +vnc: 0.525 +socket: 0.403 +mistranslation: 0.372 +boot: 0.353 +KVM: 0.174 +other: 0.166 +assembly: 0.125 + +Calling FUTEX_LOCK_PI with qemu-x86_64-static caused ENOSYS error. +Description of problem: +When I executed the command "perf bench futex lock-pi" in amd64 docker image on s390x, I got the following error. +``` +perf: thread 2: Could not lock pi-lock for 0x40006c4480 (-1): Function not implemented +perf: thread 2: Could not lock pi-lock for 0x40006c4480 (-1): Function not implemented +perf: thread 2: Could not lock pi-lock for 0x40006c4480 (-1): Function not implemented +perf: thread 2: Could not lock pi-lock for 0x40006c4480 (-1): Function not implemented +``` + +I searched for this error message in the source code of perf-bench. I think that the following system call caused ENOSYS error. +` syscall(SYS_futex, uaddr, FUTEX_LOCK_PI | opflags, val, timeout, uaddr2, val3)` +Steps to reproduce: +1. Execute the command "perf bench futex lock-pi" in amd64 docker image on s390x +2. +3. +Additional information: + |