NetBSD-Bugs archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: kern/55229: cond_timedwait_race now fails randomly on real hardware



The following reply was made to PR kern/55229; it has been noted by GNATS.

From: Andreas Gustafsson <gson%gson.org@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: gson%gson.org@localhost (Andreas Gustafsson)
Subject: Re: kern/55229: cond_timedwait_race now fails randomly on real hardware
Date: Wed, 6 May 2020 10:40:07 +0300

 I have now narrowed this down to the following period of build
 breakage:
 
   2020.04.24.03.22.06 thorpej src/sys/compat/linux/common/linux_exec.c 1.122
   2020.04.24.03.22.06 thorpej src/sys/compat/linux/common/linux_sched.c 1.75
   2020.04.24.03.22.06 thorpej src/sys/kern/kern_exec.c 1.499
   2020.04.24.03.22.06 thorpej src/sys/kern/kern_exit.c 1.289
   2020.04.24.03.22.06 thorpej src/sys/kern/kern_fork.c 1.223
   2020.04.24.03.22.06 thorpej src/sys/kern/kern_lwp.c 1.235
   2020.04.24.03.22.06 thorpej src/sys/kern/kern_proc.c 1.247
   2020.04.24.03.22.06 thorpej src/sys/kern/sys_lwp.c 1.79
   2020.04.24.03.22.06 thorpej src/sys/sys/lwp.h 1.207
   2020.04.24.03.22.06 thorpej src/sys/sys/proc.h 1.363
   2020.04.24.03.22.52 thorpej src/sys/sys/param.h 1.662
   2020.04.24.03.25.20 thorpej src/tests/lib/libc/sys/t_ptrace_wait.c 1.172
   2020.04.24.03.25.20 thorpej src/tests/lib/libc/sys/t_ptrace_x86_wait.h 1.25
   2020.04.24.03.56.12 thorpej src/sys/rump/librump/rumpkern/lwproc.c 1.46
   2020.04.24.04.34.57 ryo src/sys/dev/pci/if_aq.c 1.15
   2020.04.24.04.37.27 ryo src/share/man/man4/aq.4 1.4
   2020.04.24.04.55.40 ryo src/sys/dev/pci/if_aq.c 1.16
   2020.04.24.05.21.18 thorpej src/sys/kern/kern_proc.c 1.248
 
 To reproduce, build -current from date-based checkouts from before and
 after the above commits, boot it on real multiprocessor hardware,
 cd to /usr/tests/lib/libpthread/ and run the following command repeatedly:
 
   atf-run t_cond | atf-report
 
 For me, this works reliably (no failure in 20 runs) before the
 above commits, and fails quickly (in about 50% of the runs) after
 the above commits.
 -- 
 Andreas Gustafsson, gson%gson.org@localhost
 


Home | Main Index | Thread Index | Old Index