Crashes on larger files when scrolling on OpenBSD with sysctl vm.malloc_conf=S about vim

[chrisbra]

since you seem to be able to easily reproduce, can you please share a stack trace?

[dante-robinson]

since you seem to be able to easily reproduce, can you please share a stack trace?

Not sure how to do that sorry

[byteskeptical]

@dante-robinson you can run ktrace vim yourlargefile.txt and provide the dump file after triggering the behavior. I'm running on current (7.5 build 10) also with the aforementioned sysctl flag and can't seem to reproduce this on any flavor of vim (gtk or nox11*) on a ~2400 line file. Can you reproduce the behavior in xterm? Lastly what class is the user your running this as a part of (userinfo $USER)?

[dante-robinson]

@dante-robinson you can run ktrace vim yourlargefile.txt and provide the dump file after triggering the behavior. I'm running on current (7.5 build 10) also with the aforementioned sysctl flag and can't seem to reproduce this on any flavor of vim (gtk or nox11*) on a ~2400 line file. Can you reproduce the behavior in xterm? Lastly what class is the user your running this as a part of (userinfo $USER)?

xterm causes the same issue.

groups user wheel operator wsrc staff _shutdown

ktrace.tar.gz

[byteskeptical]

@dante-robinson are you all syspatch 'ed up on this system? Given the memory access error and the trapno=6 I would suspect the recent syscall changes may be playing a role here. You wouldn't happen to be running a current flavor package on a stable system by any chance?

 67101 vim      CALL  mmap(0,0x4000,0x3<PROT_READ|PROT_WRITE>,0x1002<MAP_PRIVATE|MAP_ANON>,-1,0)
 67101 vim      RET   mmap 7136648458240/0x67da1657000
 67101 vim      CALL  mprotect(0x67da165a000,0x1000,0<PROT_NONE>)
 67101 vim      RET   mprotect 0
 67101 vim      CALL  munmap(0x67ccdc92000,0x4000)
 67101 vim      RET   munmap 0
 67101 vim      CALL  munmap(0x67da1657000,0x4000)
 67101 vim      RET   munmap 0
 67101 vim      PSIG  SIGSEGV caught handler=0x67ac95e52d0 mask=0<> code=SEGV_ACCERR addr=0x67d702a1003 trapno=6
 67101 vim      CALL  sigprocmask(SIG_BLOCK,0<>)
 67101 vim      RET   sigprocmask 0x400<SIGSEGV>
 67101 vim      CALL  sigaction(SIGHUP,0x67dc66d5b60,0x67dc66d5b70)
 67101 vim      STRU  struct sigaction { handler=SIG_IGN, mask=0<>, flags=0x2<SA_RESTART> }
 67101 vim      STRU  struct sigaction { handler=0x67ac95e52d0, mask=0<>, flags=0x1<SA_ONSTACK> }
 67101 vim      RET   sigaction 0

[chrisbra]

Please see here: https://github.com/vim/vim/wiki/Debugging-Vim-Bugs for a stack trace

[dante-robinson]

@dante-robinson are you all syspatch 'ed up on this system? Given the memory access error and the trapno=6 I would suspect the recent syscall changes may be playing a role here. You wouldn't happen to be running a current flavor package on a stable system by any chance?

 67101 vim      CALL  mmap(0,0x4000,0x3<PROT_READ|PROT_WRITE>,0x1002<MAP_PRIVATE|MAP_ANON>,-1,0)
 67101 vim      RET   mmap 7136648458240/0x67da1657000
 67101 vim      CALL  mprotect(0x67da165a000,0x1000,0<PROT_NONE>)
 67101 vim      RET   mprotect 0
 67101 vim      CALL  munmap(0x67ccdc92000,0x4000)
 67101 vim      RET   munmap 0
 67101 vim      CALL  munmap(0x67da1657000,0x4000)
 67101 vim      RET   munmap 0
 67101 vim      PSIG  SIGSEGV caught handler=0x67ac95e52d0 mask=0<> code=SEGV_ACCERR addr=0x67d702a1003 trapno=6
 67101 vim      CALL  sigprocmask(SIG_BLOCK,0<>)
 67101 vim      RET   sigprocmask 0x400<SIGSEGV>
 67101 vim      CALL  sigaction(SIGHUP,0x67dc66d5b60,0x67dc66d5b70)
 67101 vim      STRU  struct sigaction { handler=SIG_IGN, mask=0<>, flags=0x2<SA_RESTART> }
 67101 vim      STRU  struct sigaction { handler=0x67ac95e52d0, mask=0<>, flags=0x1<SA_ONSTACK> }
 67101 vim      RET   sigaction 0

Have all the current syspatches and am using the stable package from pkg_add not the -current package.