Re: valgrind complains about WaitEventSetWaitBlock on HEAD (fe9ba28e)

On 2017-10-06 11:20:05 +0900, Michael Paquier wrote:
> Hi all,
>
> While running valgrind on latest HEAD (suppression list included), I
> am seeing complains with epoll_pwait() on Linux:
> ==12692== Syscall param epoll_pwait(sigmask) points to unaddressable byte(s)
> ==12692== at 0x62F72D0: epoll_pwait (in /usr/lib/libc-2.26.so)
> ==12692== by 0x5D819C: WaitEventSetWaitBlock (latch.c:1048)
> ==12692== by 0x5D8060: WaitEventSetWait (latch.c:1000)
> ==12692== by 0x5D774D: WaitLatchOrSocket (latch.c:385)
> ==12692== by 0x5D7619: WaitLatch (latch.c:339)
> ==12692== by 0x582E57: ApplyLauncherMain (launcher.c:976)
> ==12692== by 0x550BA7: StartBackgroundWorker (bgworker.c:841)
> ==12692== by 0x5658BC: do_start_bgworker (postmaster.c:5693)
> ==12692== by 0x565C37: maybe_start_bgworkers (postmaster.c:5897)
> ==12692== by 0x5620FC: reaper (postmaster.c:2887)
> ==12692== by 0x4E4AD9F: ??? (in /usr/lib/libpthread-2.26.so)
> ==12692== by 0x62EEAA6: select (in /usr/lib/libc-2.26.so)
> ==12692== Address 0x0 is not stack'd, malloc'd or (recently) free'd
>
> I have not looked at that very closely, but this does not look normal,
> hence this post.

I think this might be more a valgrind bug than anything. Note the
"sigmask points to to unaddressable byte" and "Address 0x0 is not
stack'd, malloc'd or (recently) free'd" bits. It's valid to pass a NULL
sigmask argument.

Greetings,

Andres Freund