Re: GIT get corrupted on lustre

From: Thomas Rast <trast@student.ethz.ch>
Date: Jan 21, 2013, 19:29 UTC
Message-ID: <87r4lejpx8.fsf@pctrast.inf.ethz.ch>
In-Reply-To: <kdk2ss$498$1@ger.gmane.org>

Please don't drop the Cc list!

"Brian J. Murrell" <brian@interlinx.bc.ca> writes:

Show 6 quoted lines

>> What's odd is that while I cannot reproduce the original problem, there
>> seems to be another issue/bug with utime():
>
> I wonder if this is related to http://jira.whamcloud.com/browse/LU-305.
>  That was reported as fixed in Lustre 2.0.0 and 2.1.0 but I thought I
> saw it on 2.1.1 and added a comment to the above ticket about that.

Aha, that's a very interesting bug report. My observations support yours: I managed to get EINTR during utime().

>> In the absence of it, wouldn't we in theory have to write a simple
>> loop-on-EINTR wrapper for *all* syscalls?
>
> IIUC, that's what SA_RESTART is all about.

Yes, but there's precious little clear language on when SA_RESTART is supposed to act. In all cases?

The wording on

  http://www.delorie.com/gnu/docs/glibc/libc_485.html
  http://www.delorie.com/gnu/docs/glibc/libc_498.html

leads me to believe that SA_RESTART is actually used on the glibc side of things, so that any glibc syscall wrapper not specifically equipped with the restarting behavior would return EINTR unmodified. This might explain why utime() doesn't restart like it should (assuming we work on the theory that POSIX doesn't allow an EINTR from utime() to begin with).

-- 
Thomas Rast
trast@{inf,student}.ethz.ch