[1062] in linux-net channel archive
Re: zero-length fragments strike again
daemon@ATHENA.MIT.EDU (Werner Almesberger)
Wed Sep 6 15:51:07 1995
From: Werner Almesberger <almesber@lrc.epfl.ch>
To: linux-net@vger.rutgers.edu
Date: Wed, 6 Sep 1995 10:59:43 +0200 (MET DST)
Rick Sladkey <jrs@world.std.com> wrote:
> Actually, this is correct, I think. The C library should ignore
> aEAGAIN when it occurs beause of interrupts. If the call returned EIO
> as it should have all would be well.
Yup, occasional EAGAINs would have been fine, but the zero-length
fragments plus Solaris silently discarding the packet made fail _all_
write attempts, so libc should either have given up after a while
(returning an error) or looped forever (at least in fclose).
According to H.J. Lu, fclose is fixed in libc 5.0.9. (That libc problem
probably didn't show up on many other occasions, though.)
- Werner
--
_________________________________________________________________________
/ Werner Almesberger, DI-LRC,EPFL,CH werner.almesberger@lrc.di.epfl.ch /
/_IN_R_311__Tel_+41_21_693_6621__Fax_+41_21_693_6610_____________________/
