This is the mail archive of the libc-hacker@sources.redhat.com mailing list for the glibc project.

Note that libc-hacker is a closed list. You may look at the archives of this list, but subscription and posting are not open.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: elm 2.5.3 and glibc 2.1.93


>>>>> H J Lu writes:

 > On Fri, Sep 22, 2000 at 05:33:28PM +0200, Jakub Jelinek wrote:
>> On Fri, Sep 22, 2000 at 07:31:41AM -0700, H . J . Lu wrote:
>> > On Fri, Sep 22, 2000 at 10:13:46AM +0200, Andreas Schwab wrote:
>> > > |> 
>> > > |> It looks like elm 2.5.3 may check errno even when fcntl returns 0.
>> > > 
>> > > This is broken.  The value of errno is only defined if the syscall failed.
>> > > 
>> > 
>> > Elm may be broken. But it is no excuse for glibc to mess with errno.
>> 
>> Why should glibc work around bugs in broken packages?
>> 

 > Take a look at glibc and ask why we have to save/restore errno all
 > over places where we support both new and old system calls. There are
 > reasons behind that. I don't recall the details. Someone may remember
 > it.
Initially we used to save/restore errno, later we noticed that this is
not needed.  All new code (at least code that Ulrich and I recently
added) doesn't save/restore errno - only older code does.  I agree, we
should unify this, but so far nobody spoke up.

Btw. this issue of saving/restoring errno has been discussed already
on the list and we came to the conclusion that it's not needed in this
situation.

elm seems to be broken, glibc is fine.

Andreas
-- 
 Andreas Jaeger
  SuSE Labs aj@suse.de
   private aj@arthur.inka.de
    http://www.suse.de/~aj

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]