A Saturday 14 June 2008 01:15:45, Pierre Muller wrote:

>   Several answers concerned the reasons of the tty problems,
> but I was mainly interested in getting
> more reliable testsuite results for windows targets.
>   Lots of tests fail simply because the programs think that
> they are not on a terminal and thus use full buffering, which is
> incompatible with the testsuite logic.
>
> But, as Pedro said, individual files are not the right location to
> fix this.
>
>   I finally came up with a different idea, that seems to work
> now quite well.
>
>   The idea is simply to add an object that call setbuf(stdout,NULL)
> in a function that is added to the ctor section.
>   The source
> gdb.arch/cygwin-ctor.c is simply:
> $ cat ../gdb.arch/cygwin-ctor.c
>
> #include <stdio.h>
>
> void
> disableoutbuffer ()
> {
>   setbuf (stdout, 0);
> }
>
> asm (".section .ctor");
> asm(" .long     _disableoutbuffer");
>
> <<< End of cygwin-ctor.c file >>>

Pierre, this is exactly how we have fixed this issue for our
internal mingw testing.  I apologize for not telling you about it
before, and not sparring you the trouble.

We have the attached patch installed in our tree for a few
months now, and never seen a problem with it -- and we use
the same tree for linux testing too.  It works OK in remote
host testing too, and avoids as much as possible interfering
with test flags, and also avoids compiling the magic
object more than once per testsuite run.

> This file can be compile with gcc as cygwin-ctor.o
> or with gcc -mno-cygwin as mingw-ctor.o (allowing to
> use it for mingw tests, even if those still have lots of other problems).
>

>   The gdb_compile procedure change is rather simple,
> but it could probably be made more universal by converting this into
> a target info property that would be set in site.exp file for instance.

That's the reason I didn't submit this patch right away.  On the one
hand, this is a problem with the test environment, e.g., using ssh
to test mingw.  I'm not sure if running tests under MSFT's
telnetd/cmd.exe would still have this problem or not.  OTOH, cygwin
inferiors in a cygwin superior don't recognize they are running
under a tty, because GDB invokes them with CreateProcess, so
this is a GDB/Cygwin problem, not a test environment problem.

All in all, my opinion is that this is hitting everyone that tests
on Windows, so deserves to go into common GDB testsuite code.

What do others think?

-- 
Pedro Alves