This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH] Unbuffer stdout and stderr on windows
- From: Pedro Alves <palves at redhat dot com>
- To: Eli Zaretskii <eliz at gnu dot org>
- Cc: yao at codesourcery dot com, gdb-patches at sourceware dot org, brobecker at adacore dot com
- Date: Fri, 16 Aug 2013 15:21:50 +0100
- Subject: Re: [PATCH] Unbuffer stdout and stderr on windows
- References: <51EE23F8 dot 1070905 at codesourcery dot com> <83wqohw4ee dot fsf at gnu dot org> <20130729192559 dot GA5348 at ednor dot casa dot cgf dot cx> <83d2q1xiyv dot fsf at gnu dot org> <51F6C7B2 dot 3020400 at redhat dot com> <20130731034045 dot GA5565 at ednor dot casa dot cgf dot cx> <20130812211105 dot GA11128 at adacore dot com> <8361v9piop dot fsf at gnu dot org> <20130815173618 dot GA6955 at ednor dot casa dot cgf dot cx> <83eh9uonlg dot fsf at gnu dot org> <20130815175940 dot GD6955 at ednor dot casa dot cgf dot cx> <520E1109 dot 7000304 at redhat dot com> <520E1C34 dot 2000907 at codesourcery dot com> <520E2B13 dot 8020706 at redhat dot com> <83r4dtn35q dot fsf at gnu dot org>
On 08/16/2013 03:03 PM, Eli Zaretskii wrote:
> I suggest not to broaden the issue beyond the original problem, lest
> we will be unable to solve it.
Nobody's doing that. What I'm saying is that the testsuite is not
special, and we should _zone in_ to the real problem, and fix it
completely. Making output not buffered seems to do that, assuming
GDB doesn't output half a line to stdout, then output to stderr, and
then output the rest to stderr, which very much it will want to avoid,
given stderr is unbuffered on most systems. If the output is buffered, then
something must be forcing flushes to get the same effect. Fixing
the issue for the testsuite but not for other GDB-consumers that
spawn it in a pipe, with special modes is just not the right
solution, IMO. What will probably happen with that solution is
that Windows GDB clients will end up using the special testsuite-only
flag too, because by default, GDB will remain broken... Why add
a flag for a half-fix instead of fixing the issue completely?
--
Pedro Alves