This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: gcc -v
- To: Yu Tang <carino at mail dot ustc dot edu dot cn>
- Subject: Re: gcc -v
- From: Gabriel Dos Reis <gdr at codesourcery dot com>
- Date: 18 Jan 2001 14:18:37 +0100
- Cc: gcc at gcc dot gnu dot org
- Organization: CodeSourcery, LLC
- References: <Pine.GSO.4.31L2A.0101181648080.13330-100000@mail>
Yu Tang <carino@mail.ustc.edu.cn> writes:
| On 17 Jan 2001, Gabriel Dos Reis wrote:
|
| >
| > Hi,
| >
| > Is there any technical reason why "gcc -v" should output to stderr and
| > not to "stdout"? I'm in the process of adding (finally!) the new style
| > DejaGnu framework for V3 and I'm finding that "feature" to be
| > annoying as something like (the following which ran successfuly is
| > reported as a failure)
| >
| > set output [remote_exec host $libstdcxx_options(CXX) -v]
| >
| > AFAICT, we should only use stderr for diagnostics, not for normal
| > behaviour.
|
| What about a wrap prgram?It just pass arguments to GCC,and redirect
| stderr to stdout.
GCC has better to behave correctly, that is not using stderr if it
appres that it not actually reporting a watrning nor an error.
-- Gaby
CodeSourcery, LLC http://www.codesourcery.com