Re: [PATCH] Ignore boilerplate logo from MSVC on stderr.

[Peter Rosin]

Den 2010-08-16 21:41 skrev Ralf Wildenhues:
> * Peter Rosin wrote on Mon, Aug 16, 2010 at 09:26:57PM CEST:
>> Den 2010-08-16 20:45 skrev Ralf Wildenhues:
>>> * Peter Rosin wrote on Mon, Aug 16, 2010 at 08:14:50PM CEST:
>>>> The logo of my version of MSVC is:
> 
>>> Without having looked at the patch or your post any further at all:
>>> what about other versions of this compiler?  What about localized
>>> versions of it?
>>
>> Good catch, I knew they were very similar and assumed my matching was just
>> about wide enough. But since I didn't check, Murphy made sure that it wasn't
>> working for everybody, see below...
>>
>> Other versions:
> [...]
>> Don't know about localization, sorry... Someone else?
> 
> Let's turn the question around: Do you think it would be a more
> maintainable approach to go with what Libtool uses in its compile/link
> tests: namely, compare stderr output contents without and with the flag
> to try out?

See my reply to Eric...

> This is a tricky question!  The linker that MSVC calls actually manages
> to print its arguments; or it's cl doing that when calling the linker.
> Seems on stdout though, judging from your recent "MSVC status" mail on
> a libtool list.

The "linker input" is on stdout, yes. Everything about the linker is on
stdout. Even the linker boilerplate logo is on stdout if you run it w/o
the compiler driver. I don't know which process produces the output
though, since I can't find any linker option to reproduce that output,
but my guess would be the linker anyway.

> Does cl have a -Werror sort of flag that causes its exit status to be
> usable for setting other flags?

"/WX treat warnings as errors"

But that doesn't change anything for the exit status nor the output. At
least not for the unknown option -g and MSVC 8. Go figure...

Cheers,
Peter