bug-gnulib
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: new module 'system-quote'


From: Eli Zaretskii
Subject: Re: new module 'system-quote'
Date: Thu, 10 May 2012 08:13:33 +0300

> From: Bruno Haible <address@hidden>
> Cc: address@hidden
> Date: Thu, 10 May 2012 00:35:08 +0200
> 
> > Which calls to CreateProcess that used '*' failed?  You aren't saying
> > that the second argument to CreateProcess cannot include unquoted '*'
> > characters, are you?
> 
> If I recall it correctly, the
>   check_one (interpreter, prog, "*");
> case failed: Apparently, when you pass a single unquoted '*' argument
> to CreateProcess, it can fail with GetLastError() = ERROR_PATH_NOT_FOUND.

That is expected and reasonable: the first token in the argument is
the executable to invoke, so it cannot possibly include wildcards.  If
an application does that, it's a bug in the application.

Therefore, I think you should not quote wildcard characters such as
'*' and '?'.  If you do, it means there's no way of passing to
subsidiary programs command-line arguments that include wildcards.

> > Unfortunately, that's not so.  The version of MSVCRT.DLL that ships
> > with Vista and later was updated with the new globbing code.  (Its
> > version is also different: 7.0.7600.16385 vs 7.0.2600.5512 on XP.)
> > The function in question is __getmainargs; you can verify with
> > something like pexports that this function lives inside MSVCRT.DLL.
> 
> Oh sh*t, they're doing the globbing in the callee, rather than in the
> caller! This is terribly wrong.

It was always like that in MS-land.

> > What is the command to run the test suite for the new module?  Is it
> > 
> >   $ cd gltests && make check-TESTS TESTS=test-system-quote.sh
> 
> It is
>   $ ./configure (with the appropriate options for mingw, depending on your
>                  environment)
>   $ make
>   $ make check

Yes, I already did all that (on XP, so far), I just wanted to be sure
I got it right.

Btw, "make check" seems to become stuck close to forever in the
"test_rwlock" stage; I lost patience and interrupted it.  Is that
normal?

> If you get a failure, it's useful to comment out some of the lines
>   check_all (SCI_SYSTEM, true, prog); /* equivalent to SCI_WINDOWS_CMD */
>   check_all (SCI_WINDOWS_CREATEPROCESS, false, prog);
>   check_all (SCI_WINDOWS_CMD, true, prog);
> in gltests/test-system-quote-main.c.
> 
> > ?  Also, is it sufficient to run the test program from the MSYS shell,
> > or should it also be run from cmd.exe?
> 
> The test-system-quote.sh script should be run through bash. It doesn't
> matter how this bash is invoked.

Thanks, will do.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]