Message-ID: <3E386A1C.9050303@mif.vu.lt>
Date: Thu, 30 Jan 2003 00:56:12 +0100
From: Laurynas Biveinis <laurynas DOT biveinis AT mif DOT vu DOT lt>
Organization: VU MIF
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.2.1) Gecko/20021130
X-Accept-Language: lt, en, en-us
MIME-Version: 1.0
To: Richard Dawe <rich AT phekda DOT freeserve DOT co DOT uk>
CC: djgpp-workers AT delorie DOT com
Subject: Re: Take on __solve_symlinks()
References: <3E286C07 DOT 7040200 AT mif DOT vu DOT lt> <3E2943DA DOT 523770FC AT phekda DOT freeserve DOT co DOT uk> <3E297258 DOT 50906 AT mif DOT vu DOT lt> <3E2A9396 DOT EF36FE73 AT phekda DOT freeserve DOT co DOT uk> <3E305DAB DOT 8090607 AT mif DOT vu DOT lt> <3E305D5A DOT 16693047 AT phekda DOT freeserve DOT co DOT uk>
In-Reply-To: <3E305D5A.16693047@phekda.freeserve.co.uk>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 29 Jan 2003 22:53:16.0590 (UTC) FILETIME=[35EE1CE0:01C2C7E9]
Reply-To: djgpp-workers AT delorie DOT com

Richard Dawe wrote:
>>Test 3: Solving
>>c:../../../../../../../devel/djgpp/tests/libc/compat/unistd/test1
>>Test 3 passed
> 
> [snip]
> 
> This test still fails for me.
> 
> I removed all the fail* and test* in the test directory, did a cvs update of
> the entire tree, make clean && make all, rebuilt the test.
> 
> The only change I have in this tree is for fstat and constant inode numbers.
> The test doesn't even contain the code for fstat, so I don't think that's the
> problem.
> 
> Could you try a cvs update, clean and rebuild please? What other diffs do you
> have in your tree?

Did a CVS update, copied it over my old tree, rebuild (no clean step, 
however everything got rebuilt anyway). No other diffs. Copied tests to
mimic your hard disk directory tree layout. Still no failure here...
Can you set a breakpoint on offending __solve_symlinks call and tell 
contents of both arguments? (i.e. before any _fixpath'ing)

> Maybe the test "xsymlink" should take a drive specifier. It could then create
> a cross-drive symlink to any file on that disk. E.g. c:/temp/foo ->
> d:/whatever.

BTW, I wonder if we could do something like SUBST does at runtime...

--
Laurynas