This is the mail archive of the
cygwin
mailing list for the Cygwin project.
Re: bug in .. handling
- From: Christopher Faylor <cgf-no-personal-reply-please at cygwin dot com>
- To: cygwin at cygwin dot com
- Date: Sat, 9 Apr 2005 18:00:07 -0400
- Subject: Re: bug in .. handling
- References: <42584A5D.7070405@byu.net>
- Reply-to: cygwin at cygwin dot com
On Sat, Apr 09, 2005 at 03:34:21PM -0600, Eric Blake wrote:
>Cygwin is taking too much liberty when resolving paths with .. in them.
>I've tested this with both 1.5.14 and the 20050408 snapshot. POSIX
>requires that the pathname before the .. exist, and if it is a symlink,
>that it is resolved, before going to the parent directory. For example:
>
>$ cd /tmp/example
>$ ls
>$ mkdir -p a/b
>$ ln -s a/b b
>$ touch a/c a/b/d
>$ ls b
>d
>$ ls b/.. # bug - should list contents of a, not .
>a b
>$ ls a
>b c
>$ ls none/.. # bug - should fail with ENOENT
>a b
>$ ln -s e e
>$ readlink -v e/none
>readlink: e/foo: Too many levels of symbolic links
>$ ls e/.. # bug - should fail with ELOOP
>a b e
>
>Basically, cygwin is blindly collapsing dir/.. into nothing, without first
>verifying dir, in violation of POSIX.
I think that's going to have to be a "that's too bad" since I'm not really
keen on slowing down cygwin just to blindly match POSIX behavior.
cgf
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/