Re: [PATCH] disable post-checkout test on Cygwin

From: Alex Riesen <raa.lkml@gmail.com>
Date: Mar 17, 2009, 20:28 UTC
Message-ID: <20090317202818.GA13458@blimp.localdomain>
In-Reply-To: <49BFD6DD.1010800@viscovery.net>

Johannes Sixt, Tue, Mar 17, 2009 17:59:09 +0100:

Show 13 quoted lines

> Junio C Hamano schrieb:
> > Alex Riesen <raa.lkml@gmail.com> writes:
> > 
> >> It is broken because of the tricks we have to play with
> >> lstat to get the bearable perfomance out of the call.
> >> Sadly, it disables access to Cygwin's executable attribute,
> >> which Windows filesystems do not have at all.
> > 
> > Hmm, perhaps when checking hooks to see if they are executable, Cygwin
> > port should avoid using the "tricks"?  Compared to paths inside the
> > worktree the number of hooks is a lot smaller, no?
> 
> Hmm. Nowadays, we run hooks through run_hook() in run_command.c. It uses

The problem is that copy_templates_1 does an lstat on the files in templates directory and gets 0666 mode (regular file, non-exec) for executable file under current Cygwin port. The st_mode of that lstat is passed to copy_file mentioned, which is useless now as we use the Win32 version of lstat, which doesn't do x-bit.

> access(..., X_OK), not lstat(). We don't play games with access(), do we?
>

access(..., X_OK) will return -1.