This is the mail archive of the
mailing list for the cygwin project.
Re: Fixing strace and cygcheck so that they work with mount -X
- From: Christopher Faylor <cgf-no-personal-reply-please at cygwin dot com>
- To: Talk Amongst Yourselves <cygwin-talk at cygwin dot com>
- Date: Wed, 11 May 2005 12:47:11 -0400
- Subject: Re: Fixing strace and cygcheck so that they work with mount -X
- References: <20050511160233.GG10119@trixie.casa.cgf.cx> <SERRANOVqtK1rls2j2q00000314@SERRANO.CAM.ARTIMI.COM>
- Reply-to: Talk Amongst Yourselves <cygwin-talk at cygwin dot com>
- Reply-to: cygwin-talk at cygwin dot com
On Wed, May 11, 2005 at 05:41:56PM +0100, Dave Korn wrote:
>>From: Christopher Faylor
>>Sent: 11 May 2005 17:03
>>Yeah, were you following the discussion related to this in the
>>gdb-patches mailing list?
>Nope, I missed that... gotta reference?
It was wrt "mi" and what it should do with absolute path references,
subject "[RFC] fullname attribute for GDB/MI stack frames". Eli was
maintaining that Windows had some kind of deep logic about maintaining
the pwd on every drive on your machine and I was mentioning that
the drive info is stored in the environment.
>>Anyway, now that you've read the documentation, do you want to provide
>>in-depth technical help on how to construct an environment block?
>Sounds like the sort of thing I could knock out in my sleep. Point me
>at the thread so I know what it's needed for.
You're responding to the thread. I modified the cygwin DLL so that a
non-cygwin program can get an _environ style block so that it will still
work in a "mount -X" directory. cygstart uses CreateProcess to start
programs so, even though cygstart is a cygwin process, it needs to know
about the cygwin environment. It doesn't even need any of the new logic
that I just added to the DLL. It just needs to construct a windows
environment block from cygwin's UNIX-like environment and pass that to