This is the mail archive of the
mailing list for the Cygwin project.
Re: Trouble with running cygwin dll on Vortex86MX+ CPU
- From: Colin <colin dot wall at gallagher dot co>
- To: cygwin at cygwin dot com
- Date: Fri, 4 Apr 2014 09:44:19 +0000 (UTC)
- Subject: Re: Trouble with running cygwin dll on Vortex86MX+ CPU
- Authentication-results: sourceware.org; auth=none
- References: <loom dot 20140403T003453-906 at post dot gmane dot org> <20140404083359 dot GD2508 at calimero dot vinschen dot de>
Corinna Vinschen <corinna-cygwin <at> cygwin.com> writes:
> > In discussing this with the embedded PC supplier, he suggests that the
> > cygwin1.dll is exiting because it doesn't recognise the CPU. Is this
> > explanation plausible? And if it is, is there a solution available, or
> > must I give up on using cygwin for this application?
> As far as the pre-built Cygwin goes, I fear that you can't do a lot.
> You could try to build Cygwin in i486 mode, and if that works, you have
> to make sure to build applications against the newly built crt0.o and
> libcygwin.a. But that's not all to keep in mind. When building
> applications, parts of GCC are linked in as well, namely crtbegin.o and
> crtend.o, and sometimes cyggcc_s-1.dll. All these may or may not use
> post-i486 opcodes, so if push comes to shove you'd end up having to
> rebuild GCC for i486 as well.
> I'm not trying to discourage you, but you might have a lot of
> experimenting to do to get this working.
Thanks Corinna for this information. Now I understand the underlying issue
and can decide how to move forward.
Given that my application is not so big, I suspect that the pragmatic
course will be to re-code it for Win32.
Problem reports: http://cygwin.com/problems.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple