This is the mail archive of the
newlib@sources.redhat.com
mailing list for the newlib project.
RE: Bug in stdio using dynamic_reent
- From: "Likely, Grant" <Grant dot Likely at gdcanada dot com>
- To: "Newlib mailing list (E-mail)" <newlib at sources dot redhat dot com>
- Date: Thu, 3 Apr 2003 09:27:27 -0700
- Subject: RE: Bug in stdio using dynamic_reent
I haven't tried a simulator yet. As soon as I figure out how to use psim
I'll try it on that.
Cheers,
g.
-----Original Message-----
From: J. Johnston [mailto:jjohnstn at redhat dot com]
Sent: Wednesday, April 02, 2003 4:06 PM
To: Likely, Grant
Subject: Re: Bug in stdio using dynamic_reent
Likely, Grant wrote:
> IO_Init simply sets up my serial IO channel so that calls to _read_r and
> _write_r will work. It does not affect the library calls. My _write_r
> function simply dumps the entire contents of *buf to the serial channel
when
> called.
>
> I've tested this on two different targets, both with no operating system.
> One is a powerpc (8260) and the other is a 5307 coldfire. I have not
> attempted it on my x86 linux box because I have not been able to build
> newlib natively. (The configure script seems to work, but the Makefile
> generated doesn't do anything).
>
You have to force newlib to build by specifying --with-newlib when you
configure.
By default, a native newlib is not built, except for Cygwin.
Have you tried a PPC simulator?
> I've done a little bit of debug, but the problem seems to be burried deep
> enough that I haven't found it yet.
>
> Am I initializing the impure data structures correctly?
>
Yes. I used the same code and it worked fine.
-- Jeff J.