|
From: Nicholas N. <nj...@cs...> - 2005-06-19 04:35:17
|
Hi, The overnight tests all failed, in various different ways. I've made lots of changes today, but everything builds and runs ok on my machine. The compile errors look really strange, some of the messages don't match the code, AFAICT, and some of the errors just seem completely bogus. And they're in files I haven't changed. And who knows about the cases where it compiled ok but failed to run any of the tests correctly. Julian, Tom, if you could take a look at the failures that would be great. Thanks. Nick |
|
From: Robert W. <rj...@du...> - 2005-06-19 07:09:19
|
> The overnight tests all failed, in various different ways. I've made lot=
s=20
> of changes today, but everything builds and runs ok on my machine. The=20
> compile errors look really strange, some of the messages don't match the=20
> code, AFAICT, and some of the errors just seem completely bogus. And=20
> they're in files I haven't changed. And who knows about the cases where=20
> it compiled ok but failed to run any of the tests correctly.
>=20
> Julian, Tom, if you could take a look at the failures that would be great=
.
> Thanks.
It looks like a bad #define somewhere. If you look at the area around
vki-amd64-linux.h:297, it looks like this:
struct vki_stat {
unsigned long st_dev;
unsigned long st_ino;
unsigned long st_nlink;
=20
unsigned int st_mode;
unsigned int st_uid;
unsigned int st_gid;
unsigned int __pad0;
unsigned long st_rdev;
long st_size;
long st_blksize;
long st_blocks; /* Number 512-byte blocks a=
llocated. */
=20
unsigned long st_atime;
unsigned long st_atime_nsec;
unsigned long st_mtime;
unsigned long st_mtime_nsec;
unsigned long st_ctime;
unsigned long st_ctime_nsec;
long __unused[3];
};
After this makes it through the pre-processor, it looks like this:
# 283 "../include/vki-amd64-linux.h"
struct vki_stat {
unsigned long st_dev;
unsigned long st_ino;
unsigned long st_nlink;
=20
unsigned int st_mode;
unsigned int st_uid;
unsigned int st_gid;
unsigned int __pad0;
unsigned long st_rdev;
long st_size;
long st_blksize;
long st_blocks;
=20
unsigned long st_atim.tv_sec;
unsigned long st_atime_nsec;
unsigned long st_mtim.tv_sec;
unsigned long st_mtime_nsec;
unsigned long st_ctim.tv_sec;
unsigned long st_ctime_nsec;
long __unused[3];
};
In /usr/include/bits/stat.h, I see this:
# define st_atime st_atim.tv_sec /* Backward compatibility. */
plus many more that do other similar munging. Anyway, enough
investigating for one night - long day tomorrow, so got to hit the sack.
Regards,
Robert.
--=20
Robert Walsh
Amalgamated Durables, Inc. - "We don't make the things you buy."
Email: rj...@du...
|
|
From: Nicholas N. <nj...@cs...> - 2005-06-19 15:47:21
|
On Sun, 19 Jun 2005, Robert Walsh wrote:
> It looks like a bad #define somewhere.
>
> After this makes it through the pre-processor, it looks like this:
>
> # 283 "../include/vki-amd64-linux.h"
> struct vki_stat {
> unsigned long st_dev;
> unsigned long st_ino;
> unsigned long st_nlink;
>
> unsigned int st_mode;
> unsigned int st_uid;
> unsigned int st_gid;
> unsigned int __pad0;
> unsigned long st_rdev;
> long st_size;
> long st_blksize;
> long st_blocks;
>
> unsigned long st_atim.tv_sec;
> unsigned long st_atime_nsec;
> unsigned long st_mtim.tv_sec;
> unsigned long st_mtime_nsec;
> unsigned long st_ctim.tv_sec;
> unsigned long st_ctime_nsec;
> long __unused[3];
> };
>
> In /usr/include/bits/stat.h, I see this:
>
> # define st_atime st_atim.tv_sec /* Backward compatibility. */
I just committed a change to m_main.c so it no longer #includes
<sys/stat.h>, which I guess was dragging in bits/stat.h. See if it works
now.
Nick
|
|
From: Julian S. <js...@ac...> - 2005-06-19 16:00:48
|
Similarly, it seems ume.c was messed up by <fcntl.h> on my SuSE box.
I just committed a (very crappy) "fix". (r3949). Some
de-glibc-ification of this file is needed for a proper fix.
J
On Sunday 19 June 2005 16:47, Nicholas Nethercote wrote:
> On Sun, 19 Jun 2005, Robert Walsh wrote:
> > It looks like a bad #define somewhere.
> >
> > After this makes it through the pre-processor, it looks like this:
> >
> > # 283 "../include/vki-amd64-linux.h"
> > struct vki_stat {
> > unsigned long st_dev;
> > unsigned long st_ino;
> > unsigned long st_nlink;
> >
> > unsigned int st_mode;
> > unsigned int st_uid;
> > unsigned int st_gid;
> > unsigned int __pad0;
> > unsigned long st_rdev;
> > long st_size;
> > long st_blksize;
> > long st_blocks;
> >
> > unsigned long st_atim.tv_sec;
> > unsigned long st_atime_nsec;
> > unsigned long st_mtim.tv_sec;
> > unsigned long st_mtime_nsec;
> > unsigned long st_ctim.tv_sec;
> > unsigned long st_ctime_nsec;
> > long __unused[3];
> > };
> >
> > In /usr/include/bits/stat.h, I see this:
> >
> > # define st_atime st_atim.tv_sec /* Backward compatibility. */
>
> I just committed a change to m_main.c so it no longer #includes
> <sys/stat.h>, which I guess was dragging in bits/stat.h. See if it works
> now.
>
> Nick
>
>
> -------------------------------------------------------
> SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
> from IBM. Find simple to follow Roadmaps, straightforward articles,
> informative Webcasts and more! Get everything you need to get up to
> speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
> _______________________________________________
> Valgrind-developers mailing list
> Val...@li...
> https://lists.sourceforge.net/lists/listinfo/valgrind-developers
|