This is the mail archive of the libc-hacker@sources.redhat.com mailing list for the glibc project.

Note that libc-hacker is a closed list. You may look at the archives of this list, but subscription and posting are not open.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: bits/stat.h


On Thu, Feb 07, 2002 at 11:51:03PM -0800, Ulrich Drepper wrote:
> I had to undo Jakub's patch as well.  We are back where we are
> started.  The change was not correct as can be easily seen by the fact
> that the stat and stat64 structures if _FILE_OFFSET_BITS=64 had
> different padding.  I also tried to compiel some code with such a
> patch applied to my system's include file and it failed miserably.

Can you show me what the problems were?

I've just tried attached test, with both -D_FILE_OFFSET_BITS=64 and nothing,
and for each of nothing and -fpack-struct the output was identical (minus
__alignof__ for obvious reasons). With -DOLD the output was the same too,
unless -fpack-struct was given, in which case it aborts with LFS and prints
different results without. If struct stat64 uses unsigned int __pad*, then
IMHO struct stat should do the same.
-fpack-struct is not useful anyway, since unless the user is very careful,
he'll misalign struct stat which is on several architectures SIGBUS.

	Jakub

Attachment: foo.c
Description: Text document


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]