* RE: the problem about difference between gcc2.7 and gcc2.96
@ 2001-10-15 12:03 Paolo Carlini
0 siblings, 0 replies; 3+ messages in thread
From: Paolo Carlini @ 2001-10-15 12:03 UTC (permalink / raw)
To: mikema; +Cc: gcc-help
Hi,
a few sparse considerations (not about the specific problem you mention):
1- Despite its misleading name, gcc2.96 is *not* an official GNU release (see,
f.i., http://gcc.gnu.org/ml/gcc-announce/2000/msg00003.html ) and you should more
appropriately contact the RedHat support.
2- If you want to stick with RedHat development tools, you may consider trying
gcc2.96-99 (vs 2.96-81, right?) which is part of the "unstable" Rawhide
distribution (f.i., ftp://ftp.redhat.com/pub/redhat/linux/rawhide/ ) and in
general is less buggy.
3- If you are still in the process of updating your development chain (from the
very old 2.7.x) consider testing GNU's gcc2.95.3 or the new 3.0.x series of
official releases (3.0.x is also available as *rpm packages in the Rawhide
distribution, by the way).
Regards,
Paolo Carlini.
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: the problem about difference between gcc2.7 and gcc2.96
2001-10-13 18:27 ma, yi
@ 2001-12-02 8:27 ` Alexandre Oliva
0 siblings, 0 replies; 3+ messages in thread
From: Alexandre Oliva @ 2001-12-02 8:27 UTC (permalink / raw)
To: ma,yi; +Cc: gcc-help
On Oct 13, 2001, "ma,yi" <mikema@cc.gatech.edu> wrote:
> : "=a" (*(uint32 *)dst), "=d" (*(((uint32 *)dst) + 1))
> :
> : "eax", "edx");
It is illegal to specify an output operand (=a) that is also marked as
clobbered (eax). Earlier versions of GCC didn't catch this
inconsistency.
--
Alexandre Oliva Enjoy Guarana', see http://www.ic.unicamp.br/~oliva/
Red Hat GCC Developer aoliva@{cygnus.com, redhat.com}
CS PhD student at IC-Unicamp oliva@{lsd.ic.unicamp.br, gnu.org}
Free Software Evangelist *Please* write to mailing lists, not to me
^ permalink raw reply [flat|nested] 3+ messages in thread
* the problem about difference between gcc2.7 and gcc2.96
@ 2001-10-13 18:27 ma, yi
2001-12-02 8:27 ` Alexandre Oliva
0 siblings, 1 reply; 3+ messages in thread
From: ma, yi @ 2001-10-13 18:27 UTC (permalink / raw)
To: gcc-help
Dear all,
I have a software that can been compiled and run well under redhat 5.2 (gcc 2.7). However I can not compile it now under redhat7.1(gcc 2.96). I don't know why!
The error is:
i586-msrs.h:122: Can't find a register in class 'AREG' while reloading
'asm'
the error happened in the lines below:
/* In a minimal tight loop with no unrolling, a micro benchmark shows this
as requiring 16 cycles */
extern __inline__ void
read_tsc(uint64 *dst)
{
__asm__ __volatile__("rdtsc"
: "=a" (*(uint32 *)dst), "=d" (*(((uint32 *)dst) + 1))
:
: "eax", "edx");
}
About the software, there are many other problems under gcc 2.96. I don't
know why. can you give me some suggestion or solution to solve this
problem?Thank you very much!
Best Regards,
Ma
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2001-12-02 16:27 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2001-10-15 12:03 the problem about difference between gcc2.7 and gcc2.96 Paolo Carlini
-- strict thread matches above, loose matches on Subject: below --
2001-10-13 18:27 ma, yi
2001-12-02 8:27 ` Alexandre Oliva
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).