On Fri, 2012-09-28 at 21:12 -0700, Benjamin De Kosnik wrote:
> ... found while working on arm-eabisim cross, using --enable-cxx-flags
> was not working as AM_CXXFLAGS was being over-ridden by CXXFLAGS.
> (Despite the comments warning about this.)
>
> Fixed.
>
> Also patchlet for the last commit,
Hi,
for simple loops like:
extern int a[];
extern int b[];
void foo (int l)
{
int i;
for (i = 0; i < l; i++)
a[i] = b [i];
}
you get in the .lim3 dump:
Unanalyzed memory reference 0: _5 = MEM[symbol: b, index: ivtmp.3_1, step: 4,
offset: 0B];
Memory reference 1: MEM[symbol: a, index:
Hi Thomas,
> not much to say, the patch is pretty straightforward.
> Regression-tested. OK for trunk?
I'm not much of a libgfortran or I/O expert, but after all this looks
ok to me. Thanks for the patch!
Cheers,
Janus
> 2012-09-24 Thomas König
>
> PR fortran/52724
> * list_
> this patch lets the compiler try to rewrite:
>
> (vec_concat (vec_select x [a]) (vec_select x [b]))
>
> as:
>
> vec_select x [a b]
>
> or even just "x" if appropriate.
>
> In a first iteration I was restricting it to b-a==1, but it seemed better
> not to: it helps for {v[1],v[0]} and doesn't
Hello,
this patch does 2 things (I should have split it in 2, but the questions
go together):
1) it handles constant folding of vector comparisons,
2) it fixes another place where vectors are not expected (I'll probably
wait to have front-end support and testcases to do more of those, but
t
> Bootstrapped in revision 191429 on i686-pc-linux-gnu.
Bootstrap on x86 for a reorg.c patch doesn't say much though. :-)
> The arc.c context can be seen here:
>
> http://gcc.gnu.org/viewcvs/branches/arc-4_4-20090909-branch/gcc/config/arc/a
> rc.c?content-type=text%2Fplain&view=co
OK, modulo th
> Which means that -O0 should also set optimize_debug to 1? -O0 is
> then !optimize && optimize_debug and -Og is optimize == 1 &&
> optimize_debug.
That sounds more confusing than helpful though. Adding (|| optimize_debug) in
the relevant places is probably good enough.
--
Eric Botcazou
Hi,
I checked in this patch to move GCC ChangeLog entry to gcc/ChangeLog.
H.J.
---
Index: ChangeLog
===
--- ChangeLog (revision 191852)
+++ ChangeLog (working copy)
@@ -9,12 +9,6 @@
* Makefile.def: Make all-gcc depend on
The following proposed patch disables setting, saving and restoring
the VRSAVE register on all targets except Darwin.
VRSAVE was removed from the AIX ABI and was suppose to have been
removed from the PPC SVR4 ABI. All recent versions of the Linux
kernel set and maintain VRSAVE itself, as a proce
On Sat, Sep 29, 2012 at 11:08 AM, Segher Boessenkool
wrote:
>> The following proposed patch disables setting, saving and restoring
>> the VRSAVE register on all targets except Darwin.
>>
>> VRSAVE was removed from the AIX ABI and was suppose to have been
>> removed from the PPC SVR4 ABI. All rece
AIX fcntl.h redefines "open", which conflicts with g++ fstream header.
This patch wraps fcntl.h to use aliases in the same way that stdio.h
currently is wrapped.
Okay?
Thanks, David
* inclhack.def (AAB_aix_fcntl): New fix.
* fixincl.x: Regenerate.
* tests/base/fcntl.h [AAB_AIX_FCNTL_CHECK]: New
ping!
2012/9/23 Janus Weil :
> Hi all,
>
> this patch fixes an ICE recently reported by Andrew Benson. It does so
> by rejecting polymorphic arguments to the C_F_POINTER intrinsic. For
> discussion and standard references, please refer to the PR and today's
> email thread.
>
> Apart from fixing t
Hi Janus,
I'm not much of a libgfortran or I/O expert, but after all this looks
ok to me. Thanks for the patch!
Committed, thanks. What to people think about backporting to 4.7?
It is a wrong-code issue, if a rather obscure corner.
If there's no feedback within a few days, I think I'll close
PR 54749 points out that if libbacktrace fails to read the debug info
from the executable (in the case of that PR, due to memory exhaustion)
the compiler will print multiple error messages. The library already
has a mechanism to avoid this, and this patch uses that mechanism.
Bootstrapped and ran
Does "make check" complain if a replacement/wrap fix doesn't have a test?
If so, I'll be fixing it when I have time -- towards the end of October.
(The patch I sent a few days ago ought to do it, but I've not fully tested yet.)
Anyway, looks pretty reasonable to me...
On Sat, Sep 29, 2012 at 9:45
Hello world,
the attached patch fixes the PR. The logic for processing
GFORTRAN_CONVERT_UNIT had been quite wrong. I have checked
that the original test case, plus a few more, no longer cause
assertion failures or memory corruption (also checked with
valgrind). No test case because it is not p
This removes all the code for non-GCC compilers in
Tested x86_64, "make check check-parallel", committed to trunk.
PR libstdc++/34106
* include/parallel/compatibility.h: Remove non-GCC code.
commit 9c3f1fdec9fc7be67cf2c0d45d1b791f167c0aec
Author: Jonathan Wakely
Date: Thu Sep
On Sat, Sep 29, 2012 at 2:00 PM, Bruce Korb wrote:
> Does "make check" complain if a replacement/wrap fix doesn't have a test?
> If so, I'll be fixing it when I have time -- towards the end of October.
> (The patch I sent a few days ago ought to do it, but I've not fully tested
> yet.)
> Anyway,
OK.
Jason
Hi Vlad,
Thanks for the testing and the logs. You must have good hardware, your
timings are all ~3 times faster than mine :-)
On Sat, Sep 29, 2012 at 3:01 AM, Vladimir Makarov wrote:
> --32-bit
> Reload:
> 581.85user 29.91system
On Sep 29, 2012, at 8:08 AM, Segher Boessenkool wrote:
>> The following proposed patch disables setting, saving and restoring
>> the VRSAVE register on all targets except Darwin.
>>
>> VRSAVE was removed from the AIX ABI and was suppose to have been
>> removed from the PPC SVR4 ABI. All recent
21 matches
Mail list logo
