This patch makes graphite run the scev_const_prop systematically even
when using -fno-tree-scev-cprop. When scev_const_prop is not applied,
there exist close_phi nodes for the main induction variable, making it
impossible for graphite to distinguish between reductions and the IVs.
So the main IV i
On 07/23/11 22:56, Richard Henderson wrote:
> Where are we at with the review of the actual shrink-wrap patch set?
Richard S. has started reviewing it, and I'm trying to break out some
more preliminary bits and pieces.
Bernd
On 07/14/2011 04:07 PM, Richard Henderson wrote:
> This finally brings us to something that can support shrink-wrapping.
> As mentioned in the description of the last patch, this is 95% of
> what Bernd had in his last 007-dw2cfg patch, except for the remember/
> restore_state stuff. And hopefully
On 07/23/2011 05:12 AM, Paolo Bonzini wrote:
> On Sat, Jul 23, 2011 at 02:32, Richard Henderson wrote:
>> Does anyone object to me committing the patch as-is, with the
>> understanding that either Paulo or myself will, before too long,
>> clean up the DF interface as described above?
>
> I would
> 2011-07-20 Rainer Orth
>
> * init.c [sgi] (__gnat_error_handler): Update sigaction(2) citation.
> Correct argument types.
> Extract code from reason.
> (__gnat_install_handler): Assign to act.sa_sigaction.
This breaks signal handling on our IRIX 6.5 machine though.
--
> Yeah, the testcase is invalid - that we lost the volatile was a bug, but
> you really have to fix the kernel.
Sadly, that's not a helpful suggestion. How else should the kernel
force a word-sized read? I thought volatile was the way to tell gcc
"do exactly what I tell you to do" ?
And the ke
On 07/23/2011 03:20 AM, Tobias Burnus wrote:
Jerry DeLisle wrote:
I would say OK for trunk and then backport in a few weeks. If you feel like
being conservative you could use
|| (dtp->u.p.ionml->type == BT_DERIVED && !dtp->u.p.ionml->touched)) but I do
not think it is necessary.
I have commit
>> Fixed on 4.6 with r176659. How about 4.5?
>
> Thanks for the committal. Regarding 4.5, I do not have a strong feeling, but
> as the patch is simple and 4.5 is still used, you might want to commit it to
> that branch as well.
Agreed. Committed to 4.5 as r176688.
Cheers,
Janus
In C it makes no difference if a function returning a pointer returns
false or NULL, but in when bootstrapping in C++ it's another story.
Fixed thusly:
2011-07-23 Richard Earnshaw
PR target/49816
* arm.c (aapcs_vfp_allocate_return_reg): Return NULL on failure.
diff --git a/gcc
On Sat, Jul 23, 2011 at 09:35, Sebastian Pop wrote:
> With this patch we avoid instantiating ADDR_EXPR: it makes no sense
> to translate b[i] into b[{0, +, 1}_1].
>
This should have been &b[i] and &b[{0, +, 1}_1].
Ok for trunk?
Thanks,
Sebastian
> Bootstrapped and tested on amd64-linux.
>
> 20
With this patch we avoid instantiating ADDR_EXPR: it makes no sense
to translate b[i] into b[{0, +, 1}_1].
Bootstrapped and tested on amd64-linux.
2011-07-22 Sebastian Pop
PR middle-end/48805
* tree-scalar-evolution.c (instantiate_scev_r): Return
chrec_dont_know for AD
On 21/07/11 12:30, Richard Guenther wrote:
>
> This is another place where we fail to properly keep single-use
> chains by leaving around dead code. This confuses propagation
> of comparisons which is restricted for non-single-uses.
>
> Bootstrapped and tested on x86_64-unknown-linux-gnu, applie
On Sat, Jul 23, 2011 at 2:38 AM, Uros Bizjak wrote:
> On Fri, Jul 22, 2011 at 9:03 PM, H.J. Lu wrote:
>
>> This patch adds x32 LEA insn support. The main issue is
>>
>> gen_lowpart (Pmode, operands[1]);
>>
>> doesn't work on symbol. This patch avoids it.
>>
>> Also we shouldn't generate 32bit s
Hi,
I checked in the following cleanup patches.
H.J.
commit b45118372b8e816325072760d431747354bb5f46
Author: H.J. Lu
Date: Fri Jul 22 09:37:26 2011 -0700
Rename x32_general_operand to si_general_operand.
diff --git a/gcc/ChangeLog.x32 b/gcc/ChangeLog.x32
index a4b34b9..5c2c673 100644
--
On Sat, Jul 23, 2011 at 2:38 AM, Uros Bizjak wrote:
> On Fri, Jul 22, 2011 at 9:03 PM, H.J. Lu wrote:
>
>> This patch adds x32 LEA insn support. The main issue is
>>
>> gen_lowpart (Pmode, operands[1]);
>>
>> doesn't work on symbol. This patch avoids it.
>>
>> Also we shouldn't generate 32bit s
Janus Weil wrote:
Fixed on 4.6 with r176659. How about 4.5?
Thanks for the committal. Regarding 4.5, I do not have a strong feeling,
but as the patch is simple and 4.5 is still used, you might want to
commit it to that branch as well.
(Think of long-term versions of Linux distributions; tho
On Sat, Jul 23, 2011 at 02:32, Richard Henderson wrote:
> Does anyone object to me committing the patch as-is, with the
> understanding that either Paulo or myself will, before too long,
> clean up the DF interface as described above?
I would kind of object for a pass that is enabled at some -O l
This fixes the implicit declarations of exit and free in libgfortran.
Tested on ppc-linux and checked in as obvious.
Andreas.
2011-07-23 Andreas Schwab
* intrinsics/ctime.c: Include .
* intrinsics/getlog.c: Likewise.
* runtime/stop.c: Likewise.
Index: libgfortran/intr
Jerry DeLisle wrote:
I would say OK for trunk and then backport in a few weeks. If you
feel like being conservative you could use
|| (dtp->u.p.ionml->type == BT_DERIVED && !dtp->u.p.ionml->touched))
but I do not think it is necessary.
I have committed the initial patch as Rev. 176661.
* *
On Sat, Jul 23, 2011 at 8:02 AM, DJ Delorie wrote:
>
>> I built an arm kernel with it, the USB stuff works.
>
> I take that back - booted the wrong kernel. Didn't work, but I'll
> confirm that I'm building with a patched gcc (too many kernels and
> compilers to keep track of ;)
Yeah, the testcas
On Sat, Jul 23, 2011 at 1:01 AM, Sebastian Pop wrote:
> The CLAST produced by CLooG-ISL contains an assignment and GCC chokes
> on it. The exact CLAST contains an assignment followed by an if:
>
> scat_1 = max(0,ceild(T_4-7,8));
> if (scat_1 <= min(1,floord(T_4-1,8))) {
> S7(scat_1);
> }
>
> Thi
On Fri, Jul 22, 2011 at 7:28 PM, Richard Henderson wrote:
> On 07/22/2011 10:19 AM, Ulrich Weigand wrote:
>> Richard Henderson wrote:
>>> On 07/22/2011 07:42 AM, Ulrich Weigand wrote:
Well, it works for me with just adding -lm to the dg-extra-ld-options.
This still folds cabs to sqrt in
On Fri, Jul 22, 2011 at 9:03 PM, H.J. Lu wrote:
> This patch adds x32 LEA insn support. The main issue is
>
> gen_lowpart (Pmode, operands[1]);
>
> doesn't work on symbol. This patch avoids it.
>
> Also we shouldn't generate 32bit store with x32 PIC source.
>
> Any comments?
Can you please pos
As mentioned in the PR, there's a wrong reference to a non existing file
in Makefile.in, fixed in this patch.
Committed on trunk.
2011-07-23 Arnaud Charlet
PR ada/49819
* gcc-interface/Makefile.in (powerpc-linux): Remove reference to
g-trasym-dwarf.adb.
--
Index: gcc-
On Fri, 22 Jul 2011, Steve Ellcey wrote:
> $ /proj/opensrc/nightly/gcc-ia64-hp-hpux11.23-trunk/bin/gcc
> /usr/lib/hpux32/dld.so: Unable to find library 'libgcc_s.so.0'.
>
> gcc (the driver) is now linked with C++ instead of C and
> that means it wants to use the shared libgcc instead of the
Hi Tobias,
Just a quick update so you know what I'm doing.
On 07/22/2011 10:09 PM, Tobias Burnus wrote:
However, the bug is not fully fixed. If you try the testcase in the
bugreport, you will see that it still fails - the stat= is correctly
set, but the array shape is wrong.
Working on it...
26 matches
Mail list logo
