On 05/22/2015 10:36 AM, Jason Merrill wrote:
> It also seems unnecessary to load 8 bytes on any target; we could add a
> function to optabs.c that returns the smallest mode for which there's atomic
> load support?
No, while we do use an atomic_load pattern if it exists, we assume that
all loads no
Hi,
surprisingly, for NSDMIs we don't use reshape_init and we end-up
rejecting simple testcases like the below. It seems clear to me that we
should - consistently with the comment preceding digest_init too - but
I'm not 100% sure that digest_nsdmi_init is the best place for that.
Anyway, the
On May 22, 2015 6:32:38 PM GMT+02:00, Aditya K wrote:
>Currently, when we print the passes it does not print its name. This
>becomes confusing when we want to print all the passes at once (e.g.,
>-fdump-tree-all-all=stderr &> pass.dump).
>This patch adds functionality to print the pass name. It pa
On 05/22/2015 09:39 AM, Richard Sandiford wrote:
Eric Botcazou writes:
Some pieces of code create a temporary REG or MEM and only fill it
in later when they're testing the cost of a particular rtx. This patch
makes sure that even the dummy REG or MEM is valid, rather than force
the gen_* code
On 05/22/2015 09:42 AM, Richard Sandiford wrote:
This patch adjusts the fix for PR target/65689 along the lines suggested
in https://gcc.gnu.org/ml/gcc-patches/2015-04/msg01559.html. The idea
is to reuse the existing gensupport.c routine to work out the codes
accepted by constraints.
I'd origin
On 05/22/2015 02:27 AM, Mikhail Maltsev wrote:
This patch fixes a bug introduced by refactoring. A cast from rtx_insn
to rtx_jump_insn in fix_crossing_conditional_branches was placed before
the check, and that caused ICE if the instruction is actually a call,
rather than a jump.
Bootstrapped/reg
On 05/22/2015 09:45 AM, Segher Boessenkool wrote:
On Thu, May 21, 2015 at 09:24:37AM -0600, Jeff Law wrote:
When combine needs to split a complex insn, it will canonicalize a
simple (mult X (const_int Y)) where Y is a power of 2 into the expected
(ashift X (const_int Y')) if the (mult ...) is se
These should have gone in with the first patch in the series.
Thankfully the splitters aren't terribly important anymore and thus
having them goof'd up for a couple days hasn't been a problem.
In fact, much like the hppa_legitimize_address code to handle
shift-add/scaled addressing modes, t
> Subject: Re: [PATCH] Print Pass Names
> From: richard.guent...@gmail.com
> Date: Fri, 22 May 2015 21:32:24 +0200
> To: hiradi...@msn.com; gcc-patches@gcc.gnu.org
>
> On May 22, 2015 6:32:38 PM GMT+02:00, Aditya K wrote:
>>Currently, when we print the pa
On 05/22/2015 02:38 PM, Aditya K wrote:
Subject: Re: [PATCH] Print Pass Names
From: richard.guent...@gmail.com
Date: Fri, 22 May 2015 21:32:24 +0200
To: hiradi...@msn.com; gcc-patches@gcc.gnu.org
On May 22, 2015 6:32:38 PM GMT+02:00, Aditya K wrote:
On Wed, May 20, 2015 at 3:58 PM, Yunlian Jiang wrote:
> GCC bootstraps with this patch.
Committed as follows.
Ian
include/:
2015-05-22 Yunlian Jiang
* libiberty.h (asprintf): Don't declare if HAVE_DECL_ASPRINTF is
not defined.
libiberty/:
2015-05-22 Yunlian Jiang
* configure.ac: Add
On Mon, 18 May 2015, Richard Biener wrote:
On Fri, May 15, 2015 at 7:22 PM, Marc Glisse wrote:
we already have the more complicated: x & ~(x & y) -> x & ~y (which I am
reindenting by the way) and the simpler: (~x | y) & x -> x & y, so I am
proposing this one for completeness. Regtested on ppc
On Fri, May 22, 2015 at 2:01 PM, Andrew Chambers
wrote:
>>
>> For example, I've tested on an x86, built cross compilers for ppc64 and
>> ppc64le and then can invoke the native go tool (built to run on x86) to
>> compile with either target compiler by changing my GOARCH value. I don't
>> want to h
This patch is the first in a series of patches that will eventually add support
for IEEE 128-bit floating point support to the PowerPC GCC compiler. At the
current time, we do not plan to change the default for long double. I added a
new type keyword (__float128) to get access to IEEE 128-bit flo
This is the final patch to the PA backend to cleanup its handling of
shadd insns and scaled indexed addressing modes.
First, it removes the old non-canonical shadd insns.
Second, it removes some non-canonical peephole patterns. No idea what I
was thinking when I wrote them. Given they're n
On 05/20/2015 10:41 AM, Jakub Jelinek wrote:
On Wed, May 20, 2015 at 10:36:25AM -0600, Jeff Law wrote:
These fix the remaining leaks in the threader that I'm aware of. We failed
to properly clean-up when we had to cancel certain jump threading
opportunities. So thankfully this wasn't a big le
On 05/21/2015 02:01 PM, Andreas Tobler wrote:
On 21.05.15 20:14, Andreas Tobler wrote:
On 20.05.15 22:30, Jeff Law wrote:
On 05/20/2015 11:04 AM, Andreas Tobler wrote:
Hi,
the attached patch enables some PIE tests on FreeBSD.
Ok for trunk?
Thanks,
Andreas
2015-05-20 Andreas Tobler
On 22.05.2015 12:10, Marek Polacek wrote:
> Thanks, applied. Here's the final version.
By the way, we have a feature test macro, __cpp_attributes=200809 which
can be used to determine, whether C++11 attribute syntax is supported by
the compiler.
I propose to add something similar for this extens
On Fri, May 22, 2015 at 3:11 PM, Andrew Chambers
wrote:
>
> I'm not suggesting breaking go conventions, I just think the default if no
> GOARCH is specified then it should match --target.
Sounds good to me.
> Perhaps we could check the symlink name for the target triple if no GOARCH
> is set.
W
On 05/20/2015 08:04 PM, Thomas Preud'homme wrote:
From: gcc-patches-ow...@gcc.gnu.org [mailto:gcc-patches-
ow...@gcc.gnu.org] On Behalf Of Thomas Preud'homme
From: Steven Bosscher [mailto:stevenb@gmail.com]
Sent: Tuesday, May 19, 2015 7:21 PM
Not OK.
This will break in move_invariants() whe
PING: https://gcc.gnu.org/ml/gcc-patches/2015-05/msg01511.html
This only needs approval from Fortran maintainers.
On 17 May 2015 at 20:22, Manuel López-Ibáñez wrote:
> Hi,
>
> This patch finishes the conversion of Fortran diagnostics to use the
> common diagnostics by removing all gfc_error*_1 v
The compiler currently ICEs when compiling a stdarg function with
+nofp, as reported in PR 66258.
The aarch64.md file disables FP instructions using TARGET_FLOAT, which
supports both -mgeneral-regs-only and +nofp. But there is code in
aarch64.c that checks TARGET_GENERAL_REGS_ONLY. This results
I bootstrapped this on powerpc-ibm-aix7.1.0.0 and my colleagues
bootstrapped this on powerpc64-linux and powerpc64le-linux.
It works and produces reasonable instruction sequences.
We can iterate on the syntax, but the core concept seems to work correctly.
Thanks, David
On Thu, May 21, 2015 at 5:41 AM, Tristan Gingold wrote:
>
> 2015-05-21 Tristan Gingold
>
> * pecoff.c: New file.
> * Makefile.am (FORMAT_FILES): Add pecoff.c and dependencies.
> * Makefile.in: Regenerate.
> * filetype.awk: Detect pecoff.
> * configure.ac:
101 - 124 of 124 matches
Mail list logo
