This makes it so that the size of types (Esize) is retrieved early and properly
capped for elementary types, in order to avoid obscure failures later on for
huge sizes. Nothing is changed for composite types.
Tested on i586-suse-linux, applied on the mainline.
2012-06-11 Eric Botcazou
The executable cannot be linked because of a duplicated symbol in the object
files, which comes from a conditional expression used to initialize a constant
that is the upper bound of an index type in an array.
Tested on i586-suse-linux, applied on the mainline.
2012-06-11 Eric Botcazou
Ping?
On 2012/6/1 06:24 PM, Chung-Lin Tang wrote:
> On 12/5/23 1:46 AM, Richard Henderson wrote:
>> On 05/18/12 03:48, Chung-Lin Tang wrote:
>>> @@ -2401,6 +2401,7 @@ scan_trace (dw_trace_info *trace)
>>> {
>>> /* Propagate across fallthru edges. */
>>> dwarf2out_flush_queued_reg_
On Wed, Jun 6, 2012 at 5:16 PM, Steven Bosscher wrote:
> Hello,
>
> tree-phinodes.c includes rtl.h for ... ceil_log2 (?!).
>
> Moving ceil_log2 to hwint.c/hwint.h to join its sibling floor_log2
> breaks this dependency.
>
> Bootstrapped on x86_64-unknown-linux-gnu. OK for trunk?
Ok.
Thanks,
Rich
On Wed, Jun 6, 2012 at 5:16 PM, Steven Bosscher wrote:
> Hello,
>
> cfgloop.h only depends on rtl.h for rtx_code. The patch introduces a
> new enum and a simple function to convert from one enum to another in
> a few non-performance critical places.
>
> The upside is that this makes most of the tr
On Thu, 7 Jun 2012, Jason Merrill wrote:
> The problem here was that when we instantiate G::bar, we want to instantiate
> F, and end up improperly pushing the declaration inside G, which leads to the
> ICE. Adjusting the call to pushclass is enough to fix this, which is what I
> propose doing for
This aligns the processing of constants used by reference in Identifier_to_gnu
with that of regular constants; in particular we stop returning the underlying
constant if this is the address of a constant and an lvalue is required.
Tested on i586-suse-linux, applied on the mainline.
2012-06-11
Hi,
This patch fixes several layout, formatting and wording issues in the
AArch64 target-specific documentation.
Thanks
Sofiane
-
2012-06-11 Sofiane Naci
[AArch64] Fix documentation layout.
* doc/invoke.texi: Fix white spaces after dots.
Change aarch64*be
On Thu, Jun 7, 2012 at 10:08 AM, Ian Lance Taylor wrote:
> This patch to the Go compiler fixes the case of taking the address of a
> field of a local variable. Previously taking the address of the field
> was not treated as taking the address of the variable. But, of course,
> it is. This patch
On Thu, Jun 7, 2012 at 10:18 PM, Steven Bosscher wrote:
> Hello,
>
> The attached patch removes cfglayout.h. It was included in a lot of
> places that didn't need it, and it didn't define much anyway.
>
> While checking the users of cfglayout.h, I noticed that a lot of code
> in cfglayout.c can be
On Fri, Jun 8, 2012 at 12:23 AM, Steven Bosscher wrote:
> Hello,
>
> Another patch to move around some code to break up a strange
> dependency. expr.c isn't supposed to emit assembly so it shouldn't
> need output.h, but it needs split_double from final.c. This
> split_double has nothing to do with
Committed to ARM/embedded-4_6-branch
2012-06-11 Joey Ye
Backport r184089,184180 from mainline
2012-02-10 Jan Hubicka
PR middle-end/48600
* predict.c (predict_paths_for_bb): Prevent looping.
(predict_paths_leading_to_edge, predict_paths_leading_to): U
On 11/06/12 09:54, Sofiane Naci wrote:
(-march and -mcpu feature modifiers): New subsubsection.
Copy and paste subsub?
/Marcus
The attached packages declare two big objects of the same size, but for the
first package the object is allocated statically whereas, for the second one,
the object is allocated dynamically because of the size clause.
The discrepancy stems from a bit vs byte confusion in a specific place in gigi
On Sun, Jun 10, 2012 at 5:58 PM, William J. Schmidt
wrote:
> The fix for PR53331 caused a degradation to 187.facerec on
> powerpc64-unknown-linux-gnu. The following simple patch reverses the
> degradation without otherwise affecting SPEC cpu2000 or cpu2006.
> Bootstrapped and regtested on that pl
Hi Matt,
This is just to let you know that I am currently reviewing these
patches. I do have a problem however. With the patches applied I am
seeing very bad results from the gcc testsuite when run with
-mcpu=iwmmxt. (Bad as in the testsuite takes days to run and most tests
fail). I am
Thank you for the review, with this patch I get some ICEs during the
regstest with:
gfortran.dg/coarray/poly_run_3.f90
gfortran.dg/elemental_optional_args_5.f03
gfortran.dg/select_type_26.f03
gfortran.dg/select_type_27.f03
gfortran.dg/class_48.f90
gfortran.dg/class_allocate_10.f03
gfortran.dg/clas
On 06/09/2012 01:58 AM, Mike Stump wrote:
On Jun 8, 2012, at 4:11 PM, Mike Stump wrote:
On Apr 17, 2012, at 2:08 PM, Bernd Schmidt wrote:
This patch enables GET_MODE_WIDER_MODE for MODE_PARTIAL_INT (by setting
the wider mode to the one the partial mode is based on), which is useful
for the port
On 06/11/2012 11:24 AM, Alessandro Fanfarillo wrote:
gfortran.dg/coarray/poly_run_3.f90
That one fails because I for forgot that se.expr in gfc_trans_deallocate
contains the descriptor and not the pointer to the data. That's fixed by:
tmp = se.expr;
if (GFC_DESCRIPTOR_TYP
Hi!
http://gcc.gnu.org/ml/gcc-patches/2012-06/msg00105.html
- add warning for complit with incomplte type
Jakub
On Mon, Jun 4, 2012 at 3:45 PM, William J. Schmidt
wrote:
> Hi Richard,
>
> Here's a revision of the hoist-adjacent-loads patch. I'm sorry for the
> delay since the last revision, but my performance testing has been
> blocked waiting for a fix to PR53487. I ended up applying a test
> version of
On Fri, 8 Jun 2012, William J. Schmidt wrote:
> This patch adds a heuristic to the vectorizer when estimating the
> minimum profitable number of iterations. The heuristic is
> target-dependent, and is currently disabled for all targets except
> PowerPC. However, the intent is to make it general
Hi everyone,
Ping RE: http://gcc.gnu.org/ml/gcc-patches/2012-06/msg00086.html
Thanks,
Robert Mason
On 06/01/2012 02:43 PM, rbmj wrote:
On 06/01/2012 02:40 PM, rbmj wrote:
Hi everyone,
These fixes are to allow building on vxWorks. Currently there are
two issues:
1. VxWorks does not have
On 05/22/2012 04:37 PM, rbmj wrote:
This patch adds an AC_ARG_ENABLE option to build/not build
libstdc++-v3. I wrote the patch in order to allow the user to
override the automatic disable for libstdc++-v3 for certain targets.
Ping^2 on http://gcc.gnu.org/ml/gcc-patches/2012-05/msg01525.html
R
Il 11/06/2012 13:56, rbmj ha scritto:
>>> 1. VxWorks does not have a variadic open - it must receive three
>>> arguments. gcc/gcov.c however opens a file for reading and does not
>>> pass in a mode argument, which causes an error on vxWorks. This just
>>> adds a platform-based ifdef around this.
Il 11/06/2012 13:59, rbmj ha scritto:
> On 05/22/2012 04:37 PM, rbmj wrote:
>> This patch adds an AC_ARG_ENABLE option to build/not build
>> libstdc++-v3. I wrote the patch in order to allow the user to
>> override the automatic disable for libstdc++-v3 for certain targets.
> Ping^2 on http://gcc.
When I changed empty arrays domain to use a signed sizetype [0, -1]
domain mangling of a empty-array-type broke because mangling adds
an unsigned one to the signed -1 which causes an ICE (I chose to
do that instead of shifting the range to [1, 0] because much more
code relies on a zero lower bound
On Mon, 2012-06-11 at 13:28 +0200, Richard Guenther wrote:
> On Mon, Jun 4, 2012 at 3:45 PM, William J. Schmidt
> wrote:
> > Hi Richard,
> >
> > Here's a revision of the hoist-adjacent-loads patch. I'm sorry for the
> > delay since the last revision, but my performance testing has been
> > blocke
On Mon, 11 Jun 2012, William J. Schmidt wrote:
> On Mon, 2012-06-11 at 13:28 +0200, Richard Guenther wrote:
> > On Mon, Jun 4, 2012 at 3:45 PM, William J. Schmidt
> > wrote:
> > > Hi Richard,
> > >
> > > Here's a revision of the hoist-adjacent-loads patch. I'm sorry for the
> > > delay since the
The ARM ABI states that vectors larger than 64 bits in size still have
64-bit alignment; never-the-less, the HW supports alignment hints of up
to 128-bits in some cases and will trap in a vector has an alignment
that less than the hint. GCC currently hard-codes larger vectors to be
aligned by the
OK.
Jason
On Mon, 2012-06-11 at 11:15 +0200, Richard Guenther wrote:
> On Sun, Jun 10, 2012 at 5:58 PM, William J. Schmidt
> wrote:
> > The fix for PR53331 caused a degradation to 187.facerec on
> > powerpc64-unknown-linux-gnu. The following simple patch reverses the
> > degradation without otherwise aff
On Mon, 11 Jun 2012, William J. Schmidt wrote:
>
>
> On Mon, 2012-06-11 at 11:15 +0200, Richard Guenther wrote:
> > On Sun, Jun 10, 2012 at 5:58 PM, William J. Schmidt
> > wrote:
> > > The fix for PR53331 caused a degradation to 187.facerec on
> > > powerpc64-unknown-linux-gnu. The following s
This fixes PR53470 - I for quite some time wanted to remove that
conditional clearing of TYPE_CONTEXT from free_lang_data_in_type
but failed to do so because we regress. I've debugged it down to
the C frontend having sometimes BLOCK as TYPE_CONTEXT for a type.
So, simply replace such BLOCK with t
On Mon, Jun 11, 2012 at 3:16 PM, Richard Earnshaw wrote:
> The ARM ABI states that vectors larger than 64 bits in size still have
> 64-bit alignment; never-the-less, the HW supports alignment hints of up
> to 128-bits in some cases and will trap in a vector has an alignment
> that less than the hi
Hi,
I haven't yet checked in my patch from last week about not creating
alias-set conflicts for stack slot sharing mostly due to the caveat I
mentioned. Namely that the RTL disambiguator still uses path-based means
(which belong to type-based aliasing) even in the write_dependence tests.
Tha
On Mon, Jun 11, 2012 at 4:19 PM, Michael Matz wrote:
> Hi,
>
> I haven't yet checked in my patch from last week about not creating
> alias-set conflicts for stack slot sharing mostly due to the caveat I
> mentioned. Namely that the RTL disambiguator still uses path-based means
> (which belong to
This makes ao_ref_from_mem less conservative if either MEM_OFFSET or
MEM_SIZE is not set. From other alias.c code and
set_mem_attributes_minus_bitpos one has to conclude that MEM_EXPR
is always conservatively correct (it only can cover a larger memory
area) and the MEM_OFFSET / MEM_SIZE pair can
On Mon, 2012-06-11 at 13:40 +0200, Richard Guenther wrote:
> On Fri, 8 Jun 2012, William J. Schmidt wrote:
>
> > This patch adds a heuristic to the vectorizer when estimating the
> > minimum profitable number of iterations. The heuristic is
> > target-dependent, and is currently disabled for all
On 11/06/12 15:17, Richard Guenther wrote:
> On Mon, Jun 11, 2012 at 3:16 PM, Richard Earnshaw wrote:
>> The ARM ABI states that vectors larger than 64 bits in size still have
>> 64-bit alignment; never-the-less, the HW supports alignment hints of up
>> to 128-bits in some cases and will trap in a
On Mon, Jun 11, 2012 at 7:40 AM, Richard Guenther wrote:
> Hmm. I don't like this patch or its general idea too much. Instead
> I'd like us to move more of the cost model detail to the target, giving
> it a chance to look at the whole loop before deciding on a cost. ISTR
> posting the overall
On Mon, 11 Jun 2012, William J. Schmidt wrote:
> On Mon, 2012-06-11 at 13:40 +0200, Richard Guenther wrote:
> > On Fri, 8 Jun 2012, William J. Schmidt wrote:
> >
> > > This patch adds a heuristic to the vectorizer when estimating the
> > > minimum profitable number of iterations. The heuristic i
On Mon, 11 Jun 2012, David Edelsohn wrote:
> On Mon, Jun 11, 2012 at 7:40 AM, Richard Guenther wrote:
>
> > Hmm. I don't like this patch or its general idea too much. Instead
> > I'd like us to move more of the cost model detail to the target, giving
> > it a chance to look at the whole loop b
This patch fixes the behaviour of legitimate address checking for TImode
and TFmode values.
The current implementation does not correctly distinguish between the
address modes available for ldp/stp X,X versus ldr/str Q.
/Marcus
ChangeLog:
* config/aarch64/aarch64.c
(aarch64_
Hello,
We used to have in gcc-interface/Makefile.in:
GNATLIBCFLAGS_FOR_C = -W -Wall $(GNATLIBCFLAGS) $(TARGET_LIBGCC2_CFLAGS) ...
As part of the libgcc move to toplevel,
http://gcc.gnu.org/ml/gcc-patches/2011-08/msg00917
On Mon, Jun 11, 2012 at 10:48 AM, Richard Guenther wrote:
> On Mon, 11 Jun 2012, David Edelsohn wrote:
>
>> On Mon, Jun 11, 2012 at 7:40 AM, Richard Guenther wrote:
>>
>> > Hmm. I don't like this patch or its general idea too much. Instead
>> > I'd like us to move more of the cost model detail
On Mon, Jun 11, 2012 at 4:38 PM, Richard Earnshaw wrote:
> On 11/06/12 15:17, Richard Guenther wrote:
>> On Mon, Jun 11, 2012 at 3:16 PM, Richard Earnshaw wrote:
>>> The ARM ABI states that vectors larger than 64 bits in size still have
>>> 64-bit alignment; never-the-less, the HW supports alignm
Hi,
This patch removes __float128 support in the AArch64 port.
Thanks
Sofiane
-
gcc/ChangeLog:
2012-06-11 Sofiane Naci
[AArch64] Remove __float128 support.
* config/aarch64/aarch64.c
(aarch64_mangle_type): Remove function.
(aarch64_init_builtins): Remove
On Mon, 11 Jun 2012, David Edelsohn wrote:
> On Mon, Jun 11, 2012 at 10:48 AM, Richard Guenther wrote:
> > On Mon, 11 Jun 2012, David Edelsohn wrote:
> >
> >> On Mon, Jun 11, 2012 at 7:40 AM, Richard Guenther
> >> wrote:
> >>
> >> > Hmm. I don't like this patch or its general idea too much. I
On Mon, 11 Jun 2012, Richard Guenther wrote:
> On Mon, 11 Jun 2012, William J. Schmidt wrote:
>
> > On Mon, 2012-06-11 at 13:40 +0200, Richard Guenther wrote:
> > > On Fri, 8 Jun 2012, William J. Schmidt wrote:
> > >
> > > > This patch adds a heuristic to the vectorizer when estimating the
> > >
On Mon, 2012-06-11 at 16:10 +0200, Richard Guenther wrote:
> On Mon, 11 Jun 2012, William J. Schmidt wrote:
>
> >
> >
> > On Mon, 2012-06-11 at 11:15 +0200, Richard Guenther wrote:
> > > On Sun, Jun 10, 2012 at 5:58 PM, William J. Schmidt
> > > wrote:
> > > > The fix for PR53331 caused a degr
On Mon, Jun 11, 2012 at 10:55 AM, Richard Guenther wrote:
> Well, they are at least magic numbers and heuristics that apply
> generally and not only to the single issue in sphinx. And in
> fact how it works for sphinx _is_ magic.
>
>> Second, I suggest that you need to rephrase "I can make you"
On Mon, 2012-06-11 at 16:58 +0200, Richard Guenther wrote:
> On Mon, 11 Jun 2012, Richard Guenther wrote:
>
> > On Mon, 11 Jun 2012, William J. Schmidt wrote:
> >
> > > On Mon, 2012-06-11 at 13:40 +0200, Richard Guenther wrote:
> > > > On Fri, 8 Jun 2012, William J. Schmidt wrote:
> > > >
> > >
On 11/06/12 15:53, Richard Guenther wrote:
> On Mon, Jun 11, 2012 at 4:38 PM, Richard Earnshaw wrote:
>> On 11/06/12 15:17, Richard Guenther wrote:
>>> On Mon, Jun 11, 2012 at 3:16 PM, Richard Earnshaw wrote:
The ARM ABI states that vectors larger than 64 bits in size still have
64-bit
On Mon, 2012-06-11 at 11:09 -0400, David Edelsohn wrote:
> On Mon, Jun 11, 2012 at 10:55 AM, Richard Guenther wrote:
>
> > Well, they are at least magic numbers and heuristics that apply
> > generally and not only to the single issue in sphinx. And in
> > fact how it works for sphinx _is_ magi
On 06/01/2012 02:09 PM, Florian Weimer wrote:
On 06/01/2012 11:00 AM, Florian Weimer wrote:
I'll try to warn about this case and make the transformation to the
proper operator new[] call.
Here's the version. I've added a warning for the ill-formed code.
The only remaining glitch is in g++.dg
ping?
On Wed, Jun 6, 2012 at 2:04 PM, Roland McGrath wrote:
> cf this change:
>
> 2010-11-19 Jakub Jelinek
>
> PR target/45870
> * dwarf2out.c (const_ok_for_output_1): Don't complain about
> non-delegitimized TLS UNSPECs.
>
> This case hit me
Richard Sandiford wrote:
As discussed in the context of the AARCH64 submission, this patch
rewrites the iterator handling in read-rtl.c so that we record
iterator positions using an on-the-side VEC rather than placeholder
modes and codes. We then substitute in-place for each sequence of
iterator
On Mon, 2012-06-11 at 14:59 +0200, Richard Guenther wrote:
> On Mon, 11 Jun 2012, William J. Schmidt wrote:
>
> > On Mon, 2012-06-11 at 13:28 +0200, Richard Guenther wrote:
> > > On Mon, Jun 4, 2012 at 3:45 PM, William J. Schmidt
> > > wrote:
> > > > Hi Richard,
> > > >
> > > > Here's a revision
On Mon, 2012-06-11 at 12:11 -0500, William J. Schmidt wrote:
> I found this parameter that seems to correspond to well-predicted
> conditional jumps:
>
> /* When branch is predicted to be taken with probability lower than this
>threshold (in percent), then it is considered well predictable. *
On 2012-06-11 01:34, Chung-Lin Tang wrote:
> Ping?
>
> On 2012/6/1 06:24 PM, Chung-Lin Tang wrote:
>> On 12/5/23 1:46 AM, Richard Henderson wrote:
>>> On 05/18/12 03:48, Chung-Lin Tang wrote:
@@ -2401,6 +2401,7 @@ scan_trace (dw_trace_info *trace)
{
/* Propagate across fallt
On Mon, Jun 11, 2012 at 5:38 PM, William J. Schmidt
wrote:
>
>
> On Mon, 2012-06-11 at 11:09 -0400, David Edelsohn wrote:
>> On Mon, Jun 11, 2012 at 10:55 AM, Richard Guenther wrote:
>>
>> > Well, they are at least magic numbers and heuristics that apply
>> > generally and not only to the single
On Mon, Jun 11, 2012 at 5:25 PM, Richard Earnshaw wrote:
> On 11/06/12 15:53, Richard Guenther wrote:
>> On Mon, Jun 11, 2012 at 4:38 PM, Richard Earnshaw wrote:
>>> On 11/06/12 15:17, Richard Guenther wrote:
On Mon, Jun 11, 2012 at 3:16 PM, Richard Earnshaw wrote:
> The ARM ABI states
Hello Everyone,
This patch is for the Cilkplus branch affecting both C and C++ compilers.
This patch will make the cilkrts_stack_frame struct independent of the runtime.
Thanking You,
Yours Sincerely,
Balaji V. Iyer.
Index: gcc/cilk.c
=
On Mon, Jun 11, 2012 at 5:01 PM, William J. Schmidt
wrote:
>
>
> On Mon, 2012-06-11 at 16:10 +0200, Richard Guenther wrote:
>> On Mon, 11 Jun 2012, William J. Schmidt wrote:
>>
>> >
>> >
>> > On Mon, 2012-06-11 at 11:15 +0200, Richard Guenther wrote:
>> > > On Sun, Jun 10, 2012 at 5:58 PM, William
Hello Everyone,
This patch is for the Cilkplus branch affecting both C and C++ compilers.
This patch will make the cilkrts_stack_frame structure independent of the
runtime.
Thanking You,
Yours Sincerely,
Balaji V. Iyer.
Index: gcc/cilk.c
The intent of the phases was to have a high-level but mutually exclusive
accounting of compile time. We want to track compile time in a way that
tells us which conceptual phases are taking the most time. That intent
is not currently satisfied. This patch restores that intent.
Add code to verify
The tests in gcc.dg/torture/stackalign use two sets of torture options:
the usual optimization sets used as default for torture tests, and up to
four sets of options that are specific to stack alignment. The stack
alignment options are passed via an option that is used by the dg-test
option dg_ext
Thanks for the update.
Tejas Belagod writes:
> +/* Implementations of the iterator_group callbacks for ints. */
> +
> +/* Since GCC does not construct a table of valid constants,
> + we have to accept any int as valid. No cross-checking can
> + be done. */
> +
> +static int
> +find_int (co
On 06/10/2012 02:38 PM, Bruce Korb wrote:
On Sun, Jun 10, 2012 at 10:57 AM, Nathan Sidwell wrote:
On 06/06/12 17:33, rbmj wrote:
Hi everyone,
This patch series is the result of this [1] thread about fixincludes on
VxWorks.
It resolves bugs 53457 and 53378, and a few other issues that previous
On 06/11/2012 08:02 AM, Paolo Bonzini wrote:
Il 11/06/2012 13:59, rbmj ha scritto:
On 05/22/2012 04:37 PM, rbmj wrote:
This patch adds an AC_ARG_ENABLE option to build/not build
libstdc++-v3. I wrote the patch in order to allow the user to
override the automatic disable for libstdc++-v3 for ce
On Mon, Jun 11, 2012 at 12:00 PM, Janis Johnson
wrote:
> The tests in gcc.dg/torture/stackalign use two sets of torture options:
> the usual optimization sets used as default for torture tests, and up to
> four sets of options that are specific to stack alignment. The stack
> alignment options ar
On 06/11/2012 12:15 PM, H.J. Lu wrote:
> On Mon, Jun 11, 2012 at 12:00 PM, Janis Johnson
> wrote:
>> The tests in gcc.dg/torture/stackalign use two sets of torture options:
>> the usual optimization sets used as default for torture tests, and up to
>> four sets of options that are specific to stac
This patch is to fix MIPS failures that occcur when configuring GCC with the
-with-synci option. In this case GCC will generate the synci instruction
by default if on a MIPS architecture that supports it . When on an
architecture that does not support synci, GCC will not generate it by default
bu
On Mon, Jun 11, 2012 at 12:21 PM, Janis Johnson
wrote:
> On 06/11/2012 12:15 PM, H.J. Lu wrote:
>> On Mon, Jun 11, 2012 at 12:00 PM, Janis Johnson
>> wrote:
>>> The tests in gcc.dg/torture/stackalign use two sets of torture options:
>>> the usual optimization sets used as default for torture test
I've just committed the attached patch as rev 188397 which fixes some
formatting in the gcc/ChangeLog.
Cheers,
Oleg
Index: gcc/ChangeLog
===
--- gcc/ChangeLog (revision 188396)
+++ gcc/ChangeLog (working copy)
@@ -452,13 +452,15 @@
On 06/11/2012 12:35 PM, H.J. Lu wrote:
> On Mon, Jun 11, 2012 at 12:21 PM, Janis Johnson
> wrote:
>> On 06/11/2012 12:15 PM, H.J. Lu wrote:
>>> On Mon, Jun 11, 2012 at 12:00 PM, Janis Johnson
>>> wrote:
The tests in gcc.dg/torture/stackalign use two sets of torture options:
the usual op
On Mon, Jun 11, 2012 at 12:41 PM, Janis Johnson
wrote:
> On 06/11/2012 12:35 PM, H.J. Lu wrote:
>> On Mon, Jun 11, 2012 at 12:21 PM, Janis Johnson
>> wrote:
>>> On 06/11/2012 12:15 PM, H.J. Lu wrote:
On Mon, Jun 11, 2012 at 12:00 PM, Janis Johnson
wrote:
> The tests in gcc.dg/tortu
On 06/11/2012 12:48 PM, H.J. Lu wrote:
> On Mon, Jun 11, 2012 at 12:41 PM, Janis Johnson
> wrote:
>> On 06/11/2012 12:35 PM, H.J. Lu wrote:
>>> On Mon, Jun 11, 2012 at 12:21 PM, Janis Johnson
>>> wrote:
On 06/11/2012 12:15 PM, H.J. Lu wrote:
> On Mon, Jun 11, 2012 at 12:00 PM, Janis John
On Mon, Jun 11, 2012 at 12:57 PM, Janis Johnson
wrote:
> On 06/11/2012 12:48 PM, H.J. Lu wrote:
>> On Mon, Jun 11, 2012 at 12:41 PM, Janis Johnson
>> wrote:
>>> On 06/11/2012 12:35 PM, H.J. Lu wrote:
On Mon, Jun 11, 2012 at 12:21 PM, Janis Johnson
wrote:
> On 06/11/2012 12:15 PM, H
On 2012-06-11 08:46, Ramana Radhakrishnan wrote:
> I will also point out that this implementation actually makes
> things worse at -O0 given we don't constant propagate the mask into
> the VEC_PERM_EXPR which is actually a regression compared to the
> current state of the art (yes at O0 but I'm sur
OK, once more with feeling... :)
This patch differs from the previous one in two respects: It disables
the optimization when either the then or else edge is well-predicted;
and it now uses the existing l1-cache-line-size parameter instead of a
new one (with updated commentary).
Bootstraps and te
On 2012-06-06 14:04, Roland McGrath wrote:
> 2012-06-06 Roland McGrath
>
> * dwarf2out.c (const_ok_for_output_1): Detect a TLS UNSPEC using
> SYMBOL_REF_TLS_MODEL rather than DECL_THREAD_LOCAL_P, in case it's
> not a VAR_DECL. Also don't limit it to UNSPECs with exactly one
>
On 2012-06-08 03:17, Andreas Schwab wrote:
> This is wrong. check_effective_target_powerpc64 tests whether the
> target can execute 64-bit insns, but that doesn't say anything about
> whether the compiler can make use of them (-mno-powerpc64 can't, which
> is the default for -m32 on linux).
Feh.
On Mon, Jun 11, 2012 at 1:42 PM, Richard Henderson wrote:
> Applied.
Thanks!
On 11 June 2012 21:10, Richard Henderson wrote:
> On 2012-06-11 08:46, Ramana Radhakrishnan wrote:
>> I will also point out that this implementation actually makes
>> things worse at -O0 given we don't constant propagate the mask into
>> the VEC_PERM_EXPR which is actually a regression compared to
On 2012-06-08 16:23, Roland McGrath wrote:
> 2012-06-08 Roland McGrath
>
> * gthr-posix.h [neither FreeBSD nor Solaris] (__gthread_active_p):
> If __GLIBC__ is defined, refer to __pthread_key_create instead of
> pthread_cancel.
Applied.
r~
On 2012-06-11 14:12, Ramana Radhakrishnan wrote:
> I could generate the masks with
> appropriate casts to unsigned variants instead.That doesn't seem to
> make a difference either.
No, that shouldn't make any difference.
r~
On Mon, Jun 11, 2012 at 2:16 PM, Richard Henderson wrote:
> Applied.
Thanks!
On 2012-05-15 17:24, Hans-Peter Nilsson wrote:
> gcc:
> * doc/tm.texi.in (Addressing Modes) :
> Mention that this hook needs to be defined for native TLS.
> * doc/tm.texi: Regenerate.
Ok.
r~
On 2012-06-06 23:53, Jakub Jelinek wrote:
> 2012-06-07 Jakub Jelinek
>
> PR rtl-optimization/53589
> * cfgrtl.c (force_nonfallthru_and_redirect): Do asm_goto_edge
> discovery even when e->dest != target. If any LABEL_REF points
> to e->dest label, redirect it to target'
Hello,
I'd like to add the attached SH test cases to verify the auto-inc-dec
pass operation on the SH target.
Checked by invoking the individual test cases manually with make
check-gcc on sh-sim. Some of the tests pass, some of them don't because
of the auto-inc-dec issues mentioned in the PR.
C
> From: Richard Earnshaw
> Date: Mon, 11 Jun 2012 15:16:50 +0200
> The ARM ABI states that vectors larger than 64 bits in size still have
> 64-bit alignment; never-the-less, the HW supports alignment hints of up
> to 128-bits in some cases and will trap in a vector has an alignment
> that less th
On 2012-06-10 02:18, Jay K wrote:
>
> gcc-4.7.0/gcc/config/alpha/alpha.c
>
>
> word1 = expand_and (DImode, word1, GEN_INT (0x0ffffff0), NULL);
>
>
> That "big" constant isn't portable since it doesn't fit in 32bits.
>
>
> 1) append LL
> or 2) break it up into an expression, lik
Bah. Wrong patch.
r~
* config/alpha/alpha.c (alpha_trampoline_init): Split large constants.
diff --git a/gcc/config/alpha/alpha.c b/gcc/config/alpha/alpha.c
index 6d15bf7..3dda9fb 100644
--- a/gcc/config/alpha/alpha.c
+++ b/gcc/config/alpha/alpha.c
@@ -5451,6 +5451,8 @@ alpha_trampoline
On May 4, 2012, at 8:01 AM, Ian Lance Taylor wrote:
> This patch updates libgo to the Go 1.0.1 release. This is a relatively
> small collection of bug fixes, with no API changes.
So, it is an important part of the test case for
libgo/go/mime/multipart/testdata/nested-mime to have ^Ms in the file
Thank you. I like it. May I have another?
book2:gcc jay$ grep -i epoch vms*
vmsdbgout.c:/* Difference in seconds between the VMS Epoch and the Unix Epoch */
vmsdbgout.c:static const long long vms_epoch_offset = 3506716800ll;
vmsdbgout.c:#define VMS_EPOCH_OFFSET 350671680
vmsdbgout.c:
On Jun 11, 2012, at 4:06 PM, Richard Henderson wrote:
> Bah. Wrong patch.
>
>
> r~
>
Hum, I'm trying to see how this patch works... I feel like there is something
I'm missing, like a shift?
Oops, agreed, shift missing. Also, I've been bitten, unable to find stuff
(grep) due to token pasting, so I am a little slower to use it.
But I understand it is useful in general for reuse.
- Jay
> Subject: Re: constant that doesn't fit in 32bits in al
Oleg Endo wrote:
> Some of the tests pass, some of them don't because
> of the auto-inc-dec issues mentioned in the PR.
I thought that the tests which are known to fail are marked
with XFAIL, though I'm not sure the usual way for the failing
test cases.
Regards,
kaz
1 - 100 of 117 matches
Mail list logo
