Hi Richard/Joseph,
I noticed, this test case crashes on arm-eabi already witout the patch.
extern void abort (void);
#define test_type unsigned short
#define MAGIC (unsigned short)0x102u
typedef struct s{
unsigned char Prefix[1];
test_type Type;
}__attribute((__packed__,__aligned__(4))) ss;
Hello!
> Currently GCC could not vectorize abs() operation for integers on x86
> with only SSE2 support. For int type, the reason is that the expand on
> abs() is not defined for vector type. This patch defines such an
> expand so that abs(int) will be vectorized with only SSE2.
+(define_expand "
On 10/22/2013 09:00 PM, Jeff Law wrote:
So I was poking at this a bit. It's trival to use infer_nonnull_range
and to teach infer_nonnull_range to use the returns_nonnull attribute to
pick up that return x in an appropriately decorated function implies
that x is non-null.
We'll need a better pl
On Tue, 22 Oct 2013, Tom de Vries wrote:
> Richard,
>
> This patch adds a missing check for gimple_vdef in stmt_local_def for the
> tail-merge pass.
>
> Bootstrapped and reg-tested on x86_64.
>
> OK for trunk, gcc-4_8-branch?
Ok.
Thanks,
Richard.
On Tue, 22 Oct 2013, Jeff Law wrote:
> On 10/22/13 03:58, Tom de Vries wrote:
> > Richard,
> >
> > This patch adds a missing check for gimple_vdef in stmt_local_def for the
> > tail-merge pass.
> >
> > Bootstrapped and reg-tested on x86_64.
> >
> > OK for trunk, gcc-4_8-branch?
> >
> > Thanks,
Jan Hubicka writes:
> +static void
> +expand_set_or_movmem_prologue_epilogue_by_misaligned_moves (rtx destmem, rtx
> srcmem,
> + rtx *destptr, rtx
> *srcptr,
> + enum machine_mode
>
> Jan Hubicka writes:
>
> > +static void
> > +expand_set_or_movmem_prologue_epilogue_by_misaligned_moves (rtx destmem,
> > rtx srcmem,
> > + rtx *destptr, rtx
> > *srcptr,
> > + enum mach
On Tue, 22 Oct 2013, Jakub Jelinek wrote:
> Hi!
>
> If VRP tells us that oprnd is always >= 0 or always < 0, we can generate
> better code for the divmode vectorization.
>
> Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?
Testcase...?
Ok with adding one (I suggest a x86 sp
On 10/23/2013 04:47 AM, Jason Merrill wrote:
I don't think this patch should be blocked on those bugs.
Sure. My point essentially was that as a *GNU* thing we could still
pretend to do whatever we wanted is corner cases, etc. As a Standard
feature, those are really full blown bugs. Anyway, what
On Mon, Oct 21, 2013 at 5:46 PM, Ian Lance Taylor wrote:
> According to c.opt, the -fdefault-inline patch is ignored, and only
> exists for backward compatibility. The only use of flag_default_inline
> was removed by Honza here:
> http://gcc.gnu.org/ml/gcc-patches/2008-07/msg02104.html
>
> This p
On Mon, Oct 21, 2013 at 9:39 PM, Mike Stump wrote:
> Concerning:
>
> 2008-11-20 Richard Guenther
>
> PR tree-optimization/37868
> * gcc.dg/torture/pr37868.c: New testcase.
> * gcc.c-torture/execute/pr38048-1.c: Likewise.
> * gcc.c-torture/execute/pr38048-2.c: Lik
On Tue, Oct 22, 2013 at 12:01 AM, Yufeng Zhang wrote:
> Hi,
>
> This patch changes the widening_mul pass to fuse the widening multiply with
> accumulate only when the multiply has single use. The widening_mul pass
> currently does the conversion regardless of the number of the uses, which
> can c
On Tue, Oct 22, 2013 at 4:27 PM, Jakub Jelinek wrote:
> On Fri, Oct 18, 2013 at 12:06:35PM +0200, Richard Biener wrote:
>> You can't move type conversion "out of the way" in most cases as
>> GIMPLE is stronly typed
>> and data sources and sinks can obviously not be "promoted" (nor can
>> function
On Tue, Oct 22, 2013 at 9:26 PM, Jakub Jelinek wrote:
> Hi!
>
> I've noticed that this testcase fails on i686-linux. The problem is
> that the vect.exp infrastructure is terrible and dg-options overrides
> all the required flags. So, vect/ testcases either don't have to have
> explicit options a
On Wed, Oct 23, 2013 at 1:42 AM, Igor Shevlyakov
wrote:
> I'm working on a chip with some unusual addressing modes.
> it does have [base_rage+msize*index_reg] and one can't omit base_reg like
> on x86.
> But when ivopts tries to calculate the costs of different addressing modes
> it only checks [m
On Wed, Oct 23, 2013 at 5:21 AM, Andrew MacLeod wrote:
> When moving all the prototypes out of tree-flow.h and into individual
> files, the "kitchen sink" #include list temporarily propagated it way into
> the new tree-ssa.h file. tree-ssa.h is now in a position to contain only
> it's own protot
On Wed, Oct 23, 2013 at 5:11 AM, wrote:
>
>
> Sent from my iPad
>
>> On Oct 22, 2013, at 7:23 PM, Cong Hou wrote:
>>
>> This patch aims at PR58762.
>>
>> Currently GCC could not vectorize abs() operation for integers on x86
>> with only SSE2 support. For int type, the reason is that the expand o
On Wed, Oct 23, 2013 at 11:14:54AM +0200, Richard Biener wrote:
> On Tue, 22 Oct 2013, Jakub Jelinek wrote:
>
> > Hi!
> >
> > If VRP tells us that oprnd is always >= 0 or always < 0, we can generate
> > better code for the divmode vectorization.
> >
> > Bootstrapped/regtested on x86_64-linux and
On Wed, 23 Oct 2013, Jakub Jelinek wrote:
> On Wed, Oct 23, 2013 at 11:14:54AM +0200, Richard Biener wrote:
> > On Tue, 22 Oct 2013, Jakub Jelinek wrote:
> >
> > > Hi!
> > >
> > > If VRP tells us that oprnd is always >= 0 or always < 0, we can generate
> > > better code for the divmode vectoriza
On Tue, Oct 22, 2013 at 12:47:41PM -0600, Jeff Law wrote:
> On 10/22/13 07:09, Jakub Jelinek wrote:
> >I've spent over two days looking at reassoc, fixing spots where
> >we invalidly reused SSA_NAMEs (this results in wrong-debug, as the added
> >guality testcases show, even some ICEs (pr58791-3.c)
On Tue, 2013-10-22 08:46:13 -0400, Diego Novillo wrote:
> On Tue, Oct 22, 2013 at 4:22 AM, Jan-Benedict Glaw wrote:
> > This fixes it:
> >
> > 2013-10-22 Jan-Benedict Glaw
> >
> > * config/tilepro/tilepro.c: Include "tree.h".
>
> Sure. It qualifies as obvious. Thanks.
Also committe
eOn 22 Oct 22:55, Jeff Law wrote:
> On 09/17/13 02:18, Ilya Enkovich wrote:
> >Hi,
> >
> >Here is a patch introducing new type and mode for bounds. It is a part of
> >MPX ISA support patch
> >(http://gcc.gnu.org/ml/gcc-patches/2013-07/msg01094.html).
> >
> >Bootstrapped and tested on linux-x86_64
Hi,
for details see the audit trail. Tested x86_64-linux, committed to mainline.
Thanks,
Paolo.
//
2013-10-23 Paolo Carlini
PR libstdc++/58815
* include/decimal/decimal (decimal32::operator long long(),
decimal64::operator long long(), decimal128:
On Wed, 23 Oct 2013, Jakub Jelinek wrote:
> On Tue, Oct 22, 2013 at 12:47:41PM -0600, Jeff Law wrote:
> > On 10/22/13 07:09, Jakub Jelinek wrote:
> > >I've spent over two days looking at reassoc, fixing spots where
> > >we invalidly reused SSA_NAMEs (this results in wrong-debug, as the added
> > >
2013/10/23 Paolo Carlini :
> Hi,
>
> for details see the audit trail. Tested x86_64-linux, committed to mainline.
I suggest to initialize the decimal values in the new unit tests. If
http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2012/n3407.html
would be applied these initializations would b
On Wed, Oct 23, 2013 at 01:54:05PM +0200, Richard Biener wrote:
> I'm ok with the patch. Not sure if we want to backport it at all - we
> will only have wrong-debug issues (well, "only").
I've added some comments and one assert (that the uids are non-zero).
Because of that assertion I'll bootstra
Boostrapped and tested on x86_64-unknonw-linux-gnu, applied.
Richard.
2013-10-23 Richard Biener
Backport from mainline
2013-06-24 Richard Biener
PR tree-optimization/57488
* tree-ssa-pre.c (insert): Clear NEW sets before each iteration.
* gcc.dg/t
Richard Biener writes:
>> The patch does that by adding:
>>
>> wi::address (t)
>>
>> for when we want to extend tree t to addr_wide_int precision and:
>>
>> wi::extend (t)
>>
>> for when we want to extend it to max_wide_int precision. (Better names
>> welcome.) These act just like addr_w
On Wed, 23 Oct 2013, Richard Sandiford wrote:
> Richard Biener writes:
> >> The patch does that by adding:
> >>
> >> wi::address (t)
> >>
> >> for when we want to extend tree t to addr_wide_int precision and:
> >>
> >> wi::extend (t)
> >>
> >> for when we want to extend it to max_wide_int
On 10/23/2013 08:13 AM, Richard Biener wrote:
On Wed, 23 Oct 2013, Richard Sandiford wrote:
Richard Biener writes:
The patch does that by adding:
wi::address (t)
for when we want to extend tree t to addr_wide_int precision and:
wi::extend (t)
for when we want to extend it to max_wid
On Mon, 21 Oct 2013, Tobias Burnus wrote:
> Dear all,
>
> attached is a new version of the patch. Changes:
> * "#pragma GCC ivdep" instead of "#pragma ivdep"
> * Corrections to the error message in c-parser.c and a test case for it
> * New wording in the .texi and examples
>
> I am still not com
Hi Ed,
It looks like we did reserve "assume" as a keyword, but it's not being
used... By any chance, did you configure without --disable-bootstrap?
I think it would be a better solution to remove the unused keywords --
there were a couple of others that we grabbed for some other
concepts-related
On Wed, Oct 23, 2013 at 9:11 AM, Bernd Edlinger
wrote:
> Hi Richard/Joseph,
>
> I noticed, this test case crashes on arm-eabi already witout the patch.
>
> extern void abort (void);
>
> #define test_type unsigned short
> #define MAGIC (unsigned short)0x102u
>
> typedef struct s{
> unsigned char P
This patch disallows the explicit specialization of concepts, as
required by the specification. It also fixes an ICE when comparing
overloads of non-template members.
2013-10-23 Andrew Sutton
* gcc/cp/class.c (get_member_fntemplate): New.
(are_constrained_member_overloads): Only
Hi,
On Wed, 23 Oct 2013 14:37:43, Richard Biener wrote:
> The C++ memory model says that you may not introduce a data-race
> and thus you have to access Type without touching Prefix.
Thanks.
Right now I see the following priorities:
1. make -fno-strict-volatile-bitfields respect the C++ memory
On Mon, Oct 21, 2013 at 7:27 PM, Jeff Law wrote:
> On 10/21/13 06:19, Richard Biener wrote:
>
>>
>> I wonder why this isn't part of the regular jump-threading code - after
>> all the opportunity is to thead to __builtin_unreachable () ;) Otherwise
>> the question is always where you'd place this
* config/rs6000/altivec.md (mulv8hi3): Adjust for little endian.
Okay.
Thanks, David
On 10/23/13 01:58, Florian Weimer wrote:
Could you keep in mind that there is considerable interest in a
check_nonnull attribute which marks values (parameters, return values,
maybe even struct fields) that can be NULL and need to be checked
explictly prior to dereference? GCC would then warn if
Hi,
strictly speaking this isn't a bug, is more a QoI issue, but it shows
that the current minutes and hours typedefs we have got in ,
while strictly speaking conforming, lead to "surprises":
/// minutes typedef duration> minutes;
/// hours typedef duration> hours;
that is, somewhat
Similar to tree-ssa.h, tree-ssa-loops.h became an aggregator for 3 of
the tree-ssa-loop* header files. This remedies that situation.
The average .c file required only 1 of the 3 includes from tree-ssa-loop.h.
Bootstraps on x86_64-unknown-linux-gnu (with graphite enabled :-).
Regressions are
PR target/58838
* config/rs6000/rs6000.md (mulsi3_internal1 and splitter): Add
TARGET_32BIT final condition.
(mulsi3_internal2 and splitter): Same.
Index: rs6000.md
===
--- rs6000.md (revision 203930)
On Wed, Oct 23, 2013 at 10:27 AM, Andrew MacLeod wrote:
> Similar to tree-ssa.h, tree-ssa-loops.h became an aggregator for 3 of the
> tree-ssa-loop* header files. This remedies that situation.
>
> The average .c file required only 1 of the 3 includes from tree-ssa-loop.h.
>
> Bootstraps on x86_6
During testing of the generalized FSA opt I ran into a case where we
tried to thread through a joiner with abnormal outgoing edges. Since we
can't reliably clone the joiner and redirect just one of its outgoing
edges in this case, we need to avoid even trying to thread through such
blocks.
Eric Botcazou writes:
> This finally converts the Ada front-end.
>
> Tested on x86_64-suse-linux, applied on the mainline.
>
>
> 2013-10-13 Nicolas Roche
> Eric Botcazou
>
> * gcc-interface/Make-lang.in (ada/%.o): Replace individual rules with
> generic rule and add $
Hi,
ping ;)
Trev
- Forwarded message from tsaund...@mozilla.com -
Date: Thu, 10 Oct 2013 14:07:22 -0400
From: tsaund...@mozilla.com
To: dnovi...@google.com, gcc-patches@gcc.gnu.org
Cc: Trevor Saunders
Subject: [PATCH] rewrite stack vectors
X-Mailer: git-send-email 1.8.4.rc3
From: Trev
On Tue, Oct 22, 2013 at 3:50 PM, Bernd Edlinger
wrote:
> Hi,
>
> On Tue, 17 Sep 2013 01:09:45, Martin Jambor wrote:
>>> @@ -4773,6 +4738,8 @@ expand_assignment (tree to, tree from, b
>>>if (MEM_P (to_rtx)
>>>&& GET_MODE (to_rtx) == BLKmode
>>>&& GET_MODE (XEXP (to_r
On Tue, 22 Oct 2013, Cong Hou wrote:
> For abs(char/short), type conversions are needed as the current abs()
> function/operation does not accept argument of char/short type.
> Therefore when we want to get the absolute value of a char_val using
> abs (char_val), it will be converted into abs ((in
On 22/10/13 20:50, Jeff Law wrote:
> On 10/22/13 03:58, Tom de Vries wrote:
>> Richard,
>>
>> This patch adds a missing check for gimple_vdef in stmt_local_def for the
>> tail-merge pass.
>>
>> Bootstrapped and reg-tested on x86_64.
>>
>> OK for trunk, gcc-4_8-branch?
>>
>> Thanks,
>> - Tom
>>
>> 2
On Wed, Oct 23, 2013 at 5:52 PM, Joseph S. Myers
wrote:
> On Tue, 22 Oct 2013, Cong Hou wrote:
>
>> For abs(char/short), type conversions are needed as the current abs()
>> function/operation does not accept argument of char/short type.
>> Therefore when we want to get the absolute value of a char
> +static enum rtx_code
> +arm_ccmode_to_code (enum machine_mode mode)
> +{
> + switch (mode)
> +{
> +case CC_DNEmode:
> + return NE;
Why would you need to encode comparisons in CCmodes?
That looks like a mis-design to me.
> +Conditional compare instruction. Operand 2 and 5 are RTL
On Tue, Oct 22, 2013 at 12:25 PM, Bernd Edlinger
wrote:
> Hi,
>
>> On Tue, 8 Oct 2013 22:50:21, Eric Botcazou wrote:
>>>
I agree, that assigning a non-BLKmode to structures with zero-sized arrays
should be considered a bug.
>>>
>>> Fine, then let's apply Martin's patch, on mainline at le
Committed to dmalcolm/jit:
gcc/jit/
* internal-api.c: Add missing include of diagnostic-core.h
---
gcc/jit/ChangeLog.jit | 4
gcc/jit/internal-api.c | 1 +
2 files changed, 5 insertions(+)
diff --git a/gcc/jit/ChangeLog.jit b/gcc/jit/ChangeLog.jit
index 5e8d0f9..8296b62 100644
---
Hi,
On Mon, Oct 21, 2013 at 11:00:38PM -0400, Vladimir Makarov wrote:
> On 13-10-21 6:56 PM, Steven Bosscher wrote:
> >
> >+ {
> >+ bitmap_clear (&need_new);
> >+ bitmap_clear (&reachable);
> >+ return;
> >+ }
> >+
> >+ for (df_ref use = DF_REG_USE_CHAIN (R
On 10/23/13 04:35, Jakub Jelinek wrote:
For debug info quality it actually isn't just about using different
SSA_NAME, but also not reusing the defining stmt; only then the code
will magically try to create debug temporaries and expressions from the old
dead defining stmt.
Joys. Something else t
Hi,
On Wed, Oct 23, 2013 at 06:00:43PM +0200, Richard Biener wrote:
>
...
> ? And why should the same issue not exist for
>
> struct { V a[1]; }
>
indeed, it does and it's simple to trigger (on x86_64):
/* { dg-do run } */
#include
typedef long lo
> That looks backwards. Why should
>
> struct { V i; V j[0]; }
>
> have a different mode than
>
> struct { V j[0]; V i; }
>
> ?
Because we support out-of-bounds access for the array in the former case and
not in the latter case.
> And why should the same issue not exist for
>
> struct { V
> The following trivial patch avoids this and allows the
> i386-pc-solaris2.1[01] and sparc-sun-solaris2.11 bootstraps to finish.
>
> Ok for mainline?
Sure, thanks for fixing this.
--
Eric Botcazou
> While I was willing to discount the zero sized array as an
> insufficiently specified oddity, this seems to be much more serious
> and potentially common. It seems we really need to be able to detect
> these out-of-bounds situations and tell lower levels of expander that
> "whatever mode you thi
I've merged trunk r203980 into "dmalcolm/jit", in particular bringing in
the fix for the segfault seen in libbacktrace when an ICE happened in
libgccjit (fix was on trunk on r203810).
gcc/jit/
* internal-api.c: Update for rename of tree-flow.h to tree-cfg.h
in r203320, for
--- Chris,
I went ahead and applied to my local trees the various tweaks I
mentioned. I tested the below make check, check-debug and again both
-m32/-m64. See if you can spot something you don't like, otherwise I
will commit it later today.
Thanks,
Paolo.
Index: testsui
Late in the C++11 process it was decided that a constructor or
destructor can be trivial but not callable; as a result, everywhere that
assumed that a call to a trivial function didn't need any processing
needed to be updated. This patch does that.
Tested x86_64-pc-linux-gnu, applying to trun
On 10/23/2013 08:05 PM, Jason Merrill wrote:
@@ -4674,15 +4674,8 @@ deduce_noexcept_on_destructors (tree t)
if (!CLASSTYPE_METHOD_VEC (t))
return;
- bool saved_nontrivial_dtor = TYPE_HAS_NONTRIVIAL_DESTRUCTOR (t);
-
- /* Avoid early exit from synthesized_method_walk (c++/57645).
On 10/08/2013 05:54 AM, Gary Benson wrote:
diff --git a/test.c b/test.c
I don't think we want to add this to the top level directory. :)
Other than that, the patch is OK.
Jason
OK to commit to CVS?
(fwiw the pre-existing page content doesn't validate as per
http://gcc.gnu.org/contribute.html#webchanges, due to issues with the
)
Index: htdocs/lists.html
===
RCS file: /cvs/gcc/wwwdocs/htdocs/lists.html,v
retr
LWG 2165, submitted by Jonathan, argues that declaring a defaulted
constructor noexcept should not be ill-formed if the implicitly-
declared constructor would not be noexcept. At the Chicago meeting, CWG
agreed. This patch makes it deleted instead.
The second hunk adds %X for printing an exce
On Oct 23, 2013, at 2:09 AM, Richard Biener wrote:
> Good. Better names - ah well, wi::to_max_wide_int (t) and
> wi::to_addr_wide_int (t)? Btw, "addr_wide_int" is an odd name
The idea was to have one type to rule them all… everything address related...
bit offsets, byte offsets… Bit offset
On Oct 23, 2013, at 5:00 AM, Richard Sandiford
wrote:
> offset_int, max_int, wi::to_offset and wi::to_max sound OK to me.
> Kenny? Mike?
Those two names seem reasonable. to_offset should document that these are for
address offsets (and address constants) exclusively.
>
> >> Can you take a look at calls.c::special_function_p and determine if we need
> to
> >> do something special for spawn here?
> >>
> >
> > I will look into it and let you know.
> Any word on this?
>
Hi Jeff,
I looked into this function and from what I can tell, it is used to
mark ce
On 10/23/13 13:46, Iyer, Balaji V wrote:
Can you take a look at calls.c::special_function_p and
determine if we need
to
do something special for spawn here?
I will look into it and let you know.
Any word on this?
Hi Jeff, I looked into this function and from what I can tell, it is
used
On 10/23/13 10:02, Tom de Vries wrote:
On 22/10/13 20:50, Jeff Law wrote:
On 10/22/13 03:58, Tom de Vries wrote:
Richard,
This patch adds a missing check for gimple_vdef in stmt_local_def for the
tail-merge pass.
Bootstrapped and reg-tested on x86_64.
OK for trunk, gcc-4_8-branch?
Thanks,
-
> -Original Message-
> From: Jeff Law [mailto:l...@redhat.com]
> Sent: Wednesday, October 23, 2013 3:53 PM
> To: Iyer, Balaji V; r...@redhat.com; Jason Merrill (ja...@redhat.com); Aldy
> Hernandez (al...@redhat.com)
> Cc: gcc-patches@gcc.gnu.org
> Subject: Re: [PATCH] Cilk Keywords (_Cilk
On 10/23/13 02:16, Richard Biener wrote:
On Tue, 22 Oct 2013, Jeff Law wrote:
On 10/22/13 03:58, Tom de Vries wrote:
Richard,
This patch adds a missing check for gimple_vdef in stmt_local_def for the
tail-merge pass.
Bootstrapped and reg-tested on x86_64.
OK for trunk, gcc-4_8-branch?
Than
On 10/23/13 04:57, Ilya Enkovich wrote:
2013-10-23 Ilya Enkovich
* mode-classes.def (MODE_POINTER_BOUNDS): New.
* tree.def (POINTER_BOUNDS_TYPE): New.
* genmodes.c (complete_mode): Support MODE_POINTER_BOUNDS.
(POINTER_BOUNDS_MODE): New.
(make_pointer_
On Mon, 7 Oct 2013, Cong Hou wrote:
> + if (type != newtype)
> +break;
That comparison would wrongly treat as different cases where the types
differ only in one being a typedef, having qualifiers, etc. - or if in
future GCC implemented proposed TS 18661-3, cases where they differ in
e.g.
I'm not sure why this did not trigger on GNU/Linux distributions,
on FreeBSD 10 one does need to #include to get to unlink()
which is used by this test.
With this, the test compiles on my i386-unknown-freebsd10.0 tester;
without it, it doesn't.
Okay?
Gerald
2013-10-22 Gerald Pfeifer
Ok.
Hi,
Thank you both for the reviewing. I've updated the patch and also added
a test (to the gcc.dg to avoid duplication). I'll commit the patch shortly.
Thanks,
Yufeng
gcc/
* tree-ssa-math-opts.c (convert_plusminus_to_widen): Call
has_single_use () and not do the conversion
This patch stores two array lengths in an INTEGER_CST: the length that
should be used when accessing the constant in its TYPE_PRECISION,
and the one that should be used for wider precisions. This means that
the main tree decompose routine is just a simple storage_ref constructor.
It also means tha
Mike Stump writes:
> On Oct 23, 2013, at 5:00 AM, Richard Sandiford
> wrote:
>> offset_int, max_int, wi::to_offset and wi::to_max sound OK to me.
>> Kenny? Mike?
>
> Those two names seem reasonable. to_offset should document that these
> are for address offsets (and address constants) exclusive
On 10/23/2013 12:37 AM, Paolo Carlini wrote:
Hi,
"François Dumont" ha scritto:
Hi
Here is a patch to clean up a little some debug functions. I got
rid of the __check_singular_aux, simply playing with __check_singular
overloads was enough. I also added the missing __check_dereferenceable
On 10/21/13 08:20, Ilya Enkovich wrote:
diff --git a/gcc/doc/tm.texi b/gcc/doc/tm.texi
index 8d220f3..79bd0f9 100644
--- a/gcc/doc/tm.texi
+++ b/gcc/doc/tm.texi
+@deftypefn {Target Hook} rtx TARGET_LOAD_BOUNDS_FOR_ARG (rtx @var{slot}, rtx
@var{arg}, rtx @var{slot_no})
+This hook is used to emit
On 10/21/13 05:49, Ilya Enkovich wrote:
Hi,
This patch introduces built-in functions used by Pointers Checker and flag to
enable Pointers Checker. Builtins available for user are expanded in
expand_builtin. All other builtins are not allowed in expand until generic
version of Pointers Cheker
The following fixes the gcc.target/powerpc/direct-move-*2.c executable
test failures. Committed as obvious.
2013-10-23 Pat Haugen
* gcc.target/powerpc/direct-move.h: Fix header for executable tests.
Index: gcc/testsuite/gcc.target/powerpc/direct-move.h
=
A rather simple patch, which tries to implement Fortran 2008's C1116
(see also PR for the wording).
While creating the patch, I found a reject-valid issue, which is now
tracked as PR fortran/58857.
Build and regtested on x86-64-gnu-linux.
OK for the trunk?
Tobias
2013-10-23 Tobias Burnus
On 10/23/2013 11:22 PM, François Dumont wrote:
You are right, I am preparing a test case. However you have to know
that __check_dereferenceable is simply not used for the moment. It is
only because I have started using it for a debug mode evolution that I
discovered the issue.
Ok, thanks. Now h
Hi!
When building on gcc111, I tried to use different compilers for my
build robot. This machine has IBM's XL C compiler installed, but not
the C++ compiler. So I used CC=xlC, g++ was auto-detected for CXX.
This led to depmode=aix, but g++ was actually used later on for
calculating dependencies, t
On 10/23/2013 02:41 PM, Jeff Law wrote:
> Out of curiosity, did you consider and/or discuss with Richard whether or not
> to make these target-dependent or target-independent builtins? I realize it's
> a bit problematic with Richard being involved during the NDA portion and
> someone else during t
On Wed, Oct 23, 2013 at 6:46 PM, Martin Jambor wrote:
> /* Perform the second half of the transformation started in
> @@ -4522,7 +4704,15 @@ ira (FILE *f)
> allocation because of -O0 usage or because the function is too
> big. */
>if (ira_conflicts_p)
> -find_moveable_pseudos
Hi,
arm testcases neon-vcond-ltgt.c and neon-vcond-unordered.c fails in
Linaro 4.8 branch. It is not reproducable with trunk but it can happen.
Both neon-vcond-ltgt.c and neon-vcond-unordered.c scans for vbsl
instruction, with other vector instructions. However, as per the comment
for "neon_vbsl_
On Wed, 23 Oct 2013, David Malcolm wrote:
> OK to commit to CVS?
Yes, thank you, David.
> (fwiw the pre-existing page content doesn't validate as per
> http://gcc.gnu.org/contribute.html#webchanges, due to issues with the
> )
I was going to say "Go ahead and commit, and if my automated checker
On, Wed, 23 Oct 2013 17:36:41Richard Biener wrote:
>
> if bitregion_start/end are used after the adjust_address call then they have
> to be adjusted (or bitpos left in place). In fact why we apply byte-parts of
> bitpos here only if offset != 0 is weird. OTOH this code is _very_ old...
> what happe
This patch from Chris Manghane changes the Go frontend to use the
backend interface for a couple of runtime calls. Bootstrapped and ran
Go testsuite on x86_64-unknown-linux-gnu. Committed to mainline.
Ian
diff -r 46c1804a8e17 go/expressions.cc
--- a/go/expressions.cc Fri Oct 18 06:25:23 2013 -0
On 10/23/2013 08:36 AM, Andrew Sutton wrote:
Hi Ed,
It looks like we did reserve "assume" as a keyword, but it's not being
used... By any chance, did you configure without --disable-bootstrap?
I think it would be a better solution to remove the unused keywords --
there were a couple of others t
In this pr, we have a -1 in type __int128. Since this value can be
represented in a HOST_WIDE_INT, we expand this to a const_int.
The expansion from tree to rtl happens in expand_builtin_atomic_store. And as
with most of our builtins, we then pass off the rtl to another routine for
expansion. W
On Wed, Oct 23, 2013 at 6:21 PM, Jan-Benedict Glaw wrote:
> Hi!
>
> When building on gcc111, I tried to use different compilers for my
> build robot. This machine has IBM's XL C compiler installed, but not
> the C++ compiler. So I used CC=xlC, g++ was auto-detected for CXX.
> This led to depmode=a
On 10/21/2013 03:01 PM, Yufeng Zhang wrote:
>
> This patch changes the widening_mul pass to fuse the widening multiply with
> accumulate only when the multiply has single use. The widening_mul pass
> currently does the conversion regardless of the number of the uses, which can
> cause poor code-g
I've committed this patch to C11-atomic branch with further
miscellaneous fixes. The testcases are expanded to cover everything I
found in going through the language parts of C11 for atomics issues
that can be covered through compile rather than execute tests.
Corresponding compiler fixes are made
On Thu, 2013-10-24 at 01:46 +0200, Gerald Pfeifer wrote:
> On Wed, 23 Oct 2013, David Malcolm wrote:
> > OK to commit to CVS?
>
> Yes, thank you, David.
Thanks; committed.
> > (fwiw the pre-existing page content doesn't validate as per
> > http://gcc.gnu.org/contribute.html#webchanges, due to is
On Tue, 2013-10-15 at 12:45 -0600, Jeff Law wrote:
> On 09/24/13 11:49, David Malcolm wrote:
> > Extend gengtype (and gtype.state reading/writing) so that it is able to
> > parse base classes in simple cases, and only attempt to do it for
> > GTY-marked types.
> >
> > * gengtype-parse.c (requir
> "Jan-Benedict" == Jan-Benedict Glaw writes:
Jan-Benedict> I'd like to get some comments on this patch, which works
Jan-Benedict> for me for stage-1 builds with gcc/g++ (on a amd64-linux
Jan-Benedict> box) as well as with xlC/g++ on gcc111 (the mentioned AIX
Jan-Benedict> machine):
FWIW,
1 - 100 of 106 matches
Mail list logo
