Hello,
The SH2A is actually superscalar and thus it might be better to treat it
as such for scheduling. It also is a harvard arch, but after adding it
to the TARGET_HARVARD arch, I've noticed that the macro is only used as
a parameter for instruction cache invalidation. Maybe it's better to
remo
Hi,
I checked in this patch to sync configure.ac with src.
H.J.
---
commit a342ff81fdd357cd5e83c51ebff0cc69f8259be2
Author: Corinna Vinschen
Date: Wed Oct 24 13:41:35 2012 +
* configure.ac (FLAGS_FOR_TARGET,target=cygwin): Fix for building
against Mingw64 w32api.
*
On Sat, Nov 3, 2012 at 8:54 AM, Bruce Korb wrote:
>
I checked in this patch to regenerate configure.
--
H.J.
---
Index: ChangeLog
===
--- ChangeLog (revision 193127)
+++ ChangeLog (working copy)
@@ -1,3 +1,7 @@
+2012-11-03 H.J
On Thu, Nov 1, 2012 at 5:17 PM, Michael Meissner
wrote:
> This patch adds 4 additional configuration switches, that allow the person
> building the compiler to add additional prefixes to search for additional
> executables and startfiles.
>
> If the backend has the appropriate tweaks installed, i
On Tue, Oct 30, 2012 at 10:48 AM, Steven Bosscher wrote:
> On Tue, Oct 30, 2012 at 6:20 AM, Teresa Johnson wrote:
>> Index: bb-reorder.c
>> ===
>> --- bb-reorder.c(revision 192692)
>> +++ bb-reorder.c(working copy)
>>
On Thu, Nov 1, 2012 at 2:26 PM, Teresa Johnson wrote:
> On Thu, Nov 1, 2012 at 10:19 AM, Teresa Johnson wrote:
>> On Wed, Oct 31, 2012 at 4:02 PM, Steven Bosscher
>> wrote:
>>> On Tue, Oct 30, 2012 at 10:43 PM, Teresa Johnson wrote:
Sure, I will give this a try after your verification patc
On Sat, Nov 3, 2012 at 8:52 PM, Richard Sandiford wrote:
> This patch rewrites sched-vis.c to use obstacks rather than fixed-length
> buffers.
Hmm, nice but...
I have some patches for GCC 4.9 to use the pretty-print infrastructure
instead (and also some half-finished ones for the non-slim RTL dum
This patch rewrites sched-vis.c to use obstacks rather than fixed-length
buffers. It then uses that file to provide a SLIM_RTX_STRING macro,
so that dump routines can do something like:
fprintf (dump_file, "convert from %s to %s",
SLIM_RTX_STRING (from, 1), SLIM_RTX_STRING (to, 1
Harshit, Nov 5 is the gcc48 cutoff date. If you want to have the x-ray
instrumentation feature into this release, you will need to port your
patch and submit for trunk review now.
On Tue, Oct 30, 2012 at 5:15 PM, Harshit Chopra wrote:
> Adding function attributes: 'always_patch_for_instrumentati
Hi,
After some of the recent patches which does cross jumping at the
tree level (tail merge) and the patch for tree-ssa-ifcombine, this
testcase fails as we merge the if statements that lead to abort. This
patch fixes the problem by changing one of the calls to abort to call
a different function
Hi,
On 11/03/2012 03:43 PM, François Dumont wrote:
On 11/03/2012 03:12 PM, Jonathan Wakely wrote:
On 3 November 2012 14:11, Jonathan Wakely wrote:
On 3 November 2012 13:53, François Dumont wrote:
Here is an updated proposal using default implementations as much as
possible.
Please check the i
On Sat, Nov 03, 2012 at 10:05:35AM -0700, Wei Mi wrote:
> --- gcc/sanitizer.def (revision 0)
> +++ gcc/sanitizer.def (revision 0)
> @@ -0,0 +1,31 @@
> +DEF_SANITIZER_BUILTIN(BUILT_IN_TSAN_WRITE_16, "__tsan_write16",
> + BT_FN_VOID_PTR, ATTR_NOTHROW_LEAF_LIST)
> +
> +
> +
Pleas
Ping ** 0.2857?
I won't be able to commit this for about a week starting tomorrow, so
I would be obliged if I could so so now.
Regards
Thomas
after the dicsussion on c.l.f, it become clear that passing a DO loop
variable to an INTENT(OUT) or INTENT(INOUT) dummy argument is an error.
On Sat, 3 Nov 2012, Andi Kleen wrote:
> "Joseph S. Myers" writes:
> >>
> >> 2012-10-31 Andi Kleen
> >>
> >>* g++.dg/fno-instrument-function.C: Add.
> >>* gcc.dg/fno-instrument-function.c: Add.
> >
> > You have a C++ test but no corresponding C++ front-end change; how does
> > that wo
On Sat, Nov 03, 2012 at 02:11:22PM -0200, Alexandre Oliva wrote:
> I didn't try delaying the creation of the pointer set as you did, but
> I'm now thinking using a linear on-stack vector for searches up to some
> size might be even faster than creating the pointer set at the first
> var.
THat woul
"Joseph S. Myers" writes:
>>
>> 2012-10-31 Andi Kleen
>>
>> * g++.dg/fno-instrument-function.C: Add.
>> * gcc.dg/fno-instrument-function.c: Add.
>
> You have a C++ test but no corresponding C++ front-end change; how does
> that work?
The code is in start_function, which is called b
On Wed, 31 Oct 2012, Andi Kleen wrote:
> gcc/:
>
> 2012-10-31 Andi Kleen
>
> * c.opt (fno-instrument-function): Document.
>
> gcc/c:
>
> 2012-10-31 Andi Kleen
>
> * c-decl.c (start_function): Handle force_no_instrument_function
>
> gcc/testsuite:
>
> 2012-10-31 Andi Kleen
On Sat, 27 Oct 2012, Manuel L?pez-Ib??ez wrote:
> More mostly easy conversions to LangEnabledBy. Regtested on x86_64-linux-gnu.
>
> OK?
OK, but I note that fortran/options.c contains a reference to warn_switch,
a variable for an option you are moving to c.opt. It looks like that
setting in fo
On Sat, 27 Oct 2012, Manuel L?pez-Ib??ez wrote:
> On 19 October 2012 18:17, Joseph S. Myers wrote:
> > On Wed, 17 Oct 2012, Manuel L?pez-Ib??ez wrote:
> >
> >> documentation but I can also implement -Wformat=0 being an alias for
> >> -Wno-format and -Wformat=1 an alias for -Wformat and simply rej
Sorry, I attached an incorrect patch.txt yesterday. This is the correct one.
Thanks,
Wei.
On Fri, Nov 2, 2012 at 6:31 PM, Wei Mi wrote:
> Hi,
>
> Thanks for so many useful comments! I update the file according to the
> comments. The major changes include adding sanitizer.def and
> generating gim
On Nov 2, 2012, Alexandre Oliva wrote:
> On Nov 1, 2012, Jakub Jelinek wrote:
>> Even for stmt frontiers it is IMHO undesirable to duplicate
>> (perhaps many times) the whole sequence, as one would then reply the var
>> changing sequence in the debugger perhaps once in the original bb, then
>>
Index: gcc/gcov-io.c
===
--- gcc/gcov-io.c (revision 193123)
+++ gcc/gcov-io.c (working copy)
@@ -92,7 +92,8 @@
{
/* Read-only mode - acquire a read-lock. */
s_flock.l_type = F_RDLCK;
- fd = open (name, O_RDONL
On Wed, 2012-10-31 at 21:54 -0500, Peter Bergner wrote:
> On Wed, 2012-10-31 at 09:02 -0500, Peter Bergner wrote:
> > On Wed, 2012-10-31 at 14:55 +0100, Jakub Jelinek wrote:
> > > On Wed, Oct 31, 2012 at 08:53:31AM -0500, Peter Bergner wrote:
> > > > Great. Jakub, were you going to commit your cha
On 11/03/2012 03:12 PM, Jonathan Wakely wrote:
On 3 November 2012 14:11, Jonathan Wakely wrote:
On 3 November 2012 13:53, François Dumont wrote:
Here is an updated proposal using default implementations as much as
possible.
Please check the indentation of the "struct emplace_point" specializat
Quoting Richard Sandiford :
Joern Rennecke writes:
So, if we want to go with a declarative approach, each instruction can have
a plurality of variants, each described by a size, an instruction fetch
alignment base, an instruction fetch alignment offset set, a cache
line base,
a cache lin
On 3 November 2012 14:11, Jonathan Wakely wrote:
> On 3 November 2012 13:53, François Dumont wrote:
>> Here is an updated proposal using default implementations as much as
>> possible.
>
> Please check the indentation of the "struct emplace_point" specializations.
>
> In testsuite/util/testsuite_co
On 3 November 2012 13:53, François Dumont wrote:
> Here is an updated proposal using default implementations as much as
> possible.
Please check the indentation of the "struct emplace_point" specializations.
In testsuite/util/testsuite_container_traits.h
+#ifdef __GXX_EXPERIMENTAL_CXX0X__
+
On 09/28/2012 10:14 PM, François Dumont wrote:
Hi
I enhance scope of the small exception safety test framework used
to validate different levels of exception safety through the
containers operations. I have added calls to C++ methods
emplace/emplace_front/emplace_back/emplace_hint in C++1
final_scan_insn didn't handle the IEEE-aware compares in a
non-jump/cmove insn, causing wrong code if it was generated with swapped
operands (CC_REVERSED set). Tested on m68k-linux.
Andreas.
* final.c (final_scan_insn) [HAVE_cc0]: Handle all comparison
codes in non-jump and cmove
While working on the insertion and extraction optabs, I came across
some strange-looking behaviour relatd to set_mem_attributes_minus_bitpos.
In gcc.c-torture/execute/20030714-1.c we have:
struct RenderStyle
{
struct NonInheritedFlags
{
union
{
struct
This patch adds the new optabs and makes MIPS use them.
There are probably some clean-up opportunities in mips_use_ins_ext_p
after this change, but I'll look at that separately.
Tested as described in the covering note. OK to install?
Richard
gcc/
* doc/md.texi (extv@var{m}, extvmisal
This patch adds a new interface for asking about insertion and extraction
instructions, to replace mode_for_extraction. This first version just uses
the traditional patterns; the next patch actually adds the new optabs.
recog has some odd code to change the mode of a memory involved in a
sign_ext
This patch makes bit_field_mode_iterator take -fstrict-volatile-bitfields
into account, in cases where the size of the underlying object is known.
This is used in the next patch.
Tested as described in the covering note. OK to install?
Richard
gcc/
* machmode.h (bit_field_mode_iterator
This patch replaces:
/* Stop if the mode is wider than the alignment of the containing
object.
It is tempting to omit the following line unless STRICT_ALIGNMENT
is true. But that is incorrect, since if the bitfield uses part
of 3 bytes and we use a 4-byt
get_best_mode has various checks to decide what counts as an acceptable
bitfield mode. It actually has two copies of them, with slightly different
alignment checks:
MIN (unit, BIGGEST_ALIGNMENT) > align
vs.
unit <= MIN (align, BIGGEST_ALIGNMENT)
The second looks more correct, since we can'
This patch is the return of narrow_bit_field_mem, originally posted as
part of the preliminary patches. This time I've used a pointer rather
than a reference for the final parameter. (I agree consistency is good.)
The new version also takes the size of the field as a parameter and
handles BLKmod
As mentioned in the covering note, the new memory optabs take a BLKmode
reference to the bitfield. That requires a new adjust_bitfield_address
interface that takes the size of the reference as a parameter.
Tested as described in the covering note. OK to install?
Richard
gcc/
* expr.h
combine.c:make_extraction peels off truncating lowpart subregs
from the structure operand. On !TRULY_NOOP_TRUNCATION targets,
truncations are represented as TRUNCATE instead, so this patch
handles them too.
As for why this patch is suddenly needed now: at the moment, we start
out with INNER being
This series was inspired by Andrew's patch from September:
http://gcc.gnu.org/ml/gcc-patches/2012-09/msg00126.html
and by a general frustration with the insv, extv and extzv interface.
The patches add 6 new optabs:
insvM: an M<-M register insertion
insvmisalignM: a BLK<-M m
Adam added octeon-exts-7.c in:
http://gcc.gnu.org/ml/gcc-patches/2009-07/msg00406.html
to test that a sign_extract, truncate, sign_extend sequence got
folded into a single sign_extract on targets where truncation is a
sign-extension operation. The patches that I'm about to post defeat
the te
Hi,
any comments appreciated, also if you think it's unnecessary. This
issue came up when I submitted a patch for PR fortran/51727 that uses
std::map. If this isn't approved within the next few days, I'll fix PR
fortran/51727 in a way that doesn't use C++ and thus needs more
boilerplate co
Andrew Pinski writes:
> Hi,
> For some reason cse produced:
> (insn 27 43 28 2 (set (reg:SI 231 [+4 ])
> (plus:SI (reg:SI 229 [+4 ])
> (const_int 0 [0]))) t.c:24 10 {*addsi3}
> (nil))
>
> (insn 28 27 29 2 (set (reg:SI 212)
> (ltu:SI (reg:SI 231 [+4 ])
>
[After multiple failed attempts at compressing this huge patch enough to
let it pass through the drastic mailing engine's 400KB size limit, I am
sending a link to the patch (at the end of this message) to let people
download it instead. Sorry for spamming the people in the CC list.]
This patch
Hi,
The testcase shows -O -mx32 -maddress-mode=long -fPIC -S generates;
x.i:22:37: internal compiler error: in plus_constant, at explow.c:88
info[0x6eff - dyn->d_tag + 12] = dyn;
expand_expr_real_2 has
/* No sense saving up arithmetic to be done
if it's all in the wrong mod
44 matches
Mail list logo
