My main concern is that the precise collector we have in place now
requires substantial care to use. It is supported by a tool that
does not really understand C, let alone C++. We avoid the problems
when annotations are unnecessary. We can do that in one of two
ways, either by upgrading
The attached patch moves the code from integrate.c to (what I hope you
agree to be) better places:
* inliner code goes to tree-inline.c
* functions only called from dwarf2out.c are moved there.
* allocate_initial_values is moved to ira.c
* the initial-value stuff is moved to function.c
This patch documents the AArch64 branch in wwwdocs/htdocs/svn.html.
OK?
/Marcus
Proposed ChangeLog:
* htdocs/svn.html: Document aarch64 branch.diff -u -p -r1.173 svn.html
--- svn.html12 Apr 2012 01:26:33 - 1.173
+++ svn.html29 May 2012 07:28:17 -
@@ -395,6
Since in 4.7, libgcc/config/aarch64/t-aarch64 only contains makefile rules for
crti.o and crtn.o and these rules are automatically added by the generic make
system, we can remove it. I have verified that ctri.o and ctrn.o are still
generated correctly.
Addition to libgcc/ChangeLog:
2012-05-28
On Tue, May 29, 2012 at 8:43 AM, Eric Botcazou ebotca...@adacore.com wrote:
The attached patch moves the code from integrate.c to (what I hope you
agree to be) better places:
* inliner code goes to tree-inline.c
* functions only called from dwarf2out.c are moved there.
*
Hello,
The attached patch does some maintenance on alias.c:
* Change reg_known_value from rtx * to a VEC(rtx).
* Change reg_known_equiv_p from char * to an sbitmap.
Nothing spectacular, just maintenance.
Bootstrapped and regtested on x86_64-unknown-linux-gnu. OK?
Ciao!
Steven
alias_tlc.diff
On 05/28/2012 06:27 PM, Joseph S. Myers wrote:
On Mon, 28 May 2012, Christian Bruel wrote:
On 05/28/2012 01:11 PM, Joseph S. Myers wrote:
On Mon, 28 May 2012, Christian Bruel wrote:
I shared the same concern, however, after playing bits with spec toys, I
couldn't a find a way to get a
domi...@lps.ens.fr (Dominique Dhumieres) writes:
I have regstapped r187893 with the following patch
[karma] gcc/darwin_buildw% diff -up ../_gcc_clean/libcpp/lex.c
../work/libcpp/lex.c
--- ../_gcc_clean/libcpp/lex.c2012-05-25 08:54:05.0 +0200
+++ ../work/libcpp/lex.c
On Tue, May 29, 2012 at 10:57:28AM +0200, Dodji Seketeli wrote:
So, dear maintainers, is the patch below OK? It bootstraps on
x86_64-unknown-linux-gnu as well.
PR bootstrap/53459
* lex.c (search_line_fast): Avoid unused local typedefs to simulate
a static assertion.
---
Ping for:
On May 15, 2012, at 11:26 AM, Tristan Gingold wrote:
Hi,
the end_prologue debug hook (only used by dwarf on VMS) is currently called
at the NOTE_INSN_FUNCTION_BEG. This is not what its name implies, neither
what the VMS debugger expect and neither what a comment in dwarf2out.c
Ping for:
On May 15, 2012, at 10:59 AM, Tristan Gingold wrote:
Hi,
I got ICE in plus_constant (after the assertions were added) due to
expand_expr_addr_expr_1 during build on ia64/Openvms.
This function is called with TMODE == SImode (32 bit pointers) but EXP
designating a variable
Yes, I realize that, but it looks like that code is not doing
something because old integrate.c choked on it. Quoting that part of
the patch:
Index: expmed.c
===
--- expmed.c (revision 187936)
+++ expmed.c (working copy)
@@
On Sun, May 27, 2012 at 2:48 AM, Michael Matz m...@suse.de wrote:
Hi,
and this is a further small cleanup. pop_temp_slots is now the same as
free_temp_slots (modulo the level-- of course), so there's no need in
calling both. (and preserve_temp_slots(NULL) is useless now).
Regstrapped on
On Sun, May 27, 2012 at 5:03 AM, Michael Matz m...@suse.de wrote:
Hi,
[for certain test cases :-) ]
the temp slot cleanups I just sent where actually motivated by PR38474.
It exposes many slownesses in the compiler, but at -O0 the only remaining
one is the expand phase. expanding variables
On Tue, May 29, 2012 at 10:46 AM, Steven Bosscher stevenb@gmail.com wrote:
Hello,
The attached patch does some maintenance on alias.c:
* Change reg_known_value from rtx * to a VEC(rtx).
* Change reg_known_equiv_p from char * to an sbitmap.
Nothing spectacular, just maintenance.
On Fri, May 25, 2012 at 8:43 PM, Mike Stump mikest...@comcast.net wrote:
On May 25, 2012, at 10:50 AM, Lawrence Crowl wrote:
Diego and I looked long and hard at this issue. It all came down
to a sequence of problems. First, libstdc++ isn't rigged for GTY,
If portability to other C++
On 05/28/2012 06:39 PM, Jason Merrill wrote:
OK.
Thanks. I also applied as obvious the below, provided by Manuel, and
changing this warning too to use the new LangEnabledBy.
Paolo.
/
2012-05-29 Manuel López-Ibáñez m...@gcc.gnu.org
* c.opt (Wmissing-braces):
Hello,
Consider this short test snippet:
-8---
#define STRINGIFY(x) #x
#define TEST(x) \
_Pragma(STRINGIFY(GCC diagnostic ignored -Wunused-local-typedefs)) \
typedef int myint;
void bar ()
{
TEST(myint)
}
On Tue, 29 May 2012, Christian Bruel wrote:
The existing rule is supposed to be: options are only accepted if in
*both* a .opt file *and* a spec. If not in a .opt file, the common
machinery will reject them; if in a .opt file but not a spec, the driver's
own validation machinery will
Hi,
On 05/29/2012 12:19 PM, Dodji Seketeli wrote:
The _Pragma is effectively ignored,
hi Dodji, and sorry for taking the occasion to mention other issues with
pragmas which definitely can be handled separately, but I can't resist
;) In fact we have a number of open PRs:
Hi!
As soon as line length goes over 200 bytes, read_line starts leaking
memory (for line length x where x 200 bytes it leaks
((1UL ((x - 199) / 2)) - 1) * 200 bytes).
This patch fixes it. As noted by Manuel, he copied that buggy code
from gcov.c, so this fixes gcov too.
On Mon, 21 May 2012, Aldy Hernandez wrote:
On 05/16/12 07:53, Richard Guenther wrote:
On Mon, 7 May 2012, Aldy Hernandez wrote:
[Sorry for the delay; I was on vacation.]
I am forgoing the load avoidance code altogether to simplify things. Thanks.
+ /* Emit the load code into the
On Tue, May 29, 2012 at 1:12 PM, Jakub Jelinek ja...@redhat.com wrote:
Hi!
As soon as line length goes over 200 bytes, read_line starts leaking
memory (for line length x where x 200 bytes it leaks
((1UL ((x - 199) / 2)) - 1) * 200 bytes).
This patch fixes it. As noted by Manuel, he
On 05/29/2012 12:50 PM, Joseph S. Myers wrote:
On Tue, 29 May 2012, Christian Bruel wrote:
The existing rule is supposed to be: options are only accepted if in
*both* a .opt file *and* a spec. If not in a .opt file, the common
machinery will reject them; if in a .opt file but not a
On May 25, 2012, at 2:42 PM, Lawrence Crowl wrote:
On 5/24/12, Jakub Jelinek ja...@redhat.com wrote:
You haven't looked at the most important problem of that approach
- code bloat.
Are you claiming that the size of the binary is more important than
run-time performance, type safety, and
On May 26, 2012, at 11:25 PM, Gabriel Dos Reis wrote:
how do I want the codebase to look like in 10, 15, 20, 25 years.
Gee, if the next 25 look just like the last 25, it will look very similar to
what it looks like today. :-)
Hi,
On Tue, 29 May 2012, Richard Guenther wrote:
The other change in function.c further improves the temp slot
machinery. While expanding free_temp_slots is called after each
statement, and if it is able to free a slot it needs to update the
RTX-slot mapping (a htab_t). The freeing
This puts checking code into place that we do not put globals
into referenced-vars (nor allocate var-annotations for them).
It also fixes remaining cases in mudflap and matrix-reorg.
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied.
Richard.
2012-05-29 Richard Guenther
Also note: the line numbers listed below work for me with yesterday's
up-to-date trunk, but if you want to use these commands, you should rerun
the commands above so that the line numbers reflect your actual sources.
)
This is just a heads up.
I'm not volunteering to make these changes.
On May 29, 2012, at 2:49 PM, Jim Meyering wrote:
Running the following command spots over 1500 typos, and suggests fixes:
$ git ls-files|misspellings -f -|grep -v '^ERROR:'|perl -pe \
's/^(.*?)\[(\d+)\]: (\w+) - (.*?)$/sed -i '\''${2}s!$3!$4!'\'' $1/' k
The misspellings command comes
On 12-05-25 18:29 , Cary Coutant wrote:
2012-05-25 Sterling Augustinesaugust...@google.com
Cary Coutantccout...@google.com
* gcc/dwarf2out.c (remove_loc_list_addr_table_entries): New function.
(is_class_die): Return TRUE for DW_TAG_structure_type.
On Tue, 29 May 2012, Christian Bruel wrote:
I agree. I see two potential settings of the validated field. The %
that we just reviewed, and the case : /* We have Xno-YYY, search for
XYYY. */
It seems possible to remove those, I've just checked this during lunch
:-) with the scenarios
Arnaud Charlet wrote:
Also note: the line numbers listed below work for me with yesterday's
up-to-date trunk, but if you want to use these commands, you should rerun
the commands above so that the line numbers reflect your actual sources.
)
This is just a heads up.
I'm not volunteering to
A lot of the changes are to files that GCC either imports strictly
verbatim with no local changes at all, or will make changes locally to
only for serious issues relating to integration with GCC; in either case,
such typos cannot justify making changes locally in GCC so if you want
those
On Tue, 29 May 2012, Jim Meyering wrote:
sed -i '464s!occurence!occurrence!' gcc/doc/include/texinfo.tex
sed -i '4802s!achive!archive,achieve!' gcc/doc/include/texinfo.tex
I missed these when first listing files from upstream projects; anyway,
I've now updated texinfo.tex from upstream to fix
This fixes PR53516 - we are happily vectorizing/memsetting stores
to bitfields. That's obviously wrong.
Bootstrapped on x86_64-unknown-linux-gnu, testing in progress.
Richard.
2012-05-29 Richard Guenther rguent...@suse.de
PR tree-optimization/53516
* tree-data-ref.c
Joseph S. Myers wrote:
I've applied this patch to fix what seemed to be some of the more
straightforward cases covered by gcc/ChangeLog.
Thanks. I've run the same thing on gettext and reported the results upstream.
Hi!
The subject already says it all. It got me confused what »funcs« that
might be.
OK to commit with regenerating configure?
diff --git a/libatomic/acinclude.m4 b/libatomic/acinclude.m4
index 282a992..a236a6d 100644
--- a/libatomic/acinclude.m4
+++ b/libatomic/acinclude.m4
@@ -188,7 +188,7 @@
On Tue, May 29, 2012 at 11:17 AM, Eric Botcazou ebotca...@adacore.com wrote:
Yes, I realize that, but it looks like that code is not doing
something because old integrate.c choked on it. Quoting that part of
the patch:
Index: expmed.c
On Tue, 29 May 2012, Jim Meyering wrote:
sed -i '167s!Millenial!Millennial!' libgcc/config/libbid/bid128_string.c
sed -i '30s!guranteed!guaranteed!' libgcc/config/libbid/bid64_fma.c
sed -i '174s!oposite!opposite!' libgcc/config/libbid/bid64_fma.c
sed -i '218s!oposite!opposite!'
On Tue, 29 May 2012, Joseph S. Myers wrote:
For libjava and gcc/java changes, CC j...@gcc.gnu.org on the list (they
may well be able to deal with the classpath typos although properly it's a
separate project). For libstdc++-v3 changes (other than
libstdc++-v3/doc/html/ext/ which are files
Done, updated diff attached.
You also need to adjust the ChangeLog:
* expmed.c (extract_fixed_bit_field): Add ??? for seemingly
outdated comment about integrate.c.
OK for trunk?
I can only formally approve the cse.c and expmed.c hunks, so you probably need
to run this by a
On 12-05-29 11:09 , Eric Botcazou wrote:
I can only formally approve the cse.c and expmed.c hunks, so you probably need
to run this by a GR.
The other changes are OK.
Nice cleanup in any case!
Indeed. Though you probably made my next cxx-conversion merge more
difficult ;)
Thanks for
Hi!
On Tue, 29 May 2012 16:35:31 +0200, I wrote:
The subject already says it all. It got me confused what »funcs« that
might be.
OK to commit with regenerating configure?
diff --git a/libatomic/acinclude.m4 b/libatomic/acinclude.m4
index 282a992..a236a6d 100644
---
Hi!
While looking at libatomic's, I noticed that fixincludes' configure can't
be regenerated without errors. Here is a patch, in spirit of
749dea2a0549c126a0e992a6dd8e9b5eb28e1cee. OK to commit?
fixincludes/
* configure.ac: Use GCC_AC_FUNC_MMAP_BLACKLIST instead of
Ian Lance Taylor wrote:
Georg-Johann Lay a...@gjlay.de writes:
The avr backend auto-generates a part of the texi documentation by
means of a small C program. The relevant part of t-avr reads:
s-avr-mmcu-texi: gen-avr-mmcu-texi$(build_exeext)
$(RUN_GEN) $ | sed -e 's:\r::g'
This patch evaluates _FORTIFY_SOURCE in a way similar to GNU libc.
If set, std::vector::operator[] throws if the index is out of bounds.
This is compliant with the standard because such usage triggers
undefined behavior. _FORTIFY_SOURCE users expect some performance hit.
Okay for trunk?
On May 29, 2012, at 5:27 PM, Thomas Schwinge wrote:
Hi!
While looking at libatomic's, I noticed that fixincludes' configure can't
be regenerated without errors. Here is a patch, in spirit of
749dea2a0549c126a0e992a6dd8e9b5eb28e1cee. OK to commit?
Humm, looks like the same as Tobias
On Tue, 29 May 2012, Jim Meyering wrote:
sed -i '8s!compability!compatibility!' config/mt-sde
sed -i '21s!enviroments!environments!' config/stdint.m4
sed -i '293s!arbitary!arbitrary!' config/tcl.m4
sed -i '376s!arbitary!arbitrary!' config/tcl.m4
sed -i '818s!apropriate!appropriate!'
This patch flags operator new on variably modified types as an error.
If this is acceptable, this will simplify the implementation of the
C++11 requirement to throw std::bad_array_new_length instead of
allocating a memory region which is too short.
Okay for trunk? Or should I guard this with
Please do send such a patch (with an explanation in each case of how
validated still gets set with that redundant setting removed).
'validated' can only be removed for user --specs switches, this is why I
think it doesn't need to be propagated to do_specs. My mistake from the
previous mail:
Hi All,
Last week I went to build a mips-gnu-linux toolchain and the built compiler
seg faulted while building glibc. Bisecting the change pointed to r187757. The
problem really goes back to r178692, which added support for
-Wunused-local-typedefs. r187757 just made it easier to hit by
On Tue, May 29, 2012 at 11:00 AM, Florian Weimer fwei...@redhat.com wrote:
This patch flags operator new on variably modified types as an error.
If this is acceptable, this will simplify the implementation of the
C++11 requirement to throw std::bad_array_new_length instead of
allocating a
Hi,
This patch evaluates _FORTIFY_SOURCE in a way similar to GNU libc.
If set, std::vector::operator[] throws if the index is out of bounds.
This is compliant with the standard because such usage triggers
undefined behavior. _FORTIFY_SOURCE users expect some performance hit.
Indeed. But at
On 05/26/2012 01:27 AM, Ramana Radhakrishnan wrote:
- for (i = 0; i nelt; i += diff)
+ for (i = 0; i nelt ; i += (diff + 1))
for (j = 0; j= diff; j += 1)
- if (d-perm[i + j] != i + diff - j)
- return false;
+ {
+ /* This is guaranteed to be true as the value of
Hi,
in this pretty old issue we crash for this testcase:
namespace N
{
namespace M = N;
namespace M {}// { dg-error namespace alias }
}
after the error message, because error recovery after error fails. We
try to do:
error (namespace alias %qD not allowed here, assuming
Hi, Uros!
Sorry, I didn't realize that patch was missed. I attached new version.
Changelog:
2012-05-29 Yuri Rumyantsev yuri.s.rumyant...@intel.com
* config/i386/i386.c (x86_sched_reorder): New function.
Added new function x86_sched_reorder.
As for multiply modes, currently we
On Tue, 29 May 2012, Christian Bruel wrote:
So I tested the following semantics, with the ones that you pointed.
Thanks for the testing. The patch is OK with the ChangeLog conflict
markers removed (and the dates on log entries updated to the date of
commit).
--
Joseph S. Myers
On Tue, 29 May 2012, Meador Inge wrote:
2012-05-29 Meador Inge mead...@codesourcery.com
* c-decl.c (c_push_function_context): Always create a new language
function.
(c_pop_function_context): Clear the language function created in
c_push_function_context.
Thanks,
On Tue, 29 May 2012, Jim Meyering wrote:
Running the following command spots over 1500 typos, and suggests fixes:
$ git ls-files|misspellings -f -|grep -v '^ERROR:'|perl -pe \
's/^(.*?)\[(\d+)\]: (\w+) - (.*?)$/sed -i '\''${2}s!$3!$4!'\'' $1/' k
This command has some bugs in how it
I've applied these fixes for three miscellaneous directories with a single
typo each to fix (gcc/c-family, libmudflap, lto-plugin).
Index: libmudflap/mf-impl.h
===
--- libmudflap/mf-impl.h(revision 187979)
+++
Alan,
I think the following patch
--- ../_gcc_clean/gcc/testsuite/gcc.target/powerpc/powerpc.exp 2012-05-02
14:25:40.0 +0200
+++ ../work/gcc/testsuite/gcc.target/powerpc/powerpc.exp2012-05-29
21:14:48.0 +0200
@@ -47,4 +47,5 @@ set-torture-options $SAVRES_TEST_OPTS
Hi!
On Tue, 22 May 2012 11:38:55 +0200, Tristan Gingold ging...@adacore.com wrote:
On May 22, 2012, at 11:20 AM, Tobias Burnus wrote:
an --enable-maintainers-build fails here with:
configure.ac:99: error: possibly undefined macro: gcc_AC_FUNC_MMAP_BLACKLIST
If this token and others
Attached patch applied then.
2012-05-29 François Dumont fdum...@gcc.gnu.org
* include/bits/stl_tempbuf.h (__uninitialized_construct_buf)
(__uninitialized_construct_buf_dispatch::__ucr): Fix to work
with iterator returning rvalue.
* testsuite/25_algorithms/stable_sort/3.cc: New.
I have just committed the patch below to add myself to the
write-after-approval section of the MAINTAINERS file.
Thanks,
Edmar
ChangeLog:
2012-05-29 Edmar Wienskoskied...@freescale.com
* MAINTAINERS (Write After Approval): Add myself.
Index: MAINTAINERS
From: Hans-Peter Nilsson h...@axis.com
Date: Wed, 16 May 2012 02:24:02 +0200
Ping...
An old patch I finally came around to submit.
Verified that the DVI and info output looks ok.
Ok to commit with inherent relicensing and whatever?
gcc:
* doc/tm.texi.in (Addressing Modes)
Joseph S. Myers wrote:
On Tue, 29 May 2012, Jim Meyering wrote:
Running the following command spots over 1500 typos, and suggests fixes:
$ git ls-files|misspellings -f -|grep -v '^ERROR:'|perl -pe \
's/^(.*?)\[(\d+)\]: (\w+) - (.*?)$/sed -i '\''${2}s!$3!$4!'\'' $1/' k
This command
As per libstdc++/52700, this fixes the configure bits for
libgfortran/libssp/libquadmath. With these fixes, I believe all the
libs are safe for --enable-symvers=gnu* variants.
Super simple patches...
I intend to put this on the 4.7 branch as well.
tested x86/linux
-benjamin
2012-05-29
On Wed, May 16, 2012 at 1:03 PM, Sterling Augustine
saugust...@google.com wrote:
This patch adds new flags and defines such that the C++ decl pretty printer
prints both canonical dwarf names for decls without perturbing normal error
message output.
It addresses the issues with the earlier
.. Ah! I think I have a much better tentative fix, which, as it happens,
also leads to a diagnostics quite close to that produced by the EDG
front-end.
The idea is checking for the offending case and not handling it in any
special way besides turning need_new = 1, as if for a normal new
Yes, I understand that's broken, but there are no consumers at this
point that make any use of that offset. Would it be acceptable if we
just put 0 there? (Given that I expect .debug_pub* to go away soon, I
don't think it's worth the trouble of filling in the offset with
anything more
On May 29, 2012, at 5:49 AM, Jim Meyering wrote:
Running the following command spots over 1500 typos, and suggests fixes:
$ git ls-files|misspellings -f -|grep -v '^ERROR:'|perl -pe \
's/^(.*?)\[(\d+)\]: (\w+) - (.*?)$/sed -i '\''${2}s!$3!$4!'\'' $1/' k
Hum, maybe a make rule (only run by
On 5/24/12, Gabriel Dos Reis g...@integrable-solutions.net wrote:
On May 24, 2012 Lawrence Crowl cr...@google.com wrote:
Add a type-safe hash table, typed_htab. Uses of this table
replace uses of libiberty's htab_t. The benefits include less
boiler-plate code, full type safety, and
On 5/28/12, Eric Botcazou ebotca...@adacore.com wrote:
My main concern is that the precise collector we have in place
now requires substantial care to use. It is supported by a tool
that does not really understand C, let alone C++. We avoid the
problems when annotations are unnecessary.
On 5/29/12, Richard Guenther richard.guent...@gmail.com wrote:
On May 25, 2012 Mike Stump mikest...@comcast.net wrote:
On May 25, 2012, at 10:50 AM, Lawrence Crowl wrote:
Diego and I looked long and hard at this issue. It all came
down to a sequence of problems. First, libstdc++ isn't
Hi,
On Sun, 27 May 2012, Gabriel Dos Reis wrote:
people actually working on it and used to that style. We don't want to
have a mixture of several different styles in the compiler. I (and I
expect many others) don't want anyone working around the latter by going
over the whole source
On Thu, May 3, 2012 at 3:39 PM, H.J. Lu hjl.to...@gmail.com wrote:
On Sat, Nov 12, 2011 at 4:42 PM, Jason Merrill ja...@redhat.com wrote:
Now that we have a way of explicitly marking a variable as dead, we can use
that to indicate the end of a temporary's lifetime by adding it as a cleanup
for
On 29 May 2012 18:30, Richard Henderson r...@redhat.com wrote:
On 05/26/2012 01:27 AM, Ramana Radhakrishnan wrote:
- for (i = 0; i nelt; i += diff)
+ for (i = 0; i nelt ; i += (diff + 1))
for (j = 0; j= diff; j += 1)
- if (d-perm[i + j] != i + diff - j)
- return false;
From: Hans-Peter Nilsson h...@axis.com
Date: Wed, 23 May 2012 06:41:58 +0200
From: Hans-Peter Nilsson h...@axis.com
Date: Wed, 16 May 2012 08:24:41 +0200
From: Hans-Peter Nilsson h...@axis.com
Date: Wed, 9 May 2012 08:02:25 +0200
Ping. I missed the PR number decoration on the
Hi!
GEN_INT (GET_MODE_MASK (mode)) isn't canonical CONST_INT for mode,
because it is zero extended rather than sign-extended for GET_MODE_BITSIZE
(mode) HOST_BITS_PER_WIDE_INT modes.
Fixed thusly, bootstrapped/regtested on x86_64-linux and i686-linux,
ok for trunk?
2012-05-30 Jakub Jelinek
80 matches
Mail list logo