On Thu, Mar 27, 2014 at 09:56:22AM +0400, Yury Gribov wrote:
> Now what bothers me is zero feedback from Dejagnu folks...
Come on, it's only been a couple of days. :-)
Ben
signature.asc
Description: Digital signature
> But if somebody is willing to maintain it for
> cross-testing
Yup, we'll do this.
Now what bothers me is zero feedback from Dejagnu folks...
-Y
> My earlier patch for 58678 caused this problem: even if we aren't
> going to refer to the dtor thunks from the vtable, we need to emit
> them in case other translation units use them.
>
> I'm adding xfails to two testcases: devirt-21.C and devirt-23.C.
> The fails aren't a new bug; we are no lon
On Wed, Mar 26, 2014 at 09:43:08PM +, Maciej W. Rozycki wrote:
> Alan,
>
> On Tue, 17 Dec 2013, Alan Modra wrote:
>
> > On Tue, Dec 17, 2013 at 01:14:23PM +0100, Bernd Edlinger wrote:
> > > the reason for this is overwriting GMPINC for the auto-build generation,
> > > because
> > > many test
PR c++/60573
* parser.c (synthesize_implicit_template_parm): Use cp_binding_level::
class_shadowed rather than TYPE_BEING_DEFINED as the predicate for
unwinding to class-defining scope to handle the erroneous definition of
a generic function of an arbitrarily
On Wed, Mar 26, 2014 at 3:50 PM, Michael Meissner
wrote:
> This patch adds support for adding a builtin to generate the vbpermq
> instruction on ISA 2.07. This instruction takes a vector in the Altivec
> register set, and returns a 64-bit value in the upper part of the register,
> and
> 0 in the
On Wed, Mar 26, 2014 at 4:55 PM, Tobias Burnus wrote:
>
> Rainer Orth wrote:
>>
>> Tobias Burnus writes:
>>>
>>> H.J. Lu wrote:
xgcc: error: libcilkrts.spec: No such file or directory
>>>
>>> Hmm, I really wonder why it fails for you while it works for me:
>>
>> Do you happen to have th
Rainer Orth wrote:
Tobias Burnus writes:
H.J. Lu wrote:
xgcc: error: libcilkrts.spec: No such file or directory
Hmm, I really wonder why it fails for you while it works for me:
Do you happen to have the same/a recent version installed at the same
prefix your build under test is configured f
is cgraph_init_gid_map called after linking?
David
On Wed, Mar 26, 2014 at 3:54 PM, Dehao Chen wrote:
> Patch updated, passed performance tests.
>
> Dehao
>
> On Tue, Mar 25, 2014 at 4:03 PM, Xinliang David Li wrote:
>> Add comment to the new function. init_node_map is better invoked after
>> t
Patch updated, passed performance tests.
Dehao
On Tue, Mar 25, 2014 at 4:03 PM, Xinliang David Li wrote:
> Add comment to the new function. init_node_map is better invoked after
> the link step to avoid creating entries with for dead nodes.
>
> Ok if large perf testing is fine.
>
> David
>
> On
Tobias Burnus writes:
> H.J. Lu wrote:
>> All cilk-plus link tests failed with
>> xgcc: error: libcilkrts.spec: No such file or directory
>> Don't you need to add -B/libcilkrts for this?
>
> Hmm, I really wonder why it fails for you while it works for me:
[...]
> I cross-checked: I ha
On Wed, Mar 26, 2014 at 3:35 PM, Tobias Burnus wrote:
> H.J. Lu wrote:
>>
>> All cilk-plus link tests failed with
>> xgcc: error: libcilkrts.spec: No such file or directory
>> Don't you need to add -B/libcilkrts for this?
>
>
> Hmm, I really wonder why it fails for you while it works f
H.J. Lu wrote:
All cilk-plus link tests failed with
xgcc: error: libcilkrts.spec: No such file or directory
Don't you need to add -B/libcilkrts for this?
Hmm, I really wonder why it fails for you while it works for me:
Running
/home/tob/projects/gcc/gcc/testsuite/gcc.dg/cilk-plus/
The first patch silences hundreds of harmless warnings from Xcode 3.2.x
due to r205679 and of the kind
warning: DWARFDebugInfoEntry::AppendDependants() -- check on this item
TAG_namelist_item: attr = AT_namelist_item form = FORM_ref4
The following three patches fix pr54083 for darwin 8 and 9.
> > Bootstrapped/regtested x86_64-linux, comitted.
>
> Not with Ada apparently, resulting in
>
> === acats tests ===
> FAIL: c34007d
> FAIL: c34007g
> FAIL: c34007s
> FAIL: c37213j
> FAIL: c37213k
> FAIL: c37213l
> FAIL: ce2201g
> FAIL: cxa5a03
> FAIL: cxa5a04
>
On Mon, Mar 10, 2014 at 11:42 PM, Tobias Burnus wrote:
> When using Cilk Plus (-fcilkplus), it makes sense to automatically link the
> run-time library (-lcilkrts).
>
> This patch mimics libgomp by adding a .spec file; I am not 100% sure whether
> the .spec file is needed, but the pthread tests in
Richard Henderson writes:
> On 03/26/2014 12:40 PM, Jakub Jelinek wrote:
>> On Wed, Mar 26, 2014 at 01:32:44PM -0600, Jeff Law wrote:
>>> On 03/26/14 12:28, Jakub Jelinek wrote:
(mult:SI (const_int 0) (const_int 4)) is IMHO far from being canonical.
And, I'd say it is likely other target
Alan,
On Tue, 17 Dec 2013, Alan Modra wrote:
> On Tue, Dec 17, 2013 at 01:14:23PM +0100, Bernd Edlinger wrote:
> > the reason for this is overwriting GMPINC for the auto-build generation,
> > because
> > many test scripts include which fails now completely (it is not
> > installed,
> > I have
On Wed, 2014-03-26 at 22:19 +0100, Jakub Jelinek wrote:
> Hi!
>
> The sys_futex0 caller expects return values as returned by raw syscalls,
> i.e. value >= 0 success, negative value are errors -errorval.
> But, the syscall function returns value >= 0 on success, and -1 on error,
> with errno set to
On Wed, Mar 26, 2014 at 10:27:37PM +0100, Marc Glisse wrote:
> On Tue, 25 Mar 2014, Marek Polacek wrote:
>
> >This is a temporary fix for UB in IRA, where ubsan complains because
> >there's signed iteger overflow in the multiplication. To shut this
> >error up, we can perform the multiplication i
> Bootstrapped/regtested x86_64-linux, comitted.
Not with Ada apparently, resulting in
=== acats tests ===
FAIL: c34007d
FAIL: c34007g
FAIL: c34007s
FAIL: c37213j
FAIL: c37213k
FAIL: c37213l
FAIL: ce2201g
FAIL: cxa5a03
FAIL: cxa5a04
FAIL: cxa5a06
FAIL: cxg20
On Tue, 25 Mar 2014, Marek Polacek wrote:
This is a temporary fix for UB in IRA, where ubsan complains because
there's signed iteger overflow in the multiplication. To shut this
error up, we can perform the multiplication in unsigned and only then
cast the result of the multiplication to int.
Various Texinfo manuals have had their indexes given node names other than
"Index" to avoid conflicts with index.html when HTML output is generated
and then used on a case-insensitive filesystem. I've applied this patch
to make such a renaming in the libitm manual, using the name Library Index
Hi!
The sys_futex0 caller expects return values as returned by raw syscalls,
i.e. value >= 0 success, negative value are errors -errorval.
But, the syscall function returns value >= 0 on success, and -1 on error,
with errno set to errorval. This means if e.g. futex syscall fails with
EAGAIN, and
On Wed, Mar 26, 2014 at 01:50:38PM -0700, H.J. Lu wrote:
> All plug tests failed with
There are some testsuite issues with both both bootstrap-ubsan and
bootstrap-asan, e.g. host libiberty tests fail, and host libbacktrace
btest fails too (in both cases because the host libiberty or libbacktrace
i
On Wed, Mar 26, 2014 at 9:00 AM, Jakub Jelinek wrote:
> On Tue, Mar 25, 2014 at 05:24:40PM -0700, H.J. Lu wrote:
>> Doesn't work:
>>
>> libtool: link:
>> /export/build/gnu/gcc-asan/build-x86_64-linux/./prev-gcc/xgcc
>> -B/export/build/gnu/gcc-asan/build-x86_64-linux/./prev-gcc/
>> -B/usr/local/x86
Sorry about the delay on this. As GCC is in pre-release and it would
be bad for 4.9 to pass options that a released binutils doesn't
support, I hesitate to approve this at this time. It looks OK, it's
just the timing is bad. Please remind us after GCC is back in stage1.
I would also like to se
On 2014-03-25 15:48, Jason Merrill wrote:
I think we need some way to designate a scope that actually
corresponds to a class-specifier.
Agreed. I'll look into it.
Adam
> I have a minor nit: can you add a @code{} around the second VOLATILE?
What about the COMMON in "variables in COMMON blocks since revision 4.3."?
Dominique
On Wed, Mar 26, 2014 at 12:26:02PM -0700, Steve Kargl wrote:
> On Wed, Mar 26, 2014 at 06:56:15PM +0100, Dominique Dhumieres wrote:
> > Updated patch. OK to commit?
> >
> > Dominique
> >
> > Index: gcc/fortran/ChangeLog
> > ===
> > -
On 03/26/2014 12:40 PM, Jakub Jelinek wrote:
> On Wed, Mar 26, 2014 at 01:32:44PM -0600, Jeff Law wrote:
>> On 03/26/14 12:28, Jakub Jelinek wrote:
>>> (mult:SI (const_int 0) (const_int 4)) is IMHO far from being canonical.
>>> And, I'd say it is likely other target legitimization hooks would also
This patch adds support for adding a builtin to generate the vbpermq
instruction on ISA 2.07. This instruction takes a vector in the Altivec
register set, and returns a 64-bit value in the upper part of the register, and
0 in the lower part of the register.
The output is explicitly a vector, sinc
On 24/01/14 20:08 +, Jonathan Wakely wrote:
The debug-mode container base classes need copy constructors that
zero-init their members rather than copy them from the source.
_Safe_unordered_container_base also needs its move constructor to be
non-throwing to fix some FAILs for the PR 55043 te
On Mar 26, 2014, at 11:33 AM, Jeff Law wrote:
> On 03/26/14 12:28, Jakub Jelinek wrote:
>> On Wed, Mar 26, 2014 at 12:17:43PM -0600, Jeff Law wrote:
>>> On 03/26/14 12:12, Jakub Jelinek wrote:
On Wed, Mar 26, 2014 at 11:02:48AM -0600, Jeff Law wrote:
> Bootstrapped and regression tested o
On Wed, Mar 26, 2014 at 01:32:44PM -0600, Jeff Law wrote:
> On 03/26/14 12:28, Jakub Jelinek wrote:
> >(mult:SI (const_int 0) (const_int 4)) is IMHO far from being canonical.
> >And, I'd say it is likely other target legitimization hooks would also try
> >to simplify it similarly.
> >simplify_gen_b
Steve Kargl wrote:
On Wed, Mar 26, 2014 at 06:56:15PM +0100, Dominique Dhumieres wrote:
Updated patch. OK to commit?
OK.
I have a minor nit: can you add a @code{} around the second VOLATILE?
+Some Fortran compilers, including @command{g77}, let the user declare
+@code{COMMON} with the @code
On 03/26/14 12:28, Jakub Jelinek wrote:
(mult:SI (const_int 0) (const_int 4)) is IMHO far from being canonical.
And, I'd say it is likely other target legitimization hooks would also try
to simplify it similarly.
simplify_gen_binary is used in several other places during expansion,
so I don't see
On Wed, Mar 26, 2014 at 06:56:15PM +0100, Dominique Dhumieres wrote:
> Updated patch. OK to commit?
>
> Dominique
>
> Index: gcc/fortran/ChangeLog
> ===
> --- gcc/fortran/ChangeLog (revision 208846)
> +++ gcc/fortran/ChangeLog
On March 26, 2014 5:06:20 PM CET, Jakub Jelinek wrote:
>Hi!
>
>Honza's r208831 change apparently broke bootstrap-ubsan.
>The problem is that it now creates __builtin_unreachable with
>gimple_location where LOCATION_LOCUS is UNKNOWN_LOCATION (comes from
>fnsplit, any ideas what gimple_location to u
On March 26, 2014 4:57:28 PM CET, Jakub Jelinek wrote:
>Hi!
>
>Doing low = -low; on signed HOST_WIDE_INT results in undefined behavior
>if the low HWI is LONG_LONG_MIN. Fixed thusly, ok for trunk?
OK.
Thanks,
Richard.
>2014-03-26 Jakub Jelinek
>
> PR other/59545
> * real.c (real_
On March 26, 2014 4:51:58 PM CET, Ian Lance Taylor wrote:
>On Wed, Mar 26, 2014 at 8:38 AM, Richard Biener
>wrote:
>>
>> - got = read (descriptor, buffer, size);
>> - if (got < 0)
>> + do
>> {
>> - *errmsg = "read";
>> - *err = errno;
>> - return 0;
>> + got = read (de
On 2014-03-26 15:17, Jason Merrill wrote:
I meant
struct A
{
struct X
{
struct B
{
void foo(auto);
};
void B::foo(auto) {} // { dg-error "cannot define" }
};
};
Here we push both A and X for the declarator. When we get to the
pushed X, we se
On irc Tom Tromey pointed out that the patch generates duplicate base
type DIEs for enums with the same underlying base type. This is because
it was calling base_type_die () and adding the DW_AT_type by hand
instead of calling add_type_attribute () to add it to the enumeration
DIE. The only differe
On Wed, Mar 26, 2014 at 12:17:43PM -0600, Jeff Law wrote:
> On 03/26/14 12:12, Jakub Jelinek wrote:
> >On Wed, Mar 26, 2014 at 11:02:48AM -0600, Jeff Law wrote:
> >>Bootstrapped and regression tested on x86_64-unknown-linux-gnu.
> >>Verified it fixes the original and reduced testcase.
> >
> >Note,
On 03/26/14 12:28, Jakub Jelinek wrote:
On Wed, Mar 26, 2014 at 12:17:43PM -0600, Jeff Law wrote:
On 03/26/14 12:12, Jakub Jelinek wrote:
On Wed, Mar 26, 2014 at 11:02:48AM -0600, Jeff Law wrote:
Bootstrapped and regression tested on x86_64-unknown-linux-gnu.
Verified it fixes the original and
On 03/26/14 12:12, Jakub Jelinek wrote:
On Wed, Mar 26, 2014 at 11:02:48AM -0600, Jeff Law wrote:
Bootstrapped and regression tested on x86_64-unknown-linux-gnu.
Verified it fixes the original and reduced testcase.
Note, the testcase is missing from your patch.
But I'd question if this is the
On Wed, Mar 26, 2014 at 10:42:12AM -0700, Mike Stump wrote:
> On Mar 25, 2014, at 7:31 AM, Yury Gribov wrote:
> > On 03/24/2014 01:46 PM, Maxim Ostapenko wrote:
> >>
> >> On 03/24/2014 12:28 PM, Maxim Ostapenko wrote:
> >>>
> >>> When porting Lsan on arm, I ran into problem with testing, because
On Wed, Mar 26, 2014 at 11:02:48AM -0600, Jeff Law wrote:
> Bootstrapped and regression tested on x86_64-unknown-linux-gnu.
> Verified it fixes the original and reduced testcase.
Note, the testcase is missing from your patch.
But I'd question if this is the right place to canonicalize it.
The non
Updated patch. OK to commit?
Dominique
Index: gcc/fortran/ChangeLog
===
--- gcc/fortran/ChangeLog (revision 208846)
+++ gcc/fortran/ChangeLog (working copy)
@@ -1,3 +1,9 @@
+2014-03-26 Dominique d'Humieres
+
+ PR f
On Mar 26, 2014, at 10:35 AM, Dominique Dhumieres wrote:
> Adding myself as Write After Approval.
Welcome. :-)
On Mar 25, 2014, at 7:31 AM, Yury Gribov wrote:
> On 03/24/2014 01:46 PM, Maxim Ostapenko wrote:
>>
>> On 03/24/2014 12:28 PM, Maxim Ostapenko wrote:
>>>
>>> When porting Lsan on arm, I ran into problem with testing, because
>>> almost all lsan tests require environment variables forwarding for
Adding myself as Write After Approval.
Dominique
Index: ChangeLog
===
--- ChangeLog (revision 208845)
+++ ChangeLog (working copy)
@@ -1,3 +1,7 @@
+2014-03-26 Dominique d'Humieres
+
+ * MAINTAINERS (Write After Approval):
The x86 backend can generate non-canonical RTL when it simplifies
address expressions.
In particular addresses which have the form
(plus (mult) (A) (B) (label_ref))
If the multiplication can be simplified to a constant, the x86 backend
will end up generating
(plus (constant) (label_ref))
On 03/26/2014 12:49 PM, Jason Merrill wrote:
But it seems odd to me that it
isn't happening on path through a non-virtual call to ~MultiTermDocs, so
I'm filing a bug about that.
60674.
Jason
My earlier patch for 58678 caused this problem: even if we aren't going
to refer to the dtor thunks from the vtable, we need to emit them in
case other translation units use them.
I'm adding xfails to two testcases: devirt-21.C and devirt-23.C. The
fails aren't a new bug; we are no longer see
> "Julian" == Julian Brown writes:
Julian> include/
Julian> * dwarf2.h (enum dwarf_calling_convention): Add
Julian> DW_CC_GNU_thiscall_i386.
We've been trying to ensure that all GNU DWARF extensions are
documented. In the past we had problems where an extension was added
and the
Hi!
Honza's r208831 change apparently broke bootstrap-ubsan.
The problem is that it now creates __builtin_unreachable with
gimple_location where LOCATION_LOCUS is UNKNOWN_LOCATION (comes from
fnsplit, any ideas what gimple_location to use for the calls if any?)
and with -fsanitize=undefined we fol
Hi!
Doing low = -low; on signed HOST_WIDE_INT results in undefined behavior
if the low HWI is LONG_LONG_MIN. Fixed thusly, ok for trunk?
2014-03-26 Jakub Jelinek
PR other/59545
* real.c (real_to_integer2): Change type of low to UHWI.
--- gcc/real.c.jj 2014-01-03 11:40:
On Wed, Mar 26, 2014 at 8:38 AM, Richard Biener wrote:
>
> - got = read (descriptor, buffer, size);
> - if (got < 0)
> + do
> {
> - *errmsg = "read";
> - *err = errno;
> - return 0;
> + got = read (descriptor, buffer, size);
> + if (got < 0
> + && errno !=
The following tries to cure random plugin claim failures I see when
building in KVM (still to be verified that this is the actual problem).
It seems that lto-plugin and simple-object fail to handle short reads
and read returning EINTR which the following fixes by teaching that
to simple_object_int
On 03/25/2014 03:48 PM, Adam Butcher wrote:
I don't follow. Are you suggesting a case like the following?
struct A
{
struct X
{
struct B
{
void foo(auto);
};
};
void X::B::foo(auto) {} // { dg-error "cannot define" }
};
I meant
s
Hi,
In gcc/config/linux-android.h we have "builtin_define ("__ANDROID__");"
So "ANDROID" as in libcilkrts now is not the correct macro to check.
Bootstrapped and passed cilk testsuite on x86_64-unknown-linux-gnu.
diff --git a/libcilkrts/ChangeLog b/libcilkrts/ChangeLog
index eb0d6ec..65efef0 100
This patch fixes an issue where only some rotate immediate operations
are merged with logical operations during combine. The problem is due
to canonicalization. The architecture only has a rotate-right
operation, so rotate-left has to be converted into rotate-right. To
avoid fighting the mid-end
Hi,
This is the GCC part of my fix for PR15559, the GDB part of which was
posted here:
https://sourceware.org/ml/gdb-patches/2014-03/msg00610.html
Cross-tested (gcc/g++/libstdc++) from Linux to Mingw32: results are the
same, apart from (what I presume to be) noise in guality.exp/pr55665.C.
Acco
On 26 March 2014 12:35, Joern Rennecke wrote:
> bootstrapped on i686-pc-linux-gnu, regtest in progress.
Passed now.
On 26 March 2014 08:15, Eric Botcazou wrote:
>> As described in the PR, this patch fixes a wrong-code bug by making the
>> order of emitted mode switching instructions more consistet & predictable.
>
> I don't understand this change (but I'm not a specialist of mode switching):
> currently the mod
xes a miscompile of the 433.milc SpecCPU testcase.
Bye,
-Andreas-
2014-03-26 Andreas Krebbel
* config/s390/s390.c (s390_can_use_return_insn): Check for
call-saved FPRs on 31 bit.
2014-03-26 Andreas Krebbel
* gcc.target/s390/20140326-1.c: New testcase.
diff --git a/gcc/c
Christian Bruel wrote:
> This patch inlines builtin_memset whose size is a constant 128 < size <
> 15. Small sizes are better unrolled with mov_insn sequences. Big size
> (or non constants) are better handled with a libc implementation that
> does cache line aligned copying and unrolling or prefet
Christian Bruel wrote:
> This patches adds a few instructions to the inlined builtin_strlen to
> unroll the remaining bytes for word-at-a-time loop. This enables to have
> 2 distinct execution paths (no fall-thru in the byte-at-a-time loop),
> allowing block alignment assignation. This partially i
On Wed, 2014-03-26 at 00:25 +0100, Eric Botcazou wrote:
> > Should I file a bug to get feedback on this issue? I know Ada is not the
> > most prioritized language for gcc, but anyway. The current
> > implementation is not POSIX-compliant.
>
> Can you post a complete patch doing the reversion? Bre
This avoids the (int) cast for switches on enums (which are
appearantly using an unsigned type in C++). It does so by
enhancing the code in simplify_gimple_switch to see if the
case value checks are preserved with a sign-change.
Bootstrap / regtest running on x86_64-unkown-linux-gnu, queued for
On 03/26/2014 11:22 AM, Christian Bruel wrote:
> Hello,
>
> This patch inlines builtin_memset whose size is a constant 128 < size <
> 15. Small sizes are better unrolled with mov_insn sequences. Big size
> (or non constants) are better handled with a libc implementation that
> does cache line alig
Hello,
This patch inlines builtin_memset whose size is a constant 128 < size <
15. Small sizes are better unrolled with mov_insn sequences. Big size
(or non constants) are better handled with a libc implementation that
does cache line aligned copying and unrolling or prefetching.
No new regressio
On 26 March 2014 15:45, Jakub Jelinek wrote:
> On Wed, Mar 26, 2014 at 03:30:44PM +0800, Zhenqiang Chen wrote:
>> Agree. CSE should never modify asm insns to drop some of the outputs.
>
> So the right fix is top prevent this from happening, not papering over about
> it.
>>
>> But in this case, CSE
On Tue, 25 Mar 2014, Jakub Jelinek wrote:
> On Tue, Mar 25, 2014 at 10:15:37AM +0100, Richard Biener wrote:
> > On Tue, 25 Mar 2014, Richard Biener wrote:
> >
> > > On Tue, 25 Mar 2014, Jakub Jelinek wrote:
> > >
> > > > Hi!
> > > >
> > > > While working on previous patch, I've noticed a severe
Il 26/03/2014 00:28, Tobias Burnus ha scritto:
Paolo Bonzini wrote:
Il 11/03/2014 07:42, Tobias Burnus ha scritto:
+XPCFLAGS=""
+CFLAGS="$CFLAGS -pthread"
+AC_LINK_IFELSE(
...
+ [XPCFLAGS=" -Wc,-pthread"],
XPCFLAGS is dead, I think?
Yes - contrary to libgomp, from which I have taken that cod
Il 25/03/2014 20:24, Jakub Jelinek ha scritto:
Hi!
This patch fixes a problem where build of host fixincludes fails
with --with-build-config=bootstrap-ubsan (and bootstrap-asan).
The problem is that fixincludes is linked against host libiberty
that is bootstrapped, but fixincludes is not bootstr
On Tue, 25 Mar 2014, Jakub Jelinek wrote:
> On Tue, Mar 25, 2014 at 10:43:02AM +0100, Richard Biener wrote:
> > Yes, all transforms in fold-const would be invalid if the result doesn't
> > behave in the same way wrt overflow. Thus you really should instrument
> > ABS_EXPR - you can treat it as A
On Fri, 21 Mar 2014, Martin Jambor wrote:
> Hi,
>
> On Fri, Mar 21, 2014 at 09:41:24AM +0100, Richard Biener wrote:
> > On Thu, 20 Mar 2014, Martin Jambor wrote:
> >
> > > Hi,
> > >
> > > in PR 60419 we end up with a call graph node for a thunk that has no
> > > callee because symtab_remove_unr
Jan Hubicka writes:
> Index: testsuite/g++.dg/torture/pr60315.C
> ===
> --- testsuite/g++.dg/torture/pr60315.C(revision 0)
> +++ testsuite/g++.dg/torture/pr60315.C(revision 0)
> @@ -0,0 +1,32 @@
> +// { dg-do compile
On Tue, Mar 25, 2014 at 05:24:40PM -0700, H.J. Lu wrote:
> Doesn't work:
>
> libtool: link:
> /export/build/gnu/gcc-asan/build-x86_64-linux/./prev-gcc/xgcc
> -B/export/build/gnu/gcc-asan/build-x86_64-linux/./prev-gcc/
> -B/usr/local/x86_64-unknown-linux-gnu/bin/
> -B/usr/local/x86_64-unknown-linux
> As described in the PR, this patch fixes a wrong-code bug by making the
> order of emitted mode switching instructions more consistet & predictable.
I don't understand this change (but I'm not a specialist of mode switching):
currently the mode setting sequence is always emitted before the insn
On 26 March 2014 03:17:11 Jan Hubicka wrote:
Hi,
Just 2 nits, cannot comment on the patch itself.
s/clonning/cloning/g as usual :)
And the content of the testcase is duplicated.
Thanks,
this patch fixes compile time issue in the testcase that is caused by fact
that the inliner
is repeatedl
Hello,
This patches adds a few instructions to the inlined builtin_strlen to
unroll the remaining bytes for word-at-a-time loop. This enables to have
2 distinct execution paths (no fall-thru in the byte-at-a-time loop),
allowing block alignment assignation. This partially improves the
problem repo
On Wed, Mar 26, 2014 at 03:30:44PM +0800, Zhenqiang Chen wrote:
> Agree. CSE should never modify asm insns to drop some of the outputs.
So the right fix is top prevent this from happening, not papering over about
it.
>
> But in this case, CSE does not drop any of the outputs. It just takes
> the
On 26 March 2014 15:00, Jakub Jelinek wrote:
> On Wed, Mar 26, 2014 at 02:16:16PM +0800, Zhenqiang Chen wrote:
>> The patch checks the number of the expected operands in
>> ASM_OPERANDS_TEMPLATE with the same logic as it in output_asm_insn to
>> make sure the ASM_OPERANDS are legal.
>>
>> Bootstra
Thomas Koenig wrote:
Hello world,
print *, 'Hello Thomas'
the attached patch fixes the regression (after some thought
of what might still be optimized, which isn't much :-)
Regression-tested. OK for trunk?
Looks good to me. (I wonder whether one should also include Mikael's
test case, bu
On Wed, Mar 26, 2014 at 02:16:16PM +0800, Zhenqiang Chen wrote:
> The patch checks the number of the expected operands in
> ASM_OPERANDS_TEMPLATE with the same logic as it in output_asm_insn to
> make sure the ASM_OPERANDS are legal.
>
> Bootstrap and no make check regression on X86-64 and ARM chr
88 matches
Mail list logo
