On 05/08/2014 06:02, Roman Gareev wrote:
I've attached the patch, which sets the separate option for all
dimensions. Is it fine for trunk?
LGTM.
Tobias
Jonathan Wakely jwak...@redhat.com writes:
One part of the patch I wasn't sure about was this, where 'mgr' is a
function pointer:
func = gdb.block_for_pc(int(mgr.cast(gdb.lookup_type('intptr_t'
Is there a better way to get a pc from the function pointer?
I tried simply int(mgr) but it
On Fri, Aug 1, 2014 at 2:21 AM, pins...@gmail.com wrote:
On Jun 6, 2014, at 1:50 AM, James Greenhalgh james.greenha...@arm.com
wrote:
Hi,
The move_by_pieces infrastructure performs a copy by repeatedly trying
the largest safe copy it can make. So for a 15-byte copy we might see:
On Mon, 4 Aug 2014, Jakub Jelinek wrote:
Hi!
I've tried to bootstrap with r213515 because later revisions broke because
of the hash_map and Ada changes, but unfortunately even that revision
failus to bootstrap, the new -Wreturn-local-addr warning rightfully
warns about get_ivts_expr
When removing the MEM_ATTR unification hash I didn't grep for
MEM_ATTR equality comares which now causes PR61672, missed
optimizations at least in PRE. The following fixes that by
exporting and using mem_attrs_eq_p.
Bootstrapped and tested on x86_64-unknown-linux-gnu - as this is
a regression
On Mon, Aug 04, 2014 at 11:51:34AM -0500, Edmar wrote:
Committed on trunk, revision 213596
Committed on 4.9 branch, revision 213597
Note the ChangeLog entry was grossly misformated.
I've fixed it up in gcc/ChangeLog on the trunk, but not on the branch
nor in libgcc. There should be no space
On Mon, Aug 4, 2014 at 3:04 PM, tsaund...@mozilla.com wrote:
From: Trevor Saunders tsaund...@mozilla.com
Hi,
It used to be that edge_var_maps held pointers to embedded vectors, but
now it holds vectors. This means that now instead of copying the
address of the embedded vector from the
On Mon, Aug 4, 2014 at 10:37 PM, Bernd Schmidt ber...@codesourcery.com wrote:
Some code I added for the ptx backend triggered tree-checking failures for
gimple created in tree-nested: we can end up taking the address of an
SSA_NAME, which seems invalid. The problem is that code in tree-nested
On Mon, Aug 4, 2014 at 2:27 PM, Kyrill Tkachov kyrylo.tkac...@arm.com wrote:
Hi all,
Following up on Josephs' comments on PR 61876 this patch guards the rint +
cast - lrint transformation on -fno-math-errno.
Bootstrapped and tested on aarch64-linux and x86.
Ok for trunk?
Ok.
Thanks,
On 04/08/14 11:13, James Greenhalgh wrote:
This patch removes the aarch64_simd_builtin_type enum. This is only really
used for indexing in to an array of strings and an array of machine_mode.
We don't need that. Given all the macro pasting we presently do in this
file, just add another.
Hi,
For some large constants, ARM will split them during expanding, which
makes impossible to hoist them out the loop or shared by different
references (refer the test case in the patch).
The patch keeps some constants in registers. If the constant can not
be optimized, the cprop and combine
On Mon, Aug 4, 2014 at 4:13 PM, Prathamesh Kulkarni
bilbotheelffri...@gmail.com wrote:
Added patterns in associate_pointerplus and associate_pointerplus_diff.
* genmatch.c (capture_max): Change value to 6.
(match-plusminus.pd): Add new patterns.
[gcc/testsuite/gcc.dg/tree-ssa]
*
On 04/08/14 13:22, Kyrill Tkachov wrote:
Hi all,
While working on another patch and looking at the rbit and clz patterns,
I noticed that in aarch32 the relevant patterns are not adjusted for
-mrestrict-it.
A program like:
int
foo (int a, int b)
{
if (a + 5 == b)
return
Hi all,
The recently added -ftrapv test FAILS on bare metal targets because it
requires the usage of fork.
It does have a dg-require-fork guard but apparently that function takes
an unused argument and if you don't give it that, DejaGNU will happily
ignore the directive and not bother
On Tue, Aug 5, 2014 at 11:47 AM, Kyrill Tkachov kyrylo.tkac...@arm.com wrote:
Hi all,
The recently added -ftrapv test FAILS on bare metal targets because it
requires the usage of fork.
It does have a dg-require-fork guard but apparently that function takes an
unused argument and if you don't
Converted from my plain-text notes I did for writing the Cauldron
presntation slides.
Built and installed.
Richard.
2014-08-05 Richard Biener rguent...@suse.de
* doc/match-and-simplify.texi: New file.
* doc/gccint.texi: Include match-and-simplify.texi.
* Makefile.in
On 04/08/14 14:32, Alan Lawrence wrote:
At the moment, for two-element vectors, __builtin_shuffle (vector, (mask) {C,
C}) for identical constants C outputs a zip (with both argument vectors the
same) rather than a dup. Dup is more obvious and easier to read, so prefer it.
For big-endian,
On 04/08/14 15:01, Kyrill Tkachov wrote:
Hi all,
Now that both backends have a way of generating CRC32 instructions this
patch adds the crc type to the scheduling information for the Cortex-A53
and Cortex-A57 cores.
For both Cortex-A53 and Cortex-A57 they behave similarly to shifted
On 04/08/14 17:31, Kyrill Tkachov wrote:
Hi all,
As part of other intrinsics-related messing around due to the
float64x1_t changes I noticed these can be (re)implemented relatively
easily.
Tested on aarch64-none-elf and aarch64_be-none-elf to make sure the
lane-wise intrinsics do the
This replaces match_and_simplify with just 'simplify' everywhere,
in patterns and API. It's shorter and as descriptive.
This also pushes two minor changes to not use APIs I consider
internal from SCCVN or forwprop.
Committed.
Richard.
2014-08-05 Richard Biener rguent...@suse.de
*
On 04/08/14 17:45, Kyrill Tkachov wrote:
Hi all,
This patch implements some saturating math *laneq_s* intrinsics.
The implementation is fairly straightforward, just use more general mode
iterators, add appropriate builtins etc.
Some execution tests are added with some scan-assembly parts
On 04/08/14 17:49, Kyrill Tkachov wrote:
Hi all,
Now that uint64x1_t and uint64_t are not interchangeable these
intrinsics have an incorrect type. The reason is that these intrinsics
should have had the scalar type for some of their arguments and results
according to the NEON intrinsics
Jakub,
On Mon, Aug 04, 2014 at 11:51:34AM -0500, Edmar wrote:
Committed on trunk, revision 213596
Committed on 4.9 branch, revision 213597
Note the ChangeLog entry was grossly misformated.
I've fixed it up in gcc/ChangeLog on the trunk, but not on the branch
nor in libgcc. There
Hi all,
This is a cleanup to replace usages of GET_CODE (x) == CONST_INT with
CONST_INT_P (x) and GET_CODE (x) == REG with REG_P (x). No functional
changes.
Tested on aarch64-none-elf and bootstrapped on aarch64-linux.
Ok for trunk?
Thanks,
Kyrill
2014-08-05 Kyrylo Tkachov
On Tue, Aug 05, 2014 at 10:45:38AM +0200, Richard Biener wrote:
On Mon, Aug 4, 2014 at 3:04 PM, tsaund...@mozilla.com wrote:
From: Trevor Saunders tsaund...@mozilla.com
Hi,
It used to be that edge_var_maps held pointers to embedded vectors, but
now it holds vectors. This means that
Hi Andrew, Joseph,
thanks for looking at the patch. See my comments and updated patch below.
2014-08-05 0:54 GMT+04:00 Andrew Pinski pins...@gmail.com:
On Mon, Aug 4, 2014 at 8:29 AM, Alexander Ivchenko aivch...@gmail.com wrote:
Hi,
The following patch disables stdio_va_list fix: stdio.h is
Hi Ian,
On Tue, Jul 22, 2014 at 1:14 AM, Rainer Orth
r...@cebitec.uni-bielefeld.de wrote:
The recent Go 1.3 update (which I couldn't find posted to gcc-patches)
broke Solaris bootstrap:
/vol/gcc/src/hg/trunk/local/libgo/runtime/mem.c:50:20: error: pointer
targets in passing argument 1 of
Hi,
On 08/05/2014 03:58 AM, Jason Merrill wrote:
On 08/04/2014 07:01 PM, Paolo Carlini wrote:
In fact I wondered about that a few minutes after sending my message...
And this is what I figured out: normally we have hard errors from
composite_pointer_type (eg, try scalar types, class types),
Hi Mike,
Or do we go with the removal suggestion of Mike?
I’ll let others weigh in. I’m fine either way. Original author likely
prefers the xfail, so I’m fine with the above.
the test now XPASSes on Solaris, adding testsuite noise in the other
direction on completely innocent systems.
Attached revision #2 of the patch.
---
Add D demangling support for version 2 of the ABI.
include/ChangeLog
2014-08-05 Iain Buclaw ibuc...@gdcproject.org
* demangle.h (DMGL_DLANG): New macro.
(DMGL_STYLE_MASK): Add DMGL_DLANG.
(demangling_styles): Add dlang_demangling.
.. a clarification. As I tried to briefly explain yesterday, this kind
of change means that:
extern void z();
void il() { if (z != (void*)0) z(); }
doesn't trigger anymore the pedwarn at beginning of
composite_pointer_type about the comparison itself, for the simple
reason that we don't call
On 08/05/2014 08:10 AM, Paolo Carlini wrote:
.. a clarification. As I tried to briefly explain yesterday, this kind
of change means that:
extern void z();
void il() { if (z != (void*)0) z(); }
doesn't trigger anymore the pedwarn at beginning of
composite_pointer_type about the comparison
From: Trevor Saunders tsaund...@mozilla.com
hi,
The change to get the entry for the old edge before inserting the new
one was incorrect because if inserting the new one resized the table
then the pointer to the entry for the old one would become invalid.
gcc/
* tree-ssa.c
On Tue, Aug 5, 2014 at 1:17 PM, Trevor Saunders tsaund...@mozilla.com wrote:
On Tue, Aug 05, 2014 at 10:45:38AM +0200, Richard Biener wrote:
On Mon, Aug 4, 2014 at 3:04 PM, tsaund...@mozilla.com wrote:
From: Trevor Saunders tsaund...@mozilla.com
Hi,
It used to be that edge_var_maps
On Tue, Aug 5, 2014 at 2:32 PM, tsaund...@mozilla.com wrote:
From: Trevor Saunders tsaund...@mozilla.com
hi,
The change to get the entry for the old edge before inserting the new
one was incorrect because if inserting the new one resized the table
then the pointer to the entry for the old
Hi,
I've noticed that vec_extract_lo_modemask_name pattern has
vm/vm alternative when mask is not applied. This can lead to insn
with 2 memory operands. Patch bellow fixes it.
Ok for trunk?
2014-08-05 Ilya Tocar ilya.to...@intel.com
* common/config/i386/sse.md
eOn 31 Jul 14:07, Ilya Enkovich wrote:
2014-06-11 12:22 GMT+04:00 Jakub Jelinek ja...@redhat.com:
On Wed, Jun 11, 2014 at 12:14:14PM +0400, Ilya Enkovich wrote:
This patch adds instrumented code support for strlen optimization.
Bootstrapped and tested on linux-x86_64.
Does it look OK?
On Tue, Aug 5, 2014 at 2:43 PM, Ilya Tocar tocarip.in...@gmail.com wrote:
Hi,
I've noticed that vec_extract_lo_modemask_name pattern has
vm/vm alternative when mask is not applied. This can lead to insn
with 2 memory operands. Patch bellow fixes it.
Ok for trunk?
2014-08-05 Ilya Tocar
On Mon, Aug 4, 2014 at 11:09 AM, Zhenqiang Chen zhenqiang.c...@arm.com wrote:
-Original Message-
From: Bin.Cheng [mailto:amker.ch...@gmail.com]
Sent: Monday, August 04, 2014 4:41 PM
To: Zhenqiang Chen
Cc: gcc-patches List
Subject: Re: [PATCH, ivopt] Try aligned offset when
Hi,
On Tue, Aug 5, 2014 at 4:35 AM, Alexander Ivchenko aivch...@gmail.com wrote:
Testing for *android* is less than ideal, because of the possibility of
configuring a *-linux* toolchain to have multilibs using various different
C libraries (with -mandroid being used to select the Android
On Fri, Aug 1, 2014 at 6:03 PM, Kugan kugan.vivekanandara...@linaro.org wrote:
if (rhs_uns)
return wi::ge_p (min, 0); // if min = 0 then range contains positive
values
else
return wi::le_p (max, wi::max_value (TYPE_PRECISION (TREE_TYPE
(ssa)), SIGNED); // if max =
On Tue, Aug 05, 2014 at 04:17:41PM +0200, Richard Biener wrote:
what's the semantic of setting SRP_SIGNED_AND_UNSIGNED
on the subreg? That is, for the created (subreg:lhs_mode
(reg:PROMOTE_MODE of ssa N))?
SRP_SIGNED_AND_UNSIGNED on a subreg should mean that
the subreg is both zero and sign
On Mon, Aug 04, 2014 at 02:04:36PM +0200, Richard Biener wrote:
Looks like .fre can optimize q - (q - 1) into 1:
bb 2:
q.0_3 = (long int) MEM[(void *)i + 4B];
_5 = (long int) i;
- _6 = q.0_3 - _5;
- t.1_7 = _6 /[ex] 4;
- t ={v} t.1_7;
+ t ={v} 1;
i ={v} {CLOBBER};
I'd suggest op0: =store_mask_constraint,v and op1: v,m. This
would result in op0:=vm,v op1:v,m and op0:=v,v op1:v,m.
Uros.
Done.
2014-08-05 Ilya Tocar ilya.to...@intel.com
* common/config/i386/sse.md (vec_extract_lo_modemask_name): Fix
constraint.
---
On Tue, Aug 5, 2014 at 4:36 PM, Ilya Tocar tocarip.in...@gmail.com wrote:
I'd suggest op0: =store_mask_constraint,v and op1: v,m. This
would result in op0:=vm,v op1:v,m and op0:=v,v op1:v,m.
Uros.
Done.
2014-08-05 Ilya Tocar ilya.to...@intel.com
* common/config/i386/sse.md
Jackub,
Thanks for point this up.
I apologize for the sloppiness.
I fixed and committed the ChangeLogs on the branch, revision 213639
Also fixed the libgcc ChangeLog on trunk. Revision 213640
Edmar
On 08/05/2014 03:11 AM, Jakub Jelinek wrote:
On Mon, Aug 04, 2014 at 11:51:34AM -0500, Edmar
Hi,
On 08/05/2014 02:32 PM, Jason Merrill wrote:
On 08/05/2014 08:10 AM, Paolo Carlini wrote:
.. a clarification. As I tried to briefly explain yesterday, this kind
of change means that:
extern void z();
void il() { if (z != (void*)0) z(); }
doesn't trigger anymore the pedwarn at beginning
From: Andi Kleen a...@linux.intel.com
This patch adds a recursion check to gcc-ar/ranlib/nm. The program
avoids to call itself, so it can be directly put into the $PATH
as a wrapper for the normal ar etc.
The recursion check will only work on Linux (or systems with
Linux like /proc) for now. It
On 07/29/2014 02:07 AM, Thomas Schwinge wrote:
On Thu, 20 Mar 2014 15:42:48 +0100, I wrote:
On Tue, 18 Mar 2014 14:50:44 +0100, I wrote:
On Tue, 18 Mar 2014 16:37:24 +0400, Ilmir Usmanov i.usma...@samsung.com
wrote:
This patch introduces support of OpenACC loop directive (and combined
On Tue, Aug 5, 2014 at 7:10 AM, Bruce Korb bk...@gnu.org wrote:
Hi,
On Tue, Aug 5, 2014 at 4:35 AM, Alexander Ivchenko aivch...@gmail.com wrote:
Testing for *android* is less than ideal, because of the possibility of
configuring a *-linux* toolchain to have multilibs using various different
On Sun, 3 Aug 2014, Jan Hubicka wrote:
* diagnostic.c (warning_n): New function.
* diagnostic-core.h (warning_n): Declare.
* ipa-devirt.c (ipa_devirt): Handle singulars correctly;
output dynamic counts when available.
OK, though note the pre-existing issue that these
Applied with a few formatting/comment tweaks, thanks!
Jason
Hi,
On Tue, Aug 5, 2014 at 10:36 AM, enh e...@google.com wrote:
you can see the current version of bionic's stdio.h here:
https://android.googlesource.com/platform/bionic/+/master/libc/include/stdio.h
i'm happy to add any string to the header file that makes things
easier. if you want
Jeff == Jeff Law l...@redhat.com writes:
Jeff Obviously if there are no objections and you check in the change,
Jeff please be on the lookout for any fallout. I'm particularly concerned
Jeff about AIX, Solaris and other non-linux platforms.
I did a build on the AIX box (gcc111) in the compile
On Tue, Aug 05, 2014 at 02:42:17PM +0200, Richard Biener wrote:
On Tue, Aug 5, 2014 at 2:32 PM, tsaund...@mozilla.com wrote:
From: Trevor Saunders tsaund...@mozilla.com
hi,
The change to get the entry for the old edge before inserting the new
one was incorrect because if inserting
Hi,
On 08/05/2014 08:26 PM, Jason Merrill wrote:
Applied with a few formatting/comment tweaks, thanks!
Great. I will double check but var-templ4.C fails for me with an ICE.
Can anybody reproduce?
Thanks!
Paolo.
/.../gcc/testsuite/g++.dg/cpp1y/var-templ4.C:8:17: internal
On 08/05/14 08:36, Marek Polacek wrote:
On Mon, Aug 04, 2014 at 02:04:36PM +0200, Richard Biener wrote:
Looks like .fre can optimize q - (q - 1) into 1:
bb 2:
q.0_3 = (long int) MEM[(void *)i + 4B];
_5 = (long int) i;
- _6 = q.0_3 - _5;
- t.1_7 = _6 /[ex] 4;
- t ={v} t.1_7;
+ t
On 08/03/14 17:44, Mircea Namolaru wrote:
2014-08-03 Mircea Namolarumircea.namol...@inria.fr
Replacement of isl-int by isl_val
* graphite-clast-to-gimple.c: include isl/val.h, isl/val_gmp.h
(compute_bounds_for_param): use isl_val instead of isl_int
On 08/04/14 00:08, Alexander Ivchenko wrote:
Hi,
libcilkrts is compiled with -nostdlib, that means we have to
explicitly specify the pthread library we should link with (e.g. we
don't have such problem with libgomp, because it is C). And, indeed,
-lpthread is hard-coded in the Makefile for
On 08/03/14 22:51, Maxim Kuvyrkov wrote:
On Jul 17, 2014, at 5:34 AM, Jeff Law l...@redhat.com wrote:
On 07/13/14 22:17, Maxim Kuvyrkov wrote:
Hi,
This patch adds dump printouts for scheduling heuristics in
rank_for_schedule. Rank_for_schedule is one of the cornerstones of
haifa scheduler,
Hi Jonathan,
Thanks a lot for taking a look. The patch in question, and the GDB
support, do not yet work with Python3. If that is a necessary
requirement, I can make the changes and send a new version of the
patch.
Thanks,
Siva Chandra
On Mon, Aug 4, 2014 at 2:36 AM, Jonathan Wakely
On 08/03/14 07:39, Richard Sandiford wrote:
This patch adds the new iterators.
gcc/
* rtl-iter.h: New file.
* rtlanal.c: Include it.
(rtx_all_subrtx_bounds, rtx_nonconst_subrtx_bounds): New variables.
(generic_subrtx_iterator T::add_single_to_queue)
On 08/03/14 07:40, Richard Sandiford wrote:
gcc/
* alias.c: Include rtl-iter.h.
(refs_newer_value_cb): Delete.
(refs_newer_value_p): Use FOR_EACH_SUBRTX instead of for_each_rtx.
OK.
Just a few notes.
1. I really like when we can make these callbacks go away :-)
2.
On 08/03/14 07:48, Richard Sandiford wrote:
As noted in https://gcc.gnu.org/ml/gcc-patches/2014-02/msg01391.html
a bitmap-related cleanup turned add_used_regs_1 into a no-op for pseudo
registers, because the result of:
regno = reg_renumber[regno];
is never used. This patch does as Steven
On 08/03/14 07:48, Richard Sandiford wrote:
gcc/
* calls.c: Include rtl-iter.h.
(internal_arg_pointer_based_exp_1): Delete.
(internal_arg_pointer_based_exp): Take a const_rtx.
Use FOR_EACH_SUBRTX to iterate over subrtxes.
OK.
jeff
On 08/03/14 07:50, Richard Sandiford wrote:
gcc/
* combine.c: Include rtl-iter.h.
(unmentioned_reg_p_1): Delete.
(unmentioned_reg_p): Use FOR_EACH_SUBRTX rather than for_each_rtx.
Don't handle null rtxes.
OK.
Jeff
On 08/03/14 07:50, Richard Sandiford wrote:
gcc/
* combine.c (record_truncated_value): Turn from being a for_each_rtx
callback to a function that takes an rtx and returns a bool
(record_truncated_values): Use FOR_EACH_SUBRTX_VAR instead of
for_each_rtx.
OK.
jeff
On 08/03/14 07:53, Richard Sandiford wrote:
gcc/
* cfgcleanup.c: Include rtl-iter.h.
(mentions_nonequal_regs): Turn from being a for_each_rtx callback
to being a function that examines each subrtx itself.
(thread_jump): Update accordingly.
OK.
jeff
On 08/03/14 07:54, Richard Sandiford wrote:
gcc/
* cse.c: Include rtl-iter.h.
(approx_reg_cost_1): Delete.
(approx_reg_cost): Use FOR_EACH_SUBRTX instead of for_each_rtx.
Don't handle null rtxes.
OK.
jeff
On 08/03/14 07:57, Richard Sandiford wrote:
gcc/
* cse.c (is_dead_reg): Change argument to const_rtx.
(dead_debug_insn_data): Delete.
(is_dead_debug_insn): Expand commentary. Turn from being a
for_each_rtx callback to being a function that examines
each
On 08/03/14 07:58, Richard Sandiford wrote:
gcc/
* cse.c (change_cc_mode_args): Delete.
(cse_change_cc_mode): Turn from being a for_each_rtx callback to being
a function that examines each subrtx itself. Take the fields of
change_cc_mode_args as argument and
On 08/03/14 07:58, Richard Sandiford wrote:
gcc/
* ddg.c: Include rtl-iter.h.
(mark_mem_use_1): Rename to...
(mark_mem_use): ...deleting old mark_mem_use. Use FOR_EACH_SUBRTX
instead of for_each_rtx.
(mem_read_insn_p): Update accordingly.
OK.
jeff
On 08/03/14 07:59, Richard Sandiford wrote:
gcc/
* ddg.c (walk_mems_2, walk_mems_1): Delete.
(insns_may_alias_p): Use FOR_EACH_SUBRTX rather than for_each_rtx
to iterate over subrtxes. Return a bool rather than an int.
OK.
Jeff
On Wed, 2014-07-30 at 10:07 +0200, Richard Biener wrote:
On Tue, Jul 29, 2014 at 10:44 PM, David Malcolm dmalc...@redhat.com wrote:
A complaint I heard at Cauldron with the C++ification of GCC passes is
that it's become much more difficult to set breakpoints on the execute
hooks of a pass,
On 08/03/14 08:02, Richard Sandiford wrote:
gcc/
* dse.c: Include rtl-iter.h.
(check_mem_read_rtx): Change void * parameter to real type.
Remove return value.
(check_mem_read_use): Fix comment. Use FOR_EACH_SUBRTX_PTR instead of
for_each_rtx. Don't
On 08/03/14 08:07, Richard Sandiford wrote:
The switch statement in the old code seemed overly cautious. It's well
established elsewhere that the first operand of an RTX_AUTOINC is the
automodified register. If anyone wanted to add a new code for which that
wasn't true they should (a)
On Tue, Aug 05, 2014 at 05:02:57PM -0400, David Malcolm wrote:
On Wed, 2014-07-30 at 10:07 +0200, Richard Biener wrote:
On Tue, Jul 29, 2014 at 10:44 PM, David Malcolm dmalc...@redhat.com wrote:
A complaint I heard at Cauldron with the C++ification of GCC passes is
that it's become much
On 08/03/14 08:10, Richard Sandiford wrote:
gcc/
* fwprop.c: Include rtl-iter.h.
(varying_mem_p): Turn from being a for_each_rtx callback to being
a function that examines each subrtx itself.
(propagate_rtx): Update accordingly.
OK.
jeff
On 08/03/14 08:08, Richard Sandiford wrote:
gcc/
* final.c: Include rtl-iter.h.
(mark_symbol_ref_as_used): Delete.
(mark_symbol_refs_as_used): Use FOR_EACH_SUBRTX instead of
for_each_rtx.
OK.
jeff
On 08/03/14 08:10, Richard Sandiford wrote:
gcc/
* ira.c: Include rtl-iter.h.
(set_paradoxical_subreg): Turn from being a for_each_rtx callback
to being a function that examines each subrtx itself. Remove
handling of null rtxes.
(update_equiv_regs):
On 08/03/14 08:11, Richard Sandiford wrote:
gcc/
* jump.c: Include rtl-iter.h.
(returnjump_p_1): Delete.
(returnjump_p): Use FOR_EACH_SUBRTX rather than for_each_rtx.
Remove handling of null rtxes.
OK.
Jeff
On 08/03/14 08:12, Richard Sandiford wrote:
gcc/
* jump.c (eh_returnjump_p_1): Delete.
(eh_returnjump_p): Use FOR_EACH_SUBRTX rather than for_each_rtx.
Remove handling of null rtxes.
OK.
jeff
On 08/03/14 08:15, Richard Sandiford wrote:
gcc/
* lower-subreg.c (adjust_decomposed_uses): Delete.
(resolve_debug): Use FOR_EACH_SUBRTX_PTR rather than for_each_rtx.
Remove handling of null rtxes.
OK.
jeff
On 08/03/14 08:14, Richard Sandiford wrote:
gcc/
* loop-iv.c (altered_reg_used): Turn from being a for_each_rtx callback
to being a function that examines each subrtx itself.
(simplify_using_condition, simplify_using_initial_values): Update
accordingly.
Ok.
jeff
On 08/03/14 08:18, Richard Sandiford wrote:
gcc/
* reg-stack.c: Include rtl-iter.h.
(subst_stack_regs_in_debug_insn): Delete.
(subst_all_stack_regs_in_debug_insn): Use FOR_EACH_SUBRTX_PTR
instead of for_each_rtx.
OK.
Jeff
On 08/03/14 08:19, Richard Sandiford wrote:
gcc/
* regcprop.c (cprop_find_used_regs_1): Delete.
(cprop_find_used_regs): Use FOR_EACH_SUBRTX instead of for_each_rtx.
OK.
jeff
On 08/03/14 08:19, Richard Sandiford wrote:
gcc/
* regcprop.c: Include rtl-iter.h.
(kill_value): Take a const_rtx.
(kill_autoinc_value): Turn from being a for_each_rtx callback
to being a function that examines each subrtx itself.
On 08/03/14 08:27, Richard Sandiford wrote:
find_all_hard_regs seems like a useful function so I split it out of
the note_uses callback and exposed it in rtl.h. I have (or had)
other patches that make use of it.
gcc/
* rtl.h (find_all_hard_regs): Declare.
* rtlanal.c
On 08/03/14 08:33, Richard Sandiford wrote:
gcc/
* rtl.h (tls_referenced_p): Take a const_rtx rather than an rtx.
* rtlanal.c (tls_referenced_p_1): Delete.
(tls_referenced_p): Take a const_rtx rather than an rtx.
Use FOR_EACH_SUBRTX rather than for_each_rtx.
OK.
On 08/03/14 08:34, Richard Sandiford wrote:
gcc/
* store-motion.c: Include rtl-iter.h.
(extract_mentioned_regs_1): Delete.
(extract_mentioned_regs): Use FOR_EACH_SUBRTX_VAR rather than
for_each_rtx to iterate over subrtxes.
OK.
jeff
On 08/03/14 08:35, Richard Sandiford wrote:
gcc/
* var-tracking.c: Include rtl-iter.h.
(rtx_debug_expr_p): Turn from being a for_each_rtx callback
to being a function that examines each subrtx itself.
(use_type): Update accordingly.
OK.
jeff
On 08/03/14 08:36, Richard Sandiford wrote:
gcc/
* var-tracking.c (non_suitable_const): Turn from being a for_each_rtx
callback to being a function that examines each subrtx itself.
Remove handling of null rtxes.
(add_uses): Update accordingly.
OK.
Jeff
On 08/03/14 08:38, Richard Sandiford wrote:
gcc/
* var-tracking.c (add_uses): Take an rtx rather than an rtx *.
Give real type of data parameter. Remove return value.
(add_uses_1): Use FOR_EACH_SUBRTX_VAR rather than for_each_rtx
to iterate over subrtxes.
OK.
On 08/03/14 08:02, Richard Sandiford wrote:
This also fixes what I think is a bug: find_memory used to stop at the
first MEM it found. If that MEM was nonvolatile and nonconstant, we'd
return MEMREF_NORMAL even if there was another volatile MEM.
gcc/
* df-problems.c: Include
On 08/03/14 08:04, Richard Sandiford wrote:
gcc/
* dwarf2out.c (resolve_one_addr): Remove unused data parameter.
Return a bool, inverting the result so that 0/false means not ok.
Use FOR_EACH_SUBRTX_PTR instead of for_each_rtx to iterate over
subrtxes of a CONST.
On 08/03/14 08:03, Richard Sandiford wrote:
gcc/
* dwarf2out.c: Include rtl-iter.h.
(const_ok_for_output_1): Take the rtx instead of a pointer to it.
Remove unused data parameter. Return a bool, inverting the result
so that 0/false means not ok.
On 08/03/14 08:13, Richard Sandiford wrote:
gcc/
* loop-iv.c: Include rtl-iter.h.
(find_single_def_src): New function.
(replace_single_def_regs): Turn from being a for_each_rtx callback
to being a function that examines each subrtx itself.
On 08/03/14 08:20, Richard Sandiford wrote:
gcc/
* reload1.c: Include rtl-iter.h.
(note_reg_elim_costly): Turn from being a for_each_rtx callback
to being a function that examines each subrtx itself.
(eliminate_regs_1, elimination_costs_in_insn): Update
On 08/03/14 08:22, Richard Sandiford wrote:
The old function handled constant pool SYMBOL_REFs by going straight to the
underlying constant, which meant you couldn't test for the SYMBOL_REF itself.
gcc/
* rtl.h (get_pool_constant, rtx_referenced_p): Replace rtx parameters
with
On 08/03/14 08:25, Richard Sandiford wrote:
The main change here is to handle ADDR_VEC and ADDR_DIFF_VECs specially,
since they can have many elements and are a relatively important case
for this function. This is for speed rather than correctness.
gcc/
* rtl.h (replace_label_data):
1 - 100 of 117 matches
Mail list logo
