On Wed, 20 Nov 2019, Jan Hubicka wrote:
> Hi,
> I have noticed that for Firefox around 1GB of peak memory use goes into
> the fact that we never free memory_block_pool::freelist.
>
> This patch adds memory_block_pool::trim which reduces freelist to a given
> size. It is called from ggc_collect w
On 11/20/19 7:54 AM, Szabolcs Nagy wrote:
> On 14/11/2019 20:23, Szabolcs Nagy wrote:
>> Sorry v2 had a bug.
>>
>> v2: added documentation and tests.
>> v3: fixed expand_simd_clones so different isa variants are actually
>> generated.
>>
>> GCC currently supports two ways to declare the availa
When building an executable with LTO, GCC effectively ignores
__attribute__((section)) on C++ inline member functions. Moving such
functions into the .text section seems to be intentional, but I think
ignoring the section attribute is unintentional:
https://gcc.gnu.org/ml/gcc-patches/2010-07/msg00
While fixing some bugs in __attribute__((section)), I found it difficult
to write tests. Make testing easier: introduce the
scan-assembler-symbol-section and scan-symbol-section helpers. See
in-line documentation for details.
Testing:
* Run `make check` on x86_64-linux-gnu with --disable-multilib
Hi,
The issue seems to happen with -O1, because header only contains phi:
[local count: 118111600]:
# iter.12_9 = PHI <0(2), iter.12_10(10)>
and thus we hit segfault in following hunk in find_loop_guard:
else
{
cond = dyn_cast (last_stmt (header));
if (! con
On 11/20/19 8:37 PM, Marek Polacek wrote:
On Tue, Nov 19, 2019 at 05:33:09PM -0500, Jason Merrill wrote:
On 11/19/19 1:44 AM, Marek Polacek wrote:
It also looks like you're using the LOOKUP flag to mean two different
things:
1) try to treat parenthesized args as an aggregate initializer
(build
On Wed, Nov 20, 2019 at 4:18 PM Jakub Jelinek wrote:
>
> On Tue, Nov 05, 2019 at 05:17:10PM +, Wilco Dijkstra wrote:
> > Passes bootstrap and regress on AArch64 and x64. OK for commit?
>
> This broke bootstrap on x86_64-linux as well as i686-linux (guess all
> targets that go supports).
> The
This libgo patch declares runtime_usestackmaps in stack.c, not
runtime.h. This fixes https://gcc.gnu.org/PR92605. Bootstrapped and
ran Go testsuite on x86_64-pc-linux-gnu. Committed to mainline.
Ian
Index: gcc/go/gofrontend/MERGE
=
This patch to the libgo mksysinfo script uses type aliases for time
struct field types. It also fixes a case where grep wasn't
redirecting to /dev/null. This lets some syscall package types be
identical to some golang.org/x/sys/unix types, and fixes
https://golang.org/issue/35713. Bootstrapped a
On Thu, Nov 21, 2019 at 01:41:47AM +0100, Rainer Orth wrote:
> Same on sparc-sun-solaris2.11 and i386-pc-solaris2.11.
>
> There where quite a number of non-Go regressions all over the place.
> Many are like this:
>
> FAIL: gcc.c-torture/execute/complex-6.c -O0 (test for excess errors)
>
> ld:
On 11/20/19 8:49 PM, Marek Polacek wrote:
On Tue, Nov 19, 2019 at 05:34:45PM -0500, Jason Merrill wrote:
On 11/18/19 7:04 PM, Marek Polacek wrote:
The newly added diagnostic causes an ICE because the new grokdeclarator call
returned error_mark_node and DECL_SOURCE_LOCATION crashes on that. So
On 11/20/19 7:21 PM, Jakub Jelinek wrote:
On Mon, Nov 18, 2019 at 02:41:48PM -0500, Jason Merrill wrote:
2019-11-11 Jakub Jelinek
PR c++/92458
* constraint.cc: Include tree-hash-traits.h.
(decl_tree_cache_traits): New type.
(decl_tree_cache_map): New typedef.
Hi Jakub,
> On Tue, Nov 05, 2019 at 05:17:10PM +, Wilco Dijkstra wrote:
>> Passes bootstrap and regress on AArch64 and x64. OK for commit?
>
> This broke bootstrap on x86_64-linux as well as i686-linux (guess all
> targets that go supports).
indeed: just saw it on Solaris with the native ld.
On Tue, Nov 05, 2019 at 05:17:10PM +, Wilco Dijkstra wrote:
> Passes bootstrap and regress on AArch64 and x64. OK for commit?
This broke bootstrap on x86_64-linux as well as i686-linux (guess all
targets that go supports).
The following patch fixes it for me, though not sure which *.c file is
This patch is for PR target/92499, a bug reported by GLIBC maintainers
against nios2 target. The initial failure symptom was a linker error
resulting from use of GP-relative addressing on an object that wasn't
allocated in the small data section. It turns out that the real problem
is that the
On Wed, Nov 20, 2019 at 11:15:56PM +, Joseph Myers wrote:
> Various bad uses of the [[fallthrough]] attribute are constraint
> violations in C2x, so need pedwarns rather than warnings.
>
> This patch duly turns the relevant warnings into pedwarns. The
> relevant code is not specific to C, and
We currently expand various floating point comparisons early, to some
sequences with cror insns and the like. This doesn't optimize well.
Change that to allow any of the 14 floating point comparisons in the
instruction stream, and split them after combine (at split1).
Tested on powerpc64-linux {
Various bad uses of the [[fallthrough]] attribute are constraint
violations in C2x, so need pedwarns rather than warnings.
This patch duly turns the relevant warnings into pedwarns. The
relevant code is not specific to C, and does not know which form the
attribute was given in ([[fallthrough]] or
This started crashing with the <=> implementation but has now been
fixed by r278465.
Tested on x86_64-linux, applying to trunk.
2019-11-20 Marek Polacek
PR c++/92443
* g++.dg/cpp0x/constexpr-92443.C: New test.
diff --git gcc/testsuite/g++.dg/cpp0x/constexpr-92443.C
gcc/tests
On Tue, Nov 19, 2019 at 08:00:22PM +0100, Tim Rühsen wrote:
> Thanks. Where exactly did it go ? Still can't see it in
> git://sourceware.org/git/binutils-gdb.git. Is there another repository ?
It will appear in the binutils repo when someone syncs libiberty with
the master sources in the gcc repo.
Am 20.11.19 um 23:18 schrieb Janne Blomqvist:
On Wed, Nov 20, 2019 at 11:35 PM Thomas König wrote:
Am 20.11.19 um 21:45 schrieb Janne Blomqvist:
BTW, since this is done for the purpose of optimization, have you done
testing on some suitable benchmark suite such as polyhedron, whether
it a) ge
On 11/20/19 11:18 PM, Janne Blomqvist wrote:
Of course, Fortran language rules specify that the call to bar
cannot do anything to n
Hmm, does it? What about the following modification to your testcase:
This code violates (quote from F2018):
"15.5.2.13 Restrictions on entities associated wit
On Thu, Nov 21, 2019 at 12:18 AM Janne Blomqvist
wrote:
>
> On Wed, Nov 20, 2019 at 11:35 PM Thomas König wrote:
> > (Why do we zero %eax
> > before each call? It should not be a variadic call right?)
>
> Not sure. Maybe some belt and suspenders thing? I guess someone better
> versed in ABI minut
On Wed, 20 Nov 2019, Janne Blomqvist wrote:
> On Wed, Nov 20, 2019 at 7:53 PM Joseph Myers wrote:
> >
> > This patch is OK with http changed to https. (That is, with it changed
> > where the patch is already changing the URL. While changing http to https
> > makes sense more generally in the do
On 11/20/19 10:35 PM, Thomas König wrote:
I haven't run any actual benchmarks.
I think it would be interesting – I think there can be quite some
advantages in some cases, while in most cases, it is not really noticable.
Of course, Fortran language rules specify that the call to bar cannot
do a
On Wed, Nov 20, 2019 at 11:35 PM Thomas König wrote:
>
> Am 20.11.19 um 21:45 schrieb Janne Blomqvist:
> > BTW, since this is done for the purpose of optimization, have you done
> > testing on some suitable benchmark suite such as polyhedron, whether
> > it a) generates any different code b) does
On 20/11/2019 20:48, Bernd Schmidt wrote:
> On 11/20/19 8:27 PM, Mikael Pettersson wrote:
>> On Wed, Nov 20, 2019 at 3:16 PM Bernd Schmidt wrote:
>>> Probably best to just run tests on stage1 and hope something shows up.
>>
>> Ok, how do I did that? I've always just done 'make -k check' after
>>
Am 20.11.19 um 21:45 schrieb Janne Blomqvist:
BTW, since this is done for the purpose of optimization, have you done
testing on some suitable benchmark suite such as polyhedron, whether
it a) generates any different code b) does it make it go faster?
I haven't run any actual benchmarks.
Howeve
On Wed, 20 Nov 2019 22:38:30 +0200
Janne Blomqvist wrote:
> On Wed, Nov 20, 2019 at 8:00 PM Bernhard Reutner-Fischer
> wrote:
> >
> > On 19 November 2019 23:54:55 CET, Thomas Koenig
> > wrote:
> > >Am 19.11.19 um 11:39 schrieb Bernhard Reutner-Fischer:
> > >> + char name[GFC_MAX_SYMBO
On Wed, 2019-11-20 at 16:07 +0100, Martin Liška wrote:
> On 11/20/19 4:02 PM, David Malcolm wrote:
> > On Wed, 2019-11-20 at 15:51 +0100, Martin Liška wrote:
> > > On 11/20/19 3:49 PM, David Malcolm wrote:
> > > > On Wed, 2019-11-06 at 11:30 +0100, Martin Liska wrote:
> > > > > gcc/ChangeLog:
> > >
gcc/ChangeLog:
* doc/analyzer.texi (Analyzer Paths): Add path pruning and
precision-of-wording vfuncs.
(Debugging the Analyzer): Consolidate duplicated material.
---
gcc/doc/analyzer.texi | 47 +--
1 file changed, 37 insertions(+)
There was a bad upper limit in diagnostic_manager::prune_path's test for
pruning redundant interprocedural pass information, leading to an ICE.
Additional, the test for pruning
[..., call, function-entry, return, ...]
triples doesn't work for -fanalyzer-verbosity=0, which doesn't generate
functi
This patch adds some special-case logic to how paths are generated
so that leaks due to longjmp rewinding past a frame now show where the
longjmp rewinds to (longjmp and setjmp are already something of a
special-case so this seems reasonable).
A colorized LTO example of the output can be seen at:
This patch adds a new debugging function.
gcc/ChangeLog:
* analyzer/checker-path.cc (checker_path::debug): New member
function.
* analyzer/checker-path.h (checker_path::debug): New decl.
---
gcc/analyzer/checker-path.cc | 19 +++
gcc/analyzer/checker-path.h
gcc/testsuite/ChangeLog:
* gcc.dg/analyzer/malloc-ipa-8-lto-a.c: New test.
* gcc.dg/analyzer/malloc-ipa-8-lto-b.c: New test.
* gcc.dg/analyzer/malloc-ipa-8-lto-c.c: New test.
* gcc.dg/analyzer/malloc-ipa-8-lto.h: New test.
---
gcc/testsuite/gcc.dg/analyzer/malloc-ip
Various commits on 2019-11-12 including r278083 through r278087
reimplemented parameter-handling in terms of options, so that
params are defined in params.opt rather than params.def.
This patch adds the params for the analyzer to plugin.opt,
replacing the patch:
[PATCH 22/49] analyzer: params.d
This patch fixes the missing leak report in setjmp-7.c, which failed
to report on an unchecked "malloc" result leaking due to a longjmp
skipping past its frame to a setjmp in its caller.
The root cause issue is that impl_region_model_context::on_state_leak
would call the state machine's on_leak vf
r277284 eliminated json::number in favor of json::float_number and
json::integer_number. Update the diagnostic_path code accordingly.
gcc/ChangeLog:
* tree-diagnostic-path.cc (default_tree_make_json_for_path): Fix
overlong line. Use json::integer_number.
---
gcc/tree-diagnostic-
"convert" is not implemented in lto1, leading to an ICE when used.
This patch implements a build_cast function for the analyzer and uses it
to replace the call to "convert" when handling casts in region_model.
This fixes numerous ICEs when running the analyzer from LTO.
gcc/ChangeLog:
*
r277284 eliminated json::number in favor of json::float_number and
json::integer_number. Update the diagnostic metadata code accordingly.
gcc/ChangeLog:
* diagnostic-format-json.cc (json_from_metadata): Use
json::integer_number.
---
gcc/diagnostic-format-json.cc | 3 ++-
1 file c
I've rebased my static analysis work (from r276961 to r278495)
This patch kit contains the changes that were needed (patches 1-4),
along with various followups (patches 5-11).
These patches fix the worst of the issues with LTO compatibility;
an example LTO diagnostic is:
https://dmalcolm.fedorap
gcc/ChangeLog:
* analyzer/analysis-plan.cc: Remove include of "params.h".
(analysis_plan::use_summary_p): Update for conversion of params to
options.
* analyzer/engine.cc: Remove include of "params.h".
(exploded_graph::get_or_create_node): Update for conversi
On Wed, Nov 20, 2019 at 10:38:30PM +0200, Janne Blomqvist wrote:
> On Wed, Nov 20, 2019 at 8:00 PM Bernhard Reutner-Fischer
> wrote:
> >
> > On 19 November 2019 23:54:55 CET, Thomas Koenig
> > wrote:
> > >Am 19.11.19 um 11:39 schrieb Bernhard Reutner-Fischer:
> > >> + char name[GFC_MAX_SYMB
On 11/7/19 1:06 PM, Peter Bergner wrote:
> Before, LRA, we have an insn that sets a TImode pseudo with an integer
> constant and a following insn that copies that TImode pseudo to a PTImode
> pseudo. During LRA spilling, we generate a new insn that sets a PTImode
> pseudo to that constant directly
On Tue, Nov 19, 2019 at 05:34:45PM -0500, Jason Merrill wrote:
> On 11/18/19 7:04 PM, Marek Polacek wrote:
> > The newly added diagnostic causes an ICE because the new grokdeclarator call
> > returned error_mark_node and DECL_SOURCE_LOCATION crashes on that. So don't
> > attempt to print the new d
On 11/20/19 8:27 PM, Mikael Pettersson wrote:
> On Wed, Nov 20, 2019 at 3:16 PM Bernd Schmidt wrote:
>> Probably best to just run tests on stage1 and hope something shows up.
>
> Ok, how do I did that? I've always just done 'make -k check' after
> full bootstraps.
> I assume the stage 1 artifact
On Wed, Nov 20, 2019 at 06:20:34PM +, Richard Sandiford wrote:
> Segher Boessenkool writes:
> > So this would work if you had pseudos here, instead of the hard reg?
> > Because it is a hard reg it is the same number in both places, making it
> > hard to move.
>
> Yeah, probably. But the hard
On Sat, Nov 16, 2019 at 10:34 PM Thomas Koenig wrote:
>
> Hello world,
>
> here is an update to the patch.
>
> I have now included variables where the user did not specify INTENT(IN)
> by checking that the dummy variables to be replaced by temporaries
> are not, indeed, assigned a value. This also
On Wed, Nov 20, 2019 at 8:00 PM Bernhard Reutner-Fischer
wrote:
>
> On 19 November 2019 23:54:55 CET, Thomas Koenig wrote:
> >Am 19.11.19 um 11:39 schrieb Bernhard Reutner-Fischer:
> >> + char name[GFC_MAX_SYMBOL_LEN + 1];
> >> + snprintf (name, GFC_MAX_SYMBOL_LEN, "__dummy_%d_%s", var_
On Tue, Nov 19, 2019 at 05:33:09PM -0500, Jason Merrill wrote:
> On 11/19/19 1:44 AM, Marek Polacek wrote:
> > > It also looks like you're using the LOOKUP flag to mean two different
> > > things:
> > >
> > > 1) try to treat parenthesized args as an aggregate initializer
> > > (build_new_method_ca
On Wed, Nov 20, 2019 at 7:53 PM Joseph Myers wrote:
>
> This patch is OK with http changed to https. (That is, with it changed
> where the patch is already changing the URL. While changing http to https
> makes sense more generally in the documentation whenever a site supports
> https, that's pr
On 11/20/19 12:27 PM, Mikael Pettersson wrote:
> On Wed, Nov 20, 2019 at 3:16 PM Bernd Schmidt
> wrote:
>>
>> On 11/20/19 2:50 PM, Mikael Pettersson wrote:
>>> On Mon, Nov 18, 2019 at 9:57 PM Mikael Pettersson
>>> wrote:
On Mon, Nov 18, 2019 at 8:31 PM Bernd Schmidt
wrote:
>
On Wed, Nov 20, 2019 at 5:46 PM Tobias Burnus wrote:
>
> Hi Janne,
>
> On 11/18/19 9:34 PM, Janne Blomqvist wrote:
> > Now that we require a minimum of MPFR 3.1.0+ to build GCC, we can do
> > some modernization of the MPFR usage in the GFortran frontend.
>
> OK – thanks for the cleanup.
>
> [For r
On Wed, Nov 20, 2019 at 3:16 PM Bernd Schmidt wrote:
>
> On 11/20/19 2:50 PM, Mikael Pettersson wrote:
> > On Mon, Nov 18, 2019 at 9:57 PM Mikael Pettersson
> > wrote:
> >>
> >> On Mon, Nov 18, 2019 at 8:31 PM Bernd Schmidt
> >> wrote:
> >>>
> >>> Hi Mikael,
> >>>
> This fixed the problem
On Mon, Nov 18, 2019 at 02:41:48PM -0500, Jason Merrill wrote:
> > 2019-11-11 Jakub Jelinek
> >
> > PR c++/92458
> > * constraint.cc: Include tree-hash-traits.h.
> > (decl_tree_cache_traits): New type.
> > (decl_tree_cache_map): New typedef.
> > (decl_constraints): Change ty
On 11/20/19 10:54 AM, David Malcolm wrote:
On Wed, 2019-11-20 at 00:38 +0100, Jakub Jelinek wrote:
Hi!
The following patch is a minimal fix to avoid
cannot convert ‘‘addr_expr’ not supported by dump_type’
to ‘X*’
and similar messages. The recently added complain_about_bad_argument
function exp
On 11/20/19 4:40 PM, Paolo Carlini wrote:
Hi,
while working on improving the locations of cp_build_indirect_ref_1 &
co, I noticed this nit which seems a separate issue.
In a nutshell, at variance with many other cases, in build_x_arrow we
don't immediately check for error_mark_node the retur
On 11/19/19 11:29 PM, Jakub Jelinek wrote:
Hi!
The following patch fixes build of older GCC releases with newer ones.
In GCC 4.6 and earlier, we had:
struct cgraph_node;
struct cgraph_node *cgraph_node (tree);
and that is something on which GCC 9+ code errors on if it sees any
__gcc_diag__ and s
On 11/20/19 9:16 AM, Jakub Jelinek wrote:
On Tue, Nov 19, 2019 at 11:35:02PM -0500, Jason Merrill wrote:
It would seem better to break after consuming the token, so we just skip the
extra processing and still give the same error.
And instead of this, maybe set CP_PARSER_FLAGS_NO_TYPE_DEFINITION
When compiling __RTL functions with a start pass, `skip_pass` needs to
set global state when skipping a pass that would have marked something
for future passes to see.
Existing examples are setting `reload_completed` and
`epilogue_completed` when skipping the reload and pro_and_epilogue
passes res
Segher Boessenkool writes:
>> /* Make sure that the value that is to be substituted for the register
>> does not use any registers whose values alter in between. However,
>> If the insns are adjacent, a use can't cross a set even though we
>> think it might (this can happe
On Wed, 20 Nov 2019, Jakub Jelinek wrote:
> On Wed, Nov 20, 2019 at 10:17:38AM -0600, Segher Boessenkool wrote:
> > On Wed, Nov 20, 2019 at 03:46:29PM +, Richard Sandiford wrote:
> > > Segher Boessenkool writes:
> > > > UNLT & ORDERED is always LT. When would it not be true?
> > >
> > > LT
On Wed, 20 Nov 2019, Matthew Malcomson wrote:
> I don't have much of a plan.
>
> The most promising lead I have is that libiberty/alloca.c has a similar
> functionality but with macros to account for a special case.
The comment in libiberty/aclocal.m4 is:
# We always want a C version of alloca
The vrbit_1 test was missing a flag to disable code sharing.
Committed as obvious.
ChangeLog:
2019-11-20 Wilco Dijkstra
testsuite/
* gcc.target/aarch64/simd/vrbit_1.c: Add -fno-ipa-icf.
--
diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vrbit_1.c
b/gcc/testsuite/gcc.target/aar
On 19 November 2019 23:54:55 CET, Thomas Koenig wrote:
>Am 19.11.19 um 11:39 schrieb Bernhard Reutner-Fischer:
>> + char name[GFC_MAX_SYMBOL_LEN + 1];
>> + snprintf (name, GFC_MAX_SYMBOL_LEN, "__dummy_%d_%s", var_num++,
>> +f->sym->name);
>> +
>> + if (gfc_get_sym_tr
On 11/19/19 12:42 AM, Richard Biener wrote:
> On Mon, 18 Nov 2019, Jeff Law wrote:
>
>> On 11/18/19 3:37 AM, Richard Biener wrote:
>>> On Mon, 18 Nov 2019, Jakub Jelinek wrote:
>>>
On Mon, Nov 18, 2019 at 11:07:22AM +0100, Richard Biener wrote:
> On Mon, 18 Nov 2019, Jakub Jelinek wrote:
Since r278084 (part of the params refactoring), most of libgccjit's
test suite has been ICEing.
The root cause is that jit-playback.c injects params to its fake_args
here:
/* Aggressively garbage-collect, to shake out bugs: */
if (get_bool_option (GCC_JIT_BOOL_OPTION_SELFCHECK_GC))
{
This patch is OK with http changed to https. (That is, with it changed
where the patch is already changing the URL. While changing http to https
makes sense more generally in the documentation whenever a site supports
https, that's probably best not mixed with the move from ftp.)
--
Joseph S
On 11/18/19 9:15 AM, Martin Sebor wrote:
> On 11/17/19 12:03 PM, Jeff Law wrote:
>> On 11/12/19 12:55 PM, Martin Sebor wrote:
>>> On 11/12/19 10:54 AM, Jeff Law wrote:
On 11/12/19 1:15 AM, Richard Biener wrote:
> On Tue, Nov 12, 2019 at 6:10 AM Jeff Law wrote:
>>
>> On 11/6/19 3:3
On Wed, 20 Nov 2019, Maciej W. Rozycki wrote:
> > But even then, if you configure GCC using "--with-sysroot" or
> > "--with-build-sysroot", both of those paths are the top-level sysroot, to
> > which the sysroot suffix gets appended before GCC uses it for any purpose,
> > unless you explicitly
On 11/19/19 2:02 AM, Claudiu Zissulescu wrote:
> The patterns neg_scc_insn and not_scc_insn are not correct, leading to
> failing pr77309 test for ARC700. Add two new bic compare with zero
> patterns to improve output code.
>
> gcc/
> -xx-xx Claudiu Zissulescu
>
> * config/arc/arc.md
On 20/11/2019 14:29, Martin Liška wrote:
> On 11/7/19 7:37 PM, Matthew Malcomson wrote:
>> I have rebased this series onto Martin Liska's patches that take the most
>> recent libhwasan from upstream LLVM.
>> https://gcc.gnu.org/ml/gcc-patches/2019-11/msg00340.html
>>
>> I've also cleared up some no
On 11/20/19 4:02 AM, Martin Liška wrote:
> Hi.
>
> The patch is about removal of an unused libtool-version file,
> and comments in other libtool-version files are legacy.
>
> Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
> I'm going to install the patch if there are no ob
On 11/20/19 5:46 AM, Richard Sandiford wrote:
> After r278246, we can try building the out[] store value from scalars
> if the target has no multiplication support. That's not necessarily
> a good thing, but like most of vect/, this test is run with the cost
> model disabled.
>
> Tested on aarch6
On 11/20/19 8:15 AM, Richard Sandiford wrote:
> bb-slp-40.c fails on SPARC targets without VIS4 because it
> requires addition on vectors of bytes. There doesn't seem to be
> an existing target selector for this, so I added vect_char_add.
> (Wasn't sure whether to use vect_char_add, for consistenc
On Wed, Nov 20, 2019 at 04:59:49PM +, Richard Sandiford wrote:
> Segher Boessenkool writes:
> > On Wed, Nov 20, 2019 at 05:30:48PM +0100, Jakub Jelinek wrote:
> >> On Wed, Nov 20, 2019 at 10:17:38AM -0600, Segher Boessenkool wrote:
> >> > On Wed, Nov 20, 2019 at 03:46:29PM +, Richard Sandi
On Wed, Nov 20, 2019 at 04:35:24PM +, Richard Sandiford wrote:
> Actually, this doesn't work because *_operators want rtxes rather
> than codes. I can get around that by passing op0 and op1 for
> the existing rtxes. For the conversion at the end, I can do:
>
> machine_mode compared_mode =
Segher Boessenkool writes:
> On Wed, Nov 20, 2019 at 05:30:48PM +0100, Jakub Jelinek wrote:
>> On Wed, Nov 20, 2019 at 10:17:38AM -0600, Segher Boessenkool wrote:
>> > On Wed, Nov 20, 2019 at 03:46:29PM +, Richard Sandiford wrote:
>> > > Segher Boessenkool writes:
>> > > > UNLT & ORDERED is a
On 11/20/19 7:16 AM, Bernd Schmidt wrote:
> On 11/20/19 2:50 PM, Mikael Pettersson wrote:
>> On Mon, Nov 18, 2019 at 9:57 PM Mikael Pettersson
>> wrote:
>>>
>>> On Mon, Nov 18, 2019 at 8:31 PM Bernd Schmidt
>>> wrote:
Hi Mikael,
> This fixed the problem, thanks.
Cou
On Wed, Nov 20, 2019 at 05:30:48PM +0100, Jakub Jelinek wrote:
> On Wed, Nov 20, 2019 at 10:17:38AM -0600, Segher Boessenkool wrote:
> > On Wed, Nov 20, 2019 at 03:46:29PM +, Richard Sandiford wrote:
> > > Segher Boessenkool writes:
> > > > UNLT & ORDERED is always LT. When would it not be tr
Hi,
this patch makes inliner to update callee keys only in function within
newly inlined clone rather than in the whole function it is inlined to.
This is possible only when the remaining edges are not becoming more hot
for inlining and on this I rely on monotonocity of the badness function:
if cal
Hi,
while working on improving the locations of cp_build_indirect_ref_1 &
co, I noticed this nit which seems a separate issue.
In a nutshell, at variance with many other cases, in build_x_arrow we
don't immediately check for error_mark_node the return value of
decay_conversion. Then, for the
My patch that added initial C2X support and associated command-line
options missed documenting -Wc11-c2x-compat although the other options
were properly documented. This patch adds the missing documentation.
Tested with "make info" and "make pdf". Applied to mainline. Will apply
to GCC 9 branc
On Wed, 20 Nov 2019, Alexander Monakov wrote:
> On Wed, 20 Nov 2019, Richard Biener wrote:
>
> > On Tue, Nov 19, 2019 at 5:07 PM Vladimir Makarov
> > wrote:
> > >
> > > The following patch fixes
> > >
> > > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90007
> > >
> > > Sometime ago a code which
Richard Sandiford writes:
>>> I'd actually considered converting to signed and back instead of adding
>>> extra cases, but I thought that would be rejected as too inefficient.
>>> (That was a concern with my patch above.) It seemed like one of the selling
>>> points of doing it your way was that
On Wed, Nov 20, 2019 at 10:17:38AM -0600, Segher Boessenkool wrote:
> On Wed, Nov 20, 2019 at 03:46:29PM +, Richard Sandiford wrote:
> > Segher Boessenkool writes:
> > > UNLT & ORDERED is always LT. When would it not be true?
> >
> > LT traps on quiet NaNs for -ftrapping-math, UNLT and ORDER
Hi,
this patch avoids malloc traffic in evaluate_properties_for_edge which
allocates vectors holding known values, aggregates and polyorphic contextes.
I made the vector allocated by a caller who can place them at stack using
auto_vec.
The patch also adds logic to avoid calculation and clearing of
On Wed, Nov 20, 2019 at 03:46:29PM +, Richard Sandiford wrote:
> Segher Boessenkool writes:
> > UNLT & ORDERED is always LT. When would it not be true?
>
> LT traps on quiet NaNs for -ftrapping-math, UNLT and ORDERED don't.
No? -ftrapping-math makes nothing trap. The only thing it does is
Hi,
this patch treads ages old problem that compile time needed to estimate call
size is not constant, but a function of a number of calls of the callee.
If there is a function with many callers and many callees this triggers
quadratic behaviour.
This patch adds summary info for all callees of a n
The `--enable-version-specific-runtime-libs' configuration option is now
supported throughout all of our target library subdirectories, so update
installation documentation accordingly and also mention that the default
for the option is `yes' for libada and `no' for the remaining libraries.
LGTM.
Thanks,
Tobias
PS: For reference, mpfr_regular_p was added in MPFR 3.0.0 (as stated);
acting as follows:
mpfr_number_p = returns nonzero if ordinary number (i.e., neither NaN
nor an infinity),
mpfr_regular_p = returns nonzero if regular number (i.e., neither NaN,
nor an infinity nor ze
On Wed, 20 Nov 2019, Jakub Jelinek wrote:
> Ok with appropriate ChangeLog entry.
I've used one included with the submission. Change applied now, thanks
for your review.
Maciej
Segher Boessenkool writes:
> Hi!
>
> On Wed, Nov 20, 2019 at 09:42:46AM +, Richard Sandiford wrote:
>> Segher Boessenkool writes:
>> >> Before I resubmit, why is the simplify-rtx.c part all wrong?
>> >
>> > It was nice and simple, and it isn't anymore. 8 4 2 1 for the four of
>> > lt gt eq u
This patch builds on https://gcc.gnu.org/ml/gcc-patches/2019-11/msg01034.html.
Tie static assertion diagnostics into constraint diagnostic. For
example, this program:
template
concept iterator =
requires (I i)
{
++i;
*i;
};
static_assert(iterator);
Yields these errors:
x.cpp:11:1
Hi Janne,
On 11/18/19 9:34 PM, Janne Blomqvist wrote:
Now that we require a minimum of MPFR 3.1.0+ to build GCC, we can do
some modernization of the MPFR usage in the GFortran frontend.
OK – thanks for the cleanup.
[For reference, those changes were introduced in MPFR 3.0.0, cf.
https://www.
On 20/11/2019 14:33, Martin Liška wrote:
> On 11/13/19 4:24 PM, Matthew Malcomson wrote:
>> On 12/11/2019 12:08, Martin Liška wrote:
>>> On 11/11/19 5:03 PM, Matthew Malcomson wrote:
Ah!
My apologies -- I sent up a series with a few documentation mistakes.
>
>>
>> b) Marking 'ptr' and 'm
Hi,
sadly this patch seems to trigger false positive
../../gcc/vec.h:311:10: error: attempt to free a non-heap object �@Xa�@Y
[-Werror=free-nonheap-object]
which is about vec destructor freeing heap allocated memory if the
vector was ever resized from hits stack allocated place. In this case
it
Allowing mixed vector sizes broke the assumption in the following assert,
since it's now possible for different accesses to require different
levels of alignment:
/* FORNOW: use the same mask to test all potentially unaligned
references in the loop. The vectorizer c
Hi!
On Wed, Nov 20, 2019 at 09:42:46AM +, Richard Sandiford wrote:
> Segher Boessenkool writes:
> >> Before I resubmit, why is the simplify-rtx.c part all wrong?
> >
> > It was nice and simple, and it isn't anymore. 8 4 2 1 for the four of
> > lt gt eq un are hardly worth documenting or maki
bb-slp-40.c fails on SPARC targets without VIS4 because it
requires addition on vectors of bytes. There doesn't seem to be
an existing target selector for this, so I added vect_char_add.
(Wasn't sure whether to use vect_char_add, for consistency
with vect_no_int_add/vect_int_mult etc., or vect_add
1 - 100 of 164 matches
Mail list logo
