The variable omp_atv_sequential was replaced by omp_atv_serialized in OpenMP
5.1. This was already implemented by Jakub (C/C++, commit ea82325afe) and
Tobias (Fortran, commit fff15bad1a).
This patch adds two tests to check if omp_atv_serialized is available (one test
for C/C++ and one for Fortran). Besides that omp_atv_sequential is marked as
deprecated in C/C++ and Fortran for OpenMP 5.1.
libgomp/ChangeLog:
* allocator.c (omp_init_allocator): Replace omp_atv_sequential with
omp_atv_serialized.
* omp.h.in: Add deprecated flag for omp_atv_sequential.
* omp_lib.f90.in: Add deprecated flag for omp_atv_sequential.
* testsuite/libgomp.c-c++-common/alloc-10.c: New test.
* testsuite/libgomp.fortran/alloc-12.f90: New test.
OpenMP 5.1 adds env vars and functions to set and query new ICVs used
as fallback if thread_limit or num_teams clauses aren't specified on
teams construct.
The following patch implements those, though further work will be needed:
1) OpenMP 5.1 also changed the num_teams clause, so that it can specify
both lower and upper limit for how many teams should be created and
changed the meaning when only one expression is provided, instead of
num_teams(expr) in 5.0 meaning num_teams(1:expr) in 5.1, it now means
num_teams(expr:expr), i.e. while previously we could create 1 to expr
teams, in 5.1 we have some low limit by default equal to the single
expression provided and may not create fewer teams.
For host teams (which we don't currently implement efficiently for
NUMA hosts) we trivially satisfy it now by always honoring what the
user asked for, but for the offloading teams I think we'll need to
rethink the APIs; currently teams construct is just a call that returns
and possibly lowers the number of teams; and whenever possible we try
to evaluate num_teams/thread_limit already on the target construct
and the GOMP_teams call just sets the number of teams to the minimum
of provided and requested teams; for some cases e.g. where target
is not combined with teams and num_teams expression calls some functions
etc., we need to call those functions in the target region and so it is
late to figure number of teams, but also hw could just limit what it
is willing to create; in that case I'm afraid we need to run the target
body multiple times and arrange for omp_get_team_num () returning the
right values
2) we need to finally implement the NUMA handling for GOMP_teams_reg
3) I now realize I haven't added some testcase coverage, will do that
incrementally
4) libgomp.texi needs updates for these new APIs, but also others like
the allocator
2021-10-11 Jakub Jelinek <jakub@redhat.com>
gcc/
* omp-low.c (omp_runtime_api_call): Handle omp_get_max_teams,
omp_[sg]et_teams_thread_limit and omp_set_num_teams.
libgomp/
* omp.h.in (omp_set_num_teams, omp_get_max_teams,
omp_set_teams_thread_limit, omp_get_teams_thread_limit): Declare.
* omp_lib.f90.in (omp_set_num_teams, omp_get_max_teams,
omp_set_teams_thread_limit, omp_get_teams_thread_limit): Declare.
* omp_lib.h.in (omp_set_num_teams, omp_get_max_teams,
omp_set_teams_thread_limit, omp_get_teams_thread_limit): Declare.
* libgomp.h (gomp_nteams_var, gomp_teams_thread_limit_var): Declare.
* libgomp.map (OMP_5.1): Export omp_get_max_teams{,_},
omp_get_teams_thread_limit{,_}, omp_set_num_teams{,_,_8_} and
omp_set_teams_thread_limit{,_,_8_}.
* icv.c (omp_set_num_teams, omp_get_max_teams,
omp_set_teams_thread_limit, omp_get_teams_thread_limit): New
functions.
* env.c (gomp_nteams_var, gomp_teams_thread_limit_var): Define.
(omp_display_env): Print OMP_NUM_TEAMS and OMP_TEAMS_THREAD_LIMIT.
(initialize_env): Handle OMP_NUM_TEAMS and OMP_TEAMS_THREAD_LIMIT env
vars.
* teams.c (GOMP_teams_reg): If thread_limit is not specified, use
gomp_teams_thread_limit_var as fallback if not zero. If num_teams
is not specified, use gomp_nteams_var.
* fortran.c (omp_set_num_teams, omp_get_max_teams,
omp_set_teams_thread_limit, omp_get_teams_thread_limit): Add
ialias_redirect.
(omp_set_num_teams_, omp_set_num_teams_8_, omp_get_max_teams_,
omp_set_teams_thread_limit_, omp_set_teams_thread_limit_8_,
omp_get_teams_thread_limit_): New functions.
This patch adds support for recognizing loops which mimic the behaviour
of functions strlen and rawmemchr, and replaces those with internal
function calls in case a target provides them. In contrast to the
standard strlen and rawmemchr functions, this patch also supports
different instances where the memory pointed to is interpreted as 8, 16,
and 32-bit sized, respectively.
gcc/ChangeLog:
* builtins.c (get_memory_rtx): Change to external linkage.
* builtins.h (get_memory_rtx): Add function prototype.
* doc/md.texi (rawmemchr<mode>): Document.
* internal-fn.c (expand_RAWMEMCHR): Define.
* internal-fn.def (RAWMEMCHR): Add.
* optabs.def (rawmemchr_optab): Add.
* tree-loop-distribution.c (find_single_drs): Change return code
behaviour by also returning true if no single store was found
but a single load.
(loop_distribution::classify_partition): Respect the new return
code behaviour of function find_single_drs.
(loop_distribution::execute): Call new function
transform_reduction_loop in order to replace rawmemchr or strlen
like loops by calls into builtins.
(generate_reduction_builtin_1): New function.
(generate_rawmemchr_builtin): New function.
(generate_strlen_builtin_1): New function.
(generate_strlen_builtin): New function.
(generate_strlen_builtin_using_rawmemchr): New function.
(reduction_var_overflows_first): New function.
(determine_reduction_stmt_1): New function.
(determine_reduction_stmt): New function.
(loop_distribution::transform_reduction_loop): New function.
gcc/testsuite/ChangeLog:
* gcc.dg/tree-ssa/ldist-rawmemchr-1.c: New test.
* gcc.dg/tree-ssa/ldist-rawmemchr-2.c: New test.
* gcc.dg/tree-ssa/ldist-strlen-1.c: New test.
* gcc.dg/tree-ssa/ldist-strlen-2.c: New test.
* gcc.dg/tree-ssa/ldist-strlen-3.c: New test.
__builtin___clear_cache was able to accept constant address for the
argument, but it seems no longer accept recently, and it even not
accept constant address which is hold in variable when optimization is
enable:
```
void foo3(){
void *yy = (void*)0x1000;
__builtin___clear_cache(yy, yy);
}
```
So this patch make BEGIN and END accept VOIDmode, like cselib_lookup_mem did per
Jim Wilson's suggestion.
```
static cselib_val *
cselib_lookup_mem (rtx x, int create)
{
...
addr_mode = GET_MODE (XEXP (x, 0));
if (addr_mode == VOIDmode)
addr_mode = Pmode;
```
Changes v2 -> v3:
- Use gcc_assert rather than error, maybe_emit_call_builtin___clear_cache is
internal use only, and we already checked the type in other place.
Changes v1 -> v2:
- Check is CONST_INT intead of cehck mode, no new testcase, since
constant value with other type like CONST_DOUBLE will catched by
front-end.
e.g.
Code:
```c
void foo(){
__builtin___clear_cache(1.11, 0);
}
```
Error message:
```
clearcache-double.c: In function 'foo':
clearcache-double.c:2:27: error: incompatible type for argument 1 of '__builtin___clear_cache'
2 | __builtin___clear_cache(1.11, 0);
| ^~~~
| |
| double
clearcache-double.c:2:27: note: expected 'void *' but argument is of type 'double'
```
gcc/ChangeLog:
PR target/100316
* builtins.c (maybe_emit_call_builtin___clear_cache): Allow
CONST_INT for BEGIN and END, and use gcc_assert rather than
error.
gcc/testsuite/ChangeLog:
PR target/100316
* gcc.c-torture/compile/pr100316.c: New.
Since my r10-7665-g33c45e51b4914008064d9b77f2c1fc0eea1ad060 change, we get
wrong-debug on e.g. the following testcase at -O2 -g on x86_64-linux for the
x parameter:
void bar (int *r);
int
foo (int x)
{
int r = 0;
bar (&r);
return r;
}
At the start of function, we have
subq $24, %rsp
leaq 12(%rsp), %rdi
instructions. The x parameter is passed in %rdi, but isn't used in the
function and so the leaq instruction overwrites %rdi without remembering
%rdi anywhere. Before the r10-7665 change (which was trying to fix a large
(3% for 32-bit, 1% for 64-bit x86-64) debug info/loc growth introduced with
r10-7515), the leaq insn above resulted in a MO_VAL_SET micro-operation that
said that the value of sp + 12, a cselib_sp_derived_value_p, is stored into
the %rdi register. The r10-7665 change added a change to add_stores that
added no micro-operation for the leaq store, with the rationale that the sp
based values can be and will be always computable some other more compact
and primarily more stable way (cfa based expression like DW_OP_fbreg, that
is the same in the whole function). That is true. But by throwing the
micro-operation on the floor, we miss another important part of the
MO_VAL_SET, in particular that the destination of the store, %rdi in this
case, now has a different value from what it had before, so the vt_*
dataflow code thinks that even after the leaq instruction %rdi still holds
the x argument value (and changes it to DW_OP_entry_value (%rdi) only in the
middle of the call to bar). Previously and with the patches below,
the location for x changes already at the end of leaq instruction to
DW_OP_entry_value (%rdi).
My first attempt to fix this was instead of dropping the MO_VAL_SET add
a MO_CLOBBER operation:
--- gcc/var-tracking.c.jj 2021-05-04 21:02:24.196799586 +0200
+++ gcc/var-tracking.c 2021-09-24 19:23:16.420154828 +0200
@@ -6133,7 +6133,9 @@ add_stores (rtx loc, const_rtx expr, voi
{
if (preserve)
preserve_value (v);
- return;
+ mo.type = MO_CLOBBER;
+ mo.u.loc = loc;
+ goto log_and_return;
}
nloc = replace_expr_with_values (oloc);
so don't track that the value lives in the loc destination, but track
that the previous value doesn't live there anymore. That failed bootstrap
miserably, the vt_* code isn't prepared to see MO_CLOBBER of a MEM that
isn't tracked (e.g. has MEM_EXPR on it that the var-tracking code wants
to track, i.e. track_p in add_stores). On the other side, thinking about
it more, in the most common case where a cselib_sp_derived_value_p value
is stored into the sp register (and which is the reason why PR94495
testcase got larger), dropping the micro-operation on the floor is the
right thing, because we have that cselib_sp_derived_value_p tracking, any
reads from the sp hard register will be treated as
cselib_sp_derived_value_p.
Then I've tried 3 different patches described below and in the end
what is committed is patch2.
Additionally, I've gathered statistics from cc1plus by always reverting the
var-tracking.c change after finished bootstrap/regtest and rebuilding the
stage3 var-tracking.o and cc1plus, such that it would be comparable.
dwlocstat and .debug_{info,loclists} section sizes detailed below.
patch3 uses MO_VAL_SET (i.e. essentially reversion of the r10-7665
change) when destination is not a REG_P and !track_p, otherwise if
destination is sp drops the micro-operation on the floor (i.e. no change),
otherwise adds a MO_CLOBBER.
patch1 is similar, except it checks for destination not equal to sp and
!track_p, i.e. for !track_p REG_P destinations other than sp it will use
MO_VAL_SET rather than MO_CLOBBER.
Finally, patch2, the shortest patch, uses MO_VAL_SET whenever destination
is not sp and otherwise drops the micro-operation on the floor.
All the 3 patches don't affect the PR94495 testcase, all the changes
there were caused by stores of sp based values into %rsp.
While the patch2 (and patch1 which results in exactly the same sizes)
causes the largest debug loclists/info growth from the 3, it is still quite
minor (0.651% on 64-bit and 0.114% on 32-bit) compared
to the 1% and 3% PR94495 was trying to solve, and I actually think it is the
best thing to do. Because, if we have say
int q[10];
int *p = &q[0];
or similar and we load the &q[0] sp based value into some hard register,
by noting in the debug info that p lives in some hard reg for some part
of the function and a user is trying to change the p var in the debugger,
if we say it lives in some register or memory, there is some chance that
the changing of the value could work successfully (of course, nothing
is guaranteed, we don't have tracking of where each var lives at which
moment for changing purposes (i.e. what register, memory or else you need
to change in order to change behavior of the code)), while if we just say
that p's location is DW_OP_fbreg 16 DW_OP_stack_value, that is a read-only
value one can just print but not change. Now, for stores of variable
values into the sp register, I don't think we have such an issue, you don't
want debugger to change your stack pointer when user asks to change value
of some variable whose value lives in the stack pointer, that would pretty
much always result in misbehavior of the program.
So, my preference from these 3 is patch2 and that is being committed.
64-bit cc1plus
==============
vanilla
cov% samples cumul
0..10 1064665/37% 1064665/37%
11..20 35972/1% 1100637/38%
21..30 47969/1% 1148606/40%
31..40 45787/1% 1194393/42%
41..50 57529/2% 1251922/44%
51..60 53974/1% 1305896/46%
61..70 112055/3% 1417951/50%
71..80 79420/2% 1497371/52%
81..90 126225/4% 1623596/57%
91..100 1206682/42% 2830278/100%
[34] .debug_info PROGBITS 0000000000000000 2f1c74c a44949f 00 0 0 1
[38] .debug_loclists PROGBITS 0000000000000000 ff5d046 506e947 00 0 0 1
patch1 (same as patch2)
cov% samples cumul
0..10 1064685/37% 1064685/37%
11..20 36011/1% 1100696/38%
21..30 47975/1% 1148671/40%
31..40 45799/1% 1194470/42%
41..50 57566/2% 1252036/44%
51..60 54011/1% 1306047/46%
61..70 112068/3% 1418115/50%
71..80 79421/2% 1497536/52%
81..90 126171/4% 1623707/57%
91..100 1206571/42% 2830278/100%
[34] .debug_info PROGBITS 0000000000000000 2f1c74c a448f27 00 0 0 1
[38] .debug_loclists PROGBITS 0000000000000000 ff608bc 52070dd 00 0 0 1
patch3
cov% samples cumul
0..10 1064698/37% 1064698/37%
11..20 36018/1% 1100716/38%
21..30 47977/1% 1148693/40%
31..40 45804/1% 1194497/42%
41..50 57562/2% 1252059/44%
51..60 54018/1% 1306077/46%
61..70 112071/3% 1418148/50%
71..80 79424/2% 1497572/52%
81..90 126172/4% 1623744/57%
91..100 1206534/42% 2830278/100%
[34] .debug_info PROGBITS 0000000000000000 2f1c74c a449548 00 0 0 1
[38] .debug_loclists PROGBITS 0000000000000000 ff5df39 507acd8 00 0 0 1
So, size of .debug_info+.debug_loclists grows for vanilla -> patch1 (or patch2) by
0.651% and for vanilla -> patch3 by 0.020%.
32-bit cc1plus
==============
vanilla
cov% samples cumul
0..10 1061892/37% 1061892/37%
11..20 34002/1% 1095894/39%
21..30 43513/1% 1139407/40%
31..40 41667/1% 1181074/42%
41..50 59144/2% 1240218/44%
51..60 47009/1% 1287227/45%
61..70 105069/3% 1392296/49%
71..80 72990/2% 1465286/52%
81..90 125988/4% 1591274/56%
91..100 1208726/43% 2800000/100%
[33] .debug_info PROGBITS 00000000 351ab10 8b1c83d 00 0 0 1
[37] .debug_loclists PROGBITS 00000000 ebc816e 3fe44fd 00 0 0 1
patch1 (same as patch2)
cov% samples cumul
0..10 1061999/37% 1061999/37%
11..20 34065/1% 1096064/39%
21..30 43557/1% 1139621/40%
31..40 41690/1% 1181311/42%
41..50 59191/2% 1240502/44%
51..60 47143/1% 1287645/45%
61..70 105045/3% 1392690/49%
71..80 73021/2% 1465711/52%
81..90 125885/4% 1591596/56%
91..100 1208404/43% 2800000/100%
[33] .debug_info PROGBITS 00000000 351ab10 8b1c597 00 0 0 1
[37] .debug_loclists PROGBITS 00000000 ebca915 401ffad 00 0 0 1
patch3
cov% samples cumul
0..10 1062006/37% 1062006/37%
11..20 34073/1% 1096079/39%
21..30 43559/1% 1139638/40%
31..40 41693/1% 1181331/42%
41..50 59189/2% 1240520/44%
51..60 47142/1% 1287662/45%
61..70 105054/3% 1392716/49%
71..80 73027/2% 1465743/52%
81..90 125874/4% 1591617/56%
91..100 1208383/43% 2800000/100%
[33] .debug_info PROGBITS 00000000 351ab10 8b1c690 00 0 0 1
[37] .debug_loclists PROGBITS 00000000 ebca40a 4020a6e 00 0 0 1
So, size of .debug_info+.debug_loclists grows for vanilla -> patch1 (or patch2) by
0.114% and for vanilla -> patch3 by 0.116%.
2021-10-10 Jakub Jelinek <jakub@redhat.com>
PR debug/102441
* var-tracking.c (add_stores): For cselib_sp_derived_value_p values
use MO_VAL_SET if loc is not sp.
So it turns out this is kinda of a latent bug but not really latent.
In GCC 9 and 10, phi-opt would transform a?-1:0 (even for signed 1-bit integer)
to -(type)a but the type is an one bit integer which means the negation is
undefined. GCC 11 fixed the problem by checking for a?pow2cst:0 transformation
before a?-1:0 transformation.
When I added the transformations to match.pd, I had swapped the order not paying
attention and I didn't expect anything of it. Because there was no testcase failing
due to this.
Anyways this fixes the problem on the trunk by swapping the order in match.pd and
adding a comment of why the order is this way.
I will try to come up with a patch for GCC 9 and 10 series later on which fixes
the problem there too.
Note I didn't include the original testcase which requires the vectorizer and AVX-512f
as I can't figure out the right dg options to restrict it to avx-512f but I did come up
with a testcase which shows the problem and even more shows the problem with the 9/10
series as mentioned.
OK? Bootstrapped and tested on x86_64-linux-gnu.
PR tree-optimization/102622
gcc/ChangeLog:
* match.pd: Swap the order of a?pow2cst:0 and a?-1:0 transformations.
Swap the order of a?0:pow2cst and a?0:-1 transformations.
gcc/testsuite/ChangeLog:
* gcc.c-torture/execute/bitfld-10.c: New test.
F2018:10.1.5.5.1(2) requires the same interpretation of old and new-style
relational operators. As gfortran internally distinguishes between
these versions, we must match equivalent notations in
USE module, ONLY: OPERATOR(op)
statements when reading modules.
gcc/fortran/ChangeLog:
PR fortran/65454
* module.c (read_module): Handle old and new-style relational
operators when used in USE module, ONLY: OPERATOR(op).
gcc/testsuite/ChangeLog:
PR fortran/65454
* gfortran.dg/interface_operator_3.f90: New test.
This fixes the typos introduced by commit r12-4240.
The dg-warning format looks like:
{ dg-warning regexp
[comment [{ target/xfail selector } [line] ]] }
Some dg-warnings such as:
{ dg-warning "\\\[-Wstringop-overflow"
{ target { i?86-*-* x86_64-*-* } } }
miss the comment field, it makes target selector not take effect.
For targets which are not { i?86-*-* x86_64-*-* }, this kind of cases
fail or pass unexpectedly.
gcc/testsuite/ChangeLog:
* c-c++-common/Wstringop-overflow-2.c: Add missing comment.
* gcc.dg/Warray-bounds-51.c: Likewise.
* gcc.dg/Warray-parameter-3.c: Likewise.
* gcc.dg/Wstringop-overflow-14.c: Likewise.
* gcc.dg/Wstringop-overflow-21.c: Likewise.
* gcc.dg/Wstringop-overflow-76.c: Likewise.
Related to this is the addition of structured-block-sequence in OpenMP 5.1,
which doesn't change anything for Fortran, but for C/C++ allows multiple
statements instead of just one possibly compound around the separating
directives (section and scan).
I've also made some updates to the OpenMP 5.1 support list in libgomp.texi.
2021-10-09 Jakub Jelinek <jakub@redhat.com>
gcc/c/
* c-parser.c (c_parser_omp_structured_block_sequence): New function.
(c_parser_omp_scan_loop_body): Use it.
(c_parser_omp_sections_scope): Likewise.
gcc/cp/
* parser.c (cp_parser_omp_structured_block): Remove disallow_omp_attrs
argument.
(cp_parser_omp_structured_block_sequence): New function.
(cp_parser_omp_scan_loop_body): Use it.
(cp_parser_omp_sections_scope): Likewise.
gcc/testsuite/
* c-c++-common/gomp/sections1.c (foo): Don't expect errors on
multiple statements in between section directive(s). Add testcases
for invalid no statements in between section directive(s).
* gcc.dg/gomp/sections-2.c (foo): Don't expect errors on
multiple statements in between section directive(s).
* g++.dg/gomp/sections-2.C (foo): Likewise.
* g++.dg/gomp/attrs-6.C (foo): Add testcases for multiple
statements in between section directive(s).
(bar): Add testcases for multiple statements in between scan
directive.
* g++.dg/gomp/attrs-7.C (bar): Adjust expected error recovery.
libgomp/
* libgomp.texi (OpenMP 5.1): Mention implemented support for
structured block sequences in C/C++. Mention support for
unconstrained/reproducible modifiers on order clause.
Mention partial (C/C++ only) support of extentensions to atomics
construct. Mention partial (C/C++ on clause only) support of
align/allocator modifiers on allocate clause.
We don't need to have <wchar.h> support in order to delete overloads
for inserting wide characters into narrow streams.
libstdc++-v3/ChangeLog:
PR libstdc++/98725
* include/std/ostream (operator<<(basic_ostream<char, Tr>&, wchar_t))
(operator<<(basic_ostream<char, Tr>&, const wchar_t*)): Always
define as deleted. Do not check _GLIBCXX_USE_WCHAR_T.
The wchar_t type is defined unconditionally for C++, so there is no
reason for std::wstring_convert and std::wbuffer_convert to be disabled
when <wchar.h> is not usable. It should be possible to use those class
templates with char16_t and char32_t even if wchar_t conversions don't
work.
libstdc++-v3/ChangeLog:
PR libstdc++/98725
* include/bits/locale_conv.h (wstring_convert, wbuffer_convert):
Define unconditionally. Do not check _GLIBCXX_USE_WCHAR_T.
None of these traits depend on libc support for wchar_t, so they should
be defined unconditionally. The wchar_t type is always defined in C++.
libstdc++-v3/ChangeLog:
PR libstdc++/98725
* include/c_global/cstddef [!_GLIBCXX_USE_WCHAR_T]
(__byte_operand<wchar_t>): Define specialization.
* include/std/type_traits (__make_signed<wchar_t>)
(__make_unsigned<wchar_t>): Remove redundant check for
__WCHAR_TYPE__ being defined.
* include/tr1/type_traits [!_GLIBCXX_USE_WCHAR_T]
(__is_integral_helper<wchar_t>): Likewise.
None of these vstring specializations depend on libc support for
wchar_t, so can be enabled unconditionally now that char_traits<wchar_t>
is always available.
libstdc++-v3/ChangeLog:
PR libstdc++/98725
* include/ext/rc_string_base.h [!_GLIBCXX_USE_WCHAR_T]
(__rc_string_base<wchar_t>): Define member function.
* include/ext/vstring.h [!_GLIBCXX_USE_WCHAR_T]
(hash<__gnu_cxx::__wvstring>): Define specialization.
* include/ext/vstring_fwd.h [!_GLIBCXX_USE_WCHAR_T] (__wvstring)
(__wsso_string, __wrc_string): Declare typedefs.
The wstring and wstring_view typedefs should be enabled even if
<wchar.h> isn't supported, because char_traits<wchar_t> works
unconditionally. Similarly, the std::hash specializations for wide
strings do not depend on <wchar.h> support.
Although the primary template works OK for std::char_traits<wchar_t> in
the absence of <wchar.h> support, this patch still defines it as an
explicit specialization for compatibility with declarations that expect
it to be specialized. The explicit specialization just uses the same
__gnu_cxx::char_traits base class as the primary template.
libstdc++-v3/ChangeLog:
PR libstdc++/98725
* include/bits/char_traits.h (char_traits<wchar_t>): Define
explicit specialization unconditionally.
* include/bits/basic_string.h (hash<wstring>): Define
unconditionally. Do not check _GLIBCXX_USE_WCHAR_T.
* include/bits/stringfwd.h (wstring): Likewise.
* include/debug/string (wstring): Likewise.
* include/experimental/string_view (experimental::wstring_view)
(hash<experimental::wstring_view>): Likewise.
* include/std/string (pmr::wstring, hash<pmr::wstring>):
Likewise.
* include/std/string_view (wstring_view, hash<wstring_view>):
Likewise.
This fixes a FAIL when --disable-wchar_t is used.
libstdc++-v3/ChangeLog:
* testsuite/27_io/basic_filebuf/close/81256.cc: Moved to...
* testsuite/27_io/basic_filebuf/close/wchar_t/81256.cc: ...here.
This avoids the tuple-like API for std::pair in the unordered
containers, removing some overly generic code.
The _Select1st projection can figure out the member types of a std::pair
without using decltype(std::get<0>(...)).
We don't need _Select2nd because it's only needed in
_NodeBuilder::_S_build, and that can just access the .second member of
the pair directly. The return type of that function doesn't need to be
deduced by decltype, we can just expose the __node_type typedef of the
node generator.
libstdc++-v3/ChangeLog:
* include/bits/hashtable_policy.h (_Select1st): Replace use of
std::get.
(_Select2nd): Remove.
(_NodeBuilder::_S_build): Use _NodeGenerator::__node_type
typedef instead of deducing it. Remove unnecessary piecewise
construction.
(_ReuseOrAllocNode): Make __node_type public.
(_Map_base): Adjust partial specialization to be able to extract
the mapped_type without using tuple_element.
(_Map_base::at): Define inline
* testsuite/23_containers/unordered_map/requirements/53339.cc:
Remove XFAIL.
* testsuite/23_containers/unordered_multimap/requirements/53339.cc:
Likewise.
This is a step towards restoring support for incomplete types in
unordered containers (PR 53339).
We do not need to instantiate the node type to get its value_type
member, because we know that the value type is the first template
parameter. We can deduce that template argument using a custom trait and
a partial specialization for _Hash_node. If we wanted to support custom
hash node types we could still use typename _Tp::value_type in the
primary template of that trait, but that seems unnecessary.
The other change needed is to defer a static assert at class scope, so
that it is done when the types are complete. We must have a complete
type in the destructor, so we can do it there instead.
libstdc++-v3/ChangeLog:
* include/bits/hashtable.h: Move static assertion to destructor.
* include/bits/hashtable_policy.h: Deduce value type from node
type without instantiating it.
2021-10-08 Sandra Loosemore <sandra@codesourcery.com>
PR fortran/54753
gcc/fortran/
* interface.c (gfc_compare_actual_formal): Add diagnostic
for F2018:C839. Refactor shared code and fix bugs with class
array info lookup, and extend similar diagnostic from PR94110
to also cover class types.
gcc/testsuite/
* gfortran.dg/c-interop/c535c-1.f90: Rewrite and expand.
* gfortran.dg/c-interop/c535c-2.f90: Remove xfails.
* gfortran.dg/c-interop/c535c-3.f90: Likewise.
* gfortran.dg/c-interop/c535c-4.f90: Likewise.
* gfortran.dg/PR94110.f90: Extend to cover class types.
In the PR test case SImode was used to split live range of cx on x86-64
because it was the biggest mode for this hard reg in the function. But
all 64-bits of cx contain structure members. We need always to use at least
natural mode of hard reg in splitting to fix this problem.
gcc/ChangeLog:
PR rtl-optimization/102627
* lra-constraints.c (split_reg): Use at least natural mode of hard reg.
gcc/testsuite/ChangeLog:
PR rtl-optimization/102627
* gcc.target/i386/pr102627.c: New test.
Add a #error directive to ensure that the definitions are not compiled
as C++17, which would prevent them being emitted.
libstdc++-v3/ChangeLog:
PR libstdc++/98725
* src/c++11/limits.cc: Fail if __cpp_inline_variables is
defined.
The strlen pass changes the IL as it works with the ranger. This
causes the non_null_ref code to sometimes get asked questions about new
SSA names.
Tested on x86-64 Linux.
gcc/ChangeLog:
* gimple-range-cache.cc (non_null_ref::non_null_deref_p): Grow
bitmap if needed.
The <bits/ranges_algobase.h> header doesn't need the stream and
streambuf iterators, so don't include the whole of <iterator>.
libstdc++-v3/ChangeLog:
PR libstdc++/92546
* include/bits/ranges_algobase.h: Replace <iterator> with a
subset of the headers it includes.
In g:62acc72a957b5614 I'd stopped the unroller from using
an epilogue loop in cases where the iteration count was
known to be a multiple of the unroll factor. The epilogue
and non-epilogue cases still shared this (preexisting) code
to update the edge frequencies:
basic_block exit_bb = single_pred (loop->latch);
new_exit = find_edge (exit_bb, rest);
new_exit->probability = profile_probability::always ()
.apply_scale (1, new_est_niter + 1);
[etc]
But of course (in hindsight) that only makes sense for the
epilogue case, where we've already moved the main loop's exit edge
to be a sibling of the latch edge. For the non-epilogue case,
the exit edge stays (and needs to stay) in its original position.
I don't really understand what the code is trying to do for
the epilogue case. It has:
/* Ensure that the frequencies in the loop match the new estimated
number of iterations, and change the probability of the new
exit edge. */
profile_count freq_h = loop->header->count;
profile_count freq_e = (loop_preheader_edge (loop))->count ();
if (freq_h.nonzero_p ())
{
...
scale_loop_frequencies (loop, freq_e.probability_in (freq_h));
}
Here, freq_e.probability_in (freq_h) is freq_e / freq_h, so for the
header block, this has the effect of:
new header count = freq_h * (freq_e / freq_h)
i.e. we say that the header executes exactly as often as the
preheader edge, which would only make sense if the loop never
iterates. Also, after setting the probability of the nonexit edge
(correctly) to new_est_niter / (new_est_niter + 1), the code does:
scale_bbs_frequencies (&loop->latch, 1, prob);
for this new probability. I think that only makes sense if the
nonexit edge was previously unconditional (100%). But the code
carefully preserved the probability of the original exit edge
when creating the new one.
All I'm trying to do here though is fix the mess I created
and get the probabilities right for the non-epilogue case.
Things are simpler there since we don't have to worry about
loop versioning. Hopefully the comments explain the approach.
The function's current interface implies that it can cope with
multiple exit edges and that the function only needs the iteration
count relative to one of those edges in order to work correctly.
In practice that's not the case: it assumes there is exactly one
exit edge and all current callers also ensure that the exit test
dominates the latch. I think the function is easier to follow
if we remove the implied generality.
gcc/
PR tree-optimization/102385
* predict.h (change_edge_frequency): Declare.
* predict.c (change_edge_frequency): New function.
* tree-ssa-loop-manip.h (tree_transform_and_unroll_loop): Remove
edge argument.
(tree_unroll_loop): Likewise.
* gimple-loop-jam.c (tree_loop_unroll_and_jam): Update accordingly.
* tree-predcom.c (pcom_worker::tree_predictive_commoning_loop):
Likewise.
* tree-ssa-loop-prefetch.c (loop_prefetch_arrays): Likewise.
* tree-ssa-loop-manip.c (tree_unroll_loop): Likewise.
(tree_transform_and_unroll_loop): Likewise. Use single_dom_exit
to retrieve the exit edges. Make all the old profile update code
conditional on !single_loop_p -- the case it was written for --
and use a different approach for the single-loop case.
gcc/testsuite/
* gcc.dg/pr102385.c: New test.
This is a missing piece of the C++20 <chrono> header.
It would be good to move the code into the compiled library, so that we
don't need <sstream> in <chrono>. It could also use spanstream in C++20,
to avoid memory allocations. That can be changed at a later date.
libstdc++-v3/ChangeLog:
* include/std/chrono (__detail::__units_suffix_misc): New
helper function.
(__detail::__units_suffix): Likewise.
(chrono::operator<<(basic_ostream&, const duration&)): Define.
* testsuite/20_util/duration/io.cc: New test.
The introduction of DECL_LOCAL_DECL_ALIAS and push_local_extern_decl_alias
in r11-3699-g4e62aca0e0520e4ed2532f2d8153581190621c1a broke the following
testcase. The following patch fixes it by treating similarly not just
the variable to or link clause is put on, but also its DECL_LOCAL_DECL_ALIAS
if any. If it hasn't been created yet, when it is created it will copy
attributes and therefore should get it for free, and as it is an extern,
nothing more than attributes is needed for it.
2021-10-08 Jakub Jelinek <jakub@redhat.com>
PR c++/102640
gcc/cp/
* parser.c (handle_omp_declare_target_clause): New function.
(cp_parser_omp_declare_target): Use it.
gcc/testsuite/
* c-c++-common/gomp/pr102640.c: New test.
As reported by Sunil's tester, -march=cascadelake triggers some SUBREG
non-determinacy in the generated assembler for my new tests. Fixed
by updating the regular expressions to match either the zero or sign
extended forms. I'm testing a backend patch that may help with the
underlying cause of these differences.
2021-10-08 Roger Sayle <roger@nextmovesoftware.com>
gcc/testsuite/ChangeLog
* gcc.target/i386/sse2-mmx-paddsb-2.c: Test for -128 or 128.
* gcc.target/i386/sse2-mmx-paddusb-2.c: Test for -1 or 255.
* gcc.target/i386/sse2-mmx-psubsb-2.c: Test for -128 or 128.
gcc/ChangeLog:
PR target/102464
* config/i386/i386.c (ix86_optab_supported_p):
Return true for HFmode.
* match.pd: Simplify (_Float16) ceil ((double) x) to
__builtin_ceilf16 (a) when a is _Float16 type and
direct_internal_fn_supported_p.
gcc/testsuite/ChangeLog:
* gcc.target/i386/pr102464.c: New test.
We're performing the [temp.param]/10 adjustment at parse time but not
also at substitution time.
PR c++/61355
gcc/cp/ChangeLog:
* pt.c (convert_template_argument): Perform array/function to
pointer conversion on the substituted type of an NTTP.
gcc/testsuite/ChangeLog:
* g++.old-deja/g++.pt/nontype5.C: Adjust.
* g++.dg/template/param6.C: New test.
This moves the "classic" contents of <chrono> to a new header, so that
<future>, <thread> etc. can get use durations and clocks without
calendar types, time zones, and chrono I/O.
libstdc++-v3/ChangeLog:
* include/Makefile.am: Add new header.
* include/Makefile.in: Regenerate.
* include/std/chrono (duration, time_point, system_clock)
(steady_clock, high_resolution_clock, chrono_literals, sys_time)
(file_clock, file_time): Move to ...
* include/bits/chrono.h: New file.
* include/bits/atomic_futex.h: Include new header instead of
<chrono>.
* include/bits/atomic_timed_wait.h: Likewise.
* include/bits/fs_fwd.h: Likewise.
* include/bits/semaphore_base.h: Likewise.
* include/bits/this_thread_sleep.h: Likewise.
* include/bits/unique_lock.h: Likewise.
* include/experimental/bits/fs_fwd.h: Likewise.
* include/experimental/chrono: Likewise.
* include/experimental/io_context: Likewise.
* include/experimental/netfwd: Likewise.
* include/experimental/timer: Likewise.
* include/std/condition_variable: Likewise.
* include/std/mutex: Likewise.
* include/std/shared_mutex: Likewise.
Marcel Vollweiler
Jakub Jelinek
Stefan Schulze Frielinghaus
Martin Liska
Kito Cheng
liuhongt
GCC Administrator
Harald Anlauf
Andrew Pinski
Kewen Lin
Jonathan Wakely
Sandra Loosemore
Vladimir N. Makarov
Aldy Hernandez
H.J. Lu
Richard Sandiford
Roger Sayle
Ian Lance Taylor
Patrick Palka