[Bug c++/102799] decltype with lambda without body error cause ICE
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102799 康桓瑋 changed: What|Removed |Added CC||hewillk at gmail dot com --- Comment #2 from 康桓瑋 --- dup of PR99505 Comment 2.
[Bug c/102800] New: Incorrect UB warning with aggressive-loop-optimizations
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102800
Bug ID: 102800
Summary: Incorrect UB warning with
aggressive-loop-optimizations
Product: gcc
Version: 11.1.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
Assignee: unassigned at gcc dot gnu.org
Reporter: phil at phord dot com
Target Milestone: ---
I found this spurious warning occurs since gcc 11.1. It does not occur in 10.3
or previous versions that I tested.
I believe this is a different bug from PR100801 because that bug started
appearing in 9.1.
This bug seems to be very sensitive to changes from this minimal code:
[ Source ]-
/* Compiled with: gcc-11 -O2 -c test.c */
void bar(char x);
void foo(char * begin)
{
// end = ALIGN(begin, 8);
char * end = begin + 7 - unsigned long)begin) - 1) % 8);
if (begin != end) {
long long x = end - begin;
do bar(x);
while(--x != 0);
}
}
static char buff[1024] = { 0 };
void test(int i)
{
foo(buff + i * 8);
}
[ Compiler output ]-
In function 'void foo(char*)',
inlined from 'void test(int)' at :16:8:
:9:19: warning: iteration 0x8000 invokes undefined behavior
[-Waggressive-loop-optimizations]
9 | while(--x != 0);
| ^~~~
:9:19: note: within this loop
Notice we don't enter the loop unless x will be greater than zero because of
the `begin != end` check, and because end >= begin due to the ALIGN math.
The compiled code appears to be correct. Only the spurious warning is a
problem. (In fact, the compiled code is optimized out where the warning is
generated because the inlined function has nothing to do. But the code is
called from elsewhere, so size and correctness matter, of course.)
I found several workarounds:
> Change x to unsigned. Produces exact same obj output but no warning:
unsigned long long x = end - begin;
> Change x to short (int16_t). Produces exact same obj output but no warning:
short x = end - begin;
> Compare with > 0 instead of != 0. Compiled output is 1 opcode longer:
while(--x > 0);
> Use calculated x value to determine whether to enter loop.
long long x = end - begin;
if (x > 0)
do bar(x);
while(--x != 0);
I just noticed this bug (spurious warning) appears in 9.3 and earlier if I
change this line from != to <:
if (begin < end) {
I guess it is this subtle difference that I think is a new bug. But maybe this
is a dup of PR100801 after all.
[Bug tree-optimization/102798] [9/10/11/12 Regression] wrong code with -O3 -fno-tree-pta by r9-2475
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102798 --- Comment #8 from H.J. Lu --- Vectorizer has if (DR_PTR_INFO (dr) && TREE_CODE (addr_base) == SSA_NAME && !SSA_NAME_PTR_INFO (addr_base)) vect_duplicate_ssa_name_ptr_info (addr_base, dr_info); This fixes the crash. diff --git a/gcc/tree-vect-data-refs.c b/gcc/tree-vect-data-refs.c index 1e13148190c..fca5aebfbdb 100644 --- a/gcc/tree-vect-data-refs.c +++ b/gcc/tree-vect-data-refs.c @@ -4785,7 +4785,8 @@ vect_create_addr_base_for_vector_ref (vec_info *vinfo, stmt_vec_info stmt_info, if (DR_PTR_INFO (dr) && TREE_CODE (addr_base) == SSA_NAME - && !SSA_NAME_PTR_INFO (addr_base)) + && !SSA_NAME_PTR_INFO (addr_base) + && TREE_CODE (SSA_NAME_VAR (addr_base)) != PARM_DECL) vect_duplicate_ssa_name_ptr_info (addr_base, dr_info); if (dump_enabled_p ())
[Bug ada/100486] Ada build fails for 32bit Windows
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100486 --- Comment #53 from Óscar Fuentes --- (In reply to Christoph Reiter from comment #52) > Turns out this might be fallout from the last grep update, see > https://github.com/msys2/MINGW-packages/issues/9771#issuecomment-945007372 > Needs investigating.. "this" means the problem with C++ exceptions, which indeed is fixed by building gcc after reverting to grep 3.0. The build failure with Ada, which predated the grep upgrade, persists. Eric: I'm sorry for the time you wasted on this wild goose chase.
[Bug tree-optimization/102798] [9/10/11/12 Regression] wrong code with -O3 -fno-tree-pta by r9-2475
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102798 H.J. Lu changed: What|Removed |Added CC||aldyh at gcc dot gnu.org --- Comment #7 from H.J. Lu --- Looks like a VRP bug. For: [local count: 168730857]: if (in_16(D) != 0B) goto ; [70.00%] else goto ; [30.00%] [local count: 118111600]: _1 = size_17(D) + 18446744073709551615; out_26 = out_18(D) + _1; goto ; [100.00%] VRP reports: in_16(D) : BACK visiting block 3 for in_16(D) 2->3 has cache, const unsigned char * VARYING, update. FWD visiting block 3 for in_16(D) starting range : UNDEFINED edge 2->3 :const unsigned char * [1B, +INF] Updating range to const unsigned char * [1B, +INF] Updating blocks : DONE visiting blocks for in_16(D)
[Bug c++/102799] decltype with lambda without body error cause ICE
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102799 --- Comment #1 from qingzhe huang --- Forget to mention that obviously this only happens with "-std=c++20" because of lambda in unevaluated context support.
[Bug c++/102799] New: decltype with lambda without body error cause ICE
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102799
Bug ID: 102799
Summary: decltype with lambda without body error cause ICE
Product: gcc
Version: 12.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c++
Assignee: unassigned at gcc dot gnu.org
Reporter: nickhuang99 at hotmail dot com
Target Milestone: ---
The lambda in outside "decltype" forget to add lambda body which cause ICE
where clang clearly points out this root cause. MSVC++ at least reports error
without crash.
decltype([](decltype([]{}))) gibish;
g++: internal compiler error: Segmentation fault signal terminated program
cc1plus
Please submit a full bug report,
with preprocessed source if appropriate.
See
[Bug ada/100486] Ada build fails for 32bit Windows
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100486 --- Comment #52 from Christoph Reiter --- Turns out this might be fallout from the last grep update, see https://github.com/msys2/MINGW-packages/issues/9771#issuecomment-945007372 Needs investigating..
[Bug bootstrap/102681] [12 Regression] AArch64 bootstrap failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102681 Andrew Pinski changed: What|Removed |Added See Also||https://gcc.gnu.org/bugzill ||a/show_bug.cgi?id=102794 --- Comment #6 from Andrew Pinski --- I have not looked into the IR here but what could be happening is jump threading is happening through the loop header and uninitialized code is getting so confused. The uninitialized code with respect to conditionals is so fragile with respect to jump thread; there are other bugs which show that.
[Bug target/102639] ICE in extract_insn, at recog.c:2769 since r12-3406-ga68412117fa47786
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102639 Andrew Pinski changed: What|Removed |Added Resolution|--- |FIXED Status|NEW |RESOLVED Target Milestone|--- |12.0 --- Comment #5 from Andrew Pinski --- .
[Bug tree-optimization/102798] [9/10/11/12 Regression] wrong code with -O3 -fno-tree-pta by r9-2475
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102798 --- Comment #6 from H.J. Lu --- A pointer is known to non-null only if we know where the pointer is pointing to. Since the null field is initialized to 0, we need to check both null and anything. This works on the test case: diff --git a/gcc/value-query.cc b/gcc/value-query.cc index ab133aab114..1e8e3b96ea7 100644 --- a/gcc/value-query.cc +++ b/gcc/value-query.cc @@ -302,7 +302,7 @@ get_ssa_name_ptr_info_nonnull (const_tree name) When PTA analysis is improved, pt.anything, pt.nonlocal and pt.escaped may also has to be considered before deciding that pointer cannot point to NULL. */ - return !pi->pt.null; + return !pi->pt.null && !pi->pt.anything; } // Update the global range for NAME into the SSA_RANGE_NAME_INFO and
[Bug tree-optimization/102798] [9/10/11/12 Regression] wrong code with -O3 -fno-tree-pta by r9-2475
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102798
Andrew Pinski changed:
What|Removed |Added
Summary|[9/10/11/12 Regression] |[9/10/11/12 Regression]
|wrong code with -O3 |wrong code with -O3
|-fno-tree-pta -mavx512f by |-fno-tree-pta by r9-2475
|r9-2475 |
--- Comment #5 from Andrew Pinski ---
Here is a testcase which fails without -mavx512
typedef __SIZE_TYPE__ size_t;
__attribute__((__noipa__))
void BUF_reverse (unsigned char *out, const unsigned char *in, size_t size)
{
size_t i;
if (in)
{
out += size - 1;
for (i = 0; i < size; i++)
*out++ = *in++;
}
else
{
unsigned char *q;
char c;
q = out + size - 1;
for (i = 0; i < size ; i++)
{
*out++ = 1;
}
}
}
int
main (void)
{
unsigned char buf[40];
unsigned char buf1[40];
for (unsigned i = 0; i < sizeof (buf); i++)
buf[i] = i;
BUF_reverse (buf, 0, sizeof (buf));
for (unsigned i = 0; i < sizeof (buf); i++)
if (buf[i] != 1)
__builtin_abort ();
return 0;
}
[Bug tree-optimization/102798] [9/10/11/12 Regression] wrong code with -O3 -fno-tree-pta -mavx512f by r9-2475
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102798 Andrew Pinski changed: What|Removed |Added Target Milestone|--- |9.5
[Bug tree-optimization/102798] [9/10/11/12 Regression] wrong code with -O3 -fno-tree-pta -mavx512f by r9-2475
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102798 Andrew Pinski changed: What|Removed |Added Keywords||alias --- Comment #4 from Andrew Pinski --- Before the vectorizer: unsigned charD.28 * out_18(D) = outD.3197; const unsigned charD.28 * in_16(D) = inD.3198; size_tD.3196 size_17(D) = sizeD.3199; After: unsigned charD.28 * out_18(D) = outD.3197; # PT = anything const unsigned charD.28 * in_16(D) = inD.3198; size_tD.3196 size_17(D) = sizeD.3199; That seems to be causing the issue.
[Bug tree-optimization/102798] [9/10/11/12 Regression] wrong code with -O3 -fno-tree-pta -mavx512f by r9-2475
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102798 --- Comment #3 from H.J. Lu --- Visiting conditional with predicate: if (in_16(D) != 0B) With known ranges in_16(D): const unsigned char * [1B, +INF] 1B for lower bound is wrong.
[Bug tree-optimization/102798] [9/10/11/12 Regression] wrong code with -O3 -fno-tree-pta -mavx512f by r9-2475
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102798 --- Comment #2 from H.J. Lu --- 192t.thread3 has if (in_16(D) != 0B) goto ; [70.00%] else goto ; [30.00%] 193t.dom3 removed "if (in_16(D) != 0B)".
[Bug tree-optimization/102798] [9/10/11/12 Regression] wrong code with -O3 -fno-tree-pta -mavx512f by r9-2475
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102798
--- Comment #1 from H.J. Lu ---
Source has
__attribute__((__noipa__))
void BUF_reverse (unsigned char *out, const unsigned char *in, size_t size)
{
size_t i;
if (in)
{
out += size - 1;
for (i = 0; i < size; i++)
*out-- = *in++;
}
else
We generate
BUF_reverse:
.LFB0:
.cfi_startproc
movq%rdi, %r8
movq%rsi, %rcx
movq%rdx, %rax
testq %rdx, %rdx <<< RDX has size, not in.
je .L28
[Bug tree-optimization/102798] [9/10/11/12 Regression] wrong code with -O3 -fno-tree-pta -mavx512f by r9-2475
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102798 H.J. Lu changed: What|Removed |Added Ever confirmed|0 |1 Summary|[9/10/11/12 Regression] |[9/10/11/12 Regression] |wrong code with -O3 |wrong code with -O3 |-fno-tree-pta -mavx512f |-fno-tree-pta -mavx512f by ||r9-2475 Status|UNCONFIRMED |NEW CC||rguenth at gcc dot gnu.org Last reconfirmed||2021-10-16 Component|target |tree-optimization
[Bug tree-optimization/102796] [12 Regresson] ICE in useless_type_conversion_p at gcc/gimple-expr.c:87 since r12-4443-g93ac832f1846e4867aa6537f76f510fab8e3e87d
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102796 H.J. Lu changed: What|Removed |Added CC||hjl.tools at gmail dot com --- Comment #3 from H.J. Lu --- On Linux/x86, it also caused: FAIL: 27_io/ios_base/failure/dual_abi.cc execution test
[Bug tree-optimization/102796] [12 Regresson] ICE in useless_type_conversion_p at gcc/gimple-expr.c:87 since r12-4443-g93ac832f1846e4867aa6537f76f510fab8e3e87d
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102796 --- Comment #2 from Martin Liška --- One more test-case can be seen here: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102797#c2
[Bug tree-optimization/102796] [12 Regresson] ICE in useless_type_conversion_p at gcc/gimple-expr.c:87 since r12-4443-g93ac832f1846e4867aa6537f76f510fab8e3e87d
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102796 Martin Liška changed: What|Removed |Added CC||dcb314 at hotmail dot com --- Comment #1 from Martin Liška --- *** Bug 102797 has been marked as a duplicate of this bug. ***
[Bug middle-end/102797] ice in useless_type_conversion_p, at gimple-expr.c:87
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102797 Martin Liška changed: What|Removed |Added CC||marxin at gcc dot gnu.org Status|NEW |RESOLVED Resolution|--- |DUPLICATE --- Comment #8 from Martin Liška --- Dup. *** This bug has been marked as a duplicate of bug 102796 ***
[Bug target/102767] [12 Regression] ICE in rs6000_builtin_vectorization_cost, at config/rs6000/rs6000.c:5216
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102767 --- Comment #6 from Segher Boessenkool --- (In reply to Richard Earnshaw from comment #5) > We have the type > type size > unit-size > and movmisalign pattern is enabled for this. > > but the vectorization cost doesn't handle the case of elements=1, which is > the case when mode is TImode. > > So I think this is an inconsistency in the rs6000 backend - either add > costing support for single elements or disable the movmisalign code in this > case. But TImode is a scalar type, not a vector type, so it should hit one of the early-outs at the top of rs6000_builtin_vectorization_cost?
[Bug middle-end/102797] ice in useless_type_conversion_p, at gimple-expr.c:87
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102797 David Binderman changed: What|Removed |Added CC||amacleod at redhat dot com --- Comment #7 from David Binderman --- Andrew's commit 93ac832f1846e4867aa6537f76f510fab8e3e87d looks to me to be the culprit. Adding Andrew for best advice.
[Bug middle-end/102797] ice in useless_type_conversion_p, at gimple-expr.c:87
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102797 --- Comment #6 from David Binderman --- Range currently seems to be (a10794eafb151b92, 730f52e05a1fb5c8). Trying 1ba7adabf29eb671. Only 7 revisions left to go.
[Bug middle-end/102797] ice in useless_type_conversion_p, at gimple-expr.c:87
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102797 --- Comment #5 from David Binderman --- (In reply to David Binderman from comment #4) > I am trying a git bisect. I frequently get this wrong ;-< > > commit a10794eafb151b92 is being built. Seems fine, trying 730f52e05a1fb5c8.
[Bug middle-end/102797] ice in useless_type_conversion_p, at gimple-expr.c:87
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102797 --- Comment #4 from David Binderman --- I am trying a git bisect. I frequently get this wrong ;-< commit a10794eafb151b92 is being built.
[Bug middle-end/102797] ice in useless_type_conversion_p, at gimple-expr.c:87
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102797 Andreas Schwab changed: What|Removed |Added Component|c |middle-end Ever confirmed|0 |1 Status|UNCONFIRMED |NEW Last reconfirmed||2021-10-16 --- Comment #3 from Andreas Schwab --- Also breaks libgo. during GIMPLE pass: evrp In function 'cmd/go/internal/modget.resolver.resolveQueries': go1: internal compiler error: tree check: expected class 'type', have 'exceptional' (error_mark) in useless_type_conversion_p, at gimple-expr.c:87 0x76210f tree_class_check_failed(tree_node const*, tree_code_class, char const*, int, char const*) ../../gcc/tree.c:8739 0xa99187 tree_class_check(tree_node*, tree_code_class, char const*, int, char const*) ../../gcc/tree.h:3556 0xa99187 useless_type_conversion_p(tree_node*, tree_node*) ../../gcc/gimple-expr.c:87 0xe30d83 verify_gimple_phi ../../gcc/tree-cfg.c:5128 0xe30d83 verify_gimple_in_cfg(function*, bool) ../../gcc/tree-cfg.c:5457 0xced80f execute_function_todo ../../gcc/passes.c:2042 0xcee2a3 execute_todo ../../gcc/passes.c:2096
[Bug c/102797] ice in useless_type_conversion_p, at gimple-expr.c:87
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102797
--- Comment #2 from Dmitry G. Dyachenko ---
r12-4256 PASS
r12- FAIL
$ cat x.ii
struct b {
b(int);
};
void d() {
int c = 1;
do
try {
b a = 1;
while (1) {
c++;
a = 1;
}
} catch (...) {
}
while (1);
}
$ g++ -fpreprocessed -O2 -std=c++98 -c x.ii
during GIMPLE pass: evrp
x.ii: In function 'void d()':
x.ii:16:1: internal compiler error: tree check: expected class 'type', have
'exceptional' (error_mark) in useless_type_conversion_p, at gimple-expr.c:87
16 | }
| ^
0x90e691 tree_class_check_failed(tree_node const*, tree_code_class, char
const*, int, char const*)
/home/dimhen/src/gcc_current/gcc/tree.c:8739
0x7fcfb1 tree_class_check(tree_node*, tree_code_class, char const*, int, char
const*)
/home/dimhen/src/gcc_current/gcc/tree.h:3556
0x7fcfb1 useless_type_conversion_p(tree_node*, tree_node*)
/home/dimhen/src/gcc_current/gcc/gimple-expr.c:87
0x169677a verify_gimple_phi
/home/dimhen/src/gcc_current/gcc/tree-cfg.c:5128
0x169677a verify_gimple_in_cfg(function*, bool)
/home/dimhen/src/gcc_current/gcc/tree-cfg.c:5457
0x154ad27 execute_function_todo
/home/dimhen/src/gcc_current/gcc/passes.c:2042
0x154b73c execute_todo
/home/dimhen/src/gcc_current/gcc/passes.c:2096
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See
[Bug tree-optimization/102720] [12 regression] gcc.dg/tree-ssa/ldist-strlen-1.c and ldist-strlen-2.c fail after r12-4324
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102720 --- Comment #12 from CVS Commits --- The master branch has been updated by Jan Hubicka : https://gcc.gnu.org/g:99b287b8ef51a0be52f7400879a619dc5f993f31 commit r12-4457-g99b287b8ef51a0be52f7400879a619dc5f993f31 Author: Jan Hubicka Date: Sat Oct 16 14:45:06 2021 +0200 Fix wrong code in ldost-strlen-1.c gcc/ChangeLog: PR tree-optimization/102720 * tree-ssa-structalias.c (compute_points_to_sets): Fix producing of call used and clobbered sets.
[Bug target/102798] New: [9/10/11/12 Regression] wrong code with -O3 -fno-tree-pta -mavx512f
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102798 Bug ID: 102798 Summary: [9/10/11/12 Regression] wrong code with -O3 -fno-tree-pta -mavx512f Product: gcc Version: 12.0 Status: UNCONFIRMED Keywords: wrong-code Severity: normal Priority: P3 Component: target Assignee: unassigned at gcc dot gnu.org Reporter: zsojka at seznam dot cz Target Milestone: --- Host: x86_64-pc-linux-gnu Target: x86_64-pc-linux-gnu Created attachment 51616 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=51616=edit auto-reduced testcase (from OpenSSL sources) $ x86_64-pc-linux-gnu-gcc -O3 -fno-tree-pta -mavx512f testcase.c $ gdb ./a.out ... (gdb) r Starting program: /home/smatz/gcc-bug/58/a.out Program received signal SIGSEGV, Segmentation fault. BUF_reverse (out=, out@entry=0x7fffda60 "", in=, in@entry=0x0, size=size@entry=40) at testcase.c:11 11 *out-- = *in++; (gdb) disas Dump of assembler code for function BUF_reverse: ... 0x00401200 <+96>:and$0xffe0,%r9 0x00401204 <+100>: add%rcx,%r9 0x00401207 <+103>: nopw 0x0(%rax,%rax,1) => 0x00401210 <+112>: vmovdqu (%rdx),%ymm2 0x00401214 <+116>: add$0x20,%rdx 0x00401218 <+120>: sub$0x20,%rdi 0x0040121c <+124>: vperm2i128 $0x1,%ymm2,%ymm2,%ymm0 ... (gdb) info reg rax0x2840 rbx0x4013d04199376 rcx0x0 0 rdx0x0 0 ... $ x86_64-pc-linux-gnu-gcc -v Using built-in specs. COLLECT_GCC=/repo/gcc-trunk/binary-latest-amd64/bin/x86_64-pc-linux-gnu-gcc COLLECT_LTO_WRAPPER=/repo/gcc-trunk/binary-trunk-r12-4456-20211016001627-g93d183a5fff-checking-yes-rtl-df-extra-amd64/bin/../libexec/gcc/x86_64-pc-linux-gnu/12.0.0/lto-wrapper Target: x86_64-pc-linux-gnu Configured with: /repo/gcc-trunk//configure --enable-languages=c,c++ --enable-valgrind-annotations --disable-nls --enable-checking=yes,rtl,df,extra --with-cloog --with-ppl --with-isl --build=x86_64-pc-linux-gnu --host=x86_64-pc-linux-gnu --target=x86_64-pc-linux-gnu --with-ld=/usr/bin/x86_64-pc-linux-gnu-ld --with-as=/usr/bin/x86_64-pc-linux-gnu-as --disable-libstdcxx-pch --prefix=/repo/gcc-trunk//binary-trunk-r12-4456-20211016001627-g93d183a5fff-checking-yes-rtl-df-extra-amd64 Thread model: posix Supported LTO compression algorithms: zlib zstd gcc version 12.0.0 20211016 (experimental) (GCC)
[Bug ada/100486] Ada build fails for 32bit Windows
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100486 --- Comment #51 from Christoph Reiter --- (In reply to Eric Botcazou from comment #50) > > Yes, from 2.36.1 to 2.37, but I've already tried reverting that and it > > didn't help. I'm going to try older versions (of everything) though.. > > This could as well be a miscompilation of the linker. Can you upload ld.exe > somewhere so that I replace mine with it? If you happen to have an tool that supports extracting .tar.zst you could take the .exe from package directly: https://repo.msys2.org/mingw/mingw32/mingw-w64-i686-binutils-2.37-4-any.pkg.tar.zst
[Bug fortran/102787] ICE in new test case gfortran.dg/reshape_shape_2.f90
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102787 Dominique d'Humieres changed: What|Removed |Added Status|ASSIGNED|NEW
[Bug fortran/102787] ICE in new test case gfortran.dg/reshape_shape_2.f90
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102787 Dominique d'Humieres changed: What|Removed |Added Ever confirmed|0 |1 Last reconfirmed||2021-10-16 Status|UNCONFIRMED |ASSIGNED --- Comment #4 from Dominique d'Humieres --- Confirmed on Darwin too.
[Bug c/102797] ice in useless_type_conversion_p, at gimple-expr.c:87
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102797
--- Comment #1 from David Binderman ---
Reduced C code is
glib_autoptr_cleanup_GdkPaintable(struct _GdkPaintable **_ptr) {
glib_autoptr_clear_GdkPaintable(*_ptr);
}
glib_autoptr_clear_GdkRGBA(struct _GdkRGBA *_ptr) {
if (_ptr)
gdk_rgba_free();
}
glib_autoptr_cleanup_GdkRGBA(struct _GdkRGBA **_ptr) {
glib_autoptr_clear_GdkRGBA(*_ptr);
}
gtd_sidebar_list_row_set_property() {
__attribute__((cleanup(glib_autoptr_cleanup_GdkPaintable))) *paintable = 0;
__attribute__((cleanup(glib_autoptr_cleanup_GdkRGBA))) *color = 0;
color = gtd_task_list_get_color();
paintable = gtd_create_circular_paintable();
gtk_image_set_from_paintable();
}
Flag -march=bdver2 not required for the reduced code, only -O2 -fexceptions.
[Bug ada/100486] Ada build fails for 32bit Windows
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100486 --- Comment #50 from Eric Botcazou --- > Yes, from 2.36.1 to 2.37, but I've already tried reverting that and it > didn't help. I'm going to try older versions (of everything) though.. This could as well be a miscompilation of the linker. Can you upload ld.exe somewhere so that I replace mine with it?
[Bug c/102797] New: ice in useless_type_conversion_p, at gimple-expr.c:87
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102797 Bug ID: 102797 Summary: ice in useless_type_conversion_p, at gimple-expr.c:87 Product: gcc Version: 12.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: c Assignee: unassigned at gcc dot gnu.org Reporter: dcb314 at hotmail dot com Target Milestone: --- Created attachment 51615 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=51615=edit gzipped C source code For the attached C code, recent gcc trunk does this: ../src/plugins/task-lists-workspace/gtd-sidebar-list-row.c:264:1: internal compiler error: tree check: expected class ‘type’, have ‘exceptional’ (error_mark) in useless_type_conversion_p, at gimple-expr.c:87 Flags -w -march=bdver2 -O2 -fexceptions required. The bug first seems to occur between git hash be072bfa5bb38171 and 93d183a5fff92d80, so about 27 revisions. I have a reduction running in the other window now.
[Bug tree-optimization/102796] [12 Regresson] ICE in useless_type_conversion_p at gcc/gimple-expr.c:87 since r12-4443-g93ac832f1846e4867aa6537f76f510fab8e3e87d
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102796 Martin Liška changed: What|Removed |Added Ever confirmed|0 |1 Last reconfirmed||2021-10-16 Status|UNCONFIRMED |NEW Known to work||11.2.0 Known to fail||12.0 Target Milestone|--- |12.0 Priority|P3 |P1
[Bug tree-optimization/102796] New: [12 Regresson] ICE in useless_type_conversion_p at gcc/gimple-expr.c:87 since r12-4443-g93ac832f1846e4867aa6537f76f510fab8e3e87d
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102796
Bug ID: 102796
Summary: [12 Regresson] ICE in useless_type_conversion_p at
gcc/gimple-expr.c:87 since
r12-4443-g93ac832f1846e4867aa6537f76f510fab8e3e87d
Product: gcc
Version: 12.0
Status: UNCONFIRMED
Keywords: ice-on-valid-code
Severity: normal
Priority: P3
Component: tree-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: marxin at gcc dot gnu.org
CC: aldyh at gcc dot gnu.org, amacleod at redhat dot com
Target Milestone: ---
The following fails:
$ cat /home/marxin/Programming/testcases/ice.cpp
namespace std {
template
struct initializer_list {
const int* __begin_;
decltype(sizeof(int)) __size_;
};
} // namespace std
struct destroyme1 {};
struct witharg1 {
witharg1(const destroyme1&);
~witharg1();
};
std::initializer_list globalInitList2 = {witharg1(destroyme1()),
witharg1(destroyme1())};
$ ./xg++ -B. /home/marxin/Programming/testcases/ice.cpp -fno-tree-ccp
-fno-tree-fre -fno-tree-forwprop -O3 -c
during GIMPLE pass: evrp
/home/marxin/Programming/testcases/ice.cpp: In function ‘void
__static_initialization_and_destruction_0(int, int)’:
/home/marxin/Programming/testcases/ice.cpp:14:65: internal compiler error: tree
check: expected class ‘type’, have ‘exceptional’ (error_mark) in
useless_type_conversion_p, at gimple-expr.c:87
14 | witharg1(destroyme1())};
| ^
0x178e0aa tree_class_check_failed(tree_node const*, tree_code_class, char
const*, int, char const*)
/home/marxin/Programming/gcc/gcc/tree.c:8739
0x90f81e tree_class_check(tree_node*, tree_code_class, char const*, int, char
const*)
/home/marxin/Programming/gcc/gcc/tree.h:3556
0x90f81e useless_type_conversion_p(tree_node*, tree_node*)
/home/marxin/Programming/gcc/gcc/gimple-expr.c:87
0x149498c verify_gimple_phi
/home/marxin/Programming/gcc/gcc/tree-cfg.c:5128
0x149498c verify_gimple_in_cfg(function*, bool)
/home/marxin/Programming/gcc/gcc/tree-cfg.c:5457
0x1356421 execute_function_todo
/home/marxin/Programming/gcc/gcc/passes.c:2042
0x1355229 do_per_function
/home/marxin/Programming/gcc/gcc/passes.c:1687
0x1356654 execute_todo
/home/marxin/Programming/gcc/gcc/passes.c:2096
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See
[Bug tree-optimization/102679] Failure to optimize out 64-bit multiplication to 32-bit multiplication when possible in circumstances involving modifying a 64-bit variable that gets converted to 32-bit
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102679
Andrew Pinski changed:
What|Removed |Added
Ever confirmed|0 |1
Status|UNCONFIRMED |NEW
Last reconfirmed||2021-10-16
--- Comment #2 from Andrew Pinski ---
So GCC is able to do this in a few cases on the RTL level, just not when
loading from memory.
Take:
#include
int32_t f();
int32_t mac(int32_t *b, int64_t sqr)
{
int64_t t = f();
sqr += t * t;
return sqr;
}
CUT
GCC produces exactly what you wanted:
callf()
imull %eax, %eax
addl%ebx, %eax
Even though at the gimple level we get:
t_5 = (int64_t) _4;
_1 = t_5 * t_5;
sqr_7 = _1 + sqr_6(D);
_8 = (int32_t) sqr_7;
Of course this gets more complex.
Maybe someone can help me understand combine for a second we for the 2->2 case,
we don't do the transformation if the mode changes but the two operations stay
the same:
Trying 10 -> 12:
10: {r90:DI=r84:DI*r84:DI;clobber flags:CC;}
REG_DEAD r84:DI
REG_UNUSED flags:CC
12: r89:SI=r90:DI#0+r88:DI#0
REG_DEAD r88:DI
REG_DEAD r90:DI
Failed to match this instruction:
(set (reg:SI 89 [ sqr ])
(plus:SI (mult:SI (subreg:SI (reg/v:DI 84 [ t ]) 0)
(subreg:SI (reg/v:DI 84 [ t ]) 0))
(subreg:SI (reg/v:DI 88 [ sqr ]) 0)))
If we did allow that, then the *b case would work on the RTL.
To do this at the gimple level, we would need to introduce non-overflow based
operations or have a lower gimple where signedness does not matter except for
some operators (divide, mod and some shifts).
[Bug tree-optimization/102795] RFE: recognize !! vs branch similarity better
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102795 Andrew Pinski changed: What|Removed |Added Keywords||missed-optimization Last reconfirmed||2021-10-16 Ever confirmed|0 |1 Status|UNCONFIRMED |ASSIGNED Assignee|unassigned at gcc dot gnu.org |pinskia at gcc dot gnu.org Severity|normal |enhancement --- Comment #1 from Andrew Pinski --- Mine. For p2: if (_1 == 0) goto ; [34.00%] else goto ; [66.00%] [local count: 365072224]: _8 = x_6(D) - _10; goto ; [100.00%] [local count: 708669601]: _11 = x_6(D) + 18446744073709551615; _7 = _11 - _10; [local count: 1073741824]: # _5 = PHI <_8(3), _7(4)> this could be transformed to (which is almost p3): if (_1 == 0) goto ; [34.00%] else goto ; [66.00%] [local count: 708669601]: _11 = x_6(D) + 18446744073709551615; [local count: 1073741824]: # _n = PHI _5 = _n - _10 And then transfomed to: if (_1 == 0) goto ; [34.00%] else goto ; [66.00%] [local count: 708669601]: [local count: 1073741824]: # _n = PHI <0(2), -1(4)> _n0 = x_6(D) + _n _5 = _n0 - _10
[Bug ada/100486] Ada build fails for 32bit Windows
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100486 --- Comment #49 from Eric Botcazou --- > Fore completeness: The "exceptions not working" problem now also crept into > our v10.3 build with the last rebuild. Maybe some dependency change in the > last two months, but no idea :/ > > https://github.com/msys2/MINGW-packages/issues/9771 I can confirm that the libstdc++ DLL available from there is broken in exactly the same way as the one from GCC 11: it fails to register its unwinding tables, i.e. the call to __register_frame_info on startup is missing. That's managed by the pair of crtbegin.o/ctrend.o object files from the compiler but the intermediate __gcc_register_frame is never called.
[Bug tree-optimization/102795] New: RFE: recognize !! vs branch similarity better
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102795
Bug ID: 102795
Summary: RFE: recognize !! vs branch similarity better
Product: gcc
Version: 12.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: tree-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: jengelh at inai dot de
Target Milestone: ---
Input
=
unsigned long p1(unsigned long x)
{
return x - x / 10 - !!(x % 10);
}
unsigned long p2(unsigned long x)
{
if (x % 10 == 0)
return x - x / 10;
return x - x / 10 - 1;
}
unsigned long p3(unsigned long x)
{
unsigned long z = x - x / 10;
if (x % 10)
--z;
return z;
}
Outcome
===
► ./host-x86_64-pc-linux-gnu/gcc/xg++ -B ./host-x86_64-pc-linux-gnu/gcc -c -O3
-v t.cpp
GNU C++17 (GCC) version 12.0.0 20211016 (experimental) (x86_64-pc-linux-gnu)
compiled by GNU C version 11.2.1 20210816 [revision
056e324ce46a7924b5cf10f61010cf9dd2ca10e9], GMP version 6.2.1, MPFR version
4.1.0-p7, MPC version 1.2.1, isl version none
GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096
► objdump -Mintel -d t.o
<_Z2p1m>:
...
24: 0f 95 c2setne dl
27: 0f b6 d2movzx edx,dl
2a: 48 29 d0subrax,rdx
...
0030 <_Z2p2m>:
...
4f: 48 29 d6subrsi,rdx
52: 48 29 d0subrax,rdx
55: 48 01 c9addrcx,rcx
58: 48 39 cfcmprdi,rcx
5b: 48 0f 44 c6 cmove rax,rsi
...
0060 <_Z2p3m>:
...
81: 48 29 d7subrdi,rdx
84: 48 83 ff 01 cmprdi,0x1
88: 48 83 d0 ff adcrax,0x
...
Expected outcome
I would have hoped that the optimizer were able to reduce p1, p2 and p3 to the
same asm.
[Bug tree-optimization/102646] [12 Regression] large performance changes between 1932e1169a236849f5e7f1cd386da100d9af470f and 9cfb95f9b92326e86e99b50350ebf04fa9cd2477 (probably jump threading)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102646 Andrew Pinski changed: What|Removed |Added Summary|large performance changes |[12 Regression] large |between |performance changes between |1932e1169a236849f5e7f1cd386 |1932e1169a236849f5e7f1cd386 |da100d9af470f and |da100d9af470f and |9cfb95f9b92326e86e99b50350e |9cfb95f9b92326e86e99b50350e |bf04fa9cd2477 (probably |bf04fa9cd2477 (probably |jump threading) |jump threading) Target Milestone|--- |12.0
[Bug ada/100486] Ada build fails for 32bit Windows
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100486 --- Comment #48 from Christoph Reiter --- (In reply to Eric Botcazou from comment #47) > > Yes, everything is checksummed in our build script. We apply various patches > > and backports, they are also checksummed: > > https://github.com/msys2/MINGW-packages/blob/master/mingw-w64-gcc/PKGBUILD > > Did you change binutils, in particular the linker, in between? Yes, from 2.36.1 to 2.37, but I've already tried reverting that and it didn't help. I'm going to try older versions (of everything) though..
[Bug ada/100486] Ada build fails for 32bit Windows
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100486 --- Comment #47 from Eric Botcazou --- > Yes, everything is checksummed in our build script. We apply various patches > and backports, they are also checksummed: > https://github.com/msys2/MINGW-packages/blob/master/mingw-w64-gcc/PKGBUILD Did you change binutils, in particular the linker, in between?
[Bug ada/100486] Ada build fails for 32bit Windows
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100486 --- Comment #46 from Christoph Reiter --- (In reply to Eric Botcazou from comment #45) > > Fore completeness: The "exceptions not working" problem now also crept into > > our v10.3 build with the last rebuild. Maybe some dependency change in the > > last two months, but no idea :/ > > > > https://github.com/msys2/MINGW-packages/issues/9771 > > Can you confirm that this is with the same compiler sources as the previous > 10.3 version that works? Do you apply local changes to them? Yes, everything is checksummed in our build script. We apply various patches and backports, they are also checksummed: https://github.com/msys2/MINGW-packages/blob/master/mingw-w64-gcc/PKGBUILD
[Bug ada/100486] Ada build fails for 32bit Windows
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100486 --- Comment #45 from Eric Botcazou --- > Fore completeness: The "exceptions not working" problem now also crept into > our v10.3 build with the last rebuild. Maybe some dependency change in the > last two months, but no idea :/ > > https://github.com/msys2/MINGW-packages/issues/9771 Can you confirm that this is with the same compiler sources as the previous 10.3 version that works? Do you apply local changes to them?
[Bug tree-optimization/102794] [12 Regression] missing vrp in evrp dealing with casts and ands
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102794 --- Comment #3 from Andrew Pinski --- (In reply to Aldy Hernandez from comment #2) > I haven't looked at this, but there's a pending patch with more > restrictions for loop threading in the presence of loops. Does this help? > > https://gcc.gnu.org/pipermail/gcc-patches/2021-October/581637.html > > We really shouldn't be destroying loop info. Yes I think it does, though I have not tried it. There was a jump threading here: : if (d_10 != 0) goto ; [INV] else goto ; [INV] : foo (); : a_12 = a_3 + 1; : ;;; loop header # a_3 = PHI <-100(2), a_12(6)> if (a_3 != 0) goto ; [INV] else goto ; [INV] where d_10 is defined as: d_10 = (short unsigned int) a_3; So yes blocking jumping through the loop header would help. It is very similar to your f3 in ssa-thread-invalid.c testcase.
[Bug c/46116] Allow passing of anonymous aggregates when signature matches
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=46116 Martin Uecker changed: What|Removed |Added CC||muecker at gwdg dot de --- Comment #3 from Martin Uecker --- I have a(very preliminary) patch here: https://github.com/uecker/gcc/tree/tagcompat It implements an option -ftagcompat that makes structs with same tag (or no tag) and content compatible as proposed in N2366.
[Bug tree-optimization/102794] [12 Regression] missing vrp in evrp dealing with casts and ands
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102794 --- Comment #2 from Aldy Hernandez --- I haven't looked at this, but there's a pending patch with more restrictions for loop threading in the presence of loops. Does this help? https://gcc.gnu.org/pipermail/gcc-patches/2021-October/581637.html We really shouldn't be destroying loop info. On Sat, Oct 16, 2021, 07:01 pinskia at gcc dot gnu.org < gcc-bugzi...@gcc.gnu.org> wrote: > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102794 > > Andrew Pinski changed: > >What|Removed |Added > > > CC||aldyh at gcc dot gnu.org > > --- Comment #1 from Andrew Pinski --- > Jump threading really messes up the loop here ... > about to thread: path: 4 -> 6, 6 -> 7, 7 -> 3, > just threaded: path: 4 -> 9, 6 -> 7, 7 -> 3, > > I don't know what else to say. Maybe move ethread after evpr? > > Note with -fno-thread-jumps, evpr is able to figure out the induction > variable > goes from [-100, -1] . It does look like jump threading is full on > messing up > how induction variable detection works; we get two a = a +1; statement > after > the jump threading improvements. > Maybe there is another bug about that case already. > > -- > You are receiving this mail because: > You are on the CC list for the bug. > >
[Bug middle-end/102492] [12 Regression] ICE in scan_sharing_clauses, at omp-low.c:1205
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102492 Jakub Jelinek changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #3 from Jakub Jelinek --- Should be fixed now.
