[Bug tree-optimization/86816] [8/9 Regression] ICE: SIGSEGV in tree-ssa-pre / tail_merge_optimize
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86816 Richard Biener changed: What|Removed |Added Status|UNCONFIRMED |ASSIGNED Last reconfirmed||2018-08-02 Version|tree-ssa|8.2.0 Assignee|unassigned at gcc dot gnu.org |rguenth at gcc dot gnu.org Target Milestone|--- |8.3 Summary|[8 Regression] ICE: SIGSEGV |[8/9 Regression] ICE: |in tree-ssa-pre / |SIGSEGV in tree-ssa-pre / |tail_merge_optimize |tail_merge_optimize Ever confirmed|0 |1 Known to fail||9.0 --- Comment #1 from Richard Biener --- Confirmed. #1 0x014e9e2a in VN_INFO (name=) at /tmp/trunk/gcc/tree-ssa-sccvn.c:376 376 gcc_checking_assert (res); so we are asking for the value-number of sth that didn't get a VN_INFO comparing (gdb) p debug_gimple_stmt (s1) if (pretmp_13 == 1) $1 = void (gdb) p debug_gimple_stmt (s2) if (_16 <= 1) with _16 being the culprit here. This SSA name is generated via the CFG cleanup done between PRE and tail-merging. #0 make_ssa_name_fn (fn=0x769f8000, var=, stmt=, version=0) at /tmp/trunk/gcc/tree-ssanames.c:325 #1 0x012e7739 in make_ssa_name (var=, stmt=) at /tmp/trunk/gcc/tree-ssanames.h:115 #2 0x0130397a in generate_range_test ( bb=, index=, low=, high=, lhs=0x7fffd520, rhs=0x7fffd518) at /tmp/trunk/gcc/tree-cfg.c:9141 #3 0x0130b17a in convert_single_case_switch (swtch=0x768a9a00, gsi=...) at /tmp/trunk/gcc/tree-cfgcleanup.c:102 #4 0x0130b2f0 in cleanup_control_expr_graph ( bb=, gsi=...) at /tmp/trunk/gcc/tree-cfgcleanup.c:141 #5 0x0130b708 in cleanup_control_flow_bb ( bb=) at /tmp/trunk/gcc/tree-cfgcleanup.c:248 #6 0x0130ce48 in cleanup_control_flow_pre () at /tmp/trunk/gcc/tree-cfgcleanup.c:758 #7 0x0130d2a9 in cleanup_tree_cfg_noloop () at /tmp/trunk/gcc/tree-cfgcleanup.c:882 #8 0x0130d52d in cleanup_tree_cfg () at /tmp/trunk/gcc/tree-cfgcleanup.c:990 #9 0x0153d81f in tail_merge_optimize (todo=32) at /tmp/trunk/gcc/tree-ssa-tail-merge.c:1732 #10 0x014c3e22 in (anonymous namespace)::pass_pre::execute ( this=0x2f26300, fun=0x769f8000) at /tmp/trunk/gcc/tree-ssa-pre.c:4208
[Bug target/86820] [8 regression] gcc-8 miscompiles binutils on m68k
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86820 Richard Biener changed: What|Removed |Added Priority|P3 |P4 Known to work||9.0
[Bug c++/86823] [6/7/8/9 Regression] private member template struct/class is publicly accessible
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86823 Richard Biener changed: What|Removed |Added Keywords||accepts-invalid Target Milestone|--- |6.5
[Bug c++/86825] g++.old-deja/g++.pt/ptrmem10.C test fails on windows targets
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86825 Richard Biener changed: What|Removed |Added Version|unknown |9.0 --- Comment #1 from Richard Biener --- Just assuming this is a report for trunk.
[Bug tree-optimization/86824] [9 Regression] internal compiler error: in replace_stmt, at tree-vectorizer.c:611
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86824 rsandifo at gcc dot gnu.org changed: What|Removed |Added Status|NEW |ASSIGNED CC||rsandifo at gcc dot gnu.org Assignee|unassigned at gcc dot gnu.org |rsandifo at gcc dot gnu.org --- Comment #3 from rsandifo at gcc dot gnu.org --- Mine.
[Bug middle-end/86763] [8/9 Regression] Wrong code comparing member of copy of a 237 byte object with nontrivial default constructor on x86-64 arch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86763
Richard Biener changed:
What|Removed |Added
Keywords||alias
Status|NEW |ASSIGNED
Component|rtl-optimization|middle-end
--- Comment #6 from Richard Biener ---
(In reply to Uroš Bizjak from comment #5)
> (In reply to Richard Biener from comment #3)
> > see how the compare is moved very far away from the branch possibly across
> > flag clobbering insns (rep movsq?). Target sounds good.
>
> Scheduler (sched2) pass is moving the compare:
>
> (insn 21 17 22 2 (set (reg:CCZ 17 flags)
> (compare:CCZ (mem/c:DI (plus:DI (reg/f:DI 7 sp)
> (const_int 8 [0x8])) [1 MEM[(struct Msg &)&t].id+0 S8
> A64])
> (const_int 1001 [0x3e9]))) "t4.cpp":32 12 {*cmpdi_1}
> (nil))
>
> in front of rep movsd:
>
> (insn 11 10 12 2 (parallel [
> (set (reg:DI 2 cx [92])
> (const_int 0 [0]))
> (set (reg/f:DI 5 di [90])
> (plus:DI (ashift:DI (reg:DI 2 cx [92])
> (const_int 3 [0x3]))
> (reg/f:DI 5 di [90])))
> (set (reg/f:DI 4 si [91])
> (plus:DI (ashift:DI (reg:DI 2 cx [92])
> (const_int 3 [0x3]))
> (reg/f:DI 4 si [91])))
> (set (mem/c:BLK (reg/f:DI 5 di [90]) [7 MEM[(struct T *)&t]+0
> S232 A128])
> (mem/c:BLK (reg/f:DI 4 si [91]) [7 MEM[(struct T
> *)&D.2891]+0 S232 A128]))
> (use (reg:DI 2 cx [92]))
> ]) "t4.cpp":31 993 {*rep_movdi_rex64}
> (expr_list:REG_UNUSED (reg:DI 2 cx [92])
> (nil)))
>
> Scheduler doesn't notice that rep movsd target overlaps compare operand.
>
> Looks like rtl-optimization (scheduler) to me.
I notice that -O -fstrict-aliasing -fschedule-insns2 fails and -O2
-fno-strict-aliasing succeeds. The above shows the block move uses alias-set 7
and the
compare alias-set 1. Alias set 1 is a subset of alias set 7 but not the
other way around - this suggests that sched-deps is querying the oracle
in a wrong way. Huh, no.
Breakpoint 2, true_dependence_1 (mem=0x76a5ec00, mem_mode=E_QImode,
mem_addr=0x3106258, x=0x76a5edb0, x_addr=0x76a61780,
mem_canonicalized=true)
at /space/rguenther/src/svn/gcc-8-branch/gcc/alias.c:2905
2905 gcc_checking_assert (mem_canonicalized ? (mem_addr != NULL_RTX)
(mem/c:QI (reg/f:DI 90) [7 MEM[(struct T *)&t]+236 S1 A32])
$27 = void
(mem/c:DI (plus:DI (reg/f:DI 20 frame)
(const_int -232 [0xff18])) [1 MEM[(struct Msg &)&t].id+0 S8
A64])
$28 = void
(gdb) fin
Run till exit from #0 true_dependence_1 (mem=0x76a5ec00,
mem_mode=E_QImode, mem_addr=0x3106258, x=0x76a5edb0,
x_addr=0x76a61780, mem_canonicalized=true)
at /space/rguenther/src/svn/gcc-8-branch/gcc/alias.c:2905
canon_true_dependence (mem=0x76a5ec00, mem_mode=E_QImode,
mem_addr=0x3106258, x=0x76a5edb0, x_addr=0x76a61780)
at /space/rguenther/src/svn/gcc-8-branch/gcc/alias.c:2998
2998}
Value returned is $29 = 0
so the oracle says the load cannot alias the store. We enter with
(gdb) p debug_rtx (mem_addr)
(value:DI 10:17131 @0x3106258/0x30e6310)
(gdb) p debug_rtx (x_addr)
(plus:DI (reg/f:DI 20 frame)
(const_int -232 [0xff18]))
and find_base_term (x_addr) returns (address:DI -3) (same as mem_base)
Hmm. alias_sets_conflict_p returns false?! Ah, alias_set_subset_of
returns true because of has_zero_child. alias-set 7 has no children,
it looks like struct T has no members but it does have a member of type V.
So we do record_component_aliases arriving at that field and get_alias_set
on this field returns zero due to TYPE_TYPELESS_STORAGE (which is new
in GCC 8 which explains why there may be a regression). So we record
the ->has_zero_child relationship (instead of previously recording
a distinct alias-set).
Now - some earlier change - improvement to alias_sets_conflict_p in fact,
made us not consider ->has_zero_child there basically to avoid making
struct X { int i; }; and struct Y { float f; char x[4]; } conflict. But
this is exactly what we are trying to allow with TYPE_TYPELESS_STORAGE :/
So TYPE_TYPELESS_STORAGE doesn't work as desired and introduces wrong-code.
[Bug middle-end/86763] [7/8/9 Regression] Wrong code comparing member of copy of a 237 byte object with nontrivial default constructor on x86-64 arch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86763 Richard Biener changed: What|Removed |Added Known to work||6.4.0 Summary|[8/9 Regression] Wrong code |[7/8/9 Regression] Wrong |comparing member of copy of |code comparing member of |a 237 byte object with |copy of a 237 byte object |nontrivial default |with nontrivial default |constructor on x86-64 arch |constructor on x86-64 arch Known to fail||7.1.0 --- Comment #7 from Richard Biener --- Also fails with GCC 7 (not with -O2 but with -O -fstrict-aliasing -fschedule-insns2), same reason.
[Bug target/86774] Alpha port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86774 Michael Cree changed: What|Removed |Added CC||mcree at orcon dot net.nz --- Comment #2 from Michael Cree --- Just to note that in testing earlier this year I managed to successfully run meltdown and spectre v1 attacks on Alpha EV68 hardware. These attacks were less successful on EV67 (only a 1% to 2% success rate). I failed to get any attack on EV56---cache timings did not reveal anything useful (though I did not spend too much time analysing this case).
[Bug c++/86763] [7/8/9 Regression] Wrong code comparing member of copy of a 237 byte object with nontrivial default constructor on x86-64 arch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86763
Richard Biener changed:
What|Removed |Added
Status|ASSIGNED|NEW
CC||jason at gcc dot gnu.org,
||nathan at gcc dot gnu.org,
||rguenth at gcc dot gnu.org
Component|middle-end |c++
--- Comment #8 from Richard Biener ---
So the question is why T doesn't get TYPE_TYPELESS_STORAGE set even though
we propagate that flag from children upwards in place_field. That's what
is missing here and why things go wrong.
Ah, so we go
build_base_field (rli=0x2ea1780, binfo=0x76a2f120, offsets=0x2ea2280,
next_field=0x76a285c8)
at /space/rguenther/src/svn/gcc-8-branch/gcc/cp/class.c:4244
4244 tree t = rli->t;
(gdb) p basetype->type_common.typeless_storage
$30 = 1
OK
4262 decl = build_base_field_1 (t, basetype, next_field);
(gdb) p decl->typed.type->type_common.typeless_storage
$31 = 0
looks like the as-base type doesn't inherit this flag. If I fix that up here
where it matters rather than tracking down the gazillion places the C++ FE
seems to set CLASSTYPE_AS_BASE and where I'm unsure at that time layout
is finished the bug is fixed:
Index: gcc/cp/class.c
===
--- gcc/cp/class.c (revision 263209)
+++ gcc/cp/class.c (working copy)
@@ -4202,6 +4202,8 @@ build_base_field_1 (tree t, tree basetyp
{
/* Create the FIELD_DECL. */
gcc_assert (CLASSTYPE_AS_BASE (basetype));
+ TYPE_TYPELESS_STORAGE (CLASSTYPE_AS_BASE (basetype))
+= TYPE_TYPELESS_STORAGE (basetype);
tree decl = build_decl (input_location,
FIELD_DECL, NULL_TREE, CLASSTYPE_AS_BASE (basetype));
DECL_ARTIFICIAL (decl) = 1;
C++ folks?
[Bug c++/86763] [7/8/9 Regression] Wrong code comparing member of copy of a 237 byte object with nontrivial default constructor on x86-64 arch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86763 --- Comment #9 from Richard Biener --- The following seems to work, will test. Index: gcc/cp/class.c === --- gcc/cp/class.c (revision 263209) +++ gcc/cp/class.c (working copy) @@ -6243,6 +6243,7 @@ layout_class_type (tree t, tree *virtual bitsize_int (BITS_PER_UNIT))); SET_TYPE_ALIGN (base_t, rli->record_align); TYPE_USER_ALIGN (base_t) = TYPE_USER_ALIGN (t); + TYPE_TYPELESS_STORAGE (base_t) = TYPE_TYPELESS_STORAGE (t); /* Copy the non-static data members of T. This will include its direct non-virtual bases & vtable. */
[Bug c/86827] New: -Warray-bounds produces negative indicies
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86827 Bug ID: 86827 Summary: -Warray-bounds produces negative indicies Product: gcc Version: 8.1.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: c Assignee: unassigned at gcc dot gnu.org Reporter: liuw at liuw dot name Target Milestone: --- Created attachment 44484 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=44484&action=edit Test code The attached program fails to build with gcc 8.1 (Debian 8.1.0-12). $ gcc -m32 -march=i686 -std=gnu99 -Wall -O2 -Werror -c -o t.o t.c t.c: In function 'func': t.c:41:9: error: 'memcpy' offset [-204, -717] is out of the bounds [0, 216] of object 'ctrl' with type 'struct kdd_ctrl' [-Werror=array-bounds] memcpy(buf, ((uint8_t *)&ctrl.c32) + offset, len); ^ t.c:27:21: note: 'ctrl' declared here struct kdd_ctrl ctrl; And to quote Martin in a thread to gcc-help: It looks like a bug in the implementation of the warning. The offset is determined not to be in the range [-205, -716] (pointer offsets are in ptrdiff_t) or (since the variable is unsigned) in [4294966580, 4294967091]. That means that it can be either in the range [0, 4294966579] or in [4294967092, UINT_MAX]. But the warning code seems to get this anti-range wrong and treats it as [-204, -717].
[Bug c++/86763] [7/8/9 Regression] Wrong code comparing member of copy of a 237 byte object with nontrivial default constructor on x86-64 arch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86763 Richard Biener changed: What|Removed |Added Status|NEW |ASSIGNED Assignee|unassigned at gcc dot gnu.org |rguenth at gcc dot gnu.org
[Bug c/86827] [8/9 Regression] -Warray-bounds produces negative indicies
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86827 Richard Biener changed: What|Removed |Added Keywords||diagnostic CC||msebor at gcc dot gnu.org Target Milestone|--- |8.3 Summary|-Warray-bounds produces |[8/9 Regression] |negative indicies |-Warray-bounds produces ||negative indicies --- Comment #1 from Richard Biener --- hand-rolled range handling code again :/
[Bug target/86802] riscv port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86802 --- Comment #2 from Richard Earnshaw --- I think the best thing to do in that case is to leave the port unfixed until such time as you know what mitigation is appropriate. That way the compiler will not define __HAVE_SPECULATION_SAFE_VALUE and users will not be given a false sense of security: if they have code that checks for whether mitigation is needed, the lack of the define will tell them that it isn't known yet. It won't stop them using the new builtin, but it will lead to a warning if used in this state.
[Bug target/85434] Address of stack protector guard spilled to stack on ARM
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85434 --- Comment #18 from Thomas Preud'homme --- Author: thopre01 Date: Thu Aug 2 09:07:17 2018 New Revision: 263245 URL: https://gcc.gnu.org/viewcvs?rev=263245&root=gcc&view=rev Log: [ARM] Fix PR85434: spilling of stack protector guard's address on ARM In case of high register pressure in PIC mode, address of the stack protector's guard can be spilled on ARM targets as shown in PR85434, thus allowing an attacker to control what the canary would be compared against. This is also known as CVE-2018-12886. ARM does lack stack_protect_set and stack_protect_test insn patterns, defining them does not help as the address is expanded regularly and the patterns only deal with the copy and test of the guard with the canary. This problem does not occur for x86 targets because the PIC access and the test can be done in the same instruction. Aarch64 is exempt too because PIC access insn pattern are mov of UNSPEC which prevents it from the second access in the epilogue being CSEd in cse_local pass with the first access in the prologue. The approach followed here is to create new "combined" set and test standard pattern names that take the unexpanded guard and do the set or test. This allows the target to use an opaque pattern (eg. using UNSPEC) to hide the individual instructions being generated to the compiler and split the pattern into generic load, compare and branch instruction after register allocator, therefore avoiding any spilling. This is here implemented for the ARM targets. For targets not implementing these new standard pattern names, the existing stack_protect_set and stack_protect_test pattern names are used. To be able to split PIC access after register allocation, the functions had to be augmented to force a new PIC register load and to control which register it loads into. This is because sharing the PIC register between prologue and epilogue could lead to spilling due to CSE again which an attacker could use to control what the canary gets compared against. 2018-08-02 Thomas Preud'homme gcc/ PR target/85434 * target-insns.def (stack_protect_combined_set): Define new standard pattern name. (stack_protect_combined_test): Likewise. * cfgexpand.c (stack_protect_prologue): Try new stack_protect_combined_set pattern first. * function.c (stack_protect_epilogue): Try new stack_protect_combined_test pattern first. * config/arm/arm.c (require_pic_register): Add pic_reg and compute_now parameters to control which register to use as PIC register and force reloading PIC register respectively. Insert in the stream of insns if possible. (legitimize_pic_address): Expose above new parameters in prototype and adapt recursive calls accordingly. (arm_legitimize_address): Adapt to new legitimize_pic_address prototype. (thumb_legitimize_address): Likewise. (arm_emit_call_insn): Adapt to new require_pic_register prototype. * config/arm/arm-protos.h (legitimize_pic_address): Adapt to prototype change. * config/arm/arm.md (movsi expander): Adapt to legitimize_pic_address prototype change. (stack_protect_combined_set): New insn_and_split pattern. (stack_protect_set): New insn pattern. (stack_protect_combined_test): New insn_and_split pattern. (stack_protect_test): New insn pattern. * config/arm/unspecs.md (UNSPEC_SP_SET): New unspec. (UNSPEC_SP_TEST): Likewise. * doc/md.texi (stack_protect_combined_set): Document new standard pattern name. (stack_protect_set): Clarify that the operand for guard's address is legal. (stack_protect_combined_test): Document new standard pattern name. (stack_protect_test): Clarify that the operand for guard's address is legal. gcc/testsuite/ PR target/85434 * gcc.target/arm/pr85434.c: New test. Added: trunk/gcc/testsuite/gcc.target/arm/pr85434.c Modified: trunk/gcc/ChangeLog trunk/gcc/cfgexpand.c trunk/gcc/config/arm/arm-protos.h trunk/gcc/config/arm/arm.c trunk/gcc/config/arm/arm.md trunk/gcc/config/arm/unspecs.md trunk/gcc/doc/md.texi trunk/gcc/function.c trunk/gcc/target-insns.def trunk/gcc/testsuite/ChangeLog
[Bug sanitizer/86022] TCB size calculated in ThreadDescriptorSize() is wrong for glibc-2.14
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86022 --- Comment #4 from Martin Liška --- Author: marxin Date: Thu Aug 2 09:32:58 2018 New Revision: 263246 URL: https://gcc.gnu.org/viewcvs?rev=263246&root=gcc&view=rev Log: Cherry-pick compiler-rt revision 338606 (PR sanitizer/86022). Fix sizeof(struct pthread) in glibc 2.14. 2018-08-02 Martin Liska PR sanitizer/86022 * sanitizer_common/sanitizer_linux_libcdep.cc (ThreadDescriptorSize): Cherry-pick compiler-rt revision 338606. Modified: trunk/libsanitizer/ChangeLog trunk/libsanitizer/sanitizer_common/sanitizer_linux_libcdep.cc
[Bug sanitizer/86022] TCB size calculated in ThreadDescriptorSize() is wrong for glibc-2.14
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86022 Martin Liška changed: What|Removed |Added Known to work||9.0 Known to fail|9.0 | --- Comment #5 from Martin Liška --- Fixed on trunk so far.
[Bug target/86758] [9 Regression] ICE in replace_stmt compiling gcc.target/i386/pr84309.c
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86758 rsandifo at gcc dot gnu.org changed: What|Removed |Added CC||hjl.tools at gmail dot com --- Comment #5 from rsandifo at gcc dot gnu.org --- *** Bug 86824 has been marked as a duplicate of this bug. ***
[Bug tree-optimization/86824] [9 Regression] internal compiler error: in replace_stmt, at tree-vectorizer.c:611
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86824 rsandifo at gcc dot gnu.org changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|--- |DUPLICATE --- Comment #4 from rsandifo at gcc dot gnu.org --- Seems to be the same as PR86758, which was fixed in r263222. Sorry again for the breakage. *** This bug has been marked as a duplicate of bug 86758 ***
[Bug tree-optimization/86702] [9 Regression] SPEC CPU2006 400.perlbench, CPU2017 500.perlbench_r ~3% performance drop after r262247
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86702 --- Comment #2 from Martin Liška --- Can you please provide info about affected target and precise options used?
[Bug tree-optimization/86702] [9 Regression] SPEC CPU2006 400.perlbench, CPU2017 500.perlbench_r ~3% performance drop after r262247
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86702 --- Comment #3 from Martin Liška --- I must confirm that probabilities are lost, but as the switch will end up as jump table, so they are not used..
[Bug target/86828] New: [6/7 Regression] wrong-code bug with "-march=knl -Ofast" (invalid memory reference)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86828 Bug ID: 86828 Summary: [6/7 Regression] wrong-code bug with "-march=knl -Ofast" (invalid memory reference) Product: gcc Version: 8.2.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: target Assignee: unassigned at gcc dot gnu.org Reporter: janus at gcc dot gnu.org Target Milestone: --- Fortran test case: program knl_bug implicit none integer, parameter :: NN = 32 type tList integer, dimension(:), allocatable :: list end type integer, dimension(1:100) :: Y = 0 type(tList) :: nb allocate(nb%list(1:NN), source = 1) call s(nb) contains subroutine s(n) type(tList), intent(in) :: n integer, dimension(1:NN) :: liKOB liKOB = Y(n%list) print *,liKOB end subroutine end Compiling this with "gfortran -march=knl -Ofast -g", and then running the executable, I get: $ ./a.out Program received signal SIGSEGV: Segmentation fault - invalid memory reference. Backtrace for this error: #0 0x7f9d495c42da in ??? #1 0x7f9d495c3503 in ??? #2 0x7f9d491f6f1f in ??? #3 0x563dd9a3dd7f in s at test.f90:26 #4 0x563dd9a3da9a in knl_bug at test.f90:17 #5 0x563dd9a3da9a in main at test.f90:17 Segmentation fault (core dumped) I see this happening with: gcc version 6.4.0 20180727 (Ubuntu 6.4.0-19ubuntu1~18.04) gcc version 7.3.0 (Ubuntu 7.3.0-27ubuntu1~18.04) but not with: gcc version 5.5.0 20171010 (Ubuntu 5.5.0-12ubuntu1) gcc version 8.2.0 (Ubuntu 8.2.0-1ubuntu2~18.04) gcc version 9.0.0 20180730 (experimental) [trunk revision 263059] (GCC) Note that the segfault also happens with -march=skylake-avx512 (which is not available with gcc 5). Similar to PR 86735. The CPU where I found this is an Intel(R) Core(TM) i9-7980XE.
[Bug target/86828] [6/7 Regression] wrong-code bug with "-march=knl -Ofast" (invalid memory reference)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86828 janus at gcc dot gnu.org changed: What|Removed |Added Keywords||wrong-code See Also||https://gcc.gnu.org/bugzill ||a/show_bug.cgi?id=86735 Target Milestone|--- |6.5
[Bug gcov-profile/86817] [8/9 Regression] Misleading warning in gcov since r254672
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86817 --- Comment #1 from Martin Liška --- Author: marxin Date: Thu Aug 2 10:17:34 2018 New Revision: 263248 URL: https://gcc.gnu.org/viewcvs?rev=263248&root=gcc&view=rev Log: Fix gcov misleading error (PR gcov-profile/86817). 2018-08-02 Martin Liska PR gcov-profile/86817 * gcov.c (process_all_functions): New function. (main): Call it. (process_file): Move functions processing to process_all_functions. Modified: trunk/gcc/ChangeLog trunk/gcc/gcov.c
[Bug target/86771] [9 Regression] gfortran.dg/actual_array_constructor_1.f90 fails on arm after combine 2 insns to 2 insns patch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86771 --- Comment #1 from Segher Boessenkool --- Could you trace this down to some bad code generated, at least?
[Bug middle-end/86815] [8/9 regression] ICE on valid code on armhf
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86815 --- Comment #6 from Gianfranco --- Created attachment 44485 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=44485&action=edit another failing output I'm attaching another file suffering from the same issue (mostly every cpp file has this failure) this file is only ~2Mb, so maybe reducing it might be easier
[Bug target/86014] [AArch64] missed LDP optimization
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86014 --- Comment #2 from jcw at gcc dot gnu.org --- Author: jcw Date: Thu Aug 2 10:39:23 2018 New Revision: 263249 URL: https://gcc.gnu.org/viewcvs?rev=263249&root=gcc&view=rev Log: gcc/ 2018-08-02 Jackson Woodruff PR target/86014 * config/aarch64/aarch64.c (aarch64_operands_adjust_ok_for_ldpstp): No longer check last store for clobber of address register. gcc/testsuite 2018-08-02 Jackson Woodruff PR target/86014 * gcc.target/aarch64/ldp_stp_13.c: New test. Added: trunk/gcc/testsuite/gcc.target/aarch64/ldp_stp_13.c Modified: trunk/gcc/ChangeLog trunk/gcc/config/aarch64/aarch64.c trunk/gcc/testsuite/ChangeLog
[Bug target/86771] [9 Regression] gfortran.dg/actual_array_constructor_1.f90 fails on arm after combine 2 insns to 2 insns patch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86771 --- Comment #2 from Christophe Lyon --- gfortran.log contains: STOP 4 STOP 4 STOP 4 before the execution fails I'll regenerate the 2 asm files.
[Bug target/86014] [AArch64] missed LDP optimization
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86014 jcw at gcc dot gnu.org changed: What|Removed |Added Status|NEW |RESOLVED CC||jcw at gcc dot gnu.org Resolution|--- |FIXED --- Comment #3 from jcw at gcc dot gnu.org --- Fixed by r263249.
[Bug target/86781] fr30 port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86781 --- Comment #1 from Nick Clifton --- Author: nickc Date: Thu Aug 2 11:24:47 2018 New Revision: 263253 URL: https://gcc.gnu.org/viewcvs?rev=263253&root=gcc&view=rev Log: PR target/86781 * config/fr30/fr30.c (TARGET_HAVE_SPECULATION_SAFE_VALUE): Define to speculation_safe_value_not_needed. Modified: trunk/gcc/ChangeLog trunk/gcc/config/fr30/fr30.c
[Bug target/86781] fr30 port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86781 Nick Clifton changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #2 from Nick Clifton --- I have defined TARGET_HAVE_SPECULATION_SAFE_VALUE as speculation_safe_value_not_needed.
[Bug target/86772] [meta-bug] tracking port status for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86772 Bug 86772 depends on bug 86781, which changed state. Bug 86781 Summary: fr30 port needs updating for CVE-2017-5753 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86781 What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED
[Bug tree-optimization/86829] New: Missing sin(atan(x)) and cos(atan(x)) optimizations
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86829 Bug ID: 86829 Summary: Missing sin(atan(x)) and cos(atan(x)) optimizations Product: gcc Version: unknown Status: UNCONFIRMED Severity: normal Priority: P3 Component: tree-optimization Assignee: unassigned at gcc dot gnu.org Reporter: giuliano.belinassi at usp dot br Target Milestone: --- Created attachment 44486 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=44486&action=edit add sin(atan(x)) and cos(atan(x)) substitutions rules. The file named 'match.pd' does not contain the following simplifications rules: sin(atan(x)) -> x / sqrt(x*x + 1) and cos(atan(x)) -> 1 / sqrt(x*x + 1). According to the simple brenchmark I made, these substitutions can provide a 10x speedup in the code. I wrote a patch to add these optimizations. link to the perf test: https://pastebin.com/5ujSRmhq assembly dump of the perftest: https://pastebin.com/gLJeWHY8 The code I wrote add an instruction 'CVTSS2SD'. I don't know why it happens.
[Bug target/86782] frv port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86782 --- Comment #1 from Nick Clifton --- Author: nickc Date: Thu Aug 2 11:46:06 2018 New Revision: 263254 URL: https://gcc.gnu.org/viewcvs?rev=263254&root=gcc&view=rev Log: PR target/86782 * config/frv/frv.c (TARGET_HAVE_SPECULATION_SAFE_VALUE): Define to speculation_safe_value_not_needed. Modified: trunk/gcc/ChangeLog trunk/gcc/config/frv/frv.c
[Bug target/86787] iq2000 port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86787 Nick Clifton changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #1 from Nick Clifton --- I have defined TARGET_HAVE_SPECULATION_SAFE_VALUE as speculation_safe_value_not_needed
[Bug target/86772] [meta-bug] tracking port status for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86772 Bug 86772 depends on bug 86787, which changed state. Bug 86787 Summary: iq2000 port needs updating for CVE-2017-5753 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86787 What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED
[Bug tree-optimization/86829] Missing sin(atan(x)) and cos(atan(x)) optimizations
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86829 --- Comment #1 from Marc Glisse --- You shouldn't use float_type_node directly, since the operations might be on double or long double. Do you have a copyright assignment (https://gcc.gnu.org/contribute.html) ?
[Bug target/86782] frv port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86782 Nick Clifton changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #2 from Nick Clifton --- I have defined TARGET_HAVE_SPECULATION_SAFE_VALUE speculation_safe_value_not_needed
[Bug target/86772] [meta-bug] tracking port status for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86772 Bug 86772 depends on bug 86782, which changed state. Bug 86782 Summary: frv port needs updating for CVE-2017-5753 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86782 What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED
[Bug tree-optimization/86829] Missing sin(atan(x)) and cos(atan(x)) optimizations
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86829
Richard Biener changed:
What|Removed |Added
Keywords||missed-optimization, patch
Status|UNCONFIRMED |NEW
Last reconfirmed||2018-08-02
CC||rguenth at gcc dot gnu.org
Version|unknown |9.0
Ever confirmed|0 |1
--- Comment #2 from Richard Biener ---
+ /* Simplify sin(atan(x)) -> x / sqrt(x*x + 1). */
+ (for sins (SIN)
+ atans (ATAN)
+ sqrts (SQRT)
+ (simplify
+ (sins (atans @0))
+ (rdiv @0 ( sqrts (plus (mult @0 @0)
+ { build_one_cst (float_type_node);})
+
+
+ /* Simplify cos(atan(x)) -> 1 / sqrt(x*x + 1). */
+ (for coss (COS)
+ atans (ATAN)
+ sqrts (SQRT)
+ (simplify
+ (coss (atans @0))
+ (rdiv { build_one_cst (float_type_node);}
+ ( sqrts (plus (mult @0 @0) { build_one_cst (float_type_node);})
likely happens because you use float_type_node instead of 'type' for
build_one_cst. I think that if the atan intermediate result has additional
uses then the substitution may not be beneficial? Thus you should
write (atans:s @0) in both places.
Otherwise this looks OK. Can you please post the patch to gcc-patches
and add testcases?
[Bug target/86771] [9 Regression] gfortran.dg/actual_array_constructor_1.f90 fails on arm after combine 2 insns to 2 insns patch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86771 --- Comment #3 from Christophe Lyon --- Created attachment 44487 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=44487&action=edit Wrong code generated This is with trunk @r263197
[Bug target/86828] [6/7 Regression] wrong-code bug with "-march=knl -Ofast" (invalid memory reference)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86828 Richard Biener changed: What|Removed |Added Keywords||needs-bisection Target||x86_64-*-*, i?86-*-* --- Comment #1 from Richard Biener --- Would be nice to know what fixed this (or maybe know if it just went latent).
[Bug target/86771] [9 Regression] gfortran.dg/actual_array_constructor_1.f90 fails on arm after combine 2 insns to 2 insns patch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86771 --- Comment #4 from Christophe Lyon --- Created attachment 44488 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=44488&action=edit Good code This is with r263197 and r263067 (your patch) reverted
[Bug target/86771] [9 Regression] gfortran.dg/actual_array_constructor_1.f90 fails on arm after combine 2 insns to 2 insns patch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86771
--- Comment #5 from Christophe Lyon ---
I think in the "ok" version we have:
add ip, sp, #60
...
ldm ip, {r0, r1}
...
add r2, sp, #72
ldm r2, {r0, r1}
in the "ko" version we have:
ldr r1, [sp, #64]
...
ldr r1, [sp, #76]
So in the "ko" version we do not load r0
[Bug target/86787] iq2000 port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86787 --- Comment #2 from Nick Clifton --- Author: nickc Date: Thu Aug 2 12:14:52 2018 New Revision: 263255 URL: https://gcc.gnu.org/viewcvs?rev=263255&root=gcc&view=rev Log: PR target/86787 * config/iq2000/iq2000.c (TARGET_HAVE_SPECULATION_SAFE_VALUE): Define to speculation_safe_value_not_needed. Modified: trunk/gcc/ChangeLog trunk/gcc/config/iq2000/iq2000.c
[Bug target/86828] [6/7 Regression] wrong-code bug with "-march=knl -Ofast" (invalid memory reference)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86828 --- Comment #2 from janus at gcc dot gnu.org --- (In reply to Richard Biener from comment #1) > Would be nice to know what fixed this (or maybe know if it just went latent). I think I remember having some intermediate versions during the reduction that failed also with 8 and trunk. So I'd guess it just went dormant, or the segfault depends on memory layout or something. Wanted to try valgrind on the exe, but it seems it doesn't support AVX-512 yet (at least the version I have).
[Bug target/85434] Address of stack protector guard spilled to stack on ARM
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85434 Christophe Lyon changed: What|Removed |Added CC||clyon at gcc dot gnu.org --- Comment #19 from Christophe Lyon --- Created attachment 44489 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=44489&action=edit Source file causing ICE on aarch64 With your patch, GCC crashes with target aarch64-none-linux-gnu aarch64-none-linux-gnu-gcc gethnamaddr.i -fstack-protector during RTL pass: expand gethnamaddr.c: In function 'getanswer': gethnamaddr.c:179:1: internal compiler error: in maybe_gen_insn, at optabs.c:7307 getanswer (const querybuf *answer, int anslen, const char *qname, int qtype) ^ 0xafcef2 maybe_gen_insn(insn_code, unsigned int, expand_operand*) /home/christophe.lyon/src/GCC/sources/gcc-fsf/trunk/gcc/optabs.c:7307 0xaffb88 maybe_expand_insn(insn_code, unsigned int, expand_operand*) /home/christophe.lyon/src/GCC/sources/gcc-fsf/trunk/gcc/optabs.c:7351 0x7748a0 stack_protect_prologue /home/christophe.lyon/src/GCC/sources/gcc-fsf/trunk/gcc/cfgexpand.c:6117 0x7748a0 execute /home/christophe.lyon/src/GCC/sources/gcc-fsf/trunk/gcc/cfgexpand.c:6357 Please submit a full bug report,
[Bug target/85434] Address of stack protector guard spilled to stack on ARM
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85434 --- Comment #20 from Thomas Preud'homme --- (In reply to Christophe Lyon from comment #19) > Created attachment 44489 [details] > Source file causing ICE on aarch64 > > With your patch, GCC crashes with target aarch64-none-linux-gnu > aarch64-none-linux-gnu-gcc gethnamaddr.i -fstack-protector > > during RTL pass: expand > gethnamaddr.c: In function 'getanswer': > gethnamaddr.c:179:1: internal compiler error: in maybe_gen_insn, at > optabs.c:7307 > getanswer (const querybuf *answer, int anslen, const char *qname, int qtype) > ^ > 0xafcef2 maybe_gen_insn(insn_code, unsigned int, expand_operand*) > /home/christophe.lyon/src/GCC/sources/gcc-fsf/trunk/gcc/optabs.c:7307 > 0xaffb88 maybe_expand_insn(insn_code, unsigned int, expand_operand*) > /home/christophe.lyon/src/GCC/sources/gcc-fsf/trunk/gcc/optabs.c:7351 > 0x7748a0 stack_protect_prologue > > /home/christophe.lyon/src/GCC/sources/gcc-fsf/trunk/gcc/cfgexpand.c:6117 > 0x7748a0 execute > > /home/christophe.lyon/src/GCC/sources/gcc-fsf/trunk/gcc/cfgexpand.c:6357 > Please submit a full bug report, Thanks Christophe, It seems to have impacted x86 as well. I'll look at all those and respin the patch. I've reverted it in the meantime.
[Bug target/86789] m32r port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86789 --- Comment #1 from Nick Clifton --- Author: nickc Date: Thu Aug 2 12:31:29 2018 New Revision: 263256 URL: https://gcc.gnu.org/viewcvs?rev=263256&root=gcc&view=rev Log: PR target/86789 * config/m32r/m32r.c (TARGET_HAVE_SPECULATION_SAFE_VALUE): Define to speculation_safe_value_not_needed. Modified: trunk/gcc/ChangeLog trunk/gcc/config/m32r/m32r.c
[Bug target/86789] m32r port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86789 Nick Clifton changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #2 from Nick Clifton --- I have defined TARGET_HAVE_SPECULATION_SAFE_VALUE as speculation_safe_value_not_needed.
[Bug target/86772] [meta-bug] tracking port status for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86772 Bug 86772 depends on bug 86789, which changed state. Bug 86789 Summary: m32r port needs updating for CVE-2017-5753 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86789 What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED
[Bug tree-optimization/86829] Missing sin(atan(x)) and cos(atan(x)) optimizations
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86829 --- Comment #3 from Giuliano Belinassi --- (In reply to Marc Glisse from comment #1) > > Do you have a copyright assignment (https://gcc.gnu.org/contribute.html) ? No. Sorry, but I think I may need help getting this right. Are there any tips? (In reply to Richard Biener from comment #2) > likely happens because you use float_type_node instead of 'type' for > build_one_cst. I think that if the atan intermediate result has additional > uses then the substitution may not be beneficial? Thus you should > write (atans:s @0) in both places. > > Otherwise this looks OK. Can you please post the patch to gcc-patches > and add testcases? I will fix these issues and add some test cases. I appreciate your help!
[Bug fortran/86830] New: Contiguous array pointer function result not recognized as contiguous
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86830 Bug ID: 86830 Summary: Contiguous array pointer function result not recognized as contiguous Product: gcc Version: 8.1.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: fortran Assignee: unassigned at gcc dot gnu.org Reporter: only_for_nouse at gmx dot de Target Milestone: --- Created attachment 44490 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=44490&action=edit REproducer of the error Compiling the attached code given the error: loc_g => petsc_o%g_dv%get_ptr() 1 Error: Assignment to contiguous pointer from non-contiguous target at (1) although the get_ptr function of the dist_func_t datatype returns a contiguous pointer. Remark: Using a polymorphic variable petsc_o instead of a static type, the error does not occur. The same is true if one omits one level of nesting of types.
[Bug tree-optimization/86816] [8/9 Regression] ICE: SIGSEGV in tree-ssa-pre / tail_merge_optimize
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86816 --- Comment #2 from Richard Biener --- Author: rguenth Date: Thu Aug 2 13:19:50 2018 New Revision: 263257 URL: https://gcc.gnu.org/viewcvs?rev=263257&root=gcc&view=rev Log: 2018-08-02 Richard Biener PR tree-optimization/86816 * tree-ssa-tail-merge.c (tail_merge_valueize): New function which checks for value availability before querying it. (gvn_uses_equal): Use it. (same_succ_hash): Likewise. (gimple_equal_p): Likewise. * g++.dg/torture/pr86816.C: New testcase. Added: trunk/gcc/testsuite/g++.dg/torture/pr86816.C Modified: trunk/gcc/ChangeLog trunk/gcc/testsuite/ChangeLog trunk/gcc/tree-ssa-tail-merge.c
[Bug tree-optimization/86816] [8 Regression] ICE: SIGSEGV in tree-ssa-pre / tail_merge_optimize
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86816 Richard Biener changed: What|Removed |Added Priority|P3 |P2 Known to work||9.0 Summary|[8/9 Regression] ICE: |[8 Regression] ICE: SIGSEGV |SIGSEGV in tree-ssa-pre / |in tree-ssa-pre / |tail_merge_optimize |tail_merge_optimize Known to fail|9.0 | --- Comment #3 from Richard Biener --- Fixed on trunk sofar.
[Bug target/86828] [6/7 Regression] wrong-code bug with "-march=knl -Ofast" (invalid memory reference)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86828 H.J. Lu changed: What|Removed |Added CC||hjl.tools at gmail dot com --- Comment #3 from H.J. Lu --- I can't reproduce it with gcc version 7.3.1 20180609.
[Bug target/86831] New: three failures in gfortran.fortran-torture with "-Ofast -march=native" on skylake-avx512
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86831
Bug ID: 86831
Summary: three failures in gfortran.fortran-torture with
"-Ofast -march=native" on skylake-avx512
Product: gcc
Version: 9.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: janus at gcc dot gnu.org
Target Milestone: ---
Triggered by two recent avx512 bugs with Fortran code (PR 86735 and PR 86828),
I just tried the following:
Index: gcc/testsuite/lib/fortran-torture.exp
===
--- gcc/testsuite/lib/fortran-torture.exp (revision 263255)
+++ gcc/testsuite/lib/fortran-torture.exp (working copy)
@@ -81,6 +81,7 @@
{ -O2 -fomit-frame-pointer -finline-functions -funroll-loops } \
{ -O2 -fbounds-check } \
{ -O3 -g } \
+ { -Ofast -march=native } \
{ -Os }
if { $test_tree_vectorize } {
lappend options $vectorizer_options
Running "make check-fortran" with this mod on a skylake-avx512 CPU yields the
following failures:
FAIL: gfortran.fortran-torture/execute/in-pack.f90 execution, -Ofast
-march=native
FAIL: gfortran.fortran-torture/execute/intrinsic_nearest.f90 execution, -Ofast
-march=native
FAIL: gfortran.fortran-torture/execute/nan_inf_fmt.f90 execution, -Ofast
-march=native
[Bug target/86791] mcore port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86791 --- Comment #1 from Nick Clifton --- Author: nickc Date: Thu Aug 2 13:30:40 2018 New Revision: 263258 URL: https://gcc.gnu.org/viewcvs?rev=263258&root=gcc&view=rev Log: PR target/86791 * config/mcore/mcore.c (TARGET_HAVE_SPECULATION_SAFE_VALUE): Define to speculation_safe_value_not_needed. Modified: trunk/gcc/ChangeLog trunk/gcc/config/mcore/mcore.c
[Bug target/86772] [meta-bug] tracking port status for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86772 Bug 86772 depends on bug 86791, which changed state. Bug 86791 Summary: mcore port needs updating for CVE-2017-5753 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86791 What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED
[Bug target/86791] mcore port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86791 Nick Clifton changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #2 from Nick Clifton --- I have defined TARGET_HAVE_SPECULATION_SAFE_VALUE as speculation_safe_value_not_needed.
[Bug target/86831] three failures in gfortran.fortran-torture with "-Ofast -march=native" on skylake-avx512
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86831 --- Comment #1 from janus at gcc dot gnu.org --- (In reply to janus from comment #0) > Running "make check-fortran" with this mod on a skylake-avx512 CPU yields > the following failures: > > FAIL: gfortran.fortran-torture/execute/in-pack.f90 execution, -Ofast > -march=native this one fails with "STOP 10" > FAIL: gfortran.fortran-torture/execute/intrinsic_nearest.f90 execution, > -Ofast -march=native "STOP 1" > FAIL: gfortran.fortran-torture/execute/nan_inf_fmt.f90 execution, -Ofast > -march=native "STOP 3"
[Bug target/86797] msp430 port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86797 --- Comment #1 from Nick Clifton --- Author: nickc Date: Thu Aug 2 14:02:32 2018 New Revision: 263259 URL: https://gcc.gnu.org/viewcvs?rev=263259&root=gcc&view=rev Log: PR target/86797 * config/msp430/msp430.c (TARGET_HAVE_SPECULATION_SAFE_VALUE): Define to speculation_safe_value_not_needed. Modified: trunk/gcc/ChangeLog trunk/gcc/config/msp430/msp430.c
[Bug target/86831] three failures in gfortran.fortran-torture with "-Ofast -march=native" on skylake-avx512
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86831 --- Comment #2 from janus at gcc dot gnu.org --- (In reply to janus from comment #0) > Running "make check-fortran" with this mod on a skylake-avx512 CPU yields > the following failures: None of them fails with -march=native alone.
[Bug target/86831] three failures in gfortran.fortran-torture with "-Ofast -march=native" on skylake-avx512
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86831 --- Comment #3 from janus at gcc dot gnu.org --- (In reply to janus from comment #0) > FAIL: gfortran.fortran-torture/execute/intrinsic_nearest.f90 execution, > -Ofast -march=native > FAIL: gfortran.fortran-torture/execute/nan_inf_fmt.f90 execution, -Ofast > -march=native Those two fail already with -Ofast alone.
[Bug middle-end/86505] [6/7/8/9 Regression] __builtin_va_arg_pack_len() computes the number of arguments wrongly
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86505 Richard Biener changed: What|Removed |Added Assignee|unassigned at gcc dot gnu.org |rguenth at gcc dot gnu.org --- Comment #3 from Richard Biener --- Created attachment 44491 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=44491&action=edit candidate patch I am testing this patch which fixes the testcase. Test coverage is low though.
[Bug c++/86763] [7/8/9 Regression] Wrong code comparing member of copy of a 237 byte object with nontrivial default constructor on x86-64 arch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86763 --- Comment #10 from Richard Biener --- Author: rguenth Date: Thu Aug 2 14:25:57 2018 New Revision: 263261 URL: https://gcc.gnu.org/viewcvs?rev=263261&root=gcc&view=rev Log: 2018-08-02 Richard Biener PR c++/86763 * class.c (layout_class_type): Copy TYPE_TYPELESS_STORAGE to the CLASSTYPE_AS_BASE. * g++.dg/torture/pr86763.C: New testcase. Added: trunk/gcc/testsuite/g++.dg/torture/pr86763.C Modified: trunk/gcc/ChangeLog trunk/gcc/cp/class.c trunk/gcc/testsuite/ChangeLog
[Bug c++/86763] [7/8 Regression] Wrong code comparing member of copy of a 237 byte object with nontrivial default constructor on x86-64 arch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86763 Richard Biener changed: What|Removed |Added Known to work||9.0 Summary|[7/8/9 Regression] Wrong|[7/8 Regression] Wrong code |code comparing member of|comparing member of copy of |copy of a 237 byte object |a 237 byte object with |with nontrivial default |nontrivial default |constructor on x86-64 arch |constructor on x86-64 arch --- Comment #11 from Richard Biener --- Fixed on trunk sofar.
[Bug target/86831] three failures in gfortran.fortran-torture with "-Ofast -march=native" on skylake-avx512
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86831 --- Comment #4 from Richard Biener --- "nan_inf_fmt" suggests that the 'fast' part in -Ofast might be the issue. So try -O3 -march=native instead?
[Bug target/86772] [meta-bug] tracking port status for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86772 Bug 86772 depends on bug 86797, which changed state. Bug 86797 Summary: msp430 port needs updating for CVE-2017-5753 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86797 What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED
[Bug target/86797] msp430 port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86797 Nick Clifton changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #2 from Nick Clifton --- I have defined TARGET_HAVE_SPECULATION_SAFE_VALUE as speculation_safe_value_not_needed.
[Bug target/86803] rx port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86803 --- Comment #1 from Nick Clifton --- Author: nickc Date: Thu Aug 2 14:35:10 2018 New Revision: 263263 URL: https://gcc.gnu.org/viewcvs?rev=263263&root=gcc&view=rev Log: PR target/86803 * config/rx/rx.c (TARGET_HAVE_SPECULATION_SAFE_VALUE): Define to speculation_safe_value_not_needed. Modified: trunk/gcc/ChangeLog trunk/gcc/config/rx/rx.c
[Bug target/86831] three failures in gfortran.fortran-torture with "-Ofast -march=native" on skylake-avx512
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86831 --- Comment #5 from janus at gcc dot gnu.org --- (In reply to Richard Biener from comment #4) > "nan_inf_fmt" suggests that the 'fast' part in -Ofast might be the issue. > So try -O3 -march=native instead? All three tests pass with those flags. As just mentioned, 'intrinsic_nearest' and 'nan_inf_fmt' fail with -Ofast alone (and with basically all recent gfortran versions, at least 5 to trunk). I'm not so much worried about these. However, in-pack.f90 is the only one that is made to fail by the combination of -Ofast and -march=skylake-avx512. I think it's the same problem as PR 86735 (at least it behaves in the same way). Here is a reduction that shows the failure: program main implicit none complex(kind=8) :: a8(5),b8(5) integer :: i a8 = (/(cmplx(i,-i,kind=8),i=1,5)/) b8 = (/(2*cmplx(i,-i,kind=8),i=1,5)/) call csub8(a8(5:1:-1),b8(5:1:-1),5) end program main subroutine csub8(a,b,n) implicit none complex(kind=8), dimension(n) :: a,b complex(kind=8), dimension(n) :: aa, bb integer :: n, i aa = (/(cmplx(n-i+1,i-n-1,kind=8),i=1,n)/) if (any(aa /= a)) STOP 9 bb = (/(2*cmplx(n-i+1,i-n-1,kind=8),i=1,5)/) if (any(bb /= b)) STOP 10 end subroutine csub8
[Bug target/86735] [8/9 Regression] Bad wrong-code bug with "-march=skylake-avx512 -Ofast"
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86735 --- Comment #13 from janus at gcc dot gnu.org --- gfortran.fortran-torture/execute/in-pack.f90 in the testsuite exhibits the same problem as comment 0, I think. It also contains Fortran array operations and fails with "-march=skylake-avx512 -Ofast". See PR 86831 comment 5.
[Bug c++/86190] [6/7 Regression] -Wsign-conversion ignores explicit conversion in some cases
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86190 --- Comment #12 from Tony E Lewis --- I confirm that Godbolt's GCC trunk now handles my testcase correctly. Thanks very much for all work on this.
[Bug target/86803] rx port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86803 Nick Clifton changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #2 from Nick Clifton --- I have defined TARGET_HAVE_SPECULATION_SAFE_VALUE as speculation_safe_value_not_needed.
[Bug target/86772] [meta-bug] tracking port status for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86772 Bug 86772 depends on bug 86803, which changed state. Bug 86803 Summary: rx port needs updating for CVE-2017-5753 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86803 What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED
[Bug middle-end/86827] [8/9 Regression] -Warray-bounds produces negative indicies
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86827 Martin Sebor changed: What|Removed |Added Status|UNCONFIRMED |ASSIGNED Last reconfirmed||2018-08-02 Ever confirmed|0 |1 --- Comment #2 from Martin Sebor --- Confirmed (gcc-help thread: https://gcc.gnu.org/ml/gcc-help/2018-08/msg6.html).
[Bug c++/86190] [6/7 Regression] -Wsign-conversion ignores explicit conversion in some cases
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86190 --- Comment #13 from Marek Polacek --- Thanks for taking the time to report the bug!
[Bug target/86810] v850 port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86810 --- Comment #1 from Nick Clifton --- Author: nickc Date: Thu Aug 2 15:57:06 2018 New Revision: 263264 URL: https://gcc.gnu.org/viewcvs?rev=263264&root=gcc&view=rev Log: PR target/86810 * config/v850/v850.c (TARGET_HAVE_SPECULATION_SAFE_VALUE): Define to speculation_safe_value_not_needed. Modified: trunk/gcc/ChangeLog trunk/gcc/config/v850/v850.c
[Bug middle-end/86660] libgomp.c++/for-15.C ICEs with nvptx offloading
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86660 --- Comment #16 from Tom de Vries --- Author: vries Date: Thu Aug 2 15:59:01 2018 New Revision: 263265 URL: https://gcc.gnu.org/viewcvs?rev=263265&root=gcc&view=rev Log: [nvptx] Ignore c++ exceptions The nvptx port can't support exceptions using sjlj, because ptx does not support sjlj. However, default_except_unwind_info still returns UI_SJLJ, even even if we configure with --disable-sjlj-exceptions, because UI_SJLJ is the fallback option. The reason default_except_unwind_info doesn't return UI_DWARF2 is because DWARF2_UNWIND_INFO is not defined in defaults.h, because INCOMING_RETURN_ADDR_RTX is not defined, because there's no ptx equivalent. Testcase libgomp.c++/for-15.C currently doesn't compile unless fno-exceptions is added because: - it tries to generate sjlj exception handling code, and - it tries to generate exception tables using label-addressed .byte sequence. Ptx doesn't support generating random data at a label, nor being able to load/write data relative to a label. This patch fixes the first problem by using UI_TARGET for nvptx. The second problem is worked around by generating all .byte sequences commented out. It would be better to have a narrower workaround, and define TARGET_ASM_BYTE_OP to "error: .byte unsupported " or some such. This patch does not enable exceptions for nvptx, it merely allows c++ programs to run correctly if they do no use exception handling. Build and reg-tested on x86_64 with nvptx accelerator. 2018-08-02 Tom de Vries PR target/86660 * common/config/nvptx/nvptx-common.c (nvptx_except_unwind_info): New function. Return UI_TARGET unconditionally. (TARGET_EXCEPT_UNWIND_INFO): Redefine to nvptx_except_unwind_info. * config/nvptx/nvptx.c (TARGET_ASM_BYTE_OP): Emit commented out '.byte'. * testsuite/libgomp.oacc-c++/routine-1-auto.C: Remove -fno-exceptions. * testsuite/libgomp.oacc-c++/routine-1-template-auto.C: Same. * testsuite/libgomp.oacc-c++/routine-1-template-trailing-return-type.C: Same. * testsuite/libgomp.oacc-c++/routine-1-template.C: Same. * testsuite/libgomp.oacc-c++/routine-1-trailing-return-type.C: Same. * testsuite/libgomp.oacc-c-c++-common/routine-1.c: Same. Modified: trunk/gcc/ChangeLog trunk/gcc/common/config/nvptx/nvptx-common.c trunk/gcc/config/nvptx/nvptx.c trunk/libgomp/ChangeLog trunk/libgomp/testsuite/libgomp.oacc-c++/routine-1-auto.C trunk/libgomp/testsuite/libgomp.oacc-c++/routine-1-template-auto.C trunk/libgomp/testsuite/libgomp.oacc-c++/routine-1-template-trailing-return-type.C trunk/libgomp/testsuite/libgomp.oacc-c++/routine-1-template.C trunk/libgomp/testsuite/libgomp.oacc-c++/routine-1-trailing-return-type.C trunk/libgomp/testsuite/libgomp.oacc-c-c++-common/routine-1.c
[Bug target/86810] v850 port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86810 Nick Clifton changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #2 from Nick Clifton --- I have defined TARGET_HAVE_SPECULATION_SAFE_VALUE as speculation_safe_value_not_needed.
[Bug target/86772] [meta-bug] tracking port status for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86772 Bug 86772 depends on bug 86810, which changed state. Bug 86810 Summary: v850 port needs updating for CVE-2017-5753 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86810 What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED
[Bug target/86813] xstormy16 port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86813 --- Comment #1 from Nick Clifton --- Author: nickc Date: Thu Aug 2 16:13:32 2018 New Revision: 263266 URL: https://gcc.gnu.org/viewcvs?rev=263266&root=gcc&view=rev Log: PR target/86813 * config/stormy16/stormy16.c (TARGET_HAVE_SPECULATION_SAFE_VALUE): Define to speculation_safe_value_not_needed. Modified: trunk/gcc/ChangeLog trunk/gcc/config/stormy16/stormy16.c
[Bug target/86772] [meta-bug] tracking port status for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86772 Bug 86772 depends on bug 86813, which changed state. Bug 86813 Summary: xstormy16 port needs updating for CVE-2017-5753 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86813 What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED
[Bug target/86813] xstormy16 port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86813 Nick Clifton changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #2 from Nick Clifton --- I have defined TARGET_HAVE_SPECULATION_SAFE_VALUE as speculation_safe_value_not_needed.
[Bug tree-optimization/86702] [9 Regression] SPEC CPU2006 400.perlbench, CPU2017 500.perlbench_r ~3% performance drop after r262247
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86702 --- Comment #4 from Alexander Nesterovskiy --- I've noticed performance regressions on different targets and with different compilation options, not only highly optimized like "-march=skylake-avx512 -Ofast -flto -funroll-loops" but with "-O2" too. The simplest case is 500.perlbench_r with "-O2" on Broadwell executed in one copy. Performance drop is not in a particular place but "spread" over whole S_regmatch function which is really big. My guess was that loosing of these probabilities affects passes that follows tree-switchlower1. And it is what I see in generated assembly - some different spilling/filling and different order of blocks.
[Bug tree-optimization/86552] missing warning for reading past the end of non-string arrays
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86552 --- Comment #5 from Martin Sebor --- Yes, that's also (partly) why I submitted a solution for pr71625 comment 15: https://gcc.gnu.org/ml/gcc-patches/2018-07/msg01884.html With that patch applied as well the strlen call is diagnosed.
[Bug c/86832] New: GCC v8.2.0 tries to use native TLS with -fstack-protector-strong on Windows (mingw-w64)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86832
Bug ID: 86832
Summary: GCC v8.2.0 tries to use native TLS with
-fstack-protector-strong on Windows (mingw-w64)
Product: gcc
Version: lto
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
Assignee: unassigned at gcc dot gnu.org
Reporter: johannes.schindelin at gmx dot de
Target Milestone: ---
When I try to compile this program:
-- snip --
static void a1(void *p) { }
int main(int argc, char **argv) {
int i;
a1(&i);
return 0;
}
-- snap --
using
gcc -o a1.o -c -fstack-protector-strong a1.c
gcc -fstack-protector-strong -o a1.exe a1.o
the resulting a1.exe causes a segmentation fault in `main()`. The offending
assembler instruction is this:
0x40156f : mov %fs:0x0,%rax
The good mingw-w64 people pointed out that this looks like native TLS, but
mingw-w64 only supports emulated TLS.
When compiling without -fstack-protector-strong, everything works.
Output of `gcc -v`:
Using built-in specs.
COLLECT_GCC=C:\git-sdk-64-ci\mingw64\bin\gcc.exe
COLLECT_LTO_WRAPPER=C:/git-sdk-64-ci/mingw64/bin/../lib/gcc/x86_64-w64-mingw32/8.2.0/lto-wrapper.exe
Target: x86_64-w64-mingw32
Configured with: ../gcc-8.2.0/configure --prefix=/mingw64
--with-local-prefix=/mingw64/local --build=x86_64-w64-mingw32
--host=x86_64-w64-mingw32 --target=x86_64-w64-mingw32
--with-native-system-header-dir=/mingw64/x86_64-w64-mingw32/include
--libexecdir=/mingw64/lib --enable-bootstrap --with-arch=x86-64
--with-tune=generic --enable-languages=ada,c,lto,c++,objc,obj-c++,fortran
--enable-shared --enable-static --enable-libatomic --enable-threads=posix
--enable-graphite --enable-fully-dynamic-string
--enable-libstdcxx-filesystem-ts=yes --enable-libstdcxx-time=yes
--disable-libstdcxx-pch --disable-libstdcxx-debug --disable-isl-version-check
--enable-lto --enable-libgomp --disable-multilib --enable-checking=release
--disable-rpath --disable-win32-registry --disable-nls --disable-werror
--disable-symvers --with-libiconv --with-system-zlib --with-gmp=/mingw64
--with-mpfr=/mingw64 --with-mpc=/mingw64 --with-isl=/mingw64
--with-pkgversion='Rev1, Built by MSYS2 project'
--with-bugurl=https://sourceforge.net/projects/msys2 --with-gnu-as
--with-gnu-ld
Thread model: posix
gcc version 8.2.0 (Rev1, Built by MSYS2 project)
[Bug c++/80744] Detect Divide By Zero and give a warning in C/C++
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80744
Martin Sebor changed:
What|Removed |Added
CC||msebor at gcc dot gnu.org
--- Comment #3 from Martin Sebor ---
I agree that diagnosing (A) or (C) under -Wdiv-by-zero would not be appropriate
because the option controls compile-time division by zero and there is none
here.
Extending -Wdiv-by-zero to diagnose possible division by zero would undoubtedly
make the option exceedingly noisy. Introducing a new option such as
-Wmaybe-div-by-zero analogous to -Wmaybe-uninitialized would be fine but the
option wouldn't diagnose ether (A) or (C) because there is no evidence of
overflow. -Wmaybe-uninitialized triggers under very restrictive conditions,
when there is some evidence that an uninitialized variable is used. To avoid
excessive noise from -Wmaybe-div-by-zero some similar approach would need to
used. Such as:
int f (int i)
{
int j;
if (i == 0) // because of this test
j = 7;
else
j = 13;
return j / i; // i may be assumed to be zero here
}
[Bug target/86828] [6/7 Regression] wrong-code bug with "-march=knl -Ofast" (invalid memory reference)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86828 --- Comment #4 from janus at gcc dot gnu.org --- (In reply to H.J. Lu from comment #3) > I can't reproduce it with gcc version 7.3.1 20180609. Just made a fresh build off today's 7-branch: gcc version 7.3.1 20180802 [gcc-7-branch revision 263263] (GCC) With this I see the same segfault as with the 7.3.0 version provided by Ubuntu 18.04. Both '-march=knl' and '-march=skylake-avx512' trigger the error in my tests (when combined with -Ofast). CPU is Core(TM) i9-7980XE, OS is Ubuntu 18.04. What's yours H.J.?
[Bug libgcc/86512] Incorrect sub result for float subnormal inputs in armv7(with -msoft-float).
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86512 --- Comment #1 from Richard Earnshaw --- Author: rearnsha Date: Thu Aug 2 16:50:07 2018 New Revision: 263267 URL: https://gcc.gnu.org/viewcvs?rev=263267&root=gcc&view=rev Log: arm - correctly handle denormal results during softfp subtraction 2018-08-02 Nicolas Pitre PR libgcc/86512 * config/arm/ieee754-df.S (adddf3): Don't shortcut denormal handling when exponent goes negative. Update my email address. * config/arm/ieee754-sf.S (addsf3): Likewise. Modified: trunk/libgcc/ChangeLog trunk/libgcc/config/arm/ieee754-df.S trunk/libgcc/config/arm/ieee754-sf.S
[Bug libgcc/86512] Incorrect sub result for float subnormal inputs in armv7(with -msoft-float).
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86512 Richard Earnshaw changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #2 from Richard Earnshaw --- fixed on trunk.
[Bug libgcc/86512] Incorrect sub result for float subnormal inputs in armv7(with -msoft-float).
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86512 Richard Earnshaw changed: What|Removed |Added Target Milestone|--- |9.0
[Bug c++/80744] Detect Divide By Zero and give a warning in C/C++
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80744
--- Comment #4 from Jonny Grant ---
Hi Martin
Thank you for your reply.
I appreciate that GCC isn't a static analyser
But I think (C) is probably easier to follow though if GCC was extended, and
(A) std::string harder, as needs knowledge of that container... the static
instrumented code would need to call mystring.size() and see what the value
would be?
int main()
{
std::string empty;
test_func(0, empty)
}
I've attached a test case where (B) is not detected, because "const" is
removed.
[Bug c++/81239] std::__cxx11::string& visible in gcc warning output
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81239 --- Comment #2 from Jonny Grant --- Created attachment 44492 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=44492&action=edit Better test case
[Bug c++/81239] std::__cxx11::string& visible in gcc warning output
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81239 --- Comment #3 from Jonny Grant --- (In reply to Jonny Grant from comment #2) > Created attachment 44492 [details] > Better test case wrong ticket. please ignore
[Bug c++/80744] Detect Divide By Zero and give a warning in C/C++
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80744 --- Comment #5 from Jonny Grant --- Created attachment 44493 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=44493&action=edit Better test case
[Bug target/86820] [8 regression] gcc-8 miscompiles binutils on m68k
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86820 --- Comment #5 from Andreas Schwab --- Author: schwab Date: Thu Aug 2 17:23:13 2018 New Revision: 263269 URL: https://gcc.gnu.org/viewcvs?rev=263269&root=gcc&view=rev Log: Fix invalid cc_status after [const_][us]mulsi3_highpart Backport from mainline PR target/86820 * config/m68k/m68k.md (umulsi3_highpart+1, const_umulsi3_highpart) (smulsi3_highpart+1, const_smulsi3_highpart): Add CC_STATUS_INIT. testsuite/: PR target/86820 * gcc.target/m68k/mulsi_highpart.c: New test. Added: branches/gcc-8-branch/gcc/testsuite/gcc.target/m68k/mulsi_highpart.c Modified: branches/gcc-8-branch/gcc/ChangeLog branches/gcc-8-branch/gcc/config/m68k/m68k.md branches/gcc-8-branch/gcc/testsuite/ChangeLog
[Bug target/86820] [8 regression] gcc-8 miscompiles binutils on m68k
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86820 Andreas Schwab changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #6 from Andreas Schwab --- Fixed.
[Bug target/86784] H8 port needs updating for CVE-2017-5753
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86784 --- Comment #1 from Jeffrey A. Law --- Author: law Date: Thu Aug 2 17:24:59 2018 New Revision: 263270 URL: https://gcc.gnu.org/viewcvs?rev=263270&root=gcc&view=rev Log: PR target/86784 * config/h8300/h8300.c (TARGET_HAVE_SPECULATION_SAFE_VALUE): Define to speculation_safe_value_not_needed. Modified: trunk/gcc/ChangeLog trunk/gcc/config/h8300/h8300.c
