[Bug fortran/54070] [4.9/5 Regression] Wrong code with allocatable deferred-length (array) function results
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=54070 --- Comment #30 from neil.n.carlson at gmail dot com --- Paul, you've done a lot of great work here (a huge thanks!) and I can confirm that many of my deferred-length character issues seem to be resolved now with the trunk (r232457, 1/15/2016). I'm uncertain though whether you consider this PR resolved or if you are still working on it (there were lots of examples in PRs marked as duplicates). I still have at least one remaining issue with this trunk version: class(*), allocatable :: x(:) allocate(x, source=['foo','bar']) end Compiles fine, but produces a run time seg fault: Program received signal SIGSEGV: Segmentation fault - invalid memory reference. Backtrace for this error: #0 0x7F85F60F5517 #1 0x7F85F60F5B5E #2 0x7F85F55F695F #3 0x7F85F56507ED #4 0x400932 in __copy_character_1.3416 at bug.f90:? #5 0x400B91 in MAIN__ at bug.f90:? Segmentation fault (core dumped) Should I open another PR for this?
[Bug tree-optimization/69336] Constant value not detected
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69336 alalaw01 at gcc dot gnu.org changed: What|Removed |Added CC||alalaw01 at gcc dot gnu.org --- Comment #4 from alalaw01 at gcc dot gnu.org --- That looks reasonable, AFAICT get_ref_base_and_extent will deal with anything that is handled_component_p. The same patch enables the optimization on aarch64, with appropriate --param sra-max-scalarization-size-Ospeed to pull the constant-pool entry in.
[Bug rtl-optimization/69052] [6 Regression] Performance regression after r229402.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69052 Igor Zamyatin changed: What|Removed |Added CC||izamyatin at gmail dot com --- Comment #3 from Igor Zamyatin --- (In reply to amker from comment #2) > It's my change, I will look into it. Any plans on this?
[Bug fortran/54070] [4.9/5 Regression] Wrong code with allocatable deferred-length (array) function results
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=54070 --- Comment #31 from neil.n.carlson at gmail dot com --- Sorry, ignore the example of comment 30. I had already reported this in PR 67564 (not a duplicate of this one). I'm getting old ...
[Bug target/47122] vax-*-openbsd* configuration purports to require openbsd-pthread.h
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=47122 Bernd Schmidt changed: What|Removed |Added CC||bernds at gcc dot gnu.org --- Comment #3 from Bernd Schmidt --- Can this be closed?
[Bug pch/68176] [4.9/5/6 Regression] all pch tests fail on eglibc systems (with bits/predefs.h)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68176 --- Comment #5 from Nix --- I didn't think of that (I try to forget that fixincludes exists because it gives me nightmares). But much though I hate fixincludes, this sort of fix (to headers for an obsolete program which will by definition never be fixed) is more or less exactly what it was designed for, and seems a perfect fit for this case. The question is whether further adjustments are needed to compensate for the removal of ... I'll do some experimentation.
[Bug middle-end/68542] [6 Regression] 10% 481.wrf performance regression
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68542 --- Comment #5 from Ilya Enkovich --- Author: ienkovich Date: Mon Jan 18 14:14:35 2016 New Revision: 232518 URL: https://gcc.gnu.org/viewcvs?rev=232518=gcc=rev Log: gcc/ 2016-01-18 Yuri RumyantsevPR middle-end/68542 * fold-const.c (fold_binary_op_with_conditional_arg): Bail out for case of mixind vector and scalar types. (fold_relational_const): Add handling of vector comparison with boolean result. * tree-cfg.c (verify_gimple_comparison): Add argument CODE, allow comparison of vector operands with boolean result for EQ/NE only. (verify_gimple_assign_binary): Adjust call for verify_gimple_comparison. (verify_gimple_cond): Likewise. * tree-vrp.c (extract_code_and_val_from_cond_with_ops): Modify check on valid type of VAL. Modified: trunk/gcc/ChangeLog trunk/gcc/fold-const.c trunk/gcc/tree-cfg.c trunk/gcc/tree-vrp.c
[Bug tree-optimization/69328] [6 Regression] ice in vect_get_vec_def_for_operand, at tree-vect-stmts.c:1379 with -O3
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69328 --- Comment #4 from Ilya Enkovich --- (In reply to Richard Biener from comment #3) > > ./cc1 -quiet t.c -O3 > t.c: In function ‘fn1’: > t.c:2:6: internal compiler error: in vector_compare_rtx, at optabs.c:5290 > void fn1() { > ^~~ > > 0xca84f2 vector_compare_rtx > /space/rguenther/src/svn/trunk3/gcc/optabs.c:5290 > 0xca979f expand_vec_cond_expr(tree_node*, tree_node*, tree_node*, > tree_node*, rtx_def*) > /space/rguenther/src/svn/trunk3/gcc/optabs.c:5612 > > sth else is wrong here. Looks like vectype1 and vectype2 in vect_is_simple_cond have different number of elements and we don't catch it.
[Bug tree-optimization/69297] [6 Regression] Performance regression after r230020
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69297 --- Comment #5 from Richard Biener --- Author: rguenth Date: Mon Jan 18 14:25:56 2016 New Revision: 232519 URL: https://gcc.gnu.org/viewcvs?rev=232519=gcc=rev Log: 2016-01-18 Richard BienerPR tree-optimization/69297 * tree-vect-slp.c (vect_bb_slp_scalar_cost): Count each scalar stmt at most once. (vect_bb_vectorization_profitable_p): Clear visited flag again. * gcc.dg/vect/costmodel/x86_64/costmodel-pr69297.c: New testcase. Added: trunk/gcc/testsuite/gcc.dg/vect/costmodel/x86_64/costmodel-pr69297.c Modified: trunk/gcc/ChangeLog trunk/gcc/testsuite/ChangeLog trunk/gcc/tree-vect-slp.c
[Bug tree-optimization/69297] [6 Regression] Performance regression after r230020
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69297 Richard Biener changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|--- |FIXED --- Comment #6 from Richard Biener --- Fixed.
[Bug target/69345] [6 Regression] 459.GemsFDTD regression
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69345 --- Comment #1 from Richard Biener --- candidates are r232435 and r232401 I think (which would be both mine).
[Bug target/69345] [6 Regression] 459.GemsFDTD regression
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69345 Alexander Fomin changed: What|Removed |Added CC||afomin.mailbox at gmail dot com --- Comment #2 from Alexander Fomin --- I can confirm we've also seen it on Core i7. Now bisecting.
[Bug bootstrap/64919] bootstrap failure of gcc-4.9.2 on ia64-hpux in libgcc
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64919 --- Comment #29 from The Written Word --- (In reply to Alexander from comment #28) > this one file should recompile with -O1 optimization Thanks. I rebuilt with charset.c with -O1 and it compiled. I resumed the build and the compile now fails with: /opt/build/china/gcc-4.9.3/.obj/./prev-gcc/xg++ -B/opt/build/china/gcc-4.9.3/.obj/./prev-gcc/ -B/opt/build/gcc49/ia64-hp-hpux11.31/bin/ -nostdinc++ -B/opt/build/china/gcc-4.9.3/.obj/prev-ia64-hp-hpux11.31/libstdc++-v3/src/.libs -B/opt/build/china/gcc-4.9.3/.obj/prev-ia64-hp-hpux11.31/libstdc++-v3/libsupc++/.libs -I/opt/build/china/gcc-4.9.3/.obj/prev-ia64-hp-hpux11.31/libstdc++-v3/include/ia64-hp-hpux11.31 -I/opt/build/china/gcc-4.9.3/.obj/prev-ia64-hp-hpux11.31/libstdc++-v3/include -I/opt/build/china/gcc-4.9.3/libstdc++-v3/libsupc++ -L/opt/build/china/gcc-4.9.3/.obj/prev-ia64-hp-hpux11.31/libstdc++-v3/src/.libs -L/opt/build/china/gcc-4.9.3/.obj/prev-ia64-hp-hpux11.31/libstdc++-v3/libsupc++/.libs -c -DUSE_LIBUNWIND_EXCEPTIONS -DIN_GCC_FRONTEND -g -O2 -DIN_GCC -fno-exceptions -fno-rtti -fasynchronous-unwind-tables -W -Wall -Wno-narrowing -Wwrite-strings -Wcast-qual -Wmissing-format-attribute -pedantic -Wno-long-long -Wno-variadic-macros -Wno-overlength-strings -DHAVE_CONFIG_H -I. -Ifortran -I/opt/build/china/gcc-4.9.3/gcc -I/opt/build/china/gcc-4.9.3/gcc/fortran -I/opt/build/china/gcc-4.9.3/gcc/../include -I./../intl -I/opt/build/china/gcc-4.9.3/gcc/../libcpp/include -I/opt/TWWfsw/libgmp51/include -I/opt/TWWfsw/libmpfr31/include -I/opt/TWWfsw/libmpc10/include -I/opt/build/china/gcc-4.9.3/gcc/../libdecnumber -I/opt/build/china/gcc-4.9.3/gcc/../libdecnumber/dpd -I../libdecnumber -I/opt/build/china/gcc-4.9.3/gcc/../libbacktrace-o fortran/scanner.o -MT fortran/scanner.o -MMD -MP -MF fortran/.deps/scanner.TPo /opt/build/china/gcc-4.9.3/gcc/fortran/scanner.c /opt/build/china/gcc-4.9.3/gcc/fortran/scanner.c: In function 'const char* gfc_read_orig_filename(const char*, const char**)': /opt/build/china/gcc-4.9.3/gcc/fortran/scanner.c:2220:1: internal compiler error: in plus_constant, at explow.c:87 } ^
[Bug target/68620] ICE on gcc.target/arm/attr-neon-fp16.c
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68620 Christophe Lyon changed: What|Removed |Added Status|UNCONFIRMED |ASSIGNED Last reconfirmed|2015-12-03 00:00:00 |2016-01-18 Ever confirmed|0 |1 --- Comment #7 from Christophe Lyon --- Patch posted at https://gcc.gnu.org/ml/gcc-patches/2016-01/msg01031.html
[Bug target/69305] [5/6 Regression] wrong code with -O and int128 @ aarch64
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69305 Jakub Jelinek changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed||2016-01-18 Ever confirmed|0 |1 --- Comment #4 from Jakub Jelinek --- Reduced testcase: __attribute__((noinline, noclone)) unsigned __int128 foo (unsigned __int128 x, unsigned long long y) { return x + (unsigned long long) -y; } int main () { if (foo (70, 0) != 70) __builtin_abort (); return 0; }
[Bug target/69344] [6 Regression] 435.gromacs regression
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69344 Igor Zamyatin changed: What|Removed |Added CC||izamyatin at gmail dot com --- Comment #1 from Igor Zamyatin --- Same as PR69274
[Bug target/69344] [6 Regression] 435.gromacs regression
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69344 Richard Biener changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |DUPLICATE --- Comment #2 from Richard Biener --- Dup then. *** This bug has been marked as a duplicate of bug 69274 ***
[Bug target/69274] [6 Regression] 435.gromacs performance regression after r231814 on x86 Haswell and bdver2
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69274 Richard Biener changed: What|Removed |Added Keywords||missed-optimization Status|UNCONFIRMED |NEW Last reconfirmed||2016-01-18 Ever confirmed|0 |1 --- Comment #2 from Richard Biener --- Confirmed (see dup).
[Bug target/69274] [6 Regression] 435.gromacs performance regression after r231814 on x86 Haswell and bdver2
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69274 Richard Biener changed: What|Removed |Added CC||rguenth at gcc dot gnu.org --- Comment #1 from Richard Biener --- *** Bug 69344 has been marked as a duplicate of this bug. ***
[Bug target/43052] [4.9/5/6 Regression] Inline memcmp is *much* slower than glibc's, no longer expanded inline
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43052 Bernd Schmidt changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED --- Comment #35 from Bernd Schmidt --- Closing since PR52171 tracks the other problem.
[Bug target/69305] [5/6 Regression] wrong code with -O and int128 @ aarch64
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69305 --- Comment #7 from Jakub Jelinek --- The patterns are just weird. (insn 10 7 11 2 (parallel [ (set (reg:CC_NZ 66 cc) (compare:CC_NZ (plus:DI (reg:DI 79) (reg:DI 85 [ x ])) (const_int 0 [0]))) (set (reg:DI 81) (plus:DI (reg:DI 79) (reg:DI 85 [ x ]))) ]) pr69305-1.c:5 100 {adddi3_compare0} (expr_list:REG_DEAD (reg:DI 85 [ x ]) (expr_list:REG_DEAD (reg:DI 79) (nil (insn 11 10 25 2 (set (reg:DI 82) (plus:DI (geu:DI (reg:CC 66 cc) (const_int 0 [0])) (reg:DI 86 [ x+8 ]))) pr69305-1.c:5 452 {*csinc2di_insn} (expr_list:REG_DEAD (reg:DI 86 [ x+8 ]) (expr_list:REG_DEAD (reg:CC 66 cc) (nil This suggests that that you compare (a + b) cmp 0 and add 1 in the second insn if (a + b) u>= 0, which is always, while what you are actually interested in is whether (a + b) u< a (or equivalently (a + b) u< b), i.e. if there is an overflow in the addition. E.g. on x86_64 (though, only after split2) it is represented that way: (insn 33 13 34 2 (parallel [ (set (reg:CCC 17 flags) (compare:CCC (plus:DI (reg:DI 0 ax [97]) (reg:DI 38 r9 [orig:90 x ] [90])) (reg:DI 0 ax [97]))) (set (reg:DI 0 ax [95]) (plus:DI (reg:DI 0 ax [97]) (reg:DI 38 r9 [orig:90 x ] [90]))) ]) pr69305-1.c:5 306 {*adddi3_cc_overflow_1} (nil)) (insn 34 33 20 2 (parallel [ (set (reg:DI 1 dx [+8 ]) (plus:DI (plus:DI (ltu:DI (reg:CC 17 flags) (const_int 0 [0])) (reg:DI 1 dx [+8 ])) (reg:DI 39 r10 [ x+8 ]))) (clobber (reg:CC 17 flags)) ]) pr69305-1.c:5 284 {adddi3_carry} (nil))
[Bug bootstrap/64919] bootstrap failure of gcc-4.9.2 on ia64-hpux in libgcc
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64919 --- Comment #30 from The Written Word --- (In reply to The Written Word from comment #29) > (In reply to Alexander from comment #28) > > this one file should recompile with -O1 optimization > > Thanks. I rebuilt with charset.c with -O1 and it compiled. I resumed the > build and the compile now fails with: > /opt/build/china/gcc-4.9.3/.obj/./prev-gcc/xg++ > -B/opt/build/china/gcc-4.9.3/.obj/./prev-gcc/ > -B/opt/build/gcc49/ia64-hp-hpux11.31/bin/ -nostdinc++ > -B/opt/build/china/gcc-4.9.3/.obj/prev-ia64-hp-hpux11.31/libstdc++-v3/src/. > libs > -B/opt/build/china/gcc-4.9.3/.obj/prev-ia64-hp-hpux11.31/libstdc++-v3/ > libsupc++/.libs > -I/opt/build/china/gcc-4.9.3/.obj/prev-ia64-hp-hpux11.31/libstdc++-v3/ > include/ia64-hp-hpux11.31 > -I/opt/build/china/gcc-4.9.3/.obj/prev-ia64-hp-hpux11.31/libstdc++-v3/ > include -I/opt/build/china/gcc-4.9.3/libstdc++-v3/libsupc++ > -L/opt/build/china/gcc-4.9.3/.obj/prev-ia64-hp-hpux11.31/libstdc++-v3/src/. > libs > -L/opt/build/china/gcc-4.9.3/.obj/prev-ia64-hp-hpux11.31/libstdc++-v3/ > libsupc++/.libs -c -DUSE_LIBUNWIND_EXCEPTIONS -DIN_GCC_FRONTEND -g -O2 > -DIN_GCC-fno-exceptions -fno-rtti -fasynchronous-unwind-tables -W -Wall > -Wno-narrowing -Wwrite-strings -Wcast-qual -Wmissing-format-attribute > -pedantic -Wno-long-long -Wno-variadic-macros -Wno-overlength-strings > -DHAVE_CONFIG_H -I. -Ifortran -I/opt/build/china/gcc-4.9.3/gcc > -I/opt/build/china/gcc-4.9.3/gcc/fortran > -I/opt/build/china/gcc-4.9.3/gcc/../include -I./../intl > -I/opt/build/china/gcc-4.9.3/gcc/../libcpp/include > -I/opt/TWWfsw/libgmp51/include -I/opt/TWWfsw/libmpfr31/include > -I/opt/TWWfsw/libmpc10/include > -I/opt/build/china/gcc-4.9.3/gcc/../libdecnumber > -I/opt/build/china/gcc-4.9.3/gcc/../libdecnumber/dpd -I../libdecnumber > -I/opt/build/china/gcc-4.9.3/gcc/../libbacktrace-o fortran/scanner.o -MT > fortran/scanner.o -MMD -MP -MF fortran/.deps/scanner.TPo > /opt/build/china/gcc-4.9.3/gcc/fortran/scanner.c > /opt/build/china/gcc-4.9.3/gcc/fortran/scanner.c: In function 'const char* > gfc_read_orig_filename(const char*, const char**)': > /opt/build/china/gcc-4.9.3/gcc/fortran/scanner.c:2220:1: internal compiler > error: in plus_constant, at explow.c:87 > } > ^ Rebuilding with -O0 or with no optimization specified worked around this. Is that what others have done?
[Bug tree-optimization/69308] ifcombine joins together floating point expression with side effects
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69308 Richard Biener changed: What|Removed |Added Known to work||6.0 --- Comment #12 from Richard Biener --- Fixed for GCC 6 sofar.
[Bug tree-optimization/69308] ifcombine joins together floating point expression with side effects
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69308 --- Comment #13 from Richard Biener --- Author: rguenth Date: Mon Jan 18 13:03:54 2016 New Revision: 232516 URL: https://gcc.gnu.org/viewcvs?rev=232516=gcc=rev Log: 2016-01-18 Richard BienerPR middle-end/69308 * gimple.c (gimple_could_trap_p_1): Handle GIMPLE_COND. Modified: trunk/gcc/ChangeLog trunk/gcc/gimple.c
[Bug target/69344] New: [6 Regression] 435.gromacs regression
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69344 Bug ID: 69344 Summary: [6 Regression] 435.gromacs regression Product: gcc Version: 6.0 Status: UNCONFIRMED Keywords: missed-optimization Severity: normal Priority: P3 Component: target Assignee: unassigned at gcc dot gnu.org Reporter: rguenth at gcc dot gnu.org Target Milestone: --- Target: x86_64-*-* Our two SPEC 2k6 testers show recent (mid December) regressions for 435.gromacs: http://gcc.opensuse.org/SPEC/CFP/sb-czerny-head-64-2006/index.html http://gcc.opensuse.org/SPEC/CFP/sb-megrez-head-64-2006/index.html which can be narrowed down to r231790 (last good) r231841 (first known bad) for the first link and r231805 (last good) r231814 (first known bad) for the second. Both testers use -Ofast -march=native, one is corei-avx2 one bdver2 This effectively means (given gromas is C only) I'd blame r231814 which is 2015-12-18 Andreas Krebbel* ira.c (ira_setup_alts): Move the scan for commutative modifier to the first loop to make it work even with disabled alternatives. but this means it is a backend issue. Didn't yet try narrowing down to a testcase or try confirming the causing revision.
[Bug target/69344] [6 Regression] 435.gromacs regression
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69344 Richard Biener changed: What|Removed |Added Target Milestone|--- |6.0
[Bug target/69345] [6 Regression] 459.GemsFDTD regression
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69345 Richard Biener changed: What|Removed |Added Target Milestone|--- |6.0
[Bug sanitizer/67515] crash from ubsan for non-virtual call in initializer list with an invalid vtable
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=67515 --- Comment #11 from Yury V. Zaytsev --- Hi Roger, Thank you for the hint! I've tried the solution from the linked ticket, but I'm still getting the same problem, albeit at a different place in the code (not sure why?!). In addition I'm still struggling with undefined behaviors in other parts of boost, like this one: https://svn.boost.org/trac/boost/ticket/11204 ... for now I'll try to remove the use of boost::lexical_cast where possible :-/ Thanks!
[Bug pch/68176] [4.9/5/6 Regression] all pch tests fail on eglibc systems (with bits/predefs.h)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68176 Jakub Jelinek changed: What|Removed |Added CC||jakub at gcc dot gnu.org, ||jsm28 at gcc dot gnu.org --- Comment #4 from Jakub Jelinek --- So shouldn't we just fixinclude the bad stdc-predef.h headers and resolve this that way? IMNSHO stdc-predef.h really shouldn't include anything.
[Bug lto/69133] [6 Regression] LTO segfault in lto_get_decl_name_mapping() on 483.xalancbmk with -flto-partition=none
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69133 Jan Hubicka changed: What|Removed |Added Status|NEW |ASSIGNED Assignee|unassigned at gcc dot gnu.org |hubicka at gcc dot gnu.org
[Bug target/69305] [5/6 Regression] wrong code with -O and int128 @ aarch64
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69305 --- Comment #5 from ktkachov at gcc dot gnu.org --- Confirmed as well. If combine changed the plus-compare into a minus-compare, shouldn't it also go into the condition code usage and update that too though?
[Bug target/69305] [5/6 Regression] wrong code with -O and int128 @ aarch64
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69305 --- Comment #6 from ktkachov at gcc dot gnu.org --- (In reply to ktkachov from comment #5) > Confirmed as well. > If combine changed the plus-compare into a minus-compare, shouldn't it also > go into the condition code usage and update that too though? scratch that, I was thinking of something else
[Bug target/69345] New: [6 Regression] 459.GemsFDTD regression
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69345 Bug ID: 69345 Summary: [6 Regression] 459.GemsFDTD regression Product: gcc Version: 6.0 Status: UNCONFIRMED Keywords: missed-optimization Severity: normal Priority: P3 Component: target Assignee: unassigned at gcc dot gnu.org Reporter: rguenth at gcc dot gnu.org Target Milestone: --- Our two SPEC 2006 testers show a 459.GemsFDTD regression a few days ago, the bdver2 tester between r232400 and r232444 while the corei7-avx2 one between r232395 and r232457. No further bisection / confirmation done yet. GemsFDTD is Fortran.
[Bug target/69305] [5/6 Regression] wrong code with -O and int128 @ aarch64
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69305 --- Comment #8 from ktkachov at gcc dot gnu.org --- (In reply to Jakub Jelinek from comment #7) > The patterns are just weird. All that comes from the addti3 expander in aarch64.md If I delete it the testcase doesn't abort. I'll have a closer look there.
[Bug target/69305] [5/6 Regression] wrong code with -O and int128 @ aarch64
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69305 ktkachov at gcc dot gnu.org changed: What|Removed |Added CC||rth at gcc dot gnu.org --- Comment #9 from ktkachov at gcc dot gnu.org --- (In reply to ktkachov from comment #8) > (In reply to Jakub Jelinek from comment #7) > > The patterns are just weird. > > All that comes from the addti3 expander in aarch64.md > If I delete it the testcase doesn't abort. > I'll have a closer look there. Every approach I've tried to implement addti3 properly (that is, comparing (a + b) u< a instead of (a + b) u>= 0) doesn't give better code than 4.9 which didn't have the custom expanders. Richard, you added these expanders. Any ideas on how to implement them properly and better than the fallback? Or should we just delete them?
[Bug lto/68881] [6 Regression] UNRESOLVED/FAIL: gcc.dg/lto/attr-weakref-1 -O2 -flto
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68881 --- Comment #7 from Uroš Bizjak --- (In reply to Richard Biener from comment #2) > Works for me but > https://gcc.gnu.org/ml/gcc-testresults/2016-01/msg01237.html still has it. > > Maybe some as feature stuff? Tom, what target did you reproduce on? I'm able to reproduce this on Fedora 23 using -fno-use-linker-plugin: $ /ssd/uros/gcc-build/gcc/xgcc -B /ssd/uros/gcc-build/gcc -O2 -flto -fno-use-linker-plugin -c attr-weakref-1_0.c $ /ssd/uros/gcc-build/gcc/xgcc -B /ssd/uros/gcc-build/gcc -O2 -flto -fno-use-linker-plugin -c attr-weakref-1_1.c $ /ssd/uros/gcc-build/gcc/xgcc -B /ssd/uros/gcc-build/gcc -O2 -flto -fno-use-linker-plugin -c attr-weakref-1_2.c $ /ssd/uros/gcc-build/gcc/xgcc -B /ssd/uros/gcc-build/gcc -O2 -flto -fno-use-linker-plugin attr-weakref-1_0.o attr-weakref-1_1.o attr-weakref-1_2.o /tmp/ccydxp6I.s: Assembler messages: /tmp/ccydxp6I.s: Error: symbol definition loop encountered at `callmealias.lto_priv.1' /tmp/ccydxp6I.s: Error: symbol definition loop encountered at `callmealias.lto_priv.0' lto-wrapper: fatal error: /ssd/uros/gcc-build/gcc/xgcc returned 1 exit status compilation terminated. collect2: fatal error: lto-wrapper returned 1 exit status compilation terminated. (The compilation works OK w/o -fno-use-linker-plugin.)
[Bug fortran/54070] [4.9/5 Regression] Wrong code with allocatable deferred-length (array) function results
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=54070 --- Comment #32 from Paul Thomas --- (In reply to neil.n.carlson from comment #31) > Sorry, ignore the example of comment 30. I had already reported this in PR > 67564 (not a duplicate of this one). I'm getting old ... Thanks for your appreciation of my efforts! I was going to attack some of the remaining deferred character issues after a spell of regression fixing. I have taken note of PR67564 and will give it a poke around tonight. Best regards Paul
[Bug fortran/66680] [5 Regression] ICE with openmp, a loop and a type bound procedure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66680 Dominique d'Humieres changed: What|Removed |Added Status|WAITING |RESOLVED Resolution|--- |FIXED --- Comment #10 from Dominique d'Humieres --- Test committed on trunk and gcc-5 branch. Closing as FIXED.
[Bug tree-optimization/69328] [6 Regression] ice in vect_get_vec_def_for_operand, at tree-vect-stmts.c:1379 with -O3
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69328 --- Comment #5 from Ilya Enkovich --- This patch works for me: diff --git a/gcc/tree-vect-stmts.c b/gcc/tree-vect-stmts.c index 635c797..9d4d286 100644 --- a/gcc/tree-vect-stmts.c +++ b/gcc/tree-vect-stmts.c @@ -7441,6 +7441,10 @@ vect_is_simple_cond (tree cond, vec_info *vinfo, tree *comp_vectype) && TREE_CODE (rhs) != FIXED_CST) return false; + if (vectype1 && vectype2 + && TYPE_VECTOR_SUBPARTS (vectype1) != TYPE_VECTOR_SUBPARTS (vectype2)) +return false; + *comp_vectype = vectype1 ? vectype1 : vectype2; return true; } @@ -7544,13 +7548,9 @@ vectorizable_condition (gimple *stmt, gimple_stmt_iterator *gsi, if (!vect_is_simple_use (else_clause, stmt_info->vinfo, _stmt, )) return false; - if (VECTOR_BOOLEAN_TYPE_P (comp_vectype)) -{ - vec_cmp_type = comp_vectype; - masked = true; -} - else -vec_cmp_type = build_same_sized_truth_vector_type (comp_vectype); + masked = !COMPARISON_CLASS_P (cond_expr); + vec_cmp_type = build_same_sized_truth_vector_type (comp_vectype); + if (vec_cmp_type == NULL_TREE) return false;
[Bug tree-optimization/69342] [6 Regression] wrong code at -O1, -O2, -O3 (NOT -Os) on x86-64-linux-gnu (in 32- and 64-bit modes)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69342 Jakub Jelinek changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed||2016-01-18 CC||jakub at gcc dot gnu.org Target Milestone|--- |6.0 Summary|wrong code at -O1, -O2, -O3 |[6 Regression] wrong code |(NOT -Os) on|at -O1, -O2, -O3 (NOT -Os) |x86-64-linux-gnu (in 32-|on x86-64-linux-gnu (in 32- |and 64-bit modes) |and 64-bit modes) Ever confirmed|0 |1 --- Comment #1 from Jakub Jelinek --- Started with r232361.
[Bug tree-optimization/69342] [6 Regression] wrong code at -O1, -O2, -O3 (NOT -Os) on x86-64-linux-gnu (in 32- and 64-bit modes)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69342 Markus Trippelsdorf changed: What|Removed |Added CC||trippels at gcc dot gnu.org --- Comment #2 from Markus Trippelsdorf --- (In reply to Jakub Jelinek from comment #1) > Started with r232361. Maybe it is time to merge all the r232361 regressions into one bug?
[Bug libgcj/69333] [6 Regression] FAIL: StackTrace2 execution - source compiled test
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69333 Jakub Jelinek changed: What|Removed |Added Priority|P3 |P4 CC||jakub at gcc dot gnu.org
[Bug tree-optimization/69320] [6 Regression] wrong code generation at -O2 and higher
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69320 Jakub Jelinek changed: What|Removed |Added CC||su at cs dot ucdavis.edu --- Comment #5 from Jakub Jelinek --- *** Bug 69326 has been marked as a duplicate of this bug. ***
[Bug target/69307] [6 Regression] wrong code with -O2 -fselective-scheduling @ armv7a
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69307 Jakub Jelinek changed: What|Removed |Added Priority|P3 |P4 CC||jakub at gcc dot gnu.org
[Bug tree-optimization/69342] [6 Regression] wrong code at -O1, -O2, -O3 (NOT -Os) on x86-64-linux-gnu (in 32- and 64-bit modes)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69342 Jakub Jelinek changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |DUPLICATE --- Comment #3 from Jakub Jelinek --- . *** This bug has been marked as a duplicate of bug 69320 ***
[Bug tree-optimization/69320] [6 Regression] wrong code generation at -O2 and higher
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69320 --- Comment #6 from Jakub Jelinek --- *** Bug 69342 has been marked as a duplicate of this bug. ***
[Bug tree-optimization/69326] [6 Regression] wrong code at -O2 and -O3 on x86_64-linux-gnu
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69326 Jakub Jelinek changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |DUPLICATE --- Comment #2 from Jakub Jelinek --- . *** This bug has been marked as a duplicate of bug 69320 ***
[Bug tree-optimization/69320] [6 Regression] wrong code generation at -O2 and higher
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69320 Jakub Jelinek changed: What|Removed |Added CC||chengniansun at gmail dot com --- Comment #4 from Jakub Jelinek --- *** Bug 69325 has been marked as a duplicate of this bug. ***
[Bug tree-optimization/69325] [6 Regression] wrong code at -O3 on x86-64-linux-gnu (in 32- and 64-bit modes)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69325 Jakub Jelinek changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |DUPLICATE --- Comment #2 from Jakub Jelinek --- . *** This bug has been marked as a duplicate of bug 69320 ***
[Bug tree-optimization/69322] [6 Regression] wrong code at -O3 on x86-64-linux-gnu (in 32- and 64-bit modes)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69322 Markus Trippelsdorf changed: What|Removed |Added Status|NEW |RESOLVED CC||trippels at gcc dot gnu.org Resolution|--- |DUPLICATE --- Comment #2 from Markus Trippelsdorf --- dup. *** This bug has been marked as a duplicate of bug 69320 ***
[Bug tree-optimization/69320] [6 Regression] wrong code generation at -O2 and higher
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69320 --- Comment #7 from Markus Trippelsdorf --- *** Bug 69322 has been marked as a duplicate of this bug. ***
[Bug ada/69219] [5/6 regression] error on nested subprograms with Inline_Always and Intrinsic
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69219 --- Comment #9 from Eric Botcazou --- Author: ebotcazou Date: Mon Jan 18 10:27:10 2016 New Revision: 232498 URL: https://gcc.gnu.org/viewcvs?rev=232498=gcc=rev Log: PR ada/69219 * gcc-interface/trans.c (check_inlining_for_nested_subprog): Consider the parent function instead of the current function in order to issue the warning or the error. Add guard for ignored functions. Added: trunk/gcc/testsuite/gnat.dg/inline12.adb Modified: trunk/gcc/ada/ChangeLog trunk/gcc/ada/gcc-interface/trans.c trunk/gcc/testsuite/ChangeLog
[Bug bootstrap/69339] [6 Regression] Failed to bootstrap powerpc-e500v2-linux-gnuspe target: libitm/vect128.cc:1:0: error: AltiVec and SPE instructions cannot coexist
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69339 Richard Biener changed: What|Removed |Added Target Milestone|--- |6.0
[Bug c++/69323] [4.9/5/6 Regression] Segmentation fault when instantiating class template with inner class which declares itself as a friend
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69323 Jakub Jelinek changed: What|Removed |Added CC||jakub at gcc dot gnu.org, ||jason at gcc dot gnu.org --- Comment #5 from Jakub Jelinek --- Started with r154223.
[Bug middle-end/52173] internal compiler error: verify_ssa failed possibly caused by itm
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=52173 Richard Biener changed: What|Removed |Added Status|REOPENED|RESOLVED Resolution|--- |FIXED --- Comment #21 from Richard Biener --- Thus fixed.
[Bug tree-optimization/69297] [6 Regression] Performance regression after r230020
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69297 Richard Biener changed: What|Removed |Added Status|UNCONFIRMED |ASSIGNED Last reconfirmed||2016-01-18 Assignee|unassigned at gcc dot gnu.org |rguenth at gcc dot gnu.org Ever confirmed|0 |1 --- Comment #2 from Richard Biener --- But note that vector stmts are accounted multiple times as well for the same reason. But yes, we can make the scalar cost more precise easily.
[Bug target/69305] [5/6 Regression] wrong code with -O and int128 @ aarch64
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69305 Richard Biener changed: What|Removed |Added Target Milestone|6.0 |5.4
[Bug c++/69300] g++ segfault on silly noexcept case
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69300 Richard Biener changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed||2016-01-18 Ever confirmed|0 |1 Known to fail||5.3.0, 6.0 --- Comment #1 from Richard Biener --- Confirmed. Rejected by 4.9. t.C: In function ‘int main(int, char**)’: t.C:8:26: error: cannot resolve overloaded function ‘f’ based on conversion to type ‘bool’ F().f(); ^ t.C: In instantiation of ‘void F::f() && [with B = int; A = void]’: t.C:8:26: required from here t.C:4:19: error: cannot resolve overloaded function ‘f’ based on conversion to type ‘bool’ void f() && noexcept(::template f) {} ^
[Bug ada/69219] [5/6 regression] error on nested subprograms with Inline_Always and Intrinsic
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69219 --- Comment #10 from Eric Botcazou --- Author: ebotcazou Date: Mon Jan 18 10:29:45 2016 New Revision: 232499 URL: https://gcc.gnu.org/viewcvs?rev=232499=gcc=rev Log: PR ada/69219 * gcc-interface/trans.c (check_inlining_for_nested_subprog): Consider the parent function instead of the current function in order to issue the warning or the error. Add guard for ignored functions. Added: branches/gcc-5-branch/gcc/testsuite/gnat.dg/inline12.adb - copied unchanged from r232498, trunk/gcc/testsuite/gnat.dg/inline12.adb Modified: branches/gcc-5-branch/gcc/ada/ChangeLog branches/gcc-5-branch/gcc/ada/gcc-interface/trans.c branches/gcc-5-branch/gcc/testsuite/ChangeLog
[Bug ada/69219] [5/6 regression] error on nested subprograms with Inline_Always and Intrinsic
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69219 Eric Botcazou changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|--- |FIXED --- Comment #11 from Eric Botcazou --- Thanks for reporting the problem.
[Bug bootstrap/69329] [6 Regression] --with-build-config=bootstrap-asan fails because LSAN_OPTIONS is not honored
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69329 Richard Biener changed: What|Removed |Added Target Milestone|--- |6.0
[Bug sanitizer/68824] [6 Regression] libtsan is missing the __interceptor___tls_get_addr symbol without bumping the soname
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68824 Jakub Jelinek changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed||2016-01-18 Ever confirmed|0 |1 --- Comment #4 from Jakub Jelinek --- So, this seems to have been removed on purpose as I read: http://llvm.org/viewvc/llvm-project?view=revision=224531 to workaround an already fixed gcc bug. So, I see 3 options: 1) bump libtsan SONAME 2) re-add the __tls_get_addr interceptor, but arrange for the interceptor to only do return __interception::real___tls_get_addr(arg); and nothing else 3) re-add the __tls_get_addr interceptor, but arrange for it to be defined in a different source file, and compile that with -mincoming-stack-boundary=3 on x86_64-linux
[Bug tree-optimization/69297] [6 Regression] Performance regression after r230020
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69297 --- Comment #3 from Richard Biener --- With a fix: t.c:76:10: note: Cost model analysis: Vector inside of basic block cost: 376 Vector prologue cost: 0 Vector epilogue cost: 0 Scalar cost of basic block: 96 t.c:76:10: note: not vectorized: vectorization is not profitable. Note the reduction loop is still vectorized: t.c:74:5: note: Cost model analysis: Vector inside of loop cost: 3 Vector prologue cost: 1 Vector epilogue cost: 7 Scalar iteration cost: 3 Scalar outside cost: 0 Vector outside cost: 8 prologue iterations: 0 epilogue iterations: 0 Calculated minimum iters for profitability: 4 but likely this isn't profitable either?
[Bug lto/69337] [6 Regression] Internal compiler error fortran c lto
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69337 Richard Biener changed: What|Removed |Added Keywords||error-recovery, ||ice-on-invalid-code Status|UNCONFIRMED |ASSIGNED Last reconfirmed||2016-01-18 Assignee|unassigned at gcc dot gnu.org |rguenth at gcc dot gnu.org Target Milestone|--- |6.0 Summary|Internal compiler error |[6 Regression] Internal |fortran c lto |compiler error fortran c ||lto Ever confirmed|0 |1 --- Comment #1 from Richard Biener --- Confirmed. GCC 5 produces > gcc-5 t.f90 t.c -flto t.c:1:6: error: variable ‘j’ redeclared as function void j_ (void); ^ t.f90:5:0: note: previously declared here COMMON /j/ j_q( 14550) ^ lto1: fatal error: errors during merging of translation units compilation terminated. lto-wrapper: fatal error: gcc-5 returned 1 exit status compilation terminated. /usr/lib64/gcc/x86_64-suse-linux/5/../../../../x86_64-suse-linux/bin/ld: lto-wrapper failed collect2: error: ld returned 1 exit status Mine.
[Bug target/69307] [6 Regression] wrong code with -O2 -fselective-scheduling @ armv7a
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69307 ktkachov at gcc dot gnu.org changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed||2016-01-18 CC||ktkachov at gcc dot gnu.org Ever confirmed|0 |1 --- Comment #2 from ktkachov at gcc dot gnu.org --- Confirmed on trunk. Btw, your toolchain is not configured to generate armv7-a code by default (you need to configure it with an appropriate --with-arch or --with-cpu). This bug doesn't trigger for armv7-a, but it does for -march=armv6j or lower.
[Bug tree-optimization/69170] [6 Regression] ICE (segfault) in find_uses_to_rename_use
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69170 Richard Biener changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|--- |FIXED --- Comment #6 from Richard Biener --- Fixed.
[Bug tree-optimization/69170] [6 Regression] ICE (segfault) in find_uses_to_rename_use
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69170 --- Comment #5 from Richard Biener --- Author: rguenth Date: Mon Jan 18 09:14:14 2016 New Revision: 232496 URL: https://gcc.gnu.org/viewcvs?rev=232496=gcc=rev Log: 2016-01-18 Richard BienerPR tree-optimization/69170 * tree-vect-slp.c (vect_build_slp_tree): Verify we are not building a vector from scalar results of a pattern stmt. * gcc.dg/torture/pr69170.c: New testcase. Added: trunk/gcc/testsuite/gcc.dg/torture/pr69170.c Modified: trunk/gcc/ChangeLog trunk/gcc/testsuite/ChangeLog trunk/gcc/tree-vect-slp.c
[Bug tree-optimization/66797] [6 Regression] FAIL: gcc.dg/tree-ssa/pr65447.c scan-tree-dump-not ivopts "\\nuse 5\\n"
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66797 --- Comment #3 from amker at gcc dot gnu.org --- Author: amker Date: Mon Jan 18 09:30:10 2016 New Revision: 232497 URL: https://gcc.gnu.org/viewcvs?rev=232497=gcc=rev Log: PR tree-optimization/66797 * gcc.c-torture/execute/pr65447.c: Relax check condition. Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/gcc.dg/tree-ssa/pr65447.c
[Bug tree-optimization/69308] ifcombine joins together floating point expression with side effects
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69308 Richard Biener changed: What|Removed |Added Keywords||wrong-code Status|NEW |ASSIGNED Assignee|unassigned at gcc dot gnu.org |rguenth at gcc dot gnu.org --- Comment #11 from Richard Biener --- bb_no_side_effects_p should detect this via its call of gimple_could_trap_p. Which doesn't handle trapping GIMPLE_CONDs ...
[Bug libstdc++/69259] std::experimental::filesystem::copy does not create symlinks for directories
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69259 --- Comment #4 from Roman Puls--- Hi Jonathan, well, I agree that the standard does not cover thar particular combination of input parameters. But: is falling through the function without the error_code set what we expect? I was assuming that copy() will fail by throwing an exception (or returning an ec in the overloaded version). So yes, closing the ticket seems ok, since the test does not violate standardized behaviour, but feedback shall be given to the standardization group. Could you bring in that topic eventually, or what do you think? Thanks and regards, Roman
[Bug tree-optimization/69336] Constant value not detected
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69336 Richard Biener changed: What|Removed |Added CC||alan.lawrence at arm dot com --- Comment #2 from Richard Biener --- Yes, I think there is a duplicate (and a proposed patch).
[Bug libstdc++/69340] [6 Regression] Incorrect use of standard try catch in libstdc++-v3/src/c++11/cow-stdexcept.cc
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69340 Richard Biener changed: What|Removed |Added Priority|P3 |P1 Status|UNCONFIRMED |NEW Last reconfirmed||2016-01-18 Target Milestone|--- |6.0 Summary|Incorrect use of standard |[6 Regression] Incorrect |try catch in|use of standard try catch |libstdc++-v3/src/c++11/cow- |in |stdexcept.cc|libstdc++-v3/src/c++11/cow- ||stdexcept.cc Ever confirmed|0 |1 --- Comment #1 from Richard Biener --- Confirmed.
[Bug tree-optimization/69297] [6 Regression] Performance regression after r230020
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69297 --- Comment #4 from Yuri Rumyantsev --- Yes, this loop was added for avoiding dce phase. Thanks. Yuri. 2016-01-18 13:33 GMT+03:00 rguenth at gcc dot gnu.org: > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69297 > > --- Comment #3 from Richard Biener --- > With a fix: > > t.c:76:10: note: Cost model analysis: > Vector inside of basic block cost: 376 > Vector prologue cost: 0 > Vector epilogue cost: 0 > Scalar cost of basic block: 96 > t.c:76:10: note: not vectorized: vectorization is not profitable. > > Note the reduction loop is still vectorized: > > t.c:74:5: note: Cost model analysis: > Vector inside of loop cost: 3 > Vector prologue cost: 1 > Vector epilogue cost: 7 > Scalar iteration cost: 3 > Scalar outside cost: 0 > Vector outside cost: 8 > prologue iterations: 0 > epilogue iterations: 0 > Calculated minimum iters for profitability: 4 > > but likely this isn't profitable either? > > -- > You are receiving this mail because: > You reported the bug.
[Bug tree-optimization/69341] [6 Regression] [graphite] ICE: verify_ssa failed (error: definition in block 37 does not dominate use in block 30)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69341 Richard Biener changed: What|Removed |Added Target Milestone|--- |6.0
[Bug target/69311] [5 Regression] ICE (cc1 killed) on s390x-linux-gnu
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69311 Richard Biener changed: What|Removed |Added Version|6.0 |5.3.1 Target Milestone|6.0 |5.4
[Bug ipa/68148] Devirtualization only applies to last of multiple successive calls
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68148 James Greenhalgh changed: What|Removed |Added CC||jgreenhalgh at gcc dot gnu.org --- Comment #6 from James Greenhalgh --- This fix also fixed an ICE on ARM targets: bug.cpp:23:1: internal compiler error: in get_untransformed_body, at cgraph.c:3311 } ^ 0x8dc245 cgraph_node::get_untransformed_body() .../gcc/cgraph.c:3311 0x8e79bd cgraph_node::expand() .../gcc/cgraphunit.c:1941 0x8e9413 expand_all_functions .../gcc/cgraphunit.c:2107 0x8e9413 symbol_table::compile() ...gcc/cgraphunit.c:2463 0x8eb56f symbol_table::finalize_compilation_unit() .../gcc/cgraphunit.c:2553 The testcase for that looked like: class B { public: virtual void a (int *x) = 0; virtual void b (int *x) = 0; }; class Mouseintdapter : public virtual B { void a (int *x) {} void b (int *x) {} }; B* collection; void foo (int *x) { collection->a(x); } void bar (int *x) { collection->b(x); } And would fail at -O2 for most ARM target flag combinations I tried.
[Bug ipa/62051] [4.9/5/6 Regression] Undefined reference to vtable with -O2 and -fdevirtualize-speculatively
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62051 --- Comment #13 from Jan Hubicka --- Created attachment 37388 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=37388=edit patch Hi, this patch adds the logic to gimple-fold.c which makes the offending dtor non-refeable. It is bit uglier than I hoped for. The reason is that the dtor itself is COMDAT so it seems perfectly OK to refer it since we can provide the body, but the body itself refers to vtable that is not output in the unit. The patch simply prohibits references to all COMDAT and EXTERN methods and vtables of types with visibility attributes which will prevent optimizing of many inlines i.e. in libstdc++. It still seem to me that the the testcase is sort of invalid because it provide inline body for the dtor. Can we think of better solution (perhaps validating comdats before using them and checking if symbols they refer to are referable)?
[Bug target/67895] Wrong assembly: incorrect rounding/SAE specifier position
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=67895 --- Comment #3 from afomin at gcc dot gnu.org --- Author: afomin Date: Mon Jan 18 17:09:06 2016 New Revision: 232533 URL: https://gcc.gnu.org/viewcvs?rev=232533=gcc=rev Log: Backport to mainline 2015-10-09 Alexander Fomingcc/ PR target/67895 * config/i386/sse.md (define_insn "sse_cvtsi2ss"): Adjust embedded rounding/SAE specifier position. (define_insn "sse_cvtsi2ssq"): Likewise. (define_insn "cvtusi232"): Likewise. (define_insn "cvtusi264"): Likewise. (define_insn "sse2_cvtsi2sdq"): Likewise. (define_insn "avx512dq_rangep"): Likewise. (define_insn "avx512dq_ranges"): Likewise. gcc/testsuite PR target/67895 * gcc.target/i386/avx512dq-vrangepd-1.c: Adjust. * gcc.target/i386/avx512dq-vrangeps-1.c: Likewise. * gcc.target/i386/avx512dq-vrangesd-1.c: Likewise. * gcc.target/i386/avx512dq-vrangess-1.c: Likewise. * gcc.target/i386/avx512f-vcvtsi2sd64-1.c: Likewise. * gcc.target/i386/avx512f-vcvtsi2ss-1.c: Likewise. * gcc.target/i386/avx512f-vcvtsi2ss64-1.c: Likewise. * gcc.target/i386/avx512f-vcvtusi2sd64-1.c: Likewise. * gcc.target/i386/avx512f-vcvtusi2ss-1.c: Likewise. * gcc.target/i386/avx512f-vcvtusi2ss64-1.c: Likewise. Modified: branches/gcc-5-branch/gcc/ChangeLog branches/gcc-5-branch/gcc/config/i386/sse.md branches/gcc-5-branch/gcc/testsuite/ChangeLog branches/gcc-5-branch/gcc/testsuite/gcc.target/i386/avx512dq-vrangepd-1.c branches/gcc-5-branch/gcc/testsuite/gcc.target/i386/avx512dq-vrangeps-1.c branches/gcc-5-branch/gcc/testsuite/gcc.target/i386/avx512dq-vrangesd-1.c branches/gcc-5-branch/gcc/testsuite/gcc.target/i386/avx512dq-vrangess-1.c branches/gcc-5-branch/gcc/testsuite/gcc.target/i386/avx512f-vcvtsi2sd64-1.c branches/gcc-5-branch/gcc/testsuite/gcc.target/i386/avx512f-vcvtsi2ss-1.c branches/gcc-5-branch/gcc/testsuite/gcc.target/i386/avx512f-vcvtsi2ss64-1.c branches/gcc-5-branch/gcc/testsuite/gcc.target/i386/avx512f-vcvtusi2sd64-1.c branches/gcc-5-branch/gcc/testsuite/gcc.target/i386/avx512f-vcvtusi2ss-1.c branches/gcc-5-branch/gcc/testsuite/gcc.target/i386/avx512f-vcvtusi2ss64-1.c
[Bug libstdc++/60637] --fast-math breaks std::signbit function
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60637 --- Comment #11 from Jonathan Wakely --- Author: redi Date: Mon Jan 18 17:15:42 2016 New Revision: 232534 URL: https://gcc.gnu.org/viewcvs?rev=232534=gcc=rev Log: Add test for PR 60637 PR libstdc++/60637 * testsuite/26_numerics/headers/cmath/60637.cc: Add test. Added: trunk/libstdc++-v3/testsuite/26_numerics/headers/cmath/60637.cc Modified: trunk/libstdc++-v3/ChangeLog
[Bug target/69345] [6 Regression] 459.GemsFDTD regression
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69345 --- Comment #3 from Alexander Fomin --- Looks like it's r232401.
[Bug testsuite/69181] multiline.exp does not handle conditional compilation
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69181 --- Comment #4 from David Malcolm --- Author: dmalcolm Date: Mon Jan 18 17:26:58 2016 New Revision: 232535 URL: https://gcc.gnu.org/viewcvs?rev=232535=gcc=rev Log: PR testsuite/69181: ensure expected multiline outputs is cleared per-test gcc/testsuite/ChangeLog: PR testsuite/69181 * gcc.dg/pr69181-1.c: New test file. * gcc.dg/pr69181-2.c: New test file. * lib/gcc-dg.exp (dg-test): Consolidate post-test cleanup of globals by moving it to... (cleanup-after-saved-dg-test): ...this new function. Add "global additional_sources_used". Add reset of global multiline_expected_outputs to the empty list. * lib/multiline.exp (_multiline_expected_outputs): Rename this global to... (multiline_expected_outputs): ...this, and updated comments to note that it is modified from gcc-dg.exp. (dg-end-multiline-output): Update for the above renaming. (handle-multiline-outputs): Likewise. Remove the clearing of the expected outputs to the empty list. Added: trunk/gcc/testsuite/gcc.dg/pr69181-1.c trunk/gcc/testsuite/gcc.dg/pr69181-2.c Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/lib/gcc-dg.exp trunk/gcc/testsuite/lib/multiline.exp
[Bug middle-end/66877] [6 Regression] FAIL: gcc.dg/vect/vect-over-widen-3-big-array.c -flto -ffat-lto-objects scan-tree-dump-times vect "vect_recog_over_widening_pattern: detected" 2
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66877 Nick Clifton changed: What|Removed |Added CC||nickc at gcc dot gnu.org --- Comment #3 from Nick Clifton --- Created attachment 37387 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=37387=edit Vectorizer dump Hi Richard, Here is a dump from a failing testcase. This was from a toolchain configured as --target=arm-eabi built using today's FSF mainline sources. Cheers Nick
[Bug testsuite/69181] multiline.exp does not handle conditional compilation
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69181 David Malcolm changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|--- |FIXED --- Comment #5 from David Malcolm --- Should be fixed as of r232535; marking as resolved.
[Bug lto/61886] [4.9/5/6 Regression] LTO breaks fread with _FORTIFY_SOURCE=2
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61886 --- Comment #57 from Jan Hubicka --- The alias-2.c should be now fixed on targets with anchors.
[Bug lto/45375] [meta-bug] Issues with building Mozilla with LTO
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=45375 --- Comment #219 from Jan Hubicka --- devirtualization issue is now fixed, so we are down to -fno-lifetime-dse.
[Bug lto/69136] [6 Regression] ICE in lto_symtab_prevailing_virtual_decl, at lto/lto-symtab.c:991
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69136 --- Comment #3 from Jan Hubicka --- Hmm, we have: > QI size unit size align 8 symtab 0 alias set -1 structural equality method basetype arg-types chain chain nothrow public private abstract external virtual QI file /aux/hubicka/skia.ii line 4 col 1 align 16 context > (gdb) p decl->decl_with_vis.assembler_name $2 = (tree) 0x0 so it seems that the decl was not seen by free_lang_data but still leaked into the stream.
[Bug lto/69136] [6 Regression] ICE in lto_symtab_prevailing_virtual_decl, at lto/lto-symtab.c:991
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69136 --- Comment #4 from Jan Hubicka --- Aha, it is abstract decl. I am testing: Index: ../../gcc/lto/lto-symtab.c === --- ../../gcc/lto/lto-symtab.c (revision 232466) +++ ../../gcc/lto/lto-symtab.c (working copy) @@ -987,6 +1008,8 @@ lto_symtab_merge_symbols (void) tree lto_symtab_prevailing_virtual_decl (tree decl) { + if (DECL_ABSTRACT_P (decl)) +return decl; gcc_checking_assert (!type_in_anonymous_namespace_p (DECL_CONTEXT (decl)) && DECL_ASSEMBLER_NAME_SET_P (decl));
[Bug lto/69133] [6 Regression] LTO segfault in lto_get_decl_name_mapping() on 483.xalancbmk with -flto-partition=none
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69133 --- Comment #5 from Jan Hubicka --- The problem seems to be that cgraph_node::get_untransformed_body checks presence of body by DECL_RESULT which is NULL for thunks. Rest of places seems to check for DECL_ARGUMENTS. I am testing: Index: cgraph.c === --- cgraph.c(revision 232466) +++ cgraph.c(working copy) @@ -3305,10 +3295,11 @@ cgraph_node::get_untransformed_body (voi size_t len; tree decl = this->decl; - if (DECL_RESULT (decl)) + /* Check if body is already there. */ + if (!DECL_ARGUMENTS (decl)) return false; - gcc_assert (in_lto_p); + gcc_assert (in_lto_p && !DECL_RESULT (decl)) timevar_push (TV_IPA_LTO_GIMPLE_IN);
[Bug lto/69133] [6 Regression] LTO segfault in lto_get_decl_name_mapping() on 483.xalancbmk with -flto-partition=none
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69133 --- Comment #6 from Jan Hubicka --- The following has even chance to work :) Index: cgraph.c === --- cgraph.c(revision 232466) +++ cgraph.c(working copy) @@ -3305,10 +3295,12 @@ cgraph_node::get_untransformed_body (voi size_t len; tree decl = this->decl; - if (DECL_RESULT (decl)) + /* Check if body is already there. Either we have gimple body or + the function is thunk and in that case we set DECL_ARGUMENTS. */ + if (DECL_ARGUMENTS (decl) || gimple_has_body_p (decl)) return false; - gcc_assert (in_lto_p); + gcc_assert (in_lto_p && !DECL_RESULT (decl)); timevar_push (TV_IPA_LTO_GIMPLE_IN);
[Bug c++/69338] incorrect ctor initialization of a flexible array member
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69338 --- Comment #1 from Martin Sebor --- Another test case (derived from the test case in bug 69253), this one even shows a -Warray-bounds warning with -O2. With 69253 fixed (and the test case still accepted), 6.0 also aborts but doesn't print the -Warray-bounds warning. $ ~/bin/gcc-5.1.0/bin/g++ -O2 -Wall -Wextra -Wpedantic -fpermissive z.c && ./a.out z.c:1:22: warning: ISO C++ forbids zero-size array ‘a’ [-Wpedantic] struct A { char n, a[]; }; ^ z.c: In function ‘int main()’: z.c:11:37: warning: ISO C++ forbids compound-literals [-Wpedantic] int a0 = ((struct A){ 1, "" }).a[0]; ^ z.c:11:37: warning: initializer-string for array of chars is too long [-fpermissive] z.c:11:43: warning: array subscript is above array bounds [-Warray-bounds] int a0 = ((struct A){ 1, "" }).a[0]; ^ 12345678 Aborted (core dumped) $ /home/msebor/build/gcc-trunk-svn/gcc/xgcc -B/home/msebor/build/gcc-trunk-svn/gcc -Wall -Wextra -Wpedantic -xc++ -O2 z.c && ./a.out z.c: In function ‘int main()’: z.c:11:37: warning: ISO C++ forbids compound-literals [-Wpedantic] int a0 = ((struct A){ 1, "" }).a[0]; ^ z.c:11:37: warning: initialization of a flexible array member [-Wpedantic] 12345678 Aborted (core dumped)
[Bug lto/68820] [6 regression] FAIL: gcc.c-torture/execute/builtins/memops-asm.c execution, -O2 -flto -fuse-linker-plugin -fno-fat-lto-objects
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68820 Jan Hubicka changed: What|Removed |Added Assignee|hubicka at gcc dot gnu.org |unassigned at gcc dot gnu.org --- Comment #10 from Jan Hubicka --- OK, this reproduces for me. We get: Merging nodes for my_memcpy. Candidates: my_memcpy/4 (my_memcpy) @0x76cc7450 Type: function definition analyzed Visibility: force_output externally_visible public next sharing asm name: 36 References: Referring: Read from file: memops-asm-lib.o First run: 0 Function flags: Called by: Calls: *my_memcpy/36 (memcpy) @0x76cc9a10 Type: function Visibility: external public next sharing asm name: 35 previous sharing asm name: 4 References: Referring: Read from file: memops-asm.o First run: 0 Function flags: Called by: main_test/31 (1.00 per call) Calls: *my_memcpy/35 (__builtin_memcpy) @0x76cc9b80 Type: function Visibility: external public previous sharing asm name: 36 References: Referring: Read from file: memops-asm.o First run: 0 Function flags: Called by: main_test/31 (1.00 per call) Calls: Not merging decls; DECL_BUILT_IN mismatch Not merging decls; DECL_BUILT_IN mismatch which is expected, one memcpy and __builtin_memcpy are builtins while my_memcpy is not. Now at ltrans time we have: memcpy/8 (memcpy) @0x76ccb450 Type: function definition analyzed Visibility: externally_visible public References: inside_main/0 (read) Referring: Read from file: oo.ltrans0.o First run: 0 Function flags: Called by: Calls: abort/13 i.e. unused memcpy and my_memcpy/4 (my_memcpy) @0x76cc1e60 Type: function definition analyzed Visibility: force_output externally_visible public References: Referring: *my_memcpy/36 (alias) Read from file: oo.ltrans0.o First run: 0 Function flags: nonfreeing_fn Called by: Calls: *my_memcpy/36 (memcpy) @0x76cc1cf0 Type: function definition analyzed alias transparent_alias Visibility: externally_visible external public References: my_memcpy/4 (alias) Referring: Read from file: oo.ltrans0.o First run: 0 Function flags: Called by: main_test/31 (1.00 per call) main_test/31 (1.00 per call) Calls: i.e. used my_memcpy Then we get to: (gdb) bt #0 expand_builtin_memcpy_args (dest=0x76923fe0, src=0x76923ca0, len=0x76918b88, target=target@entry=0x769391e0, exp=exp@entry=0x7692eab0) at ../../gcc/builtins.c:2904 #1 0x005db286 in expand_builtin_memcpy (target=0x769391e0, exp=0x7692eab0) at ../../gcc/builtins.c:2991 #2 expand_builtin (exp=0x7692eab0, target=0x769391e0, subtarget=0x0, mode=DImode, ignore=0) at ../../gcc/builtins.c:5963 #3 0x006c7bf4 in expand_expr_real_1 (exp=0x7692eab0, target=, tmode=DImode, modifier=EXPAND_NORMAL, alt_rtl=, inner_reference_p=inner_reference_p@entry=false) at ../../gcc/expr.c:10563 #4 0x006c94a9 in expand_expr_real (exp=exp@entry=0x7692eab0, target=, tmode=, modifier=modifier@entry=EXPAND_NORMAL, alt_rtl=alt_rtl@entry=0x7fffe2b0, inner_reference_p=inner_reference_p@entry=false) at ../../gcc/expr.c:7947 #5 0x006d070a in store_expr_with_bounds (exp=exp@entry=0x7692eab0, target=target@entry=0x769391e0, call_param_p=call_param_p@entry=0, nontemporal=nontemporal@entry=false, reverse=reverse@entry=false, btarget=btarget@entry=0x76918bd0) at
[Bug lto/68820] [6 regression] FAIL: gcc.c-torture/execute/builtins/memops-asm.c execution, -O2 -flto -fuse-linker-plugin -fno-fat-lto-objects
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68820 --- Comment #11 from Jan Hubicka --- ... and in case anyone is curious, with -fuse-linker-plugin we inline the memcpys as we work out the block size. This is why they pass.
[Bug sanitizer/68824] [6 Regression] libtsan is missing the __interceptor___tls_get_addr symbol without bumping the soname
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68824 --- Comment #5 from Dmitry Vyukov --- In 3) did you mean -mstackrealign? 1) looks like the simplest option. Are there any downsides?
[Bug sanitizer/68824] [6 Regression] libtsan is missing the __interceptor___tls_get_addr symbol without bumping the soname
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68824 --- Comment #7 from Dmitry Vyukov --- I will land such fix in clang. Thanks!
[Bug libstdc++/69293] scoped_allocator_adaptor::construct calls wrong function
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69293 --- Comment #6 from Jonathan Wakely --- Author: redi Date: Mon Jan 18 11:43:37 2016 New Revision: 232504 URL: https://gcc.gnu.org/viewcvs?rev=232504=gcc=rev Log: Fix construction of std::function from null pointer-to-member PR libstdc++/69293 * include/std/functional (_Function_base::_M_not_empty_function): Change overloads for pointers to take arguments by value. * testsuite/20_util/function/cons/57465.cc: Add tests for pointer-to-member cases. Modified: trunk/libstdc++-v3/ChangeLog trunk/libstdc++-v3/include/std/functional trunk/libstdc++-v3/testsuite/20_util/function/cons/57465.cc
[Bug tree-optimization/69328] [6 Regression] ice in vect_get_vec_def_for_operand, at tree-vect-stmts.c:1379 with -O3
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69328 Richard Biener changed: What|Removed |Added Priority|P3 |P1 --- Comment #3 from Richard Biener --- 1378 is_simple_use = vect_is_simple_use (op, loop_vinfo, _stmt, ); 1379 gcc_assert (is_simple_use); (gdb) p op $2 = ehm. if (masked) { vec_cond_lhs = vect_get_vec_def_for_operand (cond_expr, stmt, comp_vectype); vect_is_simple_use (cond_expr, stmt_info->vinfo, , [0]); } else { vec_cond_lhs = vect_get_vec_def_for_operand (TREE_OPERAND (cond_expr, 0), stmt, comp_vectype); vect_is_simple_use (TREE_OPERAND (cond_expr, 0), loop_vinfo, , [0]); vec_cond_rhs = vect_get_vec_def_for_operand (TREE_OPERAND (cond_expr, 1), stmt, comp_vectype); vect_is_simple_use (TREE_OPERAND (cond_expr, 1), loop_vinfo, , [1]); } Probably better test ! COMPARISON_CLASS_P (cond_expr) here. OTOH we have if (VECTOR_BOOLEAN_TYPE_P (comp_vectype)) { vec_cmp_type = comp_vectype; masked = true; } else vec_cmp_type = build_same_sized_truth_vector_type (comp_vectype); not sure why that's specially only for boolean vector types. So @@ -7544,11 +7545,9 @@ vectorizable_condition (gimple *stmt, gi if (!vect_is_simple_use (else_clause, stmt_info->vinfo, _stmt, )) return false; + bool masked = ! COMPARISON_CLASS_P (cond_expr); if (VECTOR_BOOLEAN_TYPE_P (comp_vectype)) -{ - vec_cmp_type = comp_vectype; - masked = true; -} +vec_cmp_type = comp_vectype; else vec_cmp_type = build_same_sized_truth_vector_type (comp_vectype); if (vec_cmp_type == NULL_TREE) but that later ICEs then with > ./cc1 -quiet t.c -O3 t.c: In function ‘fn1’: t.c:2:6: internal compiler error: in vector_compare_rtx, at optabs.c:5290 void fn1() { ^~~ 0xca84f2 vector_compare_rtx /space/rguenther/src/svn/trunk3/gcc/optabs.c:5290 0xca979f expand_vec_cond_expr(tree_node*, tree_node*, tree_node*, tree_node*, rtx_def*) /space/rguenther/src/svn/trunk3/gcc/optabs.c:5612 sth else is wrong here.
[Bug target/69305] [5/6 Regression] wrong code with -O and int128 @ aarch64
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69305 --- Comment #3 from Jakub Jelinek --- Looks like aarch64 backend bug to me. unsigned __int128 f1 (unsigned __int128 x, unsigned __int128 y) { return x + y; } unsigned __int128 f2 (unsigned __int128 x, unsigned __int128 y) { return x - y; } unsigned __int128 f3 (unsigned __int128 x, unsigned long long y) { return x + y; } unsigned __int128 f4 (unsigned __int128 x, unsigned long long y) { return x - y; } unsigned __int128 f5 (unsigned __int128 x, unsigned long long y) { return x + (unsigned long long) -y; } This case is f5 from the above, where the value should be negated first in DImode and then added to the TImode value. But combiner sees (insn 7 4 10 2 (set (reg:DI 79) (neg:DI (reg/v:DI 77 [ y ]))) pr69305-1.c:5 331 {negdi2} (expr_list:REG_DEAD (reg/v:DI 77 [ y ]) (nil))) (insn 10 7 11 2 (parallel [ (set (reg:CC_NZ 66 cc) (compare:CC_NZ (plus:DI (reg:DI 79) (reg:DI 85 [ x ])) (const_int 0 [0]))) (set (reg:DI 81) (plus:DI (reg:DI 79) (reg:DI 85 [ x ]))) ]) pr69305-1.c:5 100 {adddi3_compare0} (expr_list:REG_DEAD (reg:DI 85 [ x ]) (expr_list:REG_DEAD (reg:DI 79) (nil (insn 11 10 25 2 (set (reg:DI 82) (plus:DI (geu:DI (reg:CC 66 cc) (const_int 0 [0])) (reg:DI 86 [ x+8 ]))) pr69305-1.c:5 452 {*csinc2di_insn} (expr_list:REG_DEAD (reg:DI 86 [ x+8 ]) (expr_list:REG_DEAD (reg:CC 66 cc) (nil and attempts to merge insn 7 into insn 10 by having a minus in there instead, and that pattern is then recognized as subdi3_compare0. But probably csinc2di_insn actually relies on the preceeding insn being result of adddi3_compare instead of subdi3_compare. Seems on arm32 or x86_64 we have a DImode (arm32) or TImode pattern for the addition that is only lowered later on, so this is something that really can't happen on those targets.
[Bug target/69307] [6 Regression] wrong code with -O2 -fselective-scheduling @ armv7a
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69307 --- Comment #3 from ktkachov at gcc dot gnu.org --- Although, the trigger is the scheduling model used For example, -mtune=arm1136j-s shows the wrong code with any -march option
[Bug libstdc++/69295] [6 Regression] New special math function failures
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69295 --- Comment #6 from Jonathan Wakely --- It's arguable whether this should be marked as a regression, as it's new code and new tests, so the tests wouldn't even compile before gcc6.
[Bug rtl-optimization/69307] [6 Regression] wrong code with -O2 -fselective-scheduling @ armv7a
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69307 ktkachov at gcc dot gnu.org changed: What|Removed |Added Component|target |rtl-optimization Known to fail|4.9.3 |4.8.5, 4.9.4, 5.3.1 --- Comment #4 from ktkachov at gcc dot gnu.org --- Fails at least as far back as 4.8 btw and looks like a sel-sched bug to me.