[Bug fortran/69524] [6 Regression] [F08] Compiler segfaults on "module subroutine"
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69524 Jeffrey A. Law changed: What|Removed |Added Priority|P3 |P4 CC||law at redhat dot com
[Bug fortran/69603] [5/6 Regression] ICE: segfault with -fimplicit-none and proc_ptr_comp_24.f90
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69603 Jeffrey A. Law changed: What|Removed |Added Priority|P3 |P4 CC||law at redhat dot com
[Bug ipa/70188] [4.9/5/6 Regression] gcc 4.9+ miscompiles code on hppa
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70188 Jeffrey A. Law changed: What|Removed |Added CC||law at redhat dot com --- Comment #7 from Jeffrey A. Law --- Unlikely related to 69532 -- the PA port is consistent WRT PROMOTE_MODE and friends IIRC. I'm leaving as P3 until we have this better analyzed though.
[Bug tree-optimization/69728] [6 Regression] internal compiler error: in outer_projection_mupa, at graphite-sese-to-poly.c:1175
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69728 Jeffrey A. Law changed: What|Removed |Added Priority|P3 |P4 CC||law at redhat dot com
[Bug middle-end/70199] [5/6 Regression] Crash at -O2 when using labels.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70199 Jeffrey A. Law changed: What|Removed |Added Priority|P3 |P1 CC||law at redhat dot com --- Comment #4 from Jeffrey A. Law --- Looks like reload writes offsets_known_at for a label on the forced_labels_list. The problem is the label # of the offending label is 0, but FIRST_LABEL_NUM is 4. This causes the out-of-bounds array accesses and ultimately heap corruption.
[Bug middle-end/70219] [6 Regression] ICE: in delete_move_and_clobber, at lra-constraints.c:5864 with -O2
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70219 Jakub Jelinek changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed||2016-03-14 CC||jakub at gcc dot gnu.org Component|rtl-optimization|middle-end Ever confirmed|0 |1
[Bug middle-end/70219] [6 Regression] ICE: in delete_move_and_clobber, at lra-constraints.c:5864 with -O2
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70219 Jakub Jelinek changed: What|Removed |Added Status|NEW |ASSIGNED Assignee|unassigned at gcc dot gnu.org |jakub at gcc dot gnu.org --- Comment #3 from Jakub Jelinek --- Created attachment 37953 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=37953&action=edit gcc6-pr70219.patch Untested fix. The code had assertion dregno > 0, but I don't see anything special on register 0 (%rax on x86_64), assuming Vlad meant >= 0 here (i.e. make sure dregno is not negative, which is not a hard register, but has special meaning).
[Bug middle-end/70219] [6 Regression] ICE: in delete_move_and_clobber, at lra-constraints.c:5864 with -O2
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70219 Jakub Jelinek changed: What|Removed |Added Priority|P3 |P1
[Bug c++/70218] [5/6 Regression] Illegal access to private fields succeeds with 5/6
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70218 Jakub Jelinek changed: What|Removed |Added Priority|P3 |P2 Status|UNCONFIRMED |NEW Last reconfirmed||2016-03-14 CC||jakub at gcc dot gnu.org, ||jason at gcc dot gnu.org Ever confirmed|0 |1 --- Comment #2 from Jakub Jelinek --- Started with r216260 fix.
[Bug c++/70204] [6 Regression] ICE on x86_64-linux-gnu in non_const_var_error, at cp/constexpr.c:2764
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70204 Jakub Jelinek changed: What|Removed |Added Priority|P3 |P4 CC||jakub at gcc dot gnu.org, ||jason at gcc dot gnu.org --- Comment #2 from Jakub Jelinek --- Started with r231197.
[Bug c++/68722] [4.9/5/6 Regression] internal compiler error: in merge_exception_specifiers, at cp/typeck2.c:2108
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68722 Jakub Jelinek changed: What|Removed |Added Priority|P3 |P4 Status|UNCONFIRMED |NEW Last reconfirmed||2016-03-14 CC||jakub at gcc dot gnu.org, ||paolo at gcc dot gnu.org Target Milestone|--- |4.9.4 Summary|internal compiler error: in |[4.9/5/6 Regression] |merge_exception_specifiers, |internal compiler error: in |at cp/typeck2.c:2108|merge_exception_specifiers, ||at cp/typeck2.c:2108 Ever confirmed|0 |1 --- Comment #1 from Jakub Jelinek --- Started with r204818.
[Bug rtl-optimization/38644] [4.6 Regression] Optimization flag -O1 -fschedule-insns2 causes wrong code
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=38644
Satish M changed:
What|Removed |Added
CC||satishbmsce at gmail dot com
--- Comment #72 from Satish M ---
This bug still exists in GCC 4.8.2 ARM. It can reproduced by adding one more
argument in 'doStreamReadBlock' function in test case.
/x86_64-unknown-linux-gnu/bin/gcc -B /x86_64-unknown-linux-gnu/bin -msoft-float
-marm -mcpu=cortex-a9 -march=armv7-a -mno-thumb-interwork -mlong-calls
-mno-unaligned-access -O2 test.c
extern int doStreamReadBlock (int *, char *, int size, int, int);
char readStream (int *s)
{
char c = 0;
doStreamReadBlock (s, &c, 1, *s, 22);
return c;
}
:
0: e1a0c00dmov ip, sp
4: e3a02000mov r2, #0
8: e92dd800push{fp, ip, lr, pc}
c: e24cb004sub fp, ip, #4
10: e24dd008sub sp, sp, #8
14: e24b100csub r1, fp, #12
18: e3a0c016mov ip, #22
1c: e5612001strbr2, [r1, #-1]!
20: e3a02001mov r2, #1
24: e5903000ldr r3, [r0]
28: e58dc000str ip, [sp]
2c: e59fc00cldr ip, [pc, #12] ; 40
30: e12fff3cblx ip
34: e24bd00csub sp, fp, #12< Stack frame
de-allocated
38: e55b000dldrbr0, [fp, #-13] < Accessing stack.
3c: e89da800ldm sp, {fp, sp, pc}
40: andeq r0, r0, r0
[Bug c++/68475] [4.9/5/6 Regression] ICE: in merge_exception_specifiers, at cp/typeck2.c:2115 with -fno-exceptions on invalid code
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68475 Jakub Jelinek changed: What|Removed |Added Keywords||ice-checking CC||jakub at gcc dot gnu.org, ||jason at gcc dot gnu.org Target Milestone|--- |4.9.4 Summary|ICE: in |[4.9/5/6 Regression] ICE: |merge_exception_specifiers, |in |at cp/typeck2.c:2115 with |merge_exception_specifiers, |-fno-exceptions on invalid |at cp/typeck2.c:2115 with |code|-fno-exceptions on invalid ||code --- Comment #1 from Jakub Jelinek --- Started with r175073, the ICE is in gcc_checking_assert that has been added in that revision.
[Bug tree-optimization/70171] Poor code generated when return struct using ternary operator
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70171 --- Comment #8 from Richard Biener --- (In reply to Oleg Endo from comment #7) > Is this a dup of PR 56365? Sounds like so ... No, not really.
[Bug c++/68723] [4.9/5/6 Regression] ice in pop_nested_namespace, at cp/name-lookup.c:3816
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68723 Jakub Jelinek changed: What|Removed |Added CC||jakub at gcc dot gnu.org, ||jason at gcc dot gnu.org Target Milestone|--- |4.9.4 Summary|ice in |[4.9/5/6 Regression] ice in |pop_nested_namespace, at|pop_nested_namespace, at |cp/name-lookup.c:3816 |cp/name-lookup.c:3816 --- Comment #1 from Jakub Jelinek --- Started with r158075.
[Bug c++/68724] [4.9/5/6 Regression] ice in unify, at cp/pt.c:19902
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68724 Jakub Jelinek changed: What|Removed |Added CC||froydnj at gcc dot gnu.org, ||jakub at gcc dot gnu.org, ||jason at gcc dot gnu.org Target Milestone|--- |4.9.4 Summary|ice in unify, at|[4.9/5/6 Regression] ice in |cp/pt.c:19902 |unify, at cp/pt.c:19902 --- Comment #1 from Jakub Jelinek --- Started with r166977.
[Bug c++/68728] [6 Regression] ice: unexpected expression ‘(const char*)__s’ of kind implicit_conv_expr
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68728 Jakub Jelinek changed: What|Removed |Added Keywords||error-recovery CC||jakub at gcc dot gnu.org Target Milestone|--- |6.0 Summary|ice: unexpected expression |[6 Regression] ice: |‘(const char*)__s’ of kind |unexpected expression |implicit_conv_expr |‘(const char*)__s’ of kind ||implicit_conv_expr --- Comment #2 from Jakub Jelinek --- Started with r230365.
[Bug tree-optimization/70221] New: graph dump flag used for constraint graph in points-to analysis
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70221
Bug ID: 70221
Summary: graph dump flag used for constraint graph in points-to
analysis
Product: gcc
Version: 6.0
Status: UNCONFIRMED
Severity: trivial
Priority: P3
Component: tree-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: vries at gcc dot gnu.org
Target Milestone: ---
At the documentation for 'fdump-tree-switch' here
(https://gcc.gnu.org/onlinedocs/gcc/Developer-Options.html ) we read:
...
‘graph’
For each of the other indicated dump files (-fdump-rtl-pass), dump a
representation of the control flow graph suitable for viewing with GraphViz to
file.passid.pass.dot. Each function in the file is pretty-printed as a
subgraph, so that GraphViz can render them all in a single plot.
This option currently only works for RTL dumps, and the RTL is always
dumped in slim form.
...
However, also tree-ssa-structalias.c uses TDF_GRAPH, to decide whether to dump
various constraint graphs:
...
$ grep -d skip TDF_GRAPH gcc/* | grep -v ChangeLog
gcc/dumpfile.c: {"graph", TDF_GRAPH},
gcc/dumpfile.c: | TDF_STMTADDR | TDF_GRAPH | TDF_DIAGNOSTIC | TDF_VERBOSE
gcc/dumpfile.h:#define TDF_GRAPH(1 << 13) /* a graph dump is
being emitted */
gcc/passes.c: && (dfi->pflags & TDF_GRAPH) != 0
gcc/passes.c: && (dump_flags & TDF_GRAPH))
gcc/passes.c: && dump_file && (dump_flags & TDF_GRAPH)
gcc/tree-ssa-structalias.c: if (dump_file && (dump_flags & TDF_GRAPH))
gcc/tree-ssa-structalias.c: if (dump_file && (dump_flags & TDF_GRAPH))
gcc/tree-ssa-structalias.c: if (dump_file && (dump_flags & TDF_GRAPH))
...
These constraint graphs are dump into the normal dump file, and can be
extracted with scripts like this:
...
$ cat split.sh
#!/bin/sh
file="test.c.079i.pta2"
cat "$file" | awk -f split.awk
$ cat split.awk
BEGIN {
n=0
outfile=sprintf("file%02d.dot",n)
doprint=0
}
/^\/\/ The constraint graph/{
doprint=1
}
{
if (doprint)
print > outfile
}
/^}/{
n=n+1
outfile=sprintf("file%02d.dot",n)
doprint=0
}
...
It would be good if the graphs were emitted into separate file(s).
And the constraint graph does not match the description of the 'graph' flag
semantics, so it's probably better to have it triggered by something else, say
TDF_DETAILS.
[Bug c++/70222] New: Test miscompiled with -O1
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70222 Bug ID: 70222 Summary: Test miscompiled with -O1 Product: gcc Version: 6.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: c++ Assignee: unassigned at gcc dot gnu.org Reporter: vsevolod.livinskij at frtk dot ru Target Milestone: --- Created attachment 37954 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=37954&action=edit Reproducer. Test case produces incorrect result starting with -O1 optimization level. Test case fails with -march=nehalem, haswell, knl and skylake-avx512 options and with gcc version 4.8.5 (Revision=226194), gcc version 4.9.4 20160311 (prerelease) (Revision=234148) and gcc version 5.3.1 20160311 (Revision=234148). Output: > g++ repr.cpp -o out -O1 -march=haswell; ./out -262143 > g++ repr.cpp -o out -O0 -march=haswell; ./out -1 GCC version: > g++ -v Using built-in specs. COLLECT_GCC=g++ COLLECT_LTO_WRAPPER=/export/users/vlivinsk/gcc-trunk/bin/libexec/gcc/x86_64-pc-linux-gnu/6.0.0/lto-wrapper Target: x86_64-pc-linux-gnu Configured with: /export/users/vlivinsk/gcc-trunk/gcc/configure --with-arch=corei7 --with-cpu=corei7 --enable-clocale=gnu --with-system-zlib --enable-shared --with-demangler-in-ld --enable-cloog-backend=isl --with-fpmath=sse --enable-checking=release --enable-languages=c,c++,lto --with-gmp=/export/users/vlivinsk/gcc-trunk/gmp-6.1.0/bin --with-mpfr=/export/users/vlivinsk/gcc-trunk/mpfr-3.1.3/bin --with-mpc=/export/users/vlivinsk/gcc-trunk/mpc-1.0.3/bin --prefix=/export/users/vlivinsk/gcc-trunk/bin Thread model: posix gcc version 6.0.0 20160314 (experimental) (Revision=234175)
[Bug tree-optimization/70221] graph dump flag used for constraint graph in points-to analysis
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70221 vries at gcc dot gnu.org changed: What|Removed |Added Priority|P3 |P5
[Bug c++/68728] [6 Regression] ice: unexpected expression ‘(const char*)__s’ of kind implicit_conv_expr
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68728 Jakub Jelinek changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED --- Comment #3 from Jakub Jelinek --- Ah, but the error-recovery bug got fixed with r232438.
[Bug tree-optimization/70045] [6 Regression] ICE error: mismatching comparison operand types
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70045 --- Comment #4 from vries at gcc dot gnu.org --- Author: vries Date: Mon Mar 14 09:19:14 2016 New Revision: 234177 URL: https://gcc.gnu.org/viewcvs?rev=234177&root=gcc&view=rev Log: Unshare create_empty_if_region_on_edge argument 2016-03-14 Tom de Vries PR tree-optimization/70045 * graphite-isl-ast-to-gimple.c (graphite_create_new_loop_guard): Unshare create_empty_if_region_on_edge argument. * gcc.dg/graphite/pr70045.c: New test. Added: trunk/gcc/testsuite/gcc.dg/graphite/pr70045.c Modified: trunk/gcc/ChangeLog trunk/gcc/graphite-isl-ast-to-gimple.c trunk/gcc/testsuite/ChangeLog
[Bug tree-optimization/70045] [6 Regression] ICE error: mismatching comparison operand types
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70045 vries at gcc dot gnu.org changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED Assignee|unassigned at gcc dot gnu.org |vries at gcc dot gnu.org --- Comment #5 from vries at gcc dot gnu.org --- patch with test-case committed, marking resolved-fixed.
[Bug fortran/45076] [OOP] gfortran.dg/dynamic_dispatch_6.f03 ICEs with -fprofile-use
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=45076 --- Comment #10 from dominiq at gcc dot gnu.org --- Author: dominiq Date: Mon Mar 14 09:21:32 2016 New Revision: 234178 URL: https://gcc.gnu.org/viewcvs?rev=234178&root=gcc&view=rev Log: 2016-03-13 Dominique d'Humieres PR fortran/45076 gfortran.dg/prof/prof.exp: New script. gfortran.dg/prof/dynamic_dispatch_6.f03: New test. Added: branches/gcc-4_9-branch/gcc/testsuite/gfortran.dg/prof/ branches/gcc-4_9-branch/gcc/testsuite/gfortran.dg/prof/dynamic_dispatch_6.f03 branches/gcc-4_9-branch/gcc/testsuite/gfortran.dg/prof/prof.exp Modified: branches/gcc-4_9-branch/gcc/testsuite/ChangeLog
[Bug libstdc++/69782] [6 Regression] defining min() macro causes thousand of lines of error messages
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69782
Jakub Jelinek changed:
What|Removed |Added
CC||jakub at gcc dot gnu.org
--- Comment #7 from Jakub Jelinek ---
So is the plan to uglify all
find libstdc++-v3/include -type f | xargs grep '[^a-zA-Z_]\(min\|max\)(' | wc
-l
461
with ()s around min/max or whatever id-expression it appears at the end of?
For 6.x or just 7.x?
[Bug c++/70178] Loop-invariant memory loads from std::string innards are not hoisted
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70178 --- Comment #4 from Jonathan Wakely --- There is no 'restrict' keyword in ISO C++ and no equivalent language feature. I think it would be OK to use __restrict__ on the _M_dataplus._M_p member.
[Bug tree-optimization/56365] Missed opportunities for smin/smax standard name patterns when compiling as C++
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56365 Richard Biener changed: What|Removed |Added Status|NEW |ASSIGNED Assignee|unassigned at gcc dot gnu.org |rguenth at gcc dot gnu.org --- Comment #7 from Richard Biener --- Ok, so let me fix this for GCC 7. It isn't realy "phi merging" but phiopt being confused by our a <= -128 to a < -127 canonicalization.
[Bug fortran/45076] [OOP] gfortran.dg/dynamic_dispatch_6.f03 ICEs with -fprofile-use
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=45076 Dominique d'Humieres changed: What|Removed |Added Status|WAITING |RESOLVED Resolution|--- |FIXED --- Comment #11 from Dominique d'Humieres --- Test committed to the trunk and the gcc 4.9 and 5 branches. Closing as FIXED.
[Bug tree-optimization/69951] wrong code at -O1 and above on x86_64-linux-gnu
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69951 --- Comment #6 from Richard Biener --- Seems I typoed the PR reference in the commit: Author: rguenth Date: Fri Feb 26 08:34:58 2016 New Revision: 233734 URL: https://gcc.gnu.org/viewcvs?rev=233734&root=gcc&view=rev Log: 2016-02-26 Richard Biener PR tree-optimization/69551 * tree-ssa-structalias.c (get_constraint_for_ssa_var): When looking through aliases adjust DECL_PT_UID to refer to the ultimate alias target. * gcc.dg/torture/pr69951.c: New testcase. Added: trunk/gcc/testsuite/gcc.dg/torture/pr69951.c Modified: trunk/gcc/ChangeLog trunk/gcc/testsuite/ChangeLog trunk/gcc/tree-ssa-structalias.c
[Bug rtl-optimization/38644] [4.6 Regression] Optimization flag -O1 -fschedule-insns2 causes wrong code
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=38644
--- Comment #73 from Richard Biener ---
(In reply to Satish M from comment #72)
> This bug still exists in GCC 4.8.2 ARM. It can reproduced by adding one more
> argument in 'doStreamReadBlock' function in test case.
>
> /x86_64-unknown-linux-gnu/bin/gcc -B /x86_64-unknown-linux-gnu/bin
> -msoft-float -marm -mcpu=cortex-a9 -march=armv7-a -mno-thumb-interwork
> -mlong-calls -mno-unaligned-access -O2 test.c
>
> extern int doStreamReadBlock (int *, char *, int size, int, int);
>
> char readStream (int *s)
> {
>char c = 0;
>doStreamReadBlock (s, &c, 1, *s, 22);
>return c;
> }
>
> :
>0: e1a0c00dmov ip, sp
>4: e3a02000mov r2, #0
>8: e92dd800push{fp, ip, lr, pc}
>c: e24cb004sub fp, ip, #4
> 10: e24dd008sub sp, sp, #8
> 14: e24b100csub r1, fp, #12
> 18: e3a0c016mov ip, #22
> 1c: e5612001strbr2, [r1, #-1]!
> 20: e3a02001mov r2, #1
> 24: e5903000ldr r3, [r0]
> 28: e58dc000str ip, [sp]
> 2c: e59fc00cldr ip, [pc, #12] ; 40
> 30: e12fff3cblx ip
> 34: e24bd00csub sp, fp, #12< Stack frame
> de-allocated
> 38: e55b000dldrbr0, [fp, #-13] < Accessing stack.
> 3c: e89da800ldm sp, {fp, sp, pc}
> 40: andeq r0, r0, r0
Please open a new bugreport.
[Bug tree-optimization/69951] wrong code at -O1 and above on x86_64-linux-gnu
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69951 --- Comment #7 from Chengnian Sun --- (In reply to Richard Biener from comment #6) > Seems I typoed the PR reference in the commit: > > Author: rguenth > Date: Fri Feb 26 08:34:58 2016 > New Revision: 233734 > > URL: https://gcc.gnu.org/viewcvs?rev=233734&root=gcc&view=rev > Log: > 2016-02-26 Richard Biener > > PR tree-optimization/69551 > * tree-ssa-structalias.c (get_constraint_for_ssa_var): When > looking through aliases adjust DECL_PT_UID to refer to the > ultimate alias target. > > * gcc.dg/torture/pr69951.c: New testcase. > > Added: > trunk/gcc/testsuite/gcc.dg/torture/pr69951.c > Modified: > trunk/gcc/ChangeLog > trunk/gcc/testsuite/ChangeLog > trunk/gcc/tree-ssa-structalias.c Thank you so much, Richard.
[Bug fortran/67451] [5/6 Regression] [F08] ICE with sourced allocation from coarray.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=67451 vehre at gcc dot gnu.org changed: What|Removed |Added Status|WAITING |RESOLVED Resolution|--- |FIXED --- Comment #16 from vehre at gcc dot gnu.org --- No complaints so far, closing.
[Bug fortran/52673] implement -fheap-arrays in gfortran
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=52673 Dominique d'Humieres changed: What|Removed |Added Status|WAITING |RESOLVED Resolution|--- |WONTFIX --- Comment #3 from Dominique d'Humieres --- > This PR has not evolved since three years and a half and now new options > are available to catch such issues. IMO This PR should be closed as WONTFIX. No feedback, closing.
[Bug libstdc++/69782] [6 Regression] defining min() macro causes thousand of lines of error messages
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69782
--- Comment #8 from Jonathan Wakely ---
(In reply to Jakub Jelinek from comment #7)
> So is the plan to uglify all
> find libstdc++-v3/include -type f | xargs grep '[^a-zA-Z_]\(min\|max\)(' |
> wc -l
> 461
> with ()s around min/max or whatever id-expression it appears at the end of?
I started doing that, and it made me want to hit myself in the face with a
hammer.
Adding a header that #undefs the macros and including it everywhere necessary
is a much smaller patch, and doesn't obfuscate hundreds of valid uses of
min/max functions in the library. It also means users can call the std min/max
functions without parentheses because any macros have been removed. It doesn't
allow people to continue using their undefined macros, but those people are
immoral anyway.
> For 6.x or just 7.x?
I find it hard to care about code with min/max macros.
[Bug libstdc++/69782] [6/7 Regression] defining min() macro causes thousand of lines of error messages
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69782 Jakub Jelinek changed: What|Removed |Added Target Milestone|6.0 |7.0 Summary|[6 Regression] defining |[6/7 Regression] defining |min() macro causes thousand |min() macro causes thousand |of lines of error messages |of lines of error messages --- Comment #9 from Jakub Jelinek --- Retargetting at GCC 7 then, feel free to close as NOTABUG if you don't want to do anything.
[Bug target/62281] gcc doesn't conform to Solaris 32-bit ABI by expecting 16-byte stack alignment
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62281 --- Comment #15 from Rainer Orth --- Author: ro Date: Mon Mar 14 09:56:47 2016 New Revision: 234179 URL: https://gcc.gnu.org/viewcvs?rev=234179&root=gcc&view=rev Log: Only assume 4-byte stack alignment on Solaris/x86 (PR target/62281) Backport from mainline 2016-03-07 Rainer Orth PR target/62281 * config/i386/sol2.h (STACK_REALIGN_DEFAULT): Define. Modified: branches/gcc-5-branch/gcc/ChangeLog branches/gcc-5-branch/gcc/config/i386/sol2.h
[Bug fortran/38979] OpenMP extension: THREADPRIVATE for EQUIVALENCEd symbols
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=38979 Dominique d'Humieres changed: What|Removed |Added CC||jakub at gcc dot gnu.org --- Comment #12 from Dominique d'Humieres --- > Any progress after six years? PING!
[Bug target/62281] gcc doesn't conform to Solaris 32-bit ABI by expecting 16-byte stack alignment
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62281 --- Comment #16 from Rainer Orth --- Author: ro Date: Mon Mar 14 10:03:12 2016 New Revision: 234180 URL: https://gcc.gnu.org/viewcvs?rev=234180&root=gcc&view=rev Log: Only assume 4-byte stack alignment on Solaris/x86 (PR target/62281) Backport from mainline: 2016-03-07 Rainer Orth PR target/62281 * config/i386/sol2.h (STACK_REALIGN_DEFAULT): Define. Revert: 2014-02-11 Rainer Orth PR libgomp/60107 * config/i386/sol2-9.h: New file. * config.gcc (i[34567]86-*-solaris2* | x86_64-*-solaris2.1[0-9]*, *-*-solaris2.9*): Use it. Removed: branches/gcc-4_9-branch/gcc/config/i386/sol2-9.h Modified: branches/gcc-4_9-branch/gcc/ChangeLog branches/gcc-4_9-branch/gcc/config.gcc branches/gcc-4_9-branch/gcc/config/i386/sol2.h
[Bug libgomp/60107] libgomp.c/pr58392.c etc. FAIL
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60107 --- Comment #8 from Rainer Orth --- Author: ro Date: Mon Mar 14 10:03:12 2016 New Revision: 234180 URL: https://gcc.gnu.org/viewcvs?rev=234180&root=gcc&view=rev Log: Only assume 4-byte stack alignment on Solaris/x86 (PR target/62281) Backport from mainline: 2016-03-07 Rainer Orth PR target/62281 * config/i386/sol2.h (STACK_REALIGN_DEFAULT): Define. Revert: 2014-02-11 Rainer Orth PR libgomp/60107 * config/i386/sol2-9.h: New file. * config.gcc (i[34567]86-*-solaris2* | x86_64-*-solaris2.1[0-9]*, *-*-solaris2.9*): Use it. Removed: branches/gcc-4_9-branch/gcc/config/i386/sol2-9.h Modified: branches/gcc-4_9-branch/gcc/ChangeLog branches/gcc-4_9-branch/gcc/config.gcc branches/gcc-4_9-branch/gcc/config/i386/sol2.h
[Bug target/62281] gcc doesn't conform to Solaris 32-bit ABI by expecting 16-byte stack alignment
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62281 Rainer Orth changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|--- |FIXED Target Milestone|5.4 |4.9.4 --- Comment #17 from Rainer Orth --- Fixed for GCC 4.9.4, 5.4, 6.1.
[Bug c/70223] New: [ARM] Optimization level -O2 results in wrong code
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70223
Bug ID: 70223
Summary: [ARM] Optimization level -O2 results in wrong code
Product: gcc
Version: 4.8.2
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
Assignee: unassigned at gcc dot gnu.org
Reporter: satishbmsce at gmail dot com
Target Milestone: ---
-fschedule-insns2 flag within optimization level -O2 results in wrong code. By
adding one more argument to test function 'doStreamReadBlock' defined in file
'gcc/testsuite/gcc.target/arm/stack-red-zone.c' issue can produced.
/x86_64-unknown-linux-gnu/bin/gcc -B /x86_64-unknown-linux-gnu/bin -msoft-float
-marm -mcpu=cortex-a9 -march=armv7-a -mno-thumb-interwork -mlong-calls
-mno-unaligned-access -O2 test.c
test.c:
extern int doStreamReadBlock (int *, char *, int size, int, int);
char readStream (int *s)
{
char c = 0;
doStreamReadBlock (s, &c, 1, *s, 22);
return c;
}
:
0: e1a0c00dmov ip, sp
4: e3a02000mov r2, #0
8: e92dd800push{fp, ip, lr, pc}
c: e24cb004sub fp, ip, #4
10: e24dd008sub sp, sp, #8
14: e24b100csub r1, fp, #12
18: e3a0c016mov ip, #22
1c: e5612001strbr2, [r1, #-1]!
20: e3a02001mov r2, #1
24: e5903000ldr r3, [r0]
28: e58dc000str ip, [sp]
2c: e59fc00cldr ip, [pc, #12] ; 40
30: e12fff3cblx ip
34: e24bd00csub sp, fp, #12< Stack frame
de-allocated
38: e55b000dldrbr0, [fp, #-13] < Accessing stack.
3c: e89da800ldm sp, {fp, sp, pc}
40: andeq r0, r0, r0
[Bug middle-end/70199] [5/6 Regression] Crash at -O2 when using labels.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70199 Richard Biener changed: What|Removed |Added Priority|P1 |P2
[Bug tree-optimization/70193] missed loop splitting support based on iteration space
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70193 Richard Biener changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed||2016-03-14 Blocks||53947 Ever confirmed|0 |1 --- Comment #1 from Richard Biener --- Confirmed. Referenced Bugs: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53947 [Bug 53947] [meta-bug] vectorizer missed-optimizations
[Bug middle-end/70189] Combine constant-pool logic from gimplify + SRA
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70189 Richard Biener changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed||2016-03-14 Ever confirmed|0 |1 --- Comment #1 from Richard Biener --- Confirmed.
[Bug rtl-optimization/70224] New: [6 regression] ICE: RTL flag check: CROSSING_JUMP_P used with unexpected rtx code 'insn' in relax_delay_slots, at reorg.c:3310
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70224 Bug ID: 70224 Summary: [6 regression] ICE: RTL flag check: CROSSING_JUMP_P used with unexpected rtx code 'insn' in relax_delay_slots, at reorg.c:3310 Product: gcc Version: 6.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: rtl-optimization Assignee: unassigned at gcc dot gnu.org Reporter: ro at gcc dot gnu.org Target Milestone: --- Host: sparc*-sun-solaris2.* Target: sparc*-sun-solaris2.* Build: sparc*-sun-solaris2.* Within the last week, gcc.dg/tree-prof/20041218-1.c started to FAIL for the 64-bit multilib only on Solaris/SPARC: FAIL: gcc.dg/tree-prof/20041218-1.c compilation, -fprofile-use -D_PROFILE_USE (internal compiler error) UNRESOLVED: gcc.dg/tree-prof/20041218-1.c execution,-fprofile-use -D_PROFILE_USE /vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/tree-prof/20041218-1.c: In function 'foo':^M /vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/tree-prof/20041218-1.c:110:1: internal compiler error: RTL flag check: CROSSING_JUMP_P used with unexpected rtx code 'insn' in relax_delay_slots, at reorg.c:3310^M 0x913433 rtl_check_failed_flag(char const*, rtx_def const*, char const*, int, char const*)^M /vol/gcc/src/hg/trunk/local/gcc/rtl.c:878^M 0x90dfcb relax_delay_slots^M /vol/gcc/src/hg/trunk/local/gcc/reorg.c:3310^M 0x90dfcb dbr_schedule^M /vol/gcc/src/hg/trunk/local/gcc/reorg.c:3725^M 0x90dfcb rest_of_handle_delay_slots^M /vol/gcc/src/hg/trunk/local/gcc/reorg.c:3866^M 0x90dfcb execute^M /vol/gcc/src/hg/trunk/local/gcc/reorg.c:3897^M Rainer
[Bug rtl-optimization/70224] [6 regression] ICE: RTL flag check: CROSSING_JUMP_P used with unexpected rtx code 'insn' in relax_delay_slots, at reorg.c:3310
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70224 Rainer Orth changed: What|Removed |Added Target Milestone|--- |6.0
[Bug c++/70205] [4.9/5/6 Regression] ICE on valid code on x86_64-linux-gnu: tree check: expected tree_binfo, have error_mark in add_candidates, at cp/call.c:5283
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70205 Richard Biener changed: What|Removed |Added Keywords||ice-on-valid-code Known to work||4.7.4 Target Milestone|--- |4.9.4 Summary|ICE on valid code on|[4.9/5/6 Regression] ICE on |x86_64-linux-gnu: tree |valid code on |check: expected tree_binfo, |x86_64-linux-gnu: tree |have error_mark in |check: expected tree_binfo, |add_candidates, at |have error_mark in |cp/call.c:5283 |add_candidates, at ||cp/call.c:5283 Known to fail||4.8.5 --- Comment #3 from Richard Biener --- Works on the 4.7 branch.
[Bug c++/70205] [4.9/5/6 Regression] ICE on valid code on x86_64-linux-gnu: tree check: expected tree_binfo, have error_mark in add_candidates, at cp/call.c:5283
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70205 Richard Biener changed: What|Removed |Added Priority|P3 |P2
[Bug sanitizer/70147] testcase from hana testsuite gets miscompiled with -fsanitize=undefined
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70147 --- Comment #6 from Martin Liška --- Thank you Jakub for the reduced test-case. As I've read the source code, current_vtt_parm is properly defined in start_preparsed_function function. vtt_parm definition: /* In a base member initializer, we cannot rely on the vtable being set up. We have to indirect via the vtt_parm. */ So as suggested by Jakub, we have to somehow replace ((struct C *) this)->D.2663.D.2604._vptr.A with read of current_vtt_parm which can be quite easily passed to cp_ubsan_dfs_initialize_vtbl_ptrs. Question is how to easily replace these? Should I process a dfs_walk a replace all occurrences, or is there any easier approach? Thanks, Martin
[Bug c++/70201] Dump C++ template instantiations
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70201 Richard Biener changed: What|Removed |Added Status|UNCONFIRMED |ASSIGNED Last reconfirmed||2016-03-14 Ever confirmed|0 |1
[Bug c++/70201] Dump C++ template instantiations
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70201 --- Comment #1 from Jonathan Wakely --- What do you plan to implement? It would be better to produce output compatible with Templight (http://plc.inf.elte.hu/templight/) rather than a GCC-specific dump.
[Bug c++/70222] Test miscompiled with -O1
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70222
Richard Biener changed:
What|Removed |Added
Status|UNCONFIRMED |NEW
Last reconfirmed||2016-03-14
Ever confirmed|0 |1
Known to fail||4.0.4, 4.1.2, 4.2.4, 4.3.6
--- Comment #1 from Richard Biener ---
Confirmed. -O2 is fine again.
void abort (void);
int a = 1;
unsigned int b = 2;
int c = 0;
int d = 0;
void foo ()
{
int e = ((-(c >= c)) < b) > ((int)(((unsigned long long int)(-1)) >> ((a / a)
* 15)));
d = -e;
}
int main ()
{
foo ();
if (d != -1)
abort ();
return 0;
}
[Bug rtl-optimization/70222] Test miscompiled with -O1
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70222 Richard Biener changed: What|Removed |Added CC||segher at gcc dot gnu.org Component|c++ |rtl-optimization --- Comment #2 from Richard Biener --- Inlining hides the RTL opt bug. Disabling RTL combine fixes it.
[Bug rtl-optimization/70224] [6 regression] ICE: RTL flag check: CROSSING_JUMP_P used with unexpected rtx code 'insn' in relax_delay_slots, at reorg.c:3310
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70224 Richard Biener changed: What|Removed |Added Priority|P3 |P1
[Bug rtl-optimization/70223] [ARM] Optimization level -O2 results in wrong code
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70223 Richard Biener changed: What|Removed |Added Keywords||wrong-code Target||arm* Component|c |rtl-optimization --- Comment #1 from Richard Biener --- Can you please check a still supported compiler version (GCC 4.9.3, GCC 5.3 or recent trunk)?
[Bug tree-optimization/56365] Missed opportunities for smin/smax standard name patterns when compiling as C++
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56365 --- Comment #8 from Oleg Endo --- (In reply to Richard Biener from comment #7) > Ok, so let me fix this for GCC 7. Thanks :)
[Bug rtl-optimization/70223] [ARM] Optimization level -O2 results in wrong code
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70223
--- Comment #2 from Ramana Radhakrishnan ---
The ARM port does not use frame pointers unless specified on the command line
or an actual need in the program (read alloca).
With 4.8.5 I get with the command line options suggested *and*
-fno-omit-frame-pointer which seems to be missed in the options provided in the
PR, a number of which are completely redundant :
readStream:
@ args = 0, pretend = 0, frame = 8
@ frame_needed = 1, uses_anonymous_args = 0
stmfd sp!, {fp, lr}
add fp, sp, #4
sub sp, sp, #16
mov lr, #0
sub r1, fp, #4
mov r2, #22
movwip, #:lower16:doStreamReadBlock
ldr r3, [r0]
movtip, #:upper16:doStreamReadBlock
strblr, [r1, #-1]!
str r2, [sp]
mov r2, #1
blx ip
ldrbr0, [fp, #-5] @ zero_extendqisi2
sub sp, fp, #4
@ sp needed
ldmfd sp!, {fp, pc}
Not sure what's going on yet, probably worth a bisect / search between versions
of the compiler in the 4.8 tree ...
Note also that I don't see the problem on the 4.9, 5.X and trunk with the
testcase provided.
[Bug rtl-optimization/70222] Test miscompiled with -O1
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70222
Jakub Jelinek changed:
What|Removed |Added
CC||jakub at gcc dot gnu.org
--- Comment #3 from Jakub Jelinek ---
Yeah, it is combine.
First of all, the -1ULL >> (... * 15) is properly optimized into
(insn 12 11 13 2 (set (reg:DI 102)
(const_int 8589934591 [0x1])) pr70222.c:9 85 {*movdi_internal}
(nil))
but the
(insn 13 12 14 2 (parallel [
(set (reg:SI 104 [ e ])
(lshiftrt:SI (subreg:SI (reg:DI 102) 0)
(const_int 31 [0x1f])))
(clobber (reg:CC 17 flags))
]) pr70222.c:9 562 {*lshrsi3_1}
(expr_list:REG_DEAD (reg:DI 102)
(expr_list:REG_UNUSED (reg:CC 17 flags)
(nil
which is supposed to yield 1 or 0 (in this case 1) is mis-simplified into:
(insn 13 12 14 2 (parallel [
(set (subreg:DI (reg:SI 104 [ e ]) 0)
(lshiftrt:DI (reg:DI 102)
(subreg:QI (reg:SI 101) 0)))
(clobber (reg:CC 17 flags))
]) pr70222.c:9 564 {*lshrdi3_1}
(expr_list:REG_DEAD (reg:SI 101)
(expr_list:REG_UNUSED (reg:CC 17 flags)
(expr_list:REG_DEAD (reg:DI 102)
(nil)
[Bug rtl-optimization/70222] Test miscompiled with -O1
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70222 Jakub Jelinek changed: What|Removed |Added Assignee|unassigned at gcc dot gnu.org |jakub at gcc dot gnu.org Target Milestone|--- |6.0 --- Comment #4 from Jakub Jelinek --- Looking at this.
[Bug libfortran/69799] FAIL: gfortran.dg/coarray_allocate_3.f08 at -O1 and above
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69799 Dominique d'Humieres changed: What|Removed |Added CC||afanfa at gcc dot gnu.org --- Comment #3 from Dominique d'Humieres --- > Needs to link against -latomic. Why? Apparently it does not need it on x86_64-apple-darwin15. What happens if you add '-latomic' to the dg-options list?
[Bug sanitizer/70147] testcase from hana testsuite gets miscompiled with -fsanitize=undefined
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70147
--- Comment #7 from Bernd Edlinger ---
But how about that?
diff -up gcc/cp/cp-ubsan.c.jj gcc/cp/cp-ubsan.c
--- gcc/cp/cp-ubsan.c.jj2016-03-04 23:10:49.0 +0100
+++ gcc/cp/cp-ubsan.c 2016-03-14 12:09:38.730344495 +0100
@@ -283,7 +283,7 @@ cp_ubsan_dfs_initialize_vtbl_ptrs (tree
if (!TYPE_CONTAINS_VPTR_P (BINFO_TYPE (binfo)))
return dfs_skip_bases;
- if (!BINFO_PRIMARY_P (binfo) || BINFO_VIRTUAL_P (binfo))
+ if (!BINFO_PRIMARY_P (binfo) && !BINFO_VIRTUAL_P (binfo))
{
tree base_ptr = TREE_VALUE ((tree) data);
Which does just do the instrumentation on the simple case,
and leaves the virtual base classes alone.
At least the test case from PR70035 does not use virtual base
classes, and will still work.
[Bug target/70117] ppc long double isinf() is wrong?
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70117 Michel Normand changed: What|Removed |Added CC||normand at linux dot vnet.ibm.com --- Comment #8 from Michel Normand --- How to make this problem to be solved between gcc and gnulib ? There is pending discussion in gnulib ML at http://lists.gnu.org/archive/html/bug-gnulib/2016-03/msg00032.html
[Bug tree-optimization/56365] [4.9/5/6 Regression] Missed opportunities for smin/smax standard name patterns when compiling as C++
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56365
Richard Biener changed:
What|Removed |Added
Keywords||missed-optimization
Priority|P3 |P2
Known to work||4.3.6
Target Milestone|--- |4.9.4
Summary|Missed opportunities for|[4.9/5/6 Regression] Missed
|smin/smax standard name |opportunities for smin/smax
|patterns when compiling as |standard name patterns when
|C++ |compiling as C++
Known to fail||4.7.3, 5.3.0, 6.0
Severity|enhancement |normal
--- Comment #9 from Richard Biener ---
In the following testcase we used to do better with GCC 4.3 at least, handling
test_02 and test_04 completely and test_01 and test_03 half-way. Compared
to GCC 5 which only handles test_04 completely and test_03 half-way and
test_01 and test_02 not at all.
So this is a regression probably caused by adding maybe_canonicalize_comparison
to comparison folding, PR26899.
int test_01 (int a)
{
if (127 <= a)
a = 127;
else if (a <= -128)
a = -128;
return a;
}
int test_02 (int a)
{
if (127 < a)
a = 127;
else if (a <= -128)
a = -128;
return a;
}
int test_03 (int a)
{
if (127 <= a)
a = 127;
else if (a < -128)
a = -128;
return a;
}
int test_04 (int a)
{
if (127 < a)
a = 127;
else if (a < -128)
a = -128;
return a;
}
[Bug c++/68732] Internal error while compiling CLRadeonExtender
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68732 Marek Polacek changed: What|Removed |Added Status|WAITING |RESOLVED Resolution|--- |INVALID --- Comment #6 from Marek Polacek --- Feedback not coming.
[Bug c++/70209] [5/6 Regression] ICE in strip_typedefs, at cp/tree.c:1377
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70209 Marek Polacek changed: What|Removed |Added CC||mpolacek at gcc dot gnu.org --- Comment #1 from Marek Polacek --- Starting with r233715 I see the ICE, before that: $ ./cc1plus.233714 -quiet ~/w.cc /home/brq/mpolacek/w.cc: In member function ‘void A< >::m_fn2(Vector&, Vector&, Vector&, Vector&, Vector&) const [with = int]’: /home/brq/mpolacek/w.cc:14:16: error: expected ‘;’ at end of input p5.m_fn1(m_data) ^ /home/brq/mpolacek/w.cc:14:16: error: expected ‘}’ at end of input
[Bug c++/70209] [5/6 Regression] ICE in strip_typedefs, at cp/tree.c:1377
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70209 --- Comment #2 from Marek Polacek --- So the testcase seem to be incomplete.
[Bug c++/70209] [5/6 Regression] ICE in strip_typedefs, at cp/tree.c:1377
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70209 --- Comment #3 from Markus Trippelsdorf --- (In reply to Marek Polacek from comment #2) > So the testcase seem to be incomplete. Sure. It needs an additional "};".
[Bug c++/70201] Dump C++ template instantiations
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70201 --- Comment #2 from Daniel Gutson --- (In reply to Jonathan Wakely from comment #1) > What do you plan to implement? It would be better to produce output > compatible with Templight (http://plc.inf.elte.hu/templight/) rather than a > GCC-specific dump. Actually I planned to implement the human-friendly format that gcc currently outputs when an instantiation fails [with T=...], which I think it will be useful for most users; however I admit that hving templight support is interesting. So I propose to implement -fdump-template-instantations=human or alike, leaving the code cleanly ready to add other dump formats (such as =templight) or =plugin to enable user-defined dump formats (so templight could be implemented as one). What do you think?
[Bug c++/70194] [6 regression] missing -Waddress on constexpr pointer
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70194 Marek Polacek changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed||2016-03-14 CC||mpolacek at gcc dot gnu.org Ever confirmed|0 |1 --- Comment #2 from Marek Polacek --- Started since delayed folding, r230365.
[Bug testsuite/68232] gcc.dg/ifcvt-4.c fails on some arm configurations
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68232 --- Comment #8 from James Greenhalgh --- (In reply to Pat Haugen from comment #6) > (In reply to James Greenhalgh from comment #5) > > "Fixed" with the testsuite skips. Feel free to add any other target triplets > > for which this test is unreliable. > > I was going to modify the powerpc64le triplet to just powerpc*-*-* since it > also fails for powerpc64 (big endian) and powerpc-ibm-aix, but looking at > gcc/config/rs6000/rs6000.h, it has BRANCH_COST defined to a non-zero value: > > #define BRANCH_COST(speed_p, predictable_p) 3 > > > So there must be something more than just "doesn't work for targets with > branch cost == 0". I'm still happy to make the change if there are other > reasons, but didn't want to do so without hearing first. Sorry that I took a while to get round to looking at this. For powerpc64 you'll need to enable conditional move instructions using "-misel" (or equivalent) for this test to pass. For hppa64, the "experimental" movdicc pattern has this restriction: if (GET_MODE (XEXP (operands[1], 0)) != DImode || GET_MODE (XEXP (operands[1], 0)) != GET_MODE (XEXP (operands[1], 1))) But, we're trying to expand with this comparison in operands[1]: (le (subreg/s/u:SI (reg/v:DI 70 [ x+-4 ]) 4) (subreg/s/u:SI (reg/v:DI 71 [ y+-4 ]) 4)) so this test fails, and we fail to ifcvt the sequence. The test should be skipped on hppa64 until more complete support for conditional moves is added.
[Bug c++/70201] Dump C++ template instantiations
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70201 Jonathan Wakely changed: What|Removed |Added CC||jason at gcc dot gnu.org --- Comment #3 from Jonathan Wakely --- As long as it doesn't make Templight output any more difficult I have no objections :-) I don't know if Jason already started looking at what would be needed for Templight output, but he might have some suggestions.
[Bug libfortran/69799] FAIL: gfortran.dg/coarray_allocate_3.f08 at -O1 and above
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69799 --- Comment #4 from dave.anglin at bell dot net --- On 2016-03-14, at 7:17 AM, dominiq at lps dot ens.fr wrote: >> Needs to link against -latomic. > > > Why? Apparently it does not need it on x86_64-apple-darwin15. x86_64 has atomic support fully implemented in header. Some targets like hppa have code in libatomic which calls kernel helpers, etc. > > What happens if you add '-latomic' to the dg-options list? It passes. Dave -- John David Anglin dave.ang...@bell.net
[Bug fortran/56440] Compiler crash for f90 with C-binded subroutine name, which is equals to module name
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56440 Dominique d'Humieres changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |DUPLICATE --- Comment #4 from Dominique d'Humieres --- It looks like a duplicate of pr53478. *** This bug has been marked as a duplicate of bug 53478 ***
[Bug fortran/53478] gfortran segfaults when module name clashes with C binding name of procedure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53478 Dominique d'Humieres changed: What|Removed |Added CC||kapojko at yandex dot ru --- Comment #7 from Dominique d'Humieres --- *** Bug 56440 has been marked as a duplicate of this bug. ***
[Bug rtl-optimization/70222] Test miscompiled with -O1
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70222 --- Comment #5 from Jakub Jelinek --- So, the bug is apparently in simplify_shift_const_1 (code=LSHIFTRT, result_mode=SImode, varop=0x7184de70, orig_count=31) where varop is: (subreg:SI (lshiftrt:DI (const_int -1 [0x]) (subreg:QI (reg:SI 100) 0)) 0) We optimize the inner shift (mode == DImode) into (lshiftrt:DI (const_int 8589934591 [0x1]) (subreg:QI (reg:SI 100) 0)) 0) which is fine (assuming we do the final masking), and turn shift count into 0. shift_mode is still SImode, because nonzero_bits doesn't say the upper bits are 0, and we don't have OUTER_OP. For this to really work, we need to mask off the bits though. We have: /* If we were doing an LSHIFTRT in a wider mode than it was originally, turn off all the bits that the shift would have turned off. */ if (orig_code == LSHIFTRT && result_mode != shift_mode) x = simplify_and_const_int (NULL_RTX, shift_mode, x, GET_MODE_MASK (result_mode) >> orig_count); that does this, unfortunately it doesn't trigger here, because shift_mode == result_mode == SImode. So, one possibility is to do if (orig_code == LSHIFTRT && (result_mode != shift_mode || (result_mode != mode && count == 0))) (and maybe "&& orig_count != 0" too?), to say that we need to mask even if the shift in shift_mode actually isn't performed at all (or shall it be orig_count != count ?). Or we'd need to force outer_op in this case somewhere, but not sure what all cases would need to do that.
[Bug target/38239] [4.9, 5, 6 regression] The function mcount in gmon.o changes register ecx
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=38239 --- Comment #3 from Rainer Orth --- Created attachment 37955 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=37955&action=edit proposed patch
[Bug target/38239] [4.9, 5, 6 regression] The function mcount in gmon.o changes register ecx
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=38239 Rainer Orth changed: What|Removed |Added Status|UNCONFIRMED |ASSIGNED Last reconfirmed||2016-03-14 Known to work||4.3.2, 4.4.2, 4.5.2, 4.6.0, ||4.7.2, 4.8.0 Assignee|unassigned at gcc dot gnu.org |ro at gcc dot gnu.org Target Milestone|--- |6.0 Summary|The function mcount in |[4.9, 5, 6 regression] The |gmon.o changes register ecx |function mcount in gmon.o ||changes register ecx Ever confirmed|0 |1 Known to fail||4.9.0, 5.1.0, 6.0 --- Comment #2 from Rainer Orth --- I can now reproduce this, starting with gcc 4.9. After some digging, I found that _mcount isn't called like a regular function, but has to take care of saving and restoring call-clobbered registers itself. Rainer
[Bug rtl-optimization/70222] Test miscompiled with -O1
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70222 --- Comment #6 from Jakub Jelinek --- Created attachment 37956 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=37956&action=edit gcc6-pr70222.patch This untested patch works. Though, if that is the right way to go, I'd bootstrap/regtest it with some statistics gathering hack to see how often does it trigger.
[Bug libfortran/69799] FAIL: gfortran.dg/coarray_allocate_3.f08 at -O1 and above
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69799
--- Comment #5 from Dominique d'Humieres ---
Could you please test the following patch
--- ../_clean/gcc/testsuite/gfortran.dg/coarray_allocate_3.f08 2016-02-03
19:58:35.0 +0100
+++ gcc/testsuite/gfortran.dg/coarray_allocate_3.f082016-03-14
14:12:28.0 +0100
@@ -1,5 +1,6 @@
! { dg-do run }
! { dg-options "-fcoarray=single" }
+! { dg-additional-options "-latomic" { target hppa-*-* } }
!
! Contributed by Ian Harvey
! Extended by Andre Vehreschild
?
[Bug middle-end/55152] MAX_EXPR(a,-a) is really ABS_EXPR(a)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=55152 Richard Biener changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed||2016-03-14 CC||rguenth at gcc dot gnu.org Ever confirmed|0 |1 --- Comment #5 from Richard Biener --- Now sth for match.pd for integers we get : _2 = -a_1(D); _3 = MAX_EXPR ; return _3; even w/o -ffast-math.
[Bug c/59424] Optimization issue on min/max
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=59424
Richard Biener changed:
What|Removed |Added
Status|UNCONFIRMED |NEW
Last reconfirmed||2016-03-14
Ever confirmed|0 |1
--- Comment #4 from Richard Biener ---
Note that phiopt could still do min/max detection of a derived value, thus
for the PHI args look through unary ops it can associate a min/max with
and re-write as op (min/max (a, b)). Not sure if that's worth it complicating
the code.
Other than that removing the
else if (TREE_CODE (arg0) == COND_EXPR)
{
from fold_unary that does the operation sinking would "fix" this as well.
[Bug target/70220] [x86] interrupt attribute optionally needs to provide read, write and control the set of saved registers
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70220
--- Comment #3 from H.J. Lu ---
(In reply to Wink Saville from comment #2)
> (In reply to H.J. Lu from comment #1)
> > (In reply to Wink Saville from comment #0)
> > > I have identified one possible problem and with this scheme, what if the
> > > compiler needs to setup a stack frame by pushing rbp and then moving rsp
> > > to
> > > rbp, how would that case be handled.
> >
> > Why should be it a problem unless you don't want to restore RSP and RBP
> > to its original values upon returning from ISR. Please provide an example
> > here.
>
> Here a possible example, I added a printf and local variables to
> timer_reschedule_isr:
>
> void timer_reschedule_isr(struct intr_frame* frame) {
> __asm__ volatile(""::: "rax", "rbx", "rcx", "rdx", "rsi", "rdi", "rbp",
> "r8", "r9", "r10", "r11", "r12", "r13", "r14",
> "r15");
>
> If I now remove "rbp" from the clobber list it compiles:
>
rbp is special. Will remove rbp from the clobber list cause any
problems here?
[Bug tree-optimization/66580] max reduction does not auto vectorize
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66580 Richard Biener changed: What|Removed |Added Target||x64-mingw Status|NEW |WAITING --- Comment #2 from Richard Biener --- Not sure, but it works for me with GCC 5.3.0 on x86_64-linux, also with GCC 4.9.
[Bug target/70220] [x86] interrupt attribute optionally needs to provide read, write and control the set of saved registers
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70220 --- Comment #4 from H.J. Lu --- (In reply to Wink Saville from comment #0) > I'm using the new C interrupt attribute for x86 and its working well. But > when I expanded its use to include handling thread context switches, I found > that its current implementation lacking. > > When using an ISR for a thread context it is necessary to save and restore > all registers but by default it saves a few as it thinks are necessary. It > is possible to use the clobber list of inline assembly to tell the interrupt > attribute code to save additional registers but you can't seem to mention > the segment registers in that clobber list, thus to save these additional > registers you must manually save them manually. > Compiler doesn't use segment registers, except for TLS, which should be used in ISR. ISR needs to save and restore any registers, which aren't used by compilers, if they are changed in ISR.
[Bug fortran/66643] Missing compilation error for formatted data transfer without format
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66643 --- Comment #5 from Dominique d'Humieres --- > Created attachment 37610 [details] > A preliminary patch > > Attached patch gives a diagnostic. Interestingly I found three test cases > in the test suite that use the invalid construct. Works as expected without regression.
[Bug tree-optimization/68894] Recognition min/max pattern with multiple arguments.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68894 --- Comment #4 from Richard Biener --- This is a missed code hoisting opportunity again. _6 = a1[i_24]; _7 = a2[i_24]; if (_6 < _7) goto ; else goto ; : _8 = a3[i_24]; _22 = MIN_EXPR <_6, _8>; goto ; : _9 = a3[i_24]; _28 = MIN_EXPR <_7, _9>; : # d_2 = PHI <_22(6), _28(7)> we fail to hoist/CSE _8/_9 here. There is another PR speaking about a similar issue of phiopt / CSE / hoisting interaction.
[Bug sanitizer/70147] [6 Regression] testcase from hana testsuite gets miscompiled with -fsanitize=undefined
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70147 Volker Reichelt changed: What|Removed |Added CC||reichelt at gcc dot gnu.org Known to work||5.3.0 Summary|testcase from hana |[6 Regression] testcase |testsuite gets miscompiled |from hana testsuite gets |with -fsanitize=undefined |miscompiled with ||-fsanitize=undefined Known to fail||6.0 --- Comment #8 from Volker Reichelt --- This is a regression that was introduced between the 20160228 snapshot and the 20160306 snapshot.
[Bug fortran/70198] simple test floating point sequence gives incorrect values-- optimizer changes them
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70198 --- Comment #7 from scott.b.dorsey at nasa dot gov --- Oh, that is lovely. Many many thanks. This code is from the "Numerical Recipes in Fortran" book and has been used around here as a standard test for the past 40 years without anyone noticing this. --scott From: dominiq at lps dot ens.fr [gcc-bugzi...@gcc.gnu.org] Sent: Saturday, March 12, 2016 4:32 AM To: s.b.dor...@larc.nasa.gov Subject: [Bug fortran/70198] simple test floating point sequence gives incorrect values-- optimizer changes them https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70198 --- Comment #6 from Dominique d'Humieres --- > If the code is compiled with the -fno-automatic option, > one generates the desired result. Note I said desired, > not correct, because the code has a significant bug. > When m=2 (see comment #4), the computed goto jumps to > label 4. The do-loop with the label 6 is executed. > The variable 'a' is undefined. It appears that this > subroutine needs a SAVE statement. Same problem with the array q. The code gives the "desired" result with the following patch --- pr70198.f 2016-03-11 22:10:39.0 +0100 +++ pr70198_db_1.f 2016-03-12 10:26:36.0 +0100 @@ -164,6 +164,7 @@ C C This routine performs Runge-Kutta calculation by Gills Method C dimension y(10),f(10),q(10) + save a, q m=m+1 go to (1,4,5,3,7), m 1 do 2 i=1,n -- You are receiving this mail because: You reported the bug.
[Bug sanitizer/70147] [6 Regression] testcase from hana testsuite gets miscompiled with -fsanitize=undefined
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70147 --- Comment #9 from Martin Liška --- (In reply to Volker Reichelt from comment #8) > This is a regression that was introduced between the 20160228 snapshot and > the 20160306 snapshot. Yeah, it was introduced by Jakub's patch: commit e968ff6d2edad40dcc7508d413d05b71a010e2ce Author: jakub Date: Fri Mar 4 22:10:49 2016 + PR c++/70035 * cp-tree.h (cp_ubsan_maybe_initialize_vtbl_ptrs): New prototype. * decl.c (start_preparsed_function): Call cp_ubsan_maybe_initialize_vtbl_ptrs if needed. * cp-ubsan.c (cp_ubsan_dfs_initialize_vtbl_ptrs, cp_ubsan_maybe_initialize_vtbl_ptrs): New functions. * g++.dg/ubsan/pr70035.C: New test. git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/trunk@233984 138bc75d-0d04-0410-961f-82ee72b054a4 Martin
[Bug libfortran/69799] FAIL: gfortran.dg/coarray_allocate_3.f08 at -O1 and above
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69799 --- Comment #6 from dave.anglin at bell dot net --- On 2016-03-14 9:19 AM, dominiq at lps dot ens.fr wrote: > Could you please test the following patch Actually, I already have but "hppa-*-*" should be "libatomic_available". There are one or two tests with this.
[Bug target/70098] PowerPC64: eigen hits ICE following invalid register assignment
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70098 --- Comment #8 from Segher Boessenkool --- Author: segher Date: Mon Mar 14 14:18:44 2016 New Revision: 234181 URL: https://gcc.gnu.org/viewcvs?rev=234181&root=gcc&view=rev Log: rs6000: Handle "d" output in the bd*z patterns (PR70098) In the rs6000 port, FLOAT_REGS can contain DImode values when compiling for 64-bit targets. Some instructions (like "fcfid" in the testcase, convert from integer to DP float) only work on floating point registers. So, we do want to allow DImode in these regs. Now, in unusual cases IRA will assign FLOAT_REGS to some allocno where some insns cannot handle FLOAT_REGS there, so they will need a reload. Maybe IRA can be made smarter, but it isn't doing anything wrong here, so we should be able to handle it. The place it goes wrong is in the output of the *ctrdi_internal[1256] pattern: the "bdz" and "bdnz" instructions. GCC refuses to do output reloads on JUMP_INSNs, probably because it is hard to do, needs different strategies than "normal" reloads do, and it cannot even be done at all for general patterns. So JUMP_INSNs need to be able to handle every possible output for the register class used. These patterns already handle writing to "c" (the base insn case), and to "r", "m", and "c" or "l"; all those via splitters. We just need to handle "d" as well. That is what this patch does. [A predicate in one of the splitters needs to be touched up so that the correct splitter is used for the FLOAT_REGS case.] But, that leaves another problem. One of the insns that are split to is a move from a GPR to an FPR. That work fine on targets with direct move (which does exactly that), i.e. power8 and up. But older targets need memory to do the move, and this splitter runs after reload so it cannot allocate memory; and allocating memory beforehand for every bdnz insn is pretty horrible as well. This patch implements the easy part. With it, power8 works, where it didn't before. PR target/70098 * config/rs6000/rs6000.md (*ctr_internal1, *ctr_internal2, *ctr_internal5, *ctr_internal6): Also allow "d" as output. (define_split for the GPR case): Use int_reg_operand instead of gpc_reg_operand for the output. gcc/testsuite/ PR target/70098 * lib/target-supports.exp (check_effective_target_powerpc64_no_dm): New function. * g++.dg/pr70098.C: New testcase. Added: trunk/gcc/testsuite/g++.dg/pr70098.C Modified: trunk/gcc/ChangeLog trunk/gcc/config/rs6000/rs6000.md trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/lib/target-supports.exp
[Bug debug/24801] -d option changes debug information
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=24801 Oleg Endo changed: What|Removed |Added Status|WAITING |RESOLVED Resolution|--- |WORKSFORME
[Bug other/29842] [meta-bug] outstanding patches / issues from STMicroelectronics
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=29842 Bug 29842 depends on bug 24801, which changed state. Bug 24801 Summary: -d option changes debug information https://gcc.gnu.org/bugzilla/show_bug.cgi?id=24801 What|Removed |Added Status|WAITING |RESOLVED Resolution|--- |WORKSFORME
[Bug middle-end/70219] [6 Regression] ICE: in delete_move_and_clobber, at lra-constraints.c:5864 with -O2
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70219 --- Comment #4 from Vladimir Makarov --- (In reply to Jakub Jelinek from comment #3) > Created attachment 37953 [details] > gcc6-pr70219.patch > > Untested fix. The code had assertion dregno > 0, but I don't see anything > special on register 0 (%rax on x86_64), assuming Vlad meant >= 0 here (i.e. > make sure dregno is not negative, which is not a hard register, but has > special meaning). Yes, I meant >=0. I am really sorry for such obvious mistake. I was to distracted to other things. Jakub, could submit your patch.
[Bug tree-optimization/56365] [4.9/5 Regression] Missed opportunities for smin/smax standard name patterns when compiling as C++
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56365 --- Comment #11 from Richard Biener --- Author: rguenth Date: Mon Mar 14 14:50:40 2016 New Revision: 234183 URL: https://gcc.gnu.org/viewcvs?rev=234183&root=gcc&view=rev Log: 2016-03-14 Richard Biener PR tree-optimization/56365 * tree-ssa-phiopt.c (minmax_replacement): Handle alternate constants to compare against. * gcc.dg/tree-ssa/phi-opt-14.c: New testcase. Added: trunk/gcc/testsuite/gcc.dg/tree-ssa/phi-opt-14.c Modified: trunk/gcc/ChangeLog trunk/gcc/testsuite/ChangeLog trunk/gcc/tree-ssa-phiopt.c
[Bug tree-optimization/56365] [4.9/5 Regression] Missed opportunities for smin/smax standard name patterns when compiling as C++
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56365 Richard Biener changed: What|Removed |Added Known to work||6.0 Summary|[4.9/5/6 Regression] Missed |[4.9/5 Regression] Missed |opportunities for smin/smax |opportunities for smin/smax |standard name patterns when |standard name patterns when |compiling as C++|compiling as C++ Known to fail|6.0 | --- Comment #10 from Richard Biener --- Fixed on trunk sofar. Not exactly planning to backport.
[Bug target/70117] ppc long double isinf() is wrong?
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70117 Richard Biener changed: What|Removed |Added Target||powerpc* Status|UNCONFIRMED |NEW Last reconfirmed||2016-03-14 Ever confirmed|0 |1 --- Comment #9 from Richard Biener --- So another way to "fix" this would be to stop the simplification of isinf to isgreater(fabs(x),DBL_MAX) by providing patterns for isinf and friends (isfinite, isnormal), see fold_builtin_interclass_mathfn. Though neither isfinite nor isnormal have associated optabs - breaking testcases for them could be easily constructed. For the long double case more optimal implementation should be possible in the backend. Given we can't easily expose long double implementation details on GENERIC/GIMPLE it needs to be done there (well, in theory we could use sth. like VIEW_CONVERT (ldbl_1).d[0] or so ... or REALPART (V_C_E <_Complex double> (ldbl_1)) whatever would work better. But then struct realformat needs to expose this fact (real.c checks fmt->pnan < fmt->p for this!?). Confirmed anyway. The question remains whether the long double value from gnulib is "valid" (it's a valid bit pattern that is _not_ inf after all). I still somewhat think that GCC is at fault here and not gnulib - even if it is technically not LDBL_MAX that is wrong but the folding to isgreater(fabs(x),DBL_MAX).
[Bug tree-optimization/70225] New: ICE at tree.c:10783
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70225 Bug ID: 70225 Summary: ICE at tree.c:10783 Product: gcc Version: 5.3.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: tree-optimization Assignee: unassigned at gcc dot gnu.org Reporter: tarasevich at cs dot uni-saarland.de Target Milestone: --- Created attachment 37957 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=37957&action=edit test case Test case pr24257.c from torture test suite triggers ICE at tree.c:10783, when compiled with "-Os -fcheck-data-deps". GCC 5.3.0 built from source on Ubuntu 14.04 ../build/gcc_530_clean_bin/bin/gcc -Os -fcheck-data-deps -v -save-temps /home/tarasevich/source/gcc_530/gcc/testsuite//gcc.dg//torture//pr24257.c Using built-in specs. COLLECT_GCC=../build/gcc_530_clean_bin/bin/gcc COLLECT_LTO_WRAPPER=/home/tarasevich/build/gcc_530_clean_bin/libexec/gcc/x86_64-unknown-linux-gnu/5.3.0/lto-wrapper Target: x86_64-unknown-linux-gnu Configured with: ../../source/gcc_530/configure --prefix=/home/tarasevich/build/gcc_530_clean_bin/ --enable-languages=c --disable-multilib --disable-bootstrap CC=/home/tarasevich/build/llvm_371_bin/bin/clang CXX=/home/tarasevich/build/llvm_371_bin/bin/clang++ Thread model: posix gcc version 5.3.0 (GCC) COLLECT_GCC_OPTIONS='-Os' '-fcheck-data-deps' '-v' '-save-temps' '-mtune=generic' '-march=x86-64' /home/tarasevich/build/gcc_530_clean_bin/libexec/gcc/x86_64-unknown-linux-gnu/5.3.0/cc1 -E -quiet -v -imultiarch x86_64-linux-gnu /home/tarasevich/source/gcc_530/gcc/testsuite//gcc.dg//torture//pr24257.c -mtune=generic -march=x86-64 -fcheck-data-deps -Os -fpch-preprocess -o pr24257.i ignoring nonexistent directory "/usr/local/include/x86_64-linux-gnu" ignoring nonexistent directory "/home/tarasevich/build/gcc_530_clean_bin/lib/gcc/x86_64-unknown-linux-gnu/5.3.0/../../../../x86_64-unknown-linux-gnu/include" #include "..." search starts here: #include <...> search starts here: /home/tarasevich/build/gcc_530_clean_bin/lib/gcc/x86_64-unknown-linux-gnu/5.3.0/include /usr/local/include /home/tarasevich/build/gcc_530_clean_bin/include /home/tarasevich/build/gcc_530_clean_bin/lib/gcc/x86_64-unknown-linux-gnu/5.3.0/include-fixed /usr/include/x86_64-linux-gnu /usr/include End of search list. COLLECT_GCC_OPTIONS='-Os' '-fcheck-data-deps' '-v' '-save-temps' '-mtune=generic' '-march=x86-64' /home/tarasevich/build/gcc_530_clean_bin/libexec/gcc/x86_64-unknown-linux-gnu/5.3.0/cc1 -fpreprocessed pr24257.i -quiet -dumpbase pr24257.c -mtune=generic -march=x86-64 -auxbase pr24257 -Os -version -fcheck-data-deps -o pr24257.s GNU C11 (GCC) version 5.3.0 (x86_64-unknown-linux-gnu) compiled by GNU C version 4.2.1 Compatible Clang 3.7.1 (tags/RELEASE_371/final 263010), GMP version 5.1.3, MPFR version 3.1.2-p3, MPC version 1.0.1 GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072 GNU C11 (GCC) version 5.3.0 (x86_64-unknown-linux-gnu) compiled by GNU C version 4.2.1 Compatible Clang 3.7.1 (tags/RELEASE_371/final 263010), GMP version 5.1.3, MPFR version 3.1.2-p3, MPC version 1.0.1 GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072 Compiler executable checksum: c53bb545b4c066cdbf5e42f0b9d9ad8b /home/tarasevich/source/gcc_530/gcc/testsuite//gcc.dg//torture//pr24257.c: In function 'oof': /home/tarasevich/source/gcc_530/gcc/testsuite//gcc.dg//torture//pr24257.c:20:6: internal compiler error: in int_cst_value, at tree.c:10783 void oof(A *s, int n) ^ 0xaed214 int_cst_value(tree_node const*) ../../../source/gcc_530/gcc/tree.c:10783 0xece07d init_omega_eq_with_af(omega_pb_d*, unsigned int, unsigned int, tree_node*, data_dependence_relation*) ../../../source/gcc_530/gcc/tree-data-ref.c:3772 0xecd417 omega_setup_subscript(tree_node*, tree_node*, data_dependence_relation*, omega_pb_d*, bool*) ../../../source/gcc_530/gcc/tree-data-ref.c:3922 0xecd417 init_omega_for_ddr_1(data_reference*, data_reference*, data_dependence_relation*, omega_pb_d*, bool*) ../../../source/gcc_530/gcc/tree-data-ref.c:3959 0xec7b63 init_omega_for_ddr(data_dependence_relation*, bool*) ../../../source/gcc_530/gcc/tree-data-ref.c:4142 0xec7b63 compute_affine_dependence(data_dependence_relation*, loop*) ../../../source/gcc_530/gcc/tree-data-ref.c:4301 0xec83ba compute_all_dependences(vec, vec*, vec, bool) ../../../source/gcc_530/gcc/tree-data-ref.c:4384 0xec8de8 compute_data_dependences_for_loop(loop*, bool, vec*, vec*, vec*) ../../../source/gcc_530/gcc/tree-data-ref.c:4712 0xec9148 analyze_all_data_dependences(loop*) ../../../source/gcc_530/gcc/tree-data-ref.c:4820 0xec9148 tree_check_data_deps() ../../../source/gcc_530/gcc/tree-data-ref.c:4868 0xa1b99d (anonymous namespace)::pass_check_data_deps::execute(function*) ../../../source/gcc_530/gcc/tree-ssa-loop.c:342 Please submit a ful
[Bug target/70117] ppc long double isinf() is wrong?
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70117
--- Comment #10 from Richard Biener ---
So, "simplest" patch:
Index: gcc/builtins.c
===
--- gcc/builtins.c (revision 234180)
+++ gcc/builtins.c (working copy)
@@ -7529,6 +7529,13 @@ fold_builtin_interclass_mathfn (location
mode = TYPE_MODE (TREE_TYPE (arg));
+ /* Do not perform this simplification for IBM extended double
+ which is made up of two IEEE doubles where only the first
+ one is relevant to the result of isinf, isfinite and isnormal. */
+ const struct real_format *fmt = FLOAT_MODE_FORMAT (mode);
+ if (fmt->pnan < fmt->p)
+return NULL_TREE;
+
/* If there is no optab, try generic code. */
switch (DECL_FUNCTION_CODE (fndecl))
{
but it will end up with libcalls for isinf, isfinite and isnormal for
IBM extended double. I'm told glibc does the right thing for the gnulib
DBL_MAX.
PPC people should evaluate if that's eventually even more efficient than
the inlining with the comparison against the large constant immediate.
[Bug target/70083] [6 Regression] ICE: in assign_stack_local_1, at function.c:409 with -fschedule-insns -mavx512* @ i686
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70083 --- Comment #5 from Bernd Schmidt --- Author: bernds Date: Mon Mar 14 15:08:54 2016 New Revision: 234184 URL: https://gcc.gnu.org/viewcvs?rev=234184&root=gcc&view=rev Log: LRA: identify biggest access mode for hard_regs and use it in split_reg PR target/70083 * lra-lives.c (process_bb_lives): Also update biggest mode for hard regs. (lra_create_live_ranges_1): initialize hard register biggest_mode to VOIDmode. * lra-constraints.c (split_reg): For hard regs, try to find the biggest single-register mode used in the function. testsuite/ PR target/70083 * gcc.dg/torture/pr70083.c: New test. * gcc.target/i386/pr70083.c: New test. Added: trunk/gcc/testsuite/gcc.dg/torture/pr70083.c trunk/gcc/testsuite/gcc.target/i386/pr70083.c Modified: trunk/gcc/ChangeLog trunk/gcc/lra-constraints.c trunk/gcc/lra-lives.c trunk/gcc/testsuite/ChangeLog
