[Bug target/84113] [7/8 Regression] libgcc/unwind.inc:136:1: unrecognizable insn: internal compiler error on Darwin

2018-02-04 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84113 --- Comment #32 from Segher Boessenkool --- Yes, run "make -k check" (add -jN to taste if you have multiple CPUs). And then run contrib/test_summary. See if that is as expected (compare it to previous powerpc-darwin results on gcc-testresults@;

[Bug target/84113] [7/8 Regression] libgcc/unwind.inc:136:1: unrecognizable insn: internal compiler error on Darwin

2018-02-04 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84113 --- Comment #29 from Segher Boessenkool --- The patch has never been sent to gcc-patches. It also still never was properly tested. Can you do that?

[Bug target/84113] [7/8 Regression] libgcc/unwind.inc:136:1: unrecognizable insn: internal compiler error on Darwin

2018-02-04 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84113 --- Comment #27 from Segher Boessenkool --- Yes.

[Bug target/84113] [7/8 Regression] libgcc/unwind.inc:136:1: unrecognizable insn: internal compiler error on Darwin

2018-02-04 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84113 --- Comment #25 from Segher Boessenkool --- Please test https://gcc.gnu.org/ml/gcc-testresults/2017-01/txtnZhWiDkC4z.txt .

[Bug target/84113] [7/8 Regression] libgcc/unwind.inc:136:1: unrecognizable insn: internal compiler error on Darwin

2018-02-04 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84113 --- Comment #23 from Segher Boessenkool --- (In reply to Douglas Mencken from comment #22) > as for https://gcc.gnu.org/ml/gcc-patches/2016-08/msg02040.html > I think that optimizing epilogues is a good idea, but not of a cost of >

[Bug target/84169] [8 Regression] wrong code with u128 multiplication result at aarch64 -O

2018-02-02 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84169 --- Comment #4 from Segher Boessenkool --- This starts to fail with r249850 (don't know yet what the problem is).

[Bug target/84169] [8 Regression] wrong code with u128 multiplication result at aarch64 -O

2018-02-02 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84169 --- Comment #3 from Segher Boessenkool --- Confirmed. Does not happen with a compiler a few months old.

[Bug rtl-optimization/84157] [8 Regression] [nvptx] ICE: RTL check: expected code 'reg', have 'lshiftrt'

2018-02-01 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84157 --- Comment #9 from Segher Boessenkool --- I've tested this on powerpc64-linux {-m64,-m32} as well (change_zero_ext is pretty important for us; we have big testcases for this and related code). No issues found.

[Bug rtl-optimization/84157] [8 Regression] [nvptx] ICE: RTL check: expected code 'reg', have 'lshiftrt'

2018-02-01 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84157 --- Comment #7 from Segher Boessenkool --- (In reply to Tom de Vries from comment #6) > (In reply to Segher Boessenkool from comment #5) > > That looks good. > that patch was by Uros though. Yes I misread. Uros: that patch is pre-approved.

[Bug target/84113] [7/8 Regression] gcc-7.3.0/libgcc/unwind.inc:136:1: internal compiler error: in extract_insn, at recog.c:2311

2018-02-01 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84113 --- Comment #16 from Segher Boessenkool --- Darwin has its own ABI, and that is what those assembler routines are for. This has nothing to do with this bug report; please use the mailing lists for other questions/discussions/etc.

[Bug rtl-optimization/84157] [8 Regression] [nvptx] ICE: RTL check: expected code 'reg', have 'lshiftrt'

2018-02-01 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84157 --- Comment #5 from Segher Boessenkool --- Hi Tom, That looks good. But maybe the REG_P check should be in the outer "if"? Slightly easier and safer to exit as early as possible.

[Bug target/84113] [7/8 Regression] gcc-7.3.0/libgcc/unwind.inc:136:1: internal compiler error: in extract_insn, at recog.c:2311

2018-01-31 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84113 --- Comment #11 from Segher Boessenkool --- It will be fixed. But not many people have access to powerpc-darwin systems to test on. Does this still happen on trunk as well?

[Bug target/84113] [7/8 Regression] gcc-7.3.0/libgcc/unwind.inc:136:1: internal compiler error: in extract_insn, at recog.c:2311

2018-01-30 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84113 Segher Boessenkool changed: What|Removed |Added CC||iains at gcc dot gnu.org ---

[Bug target/84113] [7/8 Regression] gcc-7.3.0/libgcc/unwind.inc:136:1: internal compiler error: in extract_insn, at recog.c:2311

2018-01-30 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84113 Segher Boessenkool changed: What|Removed |Added CC||segher at gcc dot gnu.org

[Bug rtl-optimization/84123] internal compiler error: in gen_rtx_SUBREG, at emit-rtl.c:908, alpha linux.

2018-01-30 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84123 --- Comment #2 from Segher Boessenkool --- That looks good, please send the patch! What about the third case? The first in the "if" tree. Does that need this as well? (The third and last case already has a similar check).

[Bug target/83496] MIPS BE: wrong code generates under "-Os -mbranch-cost=1"

2018-01-28 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83496 --- Comment #13 from Segher Boessenkool --- There are *two* conditional branches leading to that "return 1". After dbr one is a delay-slot seq, and the other has lost the assignment of the return value.

[Bug target/83496] MIPS BE: wrong code generates under "-Os -mbranch-cost=1"

2018-01-28 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83496 --- Comment #12 from Segher Boessenkool --- It's the dbr ("reorg") pass that is deleting the insn.

[Bug target/56010] Powerpc, -mcpu=native and -mtune=native use the wrong name for target 7450

2018-01-26 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56010 --- Comment #10 from Segher Boessenkool --- (In reply to Peter Bergner from comment #9) > (In reply to Segher Boessenkool from comment #8) > >> This kernel AT_PLATFORM name should strip the '+' off: > >> .platform = "power7+", -> "power7" > >

[Bug target/83946] [7/8 Regression] Safe Indirect Jumps broken on AIX

2018-01-26 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83946 --- Comment #11 from Segher Boessenkool --- Author: segher Date: Fri Jan 26 21:31:37 2018 New Revision: 257110 URL: https://gcc.gnu.org/viewcvs?rev=257110=gcc=rev Log: rs6000: Fix safe-indirect-jump-[18].c Thist patch merges the

[Bug target/56010] Powerpc, -mcpu=native and -mtune=native use the wrong name for target 7450

2018-01-25 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56010 Segher Boessenkool changed: What|Removed |Added CC||segher at gcc dot gnu.org

[Bug rtl-optimization/83985] [8 Regression] Compile time hog for 32-bit BE powerpc targets

2018-01-25 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83985 --- Comment #10 from Segher Boessenkool --- Ah, #c6 made me think you wanted to do something else. I didn't look close enough, sorry. That patch looks good.

[Bug rtl-optimization/83985] [8 Regression] Compile time hog for 32-bit BE powerpc targets

2018-01-25 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83985 Segher Boessenkool changed: What|Removed |Added CC||segher at gcc dot gnu.org

[Bug target/83926] ICE during RTL pass: ira, in elimination_costs_in_insn, at reload1.c:3633

2018-01-24 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83926 --- Comment #7 from Segher Boessenkool --- It is first V2DI in the RTL, which exists just fine (but there is no such divide insn); then when it is split to two DImode divides, it just generates div:DI etc., which does not exist for -m32. So we

[Bug target/83964] [8 Regression] ICE in extract_insn, at recog.c:2304

2018-01-23 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83964 --- Comment #3 from Segher Boessenkool --- No, many other builtins are also problematic (there are a few more PRs open).

[Bug target/81550] [8 regression] gcc.target/powerpc/loop_align.c fails starting with r250482

2018-01-23 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81550 --- Comment #8 from Segher Boessenkool --- Yes, but that does not work if ivopts decides to make a loop that cannot work with bdnz ;-) cc:ing Bin

[Bug target/81550] [8 regression] gcc.target/powerpc/loop_align.c fails starting with r250482

2018-01-23 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81550 Segher Boessenkool changed: What|Removed |Added CC||segher at gcc dot gnu.org

[Bug target/83969] [8 Regression] ICE in final_scan_insn, at final.c:2997 (error: could not split insn) for powerpc targets

2018-01-22 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83969 --- Comment #2 from Segher Boessenkool --- Where do you get TImode with this testcase?

[Bug target/83399] Power8 ICE During LRA with 2-op rtl pattern for lvx instruction

2018-01-22 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83399 Segher Boessenkool changed: What|Removed |Added Target Milestone|--- |6.5

[Bug target/83964] [8 Regression] ICE in extract_insn, at recog.c:2304

2018-01-22 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83964 Segher Boessenkool changed: What|Removed |Added Status|UNCONFIRMED |NEW Ever confirmed|0

[Bug debug/83758] ICE building gccgo on powerpc64le --with-cpu=power8

2018-01-16 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83758 Segher Boessenkool changed: What|Removed |Added CC||segher at gcc dot gnu.org

[Bug target/83789] __builtin_altivec_lvx fails for powerpc for altivec-4.c

2018-01-16 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83789 --- Comment #3 from Segher Boessenkool --- Does not reproduce with powerpc64-linux-gcc -m32 (not on trunk and not on current 7).

[Bug target/43871] -mcpu=power4 -mtune=cell emits power7/cell-only opcodes

2018-01-16 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43871 Segher Boessenkool changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|---

[Bug rtl-optimization/83424] [6/7 Regression] wrong code with -O -fno-tree-ccp -fno-tree-coalesce-vars

2018-01-16 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83424 Segher Boessenkool changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|---

[Bug rtl-optimization/83424] [6/7 Regression] wrong code with -O -fno-tree-ccp -fno-tree-coalesce-vars

2018-01-16 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83424 --- Comment #6 from Segher Boessenkool --- Author: segher Date: Tue Jan 16 16:30:43 2018 New Revision: 256752 URL: https://gcc.gnu.org/viewcvs?rev=256752=gcc=rev Log: rtlanal: dead_or_set_regno_p should handle CLOBBER (PR83424) In PR83424

[Bug rtl-optimization/83424] [6/7 Regression] wrong code with -O -fno-tree-ccp -fno-tree-coalesce-vars

2018-01-16 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83424 --- Comment #5 from Segher Boessenkool --- Author: segher Date: Tue Jan 16 16:20:10 2018 New Revision: 256750 URL: https://gcc.gnu.org/viewcvs?rev=256750=gcc=rev Log: rtlanal: dead_or_set_regno_p should handle CLOBBER (PR83424) In PR83424

[Bug target/83789] __builtin_altivec_lvx fails for powerpc for altivec-4.c

2018-01-16 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83789 --- Comment #2 from Segher Boessenkool --- (In reply to Kaushikp from comment #0) > Should we modify the testcase so it runs only for 64-bit targets, > or should this builtin work for 32-bit targets as well? It shouldn't ICE for 32-bit

[Bug target/83629] [7 Regression] ICE: in decompose_normal_address, at rtlanal.c:6329 with -O2 -fPIC -frename-registers --param=sched-autopref-queue-depth=nnn

2018-01-15 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83629 Segher Boessenkool changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|---

[Bug target/83629] [7 Regression] ICE: in decompose_normal_address, at rtlanal.c:6329 with -O2 -fPIC -frename-registers --param=sched-autopref-queue-depth=nnn

2018-01-15 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83629 --- Comment #7 from Segher Boessenkool --- Author: segher Date: Mon Jan 15 22:08:12 2018 New Revision: 256711 URL: https://gcc.gnu.org/viewcvs?rev=256711=gcc=rev Log: Backport from mainline 2018-01-10 Segher Boessenkool

[Bug target/83629] [7 Regression] ICE: in decompose_normal_address, at rtlanal.c:6329 with -O2 -fPIC -frename-registers --param=sched-autopref-queue-depth=nnn

2018-01-12 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83629 --- Comment #6 from Segher Boessenkool --- Author: segher Date: Fri Jan 12 12:10:16 2018 New Revision: 256577 URL: https://gcc.gnu.org/viewcvs?rev=256577=gcc=rev Log: rs6000: Tune new testcase (PR83629) It has some problems running on some

[Bug target/79140] gcc.target/powerpc/ssp-1.c fails starting with its introduction in r244562

2018-01-11 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=79140 Segher Boessenkool changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|---

[Bug target/78875] -fstack-protector on powerpc64 now always use TLS, won't work for kernel/firmware

2018-01-11 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78875 Segher Boessenkool changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|---

[Bug target/68834] ICE: in decompose_normal_address, at rtlanal.c:6086 with -O2 -fPIC --param=sched-autopref-queue-depth=1

2018-01-10 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68834 --- Comment #2 from Segher Boessenkool --- Likely the same issue as pr83629 .

[Bug target/83629] [7/8 Regression] ICE: in decompose_normal_address, at rtlanal.c:6329 with -O2 -fPIC -frename-registers --param=sched-autopref-queue-depth=nnn

2018-01-10 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83629 --- Comment #5 from Segher Boessenkool --- Fixed on trunk; awaiting backports.

[Bug target/83629] [7/8 Regression] ICE: in decompose_normal_address, at rtlanal.c:6329 with -O2 -fPIC -frename-registers --param=sched-autopref-queue-depth=nnn

2018-01-10 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83629 --- Comment #4 from Segher Boessenkool --- Author: segher Date: Wed Jan 10 15:13:07 2018 New Revision: 256432 URL: https://gcc.gnu.org/viewcvs?rev=256432=gcc=rev Log: rs6000: Wrap diff of immediates in const (PR83629) In various of our 32-bit

[Bug target/83743] -mcpu=native causes gcc to exit in error if cpu is not recognized

2018-01-09 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83743 --- Comment #4 from Segher Boessenkool --- Which is perfectly fine, except the error message is a bit confusing if you don't know what's going on.

[Bug target/83399] Power8 ICE During LRA with 2-op rtl pattern for lvx instruction

2018-01-09 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83399 Segher Boessenkool changed: What|Removed |Added Keywords|ice-on-valid-code |ice-on-invalid-code --- Comment

[Bug target/83743] -mcpu=native causes gcc to exit in error if cpu is not recognized

2018-01-09 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83743 Segher Boessenkool changed: What|Removed |Added CC||segher at gcc dot gnu.org

[Bug target/83399] Power8 ICE During LRA with 2-op rtl pattern for lvx instruction

2018-01-08 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83399 --- Comment #5 from Segher Boessenkool --- That last testcase needs -mabi=elfv2 to fail (but doesn't need that -O3).

[Bug rtl-optimization/83629] [7/8 Regression] ICE: in decompose_normal_address, at rtlanal.c:6329 with -O2 -fPIC -frename-registers --param=sched-autopref-queue-depth=nnn

2018-01-08 Thread segher at gcc dot gnu.org
||2018-01-08 Assignee|unassigned at gcc dot gnu.org |segher at gcc dot gnu.org Ever confirmed|0 |1 --- Comment #3 from Segher Boessenkool --- I have a fuller patch; it's testing (your variant 1).

[Bug rtl-optimization/83629] [7/8 Regression] ICE: in decompose_normal_address, at rtlanal.c:6329 with -O2 -fPIC -frename-registers --param=sched-autopref-queue-depth=nnn

2018-01-08 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83629 --- Comment #2 from Segher Boessenkool --- Needs e.g. -mcpu=603 to fail.

[Bug web/78315] "Changes" don't explain what "LRA" is

2018-01-07 Thread segher at gcc dot gnu.org
||segher at gcc dot gnu.org Resolution|--- |WORKSFORME --- Comment #3 from Segher Boessenkool --- It is more than plenty: what is the default for new ports only concerns developers of GCC itself. Everyone else is free to read some

[Bug target/83507] [8 Regression] ICE in internal_dfa_insn_code_* for powerpc targets

2018-01-05 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83507 --- Comment #8 from Segher Boessenkool --- Ah yes, I can reproduce it with all those options and -m32. I could swear to have tried exactly that yesterday; pilot error I guess. CA is not just a hard register, but a fixed register, btw.

[Bug target/83507] [8 Regression] ICE in internal_dfa_insn_code_* for powerpc targets

2018-01-04 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83507 Segher Boessenkool changed: What|Removed |Added Keywords|ice-on-valid-code |ice-on-invalid-code --- Comment #2

[Bug target/83507] [8 Regression] ICE in internal_dfa_insn_code_* for powerpc targets

2018-01-04 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83507 --- Comment #1 from Segher Boessenkool --- I cannot reproduce this on either powerpc-linux or powerpc64-linux.

[Bug rtl-optimization/83628] [8 Regression] performance regression when accessing arrays on alpha

2018-01-04 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83628 --- Comment #5 from Segher Boessenkool --- The MULT is not canonical. For some reason we do not realise we can do the ASHIFT as SI instead of DI.

[Bug target/83660] ICE with vec_extract inside expression statement

2018-01-03 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83660 Segher Boessenkool changed: What|Removed |Added Target|powerpc64*-*-* |powerpc*-*-* --- Comment #3 from

[Bug target/83387] PowerPC64: Infinite loops in do_reload() with -msoft-float

2018-01-03 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83387 --- Comment #16 from Segher Boessenkool --- > > > --- gcc/config/rs6000/rs6000.c(revision 255700) > > > +++ gcc/config/rs6000/rs6000.c(working copy) > > > @@ -11095,7 +11095,8 @@ rs6000_discover_homogeneous_aggregate (m > > >

[Bug rtl-optimization/83565] RTL combine pass breaks shift result (at least on ia64)

2017-12-29 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83565 --- Comment #24 from Segher Boessenkool --- Binary size (of vmlinux) grows by 0.09%. I don't know about performance.

[Bug rtl-optimization/83565] RTL combine pass breaks shift result (at least on ia64)

2017-12-24 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83565 --- Comment #14 from Segher Boessenkool --- (In reply to Jim Wilson from comment #11) > This part > > >r358 is known to have the high half zero: > > 22: r358:DI=r357:SI#0^r341:SI#0 > > is wrong. The upper bits of both registers are unknown

[Bug target/83315] PowerPC xmmintrin.h emulation for _mm_{min,max}_ps not semantically equivalent

2017-12-23 Thread segher at gcc dot gnu.org
|unassigned at gcc dot gnu.org |segher at gcc dot gnu.org

[Bug rtl-optimization/83565] RTL combine pass breaks shift result (at least on ia64)

2017-12-23 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83565 --- Comment #9 from Segher Boessenkool --- #0 means subreg 0; in this case, (subreg:DI (reg:SI 357) 0) etc. (The "slim" RTL dumps are lossy in places btw; in _this_ case you can deduce it has to be DImode, but not always. OTOH slim dumps

[Bug rtl-optimization/83565] RTL combine pass breaks shift result (at least on ia64)

2017-12-23 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83565 --- Comment #7 from Segher Boessenkool --- (In reply to Sergei Trofimovich from comment #5) > gcc's try_combine() scares me lots :) It's an acquired taste. > Segher, can you guide me here what changes size of zero_extract in here from > 0x1f

[Bug target/83399] Power8 ICE During LRA with 2-op rtl pattern for lvx instruction

2017-12-23 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83399 Segher Boessenkool changed: What|Removed |Added Keywords||ice-on-invalid-code

[Bug rtl-optimization/80791] [8 regression] test case gcc.dg/sms-1.c fail2 starting with r247885

2017-12-23 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80791 --- Comment #11 from Segher Boessenkool --- In the 64-bit Power ABIs, an int is passed as a sign-extended 64-bit register. Take this trivial example: int f(int x) { return x*x; } which compiles to (on powerpc64-linux, BE): .L.f: mullw

[Bug rtl-optimization/80791] [8 regression] test case gcc.dg/sms-1.c fail2 starting with r247885

2017-12-20 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80791 Segher Boessenkool changed: What|Removed |Added CC||segher at gcc dot gnu.org

[Bug tree-optimization/81010] [8 regression] test case gcc.target/powerpc/pr56605.c fails starting with r248958

2017-12-19 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81010 --- Comment #5 from Segher Boessenkool --- Oh, heh, the testcase isn't run at all on powerpc64le. That should be fixed, too.

[Bug tree-optimization/81010] [8 regression] test case gcc.target/powerpc/pr56605.c fails starting with r248958

2017-12-19 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81010 --- Comment #4 from Segher Boessenkool --- But I see the same for LE ELFv2? I wonder what's going on...

[Bug tree-optimization/81010] [8 regression] test case gcc.target/powerpc/pr56605.c fails starting with r248958

2017-12-19 Thread segher at gcc dot gnu.org
||2017-12-19 CC||segher at gcc dot gnu.org Ever confirmed|0 |1 --- Comment #3 from Segher Boessenkool --- The testcase looks for /* { dg-final { scan-rtl-dump-times "\\\(compare:CC \\\(subreg:SI \\\(r

[Bug rtl-optimization/83424] [6/7 Regression] wrong code with -O -fno-tree-ccp -fno-tree-coalesce-vars

2017-12-18 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83424 Segher Boessenkool changed: What|Removed |Added Target Milestone|8.0 |6.4 Summary|[8

[Bug rtl-optimization/83424] [8 Regression] wrong code with -O -fno-tree-ccp -fno-tree-coalesce-vars

2017-12-18 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83424 --- Comment #4 from Segher Boessenkool --- This is latent on the branches; queued for backports.

[Bug rtl-optimization/83424] [8 Regression] wrong code with -O -fno-tree-ccp -fno-tree-coalesce-vars

2017-12-18 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83424 --- Comment #3 from Segher Boessenkool --- Author: segher Date: Mon Dec 18 20:16:19 2017 New Revision: 255787 URL: https://gcc.gnu.org/viewcvs?rev=255787=gcc=rev Log: rtlanal: dead_or_set_regno_p should handle CLOBBER (PR83424) In PR83424

[Bug rtl-optimization/83361] [8 Regression] ICE: verify_flow_info failed (error: non-cold basic block 3 reachable only by paths crossing the cold partition) on 32-bit BE powerpc targets

2017-12-17 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83361 --- Comment #5 from Segher Boessenkool --- No, it probably needs backports.

[Bug rtl-optimization/83424] [8 Regression] wrong code with -O -fno-tree-ccp -fno-tree-coalesce-vars

2017-12-16 Thread segher at gcc dot gnu.org
|unassigned at gcc dot gnu.org |segher at gcc dot gnu.org --- Comment #2 from Segher Boessenkool --- Mine.

[Bug c/66425] (void) cast doesn't suppress __attribute__((warn_unused_result))

2017-12-15 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66425 --- Comment #32 from Segher Boessenkool --- Yes, it does hurt. Quite many people use casts to void automatically on all function calls where they do not use the result. They of course need to be re-educated on that. Casts to void do not

[Bug c++/83437] [8 Regression] bogus warning: cast between incompatible function types from ‘int (*)()’ to ‘V’ {aka ‘void* (*)()’ [-Wcast-function-type]

2017-12-15 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83437 Segher Boessenkool changed: What|Removed |Added CC||segher at gcc dot gnu.org

[Bug rtl-optimization/83393] [8 Regression] wrong code with -O2 -fno-forward-propagate -fno-tree-bit-ccp

2017-12-13 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83393 Segher Boessenkool changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|---

[Bug rtl-optimization/83393] [8 Regression] wrong code with -O2 -fno-forward-propagate -fno-tree-bit-ccp

2017-12-13 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83393 --- Comment #4 from Segher Boessenkool --- Author: segher Date: Wed Dec 13 14:05:57 2017 New Revision: 255606 URL: https://gcc.gnu.org/viewcvs?rev=255606=gcc=rev Log: combine: Fix PR83393 In move_deaths we move a REG_DEAD note if the

[Bug rtl-optimization/83393] [8 Regression] wrong code with -O2 -fno-forward-propagate -fno-tree-bit-ccp

2017-12-12 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83393 --- Comment #3 from Segher Boessenkool --- In the original code, r108 as set by insn 14 is dead after insn 16. The combination 15->20 should have changed that, but apparently didn't.

[Bug rtl-optimization/83364] [8 Regression] wrong code with -O -fno-forward-propagate -fno-tree-coalesce-vars -fno-tree-ter

2017-12-12 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83364 --- Comment #2 from Segher Boessenkool --- I cannot reproduce this one, either. I built with r29.

[Bug rtl-optimization/83363] [8 Regression] wrong code at -O2 -fno-forward-propagate

2017-12-12 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83363 --- Comment #2 from Segher Boessenkool --- I cannot reproduce the problem. It generates better optimised code than before (removing half the insns in foo), but it is still correct and does not abort.

[Bug rtl-optimization/83361] [8 Regression] ICE: verify_flow_info failed (error: non-cold basic block 3 reachable only by paths crossing the cold partition) on 32-bit BE powerpc targets

2017-12-11 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83361 --- Comment #3 from Segher Boessenkool --- Author: segher Date: Mon Dec 11 21:23:08 2017 New Revision: 20 URL: https://gcc.gnu.org/viewcvs?rev=20=gcc=rev Log: ifcvt: Call fixup_partitions (PR83361) After converting a conditional branch

[Bug libquadmath/83341] [8 Regression] r254899 causes about 15000 test failures on powerpc

2017-12-11 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83341 Segher Boessenkool changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|---

[Bug rtl-optimization/83361] [8 Regression] ICE: verify_flow_info failed (error: non-cold basic block 3 reachable only by paths crossing the cold partition) on 32-bit BE powerpc targets

2017-12-11 Thread segher at gcc dot gnu.org
|unassigned at gcc dot gnu.org |segher at gcc dot gnu.org --- Comment #2 from Segher Boessenkool --- I have a patch.

[Bug tree-optimization/83361] [8 Regression] ICE: verify_flow_info failed (error: non-cold basic block 3 reachable only by paths crossing the cold partition) on 32-bit BE powerpc targets

2017-12-11 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83361 Segher Boessenkool changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed|

[Bug other/83341] r254899 causes about 15000 test failures on powerpc

2017-12-10 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83341 --- Comment #1 from Segher Boessenkool --- Actually, I'm blaming the wrong patch there. Sorry.

[Bug other/83341] New: r254899 causes about 15000 test failures on powerpc

2017-12-08 Thread segher at gcc dot gnu.org
Component: other Assignee: unassigned at gcc dot gnu.org Reporter: segher at gcc dot gnu.org Target Milestone: --- Since r254899 libquadmath isn't built for the default powerpc64-linux configuration any more, causing all fortran run tests for -m32 to fail.

[Bug rtl-optimization/83304] [8 Regression] FAIL: gcc.c-torture/execute/pr61725.c -O3 -fomit-frame-pointer -funroll-loops -fpeel-loops -ftracer -finline-functions

2017-12-08 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83304 --- Comment #11 from Segher Boessenkool --- Author: segher Date: Fri Dec 8 11:26:35 2017 New Revision: 255506 URL: https://gcc.gnu.org/viewcvs?rev=255506=gcc=rev Log: combine: Fix PR83304 In PR83304 two insns are combined, where the I2 uses a

[Bug rtl-optimization/83304] [8 Regression] FAIL: gcc.c-torture/execute/pr61725.c -O3 -fomit-frame-pointer -funroll-loops -fpeel-loops -ftracer -finline-functions

2017-12-08 Thread segher at gcc dot gnu.org
||2017-12-08 Assignee|unassigned at gcc dot gnu.org |segher at gcc dot gnu.org Ever confirmed|0 |1 --- Comment #10 from Segher Boessenkool --- I have a patch.

[Bug rtl-optimization/83304] [8 Regression] FAIL: gcc.c-torture/execute/pr61725.c -O3 -fomit-frame-pointer -funroll-loops -fpeel-loops -ftracer -finline-functions

2017-12-07 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83304 --- Comment #9 from Segher Boessenkool --- At some point during combine we have insn_cost 4 for72: r127:SI=0xff81 insn_cost 4 for 9: r130:SI=ctz(r125:SI) insn_cost 4 for69: cc:CC=cmp(r125:SI,0) insn_cost 4 for70:

[Bug target/43871] -mcpu=power4 -mtune=cell emits power7/cell-only opcodes

2017-12-07 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43871 --- Comment #7 from Segher Boessenkool --- Author: segher Date: Thu Dec 7 09:36:28 2017 New Revision: 255464 URL: https://gcc.gnu.org/viewcvs?rev=255464=gcc=rev Log: rs6000: Initialise rs6000_cpu correctly (PR43871) Finally, set rs6000_cpu

[Bug rtl-optimization/83304] [8 Regression] FAIL: gcc.c-torture/execute/pr61725.c -O3 -fomit-frame-pointer -funroll-loops -fpeel-loops -ftracer -finline-functions

2017-12-06 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83304 --- Comment #7 from Segher Boessenkool --- But losing a clobber like that is just fine (even losing a SET is fine, if its dest is REG_UNUSED, and combine actually does that in certain cases).

[Bug rtl-optimization/83304] [8 Regression] FAIL: gcc.c-torture/execute/pr61725.c -O3 -fomit-frame-pointer -funroll-loops -fpeel-loops -ftracer -finline-functions

2017-12-06 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83304 --- Comment #4 from Segher Boessenkool --- With r255384 combine manages to do many more combinations. Without it, it can get rid of most of the loop body (which should have been optimised away in gimple already really). I don't see how it

[Bug target/70216] [SH] Implement __builtin_trap

2017-12-06 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70216 --- Comment #24 from Segher Boessenkool --- Send it to gcc-patches@? If it is approved, I can commit it, sure.

[Bug rtl-optimization/83304] [8 Regression] FAIL: gcc.c-torture/execute/pr61725.c -O3 -fomit-frame-pointer -funroll-loops -fpeel-loops -ftracer -finline-functions

2017-12-06 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83304 --- Comment #2 from Segher Boessenkool --- I do not see any differences in generated asm code between before r255384 and trunk. Some other options are needed as well?

[Bug target/70216] [SH] Implement __builtin_trap

2017-12-06 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70216 --- Comment #22 from Segher Boessenkool --- ? Why me? What do I have to do with this? It's SH code, I'm not an SH maintainer. /confused

[Bug target/83289] [8 regression] i386/sse2-init-v2di-2.c, i386/avx256-unaligned-load-1.c, i386/avx256-unaligned-store-*.c fails

2017-12-05 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83289 --- Comment #2 from Segher Boessenkool --- Ah. It seems to be because the testcases test for /5, /4 etc., which is now printed as /4 resp. /3, because alternatives start counting at 0.

[Bug target/83289] [8 regression] i386/sse2-init-v2di-2.c, i386/avx256-unaligned-load-1.c, i386/avx256-unaligned-store-*.c fails

2017-12-05 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83289 --- Comment #1 from Segher Boessenkool --- Please show output and output from before it failed. And/or some analysis that shows why current is wrong. Etc. Or even better, make a patch to fix the testcase ;-)

[Bug rtl-optimization/83245] [8 regression] several tests fail starting with r255258

2017-12-04 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83245 Segher Boessenkool changed: What|Removed |Added Status|NEW |RESOLVED Resolution|---

[Bug rtl-optimization/83245] [8 regression] several tests fail starting with r255258

2017-12-04 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83245 --- Comment #3 from Segher Boessenkool --- Author: segher Date: Mon Dec 4 09:30:37 2017 New Revision: 255377 URL: https://gcc.gnu.org/viewcvs?rev=255377=gcc=rev Log: lra: Clobbers in a parallel are earlyclobbers (PR83245) The documentation

[Bug bootstrap/83265] [8 Regression] Bootstrap failure on powerpc64

2017-12-04 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83265 Segher Boessenkool changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|---

<    3   4   5   6   7   8   9   10   11   12   >