[Bug target/104781] New: [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[Bug target/104781] New: [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

            Bug ID: 104781
           Summary: [12 regression] SEGV in _Unwind_GetGR during i386 Ada
                    bootstrap
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: ro at gcc dot gnu.org
                CC: hjl at gcc dot gnu.org
  Target Milestone: ---
            Target: i?86-pc-solaris2.11

Between 20220303 (c083e654bd0f29a365ec957c4c0d4e713fb0b010) and
20220304 (49bc9c6c1ef9441c3db586882e1985da3be9cfa1), 32-bit Solaris/x86 Ada
bootstrap got broken: in stage3, I see

xgcc: internal compiler error: Segmentation Fault signal terminated program
gnat1
Please submit a full bug report, with preprocessed source (by using
-freport-bug).
See <https://gcc.gnu.org/bugs/> for instructions.
make[3]: ***
[/vol/gcc/src/hg/master/local/gcc/ada/gcc-interface/Make-lang.in:167:
ada/casing.o] Error 4

 /var/gcc/regression/master/11.4-gcc/build/./prev-gcc/gnat1 -I - -I . -I
ada/generated -I ada -I /vol/gcc/src/hg/master/local/gcc/ada -I ada/libgnat -I
/vol/gcc/src/hg/master/local/gcc/ada/libgnat -I ada/gcc-interface -I
/vol/gcc/src/hg/master/local/gcc/ada/gcc-interface -quiet -nostdinc -O2 -Wextra
-Wall -dumpdir ada/ -dumpbase casing.adb -dumpbase-ext .adb -gnatwa
-fchecking=1 -g -fchecking=1 -gnatpg -gnata -mtune=generic -march=pentium4
-gnatO ada/casing.o /vol/gcc/src/hg/master/local/gcc/ada/casing.adb -o casing.s

Thread 2 received signal SIGSEGV, Segmentation fault.
[Switching to Thread 1 (LWP 1)]
_Unwind_GetGR (regno=5, context=0xfeffc110) at
/vol/gcc/src/hg/master/local/libgcc/unwind-dw2.c:251
251         return * (_Unwind_Ptr *) (_Unwind_Internal_Ptr) val;
(gdb) bt
#0  _Unwind_GetGR (regno=5, context=0xfeffc110) at
/vol/gcc/src/hg/master/local/libgcc/unwind-dw2.c:251
#1  execute_stack_op (op_ptr=0xb0209f9 "\004e\001", op_ptr@entry=0xb0209f7
"ul\004e\001", op_end=op_end@entry=0xb0209f9 "\004e\001",
context=context@entry=0xfeffc110, initial=<optimized out>) at
/vol/gcc/src/hg/master/local/libgcc/unwind-dw2.c:705
#2  0x0abd2fb7 in uw_update_context_1 (context=context@entry=0xfeffc2d0,
fs=fs@entry=0xfeffc1e0) at
/vol/gcc/src/hg/master/local/libgcc/unwind-dw2.c:1483
#3  0x0abd40bd in uw_init_context_1 (context=context@entry=0xfeffc2d0,
outer_cfa=outer_cfa@entry=0xfeffc4b0, outer_ra=<optimized out>) at
/vol/gcc/src/hg/master/local/libgcc/unwind-dw2.c:1613
#4  0x0abd4792 in _Unwind_RaiseException (exc=0xce14220) at
/vol/gcc/src/hg/master/local/libgcc/unwind.inc:93
#5  0x09223417 in ada.exceptions.exception_propagation.propagate_gcc_exception
()
#6  0x09223461 in ada.exceptions.exception_propagation.propagate_exception ()
#7  0x09223874 in ada.exceptions.complete_and_propagate_occurrence ()
#8  0x09223899 in __gnat_raise_exception ()
#9  0x0921ffcf in system.perfect_hash_generators.compute ()
#10 0x08f3fc06 in exp_imgv.build_enumeration_image_tables ()
#11 0x08eb86e6 in exp_ch13.expand_n_freeze_entity ()
#12 0x08f6c2c5 in expander.expand ()
#13 0x0900abc9 in sem.analyze ()
#14 0x0900c011 in sem.insert_list_before_and_analyze ()
#15 0x08f5b39f in exp_util.insert_actions.part ()
#16 0x08f7ee6f in freeze.freeze_expression ()
#17 0x0916ff2b in sem_res.resolve ()
#18 0x09176bc2 in sem_res.resolve_range ()
#19 0x09170b2d in sem_res.resolve ()
#20 0x09090e6d in sem_ch3.process_range_expr_in_decl ()
#21 0x09091ca0 in sem_ch3.set_scalar_range_for_subtype ()
#22 0x0909317d in sem_ch3.process_subtype ()
#23 0x09094947 in sem_ch3.analyze_subtype_declaration ()
#24 0x0900ae9d in sem.analyze ()
#25 0x09084b93 in sem_ch3.analyze_declarations ()
#26 0x090def7b in sem_ch7.analyze_package_specification ()
#27 0x0900b49b in sem.analyze ()
#28 0x090e1a5e in sem_ch7.analyze_package_declaration ()
#29 0x0900adcf in sem.analyze ()
#30 0x09039fa9 in sem_ch10.analyze_compilation_unit ()
#31 0x0900b112 in sem.analyze ()
#32 0x0900c611 in sem.semantics ()
#33 0x0903a31a in sem_ch10.analyze_compilation_unit ()
#34 0x0900b112 in sem.analyze ()
#35 0x0900c611 in sem.semantics ()
#36 0x08f82689 in frontend ()
#37 0x09230fcd in gnat1drv ()
#38 0x08da2142 in gnat_parse_file() ()
#39 0x0980c91d in compile_file() ()
#40 0x0980fe4a in toplev::main(int, char**) ()
#41 0x0a855761 in main ()
(gdb) p val
$1 = (_Unwind_Context_Reg_Val) 0x0

I'm higly suspicious of

commit 609e8c492d62d92465460eae3d43dfc4b2c68288
Author: H.J. Lu <hjl.tools@gmail.com>
Date:   Sat Feb 26 14:17:23 2022 -0800

    x86: Always return pseudo register in ix86_gen_scratch_sse_rtx

and am current re-bootstrapping with that patch reverted locally.

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

Rainer Orth <ro at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |12.0

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

--- Comment #1 from Rainer Orth <ro at gcc dot gnu.org> ---
Indeed: with the patch reverted, I'm now running make compare.

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2022-03-04
             Status|UNCONFIRMED                 |WAITING
     Ever confirmed|0                           |1
           Assignee|unassigned at gcc dot gnu.org      |hjl.tools at gmail dot com

--- Comment #2 from H.J. Lu <hjl.tools at gmail dot com> ---
Created attachment 52563
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=52563&action=edit
A patch

Try this.

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[ubizjak at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

--- Comment #3 from Uroš Bizjak <ubizjak at gmail dot com> ---
Comment on attachment 52563
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=52563
A patch

>From ba4854c13c4aaa5b50127f23cb09cf05e3eb229d Mon Sep 17 00:00:00 2001
>From: "H.J. Lu" <hjl.tools@gmail.com>
>Date: Fri, 4 Mar 2022 05:53:14 -0800
>Subject: [PATCH] x86: Disable AVX/SSE on unwind-c.c and unwind-dw2.c
>
>Since %ecx can't be used for both DRAP register and __builtin_eh_return,
>we must avoid stack realignment when __builtin_eh_return is used.  Since

Isn't this solved with PR104362?

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

--- Comment #4 from H.J. Lu <hjl.tools at gmail dot com> ---
(In reply to Uroš Bizjak from comment #3)
> Comment on attachment 52563 [details]
> A patch
> 
> >From ba4854c13c4aaa5b50127f23cb09cf05e3eb229d Mon Sep 17 00:00:00 2001
> >From: "H.J. Lu" <hjl.tools@gmail.com>
> >Date: Fri, 4 Mar 2022 05:53:14 -0800
> >Subject: [PATCH] x86: Disable AVX/SSE on unwind-c.c and unwind-dw2.c
> >
> >Since %ecx can't be used for both DRAP register and __builtin_eh_return,
> >we must avoid stack realignment when __builtin_eh_return is used.  Since
> 
> Isn't this solved with PR104362?

It may not be sufficient for Solaris.

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[ro at CeBiTec dot Uni-Bielefeld.DE]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

--- Comment #5 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot Uni-Bielefeld.DE> ---
> --- Comment #2 from H.J. Lu <hjl.tools at gmail dot com> ---
> Created attachment 52563
>   --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=52563&action=edit
> A patch
>
> Try this.

That patch certainly did fix the build.  Thanks.

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #52563|0                           |1
        is obsolete|                            |

--- Comment #6 from H.J. Lu <hjl.tools at gmail dot com> ---
Created attachment 52567
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=52567&action=edit
The v2 patch

Please test this patch.

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[ro at CeBiTec dot Uni-Bielefeld.DE]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

--- Comment #7 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot Uni-Bielefeld.DE> ---
> --- Comment #6 from H.J. Lu <hjl.tools at gmail dot com> ---
> Created attachment 52567
>   --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=52567&action=edit
> The v2 patch
>
> Please test this patch.

This version works just as well.

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P1

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

--- Comment #8 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by H.J. Lu <hjl@gcc.gnu.org>:

https://gcc.gnu.org/g:23ed4df521db9d66782d3f9cf291fc2564cf313a

commit r12-7539-g23ed4df521db9d66782d3f9cf291fc2564cf313a
Author: H.J. Lu <hjl.tools@gmail.com>
Date:   Fri Mar 4 05:53:14 2022 -0800

    x86: Disallow unsupported EH return

    Disallow stack realignment and regparm nested function with EH return
    since they don't work together.

    gcc/

            PR target/104781
            * config/i386/i386.cc (ix86_expand_epilogue): Sorry if there is
            stack realignment or regparm nested function with EH return.

    gcc/testsuite/

            PR target/104781
            * gcc.target/i386/eh_return-1.c: Add -mincoming-stack-boundary=4.
            * gcc.target/i386/eh_return-2.c: Likewise.

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

--- Comment #9 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Jakub Jelinek <jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:caa6c33c5d5223c50657b08e73177e8d54ceee51

commit r12-7565-gcaa6c33c5d5223c50657b08e73177e8d54ceee51
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Wed Mar 9 15:25:30 2022 +0100

    x86: Define LIBGCC2_UNWIND_ATTRIBUTE on ia32 [PR104781]

    On Mon, Mar 07, 2022 at 07:06:28AM -0800, H.J. Lu wrote:
    > Since eh_return doesn't work with stack realignment, disable SSE on
    > unwind-c.c and unwind-dw2.c to avoid stack realignment with the 4-byte
    > incoming stack to avoid SSE usage which is caused by

    The following change does that using LIBGCC2_UNWIND_ATTRIBUTE macro
instead,
    for ia32 only by forcing -mgeneral-regs-only on routines that call
    __builtin_eh_return in libgcc.

    2022-03-09  Jakub Jelinek  <jakub@redhat.com>

            PR target/104781
            * config/i386/i386.h (LIBGCC2_UNWIND_ATTRIBUTE): Define for ia32.

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #10 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
*** Bug 104838 has been marked as a duplicate of this bug. ***

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

--- Comment #11 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
I believe this should now be fixed, Rainer, can you please confirm?

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[ro at CeBiTec dot Uni-Bielefeld.DE]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

--- Comment #12 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot Uni-Bielefeld.DE> ---
> --- Comment #11 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
> I believe this should now be fixed, Rainer, can you please confirm?

It is: last night's i386-pc-solaris2.11 bootstraps were fine without any
local patches.

Thanks.

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|WAITING                     |RESOLVED

--- Comment #13 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Thanks.

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[drepper.fsp+rhbz at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

Ulrich Drepper <drepper.fsp+rhbz at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |drepper.fsp+rhbz at gmail dot com

--- Comment #14 from Ulrich Drepper <drepper.fsp+rhbz at gmail dot com> ---
My standard build now fails in stage 1 with this patch when compiling
unwind-dw2.c for x86:

.../gcc-builds/20220315/gcc/include/cetintrin.h: In function
‘_Unwind_RaiseException’:
.../gcc-builds/20220315/gcc/include/cetintrin.h:47:1: error: inlining failed in
call to ‘always_inline’ ‘_get_ssp’: target specific option mismatch
   47 | _get_ssp (void)
      | ^~~~~~~~
.../gnu/gcc/libgcc/config/i386/shadow-stack-unwind.h:32:26: note: called from
here
   32 |       _Unwind_Word ssp = _get_ssp ();           \
      |                          ^~~~~~~~~~~
.../gnu/gcc/libgcc/unwind-dw2.c:1654:7: note: in expansion of macro
‘_Unwind_Frames_Extra’
 1654 |       _Unwind_Frames_Extra (FRAMES);                                   
\
      |       ^~~~~~~~~~~~~~~~~~~~
.../gnu/gcc/libgcc/unwind.inc:140:3: note: in expansion of macro
‘uw_install_context’
  140 |   uw_install_context (&this_context, &cur_context, frames);


and so on for the other CET functions.  When I comment out the
LIBGCC2_UNWIND_ATTRIBUTE definition in config/i386/i386.h it works.

The compiler command line that is used is

.../gcc-builds/20220315/./gcc/xgcc -B.../gcc-builds/20220315/./gcc/
-B/usr/x86_64-redhat-linux/bin/ -B/usr/x86_64-redhat-linux/lib/ -isystem
/usr/x86_64-redhat-linux/include -isystem /usr/x86_64-redhat-linux/sys-include 
 -fno-checking -g -O2 -m32 -O2  -g -O2 -DIN_GCC    -W -Wall -Wno-narrowing
-Wwrite-strings -Wcast-qual -Wno-format -Wstrict-prototypes
-Wmissing-prototypes -Wold-style-definition  -isystem ./include  -fpic
-mlong-double-80 -DUSE_ELF_SYMVER -fcf-protection -mshstk -g -DIN_LIBGCC2
-fbuilding-libgcc -fno-stack-protector  -fpic -mlong-double-80 -DUSE_ELF_SYMVER
-fcf-protection -mshstk -I. -I. -I../../.././gcc -I.../gcc/libgcc
-I.../gcc/libgcc/. -I.../gcc/libgcc/../gcc -I.../gcc/libgcc/../include
-I.../gcc/libgcc/config/libbid -DENABLE_DECIMAL_BID_FORMAT -DHAVE_CC_TLS 
-DUSE_TLS  -o unwind-dw2.o -MT unwind-dw2.o -MD -MP -MF unwind-dw2.dep
-fexceptions -c .../gcc/libgcc/unwind-dw2.c -fvisibility=hidden -DHIDE_EXPORTS

This is on an up-to-date Fedora 35 system with gcc 11.2.1-9

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781

--- Comment #15 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
r12-7654-gda24fce323eaf088 should fix that.

[Bug target/104781] [12 regression] SEGV in _Unwind_GetGR during i386 Ada bootstrap

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104781
Bug 104781 depends on bug 104890, which changed state.

Bug 104890 Summary: [11 Regression] <x86gprintrin.h> doesn't work with __attribute__((target ("general-regs-only"))) and -march=i686
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104890

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED