[Bug ada/115106] New: [15 regression] SEGV in sem_elab.internal_representation.nts_map.mutate_and_rehash

[Bug ada/115106] New: [15 regression] SEGV in sem_elab.internal_representation.nts_map.mutate_and_rehash

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115106

            Bug ID: 115106
           Summary: [15 regression] SEGV in
                    sem_elab.internal_representation.nts_map.mutate_and_re
                    hash
           Product: gcc
           Version: 15.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: ada
          Assignee: unassigned at gcc dot gnu.org
          Reporter: ro at gcc dot gnu.org
                CC: dkm at gcc dot gnu.org
  Target Milestone: ---
              Host: i386-pc-solaris2.11
            Target: i386-pc-solaris2.11
             Build: i386-pc-solaris2.11

Between 20240514 (0a99ad5c52caa06c113b1889bbe6634812b89be5) and 20240515
(5609d77e683944439fae38323ecabc44a1eb4671),
Ada bootstrap broke in stage 3 on Solaris/x86:

+===========================GNAT BUG DETECTED==============================+
| 15.0.0 20240515 (experimental) [master
5609d77e683944439fae38323ecabc44a1eb4671] (i386-pc-solaris2.11) |
| Constraint_Error SIGSEGV                                                 |
| Error detected at table.adb:219:13 [ali.ads:315:4]                       |
| Compiling /vol/gcc/src/hg/master/local/gcc/ada/ali.adb                   |
| Please submit a bug report; see https://gcc.gnu.org/bugs/ .              |
| Use a subject line meaningful to you and us to track the bug.            |
| Include the entire contents of this bug box in the report.               |
| Include the exact command that you entered.                              |
| Also include sources listed below.                                       |
+==========================================================================+

Please include these source files with error report
Note that list may not be accurate in some cases,
so please double check that the problem can still
be reproduced with the set of files listed.
Consider also -gnatd.n switch (see debug.adb).

/vol/gcc/src/hg/master/local/gcc/ada/gcc-interface/system.ads
/vol/gcc/src/hg/master/local/gcc/ada/ali.adb
/vol/gcc/src/hg/master/local/gcc/ada/ali.ads
/vol/gcc/src/hg/master/local/gcc/ada/casing.ads
/vol/gcc/src/hg/master/local/gcc/ada/namet.ads
/vol/gcc/src/hg/master/local/gcc/ada/alloc.ads
/vol/gcc/src/hg/master/local/gcc/ada/hostparm.ads
/vol/gcc/src/hg/master/local/gcc/ada/types.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/ada.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/a-unccon.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/a-uncdea.ads
/vol/gcc/src/hg/master/local/gcc/ada/table.ads
ada/generated/gnatvsn.ads
/vol/gcc/src/hg/master/local/gcc/ada/rident.ads
ada/s-rident.ads
ada/gnat.ads
ada/g-dyntab.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/g-htable.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-htable.ads
/vol/gcc/src/hg/master/local/gcc/ada/butil.ads
/vol/gcc/src/hg/master/local/gcc/ada/debug.ads
/vol/gcc/src/hg/master/local/gcc/ada/fname.ads
/vol/gcc/src/hg/master/local/gcc/ada/opt.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-string.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-wchcon.ads
/vol/gcc/src/hg/master/local/gcc/ada/osint.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-os_lib.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-stoele.ads
/vol/gcc/src/hg/master/local/gcc/ada/output.ads
ada/snames.ads
ada/g-dynhta.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-strhas.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-stalib.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-exctab.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-unstyp.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-conca2.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-assert.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/a-assert.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-secsta.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-parame.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/a-except.ads
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-traent.ads
/vol/gcc/src/hg/master/local/gcc/ada/table.adb
/vol/gcc/src/hg/master/local/gcc/ada/libgnat/s-memory.ads

compilation abandoned
make: *** [/vol/gcc/src/hg/master/local/gcc/ada/gcc-interface/Make-lang.in:166:
ada/ali.o] Error 1

gdb shows

  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 ali.adb -dumpbase-ext .adb -gnatwa -fchecking=1
-g -fchecking=1 -gnatpg -gnatwns -gnata -fno-PIE -mtune=generic -march=pentium4
-gnatO ada/ali.o /vol/gcc/src/hg/master/local/gcc/ada/ali.adb -o ali.s

Thread 2 received signal SIGSEGV, Segmentation fault.
[Switching to Thread 1 (LWP 1)]
0x0920c153 in sem_elab.internal_representation.nts_map.mutate_and_rehash ()
(gdb) bt
#0  0x0920c153 in sem_elab.internal_representation.nts_map.mutate_and_rehash ()
#1  0x09214c60 in sem_elab.internal_representation.scenario_representation_of
    ()
#2  0x0921e024 in sem_elab.guaranteed_abe_processor.process_guaranteed_abe ()
#3  0x0921e341 in sem_elab.record_elaboration_scenario ()
#4  0x092794b6 in sem_res.resolve_call ()
#5  0x0926937d in sem_res.resolve ()
#6  0x091c1287 in sem_ch6.analyze_procedure_call ()
#7  0x090f7b9a in sem.analyze ()
#8  0x091b96e9 in sem_ch5.analyze_statements ()
#9  0x091b98b9 in sem_ch5.analyze_if_statement.analyze_cond_then ()
#10 0x091badac in sem_ch5.analyze_if_statement ()
#11 0x090f7e0e in sem.analyze ()
#12 0x091b96e9 in sem_ch5.analyze_statements ()
#13 0x0912c037 in sem_ch11.analyze_handled_statements ()
#14 0x090f7a9b in sem.analyze ()
#15 0x091ce403 in sem_ch6.analyze_subprogram_body ()
#16 0x090f7714 in sem.analyze ()
#17 0x09179713 in sem_ch3.analyze_declarations ()
#18 0x091d7109 in sem_ch7.analyze_package_body ()
#19 0x090f7725 in sem.analyze ()
#20 0x09142d6b in sem_ch12.instantiate_package_body ()
#21 0x090758e6 in inline.instantiate_bodies.instantiate_body ()
#22 0x0907e30e in inline.instantiate_bodies ()
#23 0x0906e341 in frontend ()
#24 0x0932883d in gnat1drv ()
#25 0x08e7cbd2 in gnat_parse_file() ()
#26 0x0999df1f in compile_file() ()
#27 0x099a1427 in toplev::main(int, char**) ()
#28 0x0ae228c1 in main ()

Probably needs a reghunt.

[Bug ada/115106] [15 regression] SEGV in sem_elab.internal_representation.nts_map.mutate_and_rehash

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115106

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |15.0
           Keywords|                            |build, ice-on-valid-code

[Bug ada/115106] [15 regression] SEGV in sem_elab.internal_representation.nts_map.mutate_and_rehash

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115106

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

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

--- Comment #1 from Rainer Orth <ro at gcc dot gnu.org> ---
The reghunt identified

commit 9b7cad5884f21cc5783075be0043777448db3fab
Author: Jan Hubicka <jh@suse.cz>
Date:   Wed May 15 14:14:27 2024 +0200

    Avoid pointer compares on TYPE_MAIN_VARIANT in TBAA

FWIW, none of amd64-pc-solaris2.11, i686-pc-linux-gnu, and x86_64-pc-linux-gnu
show the failure.

[Bug ada/115106] [15 regression] SEGV in sem_elab.internal_representation.nts_map.mutate_and_rehash

[ebotcazou at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115106

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2024-05-16
                 CC|                            |ebotcazou at gcc dot gnu.org

--- Comment #2 from Eric Botcazou <ebotcazou at gcc dot gnu.org> ---
> The reghunt identified
> 
> commit 9b7cad5884f21cc5783075be0043777448db3fab
> Author: Jan Hubicka <jh@suse.cz>
> Date:   Wed May 15 14:14:27 2024 +0200
> 
>     Avoid pointer compares on TYPE_MAIN_VARIANT in TBAA

Thanks for tracking this down!

> FWIW, none of amd64-pc-solaris2.11, i686-pc-linux-gnu, and
> x86_64-pc-linux-gnu show the failure.

See PR ada/115113 though.

[Bug ada/115106] [15 regression] SEGV in sem_elab.internal_representation.nts_map.mutate_and_rehash

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115106

Iain Sandoe <iains at gcc dot gnu.org> changed:

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

--- Comment #3 from Iain Sandoe <iains at gcc dot gnu.org> ---
*** Bug 115113 has been marked as a duplicate of this bug. ***

[Bug ada/115106] [15 regression] SEGV in sem_elab.internal_representation.nts_map.mutate_and_rehash

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115106

Iain Sandoe <iains at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
              Build|i386-pc-solaris2.11         |i386-pc-solaris2.11,
                   |                            |i686-darwin
               Host|i386-pc-solaris2.11         |i386-pc-solaris2.11,
                   |                            |i686-darwin
             Target|i386-pc-solaris2.11         |i386-pc-solaris2.11,
                   |                            |i686-darwin

--- Comment #4 from Iain Sandoe <iains at gcc dot gnu.org> ---
likewise, x86_64 seems unaffected on the same OS versions (even on the same
h/w).

[Bug ada/115106] [15 regression] SEGV in sem_elab.internal_representation.nts_map.mutate_and_rehash

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115106

--- Comment #5 from Iain Sandoe <iains at gcc dot gnu.org> ---
as of r15-644, Ada bootstrap succeeded on i686-darwin9 and 17.

I do not known whether that means the issue is actually fixed by recent Ada
commits, or that it's now just become latent.

[Bug ada/115106] [15 regression] SEGV in sem_elab.internal_representation.nts_map.mutate_and_rehash

[ebotcazou at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115106

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

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

--- Comment #6 from Eric Botcazou <ebotcazou at gcc dot gnu.org> ---
> as of r15-644, Ada bootstrap succeeded on i686-darwin9 and 17.

Great!

> I do not known whether that means the issue is actually fixed by recent Ada
> commits, or that it's now just become latent.

Ada commits have nothing to do with it though, the breakage and the probable
fix both came from recent optimization changes.

[Bug ada/115106] [15 regression] SEGV in sem_elab.internal_representation.nts_map.mutate_and_rehash

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115106

--- Comment #7 from Iain Sandoe <iains at gcc dot gnu.org> ---
I have not tested on solaris - hopefully that is OK too.

However, I will comment that it maybe built but there are cats regressions (1)
on x86_64, (2) on i686-darwin17 (many) on i686-darwin9.  No idea what caused
those at the moment - and my hardware is very tied up with the releases.

[Bug ada/115106] [15 regression] SEGV in sem_elab.internal_representation.nts_map.mutate_and_rehash

[ebotcazou at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115106

--- Comment #8 from Eric Botcazou <ebotcazou at gcc dot gnu.org> ---
> However, I will comment that it maybe built but there are cats regressions
> (1) on x86_64, (2) on i686-darwin17 (many) on i686-darwin9.  No idea what
> caused those at the moment - and my hardware is very tied up with the
> releases.

The ACATS regression on x86-64 (cxa4001) is an issue in the test itself, not in
the compiler.  Darwin-specific regressions might come from Ada changes though.

[Bug ada/115106] [15 regression] SEGV in sem_elab.internal_representation.nts_map.mutate_and_rehash

[ro at CeBiTec dot Uni-Bielefeld.DE]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115106

--- Comment #9 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot Uni-Bielefeld.DE> ---
> --- Comment #6 from Eric Botcazou <ebotcazou at gcc dot gnu.org> ---
>> as of r15-644, Ada bootstrap succeeded on i686-darwin9 and 17.
>
> Great!

Same on i386-pc-solaris2.11.

>> I do not known whether that means the issue is actually fixed by recent Ada
>> commits, or that it's now just become latent.
>
> Ada commits have nothing to do with it though, the breakage and the probable
> fix both came from recent optimization changes.

I'm practically certain that this is due to Richard's recent fix for PR
middle-end/115110, so it's not latent.