[Bug ada/115305] New: [15 Regression] many (162) acats regressions on i686-darwin9.

[Bug ada/115305] New: [15 Regression] many (162) acats regressions on i686-darwin9.

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115305

            Bug ID: 115305
           Summary: [15 Regression] many (162) acats regressions on
                    i686-darwin9.
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: ada
          Assignee: unassigned at gcc dot gnu.org
          Reporter: iains at gcc dot gnu.org
                CC: dkm at gcc dot gnu.org
  Target Milestone: ---

Created attachment 58318
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=58318&action=edit
totals between r15-386-889

firstly, this is not a high priority target (even for me) - but I guess these
regressions do indicate some underlying omission or similar.

up to r15-386 darwin has clean acats results (both 32b and 64b hosts).

Unfortunately between r15-386 and 644 there was a bootstrap break for Ada on
i686-darwin, so the revision history is fragmented

I will attach diffs for the revision ranges I have; listing 162 fails here can
be done - but not sure it would be helpful.

Where we are testing an unpatched trunk/branch we are now pushing the results
to GCC's bunsen project:
https://builder.sourceware.org/testruns/?commitishes=&has_expfile_glob=&has_trsfile_glob=&has_keyvalue_k=testrun.git_describe&has_keyvalue_op=glob&has_keyvalue_v=iains%2F*&has_keyvalue2_k=omnitest.results&has_keyvalue2_op=glob&has_keyvalue2_v=*&order_by=testrun.authored.time&order=desc&limit=1000&offset=0&offset=0

the tag is :

 iains /<branch>/<platform>/<branch rev>-<build time>

so you can also poke at the detailed logs there if it's helpful.

===

There are regressions also on other earlier platform versions; I will also try
to identify them.

[Bug ada/115305] [15 Regression] many (162) acats regressions on i686-darwin9.

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115305

--- Comment #1 from Iain Sandoe <iains at gcc dot gnu.org> ---
Created attachment 58319
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=58319&action=edit
between 386 and 644

the results have been somewhat turbulent - so posting the ranges I have in case
it helps

[Bug ada/115305] [15 Regression] many (162) acats regressions on i686-darwin9.

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115305

--- Comment #2 from Iain Sandoe <iains at gcc dot gnu.org> ---
Created attachment 58320
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=58320&action=edit
between 644 and 792

[Bug ada/115305] [15 Regression] many (162) acats regressions on i686-darwin9.

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115305

--- Comment #3 from Iain Sandoe <iains at gcc dot gnu.org> ---
Created attachment 58321
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=58321&action=edit
between 792 and 856

likewise

[Bug ada/115305] [15 Regression] many (162) acats regressions on i686-darwin9.

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115305

--- Comment #4 from Iain Sandoe <iains at gcc dot gnu.org> ---
Created attachment 58322
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=58322&action=edit
between 856 and 889

likewise.

[Bug ada/115305] [15 Regression] many (162) acats regressions on i686-darwin9

[ebotcazou at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115305

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Version|14.0                        |15.0
             Status|UNCONFIRMED                 |NEW
            Summary|[15 Regression] many (162)  |[15 Regression] many (162)
                   |acats regressions on        |acats regressions on
                   |i686-darwin9.               |i686-darwin9
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2024-05-31
                 CC|                            |ebotcazou at gcc dot gnu.org

--- Comment #5 from Eric Botcazou <ebotcazou at gcc dot gnu.org> ---
Thanks for creating the PR.

[Bug ada/115305] [15 Regression] many (162) acats regressions on i686-darwin9

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115305

--- Comment #6 from Iain Sandoe <iains at gcc dot gnu.org> ---
a quick look at current results for other platform versions suggests that:

 * x86_64-darwin10 [64b runtimes, but 32b kernel] is also affected but
 * from darwin11+ [64 kernel] the main fail is FAIL:   cxa4001 which is a known
issue.

my darwin10 box is v. slow (and I lost a bunch of runs on Weds from a power
outage) but will post the 889 results for Darwin10 when it's done.

[Bug ada/115305] [15 Regression] many (162) acats regressions on i686-darwin9

[ebotcazou at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115305

--- Comment #7 from Eric Botcazou <ebotcazou at gcc dot gnu.org> ---
They might come from https://gcc.gnu.org/cgi-bin/gcc-gitref.cgi?r=r15-615 and,
in particular, the change made to libgnarl/s-osinte__darwin.ads, in which case
the way out would be to duplicate libgnat/s-oslock__posix.ads into
libgnat/s-oslock__darwin.ads, add back the sig field and remove the alignment
clause there, and then do the substitution on line 2749 of Makefile.rtl.

[Bug ada/115305] [15 Regression] many (162) acats regressions on i686-darwin9

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115305

--- Comment #8 from Iain Sandoe <iains at gcc dot gnu.org> ---
(In reply to Eric Botcazou from comment #7)
> They might come from https://gcc.gnu.org/cgi-bin/gcc-gitref.cgi?r=r15-615
> and, in particular, the change made to libgnarl/s-osinte__darwin.ads, in
> which case the way out would be to duplicate libgnat/s-oslock__posix.ads
> into libgnat/s-oslock__darwin.ads, add back the sig field and remove the
> alignment clause there, and then do the substitution on line 2749 of
> Makefile.rtl.

thanks for the pointer - if I have a chance, will take a look over the weekend.

[Bug ada/115305] [15 Regression] many (162) acats regressions on i686-darwin9

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115305

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Target|                            |i686-darwin9
   Target Milestone|---                         |15.0

[Bug ada/115305] [15 Regression] many (162) acats regressions on i686-darwin9

[dkm at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115305

--- Comment #9 from Marc Poulhiès <dkm at gcc dot gnu.org> ---
cxa4001 should be fixed since
https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=e0ab5ee9bed5cbad9ae344a23ff0d302b8279d32