[Bug c++/109629] New: internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[Bug c++/109629] New: internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[linus.zhu at mavenir dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

            Bug ID: 109629
           Summary: internal compiler error: in final_scan_insn_1, at
                    final.c:3078  6480 | }
           Product: gcc
           Version: 10.2.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: linus.zhu at mavenir dot com
  Target Milestone: ---

UeDtCmi.cpp: In member function ‘virtual cm_ims_rmt_server_t*
UeDtCmi::_ZThn21812712_N7UeDtCmi12GetImsRmtSrvEv()’:
UeDtCmi.cpp:6480:1: internal compiler error: in final_scan_insn_1, at
final.c:3078
 6480 | }
      | ^
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://bugzilla.redhat.com/bugzilla> for instructions.
{standard input}: Assembler messages:
{standard input}: Error: open CFI at the end of file; missing .cfi_endproc
directive
Preprocessed source stored into /tmp/ccdNSTWA.out file, please attach this to
your bugreport.
make[1]: *** [Makefile:619: libUeDtMgr_a-UeDtCmi.o] Error 1



similar issue was reported by aarch64 gcc 8.3.1/9.3.1/10.2.1,
and the code survived aarch64 gcc 4.8.5

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |ice-on-valid-code
          Component|c++                         |target

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Note this:
See <http://bugzilla.redhat.com/bugzilla> for instructions.

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |WAITING
   Last reconfirmed|                            |2023-04-26

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Also note:
Preprocessed source stored into /tmp/ccdNSTWA.out file, please attach this to
your bugreport.


You might need to compress it.

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[linus.zhu at mavenir dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #3 from Linus <linus.zhu at mavenir dot com> ---
Created attachment 54921
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=54921&action=edit
ccdNSTWA.out.1

ccdNSTWA.out.1

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[linus.zhu at mavenir dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #4 from Linus <linus.zhu at mavenir dot com> ---
due to 1M-attachment limit, I'll raise another bug and attach ccdNSTWA.out.2

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #5 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
/* If we have a length attribute, this instruction should have
   been split in shorten_branches, to ensure that we would have
   valid length info for the splitees.  */
gcc_assert (!HAVE_ATTR_length);

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #6 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
*** Bug 109630 has been marked as a duplicate of this bug. ***

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #7 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Linus from comment #4)
> due to 1M-attachment limit, I'll raise another bug and attach ccdNSTWA.out.2

The limit is just per-attachment, not per bug.

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[linus.zhu at mavenir dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #8 from Linus <linus.zhu at mavenir dot com> ---
hi Andrew, thanks. And before raising another ticket for attaching the second
half of ccdNSTWA.out, I did try to add one more attachment to this bug, but it
hinted that previous one will be obsolete.

Anyway it's not important, another part of ccdNSTWA.out was uploaded to
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109630.

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #9 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Linus from comment #8)
> hi Andrew, thanks. And before raising another ticket for attaching the
> second half of ccdNSTWA.out, I did try to add one more attachment to this
> bug, but it hinted that previous one will be obsolete.

No it will not. It says:
"(optional) Check each existing attachment made obsolete by your new
attachment."
See the optional part. It will only make it is obsolete if you chose to do
that.

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[linus.zhu at mavenir dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #10 from Linus <linus.zhu at mavenir dot com> ---
(In reply to Andrew Pinski from comment #5)
> /* If we have a length attribute, this instruction should have
>    been split in shorten_branches, to ensure that we would have
>    valid length info for the splitees.  */
> gcc_assert (!HAVE_ATTR_length);

Yes, when I checked source code of gcc 11, I saw them.

And I saw:
"
Tue Mar 18 16:12:28 1997  Jim Wilson  <wilson@cygnus.com>

        * final.c (shorten_branches): Split all insns before computing insn
        lengths.
        (final_scan_insn, case default): If HAVE_ATTR_length defined, call
        abort for any insn that has a '#' output template.

"


I just wait for your new response, right?
Thanks for checking the issue.

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[linus.zhu at mavenir dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #11 from Linus <linus.zhu at mavenir dot com> ---
(In reply to Andrew Pinski from comment #9)
> (In reply to Linus from comment #8)
> > hi Andrew, thanks. And before raising another ticket for attaching the
> > second half of ccdNSTWA.out, I did try to add one more attachment to this
> > bug, but it hinted that previous one will be obsolete.
> 
> No it will not. It says:
> "(optional) Check each existing attachment made obsolete by your new
> attachment."
> See the optional part. It will only make it is obsolete if you chose to do
> that.

Thanks for the clarification.

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #12 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Linus from comment #10)
> (In reply to Andrew Pinski from comment #5)
> > /* If we have a length attribute, this instruction should have
> >    been split in shorten_branches, to ensure that we would have
> >    valid length info for the splitees.  */
> > gcc_assert (!HAVE_ATTR_length);
> 
> Yes, when I checked source code of gcc 11, I saw them.

> I just wait for your new response, right?
> Thanks for checking the issue.

No, I was just pointing out where the ICE is happening. But really it is
something inside the target backend's md file. Still need to reduce/debug it to
figure out which pattern is not being split.

Also Please FILE this with redhat since this is a redhat modified compiler
first.

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[linus.zhu at mavenir dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #13 from Linus <linus.zhu at mavenir dot com> ---
(In reply to Andrew Pinski from comment #12)
> (In reply to Linus from comment #10)
> > (In reply to Andrew Pinski from comment #5)
> > > /* If we have a length attribute, this instruction should have
> > >    been split in shorten_branches, to ensure that we would have
> > >    valid length info for the splitees.  */
> > > gcc_assert (!HAVE_ATTR_length);
> > 
> > Yes, when I checked source code of gcc 11, I saw them.
> 
> > I just wait for your new response, right?
> > Thanks for checking the issue.
> 
> No, I was just pointing out where the ICE is happening. But really it is
> something inside the target backend's md file. Still need to reduce/debug it
> to figure out which pattern is not being split.
> 
> Also Please FILE this with redhat since this is a redhat modified compiler
> first.

Okay and thanks.
Is there a chance that this issue will be gone if I try the original gcc 9 or
gcc 10?
By the way, same code didn't survive aarch64 gcc 9.3.1, but did survive X86 gcc
9.3.1.

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[linus.zhu at mavenir dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #14 from Linus <linus.zhu at mavenir dot com> ---
devtoolset-9-gcc-9.3.1-2.2.el7.aarch64

devtoolset-9-gcc-9.3.1-2.el7.x86_64

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

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

--- Comment #15 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Already fixed.

*** This bug has been marked as a duplicate of bug 98136 ***

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[linus.zhu at mavenir dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #16 from Linus <linus.zhu at mavenir dot com> ---
(In reply to Andrew Pinski from comment #15)
> Already fixed.
> 
> *** This bug has been marked as a duplicate of bug 98136 ***

Thanks.

1. 98136 was fixed for gcc 9.2.0. So original gcc 10 includes the fix, right?
2. I still saw the issue with gcc 10 just because of the fix was not picked by
devtoolset-10-gcc-c++-10.2.1-11.2.el7.aarch64, right?

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #17 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Linus from comment #16)
> (In reply to Andrew Pinski from comment #15)
> > Already fixed.
> > 
> > *** This bug has been marked as a duplicate of bug 98136 ***
> 
> Thanks.
> 
> 1. 98136 was fixed for gcc 9.2.0. So original gcc 10 includes the fix, right?
> 2. I still saw the issue with gcc 10 just because of the fix was not picked
> by devtoolset-10-gcc-c++-10.2.1-11.2.el7.aarch64, right?

It was NOT fixed in GCC 9.2.0 but rather 9.4.0; it was reported against 9.2.0
though.

The versions which incldue the fix: 10.4.0, 11.1.0, 8.5.0, 9.4.0 (and above
versions, e.g. 12.1.0).

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[linus.zhu at mavenir dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #18 from Linus <linus.zhu at mavenir dot com> ---
(In reply to Andrew Pinski from comment #17)
> (In reply to Linus from comment #16)
> > (In reply to Andrew Pinski from comment #15)
> > > Already fixed.
> > > 
> > > *** This bug has been marked as a duplicate of bug 98136 ***
> > 
> > Thanks.
> > 
> > 1. 98136 was fixed for gcc 9.2.0. So original gcc 10 includes the fix, right?
> > 2. I still saw the issue with gcc 10 just because of the fix was not picked
> > by devtoolset-10-gcc-c++-10.2.1-11.2.el7.aarch64, right?
> 
> It was NOT fixed in GCC 9.2.0 but rather 9.4.0; it was reported against
> 9.2.0 though.
> 
> The versions which incldue the fix: 10.4.0, 11.1.0, 8.5.0, 9.4.0 (and above
> versions, e.g. 12.1.0).

Great thanks, Andrew, for your time and effort.

By the way, same code didn't survive aarch64 gcc 9.3.1, but did survive X86 gcc
9.3.1 -- devtoolset-9-gcc-9.3.1-2.el7.x86_64.
Do you have any clue about it?

Again, thanks for the great support!

[Bug target/109629] internal compiler error: in final_scan_insn_1, at final.c:3078 6480 | }

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109629

--- Comment #19 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Linus from comment #18)
> By the way, same code didn't survive aarch64 gcc 9.3.1, but did survive X86
> gcc 9.3.1 -- devtoolset-9-gcc-9.3.1-2.el7.x86_64.
> Do you have any clue about it?

Yes it is a target specific issue. Again if you want a fix for these
devtoolset, please talk with Redhat, we don't support them here.