[Bug target/103370] New: [12 Regression] Assembler error building glibc for ColdFire soft-float

[Bug target/103370] New: [12 Regression] Assembler error building glibc for ColdFire soft-float

[jsm28 at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103370

            Bug ID: 103370
           Summary: [12 Regression] Assembler error building glibc for
                    ColdFire soft-float
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Keywords: assemble-failure
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: jsm28 at gcc dot gnu.org
                CC: rguenth at gcc dot gnu.org
  Target Milestone: ---
            Target: m68k*-*-*

Created attachment 51849
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=51849&action=edit
preprocessed source

Compile and assemble the attached preprocessed source (from glibc) with -O2
-fPIC -c, using a compiler configured for m68k-linux-gnu --with-arch=cf
--with-cpu=54455 --disable-multilib (this is m68k-linux-gnu-coldfire-soft in
build-many-glibcs.py).  This produces an assembler error:

/tmp/cctaeK7A.s: Assembler messages:
/tmp/cctaeK7A.s: Fatal error: Tried to convert PC relative branch to absolute
jump

This appeared with

commit 045206450386bcd774db3bde0c696828402361c6
Author: Richard Biener <rguenther@suse.de>
Date:   Fri Nov 12 10:21:22 2021 +0100

    tree-optimization/102880 - improve CD-DCE

though I'm guessing it was most likely to have been latent before then.

[Bug target/103370] [12 Regression] Assembler error building glibc for ColdFire soft-float

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103370

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
  if (!HAVE_LONG_BRANCH (current_architecture))
    {
      if (flag_keep_pcrel)
        as_fatal (_("Tried to convert PC relative branch to absolute jump"));


#define HAVE_LONG_BRANCH(x)     \
        ((x) & (m68020|m68030|m68040|m68060|cpu32|fido_a|mcfisa_b))

[Bug target/103370] [12 Regression] Assembler error building glibc for ColdFire soft-float

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103370

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
    case OPTION_PCREL:          /* --pcrel means never turn PC-relative
                                   branches into absolute jumps.  */
      flag_keep_pcrel = 1;
      break;


--pcrel is passed to the assembler for -fPIC/-fpic code.

[Bug target/103370] [12 Regression] Assembler error building glibc for ColdFire soft-float

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103370

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

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

[Bug target/103370] [12 Regression] Assembler error building glibc for ColdFire soft-float

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103370

--- Comment #3 from Martin Liška <marxin at gcc dot gnu.org> ---
Created attachment 51857
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=51857&action=edit
Reduced test-case

[Bug target/103370] [12 Regression] Assembler error building glibc for ColdFire soft-float

[jsm28 at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103370

--- Comment #4 from Joseph S. Myers <jsm28 at gcc dot gnu.org> ---
It seems this glibc build failure disappeared between commit
b880d1514c1e3dd75a6ea311a5cc956742bd713c and commit
8e836af61b7027c0819da62c12a8d18b7c46f3fc.  (Not verified that the GCC change
was what's relevant - my bot updated other components as well - but it looks
plausible.)

[Bug target/103370] [12 Regression] Assembler error building glibc for ColdFire soft-float

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103370

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P4

--- Comment #5 from Richard Biener <rguenth at gcc dot gnu.org> ---
The issue likely has gone latent.  m68k is not primary/secondary.

[Bug target/103370] [12/13 Regression] Assembler error building glibc for ColdFire soft-float

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103370

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|12.0                        |12.2

--- Comment #6 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
GCC 12.1 is being released, retargeting bugs to GCC 12.2.

[Bug target/103370] [12/13 Regression] Assembler error building glibc for ColdFire soft-float

[jsm28 at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103370

--- Comment #9 from Joseph S. Myers <jsm28 at gcc dot gnu.org> ---
The glibc build failure has done away and come back at least once since my
previous comment. It came back (I think the most recent time) with

commit 4fa25a7eb322f0a003c1eb15680c71ece345e01e
Author:     Martin Liska <mliska@suse.cz>
AuthorDate: Mon Jan 24 15:45:38 2022 +0100
Commit:     Martin Liska <mliska@suse.cz>
CommitDate: Wed Nov 30 14:03:43 2022 +0100

    Improve profile handling in switch lowering.

            PR tree-optimization/101301
            PR tree-optimization/103680

though that's almost surely just exposing a back-end bug. However, the reduced
test here gave a different assembler error

/tmp/cc40cNSh.s: Assembler messages:
/tmp/cc40cNSh.s:320: Error: syntax error -- statement `lea
(.LC0@GOT+3,%a5),%a0' ignored
/tmp/cc40cNSh.s:328: Error: syntax error -- statement `lea
(.LC1@GOT+3,%a5),%a2' ignored

both before and after that commit (different register numbers and .s line
numbers before versus after). Despite the different error messages from the
reduced test and building glibc, they are still likely to be the same bug in
the m68k back end.

[Bug target/103370] [12/13/14 Regression] Assembler error building glibc for ColdFire soft-float

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103370

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|12.3                        |12.4

--- Comment #10 from Richard Biener <rguenth at gcc dot gnu.org> ---
GCC 12.3 is being released, retargeting bugs to GCC 12.4.

[Bug target/103370] [12/13/14 Regression] Assembler error building glibc for ColdFire soft-float

[fw at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103370

Florian Weimer <fw at gcc dot gnu.org> changed:

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

--- Comment #11 from Florian Weimer <fw at gcc dot gnu.org> ---
*** Bug 110627 has been marked as a duplicate of this bug. ***