[Bug middle-end/113205] New: [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346

[Bug middle-end/113205] New: [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346

[doko at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

            Bug ID: 113205
           Summary: [14 Regression] internal compiler error: in
                    backward_pass, at tree-vect-slp.cc:5346
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: middle-end
          Assignee: unassigned at gcc dot gnu.org
          Reporter: doko at gcc dot gnu.org
  Target Milestone: ---

seen with trunk 20240101 on x86_64-linux-gnu, building the mlt package. It
claims to be not reproducible, but always fails to build.

[ 31%] Building C object src/modules/gdk/CMakeFiles/mltgdk.dir/pixops.c.o
cd /home/packages/gcc/tmp/mlt-7.22.0/obj-x86_64-linux-gnu/src/modules/gdk &&
/usr/bin/cc -DARCH_X86_64 -DUSE_EXIF -DUSE_MMX -DUSE_PANGO -DUSE_PIXBUF
-Dmltgdk_EXPORTS -I/home/packages/gcc/tmp/mlt-7.22.0/src/framework/.. -isystem
/usr/include/gdk-pixbuf-2.0 -isystem /usr/include/glib-2.0 -isystem
/usr/lib/x86_64-linux-gnu/glib-2.0/include -isystem /usr/include/libpng16
-isystem /usr/include/webp -isystem /usr/include/libmount -isystem
/usr/include/blkid -isystem /usr/include/pango-1.0 -isystem
/usr/include/harfbuzz -isystem /usr/include/freetype2 -isystem
/usr/include/fribidi -isystem /usr/include/cairo -isystem /usr/include/pixman-1
-g -O2 -fno-omit-frame-pointer -mno-omit-leaf-frame-pointer
-ffile-prefix-map=/home/packages/gcc/tmp/mlt-7.22.0=. -flto=auto
-ffat-lto-objects -fstack-protector-strong -fstack-clash-protection -Wformat
-Werror=format-security -fcf-protection
-fdebug-prefix-map=/home/packages/gcc/tmp/mlt-7.22.0=/usr/src/mlt-7.22.0-1
-Wdate-time -D_FORTIFY_SOURCE=3 -std=gnu11 -fPIC -mmmx -msse -msse2 -pthread
-MD -MT src/modules/gdk/CMakeFiles/mltgdk.dir/pixops.c.o -MF
CMakeFiles/mltgdk.dir/pixops.c.o.d -o CMakeFiles/mltgdk.dir/pixops.c.o -c
/home/packages/gcc/tmp/mlt-7.22.0/src/modules/gdk/pixops.c
during GIMPLE pass: slp
/home/packages/gcc/tmp/mlt-7.22.0/src/modules/gdk/pixops.c: In function
'scale_line_22_yuv':
/home/packages/gcc/tmp/mlt-7.22.0/src/modules/gdk/pixops.c:188:1: internal
compiler error: in backward_pass, at tree-vect-slp.cc:5346
  188 | scale_line_22_yuv ( int *weights, int n_x, int n_y,
      | ^~~~~~~~~~~~~~~~~
0xff857e vect_optimize_slp_pass::backward_pass()
        ../../src/gcc/tree-vect-slp.cc:5346
0x21d18aa vect_optimize_slp_pass::run()
        ../../src/gcc/tree-vect-slp.cc:5768
0x1fa5f52 vect_optimize_slp(vec_info*)
        ../../src/gcc/tree-vect-slp.cc:5787
0x1fa5f52 vect_optimize_slp(vec_info*)
        ../../src/gcc/tree-vect-slp.cc:5783
0x1de839d vect_slp_analyze_bb_1
        ../../src/gcc/tree-vect-slp.cc:7579
0x1de839d vect_slp_region
        ../../src/gcc/tree-vect-slp.cc:7667
0x1de44d9 vect_slp_bbs
        ../../src/gcc/tree-vect-slp.cc:7878
0x1de1ccd vect_slp_function(function*)
        ../../src/gcc/tree-vect-slp.cc:8000
0x1de18dc execute
        ../../src/gcc/tree-vectorizer.cc:1533
Please submit a full bug report, with preprocessed source (by using
-freport-bug).
Please include the complete backtrace with any bug report.
See <file:///usr/share/doc/gcc-14/README.Bugs> for instructions.
The bug is not reproducible, so it is likely a hardware or OS problem.
make[3]: *** [src/modules/gdk/CMakeFiles/mltgdk.dir/build.make:107:
src/modules/gdk/CMakeFiles/mltgdk.dir/pixops.c.o] Error 1

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346

[doko at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

--- Comment #1 from Matthias Klose <doko at gcc dot gnu.org> ---
also in package qt6-quick3dphysics

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

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

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

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Can you reproduce it without -flto?

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |needs-bisection,
                   |                            |needs-reduction
   Target Milestone|---                         |14.0

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
           Keywords|needs-bisection,            |
                   |needs-reduction             |
            Summary|[14 Regression] internal    |[14 Regression] internal
                   |compiler error: in          |compiler error: in
                   |backward_pass, at           |backward_pass, at
                   |tree-vect-slp.cc:5346       |tree-vect-slp.cc:5346 since
                   |                            |r14-3220
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2024-01-08

--- Comment #3 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
char a;
char *b, *c;
int d, e, f, g, h;
int *i;

void
foo (void)
{
  unsigned p;
  d = i[0];
  e = i[1];
  f = i[2];
  g = i[3];
  p = d * b[0];
  p += f * c[h];
  p += e * b[h];
  p += g * c[h];
  a = (p + 8000) >> (__SIZEOF_INT__ * __CHAR_BIT__ / 2);
}
is reduced from mlt, ICEs at -O2 since
r14-3220-gf049868d8db773da4be61d82025e97453d6e9f2b.

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

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

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

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |rguenth at gcc dot gnu.org,
                   |                            |rsandifo at gcc dot gnu.org
           See Also|                            |https://gcc.gnu.org/bugzill
                   |                            |a/show_bug.cgi?id=110935

--- Comment #4 from Richard Biener <rguenth at gcc dot gnu.org> ---
OK, so this should already reproduce before the change when removing the
invariant add (p + 8000).  The issue seems to be that SLP build ends up
with an unsupported load permutation when we try with V2SImode vectorization
after V4SImode is scrapped because of cost issues.  We have

t.c:18:10: note:   node 0x6471a48 (max_nunits=2, refcnt=2) vector(2) int
t.c:18:10: note:   op template: _3 = MEM[(int *)i.0_1 + 4B];
t.c:18:10: note:        stmt 0 _3 = MEM[(int *)i.0_1 + 4B];
t.c:18:10: note:        stmt 1 _5 = MEM[(int *)i.0_1 + 12B];
t.c:18:10: note:        stmt 2 _4 = MEM[(int *)i.0_1 + 8B];
t.c:18:10: note:        stmt 3 _2 = *i.0_1;
t.c:18:10: note:        load permutation { 1 3 2 0 }

I'm not sure whether that's a supported situation.  Changing the code
to be more graceful like

diff --git a/gcc/tree-vect-slp.cc b/gcc/tree-vect-slp.cc
index b6cce55ce90..a12214bc1ad 100644
--- a/gcc/tree-vect-slp.cc
+++ b/gcc/tree-vect-slp.cc
@@ -5343,8 +5343,8 @@ vect_optimize_slp_pass::backward_pass ()
            }
        }

-      gcc_assert (min_layout_cost.is_possible ());
-      partition.layout = min_layout_i;
+      if (min_layout_cost.is_possible ())
+       partition.layout = min_layout_i;
     }
 }

then yields

t.c:18:10: note:  SLP optimize permutations:
t.c:18:10: note:    1: { 1, 3, 2, 0 }
t.c:18:10: note:  SLP optimize partitions:
t.c:18:10: note:    -------------
t.c:18:10: note:    partition 0 (layout 0):
t.c:18:10: note:      nodes:
t.c:18:10: note:        - 0x5f0d9b0:
t.c:18:10: note:            weight: 1.000000
t.c:18:10: note:            out weight: 1.000000 (degree 1)
t.c:18:10: note:            op template: _20 = (int) _19;
t.c:18:10: note:      edges:
t.c:18:10: note:        - 0x5f0d9b0 --> [2] 0x5f0d928
t.c:18:10: note:      layout 0: rejected
t.c:18:10: note:      layout 1: rejected
t.c:18:10: note:    -------------
t.c:18:10: note:    partition 1 (layout 1):
t.c:18:10: note:      nodes:
t.c:18:10: note:        - 0x5f0da38:
t.c:18:10: note:            weight: 1.000000
t.c:18:10: note:            out weight: 1.000000 (degree 1)
t.c:18:10: note:            op template: _3 = MEM[(int *)i.0_1 + 4B];
t.c:18:10: note:      edges:
t.c:18:10: note:        - 0x5f0da38 --> [2] 0x5f0d928
t.c:18:10: note:      layout 0: rejected
t.c:18:10: note:      layout 1: rejected
t.c:18:10: note:    -------------
t.c:18:10: note:    partition 2 (layout 1):
t.c:18:10: note:      nodes:
t.c:18:10: note:        - 0x5f0d928:
t.c:18:10: note:            weight: 1.000000
t.c:18:10: note:            out weight: 1.000000 (degree 1)
t.c:18:10: note:            op template: _21 = _3 * _20;
t.c:18:10: note:      edges:
t.c:18:10: note:        - 0x5f0d928 --> [3] 0x5f0d8a0
t.c:18:10: note:        - 0x5f0d9b0 [0] --> 0x5f0d928
t.c:18:10: note:        - 0x5f0da38 [1] --> 0x5f0d928
t.c:18:10: note:      layout 0: rejected
t.c:18:10: note:      layout 1: rejected
t.c:18:10: note:    -------------
t.c:18:10: note:    partition 3 (layout 1):
t.c:18:10: note:      nodes:
t.c:18:10: note:        - 0x5f0d8a0:
t.c:18:10: note:            weight: 1.000000
t.c:18:10: note:            op template: _22 = (unsigned int) _21;
t.c:18:10: note:      edges:
t.c:18:10: note:        - 0x5f0d928 [2] --> 0x5f0d8a0
t.c:18:10: note:      layout 0:
t.c:18:10: note:          {depth: 1.000000, total: 1.000000}
t.c:18:10: note:        + {depth: 0.000000, total: 0.000000}
t.c:18:10: note:        + {depth: 0.000000, total: 0.000000}
t.c:18:10: note:        = {depth: 1.000000, total: 1.000000}
t.c:18:10: note:      layout 1: (*)
t.c:18:10: note:          {depth: 0.000000, total: 0.000000}
t.c:18:10: note:        + {depth: 0.000000, total: 0.000000}
t.c:18:10: note:        + {depth: 0.000000, total: 0.000000}
t.c:18:10: note:        = {depth: 0.000000, total: 0.000000}
t.c:18:10: note:  inserting permutation node in place of 0x5f0d9b0
t.c:18:10: note:  recording new base alignment for i.0_1
...
t.c:18:10: note:   vectorizing permutation op0[3] op0[0] op0[2] op0[1]
t.c:18:10: note:   vectorizing permutation op0[3] op0[0] op0[2] op0[1]
t.c:18:10: note:   as vops0[1][1] vops0[0][0], vops0[1][0] vops0[0][1]
t.c:18:10: missed:   unsupported vect permute { 1 2 }
t.c:18:10: note:   Building vector operands of 0x5f0db48 from scalars instead
...
t.c:18:10: note:   removing SLP instance operations starting from: _25 = _24 +
_40;
t.c:18:10: missed:  not vectorized: bad operation in basic block.
t.c:18:10: note: ***** Analysis failed with vector mode V8QI
t.c:18:10: note: ***** Re-trying analysis with vector mode V4QI

and the ICE is gone.

I'm not sure if we can "recover" in this way or whether leaving
partition.layout unchanged could lead to wrong-code if it were actually
possible to code generate it, thus whether it's really the inability
to generate the permute that triggers this issue.

Related to PR110935, with -Ofast we should elide the unsupported permute.

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[sergio at serjux dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

Sérgio Basto <sergio at serjux dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sergio at serjux dot com

--- Comment #5 from Sérgio Basto <sergio at serjux dot com> ---
(In reply to Matthias Klose from comment #0)
> seen with trunk 20240101 on x86_64-linux-gnu, building the mlt package. It
> claims to be not reproducible, but always fails to build.

I notice that mlt just fail to build on x86_64
https://koji.fedoraproject.org/koji/taskinfo?taskID=112893731

HTH

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[sergio at serjux dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

--- Comment #6 from Sérgio Basto <sergio at serjux dot com> ---
and any workaround to build mlt on x86_64 ,exist ? 

I tried disable lto, hardening and also disable strict symbols (-Wl,-z,defs )
and still not build neither in koji nor in my machine

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

--- Comment #7 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
-fno-tree-slp-vectorize, either on the command line for the affected
compilation unit or __attribute__((optimize ("no-tree-slp-vectorize"))) on the
problematic routine.

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

--- Comment #7 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
-fno-tree-slp-vectorize, either on the command line for the affected
compilation unit or __attribute__((optimize ("no-tree-slp-vectorize"))) on the
problematic routine.

--- Comment #8 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Sérgio Basto from comment #6)
> and any workaround to build mlt on x86_64 ,exist ? 
> 
> I tried disable lto, hardening and also disable strict symbols (-Wl,-z,defs
> ) and still not build neither in koji nor in my machine

`-fno-tree-slp-vectorize` should work around it ...
but that only works around the issue on the trunk which will be fixed by the
time release happens.

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[sergio at serjux dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

--- Comment #9 from Sérgio Basto <sergio at serjux dot com> ---
Thank you it worked , MLT was built successfully on Fedora Rawhide

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

--- Comment #10 from Richard Biener <rguenth at gcc dot gnu.org> ---
Btw, I was hoping Richard would chime in here ...

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

--- Comment #11 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Richard S., any comments on this?

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[rsandifo at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

Richard Sandiford <rsandifo at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at gcc dot gnu.org      |rsandifo at gcc dot gnu.org
             Status|NEW                         |ASSIGNED

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[rsandifo at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

--- Comment #12 from Richard Sandiford <rsandifo at gcc dot gnu.org> ---
Created attachment 57511
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=57511&action=edit
Candidate patch

Sorry for the very slow response on this.  I'm testing the attached.

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

--- Comment #13 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The trunk branch has been updated by Richard Sandiford <rsandifo@gcc.gnu.org>:

https://gcc.gnu.org/g:0394ae31e832c5303f3b4aad9c66710a30c097f0

commit r14-9165-g0394ae31e832c5303f3b4aad9c66710a30c097f0
Author: Richard Sandiford <richard.sandiford@arm.com>
Date:   Sat Feb 24 11:58:22 2024 +0000

    vect: Tighten check for impossible SLP layouts [PR113205]

    During its forward pass, the SLP layout code tries to calculate
    the cost of a layout change on an incoming edge.  This is taken
    as the minimum of two costs: one in which the source partition
    keeps its current layout (chosen earlier during the pass) and
    one in which the source partition switches to the new layout.
    The latter can sometimes be arranged by the backward pass.

    If only one of the costs is valid, the other cost was ignored.
    But the PR shows that this is not safe.  If the source partition
    has layout 0 (the normal layout), we have to be prepared to handle
    the case in which that ends up being the only valid layout.

    Other code already accounts for this restriction, e.g. see
    the code starting with:

        /* Reject the layout if it would make layout 0 impossible
           for later partitions.  This amounts to testing that the
           target supports reversing the layout change on edges
           to later partitions.

    gcc/
            PR tree-optimization/113205
            * tree-vect-slp.cc (vect_optimize_slp_pass::forward_cost): Reject
            the proposed layout if it does not allow a source partition with
            layout 2 to keep that layout.

    gcc/testsuite/
            PR tree-optimization/113205
            * gcc.dg/torture/pr113205.c: New test.

[Bug middle-end/113205] [14 Regression] internal compiler error: in backward_pass, at tree-vect-slp.cc:5346 since r14-3220

[rsandifo at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113205

Richard Sandiford <rsandifo at gcc dot gnu.org> changed:

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

--- Comment #14 from Richard Sandiford <rsandifo at gcc dot gnu.org> ---
Finally fixed.