[Bug lto/97508] New: lto1: internal compiler error: decompressed stream: Destination buffer is too small

[Bug lto/97508] New: lto1: internal compiler error: decompressed stream: Destination buffer is too small

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97508

            Bug ID: 97508
           Summary: lto1: internal compiler error: decompressed stream:
                    Destination buffer is too small
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: lto
          Assignee: unassigned at gcc dot gnu.org
          Reporter: hjl.tools at gmail dot com
                CC: marxin at gcc dot gnu.org
  Target Milestone: ---

[hjl@gnu-skx-1 tmp]$ cat pr15323a.c
int main (void)
{
  return 0;
}
[hjl@gnu-skx-1 tmp]$ cat doit
CFLAGS="-flto -fno-profile-use -O2"
cc $CFLAGS -c pr15323a.c -o pr15323a.o
cc $CFLAGS -r -nostdlib pr15323a.o -o pr15323a-r.o
cc $CFLAGS  -o pr15323a.exe pr15323a-r.o
[hjl@gnu-skx-1 tmp]$ sh doit
during IPA pass: cp
lto1: internal compiler error: decompressed stream: Destination buffer is too
small
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://bugzilla.redhat.com/bugzilla> for instructions.
lto-wrapper: fatal error: cc returned 1 exit status
compilation terminated.
/usr/local/bin/ld: error: lto-wrapper failed
collect2: error: ld returned 1 exit status
[hjl@gnu-skx-1 tmp]$

[Bug lto/97508] lto1: internal compiler error: decompressed stream: Destination buffer is too small

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97508

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2020-10-21
           Keywords|                            |ice-on-valid-code, lto
             Status|UNCONFIRMED                 |WAITING
     Ever confirmed|0                           |1

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
It works for me.  What compression scheme are you using?  I have

Supported LTO compression algorithms: zlib

are you using zstd, if so, which version?

[Bug lto/97508] lto1: internal compiler error: decompressed stream: Destination buffer is too small

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97508

Martin Liška <marxin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at gcc dot gnu.org      |marxin at gcc dot gnu.org
             Status|WAITING                     |ASSIGNED

--- Comment #2 from Martin Liška <marxin at gcc dot gnu.org> ---
I can reproduce it, working on that..

[Bug lto/97508] [10/11 Regression] ICE: decompressed stream: Destination buffer is too small

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97508

Martin Liška <marxin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|lto1: internal compiler     |[10/11 Regression] ICE:
                   |error: decompressed stream: |decompressed stream:
                   |Destination buffer is too   |Destination buffer is too
                   |small                       |small

--- Comment #3 from Martin Liška <marxin at gcc dot gnu.org> ---
Started with my r10-1548-g87741e51b53511be.
It's caused by the fact that .gnu.lto_.jmpfuncs.3f272ba23409d5d4 is streamed
twise: from "cp" pass and then from "fnsummary" pass.

[Bug lto/97508] [10/11 Regression] ICE: decompressed stream: Destination buffer is too small

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97508

--- Comment #4 from Martin Liška <marxin at gcc dot gnu.org> ---
Then we end up with 2 copies of the compressed stream:

$ cat good.txt

y.o:     file format elf64-x86-64

Contents of section .gnu.lto_.jmpfuncs.c09bc76fb7147437:
 0000 28b52ffd 20116d00 00380500 01000000  (./. .m..8......
 0010 0001001b c002                        ......

$ cat bad.txt

y.o:     file format elf64-x86-64

Contents of section .gnu.lto_.jmpfuncs.9e830355985c8f8c:
 0000 28b52ffd 20116d00 00380500 01000000  (./. .m..8......
 0010 0001001b c00228b5 2ffd2011 6d000038  ......(./. .m..8
 0020 05000100 00000001 001bc002           ............    

Then we tell zstd section size and decompression fails as zstd frame header
contains a smaller size.

[Bug lto/97508] [10/11 Regression] ICE: decompressed stream: Destination buffer is too small

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97508

--- Comment #5 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Martin Liska <marxin@gcc.gnu.org>:

https://gcc.gnu.org/g:568de14d2e74cfdd600b8995ff6ac08c98ddef48

commit r11-4526-g568de14d2e74cfdd600b8995ff6ac08c98ddef48
Author: Martin Liska <mliska@suse.cz>
Date:   Wed Oct 21 11:11:03 2020 +0200

    LTO: get_section: add new argument

    gcc/ChangeLog:

            PR lto/97508
            * langhooks.c (lhd_begin_section): Call get_section with
            not_existing = true.
            * output.h (get_section): Add new argument.
            * varasm.c (get_section): Fail when NOT_EXISTING is true
            and a section already exists.
            * ipa-cp.c (ipcp_write_summary): Remove.
            (ipcp_read_summary): Likewise.
            * ipa-fnsummary.c (ipa_fn_summary_read): Always read jump
            functions summary.
            (ipa_fn_summary_write): Always stream it.

[Bug lto/97508] [10 Regression] ICE: decompressed stream: Destination buffer is too small

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97508

Martin Liška <marxin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[10/11 Regression] ICE:     |[10 Regression] ICE:
                   |decompressed stream:        |decompressed stream:
                   |Destination buffer is too   |Destination buffer is too
                   |small                       |small
      Known to fail|11.0                        |
      Known to work|                            |11.0

--- Comment #6 from Martin Liška <marxin at gcc dot gnu.org> ---
Fixed on master so far.

[Bug lto/97508] [10 Regression] ICE: decompressed stream: Destination buffer is too small

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97508

--- Comment #7 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-10 branch has been updated by Martin Liska
<marxin@gcc.gnu.org>:

https://gcc.gnu.org/g:32d16af0c4ede8d008d4360cca7c805db518166f

commit r10-8954-g32d16af0c4ede8d008d4360cca7c805db518166f
Author: Martin Liska <mliska@suse.cz>
Date:   Wed Oct 21 11:11:03 2020 +0200

    LTO: get_section: add new argument

    gcc/ChangeLog:

            PR lto/97508
            * langhooks.c (lhd_begin_section): Call get_section with
            not_existing = true.
            * output.h (get_section): Add new argument.
            * varasm.c (get_section): Fail when NOT_EXISTING is true
            and a section already exists.
            * ipa-cp.c (ipcp_write_summary): Remove.
            (ipcp_read_summary): Likewise.
            * ipa-fnsummary.c (ipa_fn_summary_read): Always read jump
            functions summary.
            (ipa_fn_summary_write): Always stream it.

    (cherry picked from commit 568de14d2e74cfdd600b8995ff6ac08c98ddef48)

[Bug lto/97508] [10 Regression] ICE: decompressed stream: Destination buffer is too small

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97508

Martin Liška <marxin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|---                         |FIXED
      Known to fail|10.2.0                      |
      Known to work|                            |10.2.1

--- Comment #8 from Martin Liška <marxin at gcc dot gnu.org> ---
Fixed now.