Bug 953363 – sci-mathematics/lcalc-2.1.0 - [llvm] ld.lld: error: non-exported symbol _muldc3 in /.../libclang_rt.builtins-x86_64.a(muldc3.c.o) is referenced by DSO ../.../libLfunction

• Status: RESOLVED FIXED
• Product: Gentoo Linux

[Description Toralf Förster 2025-04-07 18:30:41 UTC]

too long lines were shrinked:

make[1]: Entering directory '/var/tmp/portage/sci-mathematics/lcalc-2.1.0/work/lcalc-2.1.0/src/lcalc'
clang++ -DHAVE_CONFIG_H -I. -I../../src  -I../../src/libLfunction   -O2 -pipe -march=native -fno-diagnostics-color -c -o Lcommandline.o Lcommandline.cc
make[1]: Leaving directory '/var/tmp/portage/sci-mathematics/lcalc-2.1.0/work/lcalc-2.1.0/src/lcalc'
make[1]: Entering directory '/var/tmp/portage/sci-mathematics/lcalc-2.1.0/work/lcalc-2.1.0/src/lcalc'
/bin/sh ../../libtool  --tag=CXX   --mode=link clang++  -O2 -pipe -march=native -fno-diagnostics-color  -Wl,-O1 -Wl,--as-needed -Wl,-z,pack-relative-relocs -Wl,--as-needed -Wl,--defsym=__gentoo_check_ldflags__=0 -o lcalc Lcommandline.o Lcommandline_globals.o Lcommandline_misc.o Lcommandline_twist.o 
libtool: link: clang++ -O2 -pipe -march=native -fno-diagnostics-color -Wl,-O1 -Wl,--as-needed -Wl,-z -Wl,pack-relative-relocs -Wl,--as-needed -Wl,--defsym=__gentoo_check_ldflags__=0 -o .libs/lcalc Lcommandline.o Lcommandline_globals.o Lcommandline_misc.o Lcommandline_twist.o Lcommandline_values_zero
ld.lld: error: non-exported symbol '__muldc3' in '/usr/lib/llvm/20/bin/../../../../lib/clang/20/lib/linux/libclang_rt.builtins-x86_64.a(muldc3.c.o)' is referenced by DSO '../../src/libLfunction/.libs/libLfunction.so'
ld.lld: error: non-exported symbol '__divdc3' in '/usr/lib/llvm/20/bin/../../../../lib/clang/20/lib/linux/libclang_rt.builtins-x86_64.a(divdc3.c.o)' is referenced by DSO '../../src/libLfunction/.libs/libLfunction.so'
clang++: error: linker command failed with exit code 1 (use -v to see invocation)

  -------------------------------------------------------------------

  This is an unstable amd64 chroot image at a tinderbox (==build bot)
  name: 23.0_llvm-20250403-002508

  Block bug #351559 if this looks like a parallel build issue.
  Possible reproducer: MAKEOPTS='...  shuffle=3353285003'

  The attached etc.portage.tar.xz has all details.
  -------------------------------------------------------------------

GNUMAKEFLAGS="$GNUMAKEFLAGS --shuffle"
gcc-config -l:
 [1] x86_64-pc-linux-gnu-14 *
clang version 20.1.2+libcxx
llvm-config: 20.1.2+libcxx
Python 3.12.9
go version go1.24.1 linux/amd64
  [1]   php8.2
  [2]   php8.4 *
Available Ruby profiles:
  [1]   ruby32 (with Rubygems)
  [2]   ruby33 (with Rubygems) *
Available Rust versions:
  [1]   rust-bin-1.85.1
  [2]   rust-1.85.1
  [3]   rust-bin-1.86
  [4]   rust-1.86 *
The following VMs are available for generation-2:
1)	Eclipse Temurin JDK 17.0.14_p7 [openjdk-bin-17]
*)	Eclipse Temurin JDK 21.0.6_p7 [openjdk-bin-21]
3)	Eclipse Temurin JDK 8.442_p06 [openjdk-bin-8]
Available Java Virtual Machines:
  [1]   openjdk-bin-8 
  [2]   openjdk-bin-17 
  [3]   openjdk-bin-21  system-vm


  HEAD of ::gentoo
commit 81186b52896e0d81c3ba341752f9223ac5c0ca03
Author: Repository mirror & CI <[e-mail]>
Date:   Sun Apr 6 23:20:18 2025 +0000

    2025-04-06 23:20:17 UTC

emerge -qpvO =sci-mathematics/lcalc-2.1.0
[ebuild  N    ] sci-mathematics/lcalc-2.1.0  USE="double -double-double -pari -quad-double"

[Comment 1 Toralf Förster 2025-04-07 18:30:42 UTC]

Created attachment 923754 [details]
emerge-info.txt

[Comment 2 Toralf Förster 2025-04-07 18:30:43 UTC]

Created attachment 923755 [details]
emerge-history.txt

[Comment 3 Toralf Förster 2025-04-07 18:30:44 UTC]

Created attachment 923756 [details]
environment

[Comment 4 Toralf Förster 2025-04-07 18:30:45 UTC]

Created attachment 923757 [details]
etc.clang.tar.xz

[Comment 5 Toralf Förster 2025-04-07 18:30:45 UTC]

Created attachment 923758 [details]
etc.portage.tar.xz

[Comment 6 Toralf Förster 2025-04-07 18:30:46 UTC]

Created attachment 923759 [details]
logs.tar.xz

[Comment 7 Toralf Förster 2025-04-07 18:30:47 UTC]

Created attachment 923760 [details]
qlist-info.txt.xz

[Comment 8 Toralf Förster 2025-04-07 18:30:48 UTC]

Created attachment 923761 [details]
sci-mathematics:lcalc-2.1.0:20250406-233309.log

[Comment 9 Toralf Förster 2025-04-07 18:30:49 UTC]

Created attachment 923762 [details]
temp.tar.xz

[Comment 10 Michael Orlitzky 2025-04-18 03:54:24 UTC]

I tried with LLVM 20 and it built OK. Can you reproduce this?

If so, can you run "objdump -t" on /usr/lib/llvm/20/bin/../../../../lib/clang/20/lib/linux/libclang_rt.builtins-x86_64.a and check for __muldc3 and __divdc3?

In mine they are hidden but it causes no problems, e.g.:

  0000000000000000 g     F .text  0000000000000810 .hidden __divdc3


This looks like the same issue as bug 932504 but there's no explanation for that one either.

CC llvm@ in case there's some C++/LLD quirk here that I wouldn't know about.

[Comment 11 Sam James 2025-04-18 04:00:13 UTC]

Try `elibtoolize` (see bug 914068).

[Comment 12 Larry the Git Cow 2025-04-18 04:20:32 UTC]

The bug has been referenced in the following commit(s):

https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=c97c57985dee6abdf0f101ca96b7347b03afaae7

commit c97c57985dee6abdf0f101ca96b7347b03afaae7
Author:     Michael Orlitzky <[e-mail]>
AuthorDate: 2025-04-18 04:15:10 +0000
Commit:     Michael Orlitzky <[e-mail]>
CommitDate: 2025-04-18 04:15:52 +0000

    sci-mathematics/lcalc: try elibtoolize to fix bug 953363
    
    Bug: https://bugs.gentoo.org/953363
    Signed-off-by: Michael Orlitzky <[e-mail]>

 .../lcalc/{lcalc-2.1.0.ebuild => lcalc-2.1.0-r1.ebuild}          | 9 +++++++++
 1 file changed, 9 insertions(+)

[Comment 13 Michael Orlitzky 2025-04-18 04:22:05 UTC]

(In reply to Sam James from comment #11)
> Try `elibtoolize` (see bug 914068).

Sure, thanks. Toralf, please test the -r1 to see if it works for you since I wasn't able to reproduce the problem on my machine.

[Comment 14 Toralf Förster 2025-04-18 09:27:35 UTC]

(In reply to Michael Orlitzky from comment #13)
> (In reply to Sam James from comment #11)
> > Try `elibtoolize` (see bug 914068).
> 
> Sure, thanks. Toralf, please test the -r1 to see if it works for you since I
> wasn't able to reproduce the problem on my machine.

works

[Comment 15 Michael Orlitzky 2025-04-18 13:30:47 UTC]

cool, thank you both