gcc-mesboot@4.9.4 has duplicate ‘gcc-mesboot1 ’ input
(address . bug-guix@gnu.org)
Hello,
The gcc-mesboot@4.9.4 package has the same ‘gcc-mesboot1’ input twice:
Toggle snippet (20 lines)
$ guix describe
Generacio 159 Sep 18 2020 08:55:27 (nuna)
guix 98b89f4
repository URL: https://git.savannah.gnu.org/git/guix.git
branch: master
commit: 98b89f432103b66efacee0bcba41a94148b8e870
$ cat $(guix gc --references $(guix build -e '(@@ (gnu packages commencement) gcc-mesboot)' -d) |grep gcc-mesboot-4.9.4-guile-builder)
[…]
(define %build-inputs
(quote
(("source" . "/gnu/store/56km255bk1s880yn6rxw4wj3crm2pshk-gcc-4.9.4.tar.xz")
("gcc-wrapper" . "/gnu/store/w527yvvbn3m8m243x7q9h6bal035ry9h-gcc-mesboot1-wrapper-4.6.4")
("gcc" . "/gnu/store/q7f07gpd9sq9krc68gxa73vbqmk06y0a-gcc-mesboot1-4.6.4") ; ⇐
("headers" . "/gnu/store/aj9plwk7k9ma17bbcwd7l8f48z1bim66-glibc-headers-mesboot-2.16.0")
("libc" . "/gnu/store/wgb70ir475v4vhg4bcjzv4h2kaa93qj7-glibc-mesboot-2.16.0")
("binutils" . "/gnu/store/mza1k0mgr8ab407yzavwwz07ly7221wy-binutils-mesboot-2.20.1a")
("xz" . "/gnu/store/r1zsxj7wlvw1aa1ifv3nyrrjag44pc9s-xz-mesboot-5.0.0")
("gcc" . "/gnu/store/q7f07gpd9sq9krc68gxa73vbqmk06y0a-gcc-mesboot1-4.6.4") ; ⇐
This is visible at the package level:
Toggle snippet (17 lines)
scheme@(gnu packages commencement)> ,pp (package-native-inputs gcc-mesboot)
$11 = (("gcc-wrapper"
#<package gcc-mesboot1-wrapper@4.6.4 gnu/packages/commencement.scm:2321 7f3f5dd12960>)
("gcc"
#<package gcc-mesboot1@4.6.4 gnu/packages/commencement.scm:1992 7f3f5dd12d20>)
("headers"
#<package glibc-headers-mesboot@2.16.0 gnu/packages/commencement.scm:2183 7f3f5dd12aa0>)
("libc"
#<package glibc-mesboot@2.16.0 gnu/packages/commencement.scm:2291 7f3f5dd12a00>)
("binutils"
#<package binutils-mesboot@2.20.1a gnu/packages/commencement.scm:2108 7f3f5dd12b40>)
("xz"
#<package xz-mesboot@5.0.0 gnu/packages/commencement.scm:2041 7f3f5dd12c80>)
("gcc"
#<package gcc-mesboot1@4.6.4 gnu/packages/commencement.scm:1992 7f3f5dd12d20>)
I noticed it while trying to address
https://issues.guix.gnu.org/38100: it prevents the simple solution of
using (delete-duplicates input-drvs) in ‘bag->derivation’. (Apart from
that this issue is harmless, but I think we’d better be strict about
this sort of things deep down in the graph.)
Ludo’.