Re: State of core-updates

[Kaelyn]

Hi,

On the topic of the state of core-updates, I wanted to mention two things 
affecting i686-linux builds (and by extension some x86_64 packages like wine64):

1) glib-networking has a 32-bit-only patch left over from the upgrade from 
2.70.0 to 2.72.2, which does not apply against the newer version, and which 
seems unneeded. I just sent in https://issues.guix.gnu.org/62209 to fix the 
package.

2) libaio 0.3.113 does not build on core-updates, though the previous version 
0.3.112 does. I'm not sure how to handle this one, as the failure is a compile 
error from one of the test cases:

gcc -Wall -Werror -I../src -g -O2 -DTEST_NAME=\"cases/23.t\" -o cases/23.p 
main.c ../src/libaio.a -lpthread
mkdir testdir
rm -f testdir/rofile
echo "test" >testdir/rofile
chmod 400 testdir/rofile
rm -f testdir/rwfile
rm -f testdir/wofile
echo "test" >testdir/rwfile
echo "test" >testdir/wofile
chmod 600 testdir/rwfile
chmod 200 testdir/wofile
In file included from main.c:24:
cases/23.t: In function ‘thrproc2’:
cases/23.t:82:35: error: passing argument 2 of ‘splice’ from incompatible 
pointer type [-Werror=incompatible-pointer-types]
   82 |                 if (splice(tmpfd, &offset, pipefds[1], NULL, 1, 0) != 1)
      |                                   ^~~~~~~
      |                                   |
      |                                   off_t * {aka long int *}
In file included from 
/gnu/store/0hr9jpczkcgpgqkhf4q4868xd57h5a62-glibc-2.35/include/bits/fcntl.h:61,
                 from 
/gnu/store/0hr9jpczkcgpgqkhf4q4868xd57h5a62-glibc-2.35/include/fcntl.h:35,
                 from main.c:9:
/gnu/store/0hr9jpczkcgpgqkhf4q4868xd57h5a62-glibc-2.35/include/bits/fcntl-linux.h:398:49:
 note: expected ‘__off64_t *’ {aka ‘long long int *’} but argument is of type 
‘off_t *’ {aka ‘long int *’}
  398 | extern __ssize_t splice (int __fdin, __off64_t *__offin, int __fdout,
      |                                      ~~~~~~~~~~~^~~~~~~
In file included from main.c:24:
cases/23.t: In function ‘thrproc3’:
cases/23.t:106:35: error: passing argument 2 of ‘splice’ from incompatible 
pointer type [-Werror=incompatible-pointer-types]
  106 |                 if (splice(tmpfd, &offset, pipefds[1], NULL, 1, 0) != 1)
      |                                   ^~~~~~~
      |                                   |
      |                                   off_t * {aka long int *}
In file included from 
/gnu/store/0hr9jpczkcgpgqkhf4q4868xd57h5a62-glibc-2.35/include/bits/fcntl.h:61,
                 from 
/gnu/store/0hr9jpczkcgpgqkhf4q4868xd57h5a62-glibc-2.35/include/fcntl.h:35,
                 from main.c:9:
/gnu/store/0hr9jpczkcgpgqkhf4q4868xd57h5a62-glibc-2.35/include/bits/fcntl-linux.h:398:49:
 note: expected ‘__off64_t *’ {aka ‘long long int *’} but argument is of type 
‘off_t *’ {aka ‘long int *’}
  398 | extern __ssize_t splice (int __fdin, __off64_t *__offin, int __fdout,
      |                                      ~~~~~~~~~~~^~~~~~~
cc1: all warnings being treated as errors
make[1]: *** [Makefile:24: cases/23.p] Error 1
make[1]: Leaving directory 
'/tmp/guix-build-libaio-0.3.113.drv-0/libaio-0.3.113/harness'
make: *** [Makefile:23: partcheck] Error 2

Test suite failed, dumping logs.
error: in phase 'check': uncaught exception:
%exception #<&invoke-error program: "make" arguments: ("partcheck" "-j" "12" 
"prefix=/gnu/store/xr6s773c3d62g9aynydp1h6231p42ixn-libaio-0.3.113" "CC=gcc") 
exit-status: 2 term-signal: #f stop-signal: #f> 
phase `check' failed after 0.3 seconds


Cheers,
Kaelyn

P.S. For context, I hit the libaio error trying to build icecat (x86_64) on 
core-updates the other day; I hit the glib-networking error this morning trying 
to build wine, and then hit the libaio error again when retrying the wine 
(i686) build with my glib-networking change applied. The same build error 
affects both x86_64 and i686 builds of libaio.