Unsolicited kernel 5.13 RC1 feedback
by Richard Shaw
After reading the Phoronic's article about the AMD improvements baked in
the 5.15 kernel I decided to give it a try on my F34 desktop.
Bootup was OK but after logging in I saw some small screen corruption on my
secondary monitor but Gnome still logged in fine. But after it fully loaded
(showing the menu at the bottom) the screen went blank and did not recover.
I tried going to a terminal and switching from graphical to command line
and back and logged in again but it kicked me back to the login screen.
Looking forward to the improvements when they're ready as I have both AMD
CPU and GPU.
Thanks,
Richard
2 years, 7 months
Introducing API changes in google-benchmark
by Vitaly Zaitsev
Hello.
google-benchmark 1.6.0 introduces API change **without** a SOVERSION
bump: https://github.com/google/benchmark/pull/1208
Builds completed both for F35 and Rawhide. Make sure your packages are
compatible.
Potentially affected packages:
* cctz
* geometry-hpp
* libscn
* python-pypet
* seqan3
* snappy
* spdlog
I'll take care of libscn and spdlog.
--
Sincerely,
Vitaly Zaitsev (vitaly(a)easycoding.org)
2 years, 7 months
Why should $ORIGIN runpaths appear first?
by Orion Poplawski
vtk is generating the following rpaths errors like this:
ERROR 0008: file
'/usr/lib64/mpich/lib/vtk/libvtkFiltersParallelGeometryJava.so' contains
the $ORIGIN runpath specifier at the wrong position in
[/usr/lib64/mpich/lib:$ORIGIN:$ORIGIN/../]
The notes on the errors state:
* 0x0008 ... the special '$ORIGIN' RPATHs are appearing after other
* RPATHs; this is just a minor issue but usually unwanted
Can someone explain to me why it is unwanted? Why should $ORIGIN RPATHs
appear first?
In this particular case I don't feel concerned with ignoring the error
because the two paths appear to be (mostly) equivalent to me, but I'm
curious about the error.
I believe the paths come about from:
if (UNIX)
if (APPLE)
set(_vtk_java_origin_rpath_prefix
"@loader_path")
else ()
set(_vtk_java_origin_rpath_prefix
"$ORIGIN")
endif ()
list(APPEND CMAKE_INSTALL_RPATH
# For sibling wrapped modules.
"${_vtk_java_origin_rpath_prefix}")
if (DEFINED _vtk_java_LIBRARY_DESTINATION AND DEFINED
_vtk_java_JNILIB_DESTINATION)
file(RELATIVE_PATH _vtk_java_relpath
"/prefix/${_vtk_java_JNILIB_DESTINATION}"
"/prefix/${_vtk_java_LIBRARY_DESTINATION}")
list(APPEND CMAKE_INSTALL_RPATH
# For libraries.
"${_vtk_java_origin_rpath_prefix}/${_vtk_java_relpath}")
endif ()
endif ()
So I could submit a patch changing "APPEND" to "PREPEND", but other than
avoiding this error I don't know why this would be preferred.
Thanks,
Orion
--
Orion Poplawski
he/him/his - surely the least important thing about me
Manager of NWRA Technical Systems 720-772-5637
NWRA, Boulder/CoRA Office FAX: 303-415-9702
3380 Mitchell Lane orion(a)nwra.com
Boulder, CO 80301 https://www.nwra.com/
2 years, 7 months
Where has the kernel-doc package gone?
by Nils K
I recently had to perform a bit of development/research where I often had to take a look in the kernel documentation.
Most of the time was spend offline so I wanted to download the `kernel-doc` package however it does not seem to exist.
Some old fedora documentations still refer to it however somewhere in the 2x iteration of fedora it seems to have gone missing.
CentOS 8 also still has it.
Would it be possible to add this back to the repos?
2 years, 7 months