Re: Fedora Linux 39 Final blocker status summary #3
Once upon a time, Adam Williamson said: > 6. distribution - https://bugzilla.redhat.com/2242759 - NEW > dnf system-upgrade fails on some RPi4 due to system boot date that pre- > dates gpg key > > We're still kinda kicking around ideas for "fixing" this, but I think > if push comes to shove, we'll wind up revoting or waiving it as not > practically fixable. Not adding to the ticket (because "me too" is not useful there), but... I think Fedora should include SOME type of "fake hwclock"-type thing for systems with no RTC (make a systemd service depend on /dev/rtc not existing?), as other RPi-targeted distros do. This isn't RPi-specific, a number of the small boards have no RTC. I do typically add an RTC to my Pis, but not always (for various reasons). -- Chris Adams ___ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
Fedora Linux 39 Final blocker status summary #3
Hi folks! We're still trying to get F39 done, so time for another status update... Action summary == Accepted blockers - 1. kexec-tools - https://bugzilla.redhat.com/2243068 - VERIFIED: releng to push the fix stable 2. mutter - https://bugzilla.redhat.com/2241632 - ASSIGNED: desktop team (and adamwill) to keep trying to come up with a fix 3. shim - https://bugzilla.redhat.com/2113005 - NEW: assume this will be waived 4. uboot-tools - https://bugzilla.redhat.com/2241252 - ASSIGNED: ARM team (pbrobinson) to fix it 5. uboot-tools - https://bugzilla.redhat.com/2244305 - ASSIGNED: ARM team to evaluate and fix if possible, ARM/QA to test on more systems if possible 6. distribution - https://bugzilla.redhat.com/2242759 - NEW: anyone at all to come up with a genius fix, otherwise we'll likely have to document this Bug-by-bug detail = Accepted blockers - 1. kexec-tools - https://bugzilla.redhat.com/2243068 - VERIFIED kdump is enabled by default on desktops This is basically fixed, just waiting to be pushed stable. 2. mutter - https://bugzilla.redhat.com/2241632 - ASSIGNED Netinstall ISO renders a black screen when using kickstart install (bare metal and VM) Well, we kinda had a fix for this, but it turns out to break something even worse (now anaconda isn't visible on the Workstation live image). So we're still stuck trying to find a perfect fix, unfortunately. Desktop team plus me to keep cranking away on it. 3. shim - https://bugzilla.redhat.com/2113005 - NEW Live image made with BOOTX64.EFI from latest shim-x64-15.6-2 fails to boot on some boards Let's just assume this is gonna be waived. 4. uboot-tools - https://bugzilla.redhat.com/2241252 - ASSIGNED Fedora-Workstation-39_Beta-1.1 boots to a black screen on Raspberry Pi 4 Peter says "So we've basically got to the bottom of the problem and worked out the issue, I now just need to come up with a fix.", so that's what we're waiting on. 5. uboot-tools - https://bugzilla.redhat.com/2244305 - ASSIGNED Fedora Server 39 does not boot on Raspberry Pi 4 (RPi4) from microSD card slot This one's also waiting on ARM team (i.e. Peter), but seems somewhat less clear-cut of a blocker, so we're kinda waiting for his take on that, plus testing from other Raspberry Pi owners would be useful. 6. distribution - https://bugzilla.redhat.com/2242759 - NEW dnf system-upgrade fails on some RPi4 due to system boot date that pre- dates gpg key We're still kinda kicking around ideas for "fixing" this, but I think if push comes to shove, we'll wind up revoting or waiving it as not practically fixable. -- Adam Williamson (he/him/his) Fedora QA Fedora Chat: @adamwill:fedora.im | Mastodon: @ad...@fosstodon.org https://www.happyassassin.net ___ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
Analysis of the overhead of frame pointers on gcc compiles
Today I've read (twice) that the overhead of frame pointers on the runtime of the compiler, GCC, is 10%. This number is nonsense. The actual overhead is 1%, and I have done the tests that show this. I want to note up front that it was also claimed that GCC slowed down by 10% between Fedora 37 & 38 (when frame pointers were introduced). However frame pointers have been disabled in our GCC build on all architectures for 9 months, since before Fedora 38 was released. So if there has been such a slowdown (which I don't know) then it cannot be anything to do with frame pointers. At the same time GCC was upgraded from 12 to 13, such a signicant new release that claims comparing versions of Fedora are not valid. - - - In my tests I recompiled GCC with frame pointers enabled (see gcc-add-fp.patch), and then built qemu with 'make clean; time make -j32'. I took the 'user' time over 3 runs. The average user time was 1135 seconds. I then recompiled binutils with frame pointers disabled (see binutils-no-fp.patch), and installed it alongside gcc-13.2.1-4.fc40.x86_64 (which as noted above has frame pointers disabled). The average user time over 3 runs was 1122 seconds. With frame pointers: 1135 seconds Without frame pointers: 1122 seconds Overhead: 1.01% Note also that the wallclock time in every case was either 41 or 42 seconds, so in practical terms for users there is no difference with or without frame pointers. You may also note that I didn't recompile any libraries (so in my tests, frame pointers are used in both cases). However using profiling we can show that this doesn't matter. The total time used during the make -j32 run by each binary and DSO was: 58.41% lto1 14.97% cc1 11.47% Linux kernel 7.68% libc.so.6 4.31% as 1.54% libbfd-2.41-7.fc40.so 0.64% libpython3.12.so.1.0 0.26% ld-linux-x86-64.so.2 0.09% ld.bfd 0.00% python3.12 (everything else under 1%) Note the only library which consumes any significant time is libc, but at under 8% it would hardly affect the test times. libbfd is actually part of binutils so it was recompiled with & without frame pointers. Full raw results are in the attached file. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests. http://libguestfs.org/virt-v2v Q: How much are libraries used? Children Self Shared Object + 97.79% 7.68% libc.so.6◆ + 65.25%58.41% lto1 ▒ + 20.62%14.97% cc1 ▒ + 12.33%11.47% [kernel.kallsyms]▒ +8.08% 4.31% as ▒ +6.03% 1.54% libbfd-2.41-7.fc40.so▒ +1.88% 0.09% ld.bfd ▒ +0.79% 0.64% libpython3.12.so.1.0 ▒ +0.78% 0.00% python3.12 ▒ +0.67% 0.26% ld-linux-x86-64.so.2 ▒ build qemu 'make -j32' (2591 targets) gcc-13.2.1-4.fc40.x86_64 (FP accidentally turned off) real0m41.297s user18m38.587s sys 2m5.460s real0m41.218s user18m36.044s sys 2m3.838s real0m41.493s user18m42.651s sys 2m4.668s above + binutils-2.41-8.1.nofp.fc40.x86_64 (FP turned off) real0m41.357s user18m39.692s sys 2m4.310s real0m41.403s user18m43.905s sys 2m4.399s real0m41.389s user18m44.637s sys 2m4.407s Average user: 1122 binutils-2.41-7.fc40.x86_64 + gcc-13.2.1-4.1.fp.fc40.x86_64 (FP turned on) real0m41.702s user18m51.091s sys 2m3.045s real0m42.000s user18m56.555s sys 2m3.301s real0m42.029s user18m58.724s sys 2m3.922s Average user: 1135 Overhead of frame pointers: 1135/1122 = 1.01% diff --git a/gcc.spec b/gcc.spec index 5d6c7c4..c7774fd 100644 --- a/gcc.spec +++ b/gcc.spec @@ -136,7 +136,7 @@ Summary: Various compilers (C, C++, Objective-C, ...) Name: gcc Version: %{gcc_version} -Release: %{gcc_release}%{?dist} +Release: %{gcc_release}.1.fp%{?dist} # libgcc, libgfortran, libgomp, libstdc++ and crtstuff have # GCC Runtime Exception. License: GPLv3+ and GPLv3+ with exceptions and GPLv2+ with exceptions and LGPLv2+ and BSD @@ -923,7 +923,6 @@ OPT_FLAGS=`echo $OPT_FLAGS|sed -e 's/-flto=auto//g;s/-flto//g;s/-ffat-lto-object OPT_FLAGS=`echo $OPT_FLAGS|sed -e 's/-m64//g;s/-m32//g;s/-m31//g'` OPT_FLAGS=`echo $OPT_FLAGS|sed -e 's/-mfpmath=sse/-mfpmath=sse -msse2/g'` OPT_FLAGS=`echo $OPT_FLAGS|sed -e 's/ -pipe / /g'` -OPT_FLA
Re: [Fedocal] Reminder meeting : ELN SIG
On Fri, Oct 20, 2023 at 8:46 AM Stephen Gallagher wrote: > > On Thu, Oct 19, 2023 at 8:00 AM wrote: > > > > Dear all, > > > > You are kindly invited to the meeting: > >ELN SIG on 2023-10-20 from 12:00:00 to 13:00:00 US/Eastern > >At fedora-meet...@irc.libera.chat > > > > The meeting will be about: > > > > No agenda set for today, but I'll hold an Open Floor meeting if anyone > has any topics. This week's meeting involved a long discussion of how to resolve LLVM major update and OCAML rebuild issues. = #fedora-meeting: ELN (2023-10-20) = Meeting started by sgallagh at 16:07:11 UTC. The full logs are available at https://meetbot.fedoraproject.org/fedora-meeting/2023-10-20/eln.2023-10-20-16.07.log.html . Meeting summary --- * Init Process (sgallagh, 16:07:24) * Agenda Topics (sgallagh, 16:09:56) * ELNBuildSync (sgallagh, 16:12:05) * LINK: https://sgallagh.wordpress.com/2023/10/13/sausage-factory-fedora-eln-rebuild-strategy/ (sgallagh, 16:12:27) Meeting ended at 17:16:11 UTC. Action Items Action Items, by person --- * **UNASSIGNED** * (none) People Present (lines said) --- * sgallagh (84) * yselkowitz (70) * tdawson (15) * jforbes (9) * zodbot_ (8) * zodbot (8) Generated by `MeetBot`_ 0.1.4 .. _`MeetBot`: http://wiki.debian.org/MeetBot ___ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
Re: Upgrades of Sundials and PETSc
Just a heads-up that whichever process you have been using to determine the list of affected dependent packages has been missing getdp; see https://bugzilla.redhat.com/show_bug.cgi?id=2245240 and https://bugzilla.redhat.com/show_bug.cgi?id=2189695. I just rebuilt getdp in Rawhide, so there’s nothing to do right now. For next time, this seems to work for me: for p in petsc{,64,-openmpi,-mpich} libpetsc.so; do repoquery -q --repo=rawhide{,-source} --whatrequires "$p"; done | sort -u Well, it doesn’t work in Rawhide right now, because the rebuilt getdp-3.5.0-10.fc40.x86_64 hasn’t been part of a compose yet, and getdp-3.5.0-9.fc40.x86_64 still requires libpetsc.so.3.19()(64bit), which petsc-0:3.20.0-1.fc40.x86_64 doesn’t provide. But at any other time, that command would reveal getdp as a package that needs rebuilding. Thanks, Ben Beasley (FAS music) On 10/7/23 06:50, Antonio T. sagitter wrote: Hi all. In 1 week, i will upgrade Sundials and PETSc to their related newer versions: PETSc-3.20.0 https://petsc.org/release/changes/320/ sundials-6.6.1 https://github.com/LLNL/sundials/releases Following packages will be rebuilt: bout++ freefem++ octave cantera dolfin Regards. ___ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue ___ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
Re: [Fedocal] Reminder meeting : ELN SIG
On Thu, Oct 19, 2023 at 8:00 AM wrote: > > Dear all, > > You are kindly invited to the meeting: >ELN SIG on 2023-10-20 from 12:00:00 to 13:00:00 US/Eastern >At fedora-meet...@irc.libera.chat > > The meeting will be about: > No agenda set for today, but I'll hold an Open Floor meeting if anyone has any topics. ___ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
Re: Self Introduction: Nils Koenig
On Thu, Oct 19, 2023 at 06:10:41PM +0200, Nils Koenig wrote: > Hi @all, > > a quick hello from my side as I would like to join the Fedora Package > Maintainers by taking over responsibility for the vhostmd package. > > I've created a fork and added my change: > > https://src.fedoraproject.org/fork/nilskoenig/rpms/vhostmd/c/ > 697e947272969f56f3f892a93e7d37c0babe6d5a?branch=rawhide > > But it seems I can't create a PR from there. Anyhow - I would like to take > over > maintainer ship from Richard Jones regarding this package. > How is that possible? You'll need to get a sponsor, and generally read & follow the steps described here: https://docs.fedoraproject.org/en-US/package-maintainers/Joining_the_Package_Maintainers/ Rich. > Another question, is it possible to fast-forward personal forks? > Or delete/recreate forks? > > Thank you, > Nils > > ___ > devel mailing list -- devel@lists.fedoraproject.org > To unsubscribe send an email to devel-le...@lists.fedoraproject.org > Fedora Code of Conduct: > https://docs.fedoraproject.org/en-US/project/code-of-conduct/ > List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines > List Archives: > https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org > Do not reply to spam, report it: > https://pagure.io/fedora-infrastructure/new_issue -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests. http://libguestfs.org/virt-v2v ___ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-le...@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue