- openSUSE Kernel

Re: Heads-up: Snapshot 20220510 will be larger than usual (GCC12 rebuild)
by Andrei Borzenkov 16 May '22

16 May '22

On 13.05.2022 13:31, Dominique Leuenberger / DimStar wrote: > Dear Tumbleweed users, > > Just a small shout-out: Snapshot 0510 is currently in progress of > syncing out to the mirrors and should become available later this > afternoon. > > This snapshot has been rebuilt entirely using gcc12 Apparently it broke kexec https://bugzilla.opensuse.org/show_bug.cgi?id=1199593

1 0

EFI pstore backend is enabled without corresponding cleanup service
by Andrei Borzenkov 11 Apr '22

11 Apr '22

In Tumbleweed EFI pstore backend is active due to P: /devices/platform/efivars.0 L: 0 E: DEVPATH=/devices/platform/efivars.0 E: MODALIAS=platform:efivars E: SUBSYSTEM=platform E: USEC_INITIALIZED=25091896 E: ID_PATH=platform-efivars.0 E: ID_PATH_TAG=platform-efivars_0 uefi:~ # modinfo platform:efivars filename: /lib/modules/5.17.1-1-kvmsmall/kernel/drivers/firmware/efi/efi-pstore.ko.zst alias: platform:efivars uefi:~ # cat /sys/module/efi_pstore/parameters/pstore_disable N uefi:~ # grep PSTORE /boot/config-5.17.1-1-kvmsmall CONFIG_EFI_VARS_PSTORE=m # CONFIG_EFI_VARS_PSTORE_DEFAULT_DISABLE is not set so dumps get written to NVRAM but nothing cleans them up - systemd provides systemd-pstore service, but it is part of systemd-experimental package that is not installed by default. Should EFI pstore backend be enabled by default? It is explicitly disabled in SUSE/Leap if (!backend && !strcmp(psi->name, "efi")) { pr_info("Efi pstore disabled, enforce via pstore.backend=efi"); pr_info("On a broken BIOS, this can severely harm your system"); pr_info("Only enable efi based pstore when you know what you are doing"); return -EINVAL; }

3 3

kernel-vanilla: incorrect shim signature
by Matwey V. Kornilov 26 Mar '22

26 Mar '22

Hi, I am running openSUSE 15.3 and I would like to test the drm-tip kernel branch on my hardware. I am trying to install and run kernel-vanilla-5.17.0-kt.5.1.g401575d from the following repo: https://download.opensuse.org/repositories/home:/tiwai:/kernel:/drm-tip/ope… In spite of the package has been installed successfully, I cannot boot due to "incorrect shim signature" at a very early stage. Secure boot is enabled currently. What is the proper way to overcome this problem? I suppose disabling secure boot should help. -- With best regards, Matwey V. Kornilov

2 2

Dracut crypt hook not waiting for initqueue
by Georg Pfuetzenreuter 24 Mar '22

24 Mar '22

Hi, probably of a topic for the dracut mailing list, but since I cannot reach it, I figured I should ask here where I feel at home. :) I am implementing a Dracut module to unlock a LUKS encrypted root partition. This works all fine, but in case the module fails, I would like to have the default systemd based crypt module as a fallback. Unfortunately this ystemd based crypt module does not seem to respect the hook priorities. I install my script with inst_hook initqueue 01 "$moddir/luksrku-script.sh" but during boot, systemd-tty-ask-password-agent comes up right in the middle of the initqueue scripts and asks for a passphrase. My script continues to unlock the disk, but the passphrase prompt stays, until someone hits the enter key. I currently work around the issue by manually failing the prompt if my script succeeds: if [ "$?" = "0" ]; then echo "" | systemd-tty-ask-password-agent fi but this is not ideal, as it causes systemd to print errors messages: Failed to start Cryptography Setup for cr_root. Dependency failed for Local Encrypted Volumes. Granted, those errors are totally ignorable, as after Dracut continues with its boot process systemd "realizes" the disks have already been unlocked (as seen in the journal after boot-up): systemd-cryptsetup[581]: Volume cr_root already active. systemd[1]: Finished Cryptography Setup for cr_root. But it is still not pretty to have these errors during boot and I feel like there has to be some way to make the crypt hook wait for my initqueue hook to finish (which I assume would make it "realize" the volume is already unlocked before calling the password agent)? Note that initqueue is the earliest I can call my script, as I need network dependencies not yet available in the cmdline or udev runs. I found that there is a basic check for $NEWROOT in /usr/lib/dracut/modules.d/90crypt/cryptroot-ask.sh, but this script seems to be only be called on non-systemd systems without systemd-tty-ask-password-agent, making it not relevant for openSUSE. Would appreciate any pointers! Thanks for reading, Georg

1 0

Kernel:stable provides 5.17.rc7
by Frank Krüger 08 Mar '22

08 Mar '22

Hi there, This repo contains kernel-default-5.17~rc7-1.1.g04b7727. Is this on purpose? Regards, Frank

2 1

RFC: switch default flavor to HZ=300
by Dirk Müller 25 Feb '22

25 Feb '22

Hi, I recently stumbled over the 300 HZ option in the kernel configuration. Currently we are using 250 HZ with IDLE_HZ. The HZ setting is used as far as I can see to determine when jiffies are advancing, so it influences the granularity of a couple of things like kvmclock, timers and so on. Also USER_HZ is 100 HZ which is the granularity to which some metrics are exported to user space. 300 is divisible without remainder by 100, unlike 250. so it appears there are good reasons for switching the default, and it is "just" 20% more timer interrupts than before, so it should not be a huge issue. I also did a test build and a 300 HZ kernel is by a few bytes smaller than a 250 HZ kernel, indicating that the compiler can optimize away a few things. I am seeing an increase in a few small functions, and I'm looking into making the code size increase go away with a source level tweak. I've benchmarked both versions in a micro benchmark that does a billion invocations of both, and while the code is larger, it runs in exactly the same runtime (+/- 3% which I consider my benchmark noise level) on a Ryzen Zen 2+. In the Kconfig description of 300 HZ option, it appears this is more recommended for multimedia usecases because it is divisible without remainder for common rates, like 30 (fps), 60 fps , 120 fps, 44.1khZ and others that are often needed. This is imho not only usable for desktop, but also for servers that are using multimedia related applications. I've seen that fedora-like distributions use 1000 HZ, arch linux uses 300 Hz and debian defaults to 250 HZ. So there is no clear trend. I'd be fine with 300 or 1000 HZ. Comments? Would like to hear your feedback before sending a change to the Tumbleweed configs. Thanks, Dirk

5 7

What's the status of the Kernel:stable:Backport project?
by Manfred Hollstein 23 Feb '22

23 Feb '22

Hi there, I use $SUBJECT on my Leap systems which utilize newer hardware than what is supported by the standard Leap kernel. This used to work up until version 5.16.9-lp153.4.1.gdedbf20, but stopped afterwards. The gcc version in that project obviously needs its own gcc11 which doesn't exist: $ osc buildinfo Kernel:stable:Backport kernel-default standard x86_64 <buildinfo project="Kernel:stable:Backport" repository="standard" package="kernel-default" downloadurl="https://download.opensuse.org/repositories"> <arch>x86_64</arch> <error>unresolvable: nothing provides gcc11-PIE needed by gcc-PIE</error> <srcmd5>f07c78b4d19d988247e4676ed04e64d2</srcmd5> <verifymd5>3a4d3852e56d62dadb7dcaa75a12c8d4</verifymd5> <rev>1</rev> <specfile>kernel-default.spec</specfile> <file>kernel-default.spec</file> <versrel>5.16.10-2</versrel> <bcnt>1</bcnt> <release>2.1</release> <debuginfo>1</debuginfo> </buildinfo> Is this project still supposed to work for Leap users who need a newer kernel? TIA, cheers. l8er manfred

2 2

zswap
by Joe Salmeri 15 Feb '22

15 Feb '22

If I run cat /boot/config-5.16.1-1-default | grep -i swap It returns CONFIG_ZSWAP=Y cat /sys/module/zswap/parameters/enabled returns N So the kernel supports zswap but it is currently NOT enabled. When system is booted the following journal message appears kernel: zswap: loaded using pool lzo/zbud ps aux | grep -i zswap returns root 135 0.0 0.0 0 0 ? I< Feb09 0:00 [zswap-shrink] If zswap is NOT enabled, why is zswap loaded with a pool at boot time and why is the zswap-shrink process running? I find that the same situation on a TW test machine I have that is running the 5.16.5.2 kernel so it is not an isolated case. Is this a bug?

2 3

Re: zswap issue?
by Frank Krüger 13 Feb '22

13 Feb '22

Am 12.02.22 um 17:07 schrieb Joe Salmeri: > cat /boot/config-5.16.1-1-default | grep -i swap > > returns CONFIG_ZSWAP=Y > > cat /sys/module/zswap/parameters/enabled > returns N > > So the kernel supports zswap but it is currently NOT enabled. > > When system is booted the following journal message appears > > kernel: zswap: loaded using pool lzo/zbud > > ps aux | grep -i zswap > > returns > > root 135 0.0 0.0 0 0 ? I< Feb09 0:00 [zswap-shrink] > > If zswap is NOT enabled, why is zswap loaded with a pool at boot time and why is the > zswap-shrink process running? > > I find that the same situation on a TW test machine I have that is running the 5.16.5.2 kernel so it is not an isolated case. > > Is this a bug? Maybe kernel(a)lists.opensuse.org is more appropriate (cc'ing). Regards, Frank

1 0

CONFIG_SECURITY_DMESG_RESTRICT differences
by Stefan Seyfried 02 Feb '22

02 Feb '22

Hi all, I just found that I could not do "dmesg" as normal user in a Leap 15.3 installation. I'm used to be able to do this from my Factory machines. Investigation showed, that Leap's kernel sets CONFIG_SECURITY_DMESG_RESTRICT=y while the Factory kernel (at least kernel-vanilla and kernel-default from Kernel:HEAD) does not. I do not care which default is chosen, but for consistency I'd suggest to settle for one of the two possible options ;-) Have fun, -- Stefan Seyfried "For a successful technology, reality must take precedence over public relations, for nature cannot be fooled." -- Richard Feynman

4 4