patch for kernel 5.15?

[RJVB]

Which patch should/could I use for a 5.15 kernel? Are they for a range (in which case I'd probably have to pick the one for 5.3) or for a single, specific version?

[edward-ibm]

Assuming you need a stable release (format 4.X.Y) I would recommend you to patch vanilla 5.15 kernel with reiser4-for-5.13

[RJVB]

(https://sourceforge.net/projects/reiser4/files/reiser4-for-linux-5.x/reiser4-for-5.13.patch.gz)

Ah, somehow I overlooked that one, thanks.

[Metztli]

Assuming you need a stable release (format 4.X.Y) I would recommend you to patch vanilla 5.15 kernel with reiser4-for-5.13

I was not aware I was following @edward-ibm ;) However, been running reiser4 for 5.17.12-2 for AMD64 for a while:

[daemonserj]

Эдуард, есть надежда на адаптацию reiser4 для kernel 6.x ?

[edward-ibm]

Где-то неделя свободного времени на это нужна, чего я пока не могу найти, к сожалению...

[Metztli]

Где-то неделя свободного времени на это нужна, чего я пока не могу найти, к сожалению...

I was able to build reiser4 SFRN 4.0.2 kernel 5.17.12-3 for Metztli Reiser4 / Debian Bookworm backports (with default) GCC12 by modifying kernel source file /scripts/pahole-flags.sh by adding following code:

if [ "${pahole_ver}" -ge "124" ]; then
        extra_paholeopt="${extra_paholeopt} --skip_encoding_btf_enum64"
fi

But, of course, it is a downgraded kernel from the default release 6.x.y ...

[Metztli]

Couple of days ago I used tar utility to create a ~ 7 Gig archive. I did not notice anything in my shell until the next day when I checked the log

--------------------- Kernel Begin ------------------------

WARNING: Kernel Errors Present WARNING: Flushing like mad: 16384 ...: 1 Time(s)

---------------------- Kernel End -------------------------

and executed dmesg:

... [1338199.579185] reiser4[tar(1728204)]: commit_current_atom (fs/reiser4/txnmgr.c:1051)[nikita-3176]: WARNING: Flushing like mad: 16384 ...

Notwithstanding, reiser4 SFRN 4.0.2 -patched kernel continues working appropriately.

P.S. A few days ago I had installed my newest kernel build referenced above, i.e., 5.17.12-3+reiser4.0.2, over the previous one but have not rebooted the remote server yet. Probably that had to do with the 'error' and/or 'warning'.

[Metztli]

In a couple of days, if nothing else, I will be releasing a Metztli Reiser4 hack based on Debian 12 Bookworm AMD64 with the downgraded kernel referenced above. I had some issues with Busybox 1.36.x-y and what the Debian Installer (d-i) maintainers expected in regards to the path to the busybox executable, i.e., d-i generated the wrong path and thus 'failed to execute /init'. I had to manually decompress the initrd with cpio to find out what was going on. Accordingly, after 24 to 36 intensive hours of analyzing the issue, I was able to build busybox with its executable located at the path expected by the d-i. Now the Debian Installer (d-i) is working ;-)

[daemonserj]

Good job, @Metztli ! Great, that you keep alive reiser4 in a modern distro. Still waiting @edward-ibm

[Metztli]

Спасибо, @daemonserj ! After several test installations to VirtualBox 7.0.20, a couple of days ago I used the Metztli Reiser4 netboot d-i to install Bookworm to my local development machine,

where I, again, built VirtualBox 7.0.20 in a reiser4 environment and fine-tuned the last installation phase to purge the newer default Linux kernel 6.x.y and only leave the downgraded reiser4 -enabled kernel 5.17.12-3+reiser4.0.2.

Too bad I can not translate the updated text strings in the relevant dialogs to other languages but the installer is made available at Metztli IT , as well as at SourceForge.