Unfixed bugs in not-shipped components...

Posted May 17, 2024 16:26 UTC (Fri) by sandeen (guest, #42852)
In reply to: White paper: Vendor Kernels, Bugs and Stability by prarit
Parent article: White paper: Vendor Kernels, Bugs and Stability

Exactly.

Just looking at filesystems, there are about 700 "unfixed bugs" in RHEL8 for filesystems which are /not shipped in RHEL8/ based on Fixes: tags in commits to those filesystems since v5.14.

Failing to filter out bugfixes for subsystems which are configured off in the RHEL kernel makes this "analysis" lazy at best, disingenuous or malicious at worst.

Unfixed bugs in not-shipped components...

Posted May 17, 2024 17:30 UTC (Fri) by sandeen (guest, #42852) [Link] (1 responses)

Oops, I didn't count everything - RHEL8 is v4.18 not v5.14. (time flies!)

There are over 1500 "not fixed bugs" since v4.18 in not-shipped filesystems alone.

That is fully 1/3 of the "unfixed bug" count in their whitepaper without even looking at other unshipped drivers, subsystems and architectures.

There may be an interesting analysis to be made, but this whitepaper doesn't seem to be it.

Unfixed bugs in not-shipped components...

Posted May 19, 2024 4:06 UTC (Sun) by wtarreau (subscriber, #51152) [Link]

And how can you be certain that no other fixes rely on these ones being present, if the patch authors themselves don't know ? That's why stable maintainers constantly say "do not cherry-pick, take them all".