9p patches in flight recap/what did I miss

Dominique Martinet <[email protected]> Thu, 16 Apr 2026 12:12:34 +0900
Newsgroups dev.linux.lists.v9fs
Message-ID <[email protected]>
Hi all,

This is mostly for myself as I can't keep up, please tell me if I'm
missing anything obvious.

I have these queued for 7.1-rc1 (will send early next week to Linus)

https://lore.kernel.org/r/[email protected]/
* 0fd76f1be20d 9p: fix memory leak in v9fs_init_fs_context error path
https://lore.kernel.org/r/[email protected]/
* da2346a48a5a 9p: fix access mode flags being ORed instead of replaced
https://lore.kernel.org/r/[email protected]/
* 890d56964c62 9p: document missing enum values in kernel-doc comments
https://lore.kernel.org/r/[email protected]/
* 72cb9ee4f6d8 9p/trans_xen: make cleanup idempotent after dataring alloc errors
* 8fc518e489c1 9p/trans_xen: replace simple_strto* with kstrtouintq

I have these queued for 7.2

https://lore.kernel.org/r/551c057c5dbcc83f19b9c8a3b508bd971598f657.1774380535.git.repk@triplefau.lt/
* fabe6e8af53a 9p: Cache negative dentries for lookup performance
* 2e624a12ba29 9p: Add mount option for negative dentry cache retention
* 060ba5a058cb 9p: Set default negative dentry retention time for cache=loose
* 47a37d360868 9p: Enable symlink caching in page cache

And these that need some work/help sorting out
- usb fixes [email protected]
** I don't have the bandwidth for such a big rework, happy to just trust
Michael here given I can't test usb...
- writeback corruption [email protected]
** David sent another patch but honestly it's too complex for a quick
review, my gut feeling is to prefer Pierre's simpler approach first and
take the time to iterate (if we ever find time to...), but I just
haven't the time to check even that.
- lifting virtio msize limit
[email protected] and/or virtio buffer
iteration [email protected]
** Once again ENOTIME stikes so I didn't check, but do your patches also
address the invalid access problem? Should we get Chris to test with
your patches?
I still think it's a shame to duplicate all this code and it should be
possible to do a bit better, but I agree improving sg_table is an
elephant I don't want to fight...


Thank you all,

by the way if anyone wants to take over for (at least) the next few
cycles I wouldn't mind at all, having a shiny new baby and a jealous 3yo
doesn't help :p
-- 
Dominique