$ cat linux/week/2026-09-21.log
the week in review · Sep 21 – Sep 27, 2026
Omarchy patches two security holes as 7.3-rc5 fixes land
By RepoJournal · composed from the cited sources · human-reviewed weekly · methodology
Kernel rc5 gathers i2c, XFS, parisc, TCP and SCSI fixes, while archinstall drops a stale reflector wait and Omarchy reworks screenshots and video.
OM-SEC-22: Isolate OCaml cleanup authorization basecamp/omarchy
Both advisories hit the same surface: local, package-manager-adjacent helpers that run with elevated rights. User-controlled code previously ran before an ordinary sudo cleanup, and now the removal path invalidates the cached timestamp first, so a detached child cannot wait for the authentication prompt and reuse it.
[Security] Treat command operands as literal values basecamp/omarchy
Several helpers forwarded command-line values without terminating the option list, so a value like --help could be read as a flag and the script would show help, fail, or act on the wrong input. Values are now passed as literals.
Merge tag 'i2c-fixes-7.3-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/andi.shyti/linux torvalds/linux
Most of the tag is DMA resource management: at91, imx and qcom-geni all fix paths where a probe error or driver detach left exclusive tx/rx channels held. A remove callback never runs after a failed probe, so each of those drivers returned without releasing anything.
i2c: at91: release DMA channels on remove and probe error torvalds/linux
at91_twi_configure_dma() requests exclusive tx/rx channels that nothing released on detach, and the probe error path after i2c_add_numbered_adapter() failed returned without releasing them either. Relevant if you reload the driver or repeatedly probe a device on those SoCs.
i2c: qcom-geni: release DMA channels on probe error torvalds/linux
geni_i2c_init() grabs exclusive GPI tx/rx channels when the serial engine runs in GPI mode, then leaks them if i2c_add_adapter() fails. The same failed-probe hole shows up in i2c: imx, where the driver falls back to PIO on errors other than -EPROBE_DEFER but still never frees the channels.
Merge tag 'xfs-fixes-7.3-rc5' of gitolite.kernel.org:/pub/scm/fs/xfs/xfs-linux torvalds/linux
Carlos Maiolino's pull request describes the set as mostly "random" bugfixes found by LLM tools on different XFS subsystems, plus a few cleanups and a NULL pointer dereference on zoned support. XFS on zoned devices is the part to watch; the rest is background noise unless you are chasing a specific hit.
Merge tag 'parisc-for-7.3-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux torvalds/linux
The 64-bit kernel stack goes to 32kB to avoid crashes, early parameters are parsed in setup_arch() so huge page settings take effect, and further fixes round out the tag. parisc is maintenance-only for almost everyone reading this.
Use Omasnap for screenshots basecamp/omarchy
Omasnap replaces Tensaku behind omarchy screenshot and in the base package set, keeping the existing capture modes, the positional copy and save forms, and the legacy OMARCHY_SCREENSHOT_DIR override. The desktop video path moves to OWE at the same time, sharing a muted lock feed across lock-screen outputs; locks keep still backgrounds and cached posters visible while the feed starts, pauses, di...
$ ls linux/week/ # the briefings behind this review
Keep up with Linux in about 3 minutes: what actually shipped — the commits, pull requests, releases, and security advisories that matter.
fair warning: these emails are deeply technical. diffs, version numbers, CVEs, benchmark deltas. if that's not your idea of a good read, this isn't your newsletter.
Folds into your digest — weekly by default, monthly if you prefer. Unsubscribe in one click.