vcswatch reports that
this package seems to have a new changelog entry (version
2.7.4+ds-1, distribution
UNRELEASED) and new commits
in its VCS. You should consider whether it's time to make
an upload.
Here are the relevant commit messages:
commit e0ae26d4ee3376d99a937db90cf5b855c77c6055
Author: dann frazier <dannf@debian.org>
Date: Thu Oct 8 17:21:27 2026 -0600
New upstream release.
commit fe9beea528b0f83b81baf076a41b736942a11e53
Merge: 8c965d8 2fd3f8c
Author: dann frazier <dannf@debian.org>
Date: Thu Oct 8 17:21:07 2026 -0600
Update upstream source from tag 'upstream/2.7.4+ds'
Update to upstream version '2.7.4+ds'
with Debian dir 3bba7825d4cceae4420d2c65833956c913ac4e6f
commit 2fd3f8c79f59eb9b7c9b6a68d6ebd701a5ac95c7
Merge: 45e0f4d e657b2d
Author: dann frazier <dannf@debian.org>
Date: Thu Oct 8 17:21:06 2026 -0600
New upstream version 2.7.4+ds
commit e657b2d9efacbf0f92212d528c00b96386f37b47
Author: Penghe Geng <pgeng@nvidia.com>
Date: Thu Jul 2 09:32:51 2026 -0400
Update version to 2.7.4
commit ca00a94f62df3bcaa4e29eb88e421f28ee7c0e99
Author: Penghe Geng <pgeng@nvidia.com>
Date: Wed Jul 1 14:55:16 2026 -0400
rshim: retry stalled boot writes
A boot write can temporarily make no progress when the BlueField boot
FIFO has no space. The current path treats that zero-progress case as
-EINTR, so writing a BFB to /dev/rshim<N>/boot can fail even though the
caller was not interrupted.
Retry zero-progress writes until the FIFO drains or the existing boot
timeout expires. Keep the wait interruptible so Ctrl-C can abort a
stalled transfer, and keep releasing the backend mutex during the retry
delay so other rshim operations can continue.
Cherry-picked from master branch.
Orig RM #5110797
RM #5124285
commit 6a9b3ae5b021b00660e418de9b38bf7922d15b95
Author: Mahmoud Nobani <mnobani@nvidia.com>
Date: Sun Jun 14 10:10:31 2026 +0300
fix mft not working in chroot env
(cherry picked from commit 307c7956ea34d7bfd9b7b46f2569b86c88bed218)
![[bug history graph]](https://qa.debian.org/data/bts/graphs/r/rshim-user-space.png){kind=link}