Mk/bsd.port.mk: Drop the strip parameter from PATCHFILES passed to do-fetch.sh
After recent changes to handling of do-fetch.sh in bsd.port.mk, the
strip parameters (e.g., ":-p1") are no longer dropped before do-fetch.sh
receives patch files.
As a result, a strip parameter of a patch file might be interpreted as a
group instead and get prepended to the patch file, resulting in warnings
like:
```
$ make makesum
===> License MIT accepted by the user
=> e3efbb5923f638b38087a3f616906d9d4fdaf6f2.patch doesn't seem to exist in /var/cache/distfiles.
=> Attempting to fetch p1e3efbb5923f638b38087a3f616906d9d4fdaf6f2.patch
fetch: p1e3efbb5923f638b38087a3f616906d9d4fdaf6f2.patch: No such file or directory
```
Reviewed by: mat
[4 lines not shown]
hn: Refresh VF RSS configuration after link recovery
A VF can finish initialization or reset recovery after hn's handoff-time
RSS query returned ENXIO. In that case hn suppresses synthetic receive
hash metadata, but previously left it disabled even after the VF could
answer the queries again.
Queue an RSS refresh on a VF link-up notification, using the existing VF
worker in both transparent and non-transparent modes. Revalidate the
association, active VF path, administrative state, and carrier under
hn_lock before querying and reconfiguring synthetic RSS. Coalesce the
requests with an atomic flag and retain a request while capability
forwarding temporarily excludes the worker. Keep the existing query
validity checks and unsupported-query fallback unchanged.
This is a one-shot refresh, not a readiness poll. Recovery without a
link-up notification, or a query that still fails during the refresh,
does not trigger another retry by itself. A later link-up or normal
handoff can query again. Consuming the request while the VF path is
[17 lines not shown]
iavf: Reschedule pending virtchnl replies after the processing budget
The VC worker ignores the pending count returned by iavf_process_adminq.
If a pass exhausts its message budget, replies can remain in the receive
ring without another interrupt to schedule the worker. During init,
the iflib admin task cannot provide a backstop because init holds the
context lock while waiting for ENABLE_QUEUES to complete.
Reschedule the VC worker after a successful pass with messages remaining.
Keep the per-pass budget and do not retry AdminQ errors. Stop
self-rescheduling once detach clears INITIALIZED; the existing polling
and reset-recovery gates continue to exclude asynchronous processing.
Validation: reproduced the timeout on a three queue E835 VF under
Hyper-V. Temporary tracing showed a successful ENABLE_QUEUES reply
already in the guest receive ring after a three-message pass left two
messages pending. With the fix and no tracing, a GENERIC kernel with
WITNESS and INVARIANTS passed three batches of 30 down/up and MTU
1500/9000 cycles without enable/disable timeouts.
[3 lines not shown]
amdsmu: Hints on what to do if IP block blocking S0i3
Reviewed by: olce
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D59674
amdsmu: Diagnose S0i3 entry after resume
Print out diagnostic information after resuming from suspend-to-idle if
we failed to enter S0i3, i.e. the IP blocks that were blocking entry to
S0i3.
Don't give detailed IP block info for other SMUs than for Phoenix, as I
have not had a chance to test these yet and the SMU seems to be very
quirky.
Reviewed by: olce
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D59673
power: Add power_resume_check eventhandler
This is intended to be used on resuming to check that everything that
happened during suspend was expected.
It is not meant to check for errors we should hard-fail from, rather it
should be used as an opportunity for drivers such as amdsmu(4) to check
e.g. whether the previous suspend-to-idle actually entered a deep sleep
state (S0i3).
Reviewed by: olce
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D59740
acpi: Don't check suspend-to-idle if suspend failed
If we e.g. failed to suspend a device and suspend bounced because of
that, then we're not expected to have entered a deep sleep state in the
first place. In this situation, don't overload the user with irrelevant
information.
Reviewed by: olce
Fixes: 5f68acc931a4 ("acpi: Warn if no amdsmu(4) loaded after suspend-to-idle resume")
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D59731
route/fib_algo: Fix nexthop index collision across families
fib_algo indexes its idx->nhop array by the nexthop index with
assumption of its uniqueness. Which is true except for IPv4 over
IPv6 nexthops.
Give each index space its own segment within the same array and
offset the index by the segment base. Segments are created on demand
and sized independently, so the rib's own family keeps base 0 and
tables without cross-family nexthops index exactly as before.
Reviewed by: melifaro
Discussed with: markj
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D59552
bhyve: Add an option to prevent automatic restart in monitor mode
Monitor mode will currently unconditionally restart the guest if it
reboots, which can be incovenient in certain situations (e.g., resetting
the guest after a panic).
Address this by introducing a way to prevent automatic guest restarts
in monitor mode.
Reviewed by: markj
Differential Revision: https://reviews.freebsd.org/D59123
db/hash: Harden hash(3) database code
The hash(3) database code does not validate the on-disk database header,
leaving it open to several OOB read and write vulnerabilities.
This change adds basic header validation and array bounds checking to
parts of the hash(3) code that can be manipulated by messing with the
database header.
Reviewed by: kevans
Sponsored by: Klara, Inc.
MFC after: 1 month
Differential Revision: https://reviews.freebsd.org/D58822
man.cgi: remove stale manpage releases
Remove some outdated manual pages releases which are not really
supported by man.cgi: Slackware, SLES, Ubuntu interim releases.
devel/llvm22: fix arm64 and FLANG plists
In the arm64 case, add a missing file. In the FLANG case, delete an
empty directory.
PR: 298493, 298511
Reported by: yuri
