routing: Retire ROUTE_MPATH compile option
The ROUTE_MPATH compile option was introduced to
test the new multipath implementation.
Since compiling it has no overhead and it's enabled
by default, remove it.
Reviewed by: melifaro, markj
Relnotes: yes
Differential Revision: https://reviews.freebsd.org/D55884
tests: Add grand link-layer event in netinet6/ndp
Add test for ndp to verity link-layer address change event
actually triggers the grand.
Differential Revision: https://reviews.freebsd.org/D55927
packages: Add dependency from rc to mtree
/etc/rc.d/var_run uses mtree, which is in the devel set and isn't
installed as part of either minimal or optional, so add a manual
dependency.
Reviewed by: des, bapt, emaste
Differential Revision: https://reviews.freebsd.org/D54540
Sponsored by: https://www.patreon.com/bsdivy
packages: Move xz to the minimal set
pkg(8), via its daily periodic script, requires xz. We don't have
a way to encode dependencies from ports on base packages right now,
so instead move xz to the minimal set so it's always installed.
This isn't an ideal solution, but it's justified in this case since
pkg(8) is always installed, so having its dependencies always
installed is acceptable.
(Following discusson on the diff, new versions of pkg have now been
changed to use zstd instead of xz, but we still think xz is useful
enough to keep in minimal.)
MFC after: 1 week (stable/15 only)
Reviewed by: des, bapt, emaste
Differential Revision: https://reviews.freebsd.org/D55630
Sponsored by: https://www.patreon.com/bsdivy
DEVICE_IDENTIFY.9: Fix function call to detect driver in example code
Fixes: ccabc7c2e556 ("DEVICE_IDENTIFY.9: Modernize description and use cases")
MFC after: 3 days
Sponsored by: The FreeBSD Foundation
tests/netlink: Bugfix on snl_parse_errmsg_capped test
Turn off the NETLINK_EXT_ACK flag to fix bug of snl_parse_errmsg_capped.
Reviewed by: pouria
Pull Request: https://github.com/freebsd/freebsd-src/pull/1660
ix(4): Add EEE support for E610 adapters
The ix driver now supports Energy Efficient Ethernet (EEE) on Intel
E610 devices. EEE allows the network interface to enter low-power
states during periods of low link utilization, reducing power
consumption while maintaining full performance when needed.
E610 adapters provide EEE support through BASE-T PHY functionality.
Due to this PHY-based implementation, EEE is supported only
on 2.5Gb speeds and above.
Signed-off-by: Yogesh Bhosale <yogesh.bhosale at intel.com>
Signed-off-by: Krzysztof Galazka <krzysztof.galazka at intel.com>
Authored-by: Yogesh Bhosale <yogesh.bhosale at intel.com>
Approved by: kbowling (mentor)
Tested by: Mateusz Moga <mateusz.moga at intel.com>
Sponsored by: Intel Corporation
[3 lines not shown]
Revert superfluous mis-MFC of "Add ASMC_DEBUG make option"
Revert commit 12e1ab887d58 ("Add ASMC_DEBUG make option, 2026-01-04"),
as the original commit it intended to MFC has already been MFCed as
ab3eaa6ea29d846d on 2026/02/22 (UTC). The reverted commit introduced
some (superfluous) duplicate lines and an unrelated change.
This is a direct commit to stable/14.
Fixes: 12e1ab887d58 ("Add ASMC_DEBUG make option, 2026-01-04")
Sponsored by: The FreeBSD Foundation
uart/pci: support 16550A PCI serial devices
Expand the current check to also attach the ns8250 driver to devices
reporting as 16550A. This has been tested to work on a real device.
From an inspection of the code in the ns8250 driver it seems like it should
support up to 16950A devices, but I don't have hardware to ensure that,
hence be conservative with the change.
MFC: 2 weeks
Reviewed by: imp
Differential revision: https://reviews.freebsd.org/D56095
uart/pci: always disable MSI for generic devices
The generic device pci_id structure in uart_pci_probe() already has
PCI_NO_MSI appended to it's flags, however that information is not
propagated into uart_pci_attach().
Assume that any device that doesn't match the known IDs is a generic UART
device, and hence prevent the usage of MSIs.
MFC: 2 weeks
Reviewed by: imp
Differential revision: https://reviews.freebsd.org/D56097
x86: move the NUM_ISA_IRQS symbol from atpic.c into x86/isa/icu.h
This is not the best location, but works for now.
Reviewed by: markj
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D56003
vm_fault: Avoid creating clean, writeable superpage mappings
The pmap layer requires writeable superpage mappings to be dirty.
Otherwise, during demotion, we may miss a hw update of the PDE which
sets the dirty bit.
When creating a managed superpage mapping without promotion, i.e., with
pmap_enter(psind == 1), we must therefore ensure that a writeable
mapping is created with the dirty bit pre-set. To that end,
vm_fault_soft_fast(), when handling a map entry with write permissions,
checks whether all constituent pages are dirty, and if so, converts the
fault to a write fault, so that pmap_enter() does the right thing. If
one or more pages is not dirty, we simply create a 4K mapping.
vm_fault_populate(), which may also create superpage mappings, did not
do this, and thus could create mappings which violate the invariant
described above. Modify it to instead check whether all constituent
pages are already dirty, and if so, convert the fault to a write fault.
Otherwise the mapping is downgraded to read-only.
[5 lines not shown]
kqueue: Fix a race when adding an fd-based knote to a queue
When registering a new kevent backed by a file descriptor, we first look
up the file description with fget(), then lock the kqueue, then see if a
corresponding knote is already registered. If not, and KN_ADD is
specified, we add the knote to the kqueue.
closefp_impl() interlocks with this process by calling knote_fdclose(),
which locks each kqueue and checks to see if the fd is registered with a
knote. But, if userspace closes an fd while a different thread is
registering it, i.e., after fget() succeeds but before the kqueue is
locked, then we may end up with a mismatch in the knote table, where the
knote kn_fp field points to a different file description than the knote
ident.
Fix the problem by double-checking before registering a knote. Add a
new fget_noref_unlocked() helper for this purpose. It is a clone of
fget_noref(). We could simply use fget_noref(), but I like having an
explicit unlocked variant.
[5 lines not shown]
