FreeBSD git weekly: 2026-03-16 to 2026-03-22

We generally use TOOLS_PREFIX being set to indicate that we are building
in the cross-build tools stage.  This check was missing for llvm.prog.mk
consumers other than *tblgen, which have not previously been built as
cross tools.

Reviewed by:    dim
Sponsored by:   The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D55930

* Try to make the RETURN VALUES section flow better.

* Add basename(3), dirname(3), free(3) to the SEE ALSO section.

* Drop the CAVEATS section, which was obsolete the moment realpath(3)
  was added to the Single Unix Specification in 1994.

MFC after:      1 week
Sponsored by:   Klara, Inc.
Reviewed by:    kevans
Differential Revision:  https://reviews.freebsd.org/D55928

Fixes:        https://cgit.freebsd.org/src/commit/?id=1a7151f79664 cam: Add probes for xpt actions
Event:  AsiaBSDCon 2026

The previous S1-S5 options are still accepted for compatibility, but
they are now deprecated in favour of the new generic sleep types.

Reported by:    markj
Reviewed by:    markj
Approved by:    markj
Fixes:          https://cgit.freebsd.org/src/commit/?id=97d152698f48 ("acpi: Use sleep types defined in sys/power.h")
Event:          AsiaBSDCon 2026
Sponsored by:   The FreeBSD Foundation
Differential Revision:  https://reviews.freebsd.org/D56008

A race condition in the MII layer causes spurious link down events.
In `statchg`, on link down, check if the PHY reports the link as
actually down using the BMSR register, if not, force the status of the
link to back up and restart TX. Do the same in a MII `linkchg` handler.
On actual link up, restart TX in case it went idle and down.

PR:             https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=252165
Signed-off-by:  Rafael Kitover <rkitover@gmail.com>
Reviewed by:    pouria
Differential Revision: https://reviews.freebsd.org/D55682

The private struct has a 'regs' member that's only written to
and otherwise unused. This change removes it.

Fixes:                  https://cgit.freebsd.org/src/commit/?id=b3d9e5013f3e5 (Don't active memory space)
Requested by:           jhb
Reviewed by:            imp, jhb, jrtc27
Differential Revision:  https://reviews.freebsd.org/D55817

The goal this change is to remove the use of the tag and handle
needed by bus_space_[read|write]_4.

Fixes:                  https://cgit.freebsd.org/src/commit/?id=b3d9e5013f3e5 (Don't active memory space)
Requested by:           jhb
Reviewed by:            gallatin, imp, jhb, jrtc27
Differential Revision:  https://reviews.freebsd.org/D55818

A small enhancement.

Reviewed by:    Ali Mashtizadeh <ali@mashtizadeh.com>
Sponsored by:   The FreeBSD Foundation
Differential Revision:  https://reviews.freebsd.org/D55830

This change allows for fine-grained capabilities per counter index. This
is particularly useful for AMD where subclasses are not exposed to the
general PMC code, but other architectures also have asymmetric behaviors
when it comes to specific counter indices.

A new PMC_OP_GETCAPS op is added to the hwpmc(4) ioctl interface.

Reviewed by:    mhorne
Sponsored by:   Netflix
Pull Request:   https://github.com/freebsd/freebsd-src/pull/2058

Bump for the addition of PMC_OP_GETCAPS and the recently added Intel
CPUs.

Sponsored by:   The FreeBSD Foundation

Some platforms enforce power-of-two alignment for bus_dma tags. Rounding up
the natural size may result in over-alignment, but should be safe.

PR:             https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=293770
Reviewed by:    andrew
Fixes:          https://cgit.freebsd.org/src/commit/?id=c499ad6f997c ("virtio: Use bus_dma for ring and indirect buffer allocations")
Sponsored by:   Arm Ltd
Differential Revision:  https://reviews.freebsd.org/D55843

Rather than each exception calling userret use a common copy. As
syscallret already calls userret we need to skip it in that case.

Reviewed by:    kib
Sponsored by:   Arm Ltd
Differential Revision:  https://reviews.freebsd.org/D55250

- Display an appropriate alert character instead of a question mark for
  unknown bits.
- Display all fatal bits for a register and not just the ones that are
  currently set.

MFC after:      1 week
Sponsored by:   Chelsio Communications

The layout for the main INT_CAUSE is a bit different for T7.

MFC after:      1 week
Sponsored by:   Chelsio Communications

The TPM driver currently has a single buffer per instance to hold the
result of a command, and does not allow subsequent commands to be sent
until the current result is read by the same OS thread that sent the
command, with a timeout to throw away the result after a while if the
result is not read in a timely fashion.  This has a couple problems:

 - The timeout code has a bug which causes all subsequent commands to
   hang forever if a different OS thread tries to read the result
   before the OS thread which sent the command, and the OS thread
   which sent the command never tries to read the result.

 - Even if the first problem is fixed, applications expect to be able
   to read the result from a different OS thread than the OS thread
   which sent the command. The particular case that we saw was a go
   application where the go runtime scheduled the goroutine which read
   the result to a different OS thread from one where the goroutine
   that sent the command ran, and there's no way to force these to
   always run on the same OS thread.

Fix all of this by replacing the global result buffer with a per-open
result buffer via devfs_set_cdevpriv(), so that we no longer need to
block subsequent commands until the results of a previous command are
retrieved or care about which OS thread is reading the result of a
command.

Sponsored by:   Netflix
Reviewed by:    olivier, imp
Differential Revision:  https://reviews.freebsd.org/D52328

Reviewed by:  jhb, markj
Sponsored by:   The FreeBSD Foundation
MFC after:      1 week
Differential revision:  https://reviews.freebsd.org/D55809

This change only checks that the values from the user-supplied context
are not truncated by C implicit type convertions.  The validity of the
segment selectors is still checked by hardware.

Reviewed by:    jhb, markj
Sponsored by:   The FreeBSD Foundation
MFC after:      1 week
Differential revision:  https://reviews.freebsd.org/D55861

as provided by the Intel document 346446-009 AKA FRED 8.0 and SDM v90,
February 2026.

Note that the layout of the struct trapframe does not change with
FRED. The differences are in addition of two fields at the end of the
structure. Other members are carved from existing tf_cs and tf_ss by
limiting hw-written segments to 16 bits officially, and reusing rest
from the doubleword for the new data.

Reviewed by:    jhb, markj
Sponsored by:   The FreeBSD Foundation
MFC after:      1 week
Differential revision:  https://reviews.freebsd.org/D55831

We no longer put pcb on stack, it is part of the struct thread.
Similarly, we do not put user fpu save area on stack.  There is no
constraints on the pcb size due to the XSAVE area required alignment.

Reviewed by:    jhb, markj
Sponsored by:   The FreeBSD Foundation
MFC after:      1 week
Differential revision:  https://reviews.freebsd.org/D55931

Sponsored by: The FreeBSD Foundation
MFC after:      1 week

Sponsored by: The FreeBSD Foundation
MFC after:      1 week

Reported by:  markj
Fixes:          https://cgit.freebsd.org/src/commit/?id=97d152698f48 ("acpi: Use sleep types defined in sys/power.h")
Event:          AsiaBSDCon 2026
Sponsored by:   The FreeBSD Foundation

We don't need to check for NULL before calling free().

Sponsored by:   The FreeBSD Foundation

Trying to use the grown struct trapframe for IDT case broke in cases
where code supposed that hardware consumed sizeof(struct trapframe) of
the stack space when delivering interrupt or exception.  In particular,
this was broken for #NM/#DB/#MC.  Naive attempt of using IDT-trapframe
size for stack consumption caused later problems with larger C type.

Instead of pretending that IDT event delivery pushed two never-accessed
doubleword to the stack, keep it honest and provide separate type for
the FRED interrupt frame, i.e. struct trapframe_fred.

Convert between trapframe_fred and trapframe can be done by trivial pointer
arithmetic.

Sponsored by:   The FreeBSD Foundation
MFC after:      1 week

We separate the code of CPPC and legacy pstate driver to make it easier
to read.

Reviewed by:    olce
Approved by:    markj (mentor)
Sponsored by:   The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D55604

Added optional system tunable parameter to enable
4-tuple rss udp hashing.

Signed-off-by: bigJ <bigj@solanavibestation.com>
Reviewed by: adrian, pouria
Pull Request: https://github.com/freebsd/freebsd-src/pull/2057

It should be decreased only when the interface has been successfully
removed from the "active" list.

This prevents vnet_if_return() from potential OOB writes to the
allocated memory "pending".

Reviewed by:    kp, pouria
Fixes:          https://cgit.freebsd.org/src/commit/?id=a779388f8bb3 if: Protect V_ifnet in vnet_if_return()
MFC after:      3 days
Differential Revision:  https://reviews.freebsd.org/D55873

The ioctls SIOCSIFVNET and SIOCSIFRVNET are for userland only. For
SIOCSIFVNET, if_vmove_loan(), the interface is obtained from current
VNET. For SIOCSIFRVNET, if_vmove_reclaim(), a valid child prison is
held before getting the interface. In both cases the VNET of the
obtained interfaces is stable, so there's no need to check it.

No functional change intended.

Reviewed by:    glebius, jamie (for #jails)
MFC after:      2 weeks
Differential Revision:  https://reviews.freebsd.org/D55828

bridge tries to run callout_drain(9) twice under epoch
during destruction.
once for bridge_timer, which is not required to be under epoch.
second time for the BSTP callout, which is already disabled
earlier inside bridge_delete_member.

Reviewed by: glebius, zlei
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D55876

During link-layer address change event, don't send unsolicited
NA for multicast addresses.

Reviewed by: adrian, zlei
Differential Revision: https://reviews.freebsd.org/D55885

Update ECN tunneling functions from obsolete RFC 3168 to
newer RFC 6040.
Also, add ECN_COMPLETE to support dangerous packet reporting
without causing extra costs to existing caller functions.
Finally, return values are specified as macro to reduce
confusion, considering extra return values for ECN_WARN
and ECN_ALARM were added.

Reviewed By: glebius, tuexen
Differential Revision: https://reviews.freebsd.org/D53516

Multiple delayed NAs on the same ifa can occur simultaneously.
Therefore:
* Differentiate between GRAND and solicited replies.
* Cancel previous pending GRAND NA for same ifa.
* Reuse ndq memory for GRAND.
* Free non-GRAND replies immediately.
* Don't limit non-GRAND NAs.

Reviewed by: glebius
Differential Revision: https://reviews.freebsd.org/D55905

bridge doesn't require to enter epoch during destruction.

Reviewed by: zlei, glebius
Differential Revision: https://reviews.freebsd.org/D55935

IPv4-compatible IPv6 addresses are deprecated by RFC 4291.
No functional change intended.

Reviewed by: glebius, emaste
Differential Revision: https://reviews.freebsd.org/D55387

The tcp_close() will perform both actions.

Reviewed by:            pouria, tuexen, rrs
Differential Revision:  https://reviews.freebsd.org/D55965

Reviewed by:          tuexen, rrs
Differential Revision:  https://reviews.freebsd.org/D55969

Implement the PMC_CAP_SYSWIDE capability flag in the same manner as the
PMC_CAP_DOMWIDE flag.  This flag was previously introduced along side
the domain wide flag, but is yet to be used anywhere.

Reviewed by:    mhorne
Sponsored by:   Netflix
Pull Request:   https://github.com/freebsd/freebsd-src/pull/2075

It has been broken and disabled for over 10 years.  Remove mentions of the
kvm(3) from the manual page.

Determine the FAT (MS-DOS) type (FAT32, FAT16, or FAT12) and partition
scheme (fat32lba or fat16b) based on partition size.

Accept any (NetBSD) strsuftoll(3)-compatible string, as the value will
be fed to makefs(8) "-s" in a future commit.

NANO_SLICE_FAT_SIZEs defined in sectors (ending with an "s") will not be
supported.  As a workaround, the letter "b" (blocks) may be used.

Reviewed by:    imp
MFC after:      1 week
Differential Revision:  https://reviews.freebsd.org/D54975

Rely on makefs -t msdos to create the MS-DOS partition and remove the
dependency on filesystems/mtools.

Note that NANO_SLICE_FAT_SIZEs defined in sectors (ending with an "s")
are no longer supported, as a workaround the letter "b" (blocks) may be
used instead.

Reviewed by:    imp
MFC after:      1 week
Differential Revision:  https://reviews.freebsd.org/D54976

nanobsd: embedded: Fix more regressions from last cleanup

Commit c99bb5747f5e changed "()" grouping into "{}", but in these cases
we chdir and depended on the subshell not changing the directory for the
caller.  Restore the old behavior.  While things appear to work without
this change, the previous code was intentionally structured this way.

Reviewed by:    imp
Fixes:          https://cgit.freebsd.org/src/commit/?id=c99bb5747f5e ("nanobsd: minor formatting cleanup")
MFC after:      1 week
Differential Revision:  https://reviews.freebsd.org/D54978

nanobsd: Minor cleanup

No functional change intended.

Reviewed by:    imp
MFC after:      1 week
Differential Revision:  https://reviews.freebsd.org/D54979

nanobsd: embedded: Time stamp the code image

While there is a function nano_makefs() that already takes care of this
issue, we are not yet ready to consume it.

Just add the time stamp and FFS options to makefs.  This is done
primarily to use UFS2 (UFS version 2).  If the user needs further
customization, the variable NANO_MAKEFS can be overridden.  Timestamping
takes care of making the images bit-for-bit identical.

Reviewed by:    imp
MFC after:      1 week
Differential Revision:  https://reviews.freebsd.org/D54980

Pointed out by: Alex "the PS/2 guy"

A loader built with MD_IMAGE_SIZE is almost always meant for use with
its embedded image and should try that as currdev before anything else.
Recent changes (d69fc3a9dc71, 784150fd2535) seem to have relaxed the ZFS
code's search for a rootfs and exposed this problem.

Reviewed by:    imp, tsoome
MFC after:      1 week
Sponsored by:   Chelsio Communications
Differential Revision:  https://reviews.freebsd.org/D55979

GEOM journaling shouldn't be used at the same time as soft updates, so
don't enable soft updates if GEOM journaling has been requested, and
error out if both are explicitly requested.

MFC after:      1 week
PR:             https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=293896
Reviewed by:    mckusick
Differential Revision:  https://reviews.freebsd.org/D55999

GEOM journaling shouldn't be used at the same time as soft updates, so
don't enable one if the other is already in use.

MFC after:      1 week
PR:             https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=293896
Reviewed by:    mckusick
Differential Revision:  https://reviews.freebsd.org/D56002

Ignoring the error may cause pfctl(8) to load inconsistent ruleset
preventing pf(4) to enforce desired policy.

Issue reported and fix suggested by berts _from_ fastmail _dot_ com

'Looks good.' @deraadt

MFC after:      1 week
Obtained from:  OpenBSD, sashan <sashan@openbsd.org>, 9fd28a8cca
Sponsored by:   Rubicon Communications, LLC ("Netgate")

* Save prev_len after having checked for and appended a trailing slash,
  not before.  This requires us to back up if we end up returning a
  partial result, but previously we would sometimes return a partial
  result with a trailing slash and sometimes without.

* Replace strlcat() with a faster strlcpy() since we know exactly how
  far into the buffer we are.

MFC after:      1 week
Sponsored by:   Klara, Inc.
Reviewed by:    kevans
Differential Revision:  https://reviews.freebsd.org/D55914

An NFSv4.1/4.2 client can set/clear the archive, hidden
and system flags when creating non-regular files, such
as directories.

Without this patch, the setting of va_flags causes an
EPERM failure, since they are specified for VOP_MKDIR(),
VOP_MKNOD() and VOP_SYMLINK().

This patch sets va_flags == VNOVAL for the above VOP_xxx()
calls and then sets/clears the flags after creation,
which fixes the problem.

This bug only affects the Windows NFSv4.1/4.2 client.

PR:     https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=293691
Tested by:      Dan Shelton <dan.f.shelton@gmail.com>
MFC after:      2 weeks

Notable upstream pull request merges:
 #17000 -multiple cmd/zfs: support creating parents with canmount=no
 #17926 b481a8bbb Make zpool status dedup table support raw bytes -p output
 #18238 -multiple Report space metrics per allocation class
 #18240 5dad9459d Add --no-preserve-encryption flag
 #18257 1eace5906 libzfs: use mount_setattr for selective remount including
                  legacy mounts
 #18259 1e2c94a04 More consistent use of TREE_* macros in AVL comparators
 #18262 c58b8b7dc Fix check for .cfi_negate_ra_state on aarch64
 #18264 f109c7bb9 Add the --file-layout (-f) option to zdb(8)
 #18268 304de7f19 libzfs: handle EDOM error in zpool_create
 #18275 753f1e1e2 zstream: add a drop_record subcommand
 #18276 0f90a797d Fix vdev_rebuild_range() tx commit
 #18277 1e1d64d66 Fix log vdev removal issues
 #18278 06b0abfe6 Fix the send --exclude option to work with encryption
 #18281 e73ada771 libzpool: lift zfs_file ops out to separate source file
 #18284 d97945776 zstream: consolidate shared code
 #18285 -multiple Separate SIMD compile checks for userspace and kernel
 #18288 c5905b2cb Implement lzc_send_progress
 #18289 -multiple L2ARC: Add depth cap and write budget fairness for
                  persistent markers
 #18290 d45c8d648 FreeBSD: Improve dmesg kernel message prefix
 #18294 b403040c4 draid: fix data corruption after disk clear
 #18296 -multiple libzfs: mnttab cleanup
 #18300 d35951b18 zpool clear: remove undocumented rewind flags
 #18301 ae7fcd5f9 fix libzfs diff mem leak in an error path
 #18303 65165df12 zilstat: add man page
 #18314 7f65e04ab libzfs: scrub: only include start and end nv pairs if
                  needed for scrub
 #18315 -multiple Sync ZVOLs block cloning conditions with file systems
 #18330 f80338fac zarcsummary: add man page
 #18337 f259a47c7 zpool-iostat.8: clarify first report shows per-second
                  averages

Obtained from:  OpenZFS
OpenZFS commit: 3ee08abd2fcef3123188e9ad61a9cd80b9b7831c

The current implementation briefly violates the tail invariant.  This
is not usually an issue, but if an insert is in flight when a panic
occurs, we may then trip the invariant while dumping core.

MFC after:      1 week
Sponsored by:   Klara, Inc.
Sponsored by:   NetApp, Inc.
Reviewed by:    obiwac, olce, jhb
Differential Revision:  https://reviews.freebsd.org/D55819

Include index values and object pointer in the diagnostic
to improve debugging of pindex mismatches.

No functional change.

Signed-off-by: Isa Isoux <isa@isoux.org>
Reviewed by: kib, pouria
Pull Request: https://github.com/freebsd/freebsd-src/pull/2038

Each invocation of seq_read invokes the seq_file.show callback which
writes into the sbuf.  Then it invokes sbuf_finish before copying the
data into the caller's buffer.  Without this, a second call to
seq_read on the same file would try to append data to a finished sbuf.

Reviewed by:    bz
Sponsored by:   AFRL, DARPA

Reviewed by:  bz
Sponsored by:   AFRL, DARPA
Differential Revision:  https://reviews.freebsd.org/D55879

The Linux file_operations API expects the read and write operations
to take a single user buffer pointer (along with the length and the
file offset as an in/out parameter).

However, the debugfs_fill function was violating this part of the
contract as it was passing down kernel pointers instead.  An earlier
commit (5668c22a13c6befa9b8486387d38457c40ce7af4) hacked around this
by modifying simple_read_from_buffer() to treat its user pointer
argument as a kernel pointer instead.  However, other commits keep
tripping over this same API mismatch
(e.g. 78e25e65bf381303c8bdac9a713ab7b26a854b8c passes a kernel pointer
to copy_from_user in fops_str_write).

Instead, change debugfs_fill to use the "raw" pseudofs mode where the
uio is passed down to directly to the fill callback rather than an
sbuf.  debufs_fill now iterates over the iovec in the uio similar to
the implementation of uiomove invoking the read or write operation on
each user pointer.

This also fixes a tiny bug where the initial file offset from
uio_offset was ignored.  Instead, the operations were always invoked
with a file offset of 0.

As part of this, revert the the changes to simple_read_from_buffer()
from commit 5668c22a13c6befa9b8486387d38457c40ce7af4.

Also as part of this, the simple_attr_read/write methods and seq_read
now also need to accept and handle user pointers (also matching the
API in Linux).

For simple_attr_write*(), copy the user buffer into a kernel buffer
before parsing.  Also, do not permit writes at an offset as it's
unclear what the semantics for those would even be (perhaps you would
write out the formatted value into a buffer first and then allow the
copy_from_user to overwrite/extend that buffer and then re-parse the
integer value?).  The old handling of *ppos for writes was definitely
wrong before and only worked for an offset of 0 anyway.

Reviewed by:    bz
Sponsored by:   AFRL, DARPA
Differential Revision:  https://reviews.freebsd.org/D55833

I noticed that the buf_size < 0 check can never be true (it's a
size_t) and decided to check for this condition by an alternate
expression, and I also noticed that a read_size of 0 would incorrectly
return -EFAULT.  Instead, return success for both of these cases as
reading beyond the EOF of a normal file also returns EOF, not EINVAL.

Reviewed by:    bz
Sponsored by:   AFRL, DARPA
Differential Revision:  https://reviews.freebsd.org/D55845

Near the top of kern_mmap() that implements the mmap(2) system call,
it sets

        prot = PROT_EXTRACT(prot);

with

So prot can only be the three PROT_ flags.

The following test of the user's mmap(2) parameters (near line 275
in vm/vm_mmap.c):

        if (prot != PROT_NONE &&
            (prot & ~(PROT_READ | PROT_WRITE | PROT_EXEC)) != 0) {
                return (EXTERROR(EINVAL, "invalid prot %#jx", prot));
        }

can never fail. This commit deletes it.

No functional change intended.

Reviewed by:  kib
MFC-after:    1 week
Sponsored by: Netflix

Reviewed by:  kib
Differential Revision: https://reviews.freebsd.org/D55888
MFC-after:    1 week
Sponsored by: Netflix

When a timerfd interval callout misses its scheduled activation time, a
differential is calculated based on the actual activation time and the
scheduled activation time. This differential is divided by the timerfd's
interval time and the quotient is added to the timerfd's counter.

Before this change, the next callout was scheduled to activate at:
scheduled activation time + timerfd interval.

This change fixes the scheduling of the next callout to activate at:
actual activation time + timerfd interval - remainder.

Reviewed by:            markj
Differential Revision:  https://reviews.freebsd.org/D55790
MFC after:              2 weeks

When converting from timespec to sbintime, the timespec's 64-bit tv_sec
component is shifted to the left 32 bits, causing any information in the
upper 32 bits to be lost.

This data loss during conversion can turn timespecs with very large
tv_sec counters into sbintimes that represent much smaller time
durations.

Add tstosbt_sat() and tvtosbt_sat(), which are saturating versions of
tstosbt and tvtosbt. With these routines, any overflow resulting from
the conversion is clamped to [-SBT_MAX - 1, SBT_MAX].

Reviewed by:            imp, markj
Differential Revision:  https://reviews.freebsd.org/D55791
MFC after:              2 weeks

timerfd: Use saturating sbintime conversions

Some timerfd consumers set expirations with timespec tv_sec components
larger than 2^31 - 1. In such cases, converting that timespec to
sbintime results in data loss or sign flip, yielding a shorter
expiration than desired.

To avoid this problem, use saturating timespec-to-sbintime conversion
functions. These will clamp the converted sbintime to SBT_MAX under
circumstances where the normal conversion functions would overflow.

Saturating conversions still result in data loss, but the consequences
are less severe, causing problems only after SBT_MAX (~68 years) of
system uptime elapses.

Reviewed by:            imp
Differential Revision:  https://reviews.freebsd.org/D55792
MFC after:              2 weeks

timerfd: Wake up on discontinuous jump

If a discontinous realtime clock change occurs and sets any TFD_JUMPED
bits on the timerfd, then wake up waiting readers. This fixes failures
from the timerfd_root__clock_change_notification test case.

MFC after: 2 weeks

timerfd: Suppress kqueue readability after jump read

Do not report EVFILT_READ after reading a discontinuous clock jump.
This makes the kqueue filter consistent with Linux epoll behavior
and timerfd_poll(), which already checks tfd_jumped != TFD_READ before
reporting POLLIN.

MFC after: 2 weeks

Change SS_FATAL to SS_FAIL for SCSI sense code 0x3A (Medium not present)
to suppress console spam when card readers or other removable media
devices have no media inserted.

The error is still returned to callers (ENXIO), but no longer prints to
dmesg during boot/probe.

Verified with Apple SD Card Reader - reduced boot messages from 4 to 2.

Reviewed by:    emaste, imp
Differential Revision: https://reviews.freebsd.org/D55287

Requested by: jhb
Sponsored by:   The FreeBSD Foundation
MFC after:      1 week

ZSTDIO is not just for core dumps, and it is now required by ZFS, so
move it (and GZIO, which is still just for dumps) to a new “Compression
support” section in GENERIC, and add them both to MINIMAL.

MFC after:      1 week
Reviewed by:    imp
Differential Revision:  https://reviews.freebsd.org/D55940

Saying that we are called with an (un)finished or corrupted sbuf
is like saying "error" (though [un]finished is an extra hint).

Add the pointer to the sbuf so one could easily check it,
add the flags we are checking so one could see quickly about
a possible state mismatch.

Given we already dereference the pointer in the KASSERT there is no
extra harm adding it to the message.

Found while:    testing lindebugfs changes
Sponsored by:   The FreeBSD Foundation
MFC after:      3 days
Reviewed by:    jhb
Differential Revision: https://reviews.freebsd.org/D55919

Compiling drm-kmod on !X86 does not include asm/smp.h which includes
preempt.h on FreeBSD.  In order to compile drm-kmod on other
architectures add the secondary #includes for preempt.h to
spinlock.h and hardirq.h (which now also gets included from highmem.h)
to connect the #include chain.

Sponsored by:   The FreeBSD Foundation
MFC after:      3 days
PR:             https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=279864
Reviewed by:    jhibbits, emaste
Differential Revision: https://reviews.freebsd.org/D55974

ecn.9: Add ecn(9) manual

Add ecn manual for ip_ecn tunneling functions.

Reviewed by: pauamma_gundo.com, tuexen
Differential Revision: https://reviews.freebsd.org/D53517

ecn.9: Replace non-ASCII characters

Reported by:    maxim
Fixes:          https://cgit.freebsd.org/src/commit/?id=ba4298b05564 ("ecn.9: Add ecn(9) manual")

Sponsored by:   The FreeBSD Foundation
MFC after:      3 days

Sponsored by: The FreeBSD Foundation
MFC after:      3 days

Reported by:  vexeduxr, jrtc27
Reviewed by:    imp
Sponsored by:   The FreeBSD Foundation
Pull request:   https://github.com/freebsd/freebsd-src/pull/2084
Differential Revision: https://reviews.freebsd.org/D55923

Create visibility in apropos and the Hardware Release Notes.

MFC after:              3 days
Reviewed by:            pauamma, ray
Differential Revision:  https://reviews.freebsd.org/D53287

Sponsored by: Netflix
Reviewed by:    imp
Differential Revision:  https://reviews.freebsd.org/D55335

Because of this setting we were still using ELF Tool Chain tools for
buildworld.  The sets of binary utilities are largely equivalent and
this went unnoticed after commit 1cae7121c667 ("Enable LLVM_BINUTILS
by default").

This was discovered recently because ELF Tool Chain objcopy produces
standalone debug files without phdrs and this caused an issue with a
3rd party ELF parser [1].  Remove the forced setting so that we use
LLVM's binutils to build the system.

[1] https://sourceware.org/bugzilla/show_bug.cgi?id=33876

Re-commit after fixing a bootstrapping issue with LLVM binutils (in
17494c6e6b7d "build: Boostrap LLVM_BINUTILS for cross-tools").

Reviewed by:    imp, jhb
Sponsored by:   The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D55650

The FreeBSD-clang package on amd64 contains libclang_rt.asan-i386.so,
which links to the 32-bit version of libgcc_s.so.1.  It is not clear
if that file belongs in FreeBSD-clang or if it should be placed into
a different package.

For unknown reasons, pkg *sometimes* recognizes this and decides that
it needs to install a package to supply libgcc_s.so.1:32.

In particular, when we initially install the VM with 'pkg install
[long list of FreeBSD-* packages]', pkg does not recognize that it
wants this; but the *next* time 'pkg install' runs, it decides that
it needs libgcc_s.so.1:32 -- even though that particular library is
not needed by the particular package we're trying to install -- and
goes looking for it... and ends up picking gcc12-devel as a provider.

Later in the EC2 image building process, we run 'pkg autoremove' (which
was added during 15.0 to get rid of the "bogus dependencies" which were
added while installing other packages) and pkg is once again confused:
It decides that gcc12-devel is an automatic port which is not required
by anything else, but upon determining that it should be removed it
then determines that FreeBSD-clang depends upon a port which is being
removed and deletes FreeBSD-clang as well.

It's not clear what is breaking in pkg, but as a temporary workaround
add the FreeBSD-clibs-lib32 package to the "small" and "builder" EC2
image flavours in order to satisfy the libgcc_s.so.1 demand.

This commit should be reverted if libclang_rt.asan-i386.so moves out
of FreeBSD-clang, (i.e. if it becomes possible to install a 32-bit-free
amd64 system) or when the pkg issue is resolved (i.e. pkg consistently
recognizes the shared library dependency and installs clibs-lib32
automatically as a FreeBSD-clang dependency).

MFC after:      3 days
Sponsored by:   Amazon
Differential Revision:  https://reviews.freebsd.org/D55978

MFC after:    3 days
Sponsored by:   Amazon

This helps others attempting to customize the behavior of `release.sh`.

MFC after:      1 week
Differential Revision:  https://reviews.freebsd.org/D55450 (as part of a larger change)

The chroot_cleanup routine handles any cleanup needed post-chroot_setup,
etc. This consists of purely tearing down `${CHROOTDIR}/dev` today, but
might involve additional steps, as needed for custom functions. This
allows end-users to override the various chroot functions without having
to modify code in main() or replicate the unmount procedure in an
equivalent routine setup via the trap builtin.

This change modifies the /dev unmount process to use `umount -f` instead
of `umount`. The latter can result in failures if resources are still
mounted or are running post-build, whereas the former will clean up any
resources still in use by processes running in the chroot at time of
build. Moreover, the `chroot_cleanup` routine is now called when the
script is killed with `SIGINT` and `SIGTERM`, as well as at `EXIT`,
better ensuring that the script's resources are cleaned up in relatively common
scenarios that can be detected/handled.

MFC after:      1 week
Differential Revision:  https://reviews.freebsd.org/D55450 (as part of a larger change)

tests/netinet: provide test for raw_ip.c:rip_input()

The test creates differently bound SOCK_RAW sockets and sends different
probe packets.  There is expected results matrix that every probe is
checked against.

tests/netinet/raw: fix ssize_t printf(3) format

Fixes:  https://cgit.freebsd.org/src/commit/?id=c1481c9f4a425db7295332be87f837bfb7d127df

Take Jan Kokemuller's timerfd tests from the epoll-shim project,
stripping out code that isn't directly related to FreeBSD.

Reviewed by:            markj
Differential Revision:  https://reviews.freebsd.org/D55789
MFC after:              2 weeks

After patch 9d26b82, we don't provide recursive call protection anymore.
Therefore, to pass the test, we adjust the testcase by protecting on
caller and the testcase is to make sure the dtors is properly handled.

Reported by:    siva
Reviewed by:    kib
Approved by:    markj (mentor)
Fixes:  https://cgit.freebsd.org/src/commit/?id=9d26b82826d9 ("libc: Fix dtor order in __cxa_thread_atexit")
MFC after:      2 weeks
Sponsored by:   The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D55893

Sponsored by: Orange Business Services

OpenBSD fixed this separately (in their 6a338f3f70). We appear to not
have had this bug, but test for it anyway.

Sponsored by:   Rubicon Communications, LLC ("Netgate")

The function signature of `crypto_initreq()` was
incorrectly documented.

Signed-off-by:  Zixu Wu <zx@bv3.dev>
Reviewed by:    ziaee, pouria
Pull Request:   https://github.com/freebsd/freebsd-src/pull/2077

Remove an outdated comment about NanoBSD not being able to remove the
firstboot sentinel file.

It was fixed in cb51e2bc3db7 ("rc: Allow the removal of
firstboot_sentinel on read-only file systems").

Reviewed by:    imp
MFC after:      1 week
Differential Revision:  https://reviews.freebsd.org/D54977

This version is based on
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
24d479d26b25bce5faea3ddd9fa8f3a6c3129ea7 ( tag: v6.19-rc6 ).

Sponsored by:   The FreeBSD Foundation
MFC after:      3 days

ath12k: update Atheros/QCA's ath12k driver

This version is based on
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
05f7e89ab9731565d8a62e3b5d1ec206485eeb0b ( tag: v6.19 ).

Sponsored by:   The FreeBSD Foundation
MFC after:      3 days

ath11k: restore module Makefile

During the subtree merge and checking out the updated version of ath12k
parts of the previous ath11k merge were accidentally undone.
Retore the ath11k Makefile as well after 3c4eef5d838b restored the
driver bits.

Fixes:          https://cgit.freebsd.org/src/commit/?id=a96550206e4b ("ath12k: update Atheros/QCA's ath12k ..")
Sponsored by:   The FreeBSD Foundation
MFC after:      3 days

It seems that over the time of merging and only partly working
on making it compile again a few mis-merges happened.
Reduce diff to the v6.19 upstream version so that we only have
additions to the driver (usually guarded by #ifdef __FreeBSD__).

Sponsored by:   The FreeBSD Foundation
MFC after:      3 days

libcrypto: compile all PIC objects with -DOPENSSL_PIC

This change modifies the libcrypto PIC objects to always compile with
`-DOPENSSL_PIC` to restore parity with the upstream build process. This
ensures that `-DOPENSSL_PIC` is used whenever building the auxiliary
shared objects. In this case, just the legacy provider (we no longer
distribute the fips provider).

MFC after:              2 weeks
Differential Revision:  https://reviews.freebsd.org/D44896

Revert "libcrypto: compile all PIC objects with -DOPENSSL_PIC"

This commit broke the build with some build options.

Some validation needs to be done to confirm that moving the preprocessor
argument to `secure/lib/libcrypto/Makefile.inc` works without breaking
the build, but revert for now until a `tinderbox` run can be done with
the change.

PR:             https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=293934
Reported by:    Jenkins, Trond Endrestøl

This reverts commit 14b9955e57cc28b61e785165b9effcbe620edb46.

libcrypto: compile all PIC objects with -DOPENSSL_PIC -fPIC

This change modifies the libcrypto PIC objects to always compile with
`-DOPENSSL_PIC -fPIC` to restore parity with the upstream build process.
This ensures that the legacy provider is built with parity to the
upstream legacy provider.

MFC after:              12 days
Tested with:            `make check` (legacy provider), `make universe`
Fixes:                  https://cgit.freebsd.org/src/commit/?id=14b9955e
Differential Revision:  https://reviews.freebsd.org/D44896

This reverts commit 6a13aeac3c1f98db7cf156f24a4d6bc8d3c321f6.

The "bogus dependencies being installed" issue was traced down to pkg
confusion surrounding libclang_rt.asan-i386.so linking to a 32-bit
libgcc_s.so.1, and a more minimal workaround of "install clibs-lib32"
has been applied, so this hack is no longer required.

MFC after:      3 days
Sponsored by:   Amazon