llvm-*: Move all LLVM_BINUTILS symlinks to toolchain package
Some of the LLVM binary utilities were included in the Clang package
(because they did not set an explicit PACKAGE).
Add a new Makefile under clang/toolchain to create the symlinks and man
links for ar, c++filt, nm, and so on (without the llvm-* prefix) when
LLVM_BINUTILS is enabled (as it is by default).
PR: 293610
Reviewed by: bapt, ivy, brooks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D55692
shells/xonsh: Update to 0.23.3
- Temporarily switch to GitHub as release tarball on PyPI is missing files
necessary for unit tests
- Move prompt-toolkit to RUN_DEPENDS as it is required for the interactive shell
* Thanks to Andy Kipp for assisting in improving and fixing unit tests and
improving FreeBSD support
ChangeLog: https://github.com/xonsh/xonsh/releases/tag/0.23.3
15.1: On schedule
The releng/15.1 branch was created and 15.1-BETA1 builds started
on May 1, 2026.
Approved by: re (implicit)
Sponsored by: OpenSats Initiative
powerpc: refactor dmamap setup; free dmamap in error paths
* refactor the dmamap allocation / setup path for future code changes
to align with arm64 busdma code;
* free the dmamap allocation if the dma segment list can't be allocated;
* free the dmamap allocation during the busdma dmamem allocation path
if the actual memory allocation itself fails.
Locally tested:
* POWER9 ppc64le native boot, Raptor Engineering Blackbird
* POWER8 ppc6le, IBM POWER S822LC
https://reviews.freebsd.org/D56244
powerpc: busdma declaration shuffling
Shuffle some of the declarations around to match other implementations.
Expected to be a no-op.
Differential Revision: https://reviews.freebsd.org/D56243
powerpc: use local flags for bounce buffers instead of common flags
Migrate to using local flags for bounce buffers instead of the
common flag field.
Differential Revision: https://reviews.freebsd.org/D56208
powerpc: add in some busdma domain setup
* Implement the basic tag domain set routine
* Set the domain to the parent domain if provided
This is just plumbing for eventual work to re-allocate things into
the currently configured domain.
Differential Revision: https://reviews.freebsd.org/D55315
powerpc: create a tag with the parents implementation if supplied
If a parent tag is supplied then use its implementation.
Differential Revision: https://reviews.freebsd.org/D55314
powerpc: initial straight port of busdma_machdep.c -> busdma_bounce.c
This is a straight port of the code and doesn't yet handle
different implementations (which will be in a subsequent commit.)
Locally tested:
* G5 SMP (2x PPC970mp)
* power8 / power9 pseries QEMU VM
* power8 powernv
Differential Revision: https://reviews.freebsd.org/D55313
net/libpfctl: add 15.1, remove 13.5
13.5 is end-of-life, remove it from this port.
Add 15.1, which is in the process of being released.
Deliberately do not bump PORTVERSION, because we don't change existing
libpfctl versions, we only add a new supported FreeBSD version.
PR: 294871
Sponsored by: Rubicon Communications, LLC ("Netgate")
namei: Preserve ABI root for absolute symlinks before fallback
D40479 changed namei() so that an absolute symlink target
encountered during an ABI-root lookup restarts from the native root.
This helps the native fallback case, but it also makes successful
lookups inside an ABI root escape that root while following absolute
symlinks.
Only switch absolute symlink lookup to the native root after namei()
is already in the restarted/native fallback pass.
Do not mark the lookup as restarted merely because an absolute
symlink was encountered while still resolving inside the ABI root.
This preserves the intended native fallback behavior while keeping
absolute symlinks within a successfully resolved ABI-root path in
the ABI namespace.
Signed-off-by: Ricardo Branco <rbranco at suse.de>
PR: 289739
[3 lines not shown]
sysutils/beats*: fix build on armv7
The patch "patch-go-sysinfo" ports a Linux-only go module to FreeBSD.
Due to two oversights, the patch did not build on armv7. Fix the patch,
ensuring that all beats versions build on armv7 FreeBSD. A subtle bug
is addressed, too: the patch would previously convert from microseconds
to nanoseconds for time.Unix() by multiplying with time.Microsecond,
which is not the correct value. Multiply with 1000 instead.
Approved by: portmgr (build fix blanket)
MFH: 2026Q2
(cherry picked from commit 22c1284107218bffffc201881f84cb01ed33b141)
devel/cvs-devel: fix build on armv7
Same issue as was previously fixed in 36fcf661b for devel/blame.
A 64 bit time_t with a 32 bit long confuses the gnulib mktime.c
implementation. Apply the same upstream patch to fix the build.
Obtained from: 36fcf661ba23c96004395f6ee52cf2e6133b31f0
MFH: 2026Q2
Approved by: portmgr (build fix blanket)
(cherry picked from commit 978896a191d705a495efbaf98f3d8e1e47b68ef8)
archivers/archmerge: NEW PORT, replaces archivers/zipmix
Different compression programs result in different
compression ratios and compress different types of files
unequally.
Archmerge will take the best files from each archive and
create a new output file, which is guaranteed to be equal
to, or smaller in size than both of the input files. Archmerge
can also be used to perform boolean operations on archives,
such as merging 2 archives together, or updating a new file.
NOTE: in it's current incarnation. Archmerge only operates on
zip archives. Additional compression algorithms and techniques
will be added soon.
WWW: https://codeberg.org/BSDforge/archmerge
PR: 294920, 293749
devel/cvs-devel: fix build on armv7
Same issue as was previously fixed in 36fcf661b for devel/blame.
A 64 bit time_t with a 32 bit long confuses the gnulib mktime.c
implementation. Apply the same upstream patch to fix the build.
Obtained from: 36fcf661ba23c96004395f6ee52cf2e6133b31f0
MFH: 2026Q2
Approved by: portmgr (build fix blanket)
sysutils/beats*: fix build on armv7
The patch "patch-go-sysinfo" ports a Linux-only go module to FreeBSD.
Due to two oversights, the patch did not build on armv7. Fix the patch,
ensuring that all beats versions build on armv7 FreeBSD. A subtle bug
is addressed, too: the patch would previously convert from microseconds
to nanoseconds for time.Unix() by multiplying with time.Microsecond,
which is not the correct value. Multiply with 1000 instead.
Approved by: portmgr (build fix blanket)
MFH: 2026Q2
Revert 00bee6fcd77f, which partially reverted libc++ commit aa7f377c965c
After base 966fb94cb357, this revert is no longer necessary: stdint.h
will unconditionally define macros such as `SIZE_MAX`, `UINT64_C`, and
others.
Submitted by: Nikolas Klauser <nikolasklauser at berlin.de>
MFC after: 1 month
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D56746
Define stdint.h macros unconditionally
Similar to glibc, define all the stdint.h macros such as `SIZE_MAX`,
`UINT64_C`, etc unconditionally. I.e. no longer check whether
`__STDC_CONSTANT_MACROS` or `__STDC_LIMIT_MACROS` are defined. See also
<https://sourceware.org/bugzilla/show_bug.cgi?id=15366>.
This is part of reverting base 00bee6fcd77f, which reverted an upstream
libc++ commit that eliminated libc++'s stdint.h wrapper header.
Submitted by: Nikolas Klauser <nikolasklauser at berlin.de>
MFC after: 1 week
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D56746
