editors/pluma: switch to GitHub source
Switch from MATE mirror to GitHub tarball using USE_GITHUB and
GH_ACCOUNT=mate-desktop, add autoreconf, mate-submodules GH_TUPLE,
autoconf-archive, yelp-tools, common:build, patch configure.ac and
docs/reference/Makefile.am to remove gtk-doc, and remove stale API
doc plist entries
misc/shimmy: Update to 2.0.1
While here, delete options other than DOCS.
The upstream has switched its inference engine from llama.cpp to
airframe. So all llama.cpp-related options are dropped.
math/octave-forge-octave_boost: New port.
A comprehensive open-source high performance processing package for
GNU Octave based on Boost C++ Libraries. Provides direct access to
Boost.Accumulators statistical accumulators for online computation of
count, covariance, density, error of mean, extended P^2 quantiles,
kurtosis, max/min, mean, median, moments, P^2 cumulative distribution,
P^2 quantile, peaks over threshold, POT quantile, POT tail mean,
skewness, sum, and tail statistics, as well as Boost.Chrono for
high-resolution timing and Boost.Date_Time for date and time
manipulation and Boost.Thread for multithreading support.
nanobsd: Register the boot.config file in the metalog
Ensure that the boot.config file gets an entry in the metalog by
tgt_touch()ing it. If a file is not present in the metalog, it gets
excluded from the final image when using "-U" (unprivileged builds).
MFC after: 2 weeks
nanobsd: Adjust the order of the flags in makefs
The order of the flags matter in makefs(8). The -t (type) flag must come
before the -o (options) flag; otherwise, the options are reset.
Move the -t flag before the -o flag and remove the shim function
_xxx_adjust_code_size() that was created to align to the default makefs
FFS values.
It effectively prevented us from generating NanoBSD images using
unprivileged builds with the partitions internally aligned as intended.
Reviewed by: senguptaangshuman17_gmail.com, imp
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D57226
nanobsd: Use makefs -R flag to honor rounded sizes
Switch from the min/max size flag (-s) to the round-up flag (-R) when
invoking makefs(8).
Because the partition sizes passed to nano_makefs have already been
rounded up by calculate_partitioning(), using -s can cause makefs to try
to perform sizing adjustments that usually result in failures.
Reviewed by: imp
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D57225
nanobsd: Explicitly set image size and secsz in mkimg
When building an unprivileged NanoBSD image, explicitly set the desired
image size, by passing --capacity to mkimg in bytes, and the logical
sector size (-S) to 512 bytes.
Reviewed by: imp
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D57222
nanobsd: Offset code partition to a track boundary
Ensure the primary and secondary code partitions start at a proper track
boundary by applying a NANO_SECTS offset in bytes.
While track-boundary alignment is largely obsolete on modern storage,
this change maintains compatibility with current images (legacy). A
future commit will transition to 1 MiB alignment boundaries.
Reviewed by: imp
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D57221
nanobsd: Account for metadata overhead in code slice
The code slice size (CODE_SIZE) includes 16 sectors reserved for disk
metadata (see bsdlabel(8) offset). Subtract these 16 sectors from the
total size passed to nano_makefs.
This prevents the generated filesystem from consuming the entire slice
allocation, ensuring there is enough space for the metadata without
overflowing the partition boundary.
Reviewed by: imp
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D57220
nanobsd: Use rounded sizes for cfg and data slices
Use the calculated (rounded up) cfg and data sizes from the
_.partitioning file, instead of consuming them directly from the global
variables. We obtain the size of the cfg and data slices by explicitly
searching for index 3 and 4 respectively in the _.partitioning file.
This ensures that the final image has the rounded-up sizes, and not the
raw sizes.
Reviewed by: imp
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D57219
nanobsd: Get the code size by its partition index
Previously the code was assumed to be on the on the first line of the
_.partitioning file. Instead, explicitly look up the size by its
partition index to make the parsing order-independent.
The _.partitioning file:
1. First column: starting sector.
2. Second column: size in 512-byte sectors.
3. Third column: partition index.
Get the code size by explicitly selecting when the partition index is 1.
Reviewed by: imp
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D57216
nanobsd: Use NANO_IMG1NAME
Do not hard-code _.disk.image, use NANO_IMG1NAME instead.
Reviewed by: imp
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D57040
deskutils/mozo: switch to GitHub source
Switch from MATE mirror to GitHub tarball using USE_GITHUB and
GH_ACCOUNT=mate-desktop, remove tar:xz, and bump PORTREVISION.
zzz: Rewrite to use new power device
Previous script called acpiconf(8) (or apm(8) if ACPI wasn't supported,
although this was anyway redundant because APMIO just uses ACPI now).
Since a new generic power management interface was introduced, this isn't
sufficient, as this would only work for ACPI systems and for ACPI S3 suspend
(so no way to select suspend-to-idle). Rewrite in C to take advantage of the
new power interface.
We may want to add a switch to manually override the kern.power.suspend sysctl,
which is otherwise what the power device uses to decide which suspend type to
switch to (suspend-to-idle or firmware suspend), but this will require us to
amend the power interface.
Reviewed by: olce, imp, mhorne, ziaee
Tested by: mhorne
Approved by: olce, imp, mhorne, ziaee
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D56918
