Adolf Belka [Sat, 22 Nov 2025 19:52:27 +0000 (20:52 +0100)]
btrfs-progs: Update to version 6.17.1
- Update from version 6.17 to 6.17.1
- No change in rootfile
- Changelog
6.17.1
* inspect list-chunks: more sorting keys, descending order
* fi resize: add support for offline (unmounted) growing of single device
* device stats: add support for offline (unmounted) reads
* quota status: new command, overview what mode is enabled, tunables
* fi commit-stats: new command, print various commit stats from sysfs (since
kernel 6.1)
* balance start: print warning and delay start if there's a missing device
in the filesystem
* mkfs:
* print zoned mode (native, emulated)
* check:
* verify device bytes in super block item and in chunk tree
* other
* updated CI, new and updated tests
* cleanups, refactoring
* documentation updates
Signed-off-by: Adolf Belka <adolf.belka@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Adolf Belka [Sat, 22 Nov 2025 19:45:04 +0000 (20:45 +0100)]
openvpn: Update to version 2.6.16
- Update from version 2.6.15 to 2.6.16
- No change to rootfile
- Changelog
2.6.16
Security fixes:
CVE-2025-13086: Fix memcmp check for the hmac verification in the 3way
handshake. This bug renders the HMAC based protection against state
exhaustion on receiving spoofed TLS handshake packets in the OpenVPN
server inefficient.
Bug fixes:
fix invalid pointer creation in tls_pre_decrypt() - technically this is a
memory over-read issue, in practice, the compilers optimize it away so
no negative effects could be observed.
Windows: in the interactive service, fix the "undo DNS config" handling.
Windows: in the interactive service, disallow using of "stdin" for the
config file, unless the caller is authorized OpenVPN Administrator
Windows: in the interactive service, change all netsh calls to use
interface index and not interface name - sidesteps all possible attack
avenues with special characters in interface names.
Windows: in the interactive service, improve error handling in some
"unlikely to happen" paths.
auth plugin/script handling: properly check for errors in creation on
$auth_failed_reason_file (arf).
for incoming TCP connections, close-on-exec option was applied to the wrong
socket fd, leaking socket FDs to child processes.
sitnl: set close-on-exec flag on netlink socket
ssl_mbedtls: fix missing perf_pop() call (optional performance profiling)
Windows MSI changes since 2.6.15-I001:
Built against OpenSSL 3.6.0
Included openvpn-gui updated to 11.58.0.0
Check the return value of GetProp()
Make config path check similar to that in interactive service
Escape the type id of password message received from openvpn
Add a message source for event logging
Check correct management daemon path when OpenVPN3 is enabled
Fix OpenVPN3 radio button label size when OVPN3 is enabled
Use GetTempPath() for debug file in plap as well
Migrate all saved plain usernames to encrypted format
Included win-dco driver updated to 2.8.0
Signed-off-by: Adolf Belka <adolf.belka@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Stefan Schantl [Mon, 17 Nov 2025 11:49:55 +0000 (12:49 +0100)]
ffmpeg: Re-enable lame and SSL support
The ffmpeg developers decided at some point in the past to change some
defaults during configure and therefore we need to explit enable support
for lame and openssl.
Signed-off-by: Stefan Schantl <stefan.schantl@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Matthias Fischer [Sun, 16 Nov 2025 13:32:36 +0000 (14:32 +0100)]
nano: Update to 8.7
For details see:
https://www.nano-editor.org/news.php
"2025 November 12 - GNU nano 8.7 "Blue Highways"
At the Execute prompt, preceding the command with two pipe symbols
allows implementing a copy-to-clipboard feature in your nanorc (on
terminals that support OSC 52). See the doc/sample.nanorc file."
Signed-off-by: Matthias Fischer <matthias.fischer@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Robin Roevens [Tue, 11 Nov 2025 21:11:59 +0000 (22:11 +0100)]
zabbix_agentd: Update to 7.0.21 (LTS)
- Update of rootfile not required
Improvements:
- ZBXNEXT-9902 Changed timeout range for Zabbix JS and Zabbix get utilities
Bugs fixed:
- ZBX-25148 Added adjustments for sequential data entries with the same timestamp
- ZBX-25263 Fixed Zabbix agent to attempt next refresh of active checks in 60 seconds in case of connection errors
Full changelogs:
- https://www.zabbix.com/rn/rn7.0.19
- https://www.zabbix.com/rn/rn7.0.20
- https://www.zabbix.com/rn/rn7.0.21
Signed-off-by: Robin Roevens <robin.roevens@disroot.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Michael Tremer [Fri, 7 Nov 2025 14:39:27 +0000 (14:39 +0000)]
initscripts: dhcp: Tolerate running other dhcp servers
Some users have been trying to run multiple instances of the DHCP server
and restarting the main server won't work because the initscript refuses
to launch the process if there is another one with the same command.
Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Michael Tremer [Thu, 6 Nov 2025 16:46:51 +0000 (16:46 +0000)]
lldpd: New package
Signed-off-by: Michael Tremer <michael.tremer@ipfire.org> Reviewed-by: Adolf Belka <adolf.belka@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Adolf Belka [Fri, 7 Nov 2025 09:59:43 +0000 (10:59 +0100)]
strongswan: Update to version 6.0.3
- Update from version 6.0.2 to 6.0.3
- Update of rootfile
- Changelog
6.0.3
- Fixed a vulnerability in the eap-mschapv2 plugin related to processing Failure
Request packets on the client that can lead to a heap-based buffer overflow
and potentially remote code execution.
This vulnerability has been registered as CVE-2025-62291.
- The new `alert` event for vici is raised for certain error conditions.
- Only plugins with matching version number are loaded by programs.
- IKE SAs redirected during IKE_AUTH are now properly tracked by controller and
trap-manager.
- Fallback to the IKE identity for clients that don't provide an EAP-Identity to
fix a regression in 6.0.2.
- Detecting unwrapped CKA_EC_POINTs has been improved in the pkcs11 plugin.
- The whitelist plugin uses non-blocking I/O to avoid issues with clients that
stay connected for a long time. The buffer size for IDs was increased to 256.
- The certexpire plugins also uses 256 bytes for its identity buffer.
- Convenient decorators for event handling are provided by the Python bindings
for vici.
- The openssl plugin also supports Ed25519 via AWS-LC. It also loads EdDSA keys
from PKCS#12 containers.
- The testing environment is now based on Debian 13 (trixie), by default.
Signed-off-by: Adolf Belka <adolf.belka@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Michael Tremer [Fri, 7 Nov 2025 10:20:33 +0000 (11:20 +0100)]
squid: Don't forcibly kill any redirectory processes
This is a race which might cause that when squidGuard pre-compiles any
data, it will be killed too. If that happens, squid will keep forking
squidGuard processes which will be unresponsive (because they are trying
to compile they own databases) and the whole system will become
unresponsive at some point.
There should be no need to perform this action.
Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
- Update from version 3500400 to 3510000
- Update of rootfile
- Changelog 3510000
New macros in sqlite3.h:
SQLITE_SCM_BRANCH → the name of the branch from which the source code is
taken.
SQLITE_SCM_TAGS → space-separated list of tags on the source code check-in.
SQLITE_SCM_DATETIME → ISO-8601 date and time of the source code check-in.
Two new JSON functions, jsonb_each() and jsonb_tree() work the same as the
existing json_each() and json_tree() functions except that they return JSONB
for the "value" column when the "type" is 'array' or 'object'.
The carray and percentile extensions are now built into the amalgamation,
though they are disabled by default and must be activated at
compile-time using the -DSQLITE_ENABLE_CARRAY and/or
-DSQLITE_ENABLE_PERCENTILE options, respectively.
Enhancements to TCL Interface:
Add the -asdict flag to the eval command to have it set the row data as a
dict instead of an array.
User-defined functions may now break to return an SQL NULL.
CLI enhancements:
Increase the precision of ".timer" to microseconds.
Enhance the "box" and "column" formatting modes to deal with double-wide
characters.
The ".imposter" command provides read-only imposter tables that work with
VACUUM and do not require the --unsafe-testing option.
Add the --ifexists option to the CLI command-line option and to the .open
command.
Limit columns widths set by the ".width" command to 30,000 or less, as
there is not good reason to have wider columns, but supporting wider
columns provides opportunity to malefactors.
Performance enhancements:
Use fewer CPU cycles to commit a read transaction.
Early detection of joins that return no rows due to one or more of the
tables containing no rows.
Avoid evaluation of scalar subqueries if the result of the subquery does
not change the result of the overall expression.
Faster window function queries when using
"BETWEEN :x FOLLOWING AND :y FOLLOWING" with a large :y.
Add the PRAGMA wal_checkpoint=NOOP; command and the SQLITE_CHECKPOINT_NOOP
argument for sqlite3_wal_checkpoint_v2().
Add the sqlite3_set_errmsg() API for use by extensions.
Add the sqlite3_db_status64() API, which works just like the existing
sqlite3_db_status() API except that it returns 64-bit results.
Add the SQLITE_DBSTATUS_TEMPBUF_SPILL option to the sqlite3_db_status() and
sqlite3_db_status64() interfaces.
In the session extension add the sqlite3changeset_apply_v3() interface.
For the built-in printf() and the format() SQL function, omit the leading '-'
from negative floating point numbers if the '+' flag is omitted and the "#"
flag is present and all displayed digits are '0'. Use '%#f' or similar to
avoid outputs like '-0.00' and instead show just '0.00'.
Improved error messages generated by FTS5.
Enforce STRICT typing on computed columns.
Improved support for VxWorks
JavaScript/WASM now supports 64-bit WASM. The canonical builds continue to be
32-bit but creating one's own 64-bit build is now as simple as running "make".
Improved resistance to database corruption caused by an application breaking
Posix advisory locks using close().
Signed-off-by: Adolf Belka <adolf.belka@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Adolf Belka [Thu, 6 Nov 2025 18:03:29 +0000 (19:03 +0100)]
protobuf: Update to version 33.0
- Update from version 32.1 to 33.0
- Update of rootfile
- Changelog
33.0
Bazel
Feat: update bazel central registry publish workflow (#23465) (#23913)
(d5217fd)
Add target_compatible_with parameter to proto_toolchain in Bazel rules
(#22429) (30d2332)
Bazel: add missing rules_cc loads (#23584) (d98e2ef)
Compiler
Disable symbol visibility enforcement by default in C++ runtime (ae308fc)
Ship all option dependencies to plugins along with regular ones. (abeb130)
C++
Avoid calling deprecated arena-enabled constructors in arena.h. (813a7ef)
Add a macro to make RepeatedField(Arena*) constructor private in a future
release. (768db14)
Add a macro to make Map(Arena*) constructor private in a future release.
(543a17f)
Optimize ReadPackedVarint (3d94d83)
Add a macro to make RepeatedPtrField(Arena*) constructor private in a
future release (6422b9d)
Add IsEmpty() function to reflection. (b64e490)
Refactor RuntimeAssertInBounds to remove repeated logic and make
Get/Mutable easier to read. (2f270c4)
Disable symbol visibility enforcement by default in C++ runtime (ae308fc)
Fix a bug in the main C++ JSON parser/serializer camelcasing of certain
non-style-compliant names incorrectly, in a way that would prevent it
from interoperating with any other implementation on those fields.
(e25e267)
Fail early for messages with more than 65k fields. (90824aa)
Add option to C++ JSON Parser/Serializer to allow customers to
affirmatively disable legacy bug-compatibilty behaviors. (6ea1640)
Fix mishandling on JSON serialization of Timestamp with invalid negative
and too-large nanos value. (a959f27)
Preserve features in type resolver (c7030f4)
Add a DCHECK that ArenaStringPtr::Set(char*, Arena*) is not called with
(95b1763)
Java
Restored compatibility of runtime with gencode created with protoc <3.21
With this release, compatibility of the runtime with older gencode down to
3.0.0 is restored, compared to the previous support minimum of gencode
created with 3.22+. Note that it is still strongly recommended to
regenerate your gencode with a newer protoc and to avoid using gencode
which was created with an old protoc.
Generated code from this range is covered by CVE-2022-3171 and is
potentially vulnerable to a Denial of Service risk.
JavaProto 4.x previously dropped compatibility with the potentially
vulnerable generated code, having the behavior of:
The vulnerable generated code was source-incompatible with new runtime
(would not compile when built from source)
The vulnerable generated code was ABI-incompatible with new runtime
(when using a .class file compiled against old runtime, a
NoSuchMethodException would be thrown at parse time).
Starting with this release:
The vulnerable generated code is now source-compatible (will compile).
The first time each potentially vulnerable type is parsed, an error
message will be logged noting that potentially vulnerable generated
code is in use and the name of the corresponding type.
Environment variables may be set to either throw an exception instead
(-Dcom.google.protobuf.error_on_unsafe_pre22_gencode) or to entirely
silence the logged messages
(-Dcom.google.protobuf.use_unsafe_pre22_gencode)
This change was made based on community feedback regarding the difficulty
in identifying and quickly remediating stale gencode in their
transitive dependencies weighed against a careful evaluation of the
realistic risk exposure of DoS (with no risk of other concerns
including information leak or RCE).
We strongly recommend that any users who observe the log messages to
regenerate the corresponding code with a newer protoc. We recommend
that any security-conscious services opt into the
error_on_unsafe_pre22_gencode behavior to preclude any risk of a
Denial of Service surface area being exposed.
A future 4.x release may flip the default behavior to error by default as
a measure to further help the ecosystem avoid the Denial of Service
risks, while still maintaining the ability to opt into continuing to
use insecure gencode for users who are parsing trusted inputs and
where the difficulty of regenerating is high.
Changes
Switch the pre22 warning to use CopyOnWriteArraySet. (#23969) (e55224c)
Expose helpers for checking if messages and enums are nested. (8de4002)
Fix a bug calculating the file name in the absense of directories. (c4ff7a6)
Clarify the public APIs of GeneratorNames helpers. (537ac35)
Expose helpers to predict generated class names in java. (eba6df2)
Deprecate ClassName methods in favor of new QualifiedClassName ones.
(ca4fb2f)
Restore the 3-argument internalBuildGeneratedFileFrom. (4376591)
Fix large java enums not being honored on lite runtime. (a995803)
Slightly relax Java Poison Pill on prerelease versions (-rc1, -dev, etc).
(7b0bee3)
Avoid boxing/unboxing varint, fixed32, and fixed64 fields in
UnknownFieldSet.Field (810272f)
Readd new*List() methods on GeneratedMessageV3. (badaf41)
Add Values.of(Map<String, Value> values). (c518f25)
Fix handling of optional dependencies in java generator. (8d51e34)
Restore ABI compatibility for extension methods which was previously
(knowingly) broken with 4.x: 94a2a44 (ea33ae8)
Restore Protobuf Java extension modifiers in gencode that were previously
removed in 7bff169 (f2257f5)
Ship all option dependencies to plugins along with regular ones. (abeb130)
Optimize redaction state calculation (e05db5c)
Add isPlaceholder() accessors to file, message, and enum descriptors
(f978ec2)
Improve Java gencode static initialization to avoid unnecessary
temporaries again (745e15b)
Improve Java gencode static initialization to avoid unnecessary
temporaries (b68b673)
Remove protobuf-util usages of guava except annotations. (5768acd)
Restore compatibility of runtime with pre-3.22.x gencode impacted by
CVE-2022-3171 (7c51e5b)
Expose an iterator for GeneratedMessage.ExtendableMessage.extensions
(b25d39e)
Rust
Change Rust prelude to bring in traits as _ (c3f7e8d)
Make message Muts Send (8bff944)
See also UPB changes below, which may affect Rust.
Python
Publish s390x wheels for Python/upb. (56b2b89)
Fix a crash that happens during shutdown due to looking up modules in the
cache (d57d270)
Add construction support for repeated Timestamp/Duration/Struct/ListValue.
(5f6c013)
Fix handling of repeated extension fields in PyProto JSON (07ef676)
Fixed a parser bug where closed enums are parsed incorrectly for
non-repeated extensions. (c36f728)
Fixed mypy errors by setting __slots__ to empty in .pyi files. (38ca2d3)
Raise warnings for float_precision from python json_format. (4659cd7)
Raise warnings when assign bool to int/enum field in Python Proto. This
will turn into error in 34.0 release. (4ee55d7)
PHP
Fix(php): php errors on repeated field (#23372) (6fee29b)
UPB (Python/PHP/Ruby C-Extension)
Fixed a parser bug where closed enums are parsed incorrectly for
non-repeated extensions. (c36f728)
Other
Update token for BCR release to reuse existing BOT_ACCESS_TOKEN used for
staleness_refresh.yml and update_php_repo.yml (#23925) (dcace2f)
Use the 'better' JSON parser on the conformance suite harness. (4b4e405)
Add JSON conformance test that a single value provided for a repeated
field should parse fail. (9806994)
Add conformance test cases for malformed nanos fields on Durations and
Timestamps. (a6bdd0a)
Signed-off-by: Adolf Belka <adolf.belka@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Adolf Belka [Thu, 6 Nov 2025 18:03:28 +0000 (19:03 +0100)]
lvm2: Update to version 2.03.36
- Update from version 2.03.35 to 2.03.36
- No change to rootfile
- Changelog
2.03.36
Fix uninitialized chunk_size_calc_policy in pool parameter functions.
Fix approximate allocation for Raid with insufficient extents.
Fix race in dmeventd remonitoring optimization (2.03.35).
Use -real suffix for pvmove UUID.
Add support pvmove segmentation allocation/pvmove_max_segmentation_size_mb.
Allow creating _imeta with multiple segments.
Fix driver_version() accepts NULL version buffer pointer.
Fix invalid free() call in error path of _add_metadata_area_to_pv().
Avoid destroying aio context in forked process.
Add lvs -o cache_promotions,cache_promotions fields.
Update pvmove logic when moving i.e. raid legs.
Display integrity info in lvdisplay.
Increase storage size for internal filter chain.
Add helper function display_mb_size().
Enhance code for adding and removing integrity to RAID volumes.
Add code for basic validation of integrity segment.
Use -real private suffix for integrity origin and meta volumes.
Use -real private suffix for mirror and raid legs.
Detect and use existing XFS quota mount options for lvresize --fs resize.
Signed-off-by: Adolf Belka <adolf.belka@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Adolf Belka [Thu, 6 Nov 2025 18:03:27 +0000 (19:03 +0100)]
libcap: Update to version 2.77
- Update from version 2.76 to 2.77
- Update of rootfile
- Changelog
2.77
Fix mistakes in setcap for reporting errors: report them with the
appropriate filename. Thanks to Nikolas for reporting these in
Bug 220245.
Fix bug in cap.GetIAB() reported and fix provided by Garret Kelly via
Bug 220420.
Improve libcap managed memory allocation and support CHERI RISC-V. Reported
with fix by Chris Hofer via Bug 220415.
Add (unverified) support for the PSX mechanism on microblaze, arc, openrisc
and xtensa architectures. Thanks to Tom Petazzoni for including these in
Bug 219915
Please let me know if these work or fail on these architectures.
Add C++ support to the run a .so file as an executable mechanism employed by
libcap.so, libpsx.so and pam_cap.so. Not really necessary for the libcap
build tree, but wanted to capture the details of my recent update to a
Stackoverflow answer on the topic.
Use BUILD_LDFLAGS when compiling _makenames fix contributed by Khem Raj.
Fix broke some builds, so will revert and apply a more comprehensive fix.
Fixed sendmail issue discussion link. Thanks to Ariel Otilibili for noticing
the breakage and contributing a fix.
Some debugging fixes for use of the kdebug/ testing setup,
Signed-off-by: Adolf Belka <adolf.belka@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Adolf Belka [Thu, 6 Nov 2025 18:03:25 +0000 (19:03 +0100)]
elfutils: Update to version 0.194
- Update from version 0.193 to 0.194
- Update of rootfile
- Changelog
0.194
debuginfod-find: Fixed caching bug preventing user-cancelled downloads
from being re-downloaded at a later time.
elfclassify: New options --has-debug-sections and --any-ar-member.
elflint: Presence of vendor- and application-specific ELF note types no
longer triggers compliance errors.
libdwfl_stacktrace: New function dwflst_sample_getframes. The
libdwfl_stacktrace library interface is experimental
and may be subject to API/ABI changes.
libelf: Manual pages have been added for many libelf library functions.
Additional manual pages are planned for future releases.
readelf: Up to 13% faster when using the -N option.
Signed-off-by: Adolf Belka <adolf.belka@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Adolf Belka [Thu, 6 Nov 2025 18:02:55 +0000 (19:02 +0100)]
fmt: Update to version 12.1.0
- Update from version 11.2.0 to 12.1.0
- Update of rootfile
- so-bump so mpd requires shipping
- Changelog
12.1.0
- Optimized `buffer::append`, resulting in up to ~16% improvement on spdlog
benchmarks (https://github.com/fmtlib/fmt/pull/4541). Thanks @fyrsta7.
- Worked around an ABI incompatibility in `std::locale_ref` between clang and
gcc (https://github.com/fmtlib/fmt/issues/4573).
- Made `std::variant` and `std::expected` formatters work with `format_as`
(https://github.com/fmtlib/fmt/issues/4574,
https://github.com/fmtlib/fmt/pull/4575). Thanks @phprus.
- Made `fmt::join<string_view>` work with C++ modules
(https://github.com/fmtlib/fmt/issues/4379,
https://github.com/fmtlib/fmt/pull/4577). Thanks @Arghnews.
- Exported `fmt::is_compiled_string` and `operator""_cf` from the module
(https://github.com/fmtlib/fmt/pull/4544). Thanks @CrackedMatter.
- Fixed a compatibility issue with C++ modules in clang
(https://github.com/fmtlib/fmt/pull/4548). Thanks @tsarn.
- Added support for cv-qualified types to the `std::optional` formatter
(https://github.com/fmtlib/fmt/issues/4561,
https://github.com/fmtlib/fmt/pull/4562). Thanks @OleksandrKvl.
- Added demangling support (used in exception and `std::type_info` formatters)
for libc++ and clang-cl
(https://github.com/fmtlib/fmt/issues/4542,
https://github.com/fmtlib/fmt/pull/4560,
https://github.com/fmtlib/fmt/issues/4568,
https://github.com/fmtlib/fmt/pull/4571).
Thanks @FatihBAKIR and @rohitsutreja.
- Switched to global `malloc`/`free` to enable allocator customization
(https://github.com/fmtlib/fmt/issues/4569,
https://github.com/fmtlib/fmt/pull/4570). Thanks @rohitsutreja.
- Made the `FMT_USE_CONSTEVAL` macro configurable by users
(https://github.com/fmtlib/fmt/pull/4546). Thanks @SnapperTT.
- Fixed compilation with locales disabled in the header-only mode
(https://github.com/fmtlib/fmt/issues/4550).
- Fixed compilation with clang 21 and `-std=c++20`
(https://github.com/fmtlib/fmt/issues/4552).
- Fixed a dynamic linking issue with clang-cl
(https://github.com/fmtlib/fmt/issues/4576,
https://github.com/fmtlib/fmt/pull/4584). Thanks @FatihBAKIR.
- Fixed a warning suppression leakage on gcc
(https://github.com/fmtlib/fmt/pull/4588). Thanks @ZedThree.
- Made more internal color APIs `constexpr`
(https://github.com/fmtlib/fmt/pull/4581). Thanks @ishani.
- Fixed compatibility with clang as a host compiler for NVCC
(https://github.com/fmtlib/fmt/pull/4564). Thanks @valgur.
- Fixed various warnings and lint issues
(https://github.com/fmtlib/fmt/issues/4565,
https://github.com/fmtlib/fmt/pull/4572,
https://github.com/fmtlib/fmt/pull/4557).
Thanks @LiangHuDream and @teruyamato0731.
- Improved documentation
(https://github.com/fmtlib/fmt/issues/4549,
https://github.com/fmtlib/fmt/pull/4551,
https://github.com/fmtlib/fmt/issues/4566,
https://github.com/fmtlib/fmt/pull/4567,
https://github.com/fmtlib/fmt/pull/4578,).
Thanks @teruyamato0731, @petersteneteg and @zimmerman-dev.
12.0.0
- Optimized the default floating point formatting
(https://github.com/fmtlib/fmt/issues/3675,
https://github.com/fmtlib/fmt/issues/4516). In particular, formatting a
`double` with format string compilation into a stack allocated buffer is
more than 60% faster in version 12.0 compared to 11.2 according to
[dtoa-benchmark](https://github.com/fmtlib/dtoa-benchmark):
```
Function Time (ns) Speedup
fmt11 34.471 1.00x
fmt12 21.000 1.64x
```
<img width="766" height="609" src="https://github.com/user-attachments/assets/d7d768ad-7543-468c-b0bb-449abf73b31b" />
- Added `constexpr` support to `fmt::format`. For example:
```c++
#include <fmt/compile.h>
using namespace fmt::literals;
std::string s = fmt::format(""_cf, 42);
```
now works at compile time provided that `std::string` supports `constexpr`
(https://github.com/fmtlib/fmt/issues/3403,
https://github.com/fmtlib/fmt/pull/4456). Thanks @msvetkin.
- Added `FMT_STATIC_FORMAT` that allows formatting into a string of the exact
required size at compile time.
For example:
```c++
#include <fmt/compile.h>
constexpr auto s = FMT_STATIC_FORMAT("{}", 42);
```
compiles to just
```s
__ZL1s:
.asciiz "42"
```
It can be accessed as a C string with `s.c_str()` or as a string view with
`s.str()`.
- Improved C++20 module support
(https://github.com/fmtlib/fmt/pull/4451,
https://github.com/fmtlib/fmt/pull/4459,
https://github.com/fmtlib/fmt/pull/4476,
https://github.com/fmtlib/fmt/pull/4488,
https://github.com/fmtlib/fmt/issues/4491,
https://github.com/fmtlib/fmt/pull/4495).
Thanks @arBmind, @tkhyn, @Mishura4, @anonymouspc and @autoantwort.
- Switched to using estimated display width in precision. For example:
```c++
fmt::print("|{:.4}|\n|1234|\n", "🐱🐱🐱");
```
prints
because `🐱` has an estimated width of 2
(https://github.com/fmtlib/fmt/issues/4272,
https://github.com/fmtlib/fmt/pull/4443,
https://github.com/fmtlib/fmt/pull/4475).
Thanks @nikhilreddydev and @localspook.
- Fix interaction between debug presentation, precision, and width for strings
(https://github.com/fmtlib/fmt/pull/4478). Thanks @localspook.
- Implemented allocator propagation on `basic_memory_buffer` move
(https://github.com/fmtlib/fmt/issues/4487,
https://github.com/fmtlib/fmt/pull/4490). Thanks @toprakmurat.
- Fixed an ambiguity between `std::reference_wrapper<T>` and `format_as`
formatters (https://github.com/fmtlib/fmt/issues/4424,
https://github.com/fmtlib/fmt/pull/4434). Thanks @jeremy-rifkin.
- Removed the following deprecated APIs:
- `has_formatter`: use `is_formattable` instead,
- `basic_format_args::parse_context_type`,
`basic_format_args::formatter_type` and similar aliases in context types,
- wide stream overload of `fmt::printf`,
- wide stream overloads of `fmt::print` that take text styles,
- `is_*char` traits,
- `fmt::localtime`.
- Deprecated wide overloads of `fmt::fprintf` and `fmt::sprintf`.
- Improved diagnostics for the incorrect usage of `fmt::ptr`
(https://github.com/fmtlib/fmt/pull/4453). Thanks @TobiSchluter.
- Made handling of ANSI escape sequences more efficient
(https://github.com/fmtlib/fmt/pull/4511,
https://github.com/fmtlib/fmt/pull/4528).
Thanks @localspook and @Anas-Hamdane.
- Fixed a buffer overflow on all emphasis flags set
(https://github.com/fmtlib/fmt/pull/4498). Thanks @dominicpoeschko.
- Fixed an integer overflow for precision close to the max `int` value.
- Fixed compatibility with WASI (https://github.com/fmtlib/fmt/issues/4496,
https://github.com/fmtlib/fmt/pull/4497). Thanks @whitequark.
- Fixed `back_insert_iterator` detection, preventing a fallback on slower path
that handles arbitrary iterators (https://github.com/fmtlib/fmt/issues/4454).
- Fixed handling of invalid glibc `FILE` buffers
(https://github.com/fmtlib/fmt/issues/4469).
- Added `wchar_t` support to the `std::byte` formatter
(https://github.com/fmtlib/fmt/issues/4479,
https://github.com/fmtlib/fmt/pull/4480). Thanks @phprus.
- Changed component prefix from `fmt-` to `fmt_` for compatibility with
NSIS/CPack on Windows, e.g. `fmt-doc` changed to `fmt_doc`
(https://github.com/fmtlib/fmt/issues/4441,
https://github.com/fmtlib/fmt/pull/4442). Thanks @n-stein.
- Added the `FMT_CUSTOM_ASSERT_FAIL` macro to simplify providing a custom
`fmt::assert_fail` implementation (https://github.com/fmtlib/fmt/pull/4505).
Thanks @HazardyKnusperkeks.
- Switched to `FMT_THROW` on reporting format errors so that it can be
overriden by users when exceptions are disabled
(https://github.com/fmtlib/fmt/pull/4521). Thanks @HazardyKnusperkeks.
- Improved master project detection and disabled install targets when using
{fmt} as a subproject by default (https://github.com/fmtlib/fmt/pull/4536).
Thanks @crueter.
- Made various code improvements
(https://github.com/fmtlib/fmt/pull/4445,
https://github.com/fmtlib/fmt/pull/4448,
https://github.com/fmtlib/fmt/pull/4473,
https://github.com/fmtlib/fmt/pull/4522).
Thanks @localspook, @tchaikov and @way4sahil.
- Added Conan instructions to the docs
(https://github.com/fmtlib/fmt/pull/4537). Thanks @uilianries.
- Removed Bazel files to avoid issues with downstream packaging
(https://github.com/fmtlib/fmt/pull/4530). Thanks @mering.
- Added more entries for generated files to `.gitignore`
(https://github.com/fmtlib/fmt/pull/4355,
https://github.com/fmtlib/fmt/pull/4512).
Thanks @dinomight and @localspook.
- Fixed various warnings and compilation issues
(https://github.com/fmtlib/fmt/pull/4447,
https://github.com/fmtlib/fmt/issues/4470,
https://github.com/fmtlib/fmt/pull/4474,
https://github.com/fmtlib/fmt/pull/4477,
https://github.com/fmtlib/fmt/pull/4471,
https://github.com/fmtlib/fmt/pull/4483,
https://github.com/fmtlib/fmt/pull/4515,
https://github.com/fmtlib/fmt/issues/4533,
https://github.com/fmtlib/fmt/pull/4534).
Thanks @dodomorandi, @localspook, @remyjette, @Tomek-Stolarczyk, @Mishura4,
@mattiasljungstrom and @FatihBAKIR.
Signed-off-by: Adolf Belka <adolf.belka@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Adolf Belka [Thu, 30 Oct 2025 16:23:55 +0000 (17:23 +0100)]
libvirt: Uncomment two lines in the rootfile
- These lines were new in with the previous update from 11.4.0 to 11.7.0 but I
incorrectly commented them out.
- This has been flagged up in the forum as it is stopping the user doing a virsh backup.
Signed-off-by: Adolf Belka <adolf.belka@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Adolf Belka [Thu, 30 Oct 2025 10:18:27 +0000 (11:18 +0100)]
expect: Allow build of expect for swtpm
- expect was previously only built in the toolchain. This allows expect to be built for
swtpm in the normal build.
- It is only used for the building of swtpm so the rootfile has all entries commented
out.
- LFS changed to allow build in the toolchain or in the normal build
Signed-off-by: Adolf Belka <adolf.belka@ipfire.org> Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Michael Tremer [Tue, 28 Oct 2025 15:41:43 +0000 (15:41 +0000)]
make.sh: Disable the toolchain during normal builds
This is a precaution because I have seen some files from the toolchain
being baked into the initrd. That should not happen and by keeping PATH
clean, there is no chance for any confusion.
In order to have the tools available that we need to build the initial
system, we will now have to pass a flag.
Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
Michael Tremer [Tue, 28 Oct 2025 17:13:44 +0000 (17:13 +0000)]
make.sh: Create a better PID 1 inside the environment
This is required so that we won't have any make processes locking up any
more. When the build process is getting aborted, this script will now
cleanly terminate anything inside the container and not block make as
PID 1.
Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>
projects / ipfire-2.x.git / log
