1.1.6
[bug] [lexer]
Fixed issue where control statements on multi lines with a backslash would not parse correctly if the template itself contained CR/LF pairs as on Windows.
Bugfixes:
- Improved usability for knownbugs dialog
- Batch runs from Jenkins now show "old" indicator by default when more
than a day old
- Better error handling when grid engines configured but not installed
New features:
- adding split possiblity in command line fpdiff #45
- Fix for #80, full paths in descriptions in batch report when running
in parallel
2.4.14
Added
- Added `fs.copy.copy_file_if`, `fs.copy.copy_dir_if`, and `fs.copy.copy_fs_if`.
- Added `fs.base.FS.getmodified`.
Changed
- FTP servers that do not support the MLST command now try to use the MDTM command to
retrieve the last modification timestamp of a resource.
Fixed
- Fixed performance bugs in `fs.copy.copy_dir_if_newer`. Test cases were adapted to catch those bugs in the future.
- Fixed precision bug for timestamps in `fs.OSFS.setinfo`.
-vscreen: Update _NET_CURRENT_DESKTOP atom when changing vscreens
-Move vscreen initialization further down init_screen because we need the root
window set.
Today, we're releasing 0.40 of Nu. This release is includes better table imports
and much smaller release size.
What's New
-Detecting columns (jt)
Nushell now supports a new command: detect columns. This command is intended as
an easy way to handle tabular data that's written as text. Commands like ps,
ls -l, netstat, df, and others commonly output their output as text.
With detect columns, you can run the external command, and convert them into
Nushell's tabular data format.
-Smaller binaries (fdncred)
The size of the Nushell release binaries has come up as regular feedback from
Nushell users. "Why does the shell have to be so big?" And we've heard you!
Today's release now uses a combination of strip and upx to bring considerable
savings in binary size. The Linux release (plugins incl.) has dropped from
475mb to 58mb(!!), making it 88% smaller. We're seeing similar improvements in
the sizes of the macOS and Windows releases.
v0.25.0
Changes
Improve SSL option compatibility in URIs
Add Pool methods to determine its min, max, current and idle size
Make it possible to specify a statement name in Connection.prepare()
Implement support for multirange types
Fixes
Make sure timeout callbacks always get cleaned up
Update __all__ statements to a simpler form that is better supported by typecheckers
Fix test_timetz_encoding on Python 3.10
Fix a bunch of ResourceWarnings in the test suite
Fix SSLContext deprecation warnings
Fix the description of the database argument to connect()
Fix parsing of IPv6 addresses in the connection URI
Improve diagnostics of invalid executemany() input
Git 2.34 Release Notes
======================
Updates since Git 2.33
----------------------
Backward compatibility notes
* The "--preserve-merges" option of "git rebase" has been removed.
UI, Workflows & Features
* Pathname expansion (like "~username/") learned a way to specify a
location relative to Git installation (e.g. its $sharedir which is
$(prefix)/share), with "%(prefix)".
* The `ort` strategy is used instead of `recursive` as the default
merge strategy.
* The userdiff pattern for "java" language has been updated.
* "git rebase" by default skips changes that are equivalent to
commits that are already in the history the branch is rebased onto;
give messages when this happens to let the users be aware of
skipped commits, and also teach them how to tell "rebase" to keep
duplicated changes.
* The advice message that "git cherry-pick" gives when it asks
conflicted replay of a commit to be resolved by the end user has
been updated.
* After "git clone --recurse-submodules", all submodules are cloned
but they are not by default recursed into by other commands. With
submodule.stickyRecursiveClone configuration set, submodule.recurse
configuration is set to true in a repository created by "clone"
with "--recurse-submodules" option.
* The logic for auto-correction of misspelt subcommands learned to go
interactive when the help.autocorrect configuration variable is set
to 'prompt'.
* "git maintenance" scheduler learned to use systemd timers as a
possible backend.
* "git diff --submodule=diff" showed failure from run_command() when
trying to run diff inside a submodule, when the user manually
removes the submodule directory.
* "git bundle unbundle" learned to show progress display.
* In cone mode, the sparse-index code path learned to remove ignored
files (like build artifacts) outside the sparse cone, allowing the
entire directory outside the sparse cone to be removed, which is
especially useful when the sparse patterns change.
* Taking advantage of the CGI interface, http-backend has been
updated to enable protocol v2 automatically when the other side
asks for it.
* The credential-cache helper has been adjusted to Windows.
* The error in "git help no-such-git-command" is handled better.
* The unicode character width table (used for output alignment) has
been updated.
* The ref iteration code used to optionally allow dangling refs to be
shown, which has been tightened up.
* "git add", "git mv", and "git rm" have been adjusted to avoid
updating paths outside of the sparse-checkout definition unless
the user specifies a "--sparse" option.
* "git repack" has been taught to generate multi-pack reachability
bitmaps.
* "git fsck" has been taught to report mismatch between expected and
actual types of an object better.
* In addition to GnuPG, ssh public crypto can be used for object and
push-cert signing. Note that this feature cannot be used with
ssh-keygen from OpenSSH 8.7, whose support for it is broken. Avoid
using it unless you update to OpenSSH 8.8.
* "git log --grep=string --author=name" learns to highlight hits just
like "git grep string" does.
Performance, Internal Implementation, Development Support etc.
* "git bisect" spawned "git show-branch" only to pretty-print the
title of the commit after checking out the next version to be
tested; this has been rewritten in C.
* "git add" can work better with the sparse index.
* Support for ancient versions of cURL library (pre 7.19.4) has been
dropped.
* A handful of tests that assumed implementation details of files
backend for refs have been cleaned up.
* trace2 logs learned to show parent process name to see in what
context Git was invoked.
* Loading of ref tips to prepare for common ancestry negotiation in
"git fetch-pack" has been optimized by taking advantage of the
commit graph when available.
* Remind developers that the userdiff patterns should be kept simple
and permissive, assuming that the contents they apply are always
syntactically correct.
* The current implementation of GIT_TEST_FAIL_PREREQS is broken in
that checking for the lack of a prerequisite would not work. Avoid
the use of "if ! test_have_prereq X" in a test script.
* The revision traversal API has been optimized by taking advantage
of the commit-graph, when available, to determine if a commit is
reachable from any of the existing refs.
* "git fetch --quiet" optimization to avoid useless computation of
info that will never be displayed.
* Callers from older advice_config[] based API has been updated to
use the newer advice_if_enabled() and advice_enabled() API.
* Teach "test_pause" and "debug" helpers to allow using the HOME and
TERM environment variables the user usually uses.
* "make INSTALL_STRIP=-s install" allows the installation step to use
"install -s" to strip the binaries as they get installed.
* Code that handles large number of refs in the "git fetch" code
path has been optimized.
* The reachability bitmap file used to be generated only for a single
pack, but now we've learned to generate bitmaps for history that
span across multiple packfiles.
* The code to make "git grep" recurse into submodules has been
updated to migrate away from the "add submodule's object store as
an alternate object store" mechanism (which is suboptimal).
* The tracing of process ancestry information has been enhanced.
* Reduce number of write(2) system calls while sending the
ref advertisement.
* Update the build procedure to use the "-pedantic" build when
DEVELOPER makefile macro is in effect.
* Large part of "git submodule add" gets rewritten in C.
* The run-command API has been updated so that the callers can easily
ask the file descriptors open for packfiles to be closed immediately
before spawning commands that may trigger auto-gc.
* An oddball OPTION_ARGUMENT feature has been removed from the
parse-options API.
* The mergesort implementation used to sort linked list has been
optimized.
* Remove external declaration of functions that no longer exist.
* "git multi-pack-index write --bitmap" learns to propagate the
hashcache from original bitmap to resulting bitmap.
* CI learns to run the leak sanitizer builds.
* "git grep --recurse-submodules" takes trees and blobs from the
submodule repository, but the textconv settings when processing a
blob from the submodule is not taken from the submodule repository.
A test is added to demonstrate the issue, without fixing it.
* Teach "git help -c" into helping the command line completion of
configuration variables.
* When "git cmd -h" shows more than one line of usage text (e.g.
the cmd subcommand may take sub-sub-command), parse-options API
learned to align these lines, even across i18n/l10n.
* Prevent "make sparse" from running for the source files that
haven't been modified.
* The code path to write a new version of .midx multi-pack index files
has learned to release the mmaped memory holding the current
version of .midx before removing them from the disk, as some
platforms do not allow removal of a file that still has mapping.
* A new feature has been added to abort early in the test framework.
Fixes since v2.33
-----------------
* Input validation of "git pack-objects --stdin-packs" has been
corrected.
* Bugfix for common ancestor negotiation recently introduced in "git
push" code path.
* "git pull" had various corner cases that were not well thought out
around its --rebase backend, e.g. "git pull --ff-only" did not stop
but went ahead and rebased when the history on other side is not a
descendant of our history. The series tries to fix them up.
* "git apply" miscounted the bytes and failed to read to the end of
binary hunks.
* "git range-diff" code clean-up.
* "git commit --fixup" now works with "--edit" again, after it was
broken in v2.32.
* Use upload-artifacts v1 (instead of v2) for 32-bit linux, as the
new version has a blocker bug for that architecture.
* Checking out all the paths from HEAD during the last conflicted
step in "git rebase" and continuing would cause the step to be
skipped (which is expected), but leaves MERGE_MSG file behind in
$GIT_DIR and confuses the next "git commit", which has been
corrected.
* Various bugs in "git rebase -r" have been fixed.
* mmap() imitation used to call xmalloc() that dies upon malloc()
failure, which has been corrected to just return an error to the
caller to be handled.
* "git diff --relative" segfaulted and/or produced incorrect result
when there are unmerged paths.
* The delayed checkout code path in "git checkout" etc. were chatty
even when --quiet and/or --no-progress options were given.
* "git branch -D <branch>" used to refuse to remove a broken branch
ref that points at a missing commit, which has been corrected.
* Build update for Apple clang.
* The parser for the "--nl" option of "git column" has been
corrected.
* "git upload-pack" which runs on the other side of "git fetch"
forgot to take the ref namespaces into account when handling
want-ref requests.
* The sparse-index support can corrupt the index structure by storing
a stale and/or uninitialized data, which has been corrected.
* Buggy tests could damage repositories outside the throw-away test
area we created. We now by default export GIT_CEILING_DIRECTORIES
to limit the damage from such a stray test.
* Even when running "git send-email" without its own threaded
discussion support, a threading related header in one message is
carried over to the subsequent message to result in an unwanted
threading, which has been corrected.
* The output from "git fast-export", when its anonymization feature
is in use, showed an annotated tag incorrectly.
* Recent "diff -m" changes broke "gitk", which has been corrected.
* The "git apply -3" code path learned not to bother the lower level
merge machinery when the three-way merge can be trivially resolved
without the content level merge. This fixes a regression caused by
recent "-3way first and fall back to direct application" change.
* The code that optionally creates the *.rev reverse index file has
been optimized to avoid needless computation when it is not writing
the file out.
* "git range-diff -I... <range> <range>" segfaulted, which has been
corrected.
* The order in which various files that make up a single (conceptual)
packfile has been reevaluated and straightened up. This matters in
correctness, as an incomplete set of files must not be shown to a
running Git.
* The "mode" word is useless in a call to open(2) that does not
create a new file. Such a call in the files backend of the ref
subsystem has been cleaned up.
* "git update-ref --stdin" failed to flush its output as needed,
which potentially led the conversation to a deadlock.
* When "git am --abort" fails to abort correctly, it still exited
with exit status of 0, which has been corrected.
* Correct nr and alloc members of strvec struct to be of type size_t.
* "git stash", where the tentative change involves changing a
directory to a file (or vice versa), was confused, which has been
corrected.
* "git clone" from a repository whose HEAD is unborn into a bare
repository didn't follow the branch name the other side used, which
is corrected.
* "git cvsserver" had a long-standing bug in its authentication code,
which has finally been corrected (it is unclear and is a separate
question if anybody is seriously using it, though).
* "git difftool --dir-diff" mishandled symbolic links.
* Sensitive data in the HTTP trace were supposed to be redacted, but
we failed to do so in HTTP/2 requests.
* "make clean" has been updated to remove leftover .depend/
directories, even when it is not told to use them to compute header
dependencies.
* Protocol v0 clients can get stuck parsing a malformed feature line.
* A few kinds of changes "git status" can show were not documented.
(merge d2a534c515 ja/doc-status-types-and-copies later to maint).
* The mergesort implementation used to sort linked list has been
optimized.
(merge c90cfc225b rs/mergesort later to maint).
* An editor session launched during a Git operation (e.g. during 'git
commit') can leave the terminal in a funny state. The code path
has updated to save the terminal state before, and restore it
after, it spawns an editor.
(merge 3d411afabc cm/save-restore-terminal later to maint).
* "git cat-file --batch" with the "--batch-all-objects" option is
supposed to iterate over all the objects found in a repository, but
it used to translate these object names using the replace mechanism,
which defeats the point of enumerating all objects in the repository.
This has been corrected.
(merge bf972896d7 jk/cat-file-batch-all-wo-replace later to maint).
* Recent sparse-index work broke safety against attempts to add paths
with trailing slashes to the index, which has been corrected.
(merge c8ad9d04c6 rs/make-verify-path-really-verify-again later to maint).
* The "--color-lines" and "--color-by-age" options of "git blame"
have been missing, which are now documented.
(merge 8c32856133 bs/doc-blame-color-lines later to maint).
* The PATH used in CI job may be too wide and let incompatible dlls
to be grabbed, which can cause the build&test to fail. Tighten it.
(merge 7491ef6198 js/windows-ci-path-fix later to maint).
* Avoid performance measurements from getting ruined by gc and other
housekeeping pauses interfering in the middle.
(merge be79131a53 rs/disable-gc-during-perf-tests later to maint).
* Stop "git add --dry-run" from creating new blob and tree objects.
(merge e578d0311d rs/add-dry-run-without-objects later to maint).
* "git commit" gave duplicated error message when the object store
was unwritable, which has been corrected.
(merge 4ef91a2d79 ab/fix-commit-error-message-upon-unwritable-object-store later to maint).
* Recent sparse-index addition, namely any use of index_name_pos(),
can expand sparse index entries and breaks any code that walks
cache-tree or existing index entries. One such instance of such a
breakage has been corrected.
* The xxdiff difftool backend can exit with status 128, which the
difftool-helper that launches the backend takes as a significant
failure, when it is not significant at all. Work it around.
(merge 571f4348dd da/mergetools-special-case-xxdiff-exit-128 later to maint).
* Improve test framework around unwritable directories.
(merge 5d22e18965 ab/test-cleanly-recreate-trash-directory later to maint).
* "git push" client talking to an HTTP server did not diagnose the
lack of the final status report from the other side correctly,
which has been corrected.
(merge c5c3486f38 jk/http-push-status-fix later to maint).
* Update "git archive" documentation and give explicit mention on the
compression level for both zip and tar.gz format.
(merge c4b208c309 bs/archive-doc-compression-level later to maint).
* Drop "git sparse-checkout" from the list of common commands.
(merge 6a9a50a8af sg/sparse-index-not-that-common-a-command later to maint).
* "git branch -c/-m new old" was not described to copy config, which
has been corrected.
(merge 8252ec300e jc/branch-copy-doc later to maint).
* Squelch over-eager warning message added during this cycle.
* Fix long-standing shell syntax error in the completion script.
(merge 46b0585286 re/completion-fix-test-equality later to maint).
* Teach "git commit-graph" command not to allow using replace objects
at all, as we do not use the commit-graph at runtime when we see
object replacement.
(merge 095d112f8c ab/ignore-replace-while-working-on-commit-graph later to maint).
* "git pull --no-verify" did not affect the underlying "git merge".
(merge 47bfdfb3fd ar/fix-git-pull-no-verify later to maint).
* One CI task based on Fedora image noticed a not-quite-kosher
construct recently, which has been corrected.
* "git pull --ff-only" and "git pull --rebase --ff-only" should make
it a no-op to attempt pulling from a remote that is behind us, but
instead the command errored out by saying it was impossible to
fast-forward, which may technically be true, but not a useful thing
to diagnose as an error. This has been corrected.
(merge 361cb52383 jc/fix-pull-ff-only-when-already-up-to-date later to maint).
* The way Cygwin emulates a unix-domain socket, on top of which the
simple-ipc mechanism is implemented, can race with the program on
the other side that wants to use the socket, and briefly make it
appear as a regular file before lstat(2) starts reporting it as a
socket. We now have a workaround on the side that connects to a
unix domain socket.
* Other code cleanup, docfix, build fix, etc.
(merge f188160be9 ab/bundle-remove-verbose-option later to maint).
(merge 8c6b4332b4 rs/close-pack-leakfix later to maint).
(merge 51b04c05b7 bs/difftool-msg-tweak later to maint).
(merge dd20e4a6db ab/make-compdb-fix later to maint).
(merge 6ffb990dc4 os/status-docfix later to maint).
(merge 100c2da2d3 rs/p3400-lose-tac later to maint).
(merge 76f3b69896 tb/aggregate-ignore-leading-whitespaces later to maint).
(merge 6e4fd8bfcd tz/doc-link-to-bundle-format-fix later to maint).
(merge f6c013dfa1 jc/doc-commit-header-continuation-line later to maint).
(merge ec9a37d69b ab/pkt-line-cleanup later to maint).
(merge 8650c6298c ab/fix-make-lint-docs later to maint).
(merge 1c720357ce ab/test-lib-diff-cleanup later to maint).
(merge 6b615dbece ks/submodule-add-message-fix later to maint).
(merge 203eb8381a jc/doc-format-patch-clarify-auto-base later to maint).
(merge 559664c792 ab/test-lib later to maint).
e16 1.0.24 - 2021-11-14
------------------------
- 1.0.24
- Update po
- Danish translation update
- pagers: Remember zoom setting when changed by pager menu
- Regular expression matching fix simplification
- Regular expression matching fixes
- systray: Change default icon size to 32
- Set _NET_WM_PID on WM check window
- Drop initial setting of _NET_WM_NAME = "e16" on the root window
- Refactor atoms handling
- Avoid using "delete"
- COMPLIANCE: Update
- Support _NET_WM_STATE_FOCUSED
- Add _NET_WM_STATE_FOCUSED to the EWHM atoms
- Only update _NET_WM_STATE and _NET_WM_ALLOWED_ACTIONS when changed
- xprop.c: Do some sanity checks using static asserts if possible
- Shuffle some property stuff around
- session.c: Remove unused headers
- libhack: Add XGetWindowProperty interception
- libhack: Cosmetics
- Eliminate EGetGeometry()
- Change EXGetGeometry() to EXGetSize()
- x.c: Fix LTO warnings
- x.c: Deubug stuff cleanups
- x.c: Remove ancient unused ECreateFocusWindow()
- extinitwin: Fix parent/client synchronization
- Fix EImageDefineCursor() for window mode
- edox: Avoid LTO null pointer warning
- edox: Unifdef USE_XFONT
- Backgrounds: Fix trouble when deleting multiply referenced background
- Backgrounds: Fix bad list operation in CB_ConfigureDelBG()
- Groups: Avoid compiler warning in _DlgFillGroups()
- Drop animation in EwinSlideSizeTo() if there is no change
- Focus: Debug tweaks
- Focus: Fix focus on new clients requesting max size
- autofoo: Deal a bit more consistently with C[PP]FLAGS and LDADD
- French translation update (Philippe J. Guillaumie/Carla Sensa)
- Fix bad memory access during shutdown
- FX: Use xrender when available
- FX: Fix cleanup with compositing enabled
- FX: Fix clipping with compositing enabled
- FX: Make it work in regular compositing mode
- FX: Tweaks
- FX: Some refactoring
- FX: Fix waves cleanup
- FX: Remove "correct bugs" operation in waves
- Introduce HAVE_COMPOSITE_OVERLAY_WINDOW
- glwin.c: Fix incorrect config item size
- Sound: Enable changing engine at runtime
- Don't build disabled stuff
- Sound: Change some esd to esound
- Sound: Change some pa to pulse
- Sound: Header file tweaks
- Move Eexec() and friends to util.h
- Drop (v)snprintf() implementation
- Use installed gettext version, if possible
- Eliminate AM_LANGINFO_CODESET
- 1.0.23.000
- Fix version reported by e16 in release build
3.2.1
*New:*
- Add support for Django 3.2
*Bugfix:*
- Do not override signals receivers registered in a :meth:`~factory.django.mute_signals` context.
- Change the signature for :meth:`~factory.alchemy.SQLAlchemyModelFactory._save` and
:meth:`~factory.alchemy.SQLAlchemyModelFactory._get_or_create` to avoid argument names clashes with a field named
``session``.
1.0.0
Drop python < 3.6 support
Add type annotations to code
Remove external type annotations in pyi
Switch to poetry for packaging
Simplify internals and remove old compatability code
Switch to nox for testing
Lots of small project development changes
Same fix as for hercules4sdl.
No revbump needed since the change is limited to Darwin, where no built
package previously existed (at least for Darwin 20.6.0 which I tested).
Summary for 4.99.1 tcpdump release:
Source code:
Squelch some compiler warnings
ICMP: Update the snapend for some nested IP packets.
MACsec: Update the snapend thus the ICV field is not payload
for the caller.
EIGRP: Fix packet header fields
SMB: Disable printer by default in CMake builds
OLSR: Print the protocol name even if the packet is invalid
MSDP: Print ": " before the protocol name
ESP: Remove padding, padding length and next header from the buffer
DHCPv6: Update the snapend for nested DHCPv6 packets
OpenFlow 1.0: Get snapend right for nested frames.
TCP: Update the snapend before decoding a MPTCP option
Ethernet, IEEE 802.15.4, IP, L2TP, TCP, ZEP: Add bounds checks
ForCES: Refine SPARSEDATA-TLV length check.
ASCII/hex: Use nd_trunc_longjmp() in truncation cases
GeoNet: Add a ND_TCHECK_LEN() call
Replace ND_TCHECK_/memcpy() pairs with GET_CPY_BYTES().
BGP: Fix overwrites of global 'astostr' temporary buffer
ARP: fix overwrites of static buffer in q922_string().
Frame Relay: have q922_string() handle errors better.
Building and testing:
Rebuild configure script when building release
Fix "make clean" for out-of-tree autotools builds
CMake: add stuff from CMAKE_PREFIX_PATH to PKG_CONFIG_PATH.
Documentation:
man: Update a reference as www.cifs.org is gone. [skip ci]
man: Update DNS sections
Solaris:
Fix a compile error with Sun C
Summary for 4.99.0 tcpdump release
Improve the contents, wording and formatting of the man page.
Print unsupported link-layer protocol packets in hex.
Add support for new network protocols and DLTs: Arista, Autosar SOME/IP,
Broadcom LI and Ethernet switches tag, IEEE 802.15.9, IP-over-InfiniBand
(IPoIB), Linux SLL2, Linux vsockmon, MACsec, Marvell Distributed Switch
Architecture, OpenFlow 1.3, Precision Time Protocol (PTP), SSH, WHOIS,
ZigBee Encapsulation Protocol (ZEP).
Make protocol-specific updates for: AH, DHCP, DNS, ESP, FRF.16, HNCP,
ICMP6, IEEE 802.15.4, IPv6, IS-IS, Linux SLL, LLDP, LSP ping, MPTCP, NFS,
NSH, NTP, OSPF, OSPF6, PGM, PIM, PPTP, RADIUS, RSVP, Rx, SMB, UDLD,
VXLAN-GPE.
User interface:
Make SLL2 the default for Linux "any" pseudo-device.
Add --micro and --nano shorthands.
Add --count to print a counter only instead of decoding.
Add --print, to cause packet printing even with -w.
Add support for remote capture if libpcap supports it.
Display the "wireless" flag and connection status.
Flush the output packet buffer on a SIGUSR2.
Add the snapshot length to the "reading from file ..." message.
Fix local time printing (DST offset in timestamps).
Allow -C arguments > 2^31-1 GB if they can fit into a long.
Handle very large -f files by rejecting them.
Report periodic stats only when safe to do so.
Print the number of packets captured only as often as necessary.
With no -s, or with -s 0, don't specify the snapshot length with newer
versions of libpcap.
Improve version and usage message printing.
Building and testing:
Install into bindir, not sbindir.
autoconf: replace --with-system-libpcap with --disable-local-libpcap.
Require the compiler to support C99.
Better detect and use various C compilers and their features.
Add CMake as the second build system.
Make out-of-tree builds more reliable.
Use pkg-config to detect libpcap if available.
Improve Windows support.
Add more tests and improve the scripts that run them.
Test both with "normal" and "x87" floating-point.
Eliminate dependency on libdnet.
FreeBSD:
Print a proper error message about monitor mode VAP.
Use libcasper if available.
Fix failure to capture on RDMA device.
Include the correct capsicum header.
Source code:
Start the transition to longjmp() for packet truncation handling.
Introduce new helper functions, including GET_*(), nd_print_protocol(),
nd_print_invalid(), nd_print_trunc(), nd_trunc_longjmp() and others.
Put integer signedness right in many cases.
Introduce nd_uint*, nd_mac_addr, nd_ipv4 and nd_ipv6 types to fix
alignment issues, especially on SPARC.
Fix many C compiler, Coverity, UBSan and cppcheck warnings.
Fix issues detected with AddressSanitizer.
Remove many workarounds for older compilers and OSes.
Add a sanity check on packet header length.
Add and remove plenty of bounds checks.
Clean up pcap_findalldevs() call to find the first interface.
Use a short timeout, rather than immediate mode, for text output.
Handle DLT_ENC files *not* written on the same OS and byte-order host.
Add, and use, macros to do locale-independent case mapping.
Use a table instead of getprotobynumber().
Get rid of ND_UNALIGNED and ND_TCHECK().
Make roundup2() generally available.
Resync SMI list against Wireshark.
Fix many typos.
Summary for 1.10.1 libpcap release (so far!)
Packet filtering:
Fix "type XXX subtype YYY" giving a parse error
Source code:
Add PCAP_AVAILABLE_1_11.
Building and testing:
Rename struct bpf_aux_data to avoid NetBSD compile errors
Squelch some compiler warnings
Squelch some Bison warnings
Fix cross-builds with older kernels lacking BPF_MOD and BPF_XOR
Fix Bison detection for minor version 0.
Fix parallel build with FreeBSD make.
Get DLT_MATCHING_MAX right in gencode.c on NetBSD.
Define timeradd() and timersub() if necessary.
Fix Cygwin/MSYS target directories.
Fix symlinking with DESTDIR.
Fix generation of libpcap.pc with CMake when not building a shared
library.
Check for Arm64 as well as x86-64 when looking for packet.lib on
Windows.
Documentation:
Refine Markdown in README.md.
Improve the description of portrange in filters.
README.linux.md isn't Markdown, rename it just README.linux.
pcapng:
Support reading version 1.2, which some writers produce, and which
is the same as 1.0 (some new block types were added, but
that's not sufficient reason to bump the minor version number,
as code that understands those new block types can handle them
in a 1.0 file)
Linux:
Drop support for text-mode USB captures, as we require a 2.6.27
or later kernel (credit to Chaoyuan Peng for noting the
sscanf vulnerabilities in the text-mode code that got me to
realize that we didn't need this code any more)
Bluetooth: fix non-blocking mode.
Don't assume that all compilers used to build for Linux support
the __atomic builtins
Windows:
Add more information in "interface disappeared" error messages, in
the hopes of trying to figure out the cause.
Treat ERROR_DEVICE_REMOVED as "device was removed".
Indicate in the error message which "device was removed" error
occurred.
Report the Windows error status if PacketSendPacket() fails.
Use %lu for ULONGs in error message formats.
Don't treat the inability to find airpcap.dll as an error.
Ignore spurious error reports by Microsoft Surface mobile
telephony modem driver
rpcap:
Clean up error checking and error messages for server address
lookup.
Summary for 1.10.0 libpcap release
Add support for capturing on DPDK devices
Label most APIs by the first release in which they're available
Fix some memory leaks, including in pcap_compile()
Add pcap_datalink_val_to_description_or_dlt()
Handle the pcap private data in a fashion that makes fewer
assumptions about memory layouts (might fix GitHub issue #940
on ARM)
Fix some thread safety issues
pcap_findalldevs(): don't sort interfaces by unit number
Always return a list of supported time-stamp types, even if only
host time stamps are supported
Increase the maximum snaplen for LINKTYPE_USBPCAP/DLT_USBPCAP
Report the DLT description in error messages
Add pcap_init() for first-time initialization and global option
setting; it's not required, but may be used
Remove (unused) SITA support
Capture file reading:
Correctly handle pcapng captures with more than one IDB with a
snspshot length greater than the supported maximum
Capture file writing:
Create the file in pcap_dump_open_append() if it doesn't exist
Packet filtering:
Fix "unknown ether proto 'aarp'"
Add a new filter "ifindex" for DLT_LINUX_SLL2 files on all
platforms and live Linux captures
Add a hack to the optimizer to try to catch certain optimizer
loops (should prevent GitHub issue #112)
Show special Linux BPF offsets symbolically in bpf_image() and
bpf_dump()
Added support for ICMPv6 types 1-4 as tokens with names
Remove undocumented and rather old "ether proto" protocols
Catch invalid IPv4 addresses in filters
Don't assume ARM supports unaligned accesses
Security and other issues found by analysis:
Fix various security issues reported by Charles Smith at Tangible
Security
Fix various security issues reported by Include Security
Fix some issues found by cppcheck.
Add some overflow checks in the optimizer
rpcap:
Support rpcap-over-TLS
Redo protocol version negotiation to avoid problems with old
servers (it still works with servers using the old negotiation,
as well as servers not supporting negotiation)
Error handling cleanups
Add some new authentication libpcap error codes for specific
errors
Fix some inetd issues in rpcapd
Fix rpcapd core dumps with invalid configuration file
On UN*X, don't have rpcapd tell the client why authentication
failed, so a brute-force attacker can't distinguish between
"unknown user name" and "known user name, wrong password"
Allow rpcapd to rebind more rapidly (GitHub issue #765)
Documentation:
Improve man pages, including adding backward compatibility notes
Building and testing:
Require, and assume, some level of C99 support in the C compiler
Require Visual Studio 2015 or later if using Visual Studio
Fix configure script issues, including with libnl on Linux
Fix CMake issues
Squelch complaints from Bison about "%define api.pure" being
deprecated
Fix compilation of pcap-tc.c
Linux:
Require PF_PACKET support, and kernel 2.6.27 or later
Handle systems without AF_INET or AF_UNIX socket support
Get rid of Wireless Extensions for turning monitor mode on
Proper memory sync for PACKET_MMAP (may prevent GitHub issue
#898)
Drop support for libnl 1 and 2.
Return error on interface going away, but not if it just went
down but is still present
Set socket protocol only after packet ring configured,
reducing bogus packet drop reports
Get ifdrop stats from sysfs.
When adjusting BPF programs, do not subtract the
SLL[2]_HDR_LEN if the location is negative (special metadata
offset), to preserve references to metadata; see
https://github.com/the-tcpdump-group/tcpdump/issues/480#issuecomment-486827278
Report a warning for unknown ARPHRD types
Have pcap_breakloop() forcibly break out of a sleeping
capture loop
Add support for DSA data link types
For raw USB bus capture, use the snapshot length to set the
buffer size, and set the len field to reflect the length
in the URB (GitHub issue #808)
With a timeout of zero, wait indefinitely
Clean up support for some non-GNU libc C libraries
Add DLT_LINUX_SLL2 for cooked-mode captures
Probe CONFIGURATION descriptor of connected USB devices
Treat EPERM on ethtool ioctls as meaning "not supported", as
permissions checks are done before checking whether the
ioctl is supported at all
macOS:
Cope with getting EPWROFF from SIOCGIFMEDIA
Treat EPERM on SIOCGIFMEDIA as meaning "not supported", as
permissions checks are done before checking whether the
ioctl is supported at all
Treat ENXIO when reading packets as meaning "the interface
was removed"
Report "the interface disappeared", not "the interface went
down", if the interface was removed during a capture
FreeBSD:
Treat ENXIO as meaning "the interface was removed"
Report "the interface disappeared", not "the interface went
down", if the interface was removed during a capture
NetBSD:
Treat ENXIO as meaning "the interface was removed"
Report "the interface disappeared", not "the interface went
down", if the interface was removed during a capture
OpenBSD:
Treat EIO as meaning "the interface was removed"
Report "the interface disappeared", not "the interface went
down", if the interface was removed during a capture
DragonFly BSD:
Treat ENXIO as meaning "the interface was removed"
Report "the interface disappeared", not "the interface went
down", if the interface was removed during a capture
Solaris:
Treat ENXIO as meaning "the interface was removed"
Report "the interface disappeared", not "the interface went
down", if the interface was removed during a capture
AIX:
Fix loading of BPF kernel extension
Treat ENXIO as meaning "the interface was removed"
Report "the interface disappeared", not "the interface went
down", if the interface was removed during a capture
Windows:
Make the snapshot length work even if pcap_setfilter()
isn't called
Fix compilation on Cygwin/MSYS
Add pcap_handle(), and deprecate pcap_fileno()
Report PCAP_ERROR_NO_SUCH_DEVICE for a non-existent device
Return an appropriate error message for device removed or
device unusable due to a suspend/resume
Report a warning for unknown NdisMedium types
Have pcap_breakloop() forcibly break out of a sleeping
capture loop
Clean up building DLL
Handle CRT mismatch for pcap_dump_fopen()
Map NdisMediumWirelessWan to DLT_RAW
Add AirPcap support in a module, rather than using
WinPcap/Npcap's support for it
Report the system error for PacketSetHwFilter() failures
Add support for getting and setting packet time stamp types
with Npcap
Have pcap_init() allow selecting whether the API should use
local code page strings or UTF-8 strings (including error
messages)
Haiku:
Add capture support
