Passive DNS collection using Bro
csirtgadgets/bearded-avenger 150
CIF v3 -- the fastest way to consume threat intelligence
IP Address to ASN/prefix/owner/cc lookup server
simple tcp port scanner + banner grabber
react stuff
JustinAzoff/bro-statsd-plugin 8
Statsd client for Bro.
The FASTEST way to Consume Threat Intelligence
line based tcp load balancing proxy.
JustinAzoff/asnlookup-client-python 2
Python client for asnlookup server
push eventzeek/spicy
commit sha 482b0b071b4f0695c33e5e87c2ae85a4747ddd3a
Allow to enforce sequential execution of jobs spawned during JIT. By default JIT spawns compilation jobs in parallel to optimize throughput. This can lead to issues when working with hard resource constraints, e.g., in containers, as peak memory is driven by the hidden number of launched compilation jobs is hard to control without knowing exactly which jobs will be launched. This patch modifies the JIT code so that if the environment variable `HILTI_JIT_SEQUENTIAL` is set jobs are executed sequentially.
commit sha f9a5c0e8b908f5545ce309236e8f24f3a2766712
Use sequential JIT mode in CI. Since in Cirrus we work in containers with relatively little memory. By default JIT executes jobs in parallel, but this can lead to us exceeded the memory available in CI containers leading to OOM kills. Instead use sequential JIT mode so resource usage can be controlled on the top-level via BTest's `-j` flag.
commit sha 70c3f5e5f762709e9e9c24c75f4e4f2fd0a30651
Merge remote-tracking branch 'origin/topic/bbannier/ci-sequential-jit'
commit sha 9e9178e688394ed0188fb481b3c12a28a8e5decf
Use "default" compiler if possible in Docker images. Since we now compile directly with the system compiler we do not strictly require LLVM anymore. This patch simplifies the in-tree Docker images so they use the default compiler for the platform if possible; since we require a C++17-capable compiler we still need to install custom compilers on a number of platforms.
push time in 2 hours
PR opened zeek/zeek
There have been a bunch of mostly TLS 1.3 related changes.
pr created time in 2 hours
create barnchzeek/zeek
created branch time in 2 hours
push eventzeek/spicy
commit sha 482b0b071b4f0695c33e5e87c2ae85a4747ddd3a
Allow to enforce sequential execution of jobs spawned during JIT. By default JIT spawns compilation jobs in parallel to optimize throughput. This can lead to issues when working with hard resource constraints, e.g., in containers, as peak memory is driven by the hidden number of launched compilation jobs is hard to control without knowing exactly which jobs will be launched. This patch modifies the JIT code so that if the environment variable `HILTI_JIT_SEQUENTIAL` is set jobs are executed sequentially.
commit sha f9a5c0e8b908f5545ce309236e8f24f3a2766712
Use sequential JIT mode in CI. Since in Cirrus we work in containers with relatively little memory. By default JIT executes jobs in parallel, but this can lead to us exceeded the memory available in CI containers leading to OOM kills. Instead use sequential JIT mode so resource usage can be controlled on the top-level via BTest's `-j` flag.
commit sha 70c3f5e5f762709e9e9c24c75f4e4f2fd0a30651
Merge remote-tracking branch 'origin/topic/bbannier/ci-sequential-jit'
commit sha f1005ffccf2f5d3c214d022cc03eeb4ac2748307
Add CI job to validate release tarball.
push time in 2 hours
PR opened zeek/spicy
Cirrus tasks with only_if are hidden unless there condition is met.
With that it becomes impossible to trigger them from PRs. Use skip
with an inverted condition instead so we can optionally trigger them
even from PRs (this can become handy to check more expensive tasks
before merging).
pr created time in 3 hours
create barnchzeek/spicy
branch : topic/bbannier/ci-triggerable-master-only-tasks
created branch time in 3 hours
push eventzeek/spicy
commit sha 2e9dfa116025d191288563619764c4bb764d4b17
Fix Zeek plugin build when toolchain is disabled. This got broken during the recent refactor: the plugin wouldn't compile anymore without the toolchain being available for JIT. I have now moved most of the plugin's JIT-related functionality into a separate implementation file that's only compiled when the toolchain is available; plus a small number of #ifdef to activate that code.
push time in 3 hours
Pull request review commentzeek/spicy
#include <zeek-spicy/zeek-compat.h> #include <zeek-spicy/zeek-reporter.h> +#ifdef HAVE_TOOLCHAIN namespace spicy::zeek::debug { const hilti::logging::DebugStream ZeekPlugin("zeek"); }+#endif plugin::Zeek_Spicy::Plugin SpicyPlugin; plugin::Zeek_Spicy::Plugin* ::plugin::Zeek_Spicy::OurPlugin = &SpicyPlugin;
Good catch!
comment created time in 3 hours
Pull request review commentzeek/spicy
#include <zeek-spicy/zeek-compat.h> #include <zeek-spicy/zeek-reporter.h> +#ifdef HAVE_TOOLCHAIN namespace spicy::zeek::debug { const hilti::logging::DebugStream ZeekPlugin("zeek"); }+#endif plugin::Zeek_Spicy::Plugin SpicyPlugin; plugin::Zeek_Spicy::Plugin* ::plugin::Zeek_Spicy::OurPlugin = &SpicyPlugin;
(commenting here since I cannot leave a comment on l.42)
When building with BUILD_TOOLCHAIN=OFF we do not link against hilti and hilti::util::fmt will be unknown. Let's use hilti::rt::fmt instead and include <hilti/rt/fmt.h>.
comment created time in 3 hours
push eventzeek/spicy
commit sha c1b407cb961147ced6e363583f6bbdb88959d794
Move `spicy::rt::getenv` to `hilti::rt`.
commit sha 482b0b071b4f0695c33e5e87c2ae85a4747ddd3a
Allow to enforce sequential execution of jobs spawned during JIT. By default JIT spawns compilation jobs in parallel to optimize throughput. This can lead to issues when working with hard resource constraints, e.g., in containers, as peak memory is driven by the hidden number of launched compilation jobs is hard to control without knowing exactly which jobs will be launched. This patch modifies the JIT code so that if the environment variable `HILTI_JIT_SEQUENTIAL` is set jobs are executed sequentially.
commit sha f9a5c0e8b908f5545ce309236e8f24f3a2766712
Use sequential JIT mode in CI. Since in Cirrus we work in containers with relatively little memory. By default JIT executes jobs in parallel, but this can lead to us exceeded the memory available in CI containers leading to OOM kills. Instead use sequential JIT mode so resource usage can be controlled on the top-level via BTest's `-j` flag.
commit sha 70c3f5e5f762709e9e9c24c75f4e4f2fd0a30651
Merge remote-tracking branch 'origin/topic/bbannier/ci-sequential-jit'
push time in 3 hours
push eventzeek/spicy
commit sha 482b0b071b4f0695c33e5e87c2ae85a4747ddd3a
Allow to enforce sequential execution of jobs spawned during JIT. By default JIT spawns compilation jobs in parallel to optimize throughput. This can lead to issues when working with hard resource constraints, e.g., in containers, as peak memory is driven by the hidden number of launched compilation jobs is hard to control without knowing exactly which jobs will be launched. This patch modifies the JIT code so that if the environment variable `HILTI_JIT_SEQUENTIAL` is set jobs are executed sequentially.
commit sha f9a5c0e8b908f5545ce309236e8f24f3a2766712
Use sequential JIT mode in CI. Since in Cirrus we work in containers with relatively little memory. By default JIT executes jobs in parallel, but this can lead to us exceeded the memory available in CI containers leading to OOM kills. Instead use sequential JIT mode so resource usage can be controlled on the top-level via BTest's `-j` flag.
push time in 3 hours
push eventzeek/spicy
commit sha 45f95843fac53279d30f768382903052f60b2ec2
Increase minimum fiber stack size. macOS needs more it seems when compiling with the native compiler. Closes #664.
commit sha 61147fb5bbb56d7ed603523d449670a96cd17ef9
Merge pull request #665 from zeek/topic/robin/increase-stack-size Increase minimum fiber stack size.
commit sha 25734324d28fdf8556f46f8f4c05a25ca0757818
On macOS, pass CMAKE_OSX_SYSROOT through to JIT linker.
commit sha 9036726a7751164c699dee31b5136e8b0ccb1377
Merge remote-tracking branch 'origin/topic/robin/big-sur-fix' * origin/topic/robin/big-sur-fix: On macOS, pass CMAKE_OSX_SYSROOT through to JIT linker.
commit sha 1120455b20d2a087058659a644789f8422162362
Add master CI task for macOS BigSur. This work is part of #582.
commit sha 1fc66ff348a967f63f325d5c142c56ee14b7528e
Remove pre-commit hooks which require external tool installation.
commit sha e511a5d9ec8025a6311c9e02398530de3002d9b9
Merge pull request #670 from zeek/topic/bbannier/pre-commit-no-external-tools Remove pre-commit hooks which require external tool installation.
commit sha d637872a2265086cf7c532547ff86fcf77c159fd
Merge pull request #643 from zeek/topic/bbannier/ci_macos_big_sur Add master CI task for macOS BigSur.
commit sha c1b407cb961147ced6e363583f6bbdb88959d794
Move `spicy::rt::getenv` to `hilti::rt`.
commit sha d318006170d4b4d93c764eaf2e31e1c910823d16
Allow to enforce sequential execution of jobs spawned during JIT. By default JIT spawns compilation jobs in parallel to optimize throughput. This can lead to issues when working with hard resource constraints, e.g., in containers, as peak memory is driven by the hidden number of launched compilation jobs is hard to control without knowing exactly which jobs will be launched. This patch modifies the JIT code so that if the environment variable `HILTI_JIT_SEQUENTIAL` is set jobs are executed sequentially.
commit sha cc26407b7e93fe9c6b747d3686847e13d8301fea
Use sequential JIT mode in CI. Since in Cirrus we work in containers with relatively little memory. By default JIT executes jobs in parallel, but this can lead to us exceeded the memory available in CI containers leading to OOM kills. Instead use sequential JIT mode so resource usage can be controlled on the top-level via BTest's `-j` flag.
commit sha 0fe1b718774a22a2d6ec41e52da2670069fd2644
Merge remote-tracking branch 'origin/topic/bbannier/ci-sequential-jit' into topic/bbannier/test_release_tarball # Conflicts: # .cirrus.yml
push time in 4 hours
push eventzeek/spicy
commit sha c1b407cb961147ced6e363583f6bbdb88959d794
Move `spicy::rt::getenv` to `hilti::rt`.
commit sha d318006170d4b4d93c764eaf2e31e1c910823d16
Allow to enforce sequential execution of jobs spawned during JIT. By default JIT spawns compilation jobs in parallel to optimize throughput. This can lead to issues when working with hard resource constraints, e.g., in containers, as peak memory is driven by the hidden number of launched compilation jobs is hard to control without knowing exactly which jobs will be launched. This patch modifies the JIT code so that if the environment variable `HILTI_JIT_SEQUENTIAL` is set jobs are executed sequentially.
commit sha cc26407b7e93fe9c6b747d3686847e13d8301fea
Use sequential JIT mode in CI. Since in Cirrus we work in containers with relatively little memory. By default JIT executes jobs in parallel, but this can lead to us exceeded the memory available in CI containers leading to OOM kills. Instead use sequential JIT mode so resource usage can be controlled on the top-level via BTest's `-j` flag.
commit sha fe9dd082234df3e03672a35ad02b7628b7cf5f1a
Merge remote-tracking branch 'origin/topic/bbannier/ci-sequential-jit'
commit sha a5c7fbd630c7e4e47b52a0ac06e1e8d4658dceb9
fixup! Remove oversubscription in CI.
push time in 4 hours
push eventzeek/spicy
commit sha 1120455b20d2a087058659a644789f8422162362
Add master CI task for macOS BigSur. This work is part of #582.
commit sha 1fc66ff348a967f63f325d5c142c56ee14b7528e
Remove pre-commit hooks which require external tool installation.
commit sha e511a5d9ec8025a6311c9e02398530de3002d9b9
Merge pull request #670 from zeek/topic/bbannier/pre-commit-no-external-tools Remove pre-commit hooks which require external tool installation.
commit sha d637872a2265086cf7c532547ff86fcf77c159fd
Merge pull request #643 from zeek/topic/bbannier/ci_macos_big_sur Add master CI task for macOS BigSur.
commit sha 03e1c046d4d7877736c2c9265c208229baa5d85f
Fix parsing of empty look-ahead vectors. That could trigger a parse error inside a sink. Closes #671.
commit sha 294dc4b511f5270a092cbc3d5d4179d1bfb6e6c6
Add final newline to ASCII output of spicy-dump.
push time in 5 hours
Pull request review commentzeek/spicy
hilti::Result<Nothing> JIT::_compile() { if ( auto rc = _spawnJob(hilti::configuration().cxx, std::move(args)); ! rc ) return rc.error();++ if ( sequential )+ if ( auto rc = _waitForJobs(); ! rc )+ return rc.error();
Style nit
if ( sequential ) {
if ( auto rc = _waitForJobs(); ! rc )
return rc.error();
}
comment created time in 5 hours
issue openedzeek/spicy
inout does not work with basic types
According to the documentation at https://docs.zeek.org/projects/spicy/en/latest/programming/parsing.html?highlight=inout#unit-parameters, a string parameter should be able to be declared as inout.
Trying this:
module TESTING;
public type Handshake = unit(inout version: string) {
handshakes: bytes &size=0;
};
Yields the following error message:
[error] test3.spicy:3:1-5:3: only parameters of reference type can be 'inout' for struct parameters
Expected behavior: no error and parameter is modifiable in the unit.
created time in 5 hours
push eventzeek/spicy
commit sha 05fb70f49e09bede9913fbc8267bae2044e2518a
Add final newline to ASCII output of spicy-dump.
push time in 8 hours
PR merged zeek/zeek
The docs on https://cirrus-ci.org/guide/macOS/ changed from "osx_instance" to "macos_instance", so let's reflect that.
Companion PRs are zeek/broker#169 and zeek/zeek-archiver#6.
pr closed time in 11 hours
push eventzeek/zeek
commit sha b7b63e889d4f197b9b084979f708188f44bd08e2
Update Cirrus CI naming update for MacOS images The docs on https://cirrus-ci.org/guide/macOS/ changed from "osx_instance" to "macos_instance", so let's reflect that.
commit sha a98012a8fb8cafb5940bf097db932ec52a6a8195
Merge remote-tracking branch 'origin/topic/christian/osx-macos-renaming' * origin/topic/christian/osx-macos-renaming: Update Cirrus CI naming update for MacOS images
push time in 11 hours
push eventzeek/zeek-archiver
commit sha 7fe80115aa159119febc623cec0660774bcae50c
Updating CHANGES and VERSION.
push time in 11 hours
PR merged zeek/zeek-archiver
The docs on https://cirrus-ci.org/guide/macOS/ changed from "osx_instance" to "macos_instance", so let's reflect that.
pr closed time in 11 hours