Only run cbindgen when necessary. This is a bit tricky. When
building a dist we want to unconditionally build the headers.
When going through a "make; sudo make install" type process,
cbindgen should not be run as the headers already exist, are
valid, and the environment under sudo is more often than
not suitable to pick up the Rust toolchains when installed
with rustup.
For the normal "make" case we have the gen/rust-bindings.h file
depend on library file, this will cause it to only be rebuilt
if the code was modified.
For "make dist" we unconditionally create "dist/rust-bindings.h".
This means the generated file could be in 2 locations, so update
configure.ac, and the library search find to find it.
The "gen/rust-bindings.h" should be picked up first if it exists,
for those who develop from a dist archive where "dist/rust-bindings.h"
also exists.
Not completely happy having the same file in 2 locations, but not
sure how else to get the dependency tracking correct.
Support either the __thread GNUism or the C11 _Thread_local.
Use 'thread_local' to point to the one that is used. Convert existing
__thread user to 'thread_local'.
Remove non-thread-local code from the packet pool code.
Expose UTH flow builder to new 'FUZZ' define as well. Move UTHbufferToFile
as well and rename it to a more generic 'TestHelperBufferToFile'.
This way UNITTESTS can be disabled. This leads to smaller code size
and more realistic testing as in some parts of the code things
behave slightly differently when UNITTESTS are enabled.
If autodiscovery of libnss was used (default), then the line
libnss support: yes
was never set to no.
Same behavior for libnspr.
Broken by commit 'configure: fix nspr check logic' (7ea269a212)
The modifications as part of the cbindgen commit caused issues
with distcheck, revert the Makefile to how it was with the Python
generator, but still using cbindgen.
Also always assume we'll include the generated headers in the
distribution archive to fix make distcheck from distribution
archives with headers included, but no cbindgen.
The use of AC_CHECK_FILE and AC_CHECK_FILES cause the following error
when cross-compiling:
configure: error: cannot check for file existence when cross compiling
The solution is to check for the file directly instead of using a macro.
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
pcap can depends on nl-3 so use pkg-config to find these dependencies
otherwise all AC_CHECK_LIB calls will fail when building statically
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Don't try to run suricata-update if its not installed.
The 'make install-rules' target would try to run suricata-update
when it was detected that it was bundled, but didn't consider
if suricata-update was actually installed.
If rustup is in use, and a user uses sudo or su for the make
install, the install may fail with a "no default toolchain"
error.
To prevent this, detect at configure if rustup is being used,
then set RUSTUP_HOME for all calls to cargo.
Instead of telling the user what packages to install for missing
Python modules, give generic instructions about what module
needs to be installed.
It is getting tricky to get these package names correct
across distributions.
Don't detect the Python version, it is not needed anyways,
all we need is the Python path.
Also, python2 --version prints to stderr, while python3
prints to stdout, leading to some odd output during
./configure (but fixable).
Replace index by strchr and rindex by strrchr.
index(3) states "POSIX.1-2008 removes the specifications of index() and
rindex(), recommending strchr(3) and strrchr(3) instead."
Add index/rindex to banned function check so they don't get reintroduced.
Bug #1443.
As AC_SUBST doesn't expand to anything in the shell script, this
will generate a bad script on older versions of autoconf.
Change the logic to eliminate the possibility of an empty
if or else block.
Redmine issue:
https://redmine.openinfosecfoundation.org/issues/3124
Get rid of enable-rust-debug flag and use enable-debug for acheiving the
desired functionality. From now, adding `--enable-debug` to `configure`
shall create an [unoptimitized + debuginfo] target. Rest behavior stays
the same.
Closes redmine ticket #3054
This define is used to remove reference to capture bypass in case
no capture method implementing this is active.
This patch also introduces CAPTURE_OFFLOAD_MANAGER that is defined
if we need the flow bypass manager code.
This change makes it possible to generate the eBPF programs even if
Suricata itself is built a different C compiler. It also simplifies
how the correct llc program is detected.
Implements Feature https://redmine.openinfosecfoundation.org/issues/2789
Date makes it even clearer that when was the last commit for the build
that one is running. Add this info alongwith rev. Change inspired by
rustc.
Before
```
$ suricata -V
This is Suricata version 5.0.0-dev (rev 2d217e666)
```
After
```
This is Suricata version 5.0.0-dev (2d217e666 2019-07-12)
```
Closes redmine ticket #3092
* added nss and nspr requirements for appveyor build
* added nss and nspr requirements for travis builds
* added travis build without nss and nspr
Signed-off-by: jason taylor <jtfas90@gmail.com>
- There is now an option to automatically create streams on the
correct NUMA node when using cpu affinity.
- When not using cpu affinity the user can specify streams to be
created in the suricata.yaml file. It is no longer required to
use NTPL to create streams before running suricata.
- The legacy usage model of running NTPL to create streams is still
available. This can be used for legacy configurations and complex
configurations that cannot be satisfied by the auto-config option.
* the lzma check during configure wasn't properly displaying the
additional information on how to install if --enable-lzma was passed
but lzma devel files were not present
* updated additional information blocks to include distribution
package names
* minor formatting updates to add quotes around variables
Signed-off-by: jason taylor <jtfas90@gmail.com>
Require distutils to install the Python tools. Update the logic
to only install suricatactl (and suricatasc) if Python and
distutils are found. Suricata-Update will only be installed if
bundled, and python-distutils and python-yaml are found.
When looking for Python, prefer "python3" over "python2" and
"python".
Also add information about the Python path and version to the
./configure summary.
Redmine issue:
https://redmine.openinfosecfoundation.org/issues/2808
Process multiple packets at nm_dispatch. Use zero copy for workers
recv mode.
Add configure check netmap check for API 11+ and find netmap api version.
Add netmap guide to the userguide.
Currently, if the user provides --enable-libmagic or
--disable-libmagic, libmagic will be disabled because $enableval is not
used to know if the user provided --enable or --disable
Most of the options have this issue so fix them all by using $enableval
Fixes:
- https://redmine.openinfosecfoundation.org/issues/2797
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
This is to prevent suricata-update from being installed if it
would otherwise be installed based on in being bundled, and
its dependencies being available.
Warn the user that Suricata-Update will not be installed if it
is bundled, but python-yaml is missing (this will also cover
the case where Python is missing).
Add "Install suricata-update" to the build summary. For consistency,
relable "Suricatasc install" as "Install suricatasc".
Remove 'experimental' label for Rust, and enable it by default if
rustc and cargo (and libjansson) are available.
Add rustc and cargo versions to the build-info.
Jason Ish
Jeff Lucovsky
Philippe Antoine
Shivani Bhardwaj
Victor Julien
Eric Leblond
Pierre Chifflier
Phil Young
Danny Browning
Fabrice Fontaine
Ciprian
Hilko Bengen
Bill Meeks
Andreas Herz
jason taylor
Jason Ish