We are pleased to announce that Hyperscan 4.4 has been pushed to Github.
This release adds a new feature where the Hyperscan library is able to determine the and select the runtime binary code to use for the host processor it is executing on. This feature, that we are calling the “fat runtime” embeds multiple copies of the Hyperscan runtime into the compiled application, and chooses the runtime based on the microarchitectural features supported by the host.
Also included in this release is the “hsbench” benchmarking tool, providing an easy way to measure Hyperscan’s performance for particular sets of patterns and data.
This release also contains a number of compile-time and runtime perfomance improvements, and a small number of bugfixes.
For more details please refer to the change log on Github.
The 4.4 release can be found here.