# \[cxx feature check\] what satisfies GNU\_POSIX\_REGEX | POSIX\_REGEX?

**URL:** https://discourse.cmake.org/t/cxx-feature-check-what-satisfies-gnu-posix-regex-posix-regex/2218
**Category:** Usage
**Created:** [November 23, 2020, 12:48pm UTC](https://discourse.cmake.org/t/cxx-feature-check-what-satisfies-gnu-posix-regex-posix-regex/2218 "2020-11-23T12:48:55Z")
**Posts on this page:** 3
**Page:** 1

<div class="post-metadata">

### Author: ![\_0.z](https://discourse.cmake.org/letter_avatar_proxy/v4/letter/_/5daacb/32.png) [@\_0.z](https://discourse.cmake.org/u/_0.z)
#### Post date: [November 23, 2020, 12:48pm UTC](https://discourse.cmake.org/t/cxx-feature-check-what-satisfies-gnu-posix-regex-posix-regex/2218/1 "2020-11-23T12:48:55Z")

</div>

trying to compile llvm 11 from source with cmake 3.19, former pulling in Google’s benchmarking [https://github.com/google/benchmark/blob/master/CMakeLists.txt#L263](https://github.com/google/benchmark/blob/master/CMakeLists.txt#L263) which tests for

- GNU\_POSIX\_REGEX
- POSIX\_REGEX

and the test fails as apparently neither is available on the node.

since the node is based on the Debian package feed I was trying to find a package that satisfies either one but came up empty.

---

<div class="post-metadata">

### Author: ![\_0.z](https://discourse.cmake.org/letter_avatar_proxy/v4/letter/_/5daacb/32.png) [@\_0.z](https://discourse.cmake.org/u/_0.z)
#### Post date: [November 23, 2020, 1:48pm UTC](https://discourse.cmake.org/t/cxx-feature-check-what-satisfies-gnu-posix-regex-posix-regex/2218/2 "2020-11-23T13:48:58Z")

</div>

installed `libghc-regex-posix-prof` package but that did not help

---

<div class="post-metadata">

### Author: ![\_0.z](https://discourse.cmake.org/letter_avatar_proxy/v4/letter/_/5daacb/32.png) [@\_0.z](https://discourse.cmake.org/u/_0.z)
#### Post date: [November 23, 2020, 4:06pm UTC](https://discourse.cmake.org/t/cxx-feature-check-what-satisfies-gnu-posix-regex-posix-regex/2218/3 "2020-11-23T16:06:14Z")

</div>

package _gnulib_ should be satisfactory but it seems that the benchmark code is buggy.
