apteryks

I'm also hitting that issue in GNU Guix, using GCC 14 as well.

More specifically, the exact problem I see is this, in the `config.log` file produced at configure-time when building the gtick program: ``` configure:5795: gcc -o conftest -g -O2 conftest.c -ldmalloc...

A workaround is to only build the static version of the library.

But it seems this is not the only issue, as there appear to be issues with the definitions that dmalloc makes when using GCC 14 and glibc 2.41. I've written...

In my case, I'm happy with shared libraries, so I added `-DBUILD_SHARED_LIBS=ON` to both libtommath and libtomcrypt, like so: ```scheme (define-public libtomcrypt (let ((commit "563f0fb4f2a2c6a665ce9d091fac7956dbf7c40a") (revision "0")) (package (name "libtomcrypt")...

I guess this is because the libtomcrypt build system assumes libtommath includes are somewhere relative to CMAKE_INSTALL_PREFIX, which is not the case in Guix.

I get similar errors, trying to build `sysdig` from the master branch or its latest `0.40.1` release with falcosecurity-libs 0.22.2. Here, with sysdig commit 034681784d1e8ff2b7b71d8a1d8a314f40f044f2: ``` cd /tmp/guix-build-sysdig-0.40.1.drv-0/build/userspace/sysdig && /gnu/store/wssvk5s5akbx78kx92bdkirrry87jvbj-gcc-14.3.0/bin/c++...

> This should be fixed by the latest release. I'm still getting these errors: ``` /tmp/guix-build-sysdig-0.41.2.drv-0/source/userspace/sinspui/cursescomponents.cpp: In member function ‘const char* spy_text_renderer::process_event_spy(sinsp_evt*, int64_t*)’: /tmp/guix-build-sysdig-0.41.2.drv-0/source/userspace/sinspui/cursescomponents.cpp:134:36: error: ‘const class sinsp_evt_param’ has no...

> May you point me to a distribution that ships `watcher` so I can test using it as a system library? It's available in Guix (too recent to be listed...

Since the upstream of watcher have now included new pkg-config files in their upcoming 0.13.7 release (see: https://github.com/e-dant/watcher/commit/bd02b11ae7a15e4bb10c3e3e8cea5011310dbaf4), I've simplified this PR to a single (existing) macro call. I believe...