Arch Linux User Repository

36dfe40cb6dd seems to fix the unit test issue. I was just planning to clean up the redundant CMake run, I imagine some of the artefacts from re-running it might have been causing issues since the artefacts from the original make aren't cleaned up before make test.

Regardless, we only build once and it now works when building it on my test machines.

Some progress notes.

a978babfa3fa fixes the duplicate arch entry & overrides the min. CMake version, allowing builds (tests still NG).

67b8d4eae815 should fix issues with tests that require a remote host, but it appears there are still some issues regarding the two unit tests, №9 (test_packets_static) & №10 (test_packets), and test №8 (test_dynamic_groups).

The issue with test №8 appears to be an assertion failure when testing ipv6, on the first bit.

test_dynamic_groups: /usr/src/debug/rtrlib/rtrlib-0.8.0/rtrlib/lib/ipv6.c:32: lrtr_ipv6_get_bits: Assertion `first_bit <= 127' failed.

№9 fails in two places, an assert failure on test_set_last_update and a segfault on test_rtr_send_error_pdu.

[ RUN      ] test_set_last_update
[  ERROR   ] --- 0 != -1
[   LINE   ] --- /usr/src/debug/rtrlib/rtrlib-0.8.0/tests/unittests/test_packets_static.c:70: error: Failure!
[  FAILED  ] test_set_last_update
...
[ RUN      ] test_rtr_send_error_pdu
[  ERROR   ] --- Test failed with exception: Segmentation fault(11)
[  FAILED  ] test_rtr_send_error_pdu

№10 segfaults on test_rtr_send_reset_query.

[ RUN      ] test_rtr_send_reset_query
(2026/06/11 20:51:52:923842): RTR Socket: Sending reset query
[  ERROR   ] --- Test failed with exception: Segmentation fault(11)
[  FAILED  ] test_rtr_send_reset_query

All appear to act the same across different hardware and software, tested on 3 of my Arch machines. When running the build as the PKGBUILD does on a Debian container, everything succeeds as expected.

Package is seemingly orphaned. A typo has been present in the PKGBUILD for 2 years now (armv7h *2 instead of one being armv7l). mroethke's last login was 12 days ago as of writing yet no action has been taken, maintainer w784319947's last login was over a year ago.

Tests appear to be failing due to hardcoded remote testing server, and upstream developer hosting the testing infrastructure seems to have stopped that hosting. Submitting request for orphaning, and will be looking to adopt.

Not sure how we will work around the test, or if we will have to strip those tests and run untested code.

I have gotten the git version of the package to successfully build and pass tests. I will begin to look to see if the remote testing host has changed since release 0.8.0 and latest commit.

You have a typo in your PKGBUILD file where you have duplicate arch values:

arch=(x86_64 i686 aarch64 armv7h armv7h)

Notice armv7h is listed twice. Please remove

Compatibility with CMake < 3.5 has been removed from CMake.

Some tests failed.

ref: https://github.com/Jguer/yay/issues/750

ref: https://github.com/rtrlib/rtrlib/issues/274

Solution for me:

yay -G rtrlib

edit the PKGBUILD, add "-DCMAKE_POLICY_VERSION_MINIMUM=3.5 \" in prepare(), and comment the line: "make test"

makepkg -si

Build failed with newer CMake & failed tests:

CMake Error at CMakeLists.txt:5 (cmake_minimum_required):                                                                                                                                                              
  Compatibility with CMake < 3.5 has been removed from CMake.                                                                                                                                                          

  Update the VERSION argument <min> value.  Or, use the <min>...<max> syntax                                                                                                                                           
  to tell CMake that the project requires at least <min> but has been updated                                                                                                                                          
  to work with policies introduced by <max> or earlier.                                                                                                                                                                

  Or, add -DCMAKE_POLICY_VERSION_MINIMUM=3.5 to try configuring anyway.

Failed Tests:

Running tests...
Test project /home/neo_chen/.cache/yay/rtrlib/src/rtrlib-0.8.0
Connected to MAKE jobserver
      Start  1: test_pfx
 1/10 Test  #1: test_pfx .........................   Passed    0.36 sec
      Start  2: test_trie
 2/10 Test  #2: test_trie ........................   Passed    0.00 sec
      Start  3: test_ht_spkitable
 3/10 Test  #3: test_ht_spkitable ................   Passed    0.61 sec
      Start  4: test_ht_spkitable_locks
 4/10 Test  #4: test_ht_spkitable_locks ..........   Passed    1.00 sec
      Start  5: test_live_validation
 5/10 Test  #5: test_live_validation .............***Failed    2.00 sec
      Start  6: test_ipaddr
 6/10 Test  #6: test_ipaddr ......................   Passed    0.00 sec
      Start  7: test_getbits
 7/10 Test  #7: test_getbits .....................   Passed    0.00 sec
      Start  8: test_dynamic_groups
 8/10 Test  #8: test_dynamic_groups ..............***Failed    1.00 sec
      Start  9: test_packets_static
 9/10 Test  #9: test_packets_static ..............   Passed    0.00 sec
      Start 10: test_packets
10/10 Test #10: test_packets .....................   Passed    0.00 sec

80% tests passed, 2 tests failed out of 10

Total Test time (real) =   4.98 sec

The following tests FAILED:
          5 - test_live_validation (Failed)
          8 - test_dynamic_groups (Failed)

done

Ah, interesting, makes sense; I got my semantics wrong. Well, so if you could add armv7h and armv7l that'd be great :-) Thanks

Sure.

rtrlib is compiled from C to an architecture specific binary. If you compile this binary for ARM you cannot use it on x86 and vice versa. With a properly specified arch array pacman will prevent you from installing a package build for arm on x86. If I specify "any" pacman will not do that, because it thinks the package can be used on any architecture.

Can you explain the "containing compiled binaries" part please ? I don't really understand: isn't rtrlib compiled?

For the arch, at least armv7h. Thanks