It's been a few hours since any log last got written to. PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 734 portage 20 0 789696 164424 10368 R 85.0 0.5 108:21.57 test_libaom 729 portage 20 0 789696 164388 10112 R 84.7 0.5 108:16.56 test_libaom 735 portage 20 0 789696 164172 10112 R 84.4 0.5 108:19.75 test_libaom 736 portage 20 0 889572 262516 10496 R 84.1 0.8 108:19.50 test_libaom 731 portage 20 0 889552 262980 10368 R 83.7 0.8 108:07.10 test_libaom 730 portage 20 0 789700 164428 10368 R 82.4 0.5 108:27.40 test_libaom 733 portage 20 0 789696 164604 10368 R 82.1 0.5 108:19.43 test_libaom 737 portage 20 0 889576 262340 10240 R 81.7 0.8 108:16.96 test_libaom 732 portage 20 0 889540 262336 10112 R 80.7 0.8 108:24.73 test_libaom
Created attachment 895068 [details] emerge --info '=media-libs/libaom-3.8.1
Haven't measured myself but these tests are known to take several hours, encoding with libaom is very slow by design (it's the reference implementation and not meant to be fast).
Most of my confusion stems from both the build logs not saying they're done building & the lack of reporting/logging from the tests themselves. The files in the package's logging dir haven't been touched for 5 hours with tests still running.
[==========] 9999 tests from 87 test suites ran. (26869680 ms total) [ PASSED ] 9999 tests. It took almost 7 and a half hours on my i5-9400.
Guess one could argue that skipping most tests downstream (or at least most of the slower ones) would be worthwhile, not to have unnecessary maintainer+stabilization burden potentially leading to just ignoring tests entirely. That aside, I do discourage any users from running these :)
