Ivan Maidanski

Minimal configuration (no threads) to reproduce the issue: gcc -I include -O1 -D MPROTECT_VDB -D NO_TEST_HANDLE_FORK -o gctest tests/gctest.c extra/gc.c && ./gctest (probably retry ./gctest several times) Output: Switched to...

Is it reproducible with GC_RETRY_SIGNALS? GC_RETRY_SIGNALS=1 ./gctest

> To reproduce, you can use the machine gcc113.fsffrance.org on the GCC compile farm Unfortunately, it seems to take time to get the access (I've sent a request). Meanwhile, could...

I failed to reproduce it on travis (https://app.travis-ci.com/github/ivmai/bdwgc/builds/238450296) on latest release-8_0 branch (commit 5b05db68955 plus GC_RETRY_SIGNALS=1) running gctest in a loop (>60 times). Host: Ubuntu 16.04.6 LTS / arm64

\+ @wangp @cpu82 @hboehm @magv @Constellation

> Do you think that the above pthread(3) extensions would integrate nicely into bdwgc? Seems yes. This is even simpler than GC_OPENBSD_UTHREADS case. > repeat for each thread We should...

Is this observed on latest master?

Please reproduce with configure --disable-parallel-mark

Is this 100% reproducible? Looks like suspend signal is not delivered. Could you please do "bt" for other threads (with undef PARALLEL_MARK)?

Probably [tesseract](https://github.com/tesseract-ocr/tesseract) project could be used as a sample.