Robert O'Callahan
Robert O'Callahan
b422fed53f6c8b17299c087622111a4e4b47daba should fix the `infallible_send_fd` cases but there will be more. I have no idea what the replay timeout is about :-(.
4bd2467de3b402f1edb4a8423c3a720d8f10929b to fix infallible_mmap.
``` [ERROR /home/roc/rr/rr/src/TraceStream.cc:1156:read_mapped_region()] Metadata of /cache/build/exclusive-amdci1-0/julialang/rr/obj/lib/rr/librrpage.so changed: replay divergence likely, but continuing anyway. inode: 101038655/96736484; mode: 33261/33261; uid: 1000/0; gid: 1000/0; size: 18896/18896; mtime: 0/1648257863 [ERROR /home/roc/rr/rr/src/TraceStream.cc:1156:read_mapped_region()] Metadata of /cache/build/exclusive-amdci1-0/julialang/rr/obj/my_temp_parent_dir/jl_2NFyWr/rr-test-cont_race-CsXf88Rlk/cont_race-CsXf88Rlk...
FWIW the hang looks pretty disturbing. ``` #1 0x000055dc120c3f11 in rr::Task::wait (this=0x55dc127ce870, interrupt_after_elapsed=-1) at /cache/build/exclusive-amdci1-0/julialang/rr/src/Task.cc:1869 #2 0x000055dc120c26cc in rr::Task::resume_execution (this=0x55dc127ce870, how=rr::RESUME_SYSEMU, wait_how=rr::RESUME_WAIT, tick_period=rr::RESUME_UNLIMITED_TICKS, sig=0) at /cache/build/exclusive-amdci1-0/julialang/rr/src/Task.cc:1461 #3 0x000055dc1204de55 in rr::ReplaySession::continue_or_step...
e91772880ce27861252f3bdec016c71a1bbe2330 should handle `close()` syscalls.
> Unfortunately we don't have a backtrace for this kind of assertion. Ok, b08ae8b9d2843bef8db92d5ef570a6f159adcd00 should give us backtraces for failed `DEBUG_ASSERT`.
But `mapping_of` should really return an error value instead of just dying when the mapping doesn't exist.
9224099ad61eb6c9bd96c17fc9101c7254dfffa4 should fix that one.
Fedora packages VirtualGL. Hopefully Ubuntu does too. That helps. Trying `vglrun rr record glxspheres64`, rr dies on a DRM ioctl with this stack: ``` #0 0x00007f9260b6a359 in ioctl () at...
(But maybe I'm doing it wrong.)