- f2540fc Merge "Document the update process." into main by Treehugger Robot · 9 months ago main master
- 235dc6d Document the update process. by Dan Albert · 9 months ago
- bc11e47 Upgrade rr to eda23a1c9cf89d22b15d838fdf46042b51a3e4ca by Dan Albert · 9 months ago
- eda23a1 Close race condition in ppoll test where SIGALRM was delivered before we actually reached ppoll by Robert O'Callahan · 9 months ago
- e43acfb Fix CMakeList.txt for counters test renaming by Robert O'Callahan · 9 months ago
- 961524c Add README file explaining the purpose of the standalone counters test by Robert O'Callahan · 9 months ago
- 4951612 Convert standalone counters test to C++ by Robert O'Callahan · 9 months ago
- 96affb2 Extend release-testing script to support Aarch64 by Robert O'Callahan · 9 months ago
- 8c468fb Simplify task exit handling by allowing Task::registers to be the source of truth during an unexpected exit by Robert O'Callahan · 9 months ago
- 14fb404 Avoid post-syscall state processing if the task has been unexpectedly killed. by Robert O'Callahan · 9 months ago
- 425d369 Don't try to restore register values in `AutoRestoreMem::~AutoRestoreMem()` if the task is dying by Robert O'Callahan · 9 months ago
- 5817b78 Fix `vfork_done` test on Aarch64 by checking orig_arg1() for clone parameters by Robert O'Callahan · 9 months ago
- 18f05b2 Support CentOS 8 and 9 for release-testing by Robert O'Callahan · 9 months ago
- f36191f Support release testing of Debian 10/11/12 by Robert O'Callahan · 9 months ago
- 495b1b2 Refactor distro tests by Robert O'Callahan · 9 months ago
- 2da8145 Fix test to account for change in gdb behavior after Ubuntu 18 by Robert O'Callahan · 9 months ago
- 07ebc5e Ubuntu 18 doesn't have `gettid()` in libc by Robert O'Callahan · 9 months ago
- 024064f When `Task::did_waitpid` can't get registers for a stop because we prematurely exited the stop, just ignore the stop. by Robert O'Callahan · 10 months ago
- 7f4d35a Download Firefox release for testing instead of using the distro package, because on new Ubuntus Firefox is a Snap package which we don't support in rr yet by Robert O'Callahan · 9 months ago
- 4d75114 Update .gitignore by Simon Sobisch · 9 months ago
- 1b4371d Ensure that detached tasks are kicked out of their self-induced group-stop by Robert O'Callahan · 9 months ago
- 1afb318 Rework register comparison API so that all register-difference output is emitted in a single log line by Robert O'Callahan · 9 months ago
- 13b2274 Add a test for crashing after a hlt instruction. by Kyle Huey · 9 months ago
- db8d96d Update brotli version to 1.1.0 by Simon Sobisch · 9 months ago
- 460a4d8 Convert signals across architectures so `rr dump` on x86-64 can dump an Aarch64 trace (for example) by Robert O'Callahan · 9 months ago
- 068084d Don't create Firefox profile directory in preelease testing, let Firefox do it by Robert O'Callahan · 9 months ago
- 307f2d5 Add distro-configs for Ubuntu 18 LTS, 22 LTS and 23.04 by Robert O'Callahan · 9 months ago
- fe71588 Remove unused pipe by Robert O'Callahan · 9 months ago
- 0a47339 Upgrade rr to 5ead785cfbb4c46c75324ef3caa6067c70f3e3a6 by Dan Albert · 9 months ago
- 5ead785 Do not use -fno-integrated-as for Android. (#3592) by Dan Albert · 9 months ago
- 368c7a2 Make memory checksums work with syscallbuf ABORT_COMMITs by avoiding issues with `syscallbuf_record::ret` and `syscallbuf_hdr::notify_on_syscall_hook_exit` by Robert O'Callahan · 9 months ago
- ff9ac5d Buffer fstatat family of syscalls. by Kyle Huey · 10 months ago
- 115ae00 Checking seen_ptrace_event_exit() here doesn't seem to do anything. by Robert O'Callahan · 10 months ago
- 1d642be Reverse the order of the writable and map_shared checks. by Kyle Huey · 10 months ago
- b937887 Add a test for fstatat. by Kyle Huey · 10 months ago
- 28feb3b Rename `Task::is_dying()` to `seen_ptrace_exit_event()` for clarity by Robert O'Callahan · 10 months ago
- 7bcb9f6 Rename `Task::seen_ptrace_exit_event` to `seen_ptrace_exit_event_` so we can fix the getter name by Robert O'Callahan · 10 months ago
- e4df1e3 Simplify `Task::dup_from()` and maybe make it more robust by not unmapping and then duplicating the VDSO/VVAR mappings. by Robert O'Callahan · 10 months ago
- b416b51 move_vdso_and_vvar_mappings is still giving us trouble, so fix it harder. by Robert O'Callahan · 10 months ago
- 7764413 Try to fix cont_race mapping races by detecting when a map/unmap happened even though we couldn't get the syscall result due to unexpected task death by Robert O'Callahan · 10 months ago
- b8034ce Check the result of infallible_mmap_syscall_if_alive correctly. by Robert O'Callahan · 10 months ago
- 2c1a6d4 Handle unexpected reaping in AutoRemoteSyscalls by Robert O'Callahan · 10 months ago
- 67c00f8 Make RecordTask::waiting_for_pid_namespace_tasks_to_exit more robust by Robert O'Callahan · 10 months ago
- c467397 Make handle_ptrace_exit_event handle reaping statuses by emitting the correct task exit records if we haven't seen a PTRACE_EVENT_EXIT by Robert O'Callahan · 10 months ago
- 49d6a38 Rename Task::handled_ptrace_exit_event to handled_ptrace_exit_event_ so we can add a getter by Robert O'Callahan · 10 months ago
- e0b1339 Fix comment in handle_ptrace_exit_event by Robert O'Callahan · 10 months ago
- 53caadc Remove detected_unexpected_exit logic which should no longer be needed by Robert O'Callahan · 10 months ago
- c3c1011 Don't emit a fake PTRACE_EVENT_EXIT when we see an exit in WaitAggregator::try_wait_exit, emit the real exit status by Robert O'Callahan · 10 months ago
- f68f28c Simply code around account_for_potential_ptrace_interrupt_stop by Robert O'Callahan · 10 months ago
- ea58ced Rename Task::already_reaped() to was_reaped() for consistency by Robert O'Callahan · 10 months ago
- 270c8de Rename Task::was_reaped to was_reaped_ so we can fix the getter name by Robert O'Callahan · 10 months ago
- 7650cee Don't fake a PTRACE_EVENT_EXIT status after an unexpected reap. by Robert O'Callahan · 10 months ago
- c374348 Change Task::is_running() to is_stopped() to be consistent with the rest of the state description. by Robert O'Callahan · 10 months ago
- 7fd565c Rename Task::is_stopped to is_stopped_ so we can introduce an is_stopped() getter by Robert O'Callahan · 10 months ago
- a338ea3 Don't set Task:is_stopped after we've reaped the task, only set it for ptrace-stops by Robert O'Callahan · 10 months ago
- 2c691d1 Start echoing commands and exiting on error as early as possible for easier debugging by Robert O'Callahan · 10 months ago
- b5bf0a8 testsuite: add LC_ALL default to util.sh by Simon Sobisch · 10 months ago
- 3f96e4e Revert "Don't set Task:is_stopped after we've reaped the task, only set it for ptrace-stops" by Robert O'Callahan · 10 months ago
- 9b7b15b Add more cleanup/clarification to Task::ptrace_if_stopped callers. by Robert O'Callahan · 10 months ago
- ba5751e Add some documentation for usage of ptrace_if_stopped by Robert O'Callahan · 10 months ago
- 57d0a80 Call fallible_ptrace instead of ptrace_if_stopped in AutoRemoteSyscalls::ignore_signal because this is just for logging and we don't actually care if the tracee is not stopped by Robert O'Callahan · 10 months ago
- 464e4bf Make Task::ptrace_if_stopped check that we think we're actually in a ptrace-stop by Robert O'Callahan · 10 months ago
- c32bb0b Don't set Task:is_stopped after we've reaped the task, only set it for ptrace-stops by Robert O'Callahan · 10 months ago
- 7039534 Make Task::do_ptrace_interrupt not use ptrace_if_stopped, since PTRACE_INTERRUPT doesn't require the task to be stopped by Robert O'Callahan · 10 months ago
- 5096559 Make recorded PTRACE_INTERRUPT calls use fallible_ptrace since they actually can work on non-stopped tasks by Robert O'Callahan · 10 months ago
- eb1f7ea Fix some stragger calls to ptrace_if_alive by Robert O'Callahan · 10 months ago
- 04cae0a Rename ptrace_if_alive to ptrace_if_stopped for clarification by Robert O'Callahan · 10 months ago
- 6102b34 Remove Task::xptrace now that it is no longer used by Robert O'Callahan · 10 months ago
- f9a3a3e Make get_ptrace_eventmsg_pid fallible by Robert O'Callahan · 10 months ago
- 5907b2f Make get_ptrace_eventmsg a non-template method that always returns a pid_t, since that's how it's always used by Robert O'Callahan · 10 months ago
- 5edc100 Inline xptrace in ReplaySession::detach_tasks so we can get rid of xptrace later by Robert O'Callahan · 10 months ago
- 080d452 Hack test script some more by Robert O'Callahan · 10 months ago
- 1d01261 Fix rebase mistake by Keno Fischer · 10 months ago
- cc5cee0 Create a known-patchable sequence for rdtsc trapping by Keno Fischer · 10 months ago
- 55012fb Fix detection of interfering branches for PATCH_SYSCALL_INSTRUCTION_IS_LAST by Keno Fischer · 10 months ago
- 4af8558 Refactor `ntasks_running` and make sure it doesn't go out of bounds. by Robert O'Callahan · 10 months ago
- 5442ecf Remove waiting_for_zombie state, use waiting_for_reap instead by Robert O'Callahan · 10 months ago
- 28a949b Fix option names in test script to use - consistently by Robert O'Callahan · 10 months ago
- c9afd72 Handle Task::dup_from case where other's [vdso] overlaps the current task's [vvar]. by Robert O'Callahan · 10 months ago
- bc24929 Give gdb_qpasssignals extra time because it's slow by Robert O'Callahan · 10 months ago
- 5d475a5 Add extra logging to help diagnose intermittent failure by Robert O'Callahan · 10 months ago
- 9496b8c Don't let child process run through (invalid) nanosleep and execute parent-process code by Robert O'Callahan · 10 months ago
- 2033698 Fix test script by Robert O'Callahan · 10 months ago
- 6822adb Set executable mode on test-system script by Robert O'Callahan · 10 months ago
- dc029a9 More fixes for the system-test script by Robert O'Callahan · 10 months ago
- ca3554b Don't bother recording memory areas when the process has died; the changed data is unobservable anyway. by Robert O'Callahan · 10 months ago
- 5d24e9b Fix rr-testing.sh pathnames by Robert O'Callahan · 10 months ago
- 79752d0 Dump process maps when an mremap or mmap remote syscall fails by Robert O'Callahan · 10 months ago
- 9176e25 If we fail to allocate a syscallbuf in a tracee (probably because it died), don't leave an entry for it in the AddressSpace's cached VMA list. by Robert O'Callahan · 10 months ago
- 53a1a75 Dump siginfo and signal to be delivered when we get an unexpected signal while trying to enter a user signal handler by Robert O'Callahan · 10 months ago
- ceb05ee Skip setuid test if systemd doesn't support disabling RDRAND by Robert O'Callahan · 10 months ago
- 795ba07 Bump the DH key size to 256 bytes to avoid EINVAL on my kernel. by Kyle Huey · 10 months ago
- 1850755 PR_SET_VMA has been empirically observed to return ENOMEM on kernels without CONFIG_ANON_VMA_NAME, by Kyle Huey · 10 months ago
- fcc3b95 Create AWS-based infrastructure for running rr tests on various Linux distributions/architectures/versions by Robert O'Callahan · 10 months ago
- ea26db5 Fix `unicode` test on distro/gdb combinations where breaking on a function doesn't display the function parameters correctly. by Robert O'Callahan · 10 months ago
- 73f3156 fix build error by Vincent Hilla · 11 months ago
- 1f75c4f Address review comments by Vincent Hilla · 11 months ago
- 262fd7d remove unused stdc++fs by Vincent Hilla · 11 months ago
- 041c71a reformat by Vincent Hilla · 11 months ago
- 628a283 rm and mv commands to remove and move traces by Vincent Hilla · 11 months ago