Commit Graph

9 Commits

Author SHA1 Message Date
Josh Gao 6462bb41e0 debuggerd_handler: add and use fatal_errno.
Bug: none
Test: mma
Change-Id: I24d913abdbe74f9463feda78f7817ca8b92af9cc
2017-01-31 14:59:05 -08:00
Josh Gao 4ed00c8d73 debuggerd_handler: improve nonfatal signal message.
"Fatal signal 35 (???)" -> "Requested dump for"

Bug: http://b/34809044
Test: debuggerd -b $$
Change-Id: I9ece0ee1117203d30142b843973ed7e5435e21da
2017-01-30 17:58:04 -08:00
Josh Gao e5288f292a debuggerd_handler: remove PR_SET_DUMPABLE check.
crash_dump has CAP_SYS_PTRACE and this was never obeyed by debuggerd.

Change-Id: Ifee5e94b97b1f6440ad0be79758f0db2d2aaba2e
2017-01-26 15:08:18 -08:00
Josh Gao 7e14d020f1 debuggerd_handler: don't dump PR_NO_NEW_PRIVS processes.
We can't do an selinux transition when this is on.

Bug: http://b/34472671
Test: logcat -c; debuggerd `pidof media.codec`; logcat
Change-Id: Ie6c1832ab838df48879c32a86126862de9a15420
2017-01-25 11:16:03 -08:00
Josh Gao 529b3066d5 debuggerd_handler: don't resend nonfatal signals when not dumping.
Bug: http://b/34516140
Test: debuggerd -b `pidof surfaceflinger`
Change-Id: I0275ffca24bf4840e264eaa4b79611e2404edfb0
2017-01-25 11:15:01 -08:00
Josh Gao fca7ca3585 debuggerd_handler: properly crash when PR_GET_DUMPABLE is 0.
Actually exit when receiving a signal via kill(2) or raise(2) and
PR_GET_DUMPABLE is 0.

Bug: none
Test: /data/nativetest/debuggerd_test/debuggerd_test32
Test: /data/nativetest64/bionic-unit-tests/bionic-unit-tests --gtest_filter=pthread_DeathTest.pthread_mutex_lock_null_64
Change-Id: I833a2a34238129237bd9f953959ebda51d8d04d7
2017-01-23 14:13:36 -08:00
Josh Gao 575941115e crash_dump: clear the default crash handlers.
crash_dump is a dynamic executable that gets the default crash dumping
handlers set by the linker. Turn them off to prevent crash_dump from
dumping itself.

Bug: http://b/34472671
Test: inserted an abort into crash_dump
Change-Id: Ic9d708805ad47afbb2a9ff37e2ca059f23f421de
2017-01-23 11:34:49 -08:00
Josh Gao b64dd85c94 debuggerd_handler: actually wait for pseudothread to exit.
Occasionally, the pseudothread wouldn't exit in time after unlocking
the mutex to get crash_dump to proceed, resulting in spurious error
messages. Instead of using a mutex to emulate pthread_join, just
implement it correctly.

Bug: http://b/34472671
Test: debuggerd_test
Change-Id: I5c2658a84e9407ed8cc0ef2ad0fb648c388b7ad1
2017-01-23 11:34:49 -08:00
Josh Gao cbe70cb0a8 debuggerd: advance our amazing bet.
Remove debuggerd in favor of a helper process that gets execed by
crashing processes.

Bug: http://b/30705528
Test: debuggerd_test
Change-Id: I9906c69473989cbf7fe5ea6cccf9a9c563d75906
2017-01-17 13:57:57 -08:00