cortex-m: Hard fault after a thread exits (under some circumstances) #4737
Description
... at least on these two Cortex-M4 platforms (tests/xtimer_msg_receive_timeout).
The test application is expected to just stop/hang after the final Timeout! message.
In the test's main() there's a loop that runs for 10 iterations, then a return 0.
It seems like the thread, after returning is trying to jump to 0x00000000, which is bad...
stm32f4discovery:
2016-02-04 11:12:56,827 - INFO # �main(): This is RIOT! (Version: 2016.03-devel-298-gca8db-mazarin)
2016-02-04 11:12:57,828 - INFO # Message received: 44
2016-02-04 11:12:58,829 - INFO # Timeout!
2016-02-04 11:12:59,829 - INFO # Message received: 44
2016-02-04 11:13:00,830 - INFO # Timeout!
2016-02-04 11:13:01,832 - INFO # Message received: 44
2016-02-04 11:13:02,831 - INFO # Timeout!
2016-02-04 11:13:03,832 - INFO # Message received: 44
2016-02-04 11:13:04,832 - INFO # Timeout!
2016-02-04 11:13:05,833 - INFO # Message received: 44
2016-02-04 11:13:06,833 - INFO # Timeout!
2016-02-04 11:13:06,834 - INFO #
2016-02-04 11:13:06,836 - INFO # Context before hardfault:
2016-02-04 11:13:06,838 - INFO # r0: 0x2000015c
2016-02-04 11:13:06,839 - INFO # r1: 0x40000000
2016-02-04 11:13:06,841 - INFO # r2: 0x00000000
2016-02-04 11:13:06,842 - INFO # r3: 0x2000084c
2016-02-04 11:13:06,844 - INFO # r12: 0x00000000
2016-02-04 11:13:06,845 - INFO # lr: 0x08000a81
2016-02-04 11:13:06,847 - INFO # pc: 0x00000000
2016-02-04 11:13:06,848 - INFO # psr: 0x6000002c
2016-02-04 11:13:06,852 - INFO #
2016-02-04 11:13:06,853 - INFO # FSR/FAR:
2016-02-04 11:13:06,855 - INFO # CFSR: 0x00020000
2016-02-04 11:13:06,856 - INFO # HFSR: 0x40000000
2016-02-04 11:13:06,861 - INFO # DFSR: 0x00000008
2016-02-04 11:13:06,862 - INFO # AFSR: 0x00000000
2016-02-04 11:13:06,863 - INFO # Misc
2016-02-04 11:13:06,865 - INFO # EXC_RET: 0xfffffff1
2016-02-04 11:13:06,868 - INFO # Attempting to reconstruct state for debugging...
2016-02-04 11:13:06,869 - INFO # In GDB:
2016-02-04 11:13:06,870 - INFO # set $pc=0x0
2016-02-04 11:13:06,871 - INFO # frame 0
2016-02-04 11:13:06,872 - INFO # bt
2016-02-04 11:13:06,872 - INFO #
2016-02-04 11:13:06,878 - INFO # ISR stack overflowed by at least 72 bytes.
mulle:
2016-02-04 11:09:23,462 - INFO # main(): This is RIOT! (Version: 2016.03-devel-298-gca8db-mazarin)
2016-02-04 11:09:24,442 - INFO # Message received: 44
2016-02-04 11:09:25,421 - INFO # Timeout!
2016-02-04 11:09:26,400 - INFO # Message received: 44
2016-02-04 11:09:27,378 - INFO # Timeout!
2016-02-04 11:09:28,358 - INFO # Message received: 44
2016-02-04 11:09:29,337 - INFO # Timeout!
2016-02-04 11:09:30,315 - INFO # Message received: 44
2016-02-04 11:09:31,295 - INFO # Timeout!
2016-02-04 11:09:32,274 - INFO # Message received: 44
2016-02-04 11:09:33,253 - INFO # Timeout!
2016-02-04 11:09:33,254 - INFO #
2016-02-04 11:09:33,255 - INFO # Context before hardfault:
2016-02-04 11:09:33,257 - INFO # r0: 0x1fff0184
2016-02-04 11:09:33,259 - INFO # r1: 0x1fff095c
2016-02-04 11:09:33,260 - INFO # r2: 0x00000000
2016-02-04 11:09:33,262 - INFO # r3: 0x1fff0884
2016-02-04 11:09:33,263 - INFO # r12: 0x00000000
2016-02-04 11:09:33,265 - INFO # lr: 0x00000c0d
2016-02-04 11:09:33,266 - INFO # pc: 0x00000000
2016-02-04 11:09:33,268 - INFO # psr: 0x60000055
2016-02-04 11:09:33,268 - INFO #
2016-02-04 11:09:33,269 - INFO # FSR/FAR:
2016-02-04 11:09:33,270 - INFO # CFSR: 0x00020000
2016-02-04 11:09:33,272 - INFO # HFSR: 0x40000000
2016-02-04 11:09:33,273 - INFO # DFSR: 0x00000000
2016-02-04 11:09:33,275 - INFO # AFSR: 0x00000000
2016-02-04 11:09:33,276 - INFO # Misc
2016-02-04 11:09:33,277 - INFO # EXC_RET: 0xfffffff1
2016-02-04 11:09:33,281 - INFO # Attempting to reconstruct state for debugging...
2016-02-04 11:09:33,282 - INFO # In GDB:
2016-02-04 11:09:33,283 - INFO # set $pc=0x0
2016-02-04 11:09:33,284 - INFO # frame 0
2016-02-04 11:09:33,285 - INFO # bt
2016-02-04 11:09:33,285 - INFO #
2016-02-04 11:09:33,289 - INFO # ISR stack overflowed by at least 72 bytes.
Metadata
Metadata
Assignees
Labels
This PR is a candidate for review and discussion during one of RIOT's monthly Hack'n'ACK partiesThe PR changes a significant part of the code base. It should be reviewed carefullyPlatform: This PR/issue effects ARM-based platformsThe issue reports a bug / The PR fixes a bug (including spelling errors)