[riot-notifications] [RIOT-OS/RIOT] cpu/cortexm_common: replace irq_restore by __set_PRIMASK for stm32l152re (#11919)

Gaëtan Harter notifications at github.com
Mon Jul 29 17:38:44 CEST 2019


Question, I do not see the `nop` in the "original disassembly", is that normal? Or was it the code to show the crash without it?

> As for all I can tell, and as stated in #8518, the issue was the loss of the content of r0 when jumping to irq_restore which resulted in a corrupted pc trying to execute instructions out of the flash region.

I think it was "loss of the content of r0" that, had the consequence to crash when jumping. Maybe I am wrong and missing the right pointer.
I say this as it seems to be written as "loss of `r0` because of jumping".

If my previous statement is correct and that `r0` is still lost, I think it would mean that the restored value given to `PRIMASK` may be wrong. or am I missing something?
I am not sure how we managed to show the wrong `r0` value, was it with `gdb` @kYc0o ? the other PR says we do not see it when 'breaking' at some instructions.



-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/RIOT-OS/RIOT/pull/11919#issuecomment-516044393
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.riot-os.org/pipermail/notifications/attachments/20190729/c139a660/attachment.htm>


More information about the notifications mailing list