Classical OSs usually have at least two return paths (syscall vs rfi), and both should be checking for pending signals. Bug?
Brilliant: raising a signal in trap handler, then returning to userspace via a code path that doesn't check pending signals.
-
-
-
Yes. It's arch/sh, entry-common.S. No idea how the code was supposed to work; debug traps branch directly to __restore_all. Fix was easy though.
-
Yeah, I'm familiar with this on hppa, having written the kernel atomic emulation. I actually had to go quickly and look at entry.S to see if we did signal checking at each path and we do ;-)
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.