On May 18, 2009, at 6:43 PM, Timo Kreuzer wrote:
Aleksey Bragin wrote:
Exactly what I was telling yesterday in IRC.
I can only remember that you asked, if it would really fix something. ;) And yes it does. Just not the fully correct way.
Ah yes, my bad!
This assert is caused by recursive user call, not vice versa!!
True. But the recursive usercall is valid. Please take a look at the article I linked. It should simply end at some point when no more kernel stack is available.
Yes, I mixed those two different bugs earlier too, hence the confusion, sorry.
Revert?
What about the updated code I send?
For me it looks correct, maybe Alex has comments.
WBR, Aleksey Bragin.