Messages in this thread | | | Date | Mon, 11 Apr 2011 19:19:57 +0200 | From | Oleg Nesterov <> | Subject | [RFC PATCH 0/6] signal: sigprocmask fixes |
| |
Hello,
I am not sure this is really wrong and should be fixed. Please comment.
1-2: probably make sense anyway, cleanup + optimization.
4: the main change. From the changelog,
In short, almost every changing of current->blocked is wrong, or at least can lead to the unexpected results.
5-6: If this should be fixed, then we need a lot more trivial changes like these ones.
What do you all think?
In theory this can explain the hang reported by Nikita, but this is not clear.
Oleg.
| |