.@paulmckrcu Would it be possible to do "norcu" as "synchronize on every return to userspace and every cpu_idle entry"?
-
-
Replying to @RichFelker
If you don't mind long loops in the kernel stalling RCU and OOMing your system, yes... But what are you trying to achieve?
1 reply 0 retweets 0 likes -
Replying to @paulmckrcu
Unpredictable and possibly unbounded memory usage and performance characteristics and seemingly impossible-to-verify logic.
1 reply 0 retweets 0 likes -
Replying to @RichFelker @paulmckrcu
Or rather, trying to avoid those things. So frustrating debugging this stuff when simple rwlocks & no alloc would've sufficed.
2 replies 0 retweets 0 likes -
Replying to @RichFelker @paulmckrcu
What I'm trying to say is it'd be nice to have an option that doesn't depend on the scheduler to collect, does it sync. instead.
1 reply 0 retweets 0 likes -
Replying to @RichFelker
: But the real problem was not the scheduler, but rather timers, correct? Or am I behind the times in your troubleshooting?
1 reply 0 retweets 0 likes -
Replying to @paulmckrcu
I don't know, This thread is inspired by but in some sense orthogonal to current stall problem.
3 replies 0 retweets 0 likes -
Replying to @RichFelker
If you have timer problems, and you didn't have RCU, you would likely instead be fighting softlockups. Shoot the messenger?
1 reply 0 retweets 0 likes
But they'd be debuggable, I think. :-)
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.