[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [SOLVED] Re: ARAnyM VMs with Debian hanging at 100% CPU usage

schmitz dixit:

> great work on pinning this bug down - and full kudos to Andreas for
> the fix.

And for tracing down the kernel portion of it.
I merely had a guess where the problem was and
went looking for a testcase, which could then
be reduced.

>> suggested; if someone
>> here thinks that c663600584a596b5e66258cc10716fb781a5c2c9 also
>> should be added to stable, please do the same, I cannot judge
>> that.

> Since it does not appear to cause any real harm (presumaby nfeth still works
> after the message?), I'd leave it.

Yes, it works either way. Interrupts are some black magic, and I
must admit I don't really understand the impact of the issue even
after reading the verbose commit message…

> And I seem to recall we were not going to do anything like this, for
> performance reasons on both architectures.

Mhm, okay. But as things are now, we have a function call for things
like getting the TLS base and doing an atomic operation, which both
do a syscall, already, so I wonder whether it would get worse (more
things needing a function call indirection) or better (the function
call can be made indirect, then diverted to not need the syscall).

> I hear you - and I am still working on sanitizing the driver's
> interrupt use.

Oh, sounds like fun. Good luck! (ragnar76 is the one in need of
such drivers, he'd probably help testing, too.)

21:27⎜[Natureshadow] BÄH! Wer hatn das Bier neben den Notebooklüfter
     ⎜    gestellt ...
21:27⎜>Natureshadow< lol                         21:27⎜>Natureshadow< du?
21:27⎜[Natureshadow] vermutlich ...   -- Kev^WNatureshadow allein zu Haus

Reply to: