qemu-devel
[Top][All Lists]
Advanced

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

Re: [Qemu-devel] [RFC PATCH v2] PPC: smp: autodetect numbers of threads


From: Mike Day
Subject: Re: [Qemu-devel] [RFC PATCH v2] PPC: smp: autodetect numbers of threads per core
Date: Fri, 10 Jan 2014 09:12:09 -0500
User-agent: Notmuch/0.16 (http://notmuchmail.org) Emacs/24.3.1 (x86_64-redhat-linux-gnu)

Alexander Graf <address@hidden> writes:

>> The patch as it its now is very simple and well-contained. I wonder how
>> much it would expand if we added a max thread count to the cpu class. It
>> seems like the need for a max thread count is idiomatic to powerpc. 
>
> It's only ever useful on IBM POWER. Any other PowerPC system can partition 
> vcpus by host threads, it's only IBM POWER hardware that's as broken as it is.


> I do see that the user experience is slightly suboptimal, but by creating 
> this special case there's a good chance you're doing more harm than good.

All the problems you raise about special cases are true. But, as you
point out, ibm power is uniquely broken, which would possibly justify a
special case, unless there is a better general solution. In other words,
special cases exist for unique circumstances and if I understand you
correctly this is a unique circumstance.

Alexy explained the use case in his initial posting. The user needs to
allocate all threads of a core to an instance of KVM, but has no easy
way to obtain that information (threads per core) for the Qemu threads
option. So specifying threads="max" is a more user-friendly option. In
my understanding this is strictly a usability issue.

Mike
-- 
Mike Day | "Endurance is a Virtue"



reply via email to

[Prev in Thread] Current Thread [Next in Thread]