Messages in this thread | | | From | Andi Kleen <> | Subject | Re: [PATCH v7 1/4] qrwlock: A queue read/write lock implementation | Date | Wed, 18 Dec 2013 12:16:11 -0800 |
| |
Peter Zijlstra <peterz@infradead.org> writes:
> On Wed, Dec 18, 2013 at 11:38:29AM -0800, Andi Kleen wrote: >> (maybe it would be time to get rid of the patchable LOCK though?) > > With the argument that Intel simply doesn't ship UP chips anymore, with > the exception of quark which should probably run custom UP kernels due > to size constraints anyway?
That, and:
- Anything with a single core only has very fast LOCK - LOCK generally became much faster everywhere. - The original reason was for single cpu VM guests, but even those should increasingly have at least two VCPUs.
-Andi
-- ak@linux.intel.com -- Speaking for myself only
| |