On Mon, Aug 06, 2018 at 07:06:05PM +0300, Michael S. Tsirkin wrote:> > I've done something very similar in the thread I posted a few years > > ago. > > Right so that was before spectre where a virtual call was cheaper :(Sorry, I meant days, not years. The whole point of the thread was the slowdowns due to retpolines, which are the software spectre mitigation.
On Mon, Aug 06, 2018 at 09:10:40AM -0700, Christoph Hellwig wrote:> On Mon, Aug 06, 2018 at 07:06:05PM +0300, Michael S. Tsirkin wrote: > > > I've done something very similar in the thread I posted a few years > > > ago. > > > > Right so that was before spectre where a virtual call was cheaper :( > > Sorry, I meant days, not years. The whole point of the thread was the > slowdowns due to retpolines, which are the software spectre mitigation.Oh that makes sense then. Could you post a pointer pls so this patchset is rebased on top (there are things to change about 4/4 but 1-3 could go in if they don't add overhead)? -- MST
On Mon, Aug 06, 2018 at 07:13:32PM +0300, Michael S. Tsirkin wrote:> Oh that makes sense then. Could you post a pointer pls so > this patchset is rebased on top (there are things to > change about 4/4 but 1-3 could go in if they don't add > overhead)?The dma mapping direct calls will need a major work vs what I posted. I plan to start that work in about two weeks once returning from my vacation.