At 10:56 +0000 on 09 Jan (1420797418), Andrew Cooper wrote: > On 09/01/15 10:27, Jan Beulich wrote: > > While the REP MOVS acceleration appears to have helped qemu-traditional > > based guests, qemu-upstream (or really the respective video BIOSes) > > doesn't appear to benefit from that. Instead the acceleration added > > here provides a visible performance improvement during very early HVM > > guest boot. > > > > Signed-off-by: Jan Beulich <jbeul...@suse.com> > > --- > > v2: fix asm() constraints in hvmemul_rep_stos(), as pointed out by > > Andrew. Add output operand telling the compiler that "buf" is being > > written. > > Is writing buf wise? it looks like you will xfree() a wild pointer, and > appears to interfere the "buf != p_data" logic.
I think the constraints are correct, though the 'memory' clobber makes the rest of it moot. But this: > > + default: > > + xfree(buf); > > + ASSERT(!buf); looks dodgy... Tim. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org http://lists.xen.org/xen-devel
