On 10/17/2010 03:25 PM, Nadav Har'El wrote:
On Sun, Oct 17, 2010, Avi Kivity wrote about "Re: [PATCH 10/27] nVMX: Implement VMCLEAR": > >+ vmcs12 = kmap(page); > > kmap_atomic() please (better, kvm_write_guest(), but can defer that for > later) Sorry about my ignorance, but why is kmap_atomic() better here than kmap()? While handling an exit (caused by a guest running VMCLEAR instruction), we aren't in atomic context, aren't we?
kmap() is unloved since it is deadlock-prone in some circumstances, and also much slower than kmap_atomic(), since it needs global tlb synchronization.
If I use kmap_atomic() here I'll need to kunmap_atomic() below which will break the newly combined kunmap& nested_release_page function ;-) > >+ vmcs12->launch_state = 0; > >+ kunmap(page); > >+ nested_release_page(page);
Is something preventing you from changing all kmap()s to kmap_atomic()s (like guest memory access in the mapped section)?
-- error compiling committee.c: too many arguments to function -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
