Re: Question on dirty sync before kvm memslot removal

2020-04-02 Thread Peter Xu
On Thu, Apr 02, 2020 at 04:47:58PM -0400, Peter Xu wrote: > On Wed, Apr 01, 2020 at 07:09:28PM -0400, Peter Xu wrote: > > On Wed, Apr 01, 2020 at 01:12:04AM +0200, Paolo Bonzini wrote: > > > On 31/03/20 18:51, Peter Xu wrote: > > > > On Tue, Mar 31, 2020 at 05:34:43PM +0200, Paolo Bonzini wrote: >

Re: Question on dirty sync before kvm memslot removal

2020-04-02 Thread Peter Xu
On Wed, Apr 01, 2020 at 07:09:28PM -0400, Peter Xu wrote: > On Wed, Apr 01, 2020 at 01:12:04AM +0200, Paolo Bonzini wrote: > > On 31/03/20 18:51, Peter Xu wrote: > > > On Tue, Mar 31, 2020 at 05:34:43PM +0200, Paolo Bonzini wrote: > > >> On 31/03/20 17:23, Peter Xu wrote: > > Or KVM_MEM_READON

Re: Question on dirty sync before kvm memslot removal

2020-04-01 Thread Peter Xu
On Wed, Apr 01, 2020 at 01:12:04AM +0200, Paolo Bonzini wrote: > On 31/03/20 18:51, Peter Xu wrote: > > On Tue, Mar 31, 2020 at 05:34:43PM +0200, Paolo Bonzini wrote: > >> On 31/03/20 17:23, Peter Xu wrote: > Or KVM_MEM_READONLY. > >>> Yeah, I used a new flag because I thought READONLY was a b

Re: Question on dirty sync before kvm memslot removal

2020-03-31 Thread Paolo Bonzini
On 31/03/20 18:51, Peter Xu wrote: > On Tue, Mar 31, 2020 at 05:34:43PM +0200, Paolo Bonzini wrote: >> On 31/03/20 17:23, Peter Xu wrote: Or KVM_MEM_READONLY. >>> Yeah, I used a new flag because I thought READONLY was a bit tricky to >>> be used directly here. The thing is IIUC if guest write

Re: Question on dirty sync before kvm memslot removal

2020-03-31 Thread Peter Xu
On Tue, Mar 31, 2020 at 05:34:43PM +0200, Paolo Bonzini wrote: > On 31/03/20 17:23, Peter Xu wrote: > >> Or KVM_MEM_READONLY. > > Yeah, I used a new flag because I thought READONLY was a bit tricky to > > be used directly here. The thing is IIUC if guest writes to a > > READONLY slot then KVM shou

Re: Question on dirty sync before kvm memslot removal

2020-03-31 Thread Paolo Bonzini
On 31/03/20 17:23, Peter Xu wrote: >> Or KVM_MEM_READONLY. > Yeah, I used a new flag because I thought READONLY was a bit tricky to > be used directly here. The thing is IIUC if guest writes to a > READONLY slot then KVM should either ignore the write or trigger an > error which I didn't check, ho

Re: Question on dirty sync before kvm memslot removal

2020-03-31 Thread Peter Xu
On Mon, Mar 30, 2020 at 03:11:53PM +0200, Paolo Bonzini wrote: > On 27/03/20 16:04, Peter Xu wrote: > > That makes perfect sense to me, however... What if the vcpu generates > > dirty bits _after_ we do KVM_GET_DIRTY_LOG but _before_ we send > > KVM_SET_USER_MEMORY_REGION to remove the memslot? If

Re: Question on dirty sync before kvm memslot removal

2020-03-30 Thread Paolo Bonzini
On 27/03/20 16:04, Peter Xu wrote: > That makes perfect sense to me, however... What if the vcpu generates > dirty bits _after_ we do KVM_GET_DIRTY_LOG but _before_ we send > KVM_SET_USER_MEMORY_REGION to remove the memslot? If the vcpu is in > the userspace I think it's fine because BQL is needed