Re: Proposal: use disk sequence numbers to avoid races in blkback

2022-05-11 Thread Demi Marie Obenour
On Wed, May 11, 2022 at 09:37:54AM +0200, Roger Pau Monné wrote: > On Tue, May 10, 2022 at 12:22:51PM -0400, Demi Marie Obenour wrote: > > On Tue, May 10, 2022 at 12:57:48PM +0200, Roger Pau Monné wrote: > > > On Thu, May 05, 2022 at 08:30:17PM -0400, Demi Marie Obenour wrote: > > > > Proposal: Che

Re: Proposal: use disk sequence numbers to avoid races in blkback

2022-05-11 Thread Roger Pau Monné
On Tue, May 10, 2022 at 12:22:51PM -0400, Demi Marie Obenour wrote: > On Tue, May 10, 2022 at 12:57:48PM +0200, Roger Pau Monné wrote: > > On Thu, May 05, 2022 at 08:30:17PM -0400, Demi Marie Obenour wrote: > > > Proposal: Check disk sequence numbers in blkback > > > ===

Re: Proposal: use disk sequence numbers to avoid races in blkback

2022-05-10 Thread Demi Marie Obenour
On Tue, May 10, 2022 at 12:57:48PM +0200, Roger Pau Monné wrote: > On Thu, May 05, 2022 at 08:30:17PM -0400, Demi Marie Obenour wrote: > > Proposal: Check disk sequence numbers in blkback > > > > > > Currently, adding block devices to a domain is ra

Re: Proposal: use disk sequence numbers to avoid races in blkback

2022-05-10 Thread Roger Pau Monné
On Thu, May 05, 2022 at 08:30:17PM -0400, Demi Marie Obenour wrote: > Proposal: Check disk sequence numbers in blkback > > > Currently, adding block devices to a domain is racy. libxl writes the > major and minor number of the device to XenStore, b

Proposal: use disk sequence numbers to avoid races in blkback

2022-05-05 Thread Demi Marie Obenour
Proposal: Check disk sequence numbers in blkback Currently, adding block devices to a domain is racy. libxl writes the major and minor number of the device to XenStore, but it does not keep the block device open until blkback has opened it. This c