On Tue, 16 Mar 2021 21:48:38 +0530 "Tarun Gupta (SW-GPU)" <[email protected]> wrote:
> On 3/15/2021 10:52 PM, Cornelia Huck wrote: > > > > On Thu, 11 Mar 2021 00:50:09 +0530 > > Tarun Gupta <[email protected]> wrote: > > > >> Document interfaces used for VFIO device migration. Added flow of state > >> changes > >> during live migration with VFIO device. Tested by building docs with the > >> new > >> vfio-migration.rst file. > >> > >> v2: > >> - Included the new vfio-migration.rst file in index.rst > >> - Updated dirty page tracking section, also added details about > >> 'pre-copy-dirty-page-tracking' opt-out option. > >> - Incorporated comments around wording of doc. > >> > >> Signed-off-by: Tarun Gupta <[email protected]> > >> Signed-off-by: Kirti Wankhede <[email protected]> > >> --- > >> MAINTAINERS | 1 + > >> docs/devel/index.rst | 1 + > >> docs/devel/vfio-migration.rst | 135 ++++++++++++++++++++++++++++++++++ > >> 3 files changed, 137 insertions(+) > >> create mode 100644 docs/devel/vfio-migration.rst > >> > > > > (...) > > > >> diff --git a/docs/devel/vfio-migration.rst b/docs/devel/vfio-migration.rst > >> new file mode 100644 > >> index 0000000000..6196fb132c > >> --- /dev/null > >> +++ b/docs/devel/vfio-migration.rst > >> @@ -0,0 +1,135 @@ > >> +===================== > >> +VFIO device Migration > >> +===================== > > > > Maybe add an introductory sentence or two describing the general > > approach? I.e. we have a general framework, and specific support for > > devices needs to be hooked up. > > Ummm, the below paragraph does describe the approach we're using for the > migration framework involving pre-copy and stop-and-copy phase. > Can you help elaborate more on the general approach you'd like to have? The document dives right in with how vfio devices are using an iterative approach etc. A quick overview of the general setup (before you are getting to the different phases) might be helpful, i.e. who does what. Not sure what we expect a reader of this document to know already. > > > > >> + > >> +VFIO devices use an iterative approach for migration because certain VFIO > >> +devices (e.g. GPU) have large amount of data to be transfered. The > >> iterative > >> +pre-copy phase of migration allows for the guest to continue whilst the > >> VFIO > >> +device state is transferred to the destination, this helps to reduce the > >> total > >> +downtime of the VM. VFIO devices can choose to skip the pre-copy phase of > >> +migration by returning pending_bytes as zero during the pre-copy phase. > > > > What about something like: > > > > "Migration of VFIO devices consists of two phases: the optional > > pre-copy phase, and the stop-and-copy phase. The pre-copy phase is > > iterative and allows to accommodate VFIO devices that have a large > > amount of data that needs to be transferred. The iterative pre-copy > > phase..." > > > > Thanks, this looks better. I'll update it in next version incorporating > the other comments too below.
