[1/4] migration/multifd: Avoid the final FLUSH in complete()

Commit Message

Peter Xu June 12, 2024, 2:42 p.m. UTC

We always do the flush when finishing one round of scan, and during
complete() phase we should scan one more round making sure no dirty page
existed.  In that case we shouldn't need one explicit FLUSH at the end of
complete(), as when reaching there all pages should have been flushed.

Signed-off-by: Peter Xu <[e-mail]>
---
 migration/ram.c | 4 ----
 1 file changed, 4 deletions(-)

Comments

Fabiano Rosas June 13, 2024, 1:50 p.m. UTC | #1

Peter Xu <[e-mail]> writes:

> We always do the flush when finishing one round of scan, and during
> complete() phase we should scan one more round making sure no dirty page
> existed.  In that case we shouldn't need one explicit FLUSH at the end of
> complete(), as when reaching there all pages should have been flushed.
>
> Signed-off-by: Peter Xu <[e-mail]>

Reviewed-by: Fabiano Rosas <[e-mail]>

Fabiano Rosas June 13, 2024, 2:54 p.m. UTC | #2

Fabiano Rosas <[e-mail]> writes:

> Peter Xu <[e-mail]> writes:
>
>> We always do the flush when finishing one round of scan, and during
>> complete() phase we should scan one more round making sure no dirty page
>> existed.  In that case we shouldn't need one explicit FLUSH at the end of
>> complete(), as when reaching there all pages should have been flushed.
>>
>> Signed-off-by: Peter Xu <[e-mail]>
>
> Reviewed-by: Fabiano Rosas <[e-mail]>

Actually, let's be more clear and make this a:

Tested-by: Fabiano Rosas <[e-mail]>

That way I'll remember this went through the same tests as the other
multifd sync changes we made.

Peter Xu June 13, 2024, 3:53 p.m. UTC | #3

On Thu, Jun 13, 2024 at 11:54:58AM -0300, Fabiano Rosas wrote:
> Fabiano Rosas <[e-mail]> writes:
> 
> > Peter Xu <[e-mail]> writes:
> >
> >> We always do the flush when finishing one round of scan, and during
> >> complete() phase we should scan one more round making sure no dirty page
> >> existed.  In that case we shouldn't need one explicit FLUSH at the end of
> >> complete(), as when reaching there all pages should have been flushed.
> >>
> >> Signed-off-by: Peter Xu <[e-mail]>
> >
> > Reviewed-by: Fabiano Rosas <[e-mail]>
> 
> Actually, let's be more clear and make this a:
> 
> Tested-by: Fabiano Rosas <[e-mail]>
> 
> That way I'll remember this went through the same tests as the other
> multifd sync changes we made.

Or... could I take both? :)  And thanks for checking that.

Patch

diff --git a/migration/ram.c b/migration/ram.c
index ceea586b06..edec1a2d07 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -3300,10 +3300,6 @@  static int ram_save_complete(QEMUFile *f, void *opaque)
         }
     }
 
-    if (migrate_multifd() && !migrate_multifd_flush_after_each_section() &&
-        !migrate_mapped_ram()) {
-        qemu_put_be64(f, RAM_SAVE_FLAG_MULTIFD_FLUSH);
-    }
     qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
     return qemu_fflush(f);
 }