Re: [Qemu-devel] [PATCH] migration: fix incorrect enable return path

[Peter Xu]

On Mon, Jun 19, 2017 at 01:12:36PM +0200, Laurent Vivier wrote:
> On 15/06/2017 18:54, Laurent Vivier wrote:
> > On 14/06/2017 09:55, Peter Xu wrote:
> >> 0425dc9 is actually v1 of that patch, but it was accidentally
> >> merged (while there was a v2). That will cause problem when we try to
> >> migrate to some old QEMUs when return path is not really there. Let's
> >> fix it, then squashing this patch with 0425dc9 will be exactly patch
> >> content of v2.
> >>
> >> Fixes: 0425dc9 ("migration: isolate return path on src")
> >> Signed-off-by: Peter Xu <address@hidden>
> >> ---
> >>  migration/migration.c | 6 +++++-
> >>  1 file changed, 5 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/migration/migration.c b/migration/migration.c
> >> index 8ef6d6c..f9053b0 100644
> >> --- a/migration/migration.c
> >> +++ b/migration/migration.c
> >> @@ -1814,7 +1814,11 @@ static void *migration_thread(void *opaque)
> >>  
> >>      qemu_savevm_state_header(s->to_dst_file);
> >>  
> >> -    if (s->to_dst_file) {
> >> +    /*
> >> +     * If we opened the return path, we need to make sure dst has it
> >> +     * opened as well.
> >> +     */
> >> +    if (s->rp_state.from_dst_file) {
> >>          /* Now tell the dest that it should open its end so it can reply 
> >> */
> >>          qemu_savevm_send_open_return_path(s->to_dst_file);
> >>  
> >>
> > 
> > Tested-by: Laurent Vivier <address@hidden>
> > 
> > This also fixes migration of pseries guest, using the same
> > binary/machine type on both side, the error was:
> > 
> > qemu-system-ppc64: Failed to load spapr_iommu:mig_table
> > qemu-system-ppc64: error while loading state for instance 0x80000001 of
> > device 'spapr_iommu'
> > qemu-system-ppc64: load of migration failed: Input/output error
> 
> The error is an IEO because the file descriptor is closed on source. It
> seems the source closes the stream too early without this patch.
> For test purpose, I have disabled the qemu_fclose() on the source, and
> the error disappears from the destination.

Thanks for confirming this, Laurent!

-- 
Peter Xu