[PATCH 0/9] migration: Better error handling in rp thread, allow failures in recover

[Peter Xu <peterx@redhat.com>]

This patchset supersedes below:
[PATCH v2 0/7] migration: Better error handling in return path thread

Another note is that this might conflict with Fabiano's other patchset to
fix postcopy race conditions, but maybe not.  If collapse, I can rebase.
Let me send this out still for early reviews.

I dropped the last patch there (which wasn't clear on being beneficial)
from last version, meanwhile added three more patches to address an issue
reported from our QE team that one postcopy migration can stuck in RECOVER
stage and never got kicked out.  For more information of that problem, one
can refer to the last patch commit message.

Since this one covers more issues, I renamed the subject, and let me
version it from v1.  I still collected most of R-bs from Fabiano since last
version (patches 1-6).

Please have a look, thanks.

Peter Xu (9):
  migration: Display error in query-migrate irrelevant of status
  migration: Let migrate_set_error() take ownership
  migration: Introduce migrate_has_error()
  migration: Refactor error handling in source return path
  migration: Deliver return path file error to migrate state too
  qemufile: Always return a verbose error
  migration: Remember num of ramblocks to sync during recovery
  migration: Add migration_rp_wait|kick()
  migration/postcopy: Allow network to fail even during recovery

 qapi/migration.json      |   5 +-
 migration/migration.h    |  25 ++++-
 migration/qemu-file.h    |   1 +
 migration/ram.h          |   5 +-
 migration/channel.c      |   1 -
 migration/migration.c    | 230 +++++++++++++++++++++++++++------------
 migration/multifd.c      |  10 +-
 migration/postcopy-ram.c |   1 -
 migration/qemu-file.c    |  17 ++-
 migration/ram.c          |  77 +++++++------
 migration/trace-events   |   2 +-
 11 files changed, 248 insertions(+), 126 deletions(-)

-- 
2.41.0