Skip to content

Multifd zero page support

Juan Quintela qemu-devel 2 versions – In review

Get series with b4
b4 am [email protected]

Versions newest first · linked by author and title

  1. v2 Latest11 patches 5 repliesIn review
  2. v111 patches 0 repliesSuperseded

Patches in v2 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/11migration: Update atomic stats out of the mutexmigration/multifd.c+2 -20 replies
2/11migration: Make multifd_bytes atomicmigration/migration.c migration/multifd.c migration/ram.h+5 -40 replies
3/11multifd: We already account for this packet on the multifd threadmigration/multifd.c+0 -30 replies
4/11multifd: Count the number of bytes sent correctlymigration/multifd.c migration/multifd.h+6 -22 replies
5/11migration: Make ram_save_target_page() a pointermigration/ram.c+15 -40 replies
6/11multifd: Make flags field thread localmigration/multifd.c migration/multifd.h+19 -140 replies
7/11multifd: Prepare to send a packet without the mutex heldmigration/multifd.c migration/multifd.h+8 -60 replies
8/11multifd: Add capability to enable/disable zero_pagehw/core/machine.c migration/migration.c migration/migration.h +1 more+21 -23 replies
9/11multifd: Support for zero pages transmissionmigration/multifd.c migration/multifd.h migration/ram.c +1 more+49 -140 replies
10/11multifd: Zero pages transmissionmigration/multifd.c migration/multifd.h+48 -20 replies
11/11So we use multifd to transmit zero pages.migration/migration.c migration/ram.c+32 -30 replies

Reviews trailers per patch

1 of 11 patches has a Reviewed-by or Acked-by.

PatchDr. David Alan Gilbert
1/11
2/11
3/11
4/11
5/11R-b
6/11
7/11
8/11
9/11
10/11
11/11

Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes.

Read the v2 thread

Cover letter [PATCH v2 00/11] Multifd zero page support

Based on top of my next branch.
- Rebased on top of latest upstream
- Redo a lot of the packet accounting
still not completely perfect, but much better than what is upstream

Still working continuing on that.

Please review.

[v2]
- rebased on top of latest upstream
- lots of minor fixes
- start support for atomic counters
* we need to move ram_limit_used/max to migration.c
* that means fixing rdma.c
* and test-vmstate.

So I am donig that right now.

Juan Quintela (11):
migration: Update atomic stats out of the mutex
migration: Make multifd_bytes atomic
multifd: We already account for this packet on the multifd thread
multifd: Count the number of bytes sent correctly
migration: Make ram_save_target_page() a pointer
multifd: Make flags field thread local
multifd: Prepare to send a packet without the mutex held
multifd: Add capability to enable/disable zero_page
multifd: Support for zero pages transmission
multifd: Zero pages transmission
So we use multifd to transmit zero pages.

 qapi/migration.json    |   8 ++-
 migration/migration.h  |   1 +
 migration/multifd.h    |  36 ++++++++++--
 migration/ram.h        |   1 +
 hw/core/machine.c      |   1 +
 migration/migration.c  |  16 +++++-
 migration/multifd.c    | 123 +++++++++++++++++++++++++++++++----------
 migration/ram.c        |  51 +++++++++++++++--
 migration/trace-events |   8 +--
 9 files changed, 197 insertions(+), 48 deletions(-)
Signature · 1 line
2.39.1