Merge tag 'migration-20231011-pull-request' of https://gitlab.com/juan.quintela/qemu into staging

Migration Pull request (20231011 edition)

Hi

In this pull request:

- Markus RDMA cleanup series
- recover fixes from peter
- migration capability from fabiano
- negative migration test from fabiano.

Please, pull.

Thanks, Juan.

# -----BEGIN PGP SIGNATURE-----
#
# iQIzBAABCAAdFiEEGJn/jt6/WMzuA0uC9IfvGFhy1yMFAmUmaQ8ACgkQ9IfvGFhy
# 1yN9fA//SBnea3Wl2158J673l5aaI8Vp/1PjfzvNdcr/6EQbXZBgug+haQ3n5Hhf
# USNRhemrCkpZAGCUf07g9pfF4R/Jsq1OkOrWF4e6gAaZPNU4V5F7VKBk8pmFMLtr
# Kk2XgnH2ZPaFEvts0qBrOfvDHH8gOzzjpF2HGrioM8Zr3p1JHz9OqJoSyawLF0U7
# YFTq2jJSgaOQ6ax1+L8hLLuXlmNccBaTWT8Cv0rbPEgcwrJOM/wMfmd6O39ps929
# yS5NnxqqkrprTDjmeGOgOQd0Cy/flinnzmu+BVMO6/ns9Hu6q1TGG6D+DOBdgmHH
# jq7Ej5VILtXWOoZtXLHqA1Xt73ciVlmditVupoC+5vtIJou2JseClutOp98qxxzV
# llMF7ldHbRTWnu7qIrwv2OINarowR0pIZfkJqBc6dNHHScwMCnX5L9YAvNePEo2V
# 1oJpbqW7mmgwdlFAiKFD+AE6qUWxcnzOvPf+fzWrJMi507Kv5nmxQWTHw9dsFs7k
# neWnK21t0s2t77+vVBtLlr06JESG+WndzvQsXKZu8Pd0+ASnzpX8pRVzxEPk5EiH
# fT9bhXOCvxTTHulznjkOApODE5NF+KlHAFXU87cSIkdi/6JfzcvTe6KeeIPC248Q
# jk3nVlhds1xajTcPAK7HF5Ta6R8rNdTZ6q/kFNhLaTGqv9agxDU=
# =hekO
# -----END PGP SIGNATURE-----
# gpg: Signature made Wed 11 Oct 2023 05:21:19 EDT
# gpg:                using RSA key 1899FF8EDEBF58CCEE034B82F487EF185872D723
# gpg: Good signature from "Juan Quintela <quintela@redhat.com>" [full]
# gpg:                 aka "Juan Quintela <quintela@trasno.org>" [full]
# Primary key fingerprint: 1899 FF8E DEBF 58CC EE03  4B82 F487 EF18 5872 D723

* tag 'migration-20231011-pull-request' of https://gitlab.com/juan.quintela/qemu: (65 commits)
  migration: Add migration_rp_wait|kick()
  migration: Remember num of ramblocks to sync during recovery
  qemufile: Always return a verbose error
  migration: Introduce migrate_has_error()
  migration: Display error in query-migrate irrelevant of status
  migration/rdma: Replace flawed device detail dump by tracing
  migration/rdma: Use error_report() & friends instead of stderr
  migration/rdma: Downgrade qemu_rdma_cleanup() errors to warnings
  migration/rdma: Silence qemu_rdma_register_and_get_keys()
  migration/rdma: Silence qemu_rdma_block_for_wrid()
  migration/rdma: Don't report received completion events as error
  migration/rdma: Silence qemu_rdma_reg_control()
  migration/rdma: Silence qemu_rdma_connect()
  migration/rdma: Silence qemu_rdma_resolve_host()
  migration/rdma: Convert qemu_rdma_alloc_pd_cq() to Error
  migration/rdma: Convert qemu_rdma_post_recv_control() to Error
  migration/rdma: Convert qemu_rdma_post_send_control() to Error
  migration/rdma: Convert qemu_rdma_write() to Error
  migration/rdma: Convert qemu_rdma_write_one() to Error
  migration/rdma: Convert qemu_rdma_write_flush() to Error
  ...

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

1 files changed, 21 insertions, 12 deletions

diff --git a/migration/ram.c b/migration/ram.c
index e4bfd39f08..2f5ce4d60b 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -394,6 +394,14 @@ struct RAMState {
     /* Queue of outstanding page requests from the destination */
     QemuMutex src_page_req_mutex;
     QSIMPLEQ_HEAD(, RAMSrcPageRequest) src_page_requests;
+
+    /*
+     * This is only used when postcopy is in recovery phase, to communicate
+     * between the migration thread and the return path thread on dirty
+     * bitmap synchronizations.  This field is unused in other stages of
+     * RAM migration.
+     */
+    unsigned int postcopy_bmap_sync_requested;
 };
 typedef struct RAMState RAMState;
 
@@ -4119,21 +4127,21 @@ static int ram_dirty_bitmap_sync_all(MigrationState *s, RAMState *rs)
 {
     RAMBlock *block;
     QEMUFile *file = s->to_dst_file;
-    int ramblock_count = 0;
 
     trace_ram_dirty_bitmap_sync_start();
 
+    qatomic_set(&rs->postcopy_bmap_sync_requested, 0);
     RAMBLOCK_FOREACH_NOT_IGNORED(block) {
         qemu_savevm_send_recv_bitmap(file, block->idstr);
         trace_ram_dirty_bitmap_request(block->idstr);
-        ramblock_count++;
+        qatomic_inc(&rs->postcopy_bmap_sync_requested);
     }
 
     trace_ram_dirty_bitmap_sync_wait();
 
     /* Wait until all the ramblocks' dirty bitmap synced */
-    while (ramblock_count--) {
-        qemu_sem_wait(&s->rp_state.rp_sem);
+    while (qatomic_read(&rs->postcopy_bmap_sync_requested)) {
+        migration_rp_wait(s);
     }
 
     trace_ram_dirty_bitmap_sync_complete();
@@ -4141,11 +4149,6 @@ static int ram_dirty_bitmap_sync_all(MigrationState *s, RAMState *rs)
     return 0;
 }
 
-static void ram_dirty_bitmap_reload_notify(MigrationState *s)
-{
-    qemu_sem_post(&s->rp_state.rp_sem);
-}
-
 /*
  * Read the received bitmap, revert it as the initial dirty bitmap.
  * This is only used when the postcopy migration is paused but wants
@@ -4159,6 +4162,7 @@ int ram_dirty_bitmap_reload(MigrationState *s, RAMBlock *block)
     unsigned long *le_bitmap, nbits = block->used_length >> TARGET_PAGE_BITS;
     uint64_t local_size = DIV_ROUND_UP(nbits, 8);
     uint64_t size, end_mark;
+    RAMState *rs = ram_state;
 
     trace_ram_dirty_bitmap_reload_begin(block->idstr);
 
@@ -4225,11 +4229,16 @@ int ram_dirty_bitmap_reload(MigrationState *s, RAMBlock *block)
     /* We'll recalculate migration_dirty_pages in ram_state_resume_prepare(). */
     trace_ram_dirty_bitmap_reload_complete(block->idstr);
 
+    qatomic_dec(&rs->postcopy_bmap_sync_requested);
+
     /*
-     * We succeeded to sync bitmap for current ramblock. If this is
-     * the last one to sync, we need to notify the main send thread.
+     * We succeeded to sync bitmap for current ramblock. Always kick the
+     * migration thread to check whether all requested bitmaps are
+     * reloaded.  NOTE: it's racy to only kick when requested==0, because
+     * we don't know whether the migration thread may still be increasing
+     * it.
      */
-    ram_dirty_bitmap_reload_notify(s);
+    migration_rp_kick(s);
 
     ret = 0;
 out: