[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] [PULL 5/5] migratioin/ram: leave RAMBlock->bmap blank on al
From: |
Juan Quintela |
Subject: |
[Qemu-devel] [PULL 5/5] migratioin/ram: leave RAMBlock->bmap blank on allocating |
Date: |
Wed, 5 Jun 2019 13:53:18 +0200 |
From: Wei Yang <address@hidden>
During migration, we would sync bitmap from ram_list.dirty_memory to
RAMBlock.bmap in cpu_physical_memory_sync_dirty_bitmap().
Since we set RAMBlock.bmap and ram_list.dirty_memory both to all 1, this
means at the first round this sync is meaningless and is a duplicated
work.
Leaving RAMBlock->bmap blank on allocating would have a side effect on
migration_dirty_pages, since it is calculated from the result of
cpu_physical_memory_sync_dirty_bitmap(). To keep it right, we need to
set migration_dirty_pages to 0 in ram_state_init().
Signed-off-by: Wei Yang <address@hidden>
Reviewed-by: Dr. David Alan Gilbert <address@hidden>
Reviewed-by: Juan Quintela <address@hidden>
Signed-off-by: Juan Quintela <address@hidden>
---
migration/ram.c | 18 +++++++++++++-----
1 file changed, 13 insertions(+), 5 deletions(-)
diff --git a/migration/ram.c b/migration/ram.c
index 03a9cce9f9..082aea9d23 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -3173,11 +3173,11 @@ static int ram_state_init(RAMState **rsp)
QSIMPLEQ_INIT(&(*rsp)->src_page_requests);
/*
- * Count the total number of pages used by ram blocks not including any
- * gaps due to alignment or unplugs.
+ * This must match with the initial values of dirty bitmap.
+ * Currently we initialize the dirty bitmap to all zeros so
+ * here the total dirty page count is zero.
*/
- (*rsp)->migration_dirty_pages = ram_bytes_total() >> TARGET_PAGE_BITS;
-
+ (*rsp)->migration_dirty_pages = 0;
ram_state_reset(*rsp);
return 0;
@@ -3192,8 +3192,16 @@ static void ram_list_init_bitmaps(void)
if (ram_bytes_total()) {
RAMBLOCK_FOREACH_NOT_IGNORED(block) {
pages = block->max_length >> TARGET_PAGE_BITS;
+ /*
+ * The initial dirty bitmap for migration must be set with all
+ * ones to make sure we'll migrate every guest RAM page to
+ * destination.
+ * Here we didn't set RAMBlock.bmap simply because it is already
+ * set in ram_list.dirty_memory[DIRTY_MEMORY_MIGRATION] in
+ * ram_block_add, and that's where we'll sync the dirty bitmaps.
+ * Here setting RAMBlock.bmap would be fine too but not necessary.
+ */
block->bmap = bitmap_new(pages);
- bitmap_set(block->bmap, 0, pages);
if (migrate_postcopy_ram()) {
block->unsentmap = bitmap_new(pages);
bitmap_set(block->unsentmap, 0, pages);
--
2.21.0
- [Qemu-devel] [PULL 0/5] Migration patches, Juan Quintela, 2019/06/05
- [Qemu-devel] [PULL 1/5] migration/ram.c: MultiFDSendParams.sem_sync is not really used, Juan Quintela, 2019/06/05
- [Qemu-devel] [PULL 2/5] migration/ram.c: multifd_send_state->count is not really used, Juan Quintela, 2019/06/05
- [Qemu-devel] [PULL 5/5] migratioin/ram: leave RAMBlock->bmap blank on allocating,
Juan Quintela <=
- [Qemu-devel] [PULL 3/5] migration: Fix fd protocol for incoming defer, Juan Quintela, 2019/06/05
- [Qemu-devel] [PULL 4/5] migration-test: Add a test for fd protocol, Juan Quintela, 2019/06/05
- Re: [Qemu-devel] [PULL 0/5] Migration patches, Peter Maydell, 2019/06/06