From: Juan Quintela <quintela@redhat.com>
To: qemu-devel@nongnu.org
Cc: dgilbert@redhat.com, Chao Fan <fanc.fnst@cn.fujitsu.com>,
Li Zhijian <lizhijian@cn.fujitsu.com>
Subject: [Qemu-devel] [PULL 1/6] Change the method to calculate dirty-pages-rate
Date: Thu, 16 Mar 2017 10:01:36 +0100 [thread overview]
Message-ID: <20170316090141.13421-2-quintela@redhat.com> (raw)
In-Reply-To: <20170316090141.13421-1-quintela@redhat.com>
From: Chao Fan <fanc.fnst@cn.fujitsu.com>
In function cpu_physical_memory_sync_dirty_bitmap, file
include/exec/ram_addr.h:
if (src[idx][offset]) {
unsigned long bits = atomic_xchg(&src[idx][offset], 0);
unsigned long new_dirty;
new_dirty = ~dest[k];
dest[k] |= bits;
new_dirty &= bits;
num_dirty += ctpopl(new_dirty);
}
After these codes executed, only the pages not dirtied in bitmap(dest),
but dirtied in dirty_memory[DIRTY_MEMORY_MIGRATION] will be calculated.
For example:
When ram_list.dirty_memory[DIRTY_MEMORY_MIGRATION] = 0b00001111,
and atomic_rcu_read(&migration_bitmap_rcu)->bmap = 0b00000011,
the new_dirty will be 0b00001100, and this function will return 2 but not
4 which is expected.
the dirty pages in dirty_memory[DIRTY_MEMORY_MIGRATION] are all new,
so these should be calculated also.
Signed-off-by: Chao Fan <fanc.fnst@cn.fujitsu.com>
Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
---
include/exec/ram_addr.h | 5 ++++-
migration/ram.c | 12 +++++-------
2 files changed, 9 insertions(+), 8 deletions(-)
diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
index cd432e7..b05dc84 100644
--- a/include/exec/ram_addr.h
+++ b/include/exec/ram_addr.h
@@ -355,7 +355,8 @@ static inline void cpu_physical_memory_clear_dirty_range(ram_addr_t start,
static inline
uint64_t cpu_physical_memory_sync_dirty_bitmap(unsigned long *dest,
ram_addr_t start,
- ram_addr_t length)
+ ram_addr_t length,
+ int64_t *real_dirty_pages)
{
ram_addr_t addr;
unsigned long page = BIT_WORD(start >> TARGET_PAGE_BITS);
@@ -379,6 +380,7 @@ uint64_t cpu_physical_memory_sync_dirty_bitmap(unsigned long *dest,
if (src[idx][offset]) {
unsigned long bits = atomic_xchg(&src[idx][offset], 0);
unsigned long new_dirty;
+ *real_dirty_pages += ctpopl(bits);
new_dirty = ~dest[k];
dest[k] |= bits;
new_dirty &= bits;
@@ -398,6 +400,7 @@ uint64_t cpu_physical_memory_sync_dirty_bitmap(unsigned long *dest,
start + addr,
TARGET_PAGE_SIZE,
DIRTY_MEMORY_MIGRATION)) {
+ *real_dirty_pages += 1;
long k = (start + addr) >> TARGET_PAGE_BITS;
if (!test_and_set_bit(k, dest)) {
num_dirty++;
diff --git a/migration/ram.c b/migration/ram.c
index 719425b..de1e0a3 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -576,18 +576,18 @@ static inline bool migration_bitmap_clear_dirty(ram_addr_t addr)
return ret;
}
+static int64_t num_dirty_pages_period;
static void migration_bitmap_sync_range(ram_addr_t start, ram_addr_t length)
{
unsigned long *bitmap;
bitmap = atomic_rcu_read(&migration_bitmap_rcu)->bmap;
- migration_dirty_pages +=
- cpu_physical_memory_sync_dirty_bitmap(bitmap, start, length);
+ migration_dirty_pages += cpu_physical_memory_sync_dirty_bitmap(bitmap,
+ start, length, &num_dirty_pages_period);
}
/* Fix me: there are too many global variables used in migration process. */
static int64_t start_time;
static int64_t bytes_xfer_prev;
-static int64_t num_dirty_pages_period;
static uint64_t xbzrle_cache_miss_prev;
static uint64_t iterations_prev;
@@ -620,7 +620,6 @@ uint64_t ram_pagesize_summary(void)
static void migration_bitmap_sync(void)
{
RAMBlock *block;
- uint64_t num_dirty_pages_init = migration_dirty_pages;
MigrationState *s = migrate_get_current();
int64_t end_time;
int64_t bytes_xfer_now;
@@ -646,9 +645,8 @@ static void migration_bitmap_sync(void)
rcu_read_unlock();
qemu_mutex_unlock(&migration_bitmap_mutex);
- trace_migration_bitmap_sync_end(migration_dirty_pages
- - num_dirty_pages_init);
- num_dirty_pages_period += migration_dirty_pages - num_dirty_pages_init;
+ trace_migration_bitmap_sync_end(num_dirty_pages_period);
+
end_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
/* more than 1 second = 1000 millisecons */
--
2.9.3
next prev parent reply other threads:[~2017-03-16 9:01 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-03-16 9:01 [Qemu-devel] [PULL 0/6] Migration fixes Juan Quintela
2017-03-16 9:01 ` Juan Quintela [this message]
2017-03-16 9:01 ` [Qemu-devel] [PULL 2/6] migration: use "" as the default for tls-creds/hostname Juan Quintela
2017-03-16 9:01 ` [Qemu-devel] [PULL 3/6] migration/block: Avoid invoking blk_drain too frequently Juan Quintela
2017-03-16 9:01 ` [Qemu-devel] [PULL 4/6] vmstate: fix failed iotests case 68 and 91 Juan Quintela
2017-03-16 9:01 ` [Qemu-devel] [PULL 5/6] RAMBlocks: qemu_ram_is_shared Juan Quintela
2017-03-16 9:01 ` [Qemu-devel] [PULL 6/6] postcopy: Check for shared memory Juan Quintela
2017-03-16 16:39 ` [Qemu-devel] [PULL 0/6] Migration fixes Peter Maydell
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170316090141.13421-2-quintela@redhat.com \
--to=quintela@redhat.com \
--cc=dgilbert@redhat.com \
--cc=fanc.fnst@cn.fujitsu.com \
--cc=lizhijian@cn.fujitsu.com \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).