All of lore.kernel.org
 help / color / mirror / Atom feed
From: zhanghailiang <zhang.zhanghailiang@huawei.com>
To: qemu-devel@nongnu.org, dgilbert@redhat.com,
	zhangchen.fnst@cn.fujitsu.com
Cc: lizhijian@cn.fujitsu.com, xiecl.fnst@cn.fujitsu.com,
	zhanghailiang <zhang.zhanghailiang@huawei.com>,
	Juan Quintela <quintela@redhat.com>
Subject: [Qemu-devel] [PATCH 09/15] COLO: Flush PVM's cached RAM into SVM's memory
Date: Wed, 22 Feb 2017 11:42:10 +0800	[thread overview]
Message-ID: <1487734936-43472-10-git-send-email-zhang.zhanghailiang@huawei.com> (raw)
In-Reply-To: <1487734936-43472-1-git-send-email-zhang.zhanghailiang@huawei.com>

During the time of VM's running, PVM may dirty some pages, we will transfer
PVM's dirty pages to SVM and store them into SVM's RAM cache at next checkpoint
time. So, the content of SVM's RAM cache will always be same with PVM's memory
after checkpoint.

Instead of flushing all content of PVM's RAM cache into SVM's MEMORY,
we do this in a more efficient way:
Only flush any page that dirtied by PVM since last checkpoint.
In this way, we can ensure SVM's memory same with PVM's.

Besides, we must ensure flush RAM cache before load device state.

Cc: Juan Quintela <quintela@redhat.com>
Signed-off-by: zhanghailiang <zhang.zhanghailiang@huawei.com>
Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
---
 include/migration/migration.h |  1 +
 migration/ram.c               | 41 +++++++++++++++++++++++++++++++++++++++++
 migration/trace-events        |  2 ++
 3 files changed, 44 insertions(+)

diff --git a/include/migration/migration.h b/include/migration/migration.h
index 93c6148..ba5b97b 100644
--- a/include/migration/migration.h
+++ b/include/migration/migration.h
@@ -383,4 +383,5 @@ PostcopyState postcopy_state_set(PostcopyState new_state);
 /* ram cache */
 int colo_init_ram_cache(void);
 void colo_release_ram_cache(void);
+void colo_flush_ram_cache(void);
 #endif
diff --git a/migration/ram.c b/migration/ram.c
index ed3b606..3f57fe0 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -2540,6 +2540,7 @@ static int ram_load(QEMUFile *f, void *opaque, int version_id)
      * be atomic
      */
     bool postcopy_running = postcopy_state_get() >= POSTCOPY_INCOMING_LISTENING;
+    bool need_flush = false;
 
     seq_iter++;
 
@@ -2574,6 +2575,7 @@ static int ram_load(QEMUFile *f, void *opaque, int version_id)
             /* After going into COLO, we should load the Page into colo_cache */
             if (ram_cache_enable) {
                 host = colo_cache_from_block_offset(block, addr);
+                need_flush = true;
             } else {
                 host = host_from_ram_block_offset(block, addr);
             }
@@ -2668,6 +2670,10 @@ static int ram_load(QEMUFile *f, void *opaque, int version_id)
     wait_for_decompress_done();
     rcu_read_unlock();
     trace_ram_load_complete(ret, seq_iter);
+
+    if (!ret  && ram_cache_enable && need_flush) {
+        colo_flush_ram_cache();
+    }
     return ret;
 }
 
@@ -2738,6 +2744,41 @@ void colo_release_ram_cache(void)
     rcu_read_unlock();
 }
 
+/*
+ * Flush content of RAM cache into SVM's memory.
+ * Only flush the pages that be dirtied by PVM or SVM or both.
+ */
+void colo_flush_ram_cache(void)
+{
+    RAMBlock *block = NULL;
+    void *dst_host;
+    void *src_host;
+    ram_addr_t offset = 0;
+
+    trace_colo_flush_ram_cache_begin(migration_dirty_pages);
+    rcu_read_lock();
+    block = QLIST_FIRST_RCU(&ram_list.blocks);
+
+    while (block) {
+        ram_addr_t ram_addr_abs;
+        offset = migration_bitmap_find_dirty(block, offset, &ram_addr_abs);
+        migration_bitmap_clear_dirty(ram_addr_abs);
+
+        if (offset >= block->used_length) {
+            offset = 0;
+            block = QLIST_NEXT_RCU(block, next);
+        } else {
+            dst_host = block->host + offset;
+            src_host = block->colo_cache + offset;
+            memcpy(dst_host, src_host, TARGET_PAGE_SIZE);
+        }
+    }
+
+    rcu_read_unlock();
+    trace_colo_flush_ram_cache_end();
+    assert(migration_dirty_pages == 0);
+}
+
 static SaveVMHandlers savevm_ram_handlers = {
     .save_live_setup = ram_save_setup,
     .save_live_iterate = ram_save_iterate,
diff --git a/migration/trace-events b/migration/trace-events
index fa660e3..5d4cf80 100644
--- a/migration/trace-events
+++ b/migration/trace-events
@@ -71,6 +71,8 @@ migration_throttle(void) ""
 ram_load_postcopy_loop(uint64_t addr, int flags) "@%" PRIx64 " %x"
 ram_postcopy_send_discard_bitmap(void) ""
 ram_save_queue_pages(const char *rbname, size_t start, size_t len) "%s: start: %zx len: %zx"
+colo_flush_ram_cache_begin(uint64_t dirty_pages) "dirty_pages %" PRIu64
+colo_flush_ram_cache_end(void) ""
 
 # migration/migration.c
 await_return_path_close_on_source_close(void) ""
-- 
1.8.3.1

  parent reply	other threads:[~2017-02-22  3:42 UTC|newest]

Thread overview: 42+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-02-22  3:42 [Qemu-devel] [PATCH 00/15] COLO: integrate colo frame with block replication and net compare zhanghailiang
2017-02-22  3:42 ` [Qemu-devel] [PATCH 01/15] net/colo: Add notifier/callback related helpers for filter zhanghailiang
2017-04-07 15:46   ` Dr. David Alan Gilbert
2017-04-10  7:26     ` Hailiang Zhang
2017-02-22  3:42 ` [Qemu-devel] [PATCH 02/15] colo-compare: implement the process of checkpoint zhanghailiang
2017-02-22  9:31   ` Zhang Chen
2017-02-23  1:02     ` Hailiang Zhang
2017-02-23  5:49       ` Zhang Chen
2017-04-14  5:57     ` Jason Wang
2017-04-14  6:22       ` Hailiang Zhang
2017-04-14  6:38         ` Jason Wang
2017-04-17 11:04           ` Hailiang Zhang
2017-04-18  1:32             ` Zhang Chen
2017-04-18  3:55             ` Jason Wang
2017-04-18  6:58               ` Hailiang Zhang
2017-04-20  5:15                 ` Jason Wang
2017-04-21  8:10                   ` Hailiang Zhang
2017-02-22  3:42 ` [Qemu-devel] [PATCH 03/15] colo-compare: use notifier to notify packets comparing result zhanghailiang
2017-02-22  3:42 ` [Qemu-devel] [PATCH 04/15] COLO: integrate colo compare with colo frame zhanghailiang
2017-04-07 15:59   ` Dr. David Alan Gilbert
2017-02-22  3:42 ` [Qemu-devel] [PATCH 05/15] COLO: Handle shutdown command for VM in COLO state zhanghailiang
2017-02-22 15:35   ` Eric Blake
2017-02-23  1:15     ` Hailiang Zhang
2017-02-22  3:42 ` [Qemu-devel] [PATCH 06/15] COLO: Add block replication into colo process zhanghailiang
2017-02-22  3:42 ` [Qemu-devel] [PATCH 07/15] COLO: Load PVM's dirty pages into SVM's RAM cache temporarily zhanghailiang
2017-04-07 17:06   ` Dr. David Alan Gilbert
2017-04-10  7:31     ` Hailiang Zhang
2017-02-22  3:42 ` [Qemu-devel] [PATCH 08/15] ram/COLO: Record the dirty pages that SVM received zhanghailiang
2017-02-23 18:44   ` Dr. David Alan Gilbert
2017-02-22  3:42 ` zhanghailiang [this message]
2017-02-22  3:42 ` [Qemu-devel] [PATCH 10/15] qmp event: Add COLO_EXIT event to notify users while exited from COLO zhanghailiang
2017-02-22  3:42 ` [Qemu-devel] [PATCH 11/15] savevm: split save/find loadvm_handlers entry into two helper functions zhanghailiang
2017-02-22  3:42 ` [Qemu-devel] [PATCH 12/15] savevm: split the process of different stages for loadvm/savevm zhanghailiang
2017-04-07 17:18   ` Dr. David Alan Gilbert
2017-04-10  8:26     ` Hailiang Zhang
2017-04-20  9:09       ` Dr. David Alan Gilbert
2017-04-21  6:50         ` Hailiang Zhang
2017-02-22  3:42 ` [Qemu-devel] [PATCH 13/15] COLO: Separate the process of saving/loading ram and device state zhanghailiang
2017-02-22  3:42 ` [Qemu-devel] [PATCH 14/15] COLO: Split qemu_savevm_state_begin out of checkpoint process zhanghailiang
2017-02-22  3:42 ` [Qemu-devel] [PATCH 15/15] COLO: flush host dirty ram from cache zhanghailiang
2017-04-07 17:39   ` Dr. David Alan Gilbert
2017-04-10  7:13     ` Hailiang Zhang

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=1487734936-43472-10-git-send-email-zhang.zhanghailiang@huawei.com \
    --to=zhang.zhanghailiang@huawei.com \
    --cc=dgilbert@redhat.com \
    --cc=lizhijian@cn.fujitsu.com \
    --cc=qemu-devel@nongnu.org \
    --cc=quintela@redhat.com \
    --cc=xiecl.fnst@cn.fujitsu.com \
    --cc=zhangchen.fnst@cn.fujitsu.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.