qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Denis V. Lunev" <den@openvz.org>
Cc: Igor Redko <redkoi@virtuozzo.com>,
	Juan Quintela <quintela@redhat.com>,
	Anna Melekhova <annam@virtuozzo.com>,
	qemu-devel@nongnu.org, Paolo Bonzini <pbonzini@redhat.com>,
	Amit Shah <amit.shah@redhat.com>,
	"Denis V. Lunev" <den@openvz.org>
Subject: [Qemu-devel] [PATCH 2/2] migration: fix deadlock
Date: Fri, 25 Sep 2015 12:09:40 +0300	[thread overview]
Message-ID: <1443172180-1005-3-git-send-email-den@openvz.org> (raw)
In-Reply-To: <1443172180-1005-1-git-send-email-den@openvz.org>

Release qemu global mutex before call synchronize_rcu().
synchronize_rcu() waiting for all readers to finish their critical
sections. There is at least one critical section in which we try
to get QGM (critical section is in address_space_rw() and
prepare_mmio_access() is trying to aquire QGM).

Both functions (migration_end() and migration_bitmap_extend())
are called from main thread which is holding QGM.

Thus there is a race condition that ends up with deadlock:
main thread     working thread
Lock QGA                |
|             Call KVM_EXIT_IO handler
|                       |
|        Open rcu reader's critical section
Migration cleanup bh    |
|                       |
synchronize_rcu() is    |
waiting for readers     |
|            prepare_mmio_access() is waiting for QGM
  \                   /
         deadlock

The patch changes bitmap freeing from direct g_free after synchronize_rcu
to g_free_rcu.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Reported-by: Igor Redko <redkoi@virtuozzo.com>
CC: Igor Redko <redkoi@virtuozzo.com>
CC: Anna Melekhova <annam@virtuozzo.com>
CC: Juan Quintela <quintela@redhat.com>
CC: Amit Shah <amit.shah@redhat.com>
CC: Paolo Bonzini <pbonzini@redhat.com>
CC: Wen Congyang <wency@cn.fujitsu.com>
---
 migration/ram.c | 43 ++++++++++++++++++++++++++++---------------
 1 file changed, 28 insertions(+), 15 deletions(-)

diff --git a/migration/ram.c b/migration/ram.c
index a712c68..56b6fce 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -221,12 +221,27 @@ static RAMBlock *last_seen_block;
 /* This is the last block from where we have sent data */
 static RAMBlock *last_sent_block;
 static ram_addr_t last_offset;
-static unsigned long *migration_bitmap;
 static QemuMutex migration_bitmap_mutex;
 static uint64_t migration_dirty_pages;
 static uint32_t last_version;
 static bool ram_bulk_stage;
 
+static struct BitmapRcu {
+    struct rcu_head rcu;
+    unsigned long bmap[0];
+} *migration_bitmap_rcu;
+
+static inline struct BitmapRcu *bitmap_new_rcu(long nbits)
+{
+    long len = BITS_TO_LONGS(nbits) * sizeof(unsigned long);
+    struct BitmapRcu *ptr = g_try_malloc0(len + sizeof(struct BitmapRcu));
+    if (ptr == NULL) {
+        abort();
+    }
+    return ptr;
+}
+
+
 struct CompressParam {
     bool start;
     bool done;
@@ -508,7 +523,7 @@ ram_addr_t migration_bitmap_find_and_reset_dirty(MemoryRegion *mr,
 
     unsigned long next;
 
-    bitmap = atomic_rcu_read(&migration_bitmap);
+    bitmap = atomic_rcu_read(&migration_bitmap_rcu)->bmap;
     if (ram_bulk_stage && nr > base) {
         next = nr + 1;
     } else {
@@ -526,7 +541,7 @@ ram_addr_t migration_bitmap_find_and_reset_dirty(MemoryRegion *mr,
 static void migration_bitmap_sync_range(ram_addr_t start, ram_addr_t length)
 {
     unsigned long *bitmap;
-    bitmap = atomic_rcu_read(&migration_bitmap);
+    bitmap = atomic_rcu_read(&migration_bitmap_rcu)->bmap;
     migration_dirty_pages +=
         cpu_physical_memory_sync_dirty_bitmap(bitmap, start, length);
 }
@@ -1029,12 +1044,11 @@ static void migration_end(void)
     /* caller have hold iothread lock or is in a bh, so there is
      * no writing race against this migration_bitmap
      */
-    unsigned long *bitmap = migration_bitmap;
-    atomic_rcu_set(&migration_bitmap, NULL);
+    struct BitmapRcu *bitmap = migration_bitmap_rcu;
+    atomic_rcu_set(&migration_bitmap_rcu, NULL);
     if (bitmap) {
         memory_global_dirty_log_stop();
-        synchronize_rcu();
-        g_free(bitmap);
+        g_free_rcu(bitmap, rcu);
     }
 
     XBZRLE_cache_lock();
@@ -1070,9 +1084,9 @@ void migration_bitmap_extend(ram_addr_t old, ram_addr_t new)
     /* called in qemu main thread, so there is
      * no writing race against this migration_bitmap
      */
-    if (migration_bitmap) {
-        unsigned long *old_bitmap = migration_bitmap, *bitmap;
-        bitmap = bitmap_new(new);
+    if (migration_bitmap_rcu) {
+        struct BitmapRcu *old_bitmap = migration_bitmap_rcu, *bitmap;
+        bitmap = bitmap_new_rcu(new);
 
         /* prevent migration_bitmap content from being set bit
          * by migration_bitmap_sync_range() at the same time.
@@ -1080,12 +1094,11 @@ void migration_bitmap_extend(ram_addr_t old, ram_addr_t new)
          * at the same time.
          */
         qemu_mutex_lock(&migration_bitmap_mutex);
-        bitmap_copy(bitmap, old_bitmap, old);
-        atomic_rcu_set(&migration_bitmap, bitmap);
+        bitmap_copy(bitmap->bmap, old_bitmap->bmap, old);
+        atomic_rcu_set(&migration_bitmap_rcu, bitmap);
         qemu_mutex_unlock(&migration_bitmap_mutex);
         migration_dirty_pages += new - old;
-        synchronize_rcu();
-        g_free(old_bitmap);
+        g_free_rcu(old_bitmap, rcu);
     }
 }
 
@@ -1144,7 +1157,7 @@ static int ram_save_setup(QEMUFile *f, void *opaque)
     reset_ram_globals();
 
     ram_bitmap_pages = last_ram_offset() >> TARGET_PAGE_BITS;
-    migration_bitmap = bitmap_new(ram_bitmap_pages);
+    migration_bitmap_rcu = bitmap_new_rcu(ram_bitmap_pages);
 
     /*
      * Count the total number of pages used by ram blocks not including any
-- 
2.1.4

  parent reply	other threads:[~2015-09-25  9:09 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-09-24 12:53 [Qemu-devel] [PATCH 1/1] migration: fix deadlock Denis V. Lunev
2015-09-25  1:21 ` Wen Congyang
2015-09-25  8:03   ` Denis V. Lunev
2015-09-25  8:23     ` Wen Congyang
2015-09-25  9:09       ` [Qemu-devel] [PATCH v2 0/2] " Denis V. Lunev
2015-09-25  9:09         ` [Qemu-devel] [PATCH 1/2] migration: bitmap_set is unnecessary as bitmap_new uses g_try_malloc0 Denis V. Lunev
2015-09-25  9:24           ` Wen Congyang
2015-09-25  9:31             ` Denis V. Lunev
2015-09-25  9:37               ` Wen Congyang
2015-09-25 10:05                 ` Denis V. Lunev
2015-09-25  9:09         ` Denis V. Lunev [this message]
2015-09-25  9:35           ` [Qemu-devel] [PATCH 2/2] migration: fix deadlock Wen Congyang
2015-09-25  9:46         ` [Qemu-devel] [PATCH v2 0/2] " Wen Congyang
2015-09-28 10:55           ` Igor Redko
2015-09-28 15:12             ` Igor Redko
2015-09-29  8:47             ` Dr. David Alan Gilbert
2015-09-30 14:28               ` Igor Redko
2015-09-29 15:32       ` [Qemu-devel] [PATCH 1/1] " Igor Redko

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=1443172180-1005-3-git-send-email-den@openvz.org \
    --to=den@openvz.org \
    --cc=amit.shah@redhat.com \
    --cc=annam@virtuozzo.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=quintela@redhat.com \
    --cc=redkoi@virtuozzo.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 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).