Re: [Qemu-devel] [PATCH 02/41] fix migration sync

[Paolo Bonzini]

Il 21/09/2012 10:46, Juan Quintela ha scritto:
> Signed-off-by: Juan Quintela <address@hidden>
> ---
>  arch_init.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/arch_init.c b/arch_init.c
> index f849f9b..cdd8ab7 100644
> --- a/arch_init.c
> +++ b/arch_init.c
> @@ -489,6 +489,7 @@ static int ram_save_setup(QEMUFile *f, void *opaque)
>      ram_addr_t addr;
>      RAMBlock *block;
> 
> +    memory_global_sync_dirty_bitmap(get_system_memory());

Does it make sense to call this function before memory_global_dirty_log_start()?

Also, does this call subsume this loop:

    QLIST_FOREACH(block, &ram_list.blocks, next) {
        for (addr = 0; addr < block->length; addr += TARGET_PAGE_SIZE) {
            if (!memory_region_get_dirty(block->mr, addr, TARGET_PAGE_SIZE,
                                         DIRTY_MEMORY_MIGRATION)) {
                memory_region_set_dirty(block->mr, addr, TARGET_PAGE_SIZE);
            }
        }
    }

so that the right fix is

-    QLIST_FOREACH(block, &ram_list.blocks, next) {
-        for (addr = 0; addr < block->length; addr += TARGET_PAGE_SIZE) {
-            if (!memory_region_get_dirty(block->mr, addr, TARGET_PAGE_SIZE,
-                                         DIRTY_MEMORY_MIGRATION)) {
-                memory_region_set_dirty(block->mr, addr, TARGET_PAGE_SIZE);
-            }
-        }
-    }
-
     memory_global_dirty_log_start();
+    memory_global_sync_dirty_bitmap(get_system_memory());

?

Paolo

>      bytes_transferred = 0;
>      last_block = NULL;
>      last_offset = 0;
>