On 12/11/2014 06:28 PM, Liang Li wrote: > Signed-off-by: Liang Li <liang.z...@intel.com> > Signed-off-by: Yang Zhang <yang.z.zh...@intel.com> > ---
> +void migrate_compress_threads_create(MigrationState *s)
> +{
> + int i, thread_count;
> +
> + if (!migrate_use_compression()) {
> + return;
> + }
> + quit_thread = false;
> + thread_count = migrate_compress_threads();
> + s->compress_thread = g_malloc0(sizeof(QemuThread)
> + * thread_count);
Theoretically unsafe (well, unsafe if thread_count were unbounded,
although it looks like you artificially cap it at 255 later in the
series); better would be:
s->compress_thread = g_new0(QemuThread, thread_count)
because that catches potential multiplication overflow.
> + comp_param = g_malloc0(sizeof(compress_param) * thread_count);
Likewise.
>
> +static int ram_save_compressed_page(QEMUFile *f, RAMBlock* block,
Spacing is off on the second '*'.
> + ram_addr_t offset, bool last_stage)
Indentation is off.
--
Eric Blake eblake redhat com +1-919-301-3266
Libvirt virtualization library http://libvirt.org
signature.asc
Description: OpenPGP digital signature
