From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([208.118.235.92]:40028) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1UIkLB-0000nH-I4 for qemu-devel@nongnu.org; Thu, 21 Mar 2013 14:35:56 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1UIkL7-0004Qp-7M for qemu-devel@nongnu.org; Thu, 21 Mar 2013 14:35:52 -0400 Received: from mx1.redhat.com ([209.132.183.28]:25879) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1UIk6B-0007Ld-2W for qemu-devel@nongnu.org; Thu, 21 Mar 2013 14:20:23 -0400 Date: Thu, 21 Mar 2013 20:21:00 +0200 From: "Michael S. Tsirkin" Message-ID: <20130321182100.GE4366@redhat.com> References: <1363881940-27505-1-git-send-email-owasserm@redhat.com> <1363881940-27505-9-git-send-email-owasserm@redhat.com> <87mwtwpsce.fsf@elfo.elfo> <514B4CBA.9090803@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <514B4CBA.9090803@redhat.com> Subject: Re: [Qemu-devel] [PATCH v3 8/9] Use qemu_put_buffer_no_copy for guest memory pages List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Orit Wasserman Cc: pbonzini@redhat.com, chegu_vinod@hp.com, qemu-devel@nongnu.org, quintela@redhat.com On Thu, Mar 21, 2013 at 08:08:58PM +0200, Orit Wasserman wrote: > On 03/21/2013 07:37 PM, Juan Quintela wrote: > > Orit Wasserman wrote: > >> This will remove an unneeded copy of guest memory pages. > >> For the page header and device state we still copy the data to the > >> static buffer the other option is to allocate the memory on demand > >> which is more expensive. > >> > >> Signed-off-by: Orit Wasserman > >> --- > >> arch_init.c | 2 +- > >> savevm.c | 2 +- > >> 2 files changed, 2 insertions(+), 2 deletions(-) > >> > >> diff --git a/arch_init.c b/arch_init.c > >> index 98e2bc6..27b53eb 100644 > >> --- a/arch_init.c > >> +++ b/arch_init.c > >> @@ -481,7 +481,7 @@ static int ram_save_block(QEMUFile *f, bool last_stage) > >> /* XBZRLE overflow or normal page */ > >> if (bytes_sent == -1) { > >> bytes_sent = save_block_hdr(f, block, offset, cont, RAM_SAVE_FLAG_PAGE); > >> - qemu_put_buffer(f, p, TARGET_PAGE_SIZE); > >> + qemu_put_buffer_no_copy(f, p, TARGET_PAGE_SIZE); > >> bytes_sent += TARGET_PAGE_SIZE; > >> acct_info.norm_pages++; > >> } > > > > Once here, shouldn't we also change: > > > > block-migration.c::blk_send() > > > > qemu_put_buffer(f, blk->buf, BLOCK_SIZE); > > > > to nocopy? > Sadly no, I looked at the code and I saw the buffer is freed immediately > after call blk_send :(. look at mig_save_device_dirty. > > Orit Hmm if _nocopy also implies it is not safe to free the buffer immediately, then I'd rather put this in a name e.g. _async instead of _nocopy. > > > > Again, this can be an additional patch. > > > > > >> diff --git a/savevm.c b/savevm.c > >> index fbfb9e3..50e8fb2 100644 > >> --- a/savevm.c > >> +++ b/savevm.c > >> @@ -634,7 +634,7 @@ void qemu_put_buffer_no_copy(QEMUFile *f, const uint8_t *buf, int size) > >> abort(); > >> } > >> > >> - f->iov[f->iovcnt].iov_base = f->buf + f->buf_index; > >> + f->iov[f->iovcnt].iov_base = (uint8_t *)buf; > >> f->iov[f->iovcnt++].iov_len = size; > > > > This bit should be in the previous patch, the error that I pointed? > > > > Later, Juan. > >