From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:46268) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1afW6k-00082u-0n for qemu-devel@nongnu.org; Mon, 14 Mar 2016 13:16:43 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1afW6g-0002KN-AL for qemu-devel@nongnu.org; Mon, 14 Mar 2016 13:16:41 -0400 Received: from mx1.redhat.com ([209.132.183.28]:41030) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1afW6g-0002JX-4S for qemu-devel@nongnu.org; Mon, 14 Mar 2016 13:16:38 -0400 Date: Mon, 14 Mar 2016 17:16:34 +0000 From: "Daniel P. Berrange" Message-ID: <20160314171634.GK21198@redhat.com> References: <1457753338-22089-1-git-send-email-dhannawatpooja1@gmail.com> <1457973473-21791-1-git-send-email-dhannawatpooja1@gmail.com> <20160314165032.GJ21198@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: Subject: Re: [Qemu-devel] [PATCH v2 1/1] socket: Allocating Large sized arrays to heap Reply-To: "Daniel P. Berrange" List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Pooja Dhannawat Cc: qemu-devel@nongnu.org On Mon, Mar 14, 2016 at 10:44:22PM +0530, Pooja Dhannawat wrote: > On Mon, Mar 14, 2016 at 10:20 PM, Daniel P. Berrange > wrote: > > > On Mon, Mar 14, 2016 at 10:07:53PM +0530, Pooja Dhannawat wrote: > > > net_socket_send has a huge stack usage of 69712 bytes approx. > > > Moving large arrays to heap to reduce stack usage. > > > > > > Signed-off-by: Pooja Dhannawat > > > --- > > > net/socket.c | 8 +++++--- > > > 1 file changed, 5 insertions(+), 3 deletions(-) > > > > > > diff --git a/net/socket.c b/net/socket.c > > > index e32e3cb..3fcd7a6 100644 > > > --- a/net/socket.c > > > +++ b/net/socket.c > > > @@ -147,10 +147,10 @@ static void net_socket_send(void *opaque) > > > NetSocketState *s = opaque; > > > int size, err; > > > unsigned l; > > > - uint8_t buf1[NET_BUFSIZE]; > > > + uint8_t *buf1 = g_new(uint8_t, NET_BUFSIZE); > > > > You're allocating NET_BUFSIZE worth of uint8_t's > > > > I didn't get you clear. > > > > > const uint8_t *buf; > > > > > > - size = qemu_recv(s->fd, buf1, sizeof(buf1), 0); > > > + size = qemu_recv(s->fd, (uint8_t *)buf1, sizeof(uint8_t), 0); > > > > But only reading 1 byte which is clearly wrong. You likely wanted > > NET_BUFSIZE here, not sizeof(uint8_t) > > > > Correct me If I am wrong. This should also work : > size = qemu_recv(s->fd, (uint8_t *)buf1, NET_BUFSIZE * sizeof(uint8_t), 0); Sure, if you want to add a pointless 'multiply by 1' to the code, which you didn't bother doing for the g_new call. Regards, Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|