From mboxrd@z Thu Jan 1 00:00:00 1970 From: Suman Anna Subject: Re: [PATCH] rpmsg: compute number of buffers to allocate from vrings Date: Wed, 13 Aug 2014 10:47:46 -0500 Message-ID: <53EB88A2.1070408@ti.com> References: <1404420815-42108-1-git-send-email-s-anna@ti.com> <53EA3E78.6000304@ti.com> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org To: Ohad Ben-Cohen Cc: Rusty Russell , "linux-kernel@vger.kernel.org" , "linux-omap@vger.kernel.org" List-Id: linux-omap@vger.kernel.org Hi Ohad, On 08/13/2014 08:40 AM, Ohad Ben-Cohen wrote: > Hi Suman, > > On Tue, Aug 12, 2014 at 7:19 PM, Suman Anna wrote: >> Yes, I was playing around with using less buffers in the remoteproc >> resource table for the vrings. The remoteproc virtio code creates the >> vrings using the number of buffers based on .num field value of struct >> fw_rsc_vdev_vring in the resource table. The virtio rpmsg probe code >> though tries to set up the receive buffers for the same virtqueue based >> on the current hard-coded value of 512 buffers and virtqueue_add_inbuf >> would fail as the virtqueue is created with less number of buffers and >> throws a WARN_ON. > > Gotcha - thanks for the details. > > Limiting the number of buffers in case the vrings are too small makes > sense, but let's use RPMSG_NUM_BUFS as an upper bound, so wacky > resource tables won't trigger unreasonable memory waste. > > Something in the lines of: > > vrp->num_bufs = min(PMSG_NUM_BUFS, virtqueue_get_vring_size(vrp->rvq) * 2); > > Should probably do the trick. > > Does this satisfy your requirement? Yeah, this will work for me. I will go ahead and add a WARN_ON as well to detect above wacky condition, and if someone really needs more buffers in the future, we can revisit this. regards Suman