From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Dillow Subject: Re: [RFC 0/8] Reliably generate large request from SRP Date: Tue, 15 Mar 2011 19:51:28 -0400 Message-ID: <1300233089.21807.26.camel@lap75545.ornl.gov> References: <1295411242-26148-1-git-send-email-dillowda@ornl.gov> <1300148888.2772.15.camel@lap75545.ornl.gov> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: In-reply-to: Sender: linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Bart Van Assche Cc: "roland-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org" , linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Or Gerlitz List-Id: linux-rdma@vger.kernel.org [Fixed linux-rdma address] On Tue, 2011-03-15 at 07:13 -0400, Bart Van Assche wrote: > On Tue, Mar 15, 2011 at 1:28 AM, David Dillow wrote: > > On Tue, 2011-01-18 at 23:27 -0500, David Dillow wrote: > > > Now that at least one vendor is implementing full support for the SRP > > > indirect memory descriptor tables, we can safely expand the sg_tablesize, > > > and realize some performance gains, in many cases quite large. I don't > > > have vendor code that implements the full support needed for safety, but > > > the rareness of FMR mapping failures allows the mapping code to function, > > > at a risk, with existing targets. > > > > I'm getting ready to re-roll this series to address Or's review > > comments. I'm going to leave BMME support until later, as I have a few > > concerns about the error cases. > > > > I'd like to see this go into 2.6.39 -- do you guys have any other > > questions or comments I should address? > > I still have to take a closer look at these patches. But it would be > convenient if the patches could be made available in a (branch of a) > public git repository - that makes testing easier. And as you probably > know temporary branches can be deleted at any time via git push > :. I've pushed these to git://git.kernel.org/pub/scm/linux/kernel/git/dad/srp-initiator.git external-indirect They should mirror out shortly. I'll also make another review pass over the code. Here's the diff from the last version I posted: diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index 1da8b25..376d640 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -1934,7 +1934,7 @@ static const match_table_t srp_opt_tokens = { { SRP_OPT_INITIATOR_EXT, "initiator_ext=%s" }, { SRP_OPT_CMD_SG_ENTRIES, "cmd_sg_entries=%u" }, { SRP_OPT_ALLOW_EXT_SG, "allow_ext_sg=%u" }, - { SRP_OPT_SG_TABLESIZE, "sg_tablesize=%u" }, + { SRP_OPT_SG_TABLESIZE, "sg_tablesize=%u" }, { SRP_OPT_ERR, NULL } }; @@ -2347,7 +2347,7 @@ static void srp_add_one(struct ib_device *device) fmr_param.pool_size = SRP_FMR_POOL_SIZE; fmr_param.dirty_watermark = SRP_FMR_DIRTY_SIZE; fmr_param.cache = 1; - fmr_param.max_pages_per_fmr = SRP_FMR_SIZE; + fmr_param.max_pages_per_fmr = max_pages_per_fmr; fmr_param.page_shift = fmr_page_shift; fmr_param.access = (IB_ACCESS_LOCAL_WRITE | IB_ACCESS_REMOTE_WRITE | -- Dave Dillow National Center for Computational Science Oak Ridge National Laboratory (865) 241-6602 office -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html