From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Gonzalez Monroy, Sergio" Subject: Re: [PATCH 0/2] Removal of RTE_MBUF_REFCNT Date: Tue, 17 Feb 2015 08:43:04 +0000 Message-ID: <54E2FF18.6010408@intel.com> References: <1424102913-18944-1-git-send-email-sergio.gonzalez.monroy@intel.com> <20150216154710.42bd6fe9@uryu.home.lan> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Cc: dev-VfR2kkLFssw@public.gmane.org To: Stephen Hemminger Return-path: In-Reply-To: <20150216154710.42bd6fe9-CA4OZQ/Yy2Lykuyl+CZolw@public.gmane.org> List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces-VfR2kkLFssw@public.gmane.org Sender: "dev" On 16/02/2015 20:47, Stephen Hemminger wrote: > On Mon, 16 Feb 2015 16:08:31 +0000 > Sergio Gonzalez Monroy wrote: > >> This patch tries to remove the RTE_MBUF_REFCNT config options and dependencies >> by introducing a new mbuf flag IND_ATTACHED_MBUF that would indicate when the mbuf >> is an indirect attached mbuf, to differentiate between indirect mbufs and mbufs >> with external memory buffers (ie. vhost zero copy). >> >> Previous discussion: >> http://dpdk.org/ml/archives/dev/2014-October/007127.html >> >> Currently for mbufs with refcnt, we cannot free mbufs with external memory >> buffers (ie. vhost zero copy), as they are recognized as indirect >> attached mbufs and therefore we free the direct mbuf it points to, >> resulting in an error in the case of external memory buffers. >> >> We solve the issue by introducing the IND_ATTACHED_MBUF flag, which indicates >> that the mbuf is an indirect attached mbuf pointing to another mbuf. >> When we free an mbuf, we only free the direct mbuf if the flag is set. >> Freeing an mbuf with external buffer is the same as freeing a non attached mbuf. >> The flag is set during attach and clear on detach. >> >> So in the case of vhost zero copy where we have mbufs with external >> buffers, by default we just free the mbuf and it is up to the user to deal with >> the external buffer. >> >> Sergio Gonzalez Monroy (2): >> mbuf: Introduce IND_ATTACHED_MBUF flag >> Remove RTE_MBUF_REFCNT references >> >> app/test/test_link_bonding.c | 15 ----------- >> app/test/test_mbuf.c | 17 +++---------- >> config/common_bsdapp | 1 - >> config/common_linuxapp | 1 - >> examples/Makefile | 4 +-- >> examples/ip_fragmentation/Makefile | 4 --- >> examples/ip_pipeline/Makefile | 3 --- >> examples/ip_pipeline/main.c | 5 ---- >> examples/ipv4_multicast/Makefile | 4 --- >> examples/vhost/main.c | 19 +++----------- >> lib/librte_ip_frag/Makefile | 4 --- >> lib/librte_ip_frag/rte_ip_frag.h | 4 --- >> lib/librte_mbuf/rte_mbuf.c | 2 -- >> lib/librte_mbuf/rte_mbuf.h | 45 +++++++-------------------------- >> lib/librte_pmd_bond/Makefile | 4 --- >> lib/librte_pmd_bond/rte_eth_bond.h | 2 -- >> lib/librte_pmd_bond/rte_eth_bond_args.c | 2 -- >> lib/librte_pmd_bond/rte_eth_bond_pmd.c | 10 -------- >> lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c | 8 ------ >> lib/librte_port/Makefile | 4 --- >> 20 files changed, 19 insertions(+), 139 deletions(-) >> > What about supporting a clone operation instead of and in addition > to attach? The refcnt is also useful when there are two paths for > a packet (going into mulitple rings). There is already an rte_pktmbuf_clone function, not sure if that is what you are asking. Sergio