From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ferruh Yigit Subject: Re: [PATCHv6 22/33] net/dpaa2: add support for l3 and l4 checksum offload Date: Tue, 24 Jan 2017 10:51:40 +0000 Message-ID: References: <1484832240-2048-1-git-send-email-hemant.agrawal@nxp.com> <1485172803-17288-1-git-send-email-hemant.agrawal@nxp.com> <1485172803-17288-23-git-send-email-hemant.agrawal@nxp.com> <21d59f80-2d4f-65a6-657b-cbcf0d2c1a08@intel.com> <8266122a-5a5e-a8f4-09fb-7199fed540a0@nxp.com> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Cc: thomas.monjalon@6wind.com, bruce.richardson@intel.com, shreyansh.jain@nxp.com, john.mcnamara@intel.com, jerin.jacob@caviumnetworks.com To: Hemant Agrawal , dev@dpdk.org Return-path: Received: from mga04.intel.com (mga04.intel.com [192.55.52.120]) by dpdk.org (Postfix) with ESMTP id 568D15599 for ; Tue, 24 Jan 2017 11:51:43 +0100 (CET) In-Reply-To: <8266122a-5a5e-a8f4-09fb-7199fed540a0@nxp.com> List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" On 1/24/2017 10:45 AM, Hemant Agrawal wrote: > On 1/23/2017 11:05 PM, Ferruh Yigit wrote: >> On 1/23/2017 11:59 AM, Hemant Agrawal wrote: >>> Signed-off-by: Hemant Agrawal >>> --- >> <...> >>> --- a/drivers/net/dpaa2/Makefile >>> +++ b/drivers/net/dpaa2/Makefile >>> @@ -66,6 +66,6 @@ DEPDIRS-$(CONFIG_RTE_LIBRTE_DPAA2_PMD) += lib/librte_eal lib/librte_ether >>> DEPDIRS-$(CONFIG_RTE_LIBRTE_DPAA2_PMD) += lib/librte_mempool lib/librte_mbuf >>> DEPDIRS-$(CONFIG_RTE_LIBRTE_DPAA2_PMD) += lib/librte_common_dpaa2_qbman >>> DEPDIRS-$(CONFIG_RTE_LIBRTE_DPAA2_PMD) += lib/librte_bus_fslmc >>> -DEPDIRS-$(CONFIG_RTE_LIBRTE_DPAA2_PMD) += lib/librte_pmd_dpaa2_pool >>> +DEPDIRS-$(CONFIG_RTE_LIBRTE_DPAA2_PMD) += lib/librte_pool_dpaa2 >> >> I understand these are added since DEPDIRS converted to LDLIBS, but this >> does not looks right, since these variables mainly dependency solving >> and the value provided is not correct for DEPDIRS. >> >> Did you tried adding as "LDLIBS += xx", not tested, but may work. >> > > Our intention was to create dependency to help in parallel build in case > of shared library. "LDLIBS += xx" also work, but we are still not able > to create inter driver directory dependency for parallel build. > > > We may need your help in understanding the change required in depdir > script to support directories within "driver". > >>> >>> include $(RTE_SDK)/mk/rte.lib.mk >>> diff --git a/drivers/net/dpaa2/dpaa2_ethdev.c b/drivers/net/dpaa2/dpaa2_ethdev.c >>> index 6de571a..2298246 100644 >>> --- a/drivers/net/dpaa2/dpaa2_ethdev.c >>> +++ b/drivers/net/dpaa2/dpaa2_ethdev.c >>> @@ -68,7 +68,17 @@ >>> dev_info->min_rx_bufsize = DPAA2_MIN_RX_BUF_SIZE; >>> dev_info->max_rx_queues = (uint16_t)priv->nb_rx_queues; >>> dev_info->max_tx_queues = (uint16_t)priv->nb_tx_queues; >>> - >>> + dev_info->rx_offload_capa = >>> + DEV_RX_OFFLOAD_IPV4_CKSUM | >>> + DEV_RX_OFFLOAD_UDP_CKSUM | >>> + DEV_RX_OFFLOAD_TCP_CKSUM | >>> + DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM; >> >> Is there capabilities always enabled for all devices this driver >> support? Or should these be read from some device registers? >> > Capabilities are always enabled for all devices (MC DPNI object in this > case) Thanks for the clarification, is it same for the speed capabilities? > >>> + dev_info->tx_offload_capa = >>> + DEV_TX_OFFLOAD_IPV4_CKSUM | >>> + DEV_TX_OFFLOAD_UDP_CKSUM | >>> + DEV_TX_OFFLOAD_TCP_CKSUM | >>> + DEV_TX_OFFLOAD_SCTP_CKSUM | >>> + DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM; >>> dev_info->speed_capa = ETH_LINK_SPEED_1G | >>> ETH_LINK_SPEED_2_5G | >>> ETH_LINK_SPEED_10G; >> <...> >> > >