Hi folks, [Topic] iSER Target implementation for upstream in v3.10 timeframe [Abstract] I'm currently working on a kernel level iSCSI Extensions for RDMA (iSER) target driver using OFA Verbs for Infiniband + Ethernet based transports. This naturally involves quite a bit of re-factoring + creation of a iscsit_transport template in iscsi-target code to utilize existing login + authentication path processing, incoming RX PDU handling, outgoing TX PDU encoding, and other logic that ends up being independent of underlying iSCSI/TCP + iSER/RCaP transport usage. Currently this is implemented as a secondary level fabric driver that lives (transparently) under the top-level iscsi-target configfs group in /sys/kernel/config/target/iscsi/, and is enabled + disabled on a per IPv4/IPv6 network portal context basis using a 'iser' attribute. I'll be posting the initial RFC code for review in the upcoming weeks, and at LSF would like to discuss potential performance improvements (namely pre-allocation) related to ib_isert that can be extended to other fabric drivers, along with modernizations (namely se_cmd->cmd_kref usage) for existing traditional iscsi-target code. Thank you, --nab -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
