Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758151AbYHVSeD (ORCPT ); Fri, 22 Aug 2008 14:34:03 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1758193AbYHVSdn (ORCPT ); Fri, 22 Aug 2008 14:33:43 -0400 Received: from stargate.chelsio.com ([12.22.49.110]:23946 "EHLO stargate.asicdesigners.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1758051AbYHVSda (ORCPT ); Fri, 22 Aug 2008 14:33:30 -0400 Date: Fri, 22 Aug 2008 11:40:08 -0700 From: Karen Xie Message-Id: <200808221840.m7MIe83k004429@localhost.localdomain> To: netdev@vger.kernel.org, open-iscsi@googlegroups.com, linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org Subject: [PATCH 3/4 2.6.28] open-iscsi - support for digest offload and payload DDP Cc: jgarzik@pobox.com, davem@davemloft.net, michaelc@cs.wisc.edu, swise@opengridcomputing.com, rdreier@cisco.com, daisyc@us.ibm.com, wenxiong@us.ibm.com, bhua@us.ibm.com, divy@chelsio.com, dm@chelsio.com, leedom@chelsio.com, kxie@chelsio.com Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 17386 Lines: 504 [PATCH 3/4 2.6.28] open-iscsi - support for digest offload and payload DDP. From: Karen Xie Added PDU digest offload and payload direct-placement support in open-iscsi. Signed-off-by: Karen Xie --- drivers/scsi/iscsi_tcp.c | 61 ++++++++++++++++++++++++----------- drivers/scsi/iscsi_tcp.h | 14 ++++++++ drivers/scsi/libiscsi.c | 51 +++++++++++++++++++++++------ include/scsi/iscsi_if.h | 1 + include/scsi/libiscsi.h | 2 + include/scsi/scsi_transport_iscsi.h | 9 +++++ 6 files changed, 106 insertions(+), 32 deletions(-) diff --git a/drivers/scsi/iscsi_tcp.c b/drivers/scsi/iscsi_tcp.c index 2a2f009..915416c 100644 --- a/drivers/scsi/iscsi_tcp.c +++ b/drivers/scsi/iscsi_tcp.c @@ -97,7 +97,7 @@ static int iscsi_tcp_hdr_recv_done(struct iscsi_tcp_conn *tcp_conn, * data is copied to the indicated sg entry, at the given * offset. */ -static inline void +void iscsi_tcp_segment_init_sg(struct iscsi_segment *segment, struct scatterlist *sg, unsigned int offset) { @@ -107,6 +107,7 @@ iscsi_tcp_segment_init_sg(struct iscsi_segment *segment, segment->total_size - segment->total_copied); segment->data = NULL; } +EXPORT_SYMBOL_GPL(iscsi_tcp_segment_init_sg); /** * iscsi_tcp_segment_map - map the current S/G page @@ -117,7 +118,7 @@ iscsi_tcp_segment_init_sg(struct iscsi_segment *segment, * because the iscsi passthrough and internal IO paths will never use high * mem pages. */ -static inline void +void iscsi_tcp_segment_map(struct iscsi_segment *segment, int recv) { struct scatterlist *sg; @@ -143,8 +144,9 @@ iscsi_tcp_segment_map(struct iscsi_segment *segment, int recv) segment->sg_mapped = kmap_atomic(sg_page(sg), KM_SOFTIRQ0); segment->data = segment->sg_mapped + sg->offset + segment->sg_offset; } +EXPORT_SYMBOL_GPL(iscsi_tcp_segment_map); -static inline void +void iscsi_tcp_segment_unmap(struct iscsi_segment *segment) { debug_tcp("iscsi_tcp_segment_unmap %p\n", segment); @@ -156,6 +158,7 @@ iscsi_tcp_segment_unmap(struct iscsi_segment *segment) segment->data = NULL; } } +EXPORT_SYMBOL_GPL(iscsi_tcp_segment_unmap); /* * Splice the digest buffer into the buffer @@ -376,6 +379,8 @@ static inline int iscsi_tcp_dgst_verify(struct iscsi_tcp_conn *tcp_conn, struct iscsi_segment *segment) { + if (tcp_conn->iscsi_conn->session->tt->caps & CAP_DIGEST_OFFLOAD) + return ((segment->status & ISCSI_SEGMENT_DGST_ERR) ? 0 : 1); if (!segment->digest_len) return 1; @@ -448,7 +453,7 @@ iscsi_segment_seek_sg(struct iscsi_segment *segment, * function is called we do not yet know the final size of the header and want * to delay the digest processing until we know that. */ -static void +void iscsi_tcp_hdr_recv_prep(struct iscsi_tcp_conn *tcp_conn) { debug_tcp("iscsi_tcp_hdr_recv_prep(%p%s)\n", tcp_conn, @@ -457,6 +462,7 @@ iscsi_tcp_hdr_recv_prep(struct iscsi_tcp_conn *tcp_conn) tcp_conn->in.hdr_buf, sizeof(struct iscsi_hdr), iscsi_tcp_hdr_recv_done, NULL); } +EXPORT_SYMBOL_GPL(iscsi_tcp_hdr_recv_prep); /* * Handle incoming reply to any other type of command @@ -486,7 +492,8 @@ iscsi_tcp_data_recv_prep(struct iscsi_tcp_conn *tcp_conn) struct iscsi_conn *conn = tcp_conn->iscsi_conn; struct hash_desc *rx_hash = NULL; - if (conn->datadgst_en) + if (conn->datadgst_en && + !(conn->session->tt->caps & CAP_DIGEST_OFFLOAD)) rx_hash = &tcp_conn->rx_hash; iscsi_segment_init_linear(&tcp_conn->in.segment, @@ -497,7 +504,7 @@ iscsi_tcp_data_recv_prep(struct iscsi_tcp_conn *tcp_conn) /* * must be called with session lock */ -static void +void iscsi_tcp_cleanup_task(struct iscsi_conn *conn, struct iscsi_task *task) { struct iscsi_tcp_task *tcp_task = task->dd_data; @@ -521,6 +528,7 @@ iscsi_tcp_cleanup_task(struct iscsi_conn *conn, struct iscsi_task *task) tcp_task->r2t = NULL; } } +EXPORT_SYMBOL_GPL(iscsi_tcp_cleanup_task); /** * iscsi_data_rsp - SCSI Data-In Response processing @@ -737,7 +745,7 @@ iscsi_tcp_process_data_in(struct iscsi_tcp_conn *tcp_conn, * by data, the receive buffer is set up to copy the incoming data * to the correct location. */ -static int +int iscsi_tcp_hdr_dissect(struct iscsi_conn *conn, struct iscsi_hdr *hdr) { int rc = 0, opcode, ahslen; @@ -793,7 +801,8 @@ iscsi_tcp_hdr_dissect(struct iscsi_conn *conn, struct iscsi_hdr *hdr) * we move on to the next scatterlist entry and * update the digest per-entry. */ - if (conn->datadgst_en) + if (conn->datadgst_en && + !(conn->session->tt->caps & CAP_DIGEST_OFFLOAD)) rx_hash = &tcp_conn->rx_hash; debug_tcp("iscsi_tcp_begin_data_in(%p, offset=%d, " @@ -881,6 +890,7 @@ iscsi_tcp_hdr_dissect(struct iscsi_conn *conn, struct iscsi_hdr *hdr) return rc; } +EXPORT_SYMBOL_GPL(iscsi_tcp_hdr_dissect); /** * iscsi_tcp_hdr_recv_done - process PDU header @@ -919,7 +929,8 @@ iscsi_tcp_hdr_recv_done(struct iscsi_tcp_conn *tcp_conn, /* We're done processing the header. See if we're doing * header digests; if so, set up the recv_digest buffer * and go back for more. */ - if (conn->hdrdgst_en) { + if (conn->hdrdgst_en && + !(conn->session->tt->caps & CAP_DIGEST_OFFLOAD)) { if (segment->digest_len == 0) { iscsi_tcp_segment_splice_digest(segment, segment->recv_digest); @@ -944,7 +955,8 @@ iscsi_tcp_hdr_recv_done(struct iscsi_tcp_conn *tcp_conn, * @offset: offset in skb * @len: skb->len - offset **/ -static int + +int iscsi_tcp_recv(read_descriptor_t *rd_desc, struct sk_buff *skb, unsigned int offset, size_t len) { @@ -1158,7 +1170,7 @@ iscsi_tcp_xmit_qlen(struct iscsi_conn *conn) return segment->total_copied - segment->total_size; } -static inline int +static int iscsi_tcp_flush(struct iscsi_conn *conn) { int rc; @@ -1205,7 +1217,8 @@ iscsi_tcp_send_hdr_prep(struct iscsi_conn *conn, void *hdr, size_t hdrlen) * sure that both iscsi_tcp_task and mtask have * sufficient room. */ - if (conn->hdrdgst_en) { + if (conn->hdrdgst_en && + !(conn->session->tt->caps & CAP_DIGEST_OFFLOAD)) { iscsi_tcp_dgst_header(&tcp_conn->tx_hash, hdr, hdrlen, hdr + hdrlen); hdrlen += ISCSI_DIGEST_SIZE; @@ -1243,7 +1256,8 @@ iscsi_tcp_send_data_prep(struct iscsi_conn *conn, struct scatterlist *sg, hdr_spec_len = ntoh24(tcp_conn->out.hdr->dlength); WARN_ON(iscsi_padded(len) != iscsi_padded(hdr_spec_len)); - if (conn->datadgst_en) + if (conn->datadgst_en && + !(conn->session->tt->caps & CAP_DIGEST_OFFLOAD)) tx_hash = &tcp_conn->tx_hash; return iscsi_segment_seek_sg(&tcp_conn->out.data_segment, @@ -1267,7 +1281,8 @@ iscsi_tcp_send_linear_data_prepare(struct iscsi_conn *conn, void *data, hdr_spec_len = ntoh24(tcp_conn->out.hdr->dlength); WARN_ON(iscsi_padded(len) != iscsi_padded(hdr_spec_len)); - if (conn->datadgst_en) + if (conn->datadgst_en && + !(conn->session->tt->caps & CAP_DIGEST_OFFLOAD)) tx_hash = &tcp_conn->tx_hash; iscsi_segment_init_linear(&tcp_conn->out.data_segment, @@ -1329,7 +1344,7 @@ iscsi_solicit_data_cont(struct iscsi_conn *conn, struct iscsi_task *task, * @task: scsi command task * @sc: scsi command **/ -static int +int iscsi_tcp_task_init(struct iscsi_task *task) { struct iscsi_tcp_task *tcp_task = task->dd_data; @@ -1378,6 +1393,7 @@ iscsi_tcp_task_init(struct iscsi_task *task) task->imm_count = 0; return 0; } +EXPORT_SYMBOL_GPL(iscsi_tcp_task_init); /* * iscsi_tcp_task_xmit - xmit normal PDU task @@ -1387,7 +1403,7 @@ iscsi_tcp_task_init(struct iscsi_task *task) * -EAGAIN if there's still data in the queue, or != 0 for any other kind * of error. */ -static int +int iscsi_tcp_task_xmit(struct iscsi_task *task) { struct iscsi_conn *conn = task->conn; @@ -1398,7 +1414,7 @@ iscsi_tcp_task_xmit(struct iscsi_task *task) flush: /* Flush any pending data first. */ - rc = iscsi_tcp_flush(conn); + rc = conn->session->tt->flush_conn(conn); if (rc < 0) return rc; @@ -1490,6 +1506,7 @@ fail: iscsi_conn_failure(conn, rc); return -EIO; } +EXPORT_SYMBOL_GPL(iscsi_tcp_task_xmit); static struct iscsi_cls_conn * iscsi_tcp_conn_create(struct iscsi_cls_session *cls_session, uint32_t conn_idx) @@ -1696,7 +1713,7 @@ free_socket: return err; } -static int +int iscsi_r2tpool_alloc(struct iscsi_session *session) { int i; @@ -1742,8 +1759,9 @@ r2t_alloc_fail: } return -ENOMEM; } +EXPORT_SYMBOL_GPL(iscsi_r2tpool_alloc); -static void +void iscsi_r2tpool_free(struct iscsi_session *session) { int i; @@ -1756,6 +1774,7 @@ iscsi_r2tpool_free(struct iscsi_session *session) iscsi_pool_free(&tcp_task->r2tpool); } } +EXPORT_SYMBOL_GPL(iscsi_r2tpool_free); static int iscsi_conn_set_param(struct iscsi_cls_conn *cls_conn, enum iscsi_param param, @@ -1904,6 +1923,7 @@ static void iscsi_tcp_session_destroy(struct iscsi_cls_session *cls_session) struct Scsi_Host *shost = iscsi_session_to_shost(cls_session); iscsi_r2tpool_free(cls_session->dd_data); + iscsi_session_teardown(cls_session); iscsi_host_remove(shost); iscsi_host_free(shost); @@ -1927,7 +1947,7 @@ static struct scsi_host_template iscsi_sht = { .cmd_per_lun = ISCSI_DEF_CMD_PER_LUN, .eh_abort_handler = iscsi_eh_abort, .eh_device_reset_handler= iscsi_eh_device_reset, - .eh_host_reset_handler = iscsi_eh_host_reset, + .eh_target_reset_handler= iscsi_eh_target_reset, .use_clustering = DISABLE_CLUSTERING, .slave_configure = iscsi_tcp_slave_configure, .proc_name = "iscsi_tcp", @@ -1978,6 +1998,7 @@ static struct iscsi_transport iscsi_tcp_transport = { .get_session_param = iscsi_session_get_param, .start_conn = iscsi_conn_start, .stop_conn = iscsi_tcp_conn_stop, + .flush_conn = iscsi_tcp_flush, /* iscsi host params */ .get_host_param = iscsi_host_get_param, .set_host_param = iscsi_host_set_param, diff --git a/drivers/scsi/iscsi_tcp.h b/drivers/scsi/iscsi_tcp.h index 498d8ca..62747c7 100644 --- a/drivers/scsi/iscsi_tcp.h +++ b/drivers/scsi/iscsi_tcp.h @@ -32,7 +32,9 @@ struct iscsi_segment; typedef int iscsi_segment_done_fn_t(struct iscsi_tcp_conn *, struct iscsi_segment *); +#define ISCSI_SEGMENT_DGST_ERR 0x1 struct iscsi_segment { + unsigned int status; unsigned char *data; unsigned int size; unsigned int copied; @@ -130,4 +132,16 @@ struct iscsi_tcp_task { struct iscsi_data_task unsol_dtask; /* Data-Out header buf */ }; +void iscsi_tcp_segment_init_sg(struct iscsi_segment *, struct scatterlist *, + unsigned int); +void iscsi_tcp_segment_map(struct iscsi_segment *, int); +void iscsi_tcp_segment_unmap(struct iscsi_segment *); +void iscsi_tcp_hdr_recv_prep(struct iscsi_tcp_conn *); +void iscsi_tcp_cleanup_task(struct iscsi_conn *, struct iscsi_task *); +int iscsi_tcp_hdr_dissect(struct iscsi_conn *conn, struct iscsi_hdr *); +int iscsi_tcp_task_init(struct iscsi_task *); +int iscsi_tcp_task_xmit(struct iscsi_task *); +int iscsi_r2tpool_alloc(struct iscsi_session *); +void iscsi_r2tpool_free(struct iscsi_session *); + #endif /* ISCSI_H */ diff --git a/drivers/scsi/libiscsi.c b/drivers/scsi/libiscsi.c index 299e075..0504aa1 100644 --- a/drivers/scsi/libiscsi.c +++ b/drivers/scsi/libiscsi.c @@ -218,7 +218,12 @@ static int iscsi_prep_scsi_cmd_pdu(struct iscsi_task *task) hdr->opcode = ISCSI_OP_SCSI_CMD; hdr->flags = ISCSI_ATTR_SIMPLE; int_to_scsilun(sc->device->lun, (struct scsi_lun *)hdr->lun); - hdr->itt = build_itt(task->itt, session->age); + if (session->tt->reserve_itt) { + rc = session->tt->reserve_itt(task, &hdr->itt); + if (rc) + return rc; + } else + hdr->itt = build_itt(task->itt, session->age); hdr->cmdsn = cpu_to_be32(session->cmdsn); session->cmdsn++; hdr->exp_statsn = cpu_to_be32(conn->exp_statsn); @@ -332,6 +337,9 @@ static void iscsi_complete_command(struct iscsi_task *task) struct iscsi_session *session = conn->session; struct scsi_cmnd *sc = task->sc; + if (session->tt->release_itt) + session->tt->release_itt(task, task->hdr->itt); + list_del_init(&task->running); task->state = ISCSI_TASK_COMPLETED; task->sc = NULL; @@ -442,7 +450,12 @@ static int iscsi_prep_mgmt_task(struct iscsi_conn *conn, */ nop->cmdsn = cpu_to_be32(session->cmdsn); if (hdr->itt != RESERVED_ITT) { - hdr->itt = build_itt(task->itt, session->age); + if (session->tt->reserve_itt) { + int rc = session->tt->reserve_itt(task, &hdr->itt); + if (rc) + return rc; + } else + hdr->itt = build_itt(task->itt, session->age); /* * TODO: We always use immediate, so we never hit this. * If we start to send tmfs or nops as non-immediate then @@ -691,7 +704,12 @@ static int iscsi_handle_reject(struct iscsi_conn *conn, struct iscsi_hdr *hdr, if (ntoh24(reject->dlength) >= sizeof(struct iscsi_hdr)) { memcpy(&rejected_pdu, data, sizeof(struct iscsi_hdr)); - itt = get_itt(rejected_pdu.itt); + if (conn->session->tt->parse_itt) { + conn->session->tt->parse_itt(conn, + rejected_pdu.itt, + &itt, NULL); + } else + itt = get_itt(rejected_pdu.itt); iscsi_conn_printk(KERN_ERR, conn, "itt 0x%x had pdu (op 0x%x) rejected " "due to DataDigest error.\n", itt, @@ -719,7 +737,10 @@ static struct iscsi_task *iscsi_itt_to_task(struct iscsi_conn *conn, itt_t itt) if (itt == RESERVED_ITT) return NULL; - i = get_itt(itt); + if (conn->session->tt->parse_itt) + conn->session->tt->parse_itt(conn, itt, &i, NULL); + else + i = get_itt(itt); if (i >= session->cmds_max) return NULL; @@ -750,9 +771,12 @@ int __iscsi_complete_pdu(struct iscsi_conn *conn, struct iscsi_hdr *hdr, if (rc) return rc; - if (hdr->itt != RESERVED_ITT) - itt = get_itt(hdr->itt); - else + if (hdr->itt != RESERVED_ITT) { + if (conn->session->tt->parse_itt) + conn->session->tt->parse_itt(conn, hdr->itt, &itt, NULL); + else + itt = get_itt(hdr->itt); + } else itt = ~0U; debug_scsi("[op 0x%x cid %d itt 0x%x len %d]\n", @@ -899,20 +923,25 @@ EXPORT_SYMBOL_GPL(iscsi_complete_pdu); int iscsi_verify_itt(struct iscsi_conn *conn, itt_t itt) { struct iscsi_session *session = conn->session; - uint32_t i; + uint32_t i, age; if (itt == RESERVED_ITT) return 0; - if (((__force u32)itt & ISCSI_AGE_MASK) != - (session->age << ISCSI_AGE_SHIFT)) { + if (conn->session->tt->parse_itt) + conn->session->tt->parse_itt(conn, hdr->itt, &i, &age); + else { + i = get_itt(itt); + age = ((__force u32)itt >> ISCSI_AGE_SHIFT) & ISCSI_AGE_MASK; + } + + if (age != session->age) { iscsi_conn_printk(KERN_ERR, conn, "received itt %x expected session age (%x)\n", (__force u32)itt, session->age); return ISCSI_ERR_BAD_ITT; } - i = get_itt(itt); if (i >= session->cmds_max) { iscsi_conn_printk(KERN_ERR, conn, "received invalid itt index %u (max cmds " diff --git a/include/scsi/iscsi_if.h b/include/scsi/iscsi_if.h index 16be12f..75afc58 100644 --- a/include/scsi/iscsi_if.h +++ b/include/scsi/iscsi_if.h @@ -333,6 +333,7 @@ enum iscsi_host_param { #define CAP_FW_DB 0x200 #define CAP_SENDTARGETS_OFFLOAD 0x400 #define CAP_DATA_PATH_OFFLOAD 0x800 +#define CAP_DIGEST_OFFLOAD 0x1000 /* * These flags describes reason of stop_conn() call diff --git a/include/scsi/libiscsi.h b/include/scsi/libiscsi.h index 5e75bb7..31f916f 100644 --- a/include/scsi/libiscsi.h +++ b/include/scsi/libiscsi.h @@ -75,7 +75,7 @@ enum { /* this must be a power of two greater than ISCSI_MGMT_CMDS_MAX */ #define ISCSI_TOTAL_CMDS_MIN 16 #define ISCSI_AGE_SHIFT 28 -#define ISCSI_AGE_MASK (0xf << ISCSI_AGE_SHIFT) +#define ISCSI_AGE_MASK 0xf #define ISCSI_ADDRESS_BUF_LEN 64 diff --git a/include/scsi/scsi_transport_iscsi.h b/include/scsi/scsi_transport_iscsi.h index 8b6c91d..76337bc 100644 --- a/include/scsi/scsi_transport_iscsi.h +++ b/include/scsi/scsi_transport_iscsi.h @@ -56,6 +56,10 @@ struct sockaddr; * is not supported, and a -Exx value on other error * @start_conn: set connection to be operational * @stop_conn: suspend/recover/terminate connection + * @flush_conn: flush a connection's transmit queue + * @parse_itt: parse the itt rcv'ed in BHS + * @reserve_itt: construct a task itt to be sent in BHS + * @release_itt: release a itt (constructed by reserve_itt) * @send_pdu: send iSCSI PDU, Login, Logout, NOP-Out, Reject, Text. * @session_recovery_timedout: notify LLD a block during recovery timed out * @init_task: Initialize a iscsi_task and any internal structs. @@ -113,6 +117,11 @@ struct iscsi_transport { char *data, uint32_t data_size); void (*get_stats) (struct iscsi_cls_conn *conn, struct iscsi_stats *stats); + int (*flush_conn) (struct iscsi_conn *conn); + void (*parse_itt)(struct iscsi_conn *conn, itt_t hdr_itt, + int *idx, int *age); + int (*reserve_itt)(struct iscsi_task *task, itt_t *hdr_itt); + void (*release_itt)(struct iscsi_task *task, itt_t hdr_itt); int (*init_task) (struct iscsi_task *task); int (*xmit_task) (struct iscsi_task *task); void (*cleanup_task) (struct iscsi_conn *conn, -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/