diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2015-11-07 13:33:07 -0800 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2015-11-07 13:33:07 -0800 |
commit | ab9f2faf8f40604551336e5b0a18e0910a57b92c (patch) | |
tree | 9068c73acf24452762d6e2b096df19e29436183e /drivers/infiniband/ulp | |
parent | 75021d28594d9b6fb4d05bbc41f77948a0db0e02 (diff) | |
parent | db7489e07669073970358b6cacf6a9dd8dc9275e (diff) | |
download | op-kernel-dev-ab9f2faf8f40604551336e5b0a18e0910a57b92c.zip op-kernel-dev-ab9f2faf8f40604551336e5b0a18e0910a57b92c.tar.gz |
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma
Pull rdma updates from Doug Ledford:
"This is my initial round of 4.4 merge window patches. There are a few
other things I wish to get in for 4.4 that aren't in this pull, as
this represents what has gone through merge/build/run testing and not
what is the last few items for which testing is not yet complete.
- "Checksum offload support in user space" enablement
- Misc cxgb4 fixes, add T6 support
- Misc usnic fixes
- 32 bit build warning fixes
- Misc ocrdma fixes
- Multicast loopback prevention extension
- Extend the GID cache to store and return attributes of GIDs
- Misc iSER updates
- iSER clustering update
- Network NameSpace support for rdma CM
- Work Request cleanup series
- New Memory Registration API"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma: (76 commits)
IB/core, cma: Make __attribute_const__ declarations sparse-friendly
IB/core: Remove old fast registration API
IB/ipath: Remove fast registration from the code
IB/hfi1: Remove fast registration from the code
RDMA/nes: Remove old FRWR API
IB/qib: Remove old FRWR API
iw_cxgb4: Remove old FRWR API
RDMA/cxgb3: Remove old FRWR API
RDMA/ocrdma: Remove old FRWR API
IB/mlx4: Remove old FRWR API support
IB/mlx5: Remove old FRWR API support
IB/srp: Dont allocate a page vector when using fast_reg
IB/srp: Remove srp_finish_mapping
IB/srp: Convert to new registration API
IB/srp: Split srp_map_sg
RDS/IW: Convert to new memory registration API
svcrdma: Port to new memory registration API
xprtrdma: Port to new memory registration API
iser-target: Port to new memory registration API
IB/iser: Port to new fast registration API
...
Diffstat (limited to 'drivers/infiniband/ulp')
-rw-r--r-- | drivers/infiniband/ulp/ipoib/ipoib.h | 4 | ||||
-rw-r--r-- | drivers/infiniband/ulp/ipoib/ipoib_cm.c | 4 | ||||
-rw-r--r-- | drivers/infiniband/ulp/ipoib/ipoib_ib.c | 22 | ||||
-rw-r--r-- | drivers/infiniband/ulp/ipoib/ipoib_main.c | 4 | ||||
-rw-r--r-- | drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 4 | ||||
-rw-r--r-- | drivers/infiniband/ulp/ipoib/ipoib_verbs.c | 6 | ||||
-rw-r--r-- | drivers/infiniband/ulp/iser/iscsi_iser.c | 20 | ||||
-rw-r--r-- | drivers/infiniband/ulp/iser/iscsi_iser.h | 25 | ||||
-rw-r--r-- | drivers/infiniband/ulp/iser/iser_initiator.c | 51 | ||||
-rw-r--r-- | drivers/infiniband/ulp/iser/iser_memory.c | 370 | ||||
-rw-r--r-- | drivers/infiniband/ulp/iser/iser_verbs.c | 20 | ||||
-rw-r--r-- | drivers/infiniband/ulp/isert/ib_isert.c | 269 | ||||
-rw-r--r-- | drivers/infiniband/ulp/isert/ib_isert.h | 8 | ||||
-rw-r--r-- | drivers/infiniband/ulp/srp/ib_srp.c | 262 | ||||
-rw-r--r-- | drivers/infiniband/ulp/srp/ib_srp.h | 11 | ||||
-rw-r--r-- | drivers/infiniband/ulp/srpt/ib_srpt.c | 35 |
16 files changed, 378 insertions, 737 deletions
diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h index edc5b85..3ede103 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib.h +++ b/drivers/infiniband/ulp/ipoib/ipoib.h @@ -360,7 +360,7 @@ struct ipoib_dev_priv { unsigned tx_head; unsigned tx_tail; struct ib_sge tx_sge[MAX_SKB_FRAGS + 1]; - struct ib_send_wr tx_wr; + struct ib_ud_wr tx_wr; unsigned tx_outstanding; struct ib_wc send_wc[MAX_SEND_CQE]; @@ -528,7 +528,7 @@ static inline void ipoib_build_sge(struct ipoib_dev_priv *priv, priv->tx_sge[i + off].addr = mapping[i + off]; priv->tx_sge[i + off].length = skb_frag_size(&frags[i]); } - priv->tx_wr.num_sge = nr_frags + off; + priv->tx_wr.wr.num_sge = nr_frags + off; } #ifdef CONFIG_INFINIBAND_IPOIB_DEBUG diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index c78dc16..3ae9726 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -700,9 +700,9 @@ static inline int post_send(struct ipoib_dev_priv *priv, ipoib_build_sge(priv, tx_req); - priv->tx_wr.wr_id = wr_id | IPOIB_OP_CM; + priv->tx_wr.wr.wr_id = wr_id | IPOIB_OP_CM; - return ib_post_send(tx->qp, &priv->tx_wr, &bad_wr); + return ib_post_send(tx->qp, &priv->tx_wr.wr, &bad_wr); } void ipoib_cm_send(struct net_device *dev, struct sk_buff *skb, struct ipoib_cm_tx *tx) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c index d266667..5ea0c14 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c @@ -518,19 +518,19 @@ static inline int post_send(struct ipoib_dev_priv *priv, ipoib_build_sge(priv, tx_req); - priv->tx_wr.wr_id = wr_id; - priv->tx_wr.wr.ud.remote_qpn = qpn; - priv->tx_wr.wr.ud.ah = address; + priv->tx_wr.wr.wr_id = wr_id; + priv->tx_wr.remote_qpn = qpn; + priv->tx_wr.ah = address; if (head) { - priv->tx_wr.wr.ud.mss = skb_shinfo(skb)->gso_size; - priv->tx_wr.wr.ud.header = head; - priv->tx_wr.wr.ud.hlen = hlen; - priv->tx_wr.opcode = IB_WR_LSO; + priv->tx_wr.mss = skb_shinfo(skb)->gso_size; + priv->tx_wr.header = head; + priv->tx_wr.hlen = hlen; + priv->tx_wr.wr.opcode = IB_WR_LSO; } else - priv->tx_wr.opcode = IB_WR_SEND; + priv->tx_wr.wr.opcode = IB_WR_SEND; - return ib_post_send(priv->qp, &priv->tx_wr, &bad_wr); + return ib_post_send(priv->qp, &priv->tx_wr.wr, &bad_wr); } void ipoib_send(struct net_device *dev, struct sk_buff *skb, @@ -583,9 +583,9 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb, } if (skb->ip_summed == CHECKSUM_PARTIAL) - priv->tx_wr.send_flags |= IB_SEND_IP_CSUM; + priv->tx_wr.wr.send_flags |= IB_SEND_IP_CSUM; else - priv->tx_wr.send_flags &= ~IB_SEND_IP_CSUM; + priv->tx_wr.wr.send_flags &= ~IB_SEND_IP_CSUM; if (++priv->tx_outstanding == ipoib_sendq_size) { ipoib_dbg(priv, "TX ring full, stopping kernel net queue\n"); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index babba05..7d32818 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -461,7 +461,7 @@ int ipoib_set_mode(struct net_device *dev, const char *buf) netdev_update_features(dev); dev_set_mtu(dev, ipoib_cm_max_mtu(dev)); rtnl_unlock(); - priv->tx_wr.send_flags &= ~IB_SEND_IP_CSUM; + priv->tx_wr.wr.send_flags &= ~IB_SEND_IP_CSUM; ipoib_flush_paths(dev); rtnl_lock(); @@ -1860,7 +1860,7 @@ static struct net_device *ipoib_add_port(const char *format, priv->dev->broadcast[8] = priv->pkey >> 8; priv->dev->broadcast[9] = priv->pkey & 0xff; - result = ib_query_gid(hca, port, 0, &priv->local_gid); + result = ib_query_gid(hca, port, 0, &priv->local_gid, NULL); if (result) { printk(KERN_WARNING "%s: ib_query_gid port %d failed (ret = %d)\n", hca->name, port, result); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index d750a86..f357ca6 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -245,7 +245,7 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, priv->qkey = be32_to_cpu(priv->broadcast->mcmember.qkey); spin_unlock_irq(&priv->lock); - priv->tx_wr.wr.ud.remote_qkey = priv->qkey; + priv->tx_wr.remote_qkey = priv->qkey; set_qkey = 1; } @@ -561,7 +561,7 @@ void ipoib_mcast_join_task(struct work_struct *work) } priv->local_lid = port_attr.lid; - if (ib_query_gid(priv->ca, priv->port, 0, &priv->local_gid)) + if (ib_query_gid(priv->ca, priv->port, 0, &priv->local_gid, NULL)) ipoib_warn(priv, "ib_query_gid() failed\n"); else memcpy(priv->dev->dev_addr + 4, priv->local_gid.raw, sizeof (union ib_gid)); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c index 78845b6..d48c5ba 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c @@ -221,9 +221,9 @@ int ipoib_transport_dev_init(struct net_device *dev, struct ib_device *ca) for (i = 0; i < MAX_SKB_FRAGS + 1; ++i) priv->tx_sge[i].lkey = priv->pd->local_dma_lkey; - priv->tx_wr.opcode = IB_WR_SEND; - priv->tx_wr.sg_list = priv->tx_sge; - priv->tx_wr.send_flags = IB_SEND_SIGNALED; + priv->tx_wr.wr.opcode = IB_WR_SEND; + priv->tx_wr.wr.sg_list = priv->tx_sge; + priv->tx_wr.wr.send_flags = IB_SEND_SIGNALED; priv->rx_sge[0].lkey = priv->pd->local_dma_lkey; diff --git a/drivers/infiniband/ulp/iser/iscsi_iser.c b/drivers/infiniband/ulp/iser/iscsi_iser.c index f58ff96..9080161 100644 --- a/drivers/infiniband/ulp/iser/iscsi_iser.c +++ b/drivers/infiniband/ulp/iser/iscsi_iser.c @@ -111,7 +111,7 @@ module_param_named(pi_guard, iser_pi_guard, int, S_IRUGO); MODULE_PARM_DESC(pi_guard, "T10-PI guard_type [deprecated]"); /* - * iscsi_iser_recv() - Process a successfull recv completion + * iscsi_iser_recv() - Process a successful recv completion * @conn: iscsi connection * @hdr: iscsi header * @rx_data: buffer containing receive data payload @@ -126,7 +126,6 @@ iscsi_iser_recv(struct iscsi_conn *conn, struct iscsi_hdr *hdr, { int rc = 0; int datalen; - int ahslen; /* verify PDU length */ datalen = ntoh24(hdr->dlength); @@ -141,9 +140,6 @@ iscsi_iser_recv(struct iscsi_conn *conn, struct iscsi_hdr *hdr, iser_dbg("aligned datalen (%d) hdr, %d (IB)\n", datalen, rx_data_len); - /* read AHS */ - ahslen = hdr->hlength * 4; - rc = iscsi_complete_pdu(conn, hdr, rx_data, rx_data_len); if (rc && rc != ISCSI_ERR_NO_SCSI_CMD) goto error; @@ -766,9 +762,7 @@ iscsi_iser_conn_get_stats(struct iscsi_cls_conn *cls_conn, struct iscsi_stats *s stats->r2t_pdus = conn->r2t_pdus_cnt; /* always 0 */ stats->tmfcmd_pdus = conn->tmfcmd_pdus_cnt; stats->tmfrsp_pdus = conn->tmfrsp_pdus_cnt; - stats->custom_length = 1; - strcpy(stats->custom[0].desc, "fmr_unalign_cnt"); - stats->custom[0].value = conn->fmr_unalign_cnt; + stats->custom_length = 0; } static int iscsi_iser_get_ep_param(struct iscsi_endpoint *ep, @@ -973,6 +967,13 @@ static umode_t iser_attr_is_visible(int param_type, int param) return 0; } +static int iscsi_iser_slave_alloc(struct scsi_device *sdev) +{ + blk_queue_virt_boundary(sdev->request_queue, ~MASK_4K); + + return 0; +} + static struct scsi_host_template iscsi_iser_sht = { .module = THIS_MODULE, .name = "iSCSI Initiator over iSER", @@ -985,7 +986,8 @@ static struct scsi_host_template iscsi_iser_sht = { .eh_device_reset_handler= iscsi_eh_device_reset, .eh_target_reset_handler = iscsi_eh_recover_target, .target_alloc = iscsi_target_alloc, - .use_clustering = DISABLE_CLUSTERING, + .use_clustering = ENABLE_CLUSTERING, + .slave_alloc = iscsi_iser_slave_alloc, .proc_name = "iscsi_iser", .this_id = -1, .track_queue_depth = 1, diff --git a/drivers/infiniband/ulp/iser/iscsi_iser.h b/drivers/infiniband/ulp/iser/iscsi_iser.h index a5edd6e..8a5998e 100644 --- a/drivers/infiniband/ulp/iser/iscsi_iser.h +++ b/drivers/infiniband/ulp/iser/iscsi_iser.h @@ -227,18 +227,13 @@ enum iser_data_dir { * @size: num entries of this sg * @data_len: total beffer byte len * @dma_nents: returned by dma_map_sg - * @orig_sg: pointer to the original sg list (in case - * we used a copy) - * @orig_size: num entris of orig sg list */ struct iser_data_buf { struct scatterlist *sg; - unsigned int size; + int size; unsigned long data_len; unsigned int dma_nents; - struct scatterlist *orig_sg; - unsigned int orig_size; - }; +}; /* fwd declarations */ struct iser_device; @@ -300,7 +295,11 @@ struct iser_tx_desc { int num_sge; bool mapped; u8 wr_idx; - struct ib_send_wr wrs[ISER_MAX_WRS]; + union iser_wr { + struct ib_send_wr send; + struct ib_reg_wr fast_reg; + struct ib_sig_handover_wr sig; + } wrs[ISER_MAX_WRS]; struct iser_mem_reg data_reg; struct iser_mem_reg prot_reg; struct ib_sig_attrs sig_attrs; @@ -413,7 +412,6 @@ struct iser_device { * * @mr: memory region * @fmr_pool: pool of fmrs - * @frpl: fast reg page list used by frwrs * @page_vec: fast reg page list used by fmr pool * @mr_valid: is mr valid indicator */ @@ -422,10 +420,7 @@ struct iser_reg_resources { struct ib_mr *mr; struct ib_fmr_pool *fmr_pool; }; - union { - struct ib_fast_reg_page_list *frpl; - struct iser_page_vec *page_vec; - }; + struct iser_page_vec *page_vec; u8 mr_valid:1; }; @@ -712,11 +707,11 @@ iser_reg_desc_put_fmr(struct ib_conn *ib_conn, static inline struct ib_send_wr * iser_tx_next_wr(struct iser_tx_desc *tx_desc) { - struct ib_send_wr *cur_wr = &tx_desc->wrs[tx_desc->wr_idx]; + struct ib_send_wr *cur_wr = &tx_desc->wrs[tx_desc->wr_idx].send; struct ib_send_wr *last_wr; if (tx_desc->wr_idx) { - last_wr = &tx_desc->wrs[tx_desc->wr_idx - 1]; + last_wr = &tx_desc->wrs[tx_desc->wr_idx - 1].send; last_wr->next = cur_wr; } tx_desc->wr_idx++; diff --git a/drivers/infiniband/ulp/iser/iser_initiator.c b/drivers/infiniband/ulp/iser/iser_initiator.c index d511879..ffd00c4 100644 --- a/drivers/infiniband/ulp/iser/iser_initiator.c +++ b/drivers/infiniband/ulp/iser/iser_initiator.c @@ -661,48 +661,14 @@ void iser_task_rdma_init(struct iscsi_iser_task *iser_task) void iser_task_rdma_finalize(struct iscsi_iser_task *iser_task) { - int is_rdma_data_aligned = 1; - int is_rdma_prot_aligned = 1; int prot_count = scsi_prot_sg_count(iser_task->sc); - /* if we were reading, copy back to unaligned sglist, - * anyway dma_unmap and free the copy - */ - if (iser_task->data[ISER_DIR_IN].orig_sg) { - is_rdma_data_aligned = 0; - iser_finalize_rdma_unaligned_sg(iser_task, - &iser_task->data[ISER_DIR_IN], - ISER_DIR_IN); - } - - if (iser_task->data[ISER_DIR_OUT].orig_sg) { - is_rdma_data_aligned = 0; - iser_finalize_rdma_unaligned_sg(iser_task, - &iser_task->data[ISER_DIR_OUT], - ISER_DIR_OUT); - } - - if (iser_task->prot[ISER_DIR_IN].orig_sg) { - is_rdma_prot_aligned = 0; - iser_finalize_rdma_unaligned_sg(iser_task, - &iser_task->prot[ISER_DIR_IN], - ISER_DIR_IN); - } - - if (iser_task->prot[ISER_DIR_OUT].orig_sg) { - is_rdma_prot_aligned = 0; - iser_finalize_rdma_unaligned_sg(iser_task, - &iser_task->prot[ISER_DIR_OUT], - ISER_DIR_OUT); - } - if (iser_task->dir[ISER_DIR_IN]) { iser_unreg_rdma_mem(iser_task, ISER_DIR_IN); - if (is_rdma_data_aligned) - iser_dma_unmap_task_data(iser_task, - &iser_task->data[ISER_DIR_IN], - DMA_FROM_DEVICE); - if (prot_count && is_rdma_prot_aligned) + iser_dma_unmap_task_data(iser_task, + &iser_task->data[ISER_DIR_IN], + DMA_FROM_DEVICE); + if (prot_count) iser_dma_unmap_task_data(iser_task, &iser_task->prot[ISER_DIR_IN], DMA_FROM_DEVICE); @@ -710,11 +676,10 @@ void iser_task_rdma_finalize(struct iscsi_iser_task *iser_task) if (iser_task->dir[ISER_DIR_OUT]) { iser_unreg_rdma_mem(iser_task, ISER_DIR_OUT); - if (is_rdma_data_aligned) - iser_dma_unmap_task_data(iser_task, - &iser_task->data[ISER_DIR_OUT], - DMA_TO_DEVICE); - if (prot_count && is_rdma_prot_aligned) + iser_dma_unmap_task_data(iser_task, + &iser_task->data[ISER_DIR_OUT], + DMA_TO_DEVICE); + if (prot_count) iser_dma_unmap_task_data(iser_task, &iser_task->prot[ISER_DIR_OUT], DMA_TO_DEVICE); diff --git a/drivers/infiniband/ulp/iser/iser_memory.c b/drivers/infiniband/ulp/iser/iser_memory.c index 4c46d67..ea765fb 100644 --- a/drivers/infiniband/ulp/iser/iser_memory.c +++ b/drivers/infiniband/ulp/iser/iser_memory.c @@ -88,113 +88,6 @@ int iser_assign_reg_ops(struct iser_device *device) return 0; } -static void -iser_free_bounce_sg(struct iser_data_buf *data) -{ - struct scatterlist *sg; - int count; - - for_each_sg(data->sg, sg, data->size, count) - __free_page(sg_page(sg)); - - kfree(data->sg); - - data->sg = data->orig_sg; - data->size = data->orig_size; - data->orig_sg = NULL; - data->orig_size = 0; -} - -static int -iser_alloc_bounce_sg(struct iser_data_buf *data) -{ - struct scatterlist *sg; - struct page *page; - unsigned long length = data->data_len; - int i = 0, nents = DIV_ROUND_UP(length, PAGE_SIZE); - - sg = kcalloc(nents, sizeof(*sg), GFP_ATOMIC); - if (!sg) - goto err; - - sg_init_table(sg, nents); - while (length) { - u32 page_len = min_t(u32, length, PAGE_SIZE); - - page = alloc_page(GFP_ATOMIC); - if (!page) - goto err; - - sg_set_page(&sg[i], page, page_len, 0); - length -= page_len; - i++; - } - - data->orig_sg = data->sg; - data->orig_size = data->size; - data->sg = sg; - data->size = nents; - - return 0; - -err: - for (; i > 0; i--) - __free_page(sg_page(&sg[i - 1])); - kfree(sg); - - return -ENOMEM; -} - -static void -iser_copy_bounce(struct iser_data_buf *data, bool to_buffer) -{ - struct scatterlist *osg, *bsg = data->sg; - void *oaddr, *baddr; - unsigned int left = data->data_len; - unsigned int bsg_off = 0; - int i; - - for_each_sg(data->orig_sg, osg, data->orig_size, i) { - unsigned int copy_len, osg_off = 0; - - oaddr = kmap_atomic(sg_page(osg)) + osg->offset; - copy_len = min(left, osg->length); - while (copy_len) { - unsigned int len = min(copy_len, bsg->length - bsg_off); - - baddr = kmap_atomic(sg_page(bsg)) + bsg->offset; - if (to_buffer) - memcpy(baddr + bsg_off, oaddr + osg_off, len); - else - memcpy(oaddr + osg_off, baddr + bsg_off, len); - - kunmap_atomic(baddr - bsg->offset); - osg_off += len; - bsg_off += len; - copy_len -= len; - - if (bsg_off >= bsg->length) { - bsg = sg_next(bsg); - bsg_off = 0; - } - } - kunmap_atomic(oaddr - osg->offset); - left -= osg_off; - } -} - -static inline void -iser_copy_from_bounce(struct iser_data_buf *data) -{ - iser_copy_bounce(data, false); -} - -static inline void -iser_copy_to_bounce(struct iser_data_buf *data) -{ - iser_copy_bounce(data, true); -} - struct iser_fr_desc * iser_reg_desc_get_fr(struct ib_conn *ib_conn) { @@ -238,62 +131,6 @@ iser_reg_desc_put_fmr(struct ib_conn *ib_conn, { } -/** - * iser_start_rdma_unaligned_sg - */ -static int iser_start_rdma_unaligned_sg(struct iscsi_iser_task *iser_task, - struct iser_data_buf *data, - enum iser_data_dir cmd_dir) -{ - struct ib_device *dev = iser_task->iser_conn->ib_conn.device->ib_device; - int rc; - - rc = iser_alloc_bounce_sg(data); - if (rc) { - iser_err("Failed to allocate bounce for data len %lu\n", - data->data_len); - return rc; - } - - if (cmd_dir == ISER_DIR_OUT) - iser_copy_to_bounce(data); - - data->dma_nents = ib_dma_map_sg(dev, data->sg, data->size, - (cmd_dir == ISER_DIR_OUT) ? - DMA_TO_DEVICE : DMA_FROM_DEVICE); - if (!data->dma_nents) { - iser_err("Got dma_nents %d, something went wrong...\n", - data->dma_nents); - rc = -ENOMEM; - goto err; - } - - return 0; -err: - iser_free_bounce_sg(data); - return rc; -} - -/** - * iser_finalize_rdma_unaligned_sg - */ - -void iser_finalize_rdma_unaligned_sg(struct iscsi_iser_task *iser_task, - struct iser_data_buf *data, - enum iser_data_dir cmd_dir) -{ - struct ib_device *dev = iser_task->iser_conn->ib_conn.device->ib_device; - - ib_dma_unmap_sg(dev, data->sg, data->size, - (cmd_dir == ISER_DIR_OUT) ? - DMA_TO_DEVICE : DMA_FROM_DEVICE); - - if (cmd_dir == ISER_DIR_IN) - iser_copy_from_bounce(data); - - iser_free_bounce_sg(data); -} - #define IS_4K_ALIGNED(addr) ((((unsigned long)addr) & ~MASK_4K) == 0) /** @@ -355,64 +192,6 @@ static int iser_sg_to_page_vec(struct iser_data_buf *data, return cur_page; } - -/** - * iser_data_buf_aligned_len - Tries to determine the maximal correctly aligned - * for RDMA sub-list of a scatter-gather list of memory buffers, and returns - * the number of entries which are aligned correctly. Supports the case where - * consecutive SG elements are actually fragments of the same physcial page. - */ -static int iser_data_buf_aligned_len(struct iser_data_buf *data, - struct ib_device *ibdev, - unsigned sg_tablesize) -{ - struct scatterlist *sg, *sgl, *next_sg = NULL; - u64 start_addr, end_addr; - int i, ret_len, start_check = 0; - - if (data->dma_nents == 1) - return 1; - - sgl = data->sg; - start_addr = ib_sg_dma_address(ibdev, sgl); - - if (unlikely(sgl[0].offset && - data->data_len >= sg_tablesize * PAGE_SIZE)) { - iser_dbg("can't register length %lx with offset %x " - "fall to bounce buffer\n", data->data_len, - sgl[0].offset); - return 0; - } - - for_each_sg(sgl, sg, data->dma_nents, i) { - if (start_check && !IS_4K_ALIGNED(start_addr)) - break; - - next_sg = sg_next(sg); - if (!next_sg) - break; - - end_addr = start_addr + ib_sg_dma_len(ibdev, sg); - start_addr = ib_sg_dma_address(ibdev, next_sg); - - if (end_addr == start_addr) { - start_check = 0; - continue; - } else - start_check = 1; - - if (!IS_4K_ALIGNED(end_addr)) - break; - } - ret_len = (next_sg) ? i : i+1; - - if (unlikely(ret_len != data->dma_nents)) - iser_warn("rdma alignment violation (%d/%d aligned)\n", - ret_len, data->dma_nents); - - return ret_len; -} - static void iser_data_buf_dump(struct iser_data_buf *data, struct ib_device *ibdev) { @@ -483,31 +262,6 @@ iser_reg_dma(struct iser_device *device, struct iser_data_buf *mem, return 0; } -static int fall_to_bounce_buf(struct iscsi_iser_task *iser_task, - struct iser_data_buf *mem, - enum iser_data_dir cmd_dir) -{ - struct iscsi_conn *iscsi_conn = iser_task->iser_conn->iscsi_conn; - struct iser_device *device = iser_task->iser_conn->ib_conn.device; - - iscsi_conn->fmr_unalign_cnt++; - - if (iser_debug_level > 0) - iser_data_buf_dump(mem, device->ib_device); - - /* unmap the command data before accessing it */ - iser_dma_unmap_task_data(iser_task, mem, - (cmd_dir == ISER_DIR_OUT) ? - DMA_TO_DEVICE : DMA_FROM_DEVICE); - - /* allocate copy buf, if we are writing, copy the */ - /* unaligned scatterlist, dma map the copy */ - if (iser_start_rdma_unaligned_sg(iser_task, mem, cmd_dir) != 0) - return -ENOMEM; - - return 0; -} - /** * iser_reg_page_vec - Register physical memory * @@ -683,7 +437,7 @@ iser_reg_sig_mr(struct iscsi_iser_task *iser_task, { struct iser_tx_desc *tx_desc = &iser_task->desc; struct ib_sig_attrs *sig_attrs = &tx_desc->sig_attrs; - struct ib_send_wr *wr; + struct ib_sig_handover_wr *wr; int ret; memset(sig_attrs, 0, sizeof(*sig_attrs)); @@ -693,26 +447,24 @@ iser_reg_sig_mr(struct iscsi_iser_task *iser_task, iser_set_prot_checks(iser_task->sc, &sig_attrs->check_mask); - if (!pi_ctx->sig_mr_valid) { - wr = iser_tx_next_wr(tx_desc); - iser_inv_rkey(wr, pi_ctx->sig_mr); - } - - wr = iser_tx_next_wr(tx_desc); - wr->opcode = IB_WR_REG_SIG_MR; - wr->wr_id = ISER_FASTREG_LI_WRID; - wr->sg_list = &data_reg->sge; - wr->num_sge = 1; - wr->send_flags = 0; - wr->wr.sig_handover.sig_attrs = sig_attrs; - wr->wr.sig_handover.sig_mr = pi_ctx->sig_mr; + if (!pi_ctx->sig_mr_valid) + iser_inv_rkey(iser_tx_next_wr(tx_desc), pi_ctx->sig_mr); + + wr = sig_handover_wr(iser_tx_next_wr(tx_desc)); + wr->wr.opcode = IB_WR_REG_SIG_MR; + wr->wr.wr_id = ISER_FASTREG_LI_WRID; + wr->wr.sg_list = &data_reg->sge; + wr->wr.num_sge = 1; + wr->wr.send_flags = 0; + wr->sig_attrs = sig_attrs; + wr->sig_mr = pi_ctx->sig_mr; if (scsi_prot_sg_count(iser_task->sc)) - wr->wr.sig_handover.prot = &prot_reg->sge; + wr->prot = &prot_reg->sge; else - wr->wr.sig_handover.prot = NULL; - wr->wr.sig_handover.access_flags = IB_ACCESS_LOCAL_WRITE | - IB_ACCESS_REMOTE_READ | - IB_ACCESS_REMOTE_WRITE; + wr->prot = NULL; + wr->access_flags = IB_ACCESS_LOCAL_WRITE | + IB_ACCESS_REMOTE_READ | + IB_ACCESS_REMOTE_WRITE; pi_ctx->sig_mr_valid = 0; sig_reg->sge.lkey = pi_ctx->sig_mr->lkey; @@ -720,7 +472,7 @@ iser_reg_sig_mr(struct iscsi_iser_task *iser_task, sig_reg->sge.addr = 0; sig_reg->sge.length = scsi_transfer_length(iser_task->sc); - iser_dbg("sig reg: lkey: 0x%x, rkey: 0x%x, addr: 0x%llx, length: %u\n", + iser_dbg("lkey=0x%x rkey=0x%x addr=0x%llx length=%u\n", sig_reg->sge.lkey, sig_reg->rkey, sig_reg->sge.addr, sig_reg->sge.length); err: @@ -732,69 +484,41 @@ static int iser_fast_reg_mr(struct iscsi_iser_task *iser_task, struct iser_reg_resources *rsc, struct iser_mem_reg *reg) { - struct ib_conn *ib_conn = &iser_task->iser_conn->ib_conn; - struct iser_device *device = ib_conn->device; - struct ib_mr *mr = rsc->mr; - struct ib_fast_reg_page_list *frpl = rsc->frpl; struct iser_tx_desc *tx_desc = &iser_task->desc; - struct ib_send_wr *wr; - int offset, size, plen; + struct ib_mr *mr = rsc->mr; + struct ib_reg_wr *wr; + int n; - plen = iser_sg_to_page_vec(mem, device->ib_device, frpl->page_list, - &offset, &size); - if (plen * SIZE_4K < size) { - iser_err("fast reg page_list too short to hold this SG\n"); - return -EINVAL; - } + if (!rsc->mr_valid) + iser_inv_rkey(iser_tx_next_wr(tx_desc), mr); - if (!rsc->mr_valid) { - wr = iser_tx_next_wr(tx_desc); - iser_inv_rkey(wr, mr); + n = ib_map_mr_sg(mr, mem->sg, mem->size, SIZE_4K); + if (unlikely(n != mem->size)) { + iser_err("failed to map sg (%d/%d)\n", + n, mem->size); + return n < 0 ? n : -EINVAL; } - wr = iser_tx_next_wr(tx_desc); - wr->opcode = IB_WR_FAST_REG_MR; - wr->wr_id = ISER_FASTREG_LI_WRID; - wr->send_flags = 0; - wr->wr.fast_reg.iova_start = frpl->page_list[0] + offset; - wr->wr.fast_reg.page_list = frpl; - wr->wr.fast_reg.page_list_len = plen; - wr->wr.fast_reg.page_shift = SHIFT_4K; - wr->wr.fast_reg.length = size; - wr->wr.fast_reg.rkey = mr->rkey; - wr->wr.fast_reg.access_flags = (IB_ACCESS_LOCAL_WRITE | - IB_ACCESS_REMOTE_WRITE | - IB_ACCESS_REMOTE_READ); + wr = reg_wr(iser_tx_next_wr(tx_desc)); + wr->wr.opcode = IB_WR_REG_MR; + wr->wr.wr_id = ISER_FASTREG_LI_WRID; + wr->wr.send_flags = 0; + wr->wr.num_sge = 0; + wr->mr = mr; + wr->key = mr->rkey; + wr->access = IB_ACCESS_LOCAL_WRITE | + IB_ACCESS_REMOTE_WRITE | + IB_ACCESS_REMOTE_READ; + rsc->mr_valid = 0; reg->sge.lkey = mr->lkey; reg->rkey = mr->rkey; - reg->sge.addr = frpl->page_list[0] + offset; - reg->sge.length = size; + reg->sge.addr = mr->iova; + reg->sge.length = mr->length; - iser_dbg("fast reg: lkey=0x%x, rkey=0x%x, addr=0x%llx," - " length=0x%x\n", reg->sge.lkey, reg->rkey, - reg->sge.addr, reg->sge.length); - - return 0; -} - -static int -iser_handle_unaligned_buf(struct iscsi_iser_task *task, - struct iser_data_buf *mem, - enum iser_data_dir dir) -{ - struct iser_conn *iser_conn = task->iser_conn; - struct iser_device *device = iser_conn->ib_conn.device; - int err, aligned_len; - - aligned_len = iser_data_buf_aligned_len(mem, device->ib_device, - iser_conn->scsi_sg_tablesize); - if (aligned_len != mem->dma_nents) { - err = fall_to_bounce_buf(task, mem, dir); - if (err) - return err; - } + iser_dbg("lkey=0x%x rkey=0x%x addr=0x%llx length=0x%x\n", + reg->sge.lkey, reg->rkey, reg->sge.addr, reg->sge.length); return 0; } @@ -841,10 +565,6 @@ int iser_reg_rdma_mem(struct iscsi_iser_task *task, bool use_dma_key; int err; - err = iser_handle_unaligned_buf(task, mem, dir); - if (unlikely(err)) - return err; - use_dma_key = (mem->dma_nents == 1 && !iser_always_reg && scsi_get_prot_op(task->sc) == SCSI_PROT_NORMAL); @@ -867,10 +587,6 @@ int iser_reg_rdma_mem(struct iscsi_iser_task *task, if (scsi_prot_sg_count(task->sc)) { mem = &task->prot[dir]; - err = iser_handle_unaligned_buf(task, mem, dir); - if (unlikely(err)) - goto err_reg; - err = iser_reg_prot_sg(task, mem, desc, use_dma_key, prot_reg); if (unlikely(err)) diff --git a/drivers/infiniband/ulp/iser/iser_verbs.c b/drivers/infiniband/ulp/iser/iser_verbs.c index 85132d8..a930702 100644 --- a/drivers/infiniband/ulp/iser/iser_verbs.c +++ b/drivers/infiniband/ulp/iser/iser_verbs.c @@ -293,35 +293,21 @@ iser_alloc_reg_res(struct ib_device *ib_device, { int ret; - res->frpl = ib_alloc_fast_reg_page_list(ib_device, size); - if (IS_ERR(res->frpl)) { - ret = PTR_ERR(res->frpl); - iser_err("Failed to allocate ib_fast_reg_page_list err=%d\n", - ret); - return PTR_ERR(res->frpl); - } - res->mr = ib_alloc_mr(pd, IB_MR_TYPE_MEM_REG, size); if (IS_ERR(res->mr)) { ret = PTR_ERR(res->mr); iser_err("Failed to allocate ib_fast_reg_mr err=%d\n", ret); - goto fast_reg_mr_failure; + return ret; } res->mr_valid = 1; return 0; - -fast_reg_mr_failure: - ib_free_fast_reg_page_list(res->frpl); - - return ret; } static void iser_free_reg_res(struct iser_reg_resources *rsc) { ib_dereg_mr(rsc->mr); - ib_free_fast_reg_page_list(rsc->frpl); } static int @@ -1017,7 +1003,7 @@ int iser_connect(struct iser_conn *iser_conn, ib_conn->beacon.wr_id = ISER_BEACON_WRID; ib_conn->beacon.opcode = IB_WR_SEND; - ib_conn->cma_id = rdma_create_id(iser_cma_handler, + ib_conn->cma_id = rdma_create_id(&init_net, iser_cma_handler, (void *)iser_conn, RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(ib_conn->cma_id)) { @@ -1135,7 +1121,7 @@ int iser_post_send(struct ib_conn *ib_conn, struct iser_tx_desc *tx_desc, wr->opcode = IB_WR_SEND; wr->send_flags = signal ? IB_SEND_SIGNALED : 0; - ib_ret = ib_post_send(ib_conn->qp, &tx_desc->wrs[0], &bad_wr); + ib_ret = ib_post_send(ib_conn->qp, &tx_desc->wrs[0].send, &bad_wr); if (ib_ret) iser_err("ib_post_send failed, ret:%d opcode:%d\n", ib_ret, bad_wr->opcode); diff --git a/drivers/infiniband/ulp/isert/ib_isert.c b/drivers/infiniband/ulp/isert/ib_isert.c index aa59037..dfbbbb2 100644 --- a/drivers/infiniband/ulp/isert/ib_isert.c +++ b/drivers/infiniband/ulp/isert/ib_isert.c @@ -473,10 +473,8 @@ isert_conn_free_fastreg_pool(struct isert_conn *isert_conn) list_for_each_entry_safe(fr_desc, tmp, &isert_conn->fr_pool, list) { list_del(&fr_desc->list); - ib_free_fast_reg_page_list(fr_desc->data_frpl); ib_dereg_mr(fr_desc->data_mr); if (fr_desc->pi_ctx) { - ib_free_fast_reg_page_list(fr_desc->pi_ctx->prot_frpl); ib_dereg_mr(fr_desc->pi_ctx->prot_mr); ib_dereg_mr(fr_desc->pi_ctx->sig_mr); kfree(fr_desc->pi_ctx); @@ -504,22 +502,13 @@ isert_create_pi_ctx(struct fast_reg_descriptor *desc, return -ENOMEM; } - pi_ctx->prot_frpl = ib_alloc_fast_reg_page_list(device, - ISCSI_ISER_SG_TABLESIZE); - if (IS_ERR(pi_ctx->prot_frpl)) { - isert_err("Failed to allocate prot frpl err=%ld\n", - PTR_ERR(pi_ctx->prot_frpl)); - ret = PTR_ERR(pi_ctx->prot_frpl); - goto err_pi_ctx; - } - pi_ctx->prot_mr = ib_alloc_mr(pd, IB_MR_TYPE_MEM_REG, ISCSI_ISER_SG_TABLESIZE); if (IS_ERR(pi_ctx->prot_mr)) { isert_err("Failed to allocate prot frmr err=%ld\n", PTR_ERR(pi_ctx->prot_mr)); ret = PTR_ERR(pi_ctx->prot_mr); - goto err_prot_frpl; + goto err_pi_ctx; } desc->ind |= ISERT_PROT_KEY_VALID; @@ -539,8 +528,6 @@ isert_create_pi_ctx(struct fast_reg_descriptor *desc, err_prot_mr: ib_dereg_mr(pi_ctx->prot_mr); -err_prot_frpl: - ib_free_fast_reg_page_list(pi_ctx->prot_frpl); err_pi_ctx: kfree(pi_ctx); @@ -551,34 +538,18 @@ static int isert_create_fr_desc(struct ib_device *ib_device, struct ib_pd *pd, struct fast_reg_descriptor *fr_desc) { - int ret; - - fr_desc->data_frpl = ib_alloc_fast_reg_page_list(ib_device, - ISCSI_ISER_SG_TABLESIZE); - if (IS_ERR(fr_desc->data_frpl)) { - isert_err("Failed to allocate data frpl err=%ld\n", - PTR_ERR(fr_desc->data_frpl)); - return PTR_ERR(fr_desc->data_frpl); - } - fr_desc->data_mr = ib_alloc_mr(pd, IB_MR_TYPE_MEM_REG, ISCSI_ISER_SG_TABLESIZE); if (IS_ERR(fr_desc->data_mr)) { isert_err("Failed to allocate data frmr err=%ld\n", PTR_ERR(fr_desc->data_mr)); - ret = PTR_ERR(fr_desc->data_mr); - goto err_data_frpl; + return PTR_ERR(fr_desc->data_mr); } fr_desc->ind |= ISERT_DATA_KEY_VALID; isert_dbg("Created fr_desc %p\n", fr_desc); return 0; - -err_data_frpl: - ib_free_fast_reg_page_list(fr_desc->data_frpl); - - return ret; } static int @@ -1579,7 +1550,6 @@ isert_rx_do_work(struct iser_rx_desc *rx_desc, struct isert_conn *isert_conn) struct iser_hdr *iser_hdr = &rx_desc->iser_header; uint64_t read_va = 0, write_va = 0; uint32_t read_stag = 0, write_stag = 0; - int rc; switch (iser_hdr->flags & 0xF0) { case ISCSI_CTRL: @@ -1606,8 +1576,8 @@ isert_rx_do_work(struct iser_rx_desc *rx_desc, struct isert_conn *isert_conn) break; } - rc = isert_rx_opcode(isert_conn, rx_desc, - read_stag, read_va, write_stag, write_va); + isert_rx_opcode(isert_conn, rx_desc, + read_stag, read_va, write_stag, write_va); } static void @@ -1716,10 +1686,10 @@ isert_unmap_cmd(struct isert_cmd *isert_cmd, struct isert_conn *isert_conn) isert_unmap_data_buf(isert_conn, &wr->data); } - if (wr->send_wr) { + if (wr->rdma_wr) { isert_dbg("Cmd %p free send_wr\n", isert_cmd); - kfree(wr->send_wr); - wr->send_wr = NULL; + kfree(wr->rdma_wr); + wr->rdma_wr = NULL; } if (wr->ib_sge) { @@ -1754,7 +1724,7 @@ isert_unreg_rdma(struct isert_cmd *isert_cmd, struct isert_conn *isert_conn) } wr->ib_sge = NULL; - wr->send_wr = NULL; + wr->rdma_wr = NULL; } static void @@ -1923,7 +1893,7 @@ isert_completion_rdma_write(struct iser_tx_desc *tx_desc, } device->unreg_rdma_mem(isert_cmd, isert_conn); - wr->send_wr_num = 0; + wr->rdma_wr_num = 0; if (ret) transport_send_check_condition_and_sense(se_cmd, se_cmd->pi_err, 0); @@ -1951,7 +1921,7 @@ isert_completion_rdma_read(struct iser_tx_desc *tx_desc, iscsit_stop_dataout_timer(cmd); device->unreg_rdma_mem(isert_cmd, isert_conn); cmd->write_data_done = wr->data.len; - wr->send_wr_num = 0; + wr->rdma_wr_num = 0; isert_dbg("Cmd: %p RDMA_READ comp calling execute_cmd\n", isert_cmd); spin_lock_bh(&cmd->istate_lock); @@ -2403,7 +2373,7 @@ isert_put_text_rsp(struct iscsi_cmd *cmd, struct iscsi_conn *conn) static int isert_build_rdma_wr(struct isert_conn *isert_conn, struct isert_cmd *isert_cmd, - struct ib_sge *ib_sge, struct ib_send_wr *send_wr, + struct ib_sge *ib_sge, struct ib_rdma_wr *rdma_wr, u32 data_left, u32 offset) { struct iscsi_cmd *cmd = isert_cmd->iscsi_cmd; @@ -2418,8 +2388,8 @@ isert_build_rdma_wr(struct isert_conn *isert_conn, struct isert_cmd *isert_cmd, sg_nents = min(cmd->se_cmd.t_data_nents - sg_off, isert_conn->max_sge); page_off = offset % PAGE_SIZE; - send_wr->sg_list = ib_sge; - send_wr->wr_id = (uintptr_t)&isert_cmd->tx_desc; + rdma_wr->wr.sg_list = ib_sge; + rdma_wr->wr.wr_id = (uintptr_t)&isert_cmd->tx_desc; /* * Perform mapping of TCM scatterlist memory ib_sge dma_addr. */ @@ -2444,11 +2414,11 @@ isert_build_rdma_wr(struct isert_conn *isert_conn, struct isert_cmd *isert_cmd, isert_dbg("Incrementing ib_sge pointer to %p\n", ib_sge); } - send_wr->num_sge = ++i; + rdma_wr->wr.num_sge = ++i; isert_dbg("Set outgoing sg_list: %p num_sg: %u from TCM SGLs\n", - send_wr->sg_list, send_wr->num_sge); + rdma_wr->wr.sg_list, rdma_wr->wr.num_sge); - return send_wr->num_sge; + return rdma_wr->wr.num_sge; } static int @@ -2459,7 +2429,7 @@ isert_map_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd, struct isert_cmd *isert_cmd = iscsit_priv_cmd(cmd); struct isert_conn *isert_conn = conn->context; struct isert_data_buf *data = &wr->data; - struct ib_send_wr *send_wr; + struct ib_rdma_wr *rdma_wr; struct ib_sge *ib_sge; u32 offset, data_len, data_left, rdma_write_max, va_offset = 0; int ret = 0, i, ib_sge_cnt; @@ -2484,11 +2454,11 @@ isert_map_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd, } wr->ib_sge = ib_sge; - wr->send_wr_num = DIV_ROUND_UP(data->nents, isert_conn->max_sge); - wr->send_wr = kzalloc(sizeof(struct ib_send_wr) * wr->send_wr_num, + wr->rdma_wr_num = DIV_ROUND_UP(data->nents, isert_conn->max_sge); + wr->rdma_wr = kzalloc(sizeof(struct ib_rdma_wr) * wr->rdma_wr_num, GFP_KERNEL); - if (!wr->send_wr) { - isert_dbg("Unable to allocate wr->send_wr\n"); + if (!wr->rdma_wr) { + isert_dbg("Unable to allocate wr->rdma_wr\n"); ret = -ENOMEM; goto unmap_cmd; } @@ -2496,31 +2466,31 @@ isert_map_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd, wr->isert_cmd = isert_cmd; rdma_write_max = isert_conn->max_sge * PAGE_SIZE; - for (i = 0; i < wr->send_wr_num; i++) { - send_wr = &isert_cmd->rdma_wr.send_wr[i]; + for (i = 0; i < wr->rdma_wr_num; i++) { + rdma_wr = &isert_cmd->rdma_wr.rdma_wr[i]; data_len = min(data_left, rdma_write_max); - send_wr->send_flags = 0; + rdma_wr->wr.send_flags = 0; if (wr->iser_ib_op == ISER_IB_RDMA_WRITE) { - send_wr->opcode = IB_WR_RDMA_WRITE; - send_wr->wr.rdma.remote_addr = isert_cmd->read_va + offset; - send_wr->wr.rdma.rkey = isert_cmd->read_stag; - if (i + 1 == wr->send_wr_num) - send_wr->next = &isert_cmd->tx_desc.send_wr; + rdma_wr->wr.opcode = IB_WR_RDMA_WRITE; + rdma_wr->remote_addr = isert_cmd->read_va + offset; + rdma_wr->rkey = isert_cmd->read_stag; + if (i + 1 == wr->rdma_wr_num) + rdma_wr->wr.next = &isert_cmd->tx_desc.send_wr; else - send_wr->next = &wr->send_wr[i + 1]; + rdma_wr->wr.next = &wr->rdma_wr[i + 1].wr; } else { - send_wr->opcode = IB_WR_RDMA_READ; - send_wr->wr.rdma.remote_addr = isert_cmd->write_va + va_offset; - send_wr->wr.rdma.rkey = isert_cmd->write_stag; - if (i + 1 == wr->send_wr_num) - send_wr->send_flags = IB_SEND_SIGNALED; + rdma_wr->wr.opcode = IB_WR_RDMA_READ; + rdma_wr->remote_addr = isert_cmd->write_va + va_offset; + rdma_wr->rkey = isert_cmd->write_stag; + if (i + 1 == wr->rdma_wr_num) + rdma_wr->wr.send_flags = IB_SEND_SIGNALED; else - send_wr->next = &wr->send_wr[i + 1]; + rdma_wr->wr.next = &wr->rdma_wr[i + 1].wr; } ib_sge_cnt = isert_build_rdma_wr(isert_conn, isert_cmd, ib_sge, - send_wr, data_len, offset); + rdma_wr, data_len, offset); ib_sge += ib_sge_cnt; offset += data_len; @@ -2535,45 +2505,6 @@ unmap_cmd: return ret; } -static int -isert_map_fr_pagelist(struct ib_device *ib_dev, - struct scatterlist *sg_start, int sg_nents, u64 *fr_pl) -{ - u64 start_addr, end_addr, page, chunk_start = 0; - struct scatterlist *tmp_sg; - int i = 0, new_chunk, last_ent, n_pages; - - n_pages = 0; - new_chunk = 1; - last_ent = sg_nents - 1; - for_each_sg(sg_start, tmp_sg, sg_nents, i) { - start_addr = ib_sg_dma_address(ib_dev, tmp_sg); - if (new_chunk) - chunk_start = start_addr; - end_addr = start_addr + ib_sg_dma_len(ib_dev, tmp_sg); - - isert_dbg("SGL[%d] dma_addr: 0x%llx len: %u\n", - i, (unsigned long long)tmp_sg->dma_address, - tmp_sg->length); - - if ((end_addr & ~PAGE_MASK) && i < last_ent) { - new_chunk = 0; - continue; - } - new_chunk = 1; - - page = chunk_start & PAGE_MASK; - do { - fr_pl[n_pages++] = page; - isert_dbg("Mapped page_list[%d] page_addr: 0x%llx\n", - n_pages - 1, page); - page += PAGE_SIZE; - } while (page < end_addr); - } - - return n_pages; -} - static inline void isert_inv_rkey(struct ib_send_wr *inv_wr, struct ib_mr *mr) { @@ -2599,11 +2530,9 @@ isert_fast_reg_mr(struct isert_conn *isert_conn, struct isert_device *device = isert_conn->device; struct ib_device *ib_dev = device->ib_device; struct ib_mr *mr; - struct ib_fast_reg_page_list *frpl; - struct ib_send_wr fr_wr, inv_wr; - struct ib_send_wr *bad_wr, *wr = NULL; - int ret, pagelist_len; - u32 page_off; + struct ib_reg_wr reg_wr; + struct ib_send_wr inv_wr, *bad_wr, *wr = NULL; + int ret, n; if (mem->dma_nents == 1) { sge->lkey = device->pd->local_dma_lkey; @@ -2614,45 +2543,41 @@ isert_fast_reg_mr(struct isert_conn *isert_conn, return 0; } - if (ind == ISERT_DATA_KEY_VALID) { + if (ind == ISERT_DATA_KEY_VALID) /* Registering data buffer */ mr = fr_desc->data_mr; - frpl = fr_desc->data_frpl; - } else { + else /* Registering protection buffer */ mr = fr_desc->pi_ctx->prot_mr; - frpl = fr_desc->pi_ctx->prot_frpl; - } - - page_off = mem->offset % PAGE_SIZE; - - isert_dbg("Use fr_desc %p sg_nents %d offset %u\n", - fr_desc, mem->nents, mem->offset); - - pagelist_len = isert_map_fr_pagelist(ib_dev, mem->sg, mem->nents, - &frpl->page_list[0]); if (!(fr_desc->ind & ind)) { isert_inv_rkey(&inv_wr, mr); wr = &inv_wr; } - /* Prepare FASTREG WR */ - memset(&fr_wr, 0, sizeof(fr_wr)); - fr_wr.wr_id = ISER_FASTREG_LI_WRID; - fr_wr.opcode = IB_WR_FAST_REG_MR; - fr_wr.wr.fast_reg.iova_start = frpl->page_list[0] + page_off; - fr_wr.wr.fast_reg.page_list = frpl; - fr_wr.wr.fast_reg.page_list_len = pagelist_len; - fr_wr.wr.fast_reg.page_shift = PAGE_SHIFT; - fr_wr.wr.fast_reg.length = mem->len; - fr_wr.wr.fast_reg.rkey = mr->rkey; - fr_wr.wr.fast_reg.access_flags = IB_ACCESS_LOCAL_WRITE; + n = ib_map_mr_sg(mr, mem->sg, mem->nents, PAGE_SIZE); + if (unlikely(n != mem->nents)) { + isert_err("failed to map mr sg (%d/%d)\n", + n, mem->nents); + return n < 0 ? n : -EINVAL; + } + + isert_dbg("Use fr_desc %p sg_nents %d offset %u\n", + fr_desc, mem->nents, mem->offset); + + reg_wr.wr.next = NULL; + reg_wr.wr.opcode = IB_WR_REG_MR; + reg_wr.wr.wr_id = ISER_FASTREG_LI_WRID; + reg_wr.wr.send_flags = 0; + reg_wr.wr.num_sge = 0; + reg_wr.mr = mr; + reg_wr.key = mr->lkey; + reg_wr.access = IB_ACCESS_LOCAL_WRITE; if (!wr) - wr = &fr_wr; + wr = ®_wr.wr; else - wr->next = &fr_wr; + wr->next = ®_wr.wr; ret = ib_post_send(isert_conn->qp, wr, &bad_wr); if (ret) { @@ -2662,8 +2587,8 @@ isert_fast_reg_mr(struct isert_conn *isert_conn, fr_desc->ind &= ~ind; sge->lkey = mr->lkey; - sge->addr = frpl->page_list[0] + page_off; - sge->length = mem->len; + sge->addr = mr->iova; + sge->length = mr->length; isert_dbg("sge: addr: 0x%llx length: %u lkey: %x\n", sge->addr, sge->length, sge->lkey); @@ -2733,8 +2658,8 @@ isert_reg_sig_mr(struct isert_conn *isert_conn, struct isert_rdma_wr *rdma_wr, struct fast_reg_descriptor *fr_desc) { - struct ib_send_wr sig_wr, inv_wr; - struct ib_send_wr *bad_wr, *wr = NULL; + struct ib_sig_handover_wr sig_wr; + struct ib_send_wr inv_wr, *bad_wr, *wr = NULL; struct pi_context *pi_ctx = fr_desc->pi_ctx; struct ib_sig_attrs sig_attrs; int ret; @@ -2752,20 +2677,20 @@ isert_reg_sig_mr(struct isert_conn *isert_conn, } memset(&sig_wr, 0, sizeof(sig_wr)); - sig_wr.opcode = IB_WR_REG_SIG_MR; - sig_wr.wr_id = ISER_FASTREG_LI_WRID; - sig_wr.sg_list = &rdma_wr->ib_sg[DATA]; - sig_wr.num_sge = 1; - sig_wr.wr.sig_handover.access_flags = IB_ACCESS_LOCAL_WRITE; - sig_wr.wr.sig_handover.sig_attrs = &sig_attrs; - sig_wr.wr.sig_handover.sig_mr = pi_ctx->sig_mr; + sig_wr.wr.opcode = IB_WR_REG_SIG_MR; + sig_wr.wr.wr_id = ISER_FASTREG_LI_WRID; + sig_wr.wr.sg_list = &rdma_wr->ib_sg[DATA]; + sig_wr.wr.num_sge = 1; + sig_wr.access_flags = IB_ACCESS_LOCAL_WRITE; + sig_wr.sig_attrs = &sig_attrs; + sig_wr.sig_mr = pi_ctx->sig_mr; if (se_cmd->t_prot_sg) - sig_wr.wr.sig_handover.prot = &rdma_wr->ib_sg[PROT]; + sig_wr.prot = &rdma_wr->ib_sg[PROT]; if (!wr) - wr = &sig_wr; + wr = &sig_wr.wr; else - wr->next = &sig_wr; + wr->next = &sig_wr.wr; ret = ib_post_send(isert_conn->qp, wr, &bad_wr); if (ret) { @@ -2859,7 +2784,7 @@ isert_reg_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd, struct isert_cmd *isert_cmd = iscsit_priv_cmd(cmd); struct isert_conn *isert_conn = conn->context; struct fast_reg_descriptor *fr_desc = NULL; - struct ib_send_wr *send_wr; + struct ib_rdma_wr *rdma_wr; struct ib_sge *ib_sg; u32 offset; int ret = 0; @@ -2900,26 +2825,26 @@ isert_reg_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd, memcpy(&wr->s_ib_sge, ib_sg, sizeof(*ib_sg)); wr->ib_sge = &wr->s_ib_sge; - wr->send_wr_num = 1; - memset(&wr->s_send_wr, 0, sizeof(*send_wr)); - wr->send_wr = &wr->s_send_wr; + wr->rdma_wr_num = 1; + memset(&wr->s_rdma_wr, 0, sizeof(wr->s_rdma_wr)); + wr->rdma_wr = &wr->s_rdma_wr; wr->isert_cmd = isert_cmd; - send_wr = &isert_cmd->rdma_wr.s_send_wr; - send_wr->sg_list = &wr->s_ib_sge; - send_wr->num_sge = 1; - send_wr->wr_id = (uintptr_t)&isert_cmd->tx_desc; + rdma_wr = &isert_cmd->rdma_wr.s_rdma_wr; + rdma_wr->wr.sg_list = &wr->s_ib_sge; + rdma_wr->wr.num_sge = 1; + rdma_wr->wr.wr_id = (uintptr_t)&isert_cmd->tx_desc; if (wr->iser_ib_op == ISER_IB_RDMA_WRITE) { - send_wr->opcode = IB_WR_RDMA_WRITE; - send_wr->wr.rdma.remote_addr = isert_cmd->read_va; - send_wr->wr.rdma.rkey = isert_cmd->read_stag; - send_wr->send_flags = !isert_prot_cmd(isert_conn, se_cmd) ? + rdma_wr->wr.opcode = IB_WR_RDMA_WRITE; + rdma_wr->remote_addr = isert_cmd->read_va; + rdma_wr->rkey = isert_cmd->read_stag; + rdma_wr->wr.send_flags = !isert_prot_cmd(isert_conn, se_cmd) ? 0 : IB_SEND_SIGNALED; } else { - send_wr->opcode = IB_WR_RDMA_READ; - send_wr->wr.rdma.remote_addr = isert_cmd->write_va; - send_wr->wr.rdma.rkey = isert_cmd->write_stag; - send_wr->send_flags = IB_SEND_SIGNALED; + rdma_wr->wr.opcode = IB_WR_RDMA_READ; + rdma_wr->remote_addr = isert_cmd->write_va; + rdma_wr->rkey = isert_cmd->write_stag; + rdma_wr->wr.send_flags = IB_SEND_SIGNALED; } return 0; @@ -2967,8 +2892,8 @@ isert_put_datain(struct iscsi_conn *conn, struct iscsi_cmd *cmd) isert_init_tx_hdrs(isert_conn, &isert_cmd->tx_desc); isert_init_send_wr(isert_conn, isert_cmd, &isert_cmd->tx_desc.send_wr); - isert_cmd->rdma_wr.s_send_wr.next = &isert_cmd->tx_desc.send_wr; - wr->send_wr_num += 1; + isert_cmd->rdma_wr.s_rdma_wr.wr.next = &isert_cmd->tx_desc.send_wr; + wr->rdma_wr_num += 1; rc = isert_post_recv(isert_conn, isert_cmd->rx_desc); if (rc) { @@ -2977,7 +2902,7 @@ isert_put_datain(struct iscsi_conn *conn, struct iscsi_cmd *cmd) } } - rc = ib_post_send(isert_conn->qp, wr->send_wr, &wr_failed); + rc = ib_post_send(isert_conn->qp, &wr->rdma_wr->wr, &wr_failed); if (rc) isert_warn("ib_post_send() failed for IB_WR_RDMA_WRITE\n"); @@ -3011,7 +2936,7 @@ isert_get_dataout(struct iscsi_conn *conn, struct iscsi_cmd *cmd, bool recovery) return rc; } - rc = ib_post_send(isert_conn->qp, wr->send_wr, &wr_failed); + rc = ib_post_send(isert_conn->qp, &wr->rdma_wr->wr, &wr_failed); if (rc) isert_warn("ib_post_send() failed for IB_WR_RDMA_READ\n"); @@ -3097,7 +3022,7 @@ isert_setup_id(struct isert_np *isert_np) sa = (struct sockaddr *)&np->np_sockaddr; isert_dbg("ksockaddr: %p, sa: %p\n", &np->np_sockaddr, sa); - id = rdma_create_id(isert_cma_handler, isert_np, + id = rdma_create_id(&init_net, isert_cma_handler, isert_np, RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(id)) { isert_err("rdma_create_id() failed: %ld\n", PTR_ERR(id)); diff --git a/drivers/infiniband/ulp/isert/ib_isert.h b/drivers/infiniband/ulp/isert/ib_isert.h index c5b99bc..3d7fbc4 100644 --- a/drivers/infiniband/ulp/isert/ib_isert.h +++ b/drivers/infiniband/ulp/isert/ib_isert.h @@ -84,14 +84,12 @@ enum isert_indicator { struct pi_context { struct ib_mr *prot_mr; - struct ib_fast_reg_page_list *prot_frpl; struct ib_mr *sig_mr; }; struct fast_reg_descriptor { struct list_head list; struct ib_mr *data_mr; - struct ib_fast_reg_page_list *data_frpl; u8 ind; struct pi_context *pi_ctx; }; @@ -117,9 +115,9 @@ struct isert_rdma_wr { enum iser_ib_op_code iser_ib_op; struct ib_sge *ib_sge; struct ib_sge s_ib_sge; - int send_wr_num; - struct ib_send_wr *send_wr; - struct ib_send_wr s_send_wr; + int rdma_wr_num; + struct ib_rdma_wr *rdma_wr; + struct ib_rdma_wr s_rdma_wr; struct ib_sge ib_sg[3]; struct isert_data_buf data; struct isert_data_buf prot; diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index b481490..32f7962 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -340,8 +340,6 @@ static void srp_destroy_fr_pool(struct srp_fr_pool *pool) return; for (i = 0, d = &pool->desc[0]; i < pool->size; i++, d++) { - if (d->frpl) - ib_free_fast_reg_page_list(d->frpl); if (d->mr) ib_dereg_mr(d->mr); } @@ -362,7 +360,6 @@ static struct srp_fr_pool *srp_create_fr_pool(struct ib_device *device, struct srp_fr_pool *pool; struct srp_fr_desc *d; struct ib_mr *mr; - struct ib_fast_reg_page_list *frpl; int i, ret = -EINVAL; if (pool_size <= 0) @@ -385,12 +382,6 @@ static struct srp_fr_pool *srp_create_fr_pool(struct ib_device *device, goto destroy_pool; } d->mr = mr; - frpl = ib_alloc_fast_reg_page_list(device, max_page_list_len); - if (IS_ERR(frpl)) { - ret = PTR_ERR(frpl); - goto destroy_pool; - } - d->frpl = frpl; list_add_tail(&d->entry, &pool->free_list); } @@ -849,11 +840,12 @@ static void srp_free_req_data(struct srp_target_port *target, for (i = 0; i < target->req_ring_size; ++i) { req = &ch->req_ring[i]; - if (dev->use_fast_reg) + if (dev->use_fast_reg) { kfree(req->fr_list); - else + } else { kfree(req->fmr_list); - kfree(req->map_page); + kfree(req->map_page); + } if (req->indirect_dma_addr) { ib_dma_unmap_single(ibdev, req->indirect_dma_addr, target->indirect_size, @@ -887,14 +879,15 @@ static int srp_alloc_req_data(struct srp_rdma_ch *ch) GFP_KERNEL); if (!mr_list) goto out; - if (srp_dev->use_fast_reg) + if (srp_dev->use_fast_reg) { req->fr_list = mr_list; - else + } else { req->fmr_list = mr_list; - req->map_page = kmalloc(srp_dev->max_pages_per_mr * - sizeof(void *), GFP_KERNEL); - if (!req->map_page) - goto out; + req->map_page = kmalloc(srp_dev->max_pages_per_mr * + sizeof(void *), GFP_KERNEL); + if (!req->map_page) + goto out; + } req->indirect_desc = kmalloc(target->indirect_size, GFP_KERNEL); if (!req->indirect_desc) goto out; @@ -1286,6 +1279,17 @@ static int srp_map_finish_fmr(struct srp_map_state *state, if (state->fmr.next >= state->fmr.end) return -ENOMEM; + WARN_ON_ONCE(!dev->use_fmr); + + if (state->npages == 0) + return 0; + + if (state->npages == 1 && target->global_mr) { + srp_map_desc(state, state->base_dma_addr, state->dma_len, + target->global_mr->rkey); + goto reset_state; + } + fmr = ib_fmr_pool_map_phys(ch->fmr_pool, state->pages, state->npages, io_addr); if (IS_ERR(fmr)) @@ -1297,6 +1301,10 @@ static int srp_map_finish_fmr(struct srp_map_state *state, srp_map_desc(state, state->base_dma_addr & ~dev->mr_page_mask, state->dma_len, fmr->fmr->rkey); +reset_state: + state->npages = 0; + state->dma_len = 0; + return 0; } @@ -1306,13 +1314,26 @@ static int srp_map_finish_fr(struct srp_map_state *state, struct srp_target_port *target = ch->target; struct srp_device *dev = target->srp_host->srp_dev; struct ib_send_wr *bad_wr; - struct ib_send_wr wr; + struct ib_reg_wr wr; struct srp_fr_desc *desc; u32 rkey; + int n, err; if (state->fr.next >= state->fr.end) return -ENOMEM; + WARN_ON_ONCE(!dev->use_fast_reg); + + if (state->sg_nents == 0) + return 0; + + if (state->sg_nents == 1 && target->global_mr) { + srp_map_desc(state, sg_dma_address(state->sg), + sg_dma_len(state->sg), + target->global_mr->rkey); + return 1; + } + desc = srp_fr_pool_get(ch->fr_pool); if (!desc) return -ENOMEM; @@ -1320,56 +1341,33 @@ static int srp_map_finish_fr(struct srp_map_state *state, rkey = ib_inc_rkey(desc->mr->rkey); ib_update_fast_reg_key(desc->mr, rkey); - memcpy(desc->frpl->page_list, state->pages, - sizeof(state->pages[0]) * state->npages); - - memset(&wr, 0, sizeof(wr)); - wr.opcode = IB_WR_FAST_REG_MR; - wr.wr_id = FAST_REG_WR_ID_MASK; - wr.wr.fast_reg.iova_start = state->base_dma_addr; - wr.wr.fast_reg.page_list = desc->frpl; - wr.wr.fast_reg.page_list_len = state->npages; - wr.wr.fast_reg.page_shift = ilog2(dev->mr_page_size); - wr.wr.fast_reg.length = state->dma_len; - wr.wr.fast_reg.access_flags = (IB_ACCESS_LOCAL_WRITE | - IB_ACCESS_REMOTE_READ | - IB_ACCESS_REMOTE_WRITE); - wr.wr.fast_reg.rkey = desc->mr->lkey; + n = ib_map_mr_sg(desc->mr, state->sg, state->sg_nents, + dev->mr_page_size); + if (unlikely(n < 0)) + return n; + + wr.wr.next = NULL; + wr.wr.opcode = IB_WR_REG_MR; + wr.wr.wr_id = FAST_REG_WR_ID_MASK; + wr.wr.num_sge = 0; + wr.wr.send_flags = 0; + wr.mr = desc->mr; + wr.key = desc->mr->rkey; + wr.access = (IB_ACCESS_LOCAL_WRITE | + IB_ACCESS_REMOTE_READ | + IB_ACCESS_REMOTE_WRITE); *state->fr.next++ = desc; state->nmdesc++; - srp_map_desc(state, state->base_dma_addr, state->dma_len, - desc->mr->rkey); + srp_map_desc(state, desc->mr->iova, + desc->mr->length, desc->mr->rkey); - return ib_post_send(ch->qp, &wr, &bad_wr); -} + err = ib_post_send(ch->qp, &wr.wr, &bad_wr); + if (unlikely(err)) + return err; -static int srp_finish_mapping(struct srp_map_state *state, - struct srp_rdma_ch *ch) -{ - struct srp_target_port *target = ch->target; - struct srp_device *dev = target->srp_host->srp_dev; - int ret = 0; - - WARN_ON_ONCE(!dev->use_fast_reg && !dev->use_fmr); - - if (state->npages == 0) - return 0; - - if (state->npages == 1 && target->global_mr) - srp_map_desc(state, state->base_dma_addr, state->dma_len, - target->global_mr->rkey); - else - ret = dev->use_fast_reg ? srp_map_finish_fr(state, ch) : - srp_map_finish_fmr(state, ch); - - if (ret == 0) { - state->npages = 0; - state->dma_len = 0; - } - - return ret; + return n; } static int srp_map_sg_entry(struct srp_map_state *state, @@ -1389,7 +1387,7 @@ static int srp_map_sg_entry(struct srp_map_state *state, while (dma_len) { unsigned offset = dma_addr & ~dev->mr_page_mask; if (state->npages == dev->max_pages_per_mr || offset != 0) { - ret = srp_finish_mapping(state, ch); + ret = srp_map_finish_fmr(state, ch); if (ret) return ret; } @@ -1411,51 +1409,83 @@ static int srp_map_sg_entry(struct srp_map_state *state, */ ret = 0; if (len != dev->mr_page_size) - ret = srp_finish_mapping(state, ch); + ret = srp_map_finish_fmr(state, ch); return ret; } -static int srp_map_sg(struct srp_map_state *state, struct srp_rdma_ch *ch, - struct srp_request *req, struct scatterlist *scat, - int count) +static int srp_map_sg_fmr(struct srp_map_state *state, struct srp_rdma_ch *ch, + struct srp_request *req, struct scatterlist *scat, + int count) { - struct srp_target_port *target = ch->target; - struct srp_device *dev = target->srp_host->srp_dev; struct scatterlist *sg; int i, ret; - state->desc = req->indirect_desc; - state->pages = req->map_page; - if (dev->use_fast_reg) { - state->fr.next = req->fr_list; - state->fr.end = req->fr_list + target->cmd_sg_cnt; - } else if (dev->use_fmr) { - state->fmr.next = req->fmr_list; - state->fmr.end = req->fmr_list + target->cmd_sg_cnt; - } + state->desc = req->indirect_desc; + state->pages = req->map_page; + state->fmr.next = req->fmr_list; + state->fmr.end = req->fmr_list + ch->target->cmd_sg_cnt; - if (dev->use_fast_reg || dev->use_fmr) { - for_each_sg(scat, sg, count, i) { - ret = srp_map_sg_entry(state, ch, sg, i); - if (ret) - goto out; - } - ret = srp_finish_mapping(state, ch); + for_each_sg(scat, sg, count, i) { + ret = srp_map_sg_entry(state, ch, sg, i); if (ret) - goto out; - } else { - for_each_sg(scat, sg, count, i) { - srp_map_desc(state, ib_sg_dma_address(dev->dev, sg), - ib_sg_dma_len(dev->dev, sg), - target->global_mr->rkey); - } + return ret; } + ret = srp_map_finish_fmr(state, ch); + if (ret) + return ret; + req->nmdesc = state->nmdesc; - ret = 0; -out: - return ret; + return 0; +} + +static int srp_map_sg_fr(struct srp_map_state *state, struct srp_rdma_ch *ch, + struct srp_request *req, struct scatterlist *scat, + int count) +{ + state->desc = req->indirect_desc; + state->fr.next = req->fr_list; + state->fr.end = req->fr_list + ch->target->cmd_sg_cnt; + state->sg = scat; + state->sg_nents = scsi_sg_count(req->scmnd); + + while (state->sg_nents) { + int i, n; + + n = srp_map_finish_fr(state, ch); + if (unlikely(n < 0)) + return n; + + state->sg_nents -= n; + for (i = 0; i < n; i++) + state->sg = sg_next(state->sg); + } + + req->nmdesc = state->nmdesc; + + return 0; +} + +static int srp_map_sg_dma(struct srp_map_state *state, struct srp_rdma_ch *ch, + struct srp_request *req, struct scatterlist *scat, + int count) +{ + struct srp_target_port *target = ch->target; + struct srp_device *dev = target->srp_host->srp_dev; + struct scatterlist *sg; + int i; + + state->desc = req->indirect_desc; + for_each_sg(scat, sg, count, i) { + srp_map_desc(state, ib_sg_dma_address(dev->dev, sg), + ib_sg_dma_len(dev->dev, sg), + target->global_mr->rkey); + } + + req->nmdesc = state->nmdesc; + + return 0; } /* @@ -1474,6 +1504,7 @@ static int srp_map_idb(struct srp_rdma_ch *ch, struct srp_request *req, struct srp_map_state state; struct srp_direct_buf idb_desc; u64 idb_pages[1]; + struct scatterlist idb_sg[1]; int ret; memset(&state, 0, sizeof(state)); @@ -1481,20 +1512,32 @@ static int srp_map_idb(struct srp_rdma_ch *ch, struct srp_request *req, state.gen.next = next_mr; state.gen.end = end_mr; state.desc = &idb_desc; - state.pages = idb_pages; - state.pages[0] = (req->indirect_dma_addr & - dev->mr_page_mask); - state.npages = 1; state.base_dma_addr = req->indirect_dma_addr; state.dma_len = idb_len; - ret = srp_finish_mapping(&state, ch); - if (ret < 0) - goto out; + + if (dev->use_fast_reg) { + state.sg = idb_sg; + state.sg_nents = 1; + sg_set_buf(idb_sg, req->indirect_desc, idb_len); + idb_sg->dma_address = req->indirect_dma_addr; /* hack! */ + ret = srp_map_finish_fr(&state, ch); + if (ret < 0) + return ret; + } else if (dev->use_fmr) { + state.pages = idb_pages; + state.pages[0] = (req->indirect_dma_addr & + dev->mr_page_mask); + state.npages = 1; + ret = srp_map_finish_fmr(&state, ch); + if (ret < 0) + return ret; + } else { + return -EINVAL; + } *idb_rkey = idb_desc.key; -out: - return ret; + return 0; } static int srp_map_data(struct scsi_cmnd *scmnd, struct srp_rdma_ch *ch, @@ -1563,7 +1606,12 @@ static int srp_map_data(struct scsi_cmnd *scmnd, struct srp_rdma_ch *ch, target->indirect_size, DMA_TO_DEVICE); memset(&state, 0, sizeof(state)); - srp_map_sg(&state, ch, req, scat, count); + if (dev->use_fast_reg) + srp_map_sg_fr(&state, ch, req, scat, count); + else if (dev->use_fmr) + srp_map_sg_fmr(&state, ch, req, scat, count); + else + srp_map_sg_dma(&state, ch, req, scat, count); /* We've mapped the request, now pull as much of the indirect * descriptor table as we can into the command buffer. If this @@ -3213,7 +3261,7 @@ static ssize_t srp_create_target(struct device *dev, INIT_WORK(&target->tl_err_work, srp_tl_err_work); INIT_WORK(&target->remove_work, srp_remove_work); spin_lock_init(&target->lock); - ret = ib_query_gid(ibdev, host->port, 0, &target->sgid); + ret = ib_query_gid(ibdev, host->port, 0, &target->sgid, NULL); if (ret) goto out; diff --git a/drivers/infiniband/ulp/srp/ib_srp.h b/drivers/infiniband/ulp/srp/ib_srp.h index 3608f2e..87a2a91 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.h +++ b/drivers/infiniband/ulp/srp/ib_srp.h @@ -242,7 +242,6 @@ struct srp_iu { struct srp_fr_desc { struct list_head entry; struct ib_mr *mr; - struct ib_fast_reg_page_list *frpl; }; /** @@ -294,11 +293,17 @@ struct srp_map_state { } gen; }; struct srp_direct_buf *desc; - u64 *pages; + union { + u64 *pages; + struct scatterlist *sg; + }; dma_addr_t base_dma_addr; u32 dma_len; u32 total_len; - unsigned int npages; + union { + unsigned int npages; + int sg_nents; + }; unsigned int nmdesc; unsigned int ndesc; }; diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.c b/drivers/infiniband/ulp/srpt/ib_srpt.c index f6fe041..47c4022 100644 --- a/drivers/infiniband/ulp/srpt/ib_srpt.c +++ b/drivers/infiniband/ulp/srpt/ib_srpt.c @@ -546,7 +546,8 @@ static int srpt_refresh_port(struct srpt_port *sport) sport->sm_lid = port_attr.sm_lid; sport->lid = port_attr.lid; - ret = ib_query_gid(sport->sdev->device, sport->port, 0, &sport->gid); + ret = ib_query_gid(sport->sdev->device, sport->port, 0, &sport->gid, + NULL); if (ret) goto err_query_port; @@ -2822,7 +2823,7 @@ static int srpt_cm_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event) static int srpt_perform_rdmas(struct srpt_rdma_ch *ch, struct srpt_send_ioctx *ioctx) { - struct ib_send_wr wr; + struct ib_rdma_wr wr; struct ib_send_wr *bad_wr; struct rdma_iu *riu; int i; @@ -2850,29 +2851,29 @@ static int srpt_perform_rdmas(struct srpt_rdma_ch *ch, for (i = 0; i < n_rdma; ++i, ++riu) { if (dir == DMA_FROM_DEVICE) { - wr.opcode = IB_WR_RDMA_WRITE; - wr.wr_id = encode_wr_id(i == n_rdma - 1 ? + wr.wr.opcode = IB_WR_RDMA_WRITE; + wr.wr.wr_id = encode_wr_id(i == n_rdma - 1 ? SRPT_RDMA_WRITE_LAST : SRPT_RDMA_MID, ioctx->ioctx.index); } else { - wr.opcode = IB_WR_RDMA_READ; - wr.wr_id = encode_wr_id(i == n_rdma - 1 ? + wr.wr.opcode = IB_WR_RDMA_READ; + wr.wr.wr_id = encode_wr_id(i == n_rdma - 1 ? SRPT_RDMA_READ_LAST : SRPT_RDMA_MID, ioctx->ioctx.index); } - wr.next = NULL; - wr.wr.rdma.remote_addr = riu->raddr; - wr.wr.rdma.rkey = riu->rkey; - wr.num_sge = riu->sge_cnt; - wr.sg_list = riu->sge; + wr.wr.next = NULL; + wr.remote_addr = riu->raddr; + wr.rkey = riu->rkey; + wr.wr.num_sge = riu->sge_cnt; + wr.wr.sg_list = riu->sge; /* only get completion event for the last rdma write */ if (i == (n_rdma - 1) && dir == DMA_TO_DEVICE) - wr.send_flags = IB_SEND_SIGNALED; + wr.wr.send_flags = IB_SEND_SIGNALED; - ret = ib_post_send(ch->qp, &wr, &bad_wr); + ret = ib_post_send(ch->qp, &wr.wr, &bad_wr); if (ret) break; } @@ -2881,11 +2882,11 @@ static int srpt_perform_rdmas(struct srpt_rdma_ch *ch, pr_err("%s[%d]: ib_post_send() returned %d for %d/%d\n", __func__, __LINE__, ret, i, n_rdma); if (ret && i > 0) { - wr.num_sge = 0; - wr.wr_id = encode_wr_id(SRPT_RDMA_ABORT, ioctx->ioctx.index); - wr.send_flags = IB_SEND_SIGNALED; + wr.wr.num_sge = 0; + wr.wr.wr_id = encode_wr_id(SRPT_RDMA_ABORT, ioctx->ioctx.index); + wr.wr.send_flags = IB_SEND_SIGNALED; while (ch->state == CH_LIVE && - ib_post_send(ch->qp, &wr, &bad_wr) != 0) { + ib_post_send(ch->qp, &wr.wr, &bad_wr) != 0) { pr_info("Trying to abort failed RDMA transfer [%d]\n", ioctx->ioctx.index); msleep(1000); |