| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-Id: <200808221840.m7MIeuvB004442@localhost.localdomain>
Date: Fri, 22 Aug 2008 11:40:56 -0700
From: Karen Xie <kxie@...lsio.com>
To: netdev@...r.kernel.org, open-iscsi@...glegroups.com,
linux-scsi@...r.kernel.org, linux-kernel@...r.kernel.org
Cc: jgarzik@...ox.com, davem@...emloft.net, michaelc@...wisc.edu,
swise@...ngridcomputing.com, rdreier@...co.com, daisyc@...ibm.com,
wenxiong@...ibm.com, bhua@...ibm.com, divy@...lsio.com,
dm@...lsio.com, leedom@...lsio.com, kxie@...lsio.com
Subject: [PATCH 4/4 2.6.28] cxgb3i - cxgb3i iscsi driver
[PATCH 4/4 2.6.28] cxgb3i - cxgb3i iscsi driver
From: Karen Xie <kxie@...lsio.com>
cxgb3i iSCSI driver.
Signed-off-by: Karen Xie <kxie@...lsio.com>
---
drivers/scsi/Kconfig | 2
drivers/scsi/Makefile | 1
drivers/scsi/cxgb3i/Kconfig | 6
drivers/scsi/cxgb3i/Makefile | 5
drivers/scsi/cxgb3i/cxgb3i.h | 155 +++
drivers/scsi/cxgb3i/cxgb3i_init.c | 109 ++
drivers/scsi/cxgb3i/cxgb3i_iscsi.c | 824 ++++++++++++++
drivers/scsi/cxgb3i/cxgb3i_offload.c | 1985 ++++++++++++++++++++++++++++++++++
drivers/scsi/cxgb3i/cxgb3i_offload.h | 252 ++++
drivers/scsi/cxgb3i/cxgb3i_ulp2.c | 692 ++++++++++++
drivers/scsi/cxgb3i/cxgb3i_ulp2.h | 106 ++
11 files changed, 4137 insertions(+), 0 deletions(-)
create mode 100644 drivers/scsi/cxgb3i/Kconfig
create mode 100644 drivers/scsi/cxgb3i/Makefile
create mode 100644 drivers/scsi/cxgb3i/cxgb3i.h
create mode 100644 drivers/scsi/cxgb3i/cxgb3i_init.c
create mode 100644 drivers/scsi/cxgb3i/cxgb3i_iscsi.c
create mode 100644 drivers/scsi/cxgb3i/cxgb3i_offload.c
create mode 100644 drivers/scsi/cxgb3i/cxgb3i_offload.h
create mode 100644 drivers/scsi/cxgb3i/cxgb3i_ulp2.c
create mode 100644 drivers/scsi/cxgb3i/cxgb3i_ulp2.h
diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig
index c7f0629..4308df5 100644
--- a/drivers/scsi/Kconfig
+++ b/drivers/scsi/Kconfig
@@ -1779,6 +1779,8 @@ config ZFCP
called zfcp. If you want to compile it as a module, say M here
and read <file:Documentation/kbuild/modules.txt>.
+source "drivers/scsi/cxgb3i/Kconfig"
+
config SCSI_SRP
tristate "SCSI RDMA Protocol helper library"
depends on SCSI && PCI
diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile
index 72fd504..537b4c4 100644
--- a/drivers/scsi/Makefile
+++ b/drivers/scsi/Makefile
@@ -37,6 +37,7 @@ obj-$(CONFIG_SCSI_SRP_ATTRS) += scsi_transport_srp.o
obj-$(CONFIG_SCSI_DH) += device_handler/
obj-$(CONFIG_ISCSI_TCP) += libiscsi.o iscsi_tcp.o
+obj-$(CONFIG_SCSI_CXGB3_ISCSI) += libiscsi.o cxgb3i/
obj-$(CONFIG_INFINIBAND_ISER) += libiscsi.o
obj-$(CONFIG_SCSI_A4000T) += 53c700.o a4000t.o
obj-$(CONFIG_SCSI_ZORRO7XX) += 53c700.o zorro7xx.o
diff --git a/drivers/scsi/cxgb3i/Kconfig b/drivers/scsi/cxgb3i/Kconfig
new file mode 100644
index 0000000..2762814
--- /dev/null
+++ b/drivers/scsi/cxgb3i/Kconfig
@@ -0,0 +1,6 @@
+config SCSI_CXGB3_ISCSI
+ tristate "Chelsio S3xx iSCSI support"
+ select CHELSIO_T3
+ select SCSI_ISCSI_ATTRS
+ ---help---
+ This driver supports iSCSI offload for the Chelsio S3 series devices.
diff --git a/drivers/scsi/cxgb3i/Makefile b/drivers/scsi/cxgb3i/Makefile
new file mode 100644
index 0000000..8c8a894
--- /dev/null
+++ b/drivers/scsi/cxgb3i/Makefile
@@ -0,0 +1,5 @@
+EXTRA_CFLAGS += -I$(TOPDIR)/drivers/net/cxgb3
+
+cxgb3i-y := cxgb3i_init.o cxgb3i_iscsi.o cxgb3i_ulp2.o cxgb3i_offload.o
+
+obj-$(CONFIG_SCSI_CXGB3_ISCSI) += cxgb3i.o
diff --git a/drivers/scsi/cxgb3i/cxgb3i.h b/drivers/scsi/cxgb3i/cxgb3i.h
new file mode 100644
index 0000000..3c44c3c
--- /dev/null
+++ b/drivers/scsi/cxgb3i/cxgb3i.h
@@ -0,0 +1,155 @@
+/*
+ * cxgb3i.h: Chelsio S3xx iSCSI driver.
+ *
+ * Copyright (c) 2008 Chelsio Communications, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation.
+ *
+ * Written by: Karen Xie (kxie@...lsio.com)
+ */
+
+#ifndef __CXGB3I_H__
+#define __CXGB3I_H__
+
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/netdevice.h>
+#include <linux/scatterlist.h>
+
+/* from cxgb3 LLD */
+#include "common.h"
+#include "t3_cpl.h"
+#include "t3cdev.h"
+#include "cxgb3_ctl_defs.h"
+#include "cxgb3_offload.h"
+#include "firmware_exports.h"
+#include "cxgb3i_offload.h"
+
+#define CXGB3I_SCSI_QDEPTH_DFLT 128
+#define ISCSI_PDU_HEADER_MAX (56 + 256) /* bhs + digests + ahs */
+
+struct cxgb3i_adapter;
+struct cxgb3i_hba;
+struct cxgb3i_endpoint;
+
+/**
+ * struct cxgb3i_tag_format - cxgb3i ulp tag for steering pdu payload
+ *
+ * @rsvd_bits: # of bits used by h/w
+ * @rsvd_shift: shift left
+ * @rsvd_mask: bit mask
+ *
+ */
+struct cxgb3i_tag_format {
+ unsigned char idx_bits;
+ unsigned char age_bits;
+ unsigned char rsvd_bits;
+ unsigned char rsvd_shift;
+ u32 rsvd_mask;
+};
+
+/**
+ * struct cxgb3i_ddp_info - cxgb3i direct data placement for pdu payload
+ *
+ * @llimit: lower bound of the page pod memory
+ * @ulimit: upper bound of the page pod memory
+ * @nppods: # of page pod entries
+ * @idx_last: page pod entry last used
+ * @map_lock: lock to synchonize access to the page pod map
+ * @map: page pod map
+ */
+struct cxgb3i_ddp_info {
+ unsigned int llimit;
+ unsigned int ulimit;
+ unsigned int nppods;
+ unsigned int idx_last;
+ spinlock_t map_lock;
+ u8 *map;
+};
+
+struct cxgb3i_hba {
+ struct cxgb3i_adapter *snic;
+ struct net_device *ndev;
+ struct Scsi_Host *shost;
+
+ rwlock_t cconn_rwlock;
+ struct list_head cconn_list;
+};
+
+struct cxgb3i_adapter {
+ struct list_head list_head;
+ spinlock_t lock;
+ struct t3cdev *tdev;
+ struct pci_dev *pdev;
+ unsigned char hba_cnt;
+ struct cxgb3i_hba *hba[MAX_NPORTS];
+
+ unsigned int tx_max_size;
+ unsigned int rx_max_size;
+
+ struct cxgb3i_tag_format tag_format;
+ struct cxgb3i_ddp_info ddp;
+};
+
+struct cxgb3i_conn {
+ struct list_head list_head;
+
+ struct cxgb3i_endpoint *cep;
+ struct iscsi_conn *conn;
+ struct cxgb3i_hba *hba;
+};
+
+struct cxgb3i_endpoint {
+ struct s3_conn *c3cn;
+ struct cxgb3i_hba *hba;
+ struct cxgb3i_conn *cconn;
+};
+
+int cxgb3i_iscsi_init(void);
+void cxgb3i_iscsi_cleanup(void);
+
+struct cxgb3i_adapter *cxgb3i_adapter_find_by_tdev(struct t3cdev *);
+struct cxgb3i_adapter *cxgb3i_adapter_add(struct t3cdev *);
+void cxgb3i_adapter_remove(struct cxgb3i_adapter *);
+int cxgb3i_adapter_ulp_init(struct cxgb3i_adapter *);
+void cxgb3i_adapter_ulp_cleanup(struct cxgb3i_adapter *);
+
+struct cxgb3i_hba *cxgb3i_hba_find_by_netdev(struct net_device *);
+struct cxgb3i_hba *cxgb3i_hba_host_add(struct cxgb3i_adapter *,
+ struct net_device *);
+void cxgb3i_hba_host_remove(struct cxgb3i_hba *);
+
+void cxgb3i_hba_conn_add(struct cxgb3i_conn *, struct cxgb3i_hba *);
+void cxgb3i_hba_conn_remove(struct cxgb3i_conn *);
+
+int cxgb3i_ulp2_init(void);
+void cxgb3i_ulp2_cleanup(void);
+int cxgb3i_conn_ulp_setup(struct cxgb3i_conn *, int, int);
+
+void cxgb3i_ddp_tag_release(struct cxgb3i_adapter *, u32,
+ struct scatterlist *, unsigned int);
+u32 cxgb3i_ddp_tag_reserve(struct cxgb3i_adapter *, unsigned int,
+ u32, unsigned int, struct scatterlist *,
+ unsigned int);
+static inline void cxgb3i_parse_tag(struct cxgb3i_tag_format *format,
+ u32 tag, u32 *rsvd_bits, u32 *sw_bits)
+{
+ if (rsvd_bits)
+ *rsvd_bits = (tag >> format->rsvd_shift) & format->rsvd_mask;
+ if (sw_bits) {
+ *sw_bits = (tag >> (format->rsvd_shift + format->rsvd_bits))
+ << format->rsvd_shift;
+ *sw_bits |= tag & ((1 << format->rsvd_shift) - 1);
+ }
+}
+
+int cxgb3i_conn_ulp2_xmit(struct iscsi_conn *);
+
+void cxgb3i_display_byte_string(char *, unsigned char *, int, int);
+
+#endif
diff --git a/drivers/scsi/cxgb3i/cxgb3i_init.c b/drivers/scsi/cxgb3i/cxgb3i_init.c
new file mode 100644
index 0000000..1c91bb0
--- /dev/null
+++ b/drivers/scsi/cxgb3i/cxgb3i_init.c
@@ -0,0 +1,109 @@
+/* cxgb3i_init.c: Chelsio S3xx iSCSI driver.
+ *
+ * Copyright (c) 2008 Chelsio Communications, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation.
+ *
+ * Written by: Karen Xie (kxie@...lsio.com)
+ */
+
+#include "cxgb3i.h"
+
+#define DRV_MODULE_NAME "cxgb3i"
+#define DRV_MODULE_VERSION "1.0.0"
+#define DRV_MODULE_RELDATE "May 1, 2008"
+
+static char version[] =
+ "Chelsio S3xx iSCSI Driver " DRV_MODULE_NAME
+ " v" DRV_MODULE_VERSION " (" DRV_MODULE_RELDATE ")\n";
+
+MODULE_AUTHOR("Karen Xie <kxie@...lsio.com>");
+MODULE_DESCRIPTION("Chelsio S3xx iSCSI Driver");
+MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_MODULE_VERSION);
+
+static void open_s3_dev(struct t3cdev *);
+static void close_s3_dev(struct t3cdev *);
+cxgb3_cpl_handler_func cxgb3i_cpl_handlers[NUM_CPL_CMDS];
+struct cxgb3_client t3c_client = {
+ .name = "iscsi_cxgb3",
+ .handlers = cxgb3i_cpl_handlers,
+ .add = open_s3_dev,
+ .remove = close_s3_dev,
+};
+
+/**
+ * open_s3_dev - register with cxgb3 LLD
+ * @t3dev cxgb3 adapter instance
+ */
+static void open_s3_dev(struct t3cdev *t3dev)
+{
+ static int vers_printed;
+
+ if (!vers_printed) {
+ printk(KERN_INFO "%s", version);
+ vers_printed = 1;
+ }
+
+ cxgb3i_log_debug("open cxgb3 %s.\n", t3dev->name);
+
+ cxgb3i_sdev_add(t3dev, &t3c_client);
+ cxgb3i_adapter_add(t3dev);
+}
+
+/**
+ * close_s3_dev - de-register with cxgb3 LLD
+ * @t3dev cxgb3 adapter instance
+ */
+static void close_s3_dev(struct t3cdev *t3dev)
+{
+ struct cxgb3i_adapter *snic = cxgb3i_adapter_find_by_tdev(t3dev);
+ cxgb3i_log_debug("close cxgb3 %s.\n", t3dev->name);
+ if (snic)
+ cxgb3i_adapter_remove(snic);
+ cxgb3i_sdev_remove(t3dev);
+}
+
+/**
+ * cxgb3i_init_module - module init entry point
+ *
+ * initialize any driver wide global data structures and register itself
+ * with the cxgb3 module
+ */
+static int __init cxgb3i_init_module(void)
+{
+ int err;
+
+ err = cxgb3i_sdev_init(cxgb3i_cpl_handlers);
+ if (err < 0)
+ return err;
+
+ err = cxgb3i_iscsi_init();
+ if (err < 0)
+ return err;
+
+ err = cxgb3i_ulp2_init();
+ if (err < 0)
+ return err;
+
+ cxgb3_register_client(&t3c_client);
+ return 0;
+}
+
+/**
+ * cxgb3i_exit_module - module cleanup/exit entry point
+ *
+ * go through the driver hba list and for each hba, release any resource held.
+ * and unregisters iscsi transport and the cxgb3 module
+ */
+static void __exit cxgb3i_exit_module(void)
+{
+ cxgb3_unregister_client(&t3c_client);
+ cxgb3i_ulp2_cleanup();
+ cxgb3i_iscsi_cleanup();
+}
+
+module_init(cxgb3i_init_module);
+module_exit(cxgb3i_exit_module);
diff --git a/drivers/scsi/cxgb3i/cxgb3i_iscsi.c b/drivers/scsi/cxgb3i/cxgb3i_iscsi.c
new file mode 100644
index 0000000..19a9a95
--- /dev/null
+++ b/drivers/scsi/cxgb3i/cxgb3i_iscsi.c
@@ -0,0 +1,824 @@
+/* cxgb3i_iscsi.c: Chelsio S3xx iSCSI driver.
+ *
+ * Copyright (c) 2008 Chelsio Communications, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation.
+ *
+ * Written by: Karen Xie (kxie@...lsio.com)
+ */
+
+#include <net/tcp.h>
+#include <scsi/scsi_cmnd.h>
+#include <scsi/scsi_device.h>
+#include <scsi/scsi_eh.h>
+#include <scsi/scsi_host.h>
+#include <scsi/scsi.h>
+#include <scsi/iscsi_proto.h>
+#include <scsi/libiscsi.h>
+#include <scsi/scsi_transport_iscsi.h>
+#include <linux/crypto.h>
+#include "../iscsi_tcp.h"
+
+#include "cxgb3i.h"
+
+static struct scsi_transport_template *cxgb3i_scsi_transport;
+static struct scsi_host_template cxgb3i_host_template;
+static struct iscsi_transport cxgb3i_iscsi_transport;
+
+static LIST_HEAD(cxgb3i_snic_list);
+static DEFINE_RWLOCK(cxgb3i_snic_rwlock);
+
+/**
+ * cxgb3i_adapter_add - initialize a s3 adapter structure and any h/w settings
+ * necessary
+ * @snic: pointer to adapter instance
+ */
+struct cxgb3i_adapter *cxgb3i_adapter_add(struct t3cdev *t3dev)
+{
+ struct cxgb3i_adapter *snic;
+ struct adapter *adapter = tdev2adap(t3dev);
+ int i;
+
+ snic = kzalloc(sizeof(*snic), GFP_KERNEL);
+ if (!snic) {
+ cxgb3i_log_debug("cxgb3 %s, OOM.\n", t3dev->name);
+ return NULL;
+ }
+
+ spin_lock_init(&snic->lock);
+ snic->tdev = t3dev;
+ snic->pdev = adapter->pdev;
+
+ if (cxgb3i_adapter_ulp_init(snic))
+ goto free_snic;
+
+ for_each_port(adapter, i) {
+ snic->hba[i] = cxgb3i_hba_host_add(snic, adapter->port[i]);
+ if (!snic->hba[i])
+ goto ulp_cleanup;
+ }
+ snic->hba_cnt = adapter->params.nports;
+
+ /* add to the list */
+ write_lock(&cxgb3i_snic_rwlock);
+ list_add_tail(&snic->list_head, &cxgb3i_snic_list);
+ write_unlock(&cxgb3i_snic_rwlock);
+
+ return snic;
+
+ulp_cleanup:
+ cxgb3i_adapter_ulp_cleanup(snic);
+free_snic:
+ kfree(snic);
+ return NULL;
+}
+
+/**
+ * cxgb3i_snic_cleanup - release all the resources held and cleanup any h/w
+ * settings necessary
+ * @snic: pointer to adapter instance
+ */
+void cxgb3i_adapter_remove(struct cxgb3i_adapter *snic)
+{
+ int i;
+
+ /* remove from the list */
+ write_lock(&cxgb3i_snic_rwlock);
+ list_del(&snic->list_head);
+ write_unlock(&cxgb3i_snic_rwlock);
+
+ for (i = 0; i < snic->hba_cnt; i++) {
+ if (snic->hba[i]) {
+ cxgb3i_hba_host_remove(snic->hba[i]);
+ snic->hba[i] = NULL;
+ }
+ }
+
+ /* release ddp resources */
+ cxgb3i_adapter_ulp_cleanup(snic);
+ kfree(snic);
+}
+
+struct cxgb3i_adapter *cxgb3i_adapter_find_by_tdev(struct t3cdev *t3dev)
+{
+ struct cxgb3i_adapter *snic;
+
+ read_lock(&cxgb3i_snic_rwlock);
+ list_for_each_entry(snic, &cxgb3i_snic_list, list_head) {
+ if (snic->tdev == t3dev) {
+ read_unlock(&cxgb3i_snic_rwlock);
+ return snic;
+ }
+ }
+ read_unlock(&cxgb3i_snic_rwlock);
+
+ return NULL;
+}
+
+struct cxgb3i_hba *cxgb3i_hba_find_by_netdev(struct net_device *ndev)
+{
+ struct cxgb3i_adapter *snic;
+ int i;
+
+ read_lock(&cxgb3i_snic_rwlock);
+ list_for_each_entry(snic, &cxgb3i_snic_list, list_head) {
+ for (i = 0; i < snic->hba_cnt; i++) {
+ if (snic->hba[i]->ndev == ndev) {
+ read_unlock(&cxgb3i_snic_rwlock);
+ return (snic->hba[i]);
+ }
+ }
+ }
+ read_unlock(&cxgb3i_snic_rwlock);
+ return NULL;
+}
+
+void cxgb3i_hba_conn_add(struct cxgb3i_conn *cconn, struct cxgb3i_hba *hba)
+{
+ cconn->hba = hba;
+ write_lock(&hba->cconn_rwlock);
+ list_add_tail(&cconn->list_head, &hba->cconn_list);
+ write_unlock(&hba->cconn_rwlock);
+}
+
+void cxgb3i_hba_conn_remove(struct cxgb3i_conn *cconn)
+{
+ struct cxgb3i_hba *hba = cconn->hba;
+
+ if (hba) {
+ write_lock(&hba->cconn_rwlock);
+ list_del(&cconn->list_head);
+ write_unlock(&hba->cconn_rwlock);
+ }
+}
+
+struct cxgb3i_hba *cxgb3i_hba_host_add(struct cxgb3i_adapter *snic,
+ struct net_device *ndev)
+{
+ struct cxgb3i_hba *hba;
+ struct Scsi_Host *shost;
+ int err;
+
+ shost = iscsi_host_alloc(&cxgb3i_host_template,
+ sizeof(struct cxgb3i_hba),
+ CXGB3I_SCSI_QDEPTH_DFLT);
+ if (!shost) {
+ cxgb3i_log_info("iscsi_host_alloc failed.\n");
+ return NULL;
+ }
+
+ shost->transportt = cxgb3i_scsi_transport;
+ shost->max_lun = 512;
+ shost->max_id = 0;
+ shost->max_channel = 0;
+ shost->max_cmd_len = 16;
+
+ hba = iscsi_host_priv(shost);
+ INIT_LIST_HEAD(&hba->cconn_list);
+ rwlock_init(&hba->cconn_rwlock);
+ hba->snic = snic;
+ hba->ndev = ndev;
+ hba->shost = shost;
+
+ pci_dev_get(snic->pdev);
+ err = iscsi_host_add(shost, &snic->pdev->dev);
+ if (err) {
+ cxgb3i_log_info("iscsi_host_add failed.\n");
+ goto pci_dev_put;
+ }
+
+ cxgb3i_log_debug("shost 0x%p, hba 0x%p, no %u.\n",
+ shost, hba, shost->host_no);
+
+ return hba;
+
+pci_dev_put:
+ pci_dev_put(snic->pdev);
+ scsi_host_put(shost);
+ return NULL;
+}
+
+void cxgb3i_hba_host_remove(struct cxgb3i_hba *hba)
+{
+ if (hba->shost) {
+ cxgb3i_log_debug("shost 0x%p, hba 0x%p, no %u.\n",
+ hba->shost, hba, hba->shost->host_no);
+ iscsi_host_remove(hba->shost);
+ pci_dev_put(hba->snic->pdev);
+ /* cleanup connections ? */
+ iscsi_host_free(hba->shost);
+ }
+}
+
+/**
+ * cxgb3i_ep_connect - establish TCP connection to target portal
+ * @dst_addr: target IP address
+ * @non_blocking: blocking or non-blocking call
+ *
+ * Initiates a TCP/IP connection to the dst_addr
+ */
+static struct iscsi_endpoint *cxgb3i_ep_connect(struct sockaddr *dst_addr,
+ int non_blocking)
+{
+ struct iscsi_endpoint *ep;
+ struct cxgb3i_endpoint *cep;
+ struct cxgb3i_hba *hba;
+ struct s3_conn *c3cn = NULL;
+ int err = 0;
+
+ c3cn = cxgb3i_c3cn_create();
+ if (!c3cn) {
+ cxgb3i_log_info("ep connect OOM.\n");
+ err = -ENOMEM;
+ goto release_conn;
+ }
+
+ err = cxgb3i_c3cn_connect(c3cn, (struct sockaddr_in *)dst_addr);
+ if (err < 0) {
+ cxgb3i_log_info("ep connect failed.\n");
+ goto release_conn;
+ }
+ hba = cxgb3i_hba_find_by_netdev(c3cn->dst_cache->dev);
+ if (!hba) {
+ err = -ENOSPC;
+ cxgb3i_log_info("NOT going through cxgbi device.\n");
+ goto release_conn;
+ }
+ if (c3cn_in_state(c3cn, C3CN_STATE_CLOSE)) {
+ err = -ENOSPC;
+ cxgb3i_log_info("ep connect unable to connect.\n");
+ goto release_conn;
+ }
+
+ ep = iscsi_create_endpoint(sizeof(*cep));
+ if (!ep) {
+ err = -ENOMEM;
+ cxgb3i_log_info("iscsi alloc ep, OOM.\n");
+ goto release_conn;
+ }
+ cep = ep->dd_data;
+ cep->c3cn = c3cn;
+ cep->hba = hba;
+
+ cxgb3i_log_debug("ep 0x%p, 0x%p, c3cn 0x%p, hba 0x%p.\n",
+ ep, cep, c3cn, hba);
+ return ep;
+
+release_conn:
+ cxgb3i_log_debug("conn failed release.\n");
+ if (c3cn)
+ c3cn_release(c3cn);
+ return ERR_PTR(err);
+}
+
+/**
+ * cxgb3i_ep_poll - polls for TCP connection establishement
+ * @ep: TCP connection (endpoint) handle
+ * @timeout_ms: timeout value in milli secs
+ *
+ * polls for TCP connect request to complete
+ */
+static int cxgb3i_ep_poll(struct iscsi_endpoint *ep, int timeout_ms)
+{
+ struct cxgb3i_endpoint *cep = (struct cxgb3i_endpoint *)ep->dd_data;
+ struct s3_conn *c3cn = cep->c3cn;
+
+ cxgb3i_log_debug("ep 0x%p, timeout %d, c3cn 0x%p, state 0x%x.\n",
+ ep, timeout_ms, c3cn, c3cn->state);
+
+ if (!c3cn_in_state(c3cn, C3CN_STATE_ESTABLISHED)) {
+ cxgb3i_log_info("not in established state.\n");
+ return 0;
+ }
+ return 1;
+}
+
+/**
+ * cxgb3i_ep_disconnect - teardown TCP connection
+ * @ep: TCP connection (endpoint) handle
+ *
+ * teardown TCP connection
+ */
+static void cxgb3i_ep_disconnect(struct iscsi_endpoint *ep)
+{
+ struct cxgb3i_endpoint *cep = (struct cxgb3i_endpoint *)ep->dd_data;
+ struct cxgb3i_conn *cconn = cep->cconn;
+
+ cxgb3i_log_debug("ep 0x%p, cep 0x%p.\n", ep, cep);
+
+ if (cconn && cconn->conn) {
+ struct iscsi_conn *conn = cconn->conn;
+ struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
+ write_lock_bh(&cep->c3cn->callback_lock);
+ cep->c3cn->user_data = NULL;
+ set_bit(ISCSI_SUSPEND_BIT, &conn->suspend_rx);
+ cconn->cep = NULL;
+ tcp_conn->sock = NULL;
+ write_unlock_bh(&cep->c3cn->callback_lock);
+ }
+
+ cxgb3i_log_debug("ep 0x%p, cep 0x%p, release c3cn 0x%p.\n",
+ ep, cep, cep->c3cn);
+ c3cn_release(cep->c3cn);
+ iscsi_destroy_endpoint(ep);
+}
+
+/**
+ * cxgb3i_session_create - create a new iscsi session
+ * @cmds_max: max # of commands
+ * @qdepth: scsi queue depth
+ * @initial_cmdsn: initial iscsi CMDSN for this session
+ * @host_no: pointer to return host no
+ *
+ * Creates a new iSCSI session
+ */
+static struct iscsi_cls_session *cxgb3i_session_create(struct iscsi_endpoint
+ *ep, uint16_t cmds_max,
+ uint16_t qdepth,
+ uint32_t initial_cmdsn,
+ uint32_t *host_no)
+{
+ struct cxgb3i_endpoint *cep;
+ struct cxgb3i_hba *hba;
+ struct Scsi_Host *shost;
+ struct iscsi_cls_session *cls_session;
+ struct iscsi_session *session;
+ int i;
+
+ if (!ep) {
+ cxgb3i_log_error("%s, missing endpoint.\n", __func__);
+ return NULL;
+ }
+
+ cep = (struct cxgb3i_endpoint *)ep->dd_data;
+ hba = cep->hba;
+ shost = hba->shost;
+ cxgb3i_log_debug("ep 0x%p, cep 0x%p, hba 0x%p.\n", ep, cep, hba);
+ BUG_ON(hba != iscsi_host_priv(shost));
+
+ *host_no = shost->host_no;
+
+ cls_session = iscsi_session_setup(&cxgb3i_iscsi_transport, shost,
+ cmds_max,
+ sizeof(struct iscsi_tcp_task),
+ initial_cmdsn, ISCSI_MAX_TARGET);
+ if (!cls_session)
+ return NULL;
+
+ session = cls_session->dd_data;
+
+ for (i = 0; i < session->cmds_max; i++) {
+ struct iscsi_task *task = session->cmds[i];
+ struct iscsi_tcp_task *tcp_task = task->dd_data;
+
+ task->hdr = &tcp_task->hdr.cmd_hdr;
+ task->hdr_max = sizeof(tcp_task->hdr) - ISCSI_DIGEST_SIZE;
+ }
+
+ if (iscsi_r2tpool_alloc(session))
+ goto remove_session;
+
+ return cls_session;
+
+remove_session:
+ iscsi_session_teardown(cls_session);
+ return NULL;
+}
+
+/**
+ * cxgb3i_session_destroy - destroys iscsi session
+ * @cls_session: pointer to iscsi cls session
+ *
+ * Destroys an iSCSI session instance and releases its all resources held
+ */
+static void cxgb3i_session_destroy(struct iscsi_cls_session *cls_session)
+{
+ cxgb3i_log_debug("sess 0x%p.\n", cls_session);
+ iscsi_r2tpool_free(cls_session->dd_data);
+ iscsi_session_teardown(cls_session);
+}
+
+/**
+ * cxgb3i_conn_create - create iscsi connection instance
+ * @cls_session: pointer to iscsi cls session
+ * @cid: iscsi cid
+ *
+ * Creates a new iSCSI connection instance for a given session
+ */
+static struct iscsi_cls_conn *cxgb3i_conn_create(struct iscsi_cls_session
+ *cls_session, uint32_t cid)
+{
+ struct iscsi_cls_conn *cls_conn;
+ struct iscsi_conn *conn;
+ struct iscsi_tcp_conn *tcp_conn;
+ struct cxgb3i_conn *cconn;
+
+ cxgb3i_log_debug("sess 0x%p, cid %u.\n", cls_session, cid);
+
+ cls_conn = iscsi_conn_setup(cls_session,
+ sizeof(*tcp_conn) + sizeof(*cconn), cid);
+ if (!cls_conn)
+ return NULL;
+ conn = cls_conn->dd_data;
+
+ conn->max_xmit_dlength = conn->max_recv_dlength = 16224 - 56 - 256;
+
+ tcp_conn = conn->dd_data;
+ tcp_conn->iscsi_conn = conn;
+
+ cconn = (struct cxgb3i_conn *)(tcp_conn + 1);
+ cconn->conn = conn;
+
+ return cls_conn;
+}
+
+/**
+ * cxgb3i_conn_bind - binds iscsi sess, conn and endpoint together
+ * @cls_session: pointer to iscsi cls session
+ * @cls_conn: pointer to iscsi cls conn
+ * @transport_eph: 64-bit EP handle
+ * @is_leading: leading connection on this session?
+ *
+ * Binds together an iSCSI session, an iSCSI connection and a
+ * TCP connection. This routine returns error code if the TCP
+ * connection does not belong on the device iSCSI sess/conn is bound
+ */
+
+static int cxgb3i_conn_bind(struct iscsi_cls_session *cls_session,
+ struct iscsi_cls_conn *cls_conn,
+ uint64_t transport_eph, int is_leading)
+{
+ struct iscsi_conn *conn = cls_conn->dd_data;
+ struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
+ struct cxgb3i_conn *cconn = (struct cxgb3i_conn *)(tcp_conn + 1);
+ struct iscsi_endpoint *ep;
+ struct cxgb3i_endpoint *cep;
+ struct s3_conn *c3cn;
+ int err;
+
+ ep = iscsi_lookup_endpoint(transport_eph);
+ if (!ep)
+ return -EINVAL;
+
+ cxgb3i_log_debug("ep 0x%p, cls sess 0x%p, cls conn 0x%p.\n",
+ ep, cls_session, cls_conn);
+
+ err = iscsi_conn_bind(cls_session, cls_conn, is_leading);
+ if (err)
+ return -EINVAL;
+
+ cep = (struct cxgb3i_endpoint *)ep->dd_data;
+ c3cn = cep->c3cn;
+
+ read_lock(&c3cn->callback_lock);
+ tcp_conn->sock = (struct socket *)c3cn;
+ c3cn->user_data = conn;
+ read_unlock(&c3cn->callback_lock);
+
+ cconn->hba = cep->hba;
+ cconn->cep = cep;
+ cep->cconn = cconn;
+
+ conn->max_recv_dlength = cconn->hba->snic->rx_max_size - ISCSI_PDU_HEADER_MAX;
+ conn->max_xmit_dlength = cconn->hba->snic->tx_max_size - ISCSI_PDU_HEADER_MAX;
+
+ spin_lock_bh(&conn->session->lock);
+ sprintf(conn->portal_address, NIPQUAD_FMT,
+ NIPQUAD(c3cn->daddr.sin_addr.s_addr));
+ conn->portal_port = ntohs(c3cn->daddr.sin_port);
+ spin_unlock_bh(&conn->session->lock);
+
+ iscsi_tcp_hdr_recv_prep(tcp_conn);
+
+ return 0;
+}
+
+/**
+ * cxgb3i_conn_flush - flush tx
+ * @conn: pointer to iscsi conn
+ */
+static int cxgb3i_conn_flush(struct iscsi_conn *conn)
+{
+ struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
+ struct iscsi_segment *segment = &tcp_conn->out.segment;
+
+ cxgb3i_log_debug("conn 0x%p, segment sent %u/%u.\n",
+ conn, segment->total_copied, segment->total_size);
+
+ if (segment->total_copied < segment->total_size)
+ return cxgb3i_conn_ulp2_xmit(conn);
+ return 0;
+}
+
+/**
+ * cxgb3i_conn_get_param - return iscsi connection parameter to caller
+ * @cls_conn: pointer to iscsi cls conn
+ * @param: parameter type identifier
+ * @buf: buffer pointer
+ *
+ * returns iSCSI connection parameters
+ */
+static int cxgb3i_conn_get_param(struct iscsi_cls_conn *cls_conn,
+ enum iscsi_param param, char *buf)
+{
+ struct iscsi_conn *conn = cls_conn->dd_data;
+ int len;
+
+ cxgb3i_log_debug("cls_conn 0x%p, param %d.\n", cls_conn, param);
+
+ switch (param) {
+ case ISCSI_PARAM_CONN_PORT:
+ spin_lock_bh(&conn->session->lock);
+ len = sprintf(buf, "%hu\n", conn->portal_port);
+ spin_unlock_bh(&conn->session->lock);
+ break;
+ case ISCSI_PARAM_CONN_ADDRESS:
+ spin_lock_bh(&conn->session->lock);
+ len = sprintf(buf, "%s\n", conn->portal_address);
+ spin_unlock_bh(&conn->session->lock);
+ break;
+ default:
+ return iscsi_conn_get_param(cls_conn, param, buf);
+ }
+
+ return len;
+}
+
+static int cxgb3i_conn_set_param(struct iscsi_cls_conn *cls_conn,
+ enum iscsi_param param, char *buf, int buflen)
+{
+ struct iscsi_conn *conn = cls_conn->dd_data;
+ struct iscsi_session *session = conn->session;
+ struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
+ struct cxgb3i_conn *cconn = (struct cxgb3i_conn *)(tcp_conn + 1);
+ int value, err = 0;
+
+ switch (param) {
+ case ISCSI_PARAM_HDRDGST_EN:
+ err = iscsi_set_param(cls_conn, param, buf, buflen);
+ if (!err && conn->hdrdgst_en)
+ cxgb3i_conn_ulp_setup(cconn, conn->hdrdgst_en,
+ conn->datadgst_en);
+ break;
+ case ISCSI_PARAM_DATADGST_EN:
+ err = iscsi_set_param(cls_conn, param, buf, buflen);
+ if (!err && conn->datadgst_en)
+ cxgb3i_conn_ulp_setup(cconn, conn->hdrdgst_en,
+ conn->datadgst_en);
+ break;
+ case ISCSI_PARAM_MAX_R2T:
+ sscanf(buf, "%d", &value);
+ if (value <= 0 || !is_power_of_2(value))
+ return -EINVAL;
+ if (session->max_r2t == value)
+ break;
+ iscsi_r2tpool_free(session);
+ err = iscsi_set_param(cls_conn, param, buf, buflen);
+ if (!err && iscsi_r2tpool_alloc(session))
+ return -ENOMEM;
+ case ISCSI_PARAM_MAX_RECV_DLENGTH:
+ err = iscsi_set_param(cls_conn, param, buf, buflen);
+ cxgb3i_log_debug("MAX_RECV %u.\n", conn->max_recv_dlength);
+ break;
+ case ISCSI_PARAM_MAX_XMIT_DLENGTH:
+ err = iscsi_set_param(cls_conn, param, buf, buflen);
+ cxgb3i_log_debug("MAX_XMIT %u.\n", conn->max_xmit_dlength);
+ break;
+ default:
+ return iscsi_set_param(cls_conn, param, buf, buflen);
+ }
+ return err;
+}
+
+/**
+ * cxgb3i_host_get_param - returns host (adapter) related parameters
+ * @shost: scsi host pointer
+ * @param: parameter type identifier
+ * @buf: buffer pointer
+ */
+static int cxgb3i_host_get_param(struct Scsi_Host *shost,
+ enum iscsi_host_param param, char *buf)
+{
+ struct cxgb3i_hba *hba = iscsi_host_priv(shost);
+ int i;
+ int len = 0;
+
+ switch (param) {
+ case ISCSI_HOST_PARAM_HWADDRESS:
+ for (i = 0; i < 6; i++)
+ len +=
+ sprintf(buf + len, "%02x.",
+ hba->ndev->dev_addr[i]);
+ len--;
+ buf[len] = '\0';
+ break;
+ case ISCSI_HOST_PARAM_NETDEV_NAME:
+ len = sprintf(buf, "%s\n", hba->ndev->name);
+ break;
+ default:
+ return iscsi_host_get_param(shost, param, buf);
+ }
+ return len;
+}
+
+/**
+ * cxgb3i_conn_get_stats - returns iSCSI stats
+ * @cls_conn: pointer to iscsi cls conn
+ * @stats: pointer to iscsi statistic struct
+ */
+static void cxgb3i_conn_get_stats(struct iscsi_cls_conn *cls_conn,
+ struct iscsi_stats *stats)
+{
+ struct iscsi_conn *conn = cls_conn->dd_data;
+
+ stats->txdata_octets = conn->txdata_octets;
+ stats->rxdata_octets = conn->rxdata_octets;
+ stats->scsicmd_pdus = conn->scsicmd_pdus_cnt;
+ stats->dataout_pdus = conn->dataout_pdus_cnt;
+ stats->scsirsp_pdus = conn->scsirsp_pdus_cnt;
+ stats->datain_pdus = conn->datain_pdus_cnt;
+ stats->r2t_pdus = conn->r2t_pdus_cnt;
+ stats->tmfcmd_pdus = conn->tmfcmd_pdus_cnt;
+ stats->tmfrsp_pdus = conn->tmfrsp_pdus_cnt;
+ stats->digest_err = 0;
+ stats->timeout_err = 0;
+ stats->custom_length = 1;
+ strcpy(stats->custom[0].desc, "eh_abort_cnt");
+ stats->custom[0].value = conn->eh_abort_cnt;
+}
+
+static inline u32 tag_base(struct cxgb3i_tag_format *format,
+ unsigned int idx, unsigned int age)
+{
+ u32 sw_bits = idx | (age << format->idx_bits);
+ u32 tag = sw_bits >> format->rsvd_shift;
+ tag <<= format->rsvd_bits + format->rsvd_shift;
+ tag |= sw_bits & ((1 << format->rsvd_shift) - 1);
+ return tag;
+}
+
+static void cxgb3i_parse_itt(struct iscsi_conn *conn, itt_t itt,
+ int *idx, int *age)
+{
+ struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
+ struct cxgb3i_conn *cconn = (struct cxgb3i_conn *)(tcp_conn + 1);
+ struct cxgb3i_adapter *snic = cconn->hba->snic;
+ u32 sw_bits;
+
+ cxgb3i_parse_tag(&snic->tag_format, itt, NULL, &sw_bits);
+ if (idx)
+ *idx = sw_bits & ISCSI_ITT_MASK;
+ if (age)
+ *age = (sw_bits >> snic->tag_format.idx_bits) & ISCSI_AGE_MASK;
+}
+
+static int cxgb3i_reserve_itt(struct iscsi_task *task, itt_t *hdr_itt)
+{
+ struct scsi_cmnd *sc = task->sc;
+ struct iscsi_conn *conn = task->conn;
+ struct iscsi_session *sess = conn->session;
+ struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
+ struct cxgb3i_conn *cconn = (struct cxgb3i_conn *)(tcp_conn + 1);
+ struct cxgb3i_adapter *snic = cconn->hba->snic;
+ u32 sw_tag = tag_base(&snic->tag_format, task->itt, sess->age);
+ u32 tag = RESERVED_ITT;
+
+ if (sc && (sc->sc_data_direction == DMA_FROM_DEVICE)) {
+ struct s3_conn *c3cn =
+ (struct s3_conn *)(tcp_conn->sock);
+ tag =
+ cxgb3i_ddp_tag_reserve(snic, c3cn->tid, sw_tag,
+ scsi_out(sc)->length,
+ scsi_out(sc)->table.sgl,
+ scsi_out(sc)->table.nents);
+ }
+ if (tag == RESERVED_ITT)
+ tag = sw_tag | (snic->tag_format.rsvd_mask <<
+ snic->tag_format.rsvd_shift);
+ *hdr_itt = htonl(tag);
+ return 0;
+}
+
+static void cxgb3i_release_itt(struct iscsi_task *task, itt_t hdr_itt)
+{
+ struct scsi_cmnd *sc = task->sc;
+ struct iscsi_conn *conn = task->conn;
+ struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
+ struct cxgb3i_conn *cconn = (struct cxgb3i_conn *)(tcp_conn + 1);
+ struct cxgb3i_adapter *snic = cconn->hba->snic;
+
+ hdr_itt = ntohl(hdr_itt);
+ if (sc && (sc->sc_data_direction == DMA_FROM_DEVICE))
+ cxgb3i_ddp_tag_release(snic, hdr_itt,
+ scsi_out(sc)->table.sgl,
+ scsi_out(sc)->table.nents);
+}
+
+/**
+ * cxgb3i_host_template -- Scsi_Host_Template structure
+ * used when registering with the scsi mid layer
+ */
+static struct scsi_host_template cxgb3i_host_template = {
+ .module = THIS_MODULE,
+ .name = "Chelsio S3xx iSCSI Initiator",
+ .proc_name = "cxgb3i",
+ .queuecommand = iscsi_queuecommand,
+ .change_queue_depth = iscsi_change_queue_depth,
+ .can_queue = 128 * (ISCSI_DEF_XMIT_CMDS_MAX - 1),
+ .sg_tablesize = SG_ALL,
+ .max_sectors = 0xFFFF,
+ .cmd_per_lun = ISCSI_DEF_CMD_PER_LUN,
+ .eh_abort_handler = iscsi_eh_abort,
+ .eh_device_reset_handler = iscsi_eh_device_reset,
+ .eh_target_reset_handler = iscsi_eh_target_reset,
+ .use_clustering = DISABLE_CLUSTERING,
+ .slave_alloc = iscsi_slave_alloc,
+ .this_id = -1,
+};
+
+static struct iscsi_transport cxgb3i_iscsi_transport = {
+ .owner = THIS_MODULE,
+ .name = "cxgb3i",
+ .caps = CAP_RECOVERY_L0 | CAP_MULTI_R2T | CAP_HDRDGST
+ | CAP_DATADGST | CAP_DIGEST_OFFLOAD,
+ .param_mask = ISCSI_MAX_RECV_DLENGTH |
+ ISCSI_MAX_XMIT_DLENGTH |
+ ISCSI_HDRDGST_EN |
+ ISCSI_DATADGST_EN |
+ ISCSI_INITIAL_R2T_EN |
+ ISCSI_MAX_R2T |
+ ISCSI_IMM_DATA_EN |
+ ISCSI_FIRST_BURST |
+ ISCSI_MAX_BURST |
+ ISCSI_PDU_INORDER_EN |
+ ISCSI_DATASEQ_INORDER_EN |
+ ISCSI_ERL |
+ ISCSI_CONN_PORT |
+ ISCSI_CONN_ADDRESS |
+ ISCSI_EXP_STATSN |
+ ISCSI_PERSISTENT_PORT |
+ ISCSI_PERSISTENT_ADDRESS |
+ ISCSI_TARGET_NAME | ISCSI_TPGT |
+ ISCSI_USERNAME | ISCSI_PASSWORD |
+ ISCSI_USERNAME_IN | ISCSI_PASSWORD_IN |
+ ISCSI_FAST_ABORT | ISCSI_ABORT_TMO |
+ ISCSI_LU_RESET_TMO |
+ ISCSI_PING_TMO | ISCSI_RECV_TMO |
+ ISCSI_IFACE_NAME | ISCSI_INITIATOR_NAME,
+ .host_param_mask = ISCSI_HOST_HWADDRESS | ISCSI_HOST_IPADDRESS |
+ ISCSI_HOST_INITIATOR_NAME | ISCSI_HOST_NETDEV_NAME,
+ .get_host_param = cxgb3i_host_get_param,
+ /* session management */
+ .create_session = cxgb3i_session_create,
+ .destroy_session = cxgb3i_session_destroy,
+ .get_session_param = iscsi_session_get_param,
+ /* connection management */
+ .create_conn = cxgb3i_conn_create,
+ .bind_conn = cxgb3i_conn_bind,
+ .destroy_conn = iscsi_conn_teardown,
+ .start_conn = iscsi_conn_start,
+ .stop_conn = iscsi_conn_stop,
+ .flush_conn = cxgb3i_conn_flush,
+ .get_conn_param = cxgb3i_conn_get_param,
+ .set_param = cxgb3i_conn_set_param,
+ .get_stats = cxgb3i_conn_get_stats,
+ /* pdu xmit req. from user space */
+ .send_pdu = iscsi_conn_send_pdu,
+ /* task */
+ .init_task = iscsi_tcp_task_init,
+ .xmit_task = iscsi_tcp_task_xmit,
+ .cleanup_task = iscsi_tcp_cleanup_task,
+ .parse_itt = cxgb3i_parse_itt,
+ .reserve_itt = cxgb3i_reserve_itt,
+ .release_itt = cxgb3i_release_itt,
+ /* TCP connect/disconnect */
+ .ep_connect = cxgb3i_ep_connect,
+ .ep_poll = cxgb3i_ep_poll,
+ .ep_disconnect = cxgb3i_ep_disconnect,
+ /* Error recovery timeout call */
+ .session_recovery_timedout = iscsi_session_recovery_timedout,
+};
+
+int cxgb3i_iscsi_init(void)
+{
+ cxgb3i_scsi_transport =
+ iscsi_register_transport(&cxgb3i_iscsi_transport);
+ if (!cxgb3i_scsi_transport) {
+ cxgb3i_log_error("Could not register cxgb3i transport.\n");
+ return -ENODEV;
+ }
+ cxgb3i_log_debug("cxgb3i transport 0x%p.\n", cxgb3i_scsi_transport);
+ return 0;
+}
+
+void cxgb3i_iscsi_cleanup(void)
+{
+ if (cxgb3i_scsi_transport) {
+ cxgb3i_log_debug("cxgb3i transport 0x%p.\n",
+ cxgb3i_scsi_transport);
+ iscsi_unregister_transport(&cxgb3i_iscsi_transport);
+ cxgb3i_scsi_transport = NULL;
+ }
+}
diff --git a/drivers/scsi/cxgb3i/cxgb3i_offload.c b/drivers/scsi/cxgb3i/cxgb3i_offload.c
new file mode 100644
index 0000000..d75e0fd
--- /dev/null
+++ b/drivers/scsi/cxgb3i/cxgb3i_offload.c
@@ -0,0 +1,1985 @@
+/*
+ * Copyright (C) 2003-2008 Chelsio Communications. All rights reserved.
+ *
+ * Written by Dimitris Michailidis (dm@...lsio.com)
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the LICENSE file included in this
+ * release for licensing terms and conditions.
+ */
+
+#include <linux/if_vlan.h>
+#include <linux/version.h>
+
+#include "cxgb3_defs.h"
+#include "cxgb3_ctl_defs.h"
+#include "firmware_exports.h"
+#include "cxgb3i_offload.h"
+#include "cxgb3i_ulp2.h"
+
+static int rcv_win = 256 * 1024;
+module_param(rcv_win, int, 0644);
+MODULE_PARM_DESC(rcv_win, "TCP receive window in bytes (default=256KB)");
+
+static int snd_win = 64 * 1024;
+module_param(snd_win, int, 0644);
+MODULE_PARM_DESC(snd_win, "TCP send window in bytes (default=64KB)");
+
+static int rx_credit_thres = 10 * 1024;
+module_param(rx_credit_thres, int, 0644);
+MODULE_PARM_DESC(snd_win, "RX credits return threshold in bytes (default=10KB)");
+
+static unsigned int max_connect = 8 * 1024;
+module_param(max_connect, uint, 0644);
+MODULE_PARM_DESC(max_connect, "Max. # of connections (default=8092)");
+
+static unsigned int sport_base = 20000;
+module_param(sport_base, uint, 0644);
+MODULE_PARM_DESC(sport_start, "starting port number (default=20000)");
+
+#ifdef __DEBUG_C3CN_CONN__
+#define c3cn_conn_debug cxgb3i_log_debug
+#else
+#define c3cn_conn_debug(fmt...)
+#endif
+
+#ifdef __DEBUG_C3CN_TX__
+#define c3cn_tx_debug cxgb3i_log_debug
+#else
+#define c3cn_tx_debug(fmt...)
+#endif
+
+/* minimal port allocation management scheme */
+spinlock_t sport_map_lock;
+unsigned int sport_map_next = 0;
+unsigned char *sport_map = NULL;
+
+/*
+ * Find a free source port in our allocation map. We use a very simple rotor
+ * scheme to look for the next free port.
+ *
+ * If a source port has been specified make sure that it doesn't collide with
+ * our normal source port allocation map. If it's outside the range of our
+ * allocation scheme just let them use it.
+ */
+static int c3cn_get_port(struct s3_conn *c3cn)
+{
+ unsigned int start;
+
+ if (!sport_map)
+ goto error_out;
+
+ if (c3cn->saddr.sin_port != 0) {
+ int sport = ntohs(c3cn->saddr.sin_port) - sport_base;
+ int err = 0;
+
+ if (sport < 0 || sport >= max_connect)
+ return 0;
+ spin_lock(&sport_map_lock);
+ err = __test_and_set_bit(sport, sport_map);
+ spin_unlock(&sport_map_lock);
+ return (err ? -EADDRINUSE : 0);
+ }
+
+ spin_lock(&sport_map_lock);
+ start = sport_map_next;
+ do {
+ unsigned int new = sport_map_next;
+ if (++sport_map_next >= max_connect)
+ sport_map_next = 0;
+ if (!(__test_and_set_bit(new, sport_map))) {
+ spin_unlock(&sport_map_lock);
+ c3cn->saddr.sin_port = htons(sport_base + new);
+ return 0;
+ }
+ } while (sport_map_next != start);
+ spin_unlock(&sport_map_lock);
+
+error_out:
+ return -EADDRNOTAVAIL;
+}
+
+/*
+ * Deallocate a source port from the allocation map. If the source port is
+ * outside our allocation range just return -- the caller is responsible for
+ * keeping track of their port usage outside of our allocation map.
+ */
+static void c3cn_put_port(struct s3_conn *c3cn)
+{
+ int old = ntohs(c3cn->saddr.sin_port) - sport_base;
+ c3cn->saddr.sin_port = 0;
+
+ if (old < 0 || old >= max_connect)
+ return;
+
+ spin_lock(&sport_map_lock);
+ __clear_bit(old, sport_map);
+ spin_unlock(&sport_map_lock);
+}
+
+static void c3cn_set_state(struct s3_conn *c3cn, int state)
+{
+ c3cn_conn_debug("c3cn 0x%p state -> 0x%x.\n", c3cn, state);
+ if (state == C3CN_STATE_CLOSE)
+ c3cn_put_port(c3cn);
+ c3cn->state = state;
+}
+
+
+void c3cn_reset_timer(struct s3_conn *c3cn, struct timer_list* timer,
+ unsigned long expires)
+{
+ if (!mod_timer(timer, expires))
+ c3cn_hold(c3cn);
+}
+
+typedef int (cxgb3_cpl_handler_decl) (struct t3cdev *,
+ struct sk_buff *, void *);
+
+static cxgb3_cpl_handler_decl do_act_establish;
+static cxgb3_cpl_handler_decl do_act_open_rpl;
+static cxgb3_cpl_handler_decl do_wr_ack;
+static cxgb3_cpl_handler_decl do_peer_close;
+static cxgb3_cpl_handler_decl do_abort_req;
+static cxgb3_cpl_handler_decl do_abort_rpl;
+static cxgb3_cpl_handler_decl do_close_con_rpl;
+static cxgb3_cpl_handler_decl do_iscsi_hdr;
+
+/*
+ * Protocol functions for our connections.
+ */
+static void process_deferq(struct work_struct *);
+
+static LIST_HEAD(cxgb3_list);
+static DEFINE_MUTEX(cxgb3_list_lock);
+
+/*
+ * For ULP connections HW may inserts digest bytes into the pdu. This array
+ * contains the compensating extra lengths for ULP packets. It is indexed by
+ * a packet's ULP submode.
+ */
+static const unsigned int cxgb3_ulp_extra_len[] = { 0, 4, 4, 8 };
+
+/*
+ * Return the length of any HW additions that will be made to a Tx packet.
+ * Such additions can happen for some types of ULP packets.
+ */
+static inline unsigned int ulp_extra_len(const struct sk_buff *skb)
+{
+ return cxgb3_ulp_extra_len[skb_ulp_mode(skb) & 3];
+}
+
+/*
+ * Size of WRs in bytes. Note that we assume all devices we are handling have
+ * the same WR size.
+ */
+static unsigned int wrlen __read_mostly;
+
+/*
+ * The number of WRs needed for an skb depends on the number of page fragments
+ * in the skb and whether it has any payload in its main body. This maps the
+ * length of the gather list represented by an skb into the # of necessary WRs.
+ */
+static unsigned int skb_wrs[MAX_SKB_FRAGS + 2] __read_mostly;
+
+static void s3_init_wr_tab(unsigned int wr_len)
+{
+ int i;
+
+ if (skb_wrs[1]) /* already initialized */
+ return;
+
+ for (i = 1; i < ARRAY_SIZE(skb_wrs); i++) {
+ int sgl_len = (3 * i) / 2 + (i & 1);
+
+ sgl_len += 3;
+ skb_wrs[i] = (sgl_len <= wr_len
+ ? 1 : 1 + (sgl_len - 2) / (wr_len - 1));
+ }
+
+ wrlen = wr_len * 8;
+}
+
+/*
+ * Initialization/cleanup cxgb3 API operations.
+ */
+/*
+ * large memory chunk allocation/release
+ */
+void *cxgb3i_alloc_big_mem(unsigned int size)
+{
+ void *p = kmalloc(size, GFP_KERNEL);
+ if (!p)
+ p = vmalloc(size);
+ if (p)
+ memset(p, 0, size);
+ return p;
+}
+
+void cxgb3i_free_big_mem(void *addr)
+{
+ unsigned long p = (unsigned long)addr;
+ if (p >= VMALLOC_START && p < VMALLOC_END)
+ vfree(addr);
+ else
+ kfree(addr);
+}
+
+void cxgb3i_sdev_cleanup(cxgb3_cpl_handler_func *cpl_handlers)
+{
+ memset(cpl_handlers, 0, NUM_CPL_CMDS*(sizeof(*cpl_handlers)));
+ if (sport_map)
+ cxgb3i_free_big_mem(sport_map);
+}
+
+int cxgb3i_sdev_init(cxgb3_cpl_handler_func *cpl_handlers)
+{
+ cpl_handlers[CPL_ACT_ESTABLISH] = do_act_establish;
+ cpl_handlers[CPL_ACT_OPEN_RPL] = do_act_open_rpl;
+ cpl_handlers[CPL_PEER_CLOSE] = do_peer_close;
+ cpl_handlers[CPL_ABORT_REQ_RSS] = do_abort_req;
+ cpl_handlers[CPL_ABORT_RPL_RSS] = do_abort_rpl;
+ cpl_handlers[CPL_CLOSE_CON_RPL] = do_close_con_rpl;
+ cpl_handlers[CPL_TX_DMA_ACK] = do_wr_ack;
+ cpl_handlers[CPL_ISCSI_HDR] = do_iscsi_hdr;
+
+ sport_map = cxgb3i_alloc_big_mem((max_connect + 7)/8);
+ if (!sport_map)
+ return -ENOMEM;
+ return 0;
+}
+
+void cxgb3i_sdev_add(struct t3cdev *cdev, struct cxgb3_client *client)
+{
+ struct cxgb3i_sdev_data *cdata;
+ struct adap_ports *ports;
+ struct ofld_page_info rx_page_info;
+ unsigned int wr_len;
+ int i;
+
+ cdata = kzalloc(sizeof *cdata, GFP_KERNEL);
+ if (!cdata)
+ return;
+ ports = kzalloc(sizeof *ports, GFP_KERNEL);
+ if (!ports)
+ goto free_ports;
+ cdata->ports = ports;
+
+ if (cdev->ctl(cdev, GET_WR_LEN, &wr_len) < 0 ||
+ cdev->ctl(cdev, GET_PORTS, cdata->ports) < 0 ||
+ cdev->ctl(cdev, GET_RX_PAGE_INFO, &rx_page_info) < 0)
+ goto free_ports;
+
+ s3_init_wr_tab(wr_len);
+
+ INIT_LIST_HEAD(&cdata->list);
+ cdata->cdev = cdev;
+ cdata->client = client;
+ cdata->rx_page_size = rx_page_info.page_size;
+ skb_queue_head_init(&cdata->deferq);
+ INIT_WORK(&cdata->deferq_task, process_deferq);
+
+ for (i = 0; i < ports->nports; i++)
+ NDEV2CDATA(ports->lldevs[i]) = cdata;
+
+ mutex_lock(&cxgb3_list_lock);
+ list_add_tail(&cdata->list, &cxgb3_list);
+ mutex_unlock(&cxgb3_list_lock);
+
+ return;
+
+free_ports:
+ kfree(ports);
+ kfree(cdata);
+}
+
+void cxgb3i_sdev_remove(struct t3cdev *cdev)
+{
+ struct cxgb3i_sdev_data *cdata = CXGB3_SDEV_DATA(cdev);
+ struct adap_ports *ports = cdata->ports;
+ int i;
+
+ for (i = 0; i < ports->nports; i++)
+ NDEV2CDATA(ports->lldevs[i]) = NULL;
+
+ mutex_lock(&cxgb3_list_lock);
+ list_del(&cdata->list);
+ mutex_unlock(&cxgb3_list_lock);
+
+ kfree(ports);
+ kfree(cdata);
+}
+
+/*
+ * Return TRUE if the specified net device is for a port on one of our
+ * registered adapters.
+ */
+static int is_cxgb3_dev(struct net_device *dev)
+{
+ struct cxgb3i_sdev_data *cdata;
+
+ mutex_lock(&cxgb3_list_lock);
+ list_for_each_entry(cdata, &cxgb3_list, list) {
+ struct adap_ports *ports = cdata->ports;
+ int i;
+
+ for (i = 0; i < ports->nports; i++)
+ if (dev == ports->lldevs[i]) {
+ mutex_unlock(&cxgb3_list_lock);
+ return 1;
+ }
+ }
+ mutex_unlock(&cxgb3_list_lock);
+ return 0;
+}
+
+/*
+ * Primary cxgb3 API operations.
+ * =============================
+ */
+
+static int s3_push_frames(struct s3_conn *, int);
+static int s3_send_reset(struct s3_conn *, int, struct sk_buff *);
+
+struct s3_conn * cxgb3i_c3cn_create(void)
+{
+ struct s3_conn *c3cn;
+
+ c3cn = kzalloc(sizeof(*c3cn), GFP_KERNEL);
+ if (c3cn == NULL)
+ return NULL;
+
+ c3cn->flags = 0;
+ spin_lock_init(&c3cn->lock);
+ atomic_set(&c3cn->refcnt, 1);
+ skb_queue_head_init(&c3cn->receive_queue);
+ skb_queue_head_init(&c3cn->write_queue);
+ setup_timer(&c3cn->retry_timer, NULL, (unsigned long)c3cn);
+ rwlock_init(&c3cn->callback_lock);
+
+ return c3cn;
+}
+
+static void mk_close_req(struct s3_conn *);
+static inline void s3_purge_write_queue(struct s3_conn *);
+
+/*
+ * Release a connection's local port if the connection is bound.
+ */
+static inline void release_port(struct s3_conn *c3cn)
+{
+ c3cn_conn_debug("c3cn 0x%p, port %u.\n", c3cn, c3cn->saddr.sin_port);
+ if (c3cn->saddr.sin_port)
+ c3cn_put_port(c3cn);
+}
+
+static void c3cn_done(struct s3_conn *c3cn)
+{
+ c3cn_conn_debug("c3cn 0x%p.\n", c3cn);
+
+ c3cn_set_state(c3cn, C3CN_STATE_CLOSE);
+ c3cn->shutdown = C3CN_SHUTDOWN_MASK;
+
+ cxgb3i_conn_closing(c3cn);
+}
+
+void c3cn_close(struct s3_conn *c3cn)
+{
+ int data_lost, old_state;
+
+ c3cn_conn_debug("c3cn 0x%p, state 0x%x, flag 0x%lx.\n",
+ c3cn, c3cn->state, c3cn->flags);
+
+ dst_confirm(c3cn->dst_cache);
+
+ spin_lock_bh(&c3cn->lock);
+ c3cn->shutdown |= C3CN_SHUTDOWN_MASK;
+
+ /*
+ * We need to flush the receive buffs. We do this only on the
+ * descriptor close, not protocol-sourced closes, because the
+ * reader process may not have drained the data yet! Make a note
+ * of whether any received data will be lost so we can decide whether
+ * to FIN or RST.
+ */
+ data_lost = skb_queue_len(&c3cn->receive_queue);
+ __skb_queue_purge(&c3cn->receive_queue);
+
+ if (c3cn->state == C3CN_STATE_CLOSE) /* Nothing if we are already closed */
+ ;
+ else if (data_lost || c3cn->state == C3CN_STATE_SYN_SENT) {
+ /* Unread data was tossed, zap the connection. */
+ s3_send_reset(c3cn, CPL_ABORT_SEND_RST, NULL);
+ release_port(c3cn);
+ goto unlock;
+ } else if (c3cn->state == C3CN_STATE_ESTABLISHED) {
+ c3cn_set_state(c3cn, C3CN_STATE_CLOSING);
+ mk_close_req(c3cn);
+ }
+
+unlock:
+ old_state = c3cn->state;
+ c3cn_hold(c3cn); /* must last past the potential destroy() */
+
+ spin_unlock_bh(&c3cn->lock); /* Final release in connection's lifetime. */
+
+ /*
+ * There are no more user references at this point. Grab the
+ * connection lock and finish the close.
+ */
+ local_bh_disable();
+ spin_lock(&c3cn->lock);
+
+ /*
+ * Because the connection was orphaned before the spin_lock()
+ * either the backlog or a BH may have already destroyed it.
+ * Bail out if so.
+ */
+ if (old_state != C3CN_STATE_CLOSE && c3cn->state == C3CN_STATE_CLOSE)
+ goto out;
+
+ if (c3cn->state == C3CN_STATE_CLOSE)
+ s3_purge_write_queue(c3cn);
+
+out:
+ spin_unlock(&c3cn->lock);
+ local_bh_enable();
+ c3cn_put(c3cn);
+}
+
+/*
+ * Local utility routines used to implement primary cxgb3 API operations.
+ * ======================================================================
+ */
+
+static int s3_connect(struct s3_conn *);
+static u32 s3_send_rx_credits(struct s3_conn *, u32, u32, int);
+static void mk_act_open_req(struct s3_conn *, struct sk_buff *,
+ unsigned int, const struct l2t_entry *);
+static void skb_entail(struct s3_conn *, struct sk_buff *, int);
+
+static inline void reset_wr_list(struct s3_conn *c3cn)
+{
+ c3cn->wr_pending_head = NULL;
+}
+
+/*
+ * Add a WR to a connections's list of pending WRs. This is a singly-linked
+ * list of sk_buffs operating as a FIFO. The head is kept in wr_pending_head
+ * and the tail in wr_pending_tail.
+ */
+static inline void enqueue_wr(struct s3_conn *c3cn,
+ struct sk_buff *skb)
+{
+ skb->sp = NULL;
+
+ /*
+ * We want to take an extra reference since both us and the driver
+ * need to free the packet before it's really freed. We know there's
+ * just one user currently so we use atomic_set rather than skb_get
+ * to avoid the atomic op.
+ */
+ atomic_set(&skb->users, 2);
+
+ if (!c3cn->wr_pending_head)
+ c3cn->wr_pending_head = skb;
+ else
+ c3cn->wr_pending_tail->sp = (void *)skb;
+ c3cn->wr_pending_tail = skb;
+}
+
+/*
+ * The next two functions calculate the option 0 value for a connection.
+ */
+static inline int compute_wscale(int win)
+{
+ int wscale = 0;
+ while (wscale < 14 && (65535<<wscale) < win)
+ wscale++;
+ return wscale;
+}
+
+static inline unsigned int calc_opt0h(struct s3_conn *c3cn)
+{
+ int wscale = compute_wscale(rcv_win);
+ return (V_KEEP_ALIVE(1) |
+ F_TCAM_BYPASS |
+ V_WND_SCALE(wscale) |
+ V_MSS_IDX(c3cn->mss_idx));
+}
+
+static inline unsigned int calc_opt0l(struct s3_conn *c3cn)
+{
+ return (V_ULP_MODE(ULP_MODE_ISCSI) |
+ V_RCV_BUFSIZ(rcv_win>>10));
+}
+
+static inline void make_tx_data_wr(struct s3_conn *c3cn,
+ struct sk_buff *skb, int len)
+{
+ struct tx_data_wr *req;
+
+ skb_reset_transport_header(skb);
+ req = (struct tx_data_wr *)__skb_push(skb, sizeof(*req));
+ req->wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_TX_DATA));
+ req->wr_lo = htonl(V_WR_TID(c3cn->tid));
+ req->sndseq = htonl(c3cn->snd_nxt);
+ /* len includes the length of any HW ULP additions */
+ req->len = htonl(len);
+ req->param = htonl(V_TX_PORT(c3cn->l2t->smt_idx));
+ /* V_TX_ULP_SUBMODE sets both the mode and submode */
+ req->flags = htonl(V_TX_ULP_SUBMODE(skb_ulp_mode(skb)) |
+ V_TX_SHOVE((skb_peek(&c3cn->write_queue) ? 0 : 1)));
+
+ if (!c3cn_flag(c3cn, C3CN_TX_DATA_SENT)) {
+
+ req->flags |= htonl(V_TX_ACK_PAGES(2) | F_TX_INIT |
+ V_TX_CPU_IDX(c3cn->qset));
+
+ /* Sendbuffer is in units of 32KB.
+ */
+ req->param |= htonl(V_TX_SNDBUF(snd_win >> 15));
+ c3cn_set_flag(c3cn, C3CN_TX_DATA_SENT);
+ }
+}
+
+static struct rtable *find_route(__be32 saddr, __be32 daddr,
+ __be16 sport, __be16 dport)
+{
+ struct rtable *rt;
+ struct flowi fl = {
+ .oif = 0,
+ .nl_u = {
+ .ip4_u = {
+ .daddr = daddr,
+ .saddr = saddr,
+ .tos = 0 } },
+ .proto = IPPROTO_TCP,
+ .uli_u = {
+ .ports = {
+ .sport = sport,
+ .dport = dport } } };
+
+ if (ip_route_output_flow(&init_net, &rt, &fl, NULL, 0))
+ return NULL;
+ return rt;
+}
+
+int cxgb3i_c3cn_connect(struct s3_conn *c3cn, struct sockaddr_in *usin)
+{
+ struct rtable *rt;
+ int err;
+
+ if (usin->sin_family != AF_INET)
+ return -EAFNOSUPPORT;
+
+ /* get a source port if one hasn't been provided */
+ err = c3cn_get_port(c3cn);
+ if (err)
+ return err;
+ c3cn_conn_debug("c3cn 0x%p get port %u.\n", c3cn, ntohs(c3cn->saddr.sin_port));
+
+ c3cn->daddr.sin_port = usin->sin_port;
+ c3cn->daddr.sin_addr.s_addr = usin->sin_addr.s_addr;
+
+ rt = find_route(c3cn->saddr.sin_addr.s_addr,
+ c3cn->daddr.sin_addr.s_addr,
+ c3cn->saddr.sin_port,
+ c3cn->daddr.sin_port);
+ if (rt == NULL) {
+ c3cn_conn_debug("NO route to 0x%x, port %u.\n",
+ c3cn->daddr.sin_addr.s_addr,
+ ntohs(c3cn->daddr.sin_port));
+ return -ENETUNREACH;
+ }
+
+ if (rt->rt_flags & (RTCF_MULTICAST | RTCF_BROADCAST)) {
+ c3cn_conn_debug("multi-cast route to 0x%x, port %u.\n",
+ c3cn->daddr.sin_addr.s_addr,
+ ntohs(c3cn->daddr.sin_port));
+ ip_rt_put(rt);
+ return -ENETUNREACH;
+ }
+
+ if (!c3cn->saddr.sin_addr.s_addr)
+ c3cn->saddr.sin_addr.s_addr = rt->rt_src;
+
+ c3cn_conn_debug("c3cn 0x%p -> SYN_SENT.\n", c3cn);
+ c3cn_set_state(c3cn, C3CN_STATE_SYN_SENT);
+
+ /* now commit destination to connection */
+ c3cn->dst_cache = &rt->u.dst;
+
+ if (s3_connect(c3cn))
+ return 0;
+ /*
+ * If we get here, we don't have an offload connection so simply
+ * return a failure.
+ */
+ err = -ENOTSUPP;
+
+ /*
+ * This trashes the connection and releases the local port,
+ * if necessary.
+ */
+ c3cn_conn_debug("c3cn 0x%p -> CLOSE.\n", c3cn);
+ c3cn_set_state(c3cn, C3CN_STATE_CLOSE);
+ ip_rt_put(rt);
+ c3cn_put_port(c3cn);
+ c3cn->daddr.sin_port = 0;
+ return err;
+}
+
+/*
+ * Set of states for which we should return RX credits.
+ */
+#define CREDIT_RETURN_STATE (C3CN_STATE_ESTABLISHED)
+
+/*
+ * Called after some received data has been read. It returns RX credits
+ * to the HW for the amount of data processed.
+ */
+void cxgb3i_c3cn_rx_credits(struct s3_conn *c3cn, int copied)
+{
+ struct t3cdev *cdev;
+ int must_send;
+ u32 credits, dack = 0;
+
+ if (!c3cn_in_state(c3cn, CREDIT_RETURN_STATE))
+ return;
+
+ credits = c3cn->copied_seq - c3cn->rcv_wup;
+ if (unlikely(!credits))
+ return;
+
+ cdev = c3cn->cdev;
+
+ if (unlikely(rx_credit_thres == 0))
+ return;
+
+ dack = F_RX_DACK_CHANGE | V_RX_DACK_MODE(1);
+
+ /*
+ * For coalescing to work effectively ensure the receive window has
+ * at least 16KB left.
+ */
+ must_send = credits + 16384 >= rcv_win;
+
+ if (must_send || credits >= rx_credit_thres)
+ c3cn->rcv_wup += s3_send_rx_credits(c3cn, credits, dack, must_send);
+}
+
+/*
+ * Generic ARP failure handler that discards the buffer.
+ */
+static void arp_failure_discard(struct t3cdev *cdev, struct sk_buff *skb)
+{
+ kfree_skb(skb);
+}
+
+/*
+ * Prepends TX_DATA_WR or CPL_CLOSE_CON_REQ headers to buffers waiting in a
+ * connection's send queue and sends them on to T3. Must be called with the
+ * connection's lock held. Returns the amount of send buffer space that was
+ * freed as a result of sending queued data to T3.
+ */
+static int s3_push_frames(struct s3_conn *c3cn, int req_completion)
+{
+ int total_size = 0;
+ struct sk_buff *skb;
+ struct t3cdev *cdev;
+ struct cxgb3i_sdev_data *cdata;
+
+ if (unlikely(c3cn_in_state(c3cn, C3CN_STATE_SYN_SENT | C3CN_STATE_CLOSE)))
+ return 0;
+
+ /*
+ * We shouldn't really be called at all after an abort but check just
+ * in case.
+ */
+ if (unlikely(c3cn_flag(c3cn, C3CN_ABORT_SHUTDOWN)))
+ return 0;
+
+ cdev = c3cn->cdev;
+ cdata = CXGB3_SDEV_DATA(cdev);
+
+ while (c3cn->wr_avail
+ && (skb = skb_peek(&c3cn->write_queue)) != NULL
+ && !c3cn_flag(c3cn, C3CN_TX_WAIT_IDLE)) {
+
+ int len = skb->len; /* length before skb_push */
+ int frags = skb_shinfo(skb)->nr_frags + (len != skb->data_len);
+ int wrs_needed = skb_wrs[frags];
+
+ if (wrs_needed > 1 && len + sizeof(struct tx_data_wr) <= wrlen)
+ wrs_needed = 1;
+
+ WARN_ON(frags >= ARRAY_SIZE(skb_wrs) || wrs_needed < 1);
+ if (c3cn->wr_avail < wrs_needed)
+ break;
+
+ __skb_unlink(skb, &c3cn->write_queue);
+ skb->priority = CPL_PRIORITY_DATA;
+ skb->csum = wrs_needed; /* remember this until the WR_ACK */
+ c3cn->wr_avail -= wrs_needed;
+ c3cn->wr_unacked += wrs_needed;
+ enqueue_wr(c3cn, skb);
+
+ if (likely(CXGB3_SKB_CB(skb)->flags & C3CB_FLAG_NEED_HDR)) {
+ len += ulp_extra_len(skb);
+ make_tx_data_wr(c3cn, skb, len);
+ c3cn->snd_nxt += len;
+ if ((req_completion
+ && c3cn->wr_unacked == wrs_needed)
+ || (CXGB3_SKB_CB(skb)->flags & C3CB_FLAG_COMPL)
+ || c3cn->wr_unacked >= c3cn->wr_max / 2) {
+ struct work_request_hdr *wr = cplhdr(skb);
+
+ wr->wr_hi |= htonl(F_WR_COMPL);
+ c3cn->wr_unacked = 0;
+ }
+ CXGB3_SKB_CB(skb)->flags &= ~C3CB_FLAG_NEED_HDR;
+ } else if (skb->data[0] == FW_WROPCODE_OFLD_CLOSE_CON)
+ c3cn_set_flag(c3cn, C3CN_CLOSE_CON_REQUESTED);
+
+ total_size += skb->truesize;
+ set_arp_failure_handler(skb, arp_failure_discard);
+ l2t_send(cdev, skb, c3cn->l2t);
+ }
+ return total_size;
+}
+
+/*
+ * Handle an ARP failure for a CPL_ABORT_REQ. Change it into a no RST variant
+ * and send it along.
+ */
+static void abort_arp_failure(struct t3cdev *cdev, struct sk_buff *skb)
+{
+ struct cpl_abort_req *req = cplhdr(skb);
+
+ req->cmd = CPL_ABORT_NO_RST;
+ cxgb3_ofld_send(cdev, skb);
+}
+
+/*
+ * Send an ABORT_REQ message. Cannot fail. This routine makes sure we do
+ * not send multiple ABORT_REQs for the same connection and also that we do
+ * not try to send a message after the connection has closed. Returns 1 if
+ * an ABORT_REQ wasn't generated after all, 0 otherwise.
+ */
+static int s3_send_reset(struct s3_conn *c3cn, int mode,
+ struct sk_buff *skb)
+{
+ struct cpl_abort_req *req;
+ unsigned int tid = c3cn->tid;
+
+ if (unlikely(c3cn_flag(c3cn, C3CN_ABORT_SHUTDOWN) || !c3cn->cdev)) {
+ if (skb)
+ __kfree_skb(skb);
+ return 1;
+ }
+
+ c3cn_conn_debug("c3cn 0x%p, mode %d.\n", c3cn, mode);
+
+ c3cn_set_flag(c3cn, C3CN_ABORT_RPL_PENDING);
+ c3cn_set_flag(c3cn, C3CN_ABORT_SHUTDOWN);
+
+ /* Purge the send queue so we don't send anything after an abort. */
+ s3_purge_write_queue(c3cn);
+
+ if (!skb)
+ skb = alloc_skb(sizeof(*req), GFP_KERNEL | __GFP_NOFAIL);
+ skb->priority = CPL_PRIORITY_DATA;
+ set_arp_failure_handler(skb, abort_arp_failure);
+
+ req = (struct cpl_abort_req *)skb_put(skb, sizeof(*req));
+ req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_HOST_ABORT_CON_REQ));
+ req->wr.wr_lo = htonl(V_WR_TID(tid));
+ OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_ABORT_REQ, tid));
+ req->rsvd0 = htonl(c3cn->snd_nxt);
+ req->rsvd1 = !c3cn_flag(c3cn, C3CN_TX_DATA_SENT);
+ req->cmd = mode;
+
+ l2t_send(c3cn->cdev, skb, c3cn->l2t);
+ return 0;
+}
+
+/*
+ * Add a list of skbs to a connection send queue. This interface is intended
+ * for use by in-kernel ULPs. The skbs must comply with the max size limit of
+ * the device and have a headroom of at least TX_HEADER_LEN bytes.
+ */
+int cxgb3i_c3cn_send_pdus(struct s3_conn *c3cn, struct sk_buff *skb, int flags)
+{
+ struct sk_buff *next;
+ int err, copied = 0;
+
+ spin_lock_bh(&c3cn->lock);
+
+ if (!c3cn_in_state(c3cn, C3CN_STATE_ESTABLISHED)) {
+ err = -EAGAIN;
+ goto out_err;
+ }
+
+ err = -EPIPE;
+ if (c3cn->err || (c3cn->shutdown & C3CN_SEND_SHUTDOWN))
+ goto out_err;
+
+ while (skb) {
+ if (unlikely(skb_headroom(skb) < TX_HEADER_LEN)) {
+ c3cn_tx_debug("c3cn 0x%p, skb head.\n", c3cn);
+ err = -EINVAL;
+ goto out_err;
+ }
+
+ next = skb->next;
+ skb->next = NULL;
+ skb_entail(c3cn, skb, C3CB_FLAG_NO_APPEND | C3CB_FLAG_NEED_HDR);
+ copied += skb->len;
+ c3cn->write_seq += skb->len + ulp_extra_len(skb);
+ skb = next;
+ }
+done:
+ if (likely(skb_queue_len(&c3cn->write_queue)))
+ s3_push_frames(c3cn, 1);
+ spin_unlock_bh(&c3cn->lock);
+ return copied;
+
+out_err:
+ if (copied == 0 && err == -EPIPE)
+ copied = c3cn->err ? c3cn->err : -EPIPE;
+ goto done;
+}
+
+/*
+ * Low-level utility routines for primary API functions.
+ * =====================================================
+ */
+/* routines to implement CPL message processing */
+static void c3cn_act_establish(struct s3_conn *, struct sk_buff *);
+static void active_open_failed(struct s3_conn *, struct sk_buff *);
+static void wr_ack(struct s3_conn *, struct sk_buff *);
+static void do_peer_fin(struct s3_conn *, struct sk_buff *);
+static void process_abort_req(struct s3_conn *, struct sk_buff *);
+static void process_abort_rpl(struct s3_conn *, struct sk_buff *);
+static void process_close_con_rpl(struct s3_conn *, struct sk_buff *);
+static void process_rx_iscsi_hdr(struct s3_conn *, struct sk_buff *);
+
+static struct sk_buff *__get_cpl_reply_skb(struct sk_buff *, size_t, gfp_t);
+
+static int act_open(struct s3_conn *, struct net_device *);
+static void fail_act_open(struct s3_conn *, int);
+static void init_offload_conn(struct s3_conn *, struct t3cdev *,
+ struct dst_entry *);
+
+/*
+ * Insert a connection into the TID table and take an extra reference.
+ */
+static inline void c3cn_insert_tid(struct cxgb3i_sdev_data *cdata,
+ struct s3_conn *c3cn,
+ unsigned int tid)
+{
+ c3cn_hold(c3cn);
+ cxgb3_insert_tid(cdata->cdev, cdata->client, c3cn, tid);
+}
+
+static inline void free_atid(struct t3cdev *cdev, unsigned int tid)
+{
+ struct s3_conn *c3cn = cxgb3_free_atid(cdev, tid);
+ if (c3cn)
+ c3cn_put(c3cn);
+}
+
+/*
+ * This function is intended for allocations of small control messages.
+ * Such messages go as immediate data and usually the pakets are freed
+ * immediately. We maintain a cache of one small sk_buff and use it whenever
+ * it is available (has a user count of 1). Otherwise we get a fresh buffer.
+ */
+#define CTRL_SKB_LEN 120
+
+static struct sk_buff *alloc_ctrl_skb(const struct s3_conn *c3cn,
+ int len)
+{
+ struct sk_buff *skb = c3cn->ctrl_skb_cache;
+
+ if (likely(skb && !skb_shared(skb) && !skb_cloned(skb))) {
+ __skb_trim(skb, 0);
+ atomic_set(&skb->users, 2);
+ } else if (likely(!in_atomic()))
+ skb = alloc_skb(len, GFP_ATOMIC | __GFP_NOFAIL);
+ else
+ skb = alloc_skb(len, GFP_ATOMIC);
+ return skb;
+}
+
+/**
+ * cxgb3_egress_dev - return the cxgb3 egress device or NULL if the egress
+ * device isn't one of our ports.
+ *
+ * @root_dev: the root device anchoring the search
+ * @c3cn: the connection used to determine egress port in bonding mode
+ * @context: in bonding mode, indicates a connection set up or failover
+ *
+ * Given a root network device it returns the physical egress device that is a
+ * descendant of the root device. The root device may be either a physical
+ * device, in which case it is the device returned, or a virtual device, such
+ * as a VLAN or bonding device. In case of a bonding device the search
+ * considers the decisions of the bonding device given its mode to locate the
+ * correct egress device.
+ */
+static struct net_device *cxgb3_egress_dev(struct net_device *root_dev,
+ struct s3_conn *c3cn,
+ int context)
+{
+ while (root_dev) {
+ if (root_dev->priv_flags & IFF_802_1Q_VLAN)
+ root_dev = vlan_dev_info(root_dev)->real_dev;
+ else if (is_cxgb3_dev(root_dev))
+ return root_dev;
+ else
+ return NULL;
+ }
+ return NULL;
+}
+
+/*
+ * Return TRUE if we're able to establish an offload connection; otherwise
+ * return FALSE.
+ */
+static int s3_connect(struct s3_conn *c3cn)
+{
+ struct net_device *dev = cxgb3_egress_dev(c3cn->dst_cache->dev,
+ c3cn, 0);
+ if (dev == NULL) {
+ c3cn_conn_debug("c3cn 0x%p, egress dev NULL.\n", c3cn);
+ return 0;
+ }
+ return act_open(c3cn, dev) == 0;
+}
+
+/*
+ * Handle an ARP failure for an active open.
+ */
+static void act_open_req_arp_failure(struct t3cdev *dev, struct sk_buff *skb)
+{
+ struct s3_conn *c3cn = (struct s3_conn *)skb->sk;
+
+ c3cn_hold(c3cn);
+ spin_lock(&c3cn->lock);
+ if (c3cn->state == C3CN_STATE_SYN_SENT) {
+ fail_act_open(c3cn, EHOSTUNREACH);
+ __kfree_skb(skb);
+ }
+ spin_unlock(&c3cn->lock);
+ c3cn_put(c3cn);
+}
+
+/*
+ * Send an active open request.
+ */
+static int act_open(struct s3_conn *c3cn, struct net_device *dev)
+{
+ struct cxgb3i_sdev_data *cdata = NDEV2CDATA(dev);
+ struct t3cdev *cdev = cdata->cdev;
+ struct dst_entry *dst = c3cn->dst_cache;
+ struct sk_buff *skb;
+
+ c3cn_conn_debug("c3cn 0x%p.\n", c3cn);
+ /*
+ * Initialize connection data. Note that the flags and ULP mode are
+ * initialized higher up ...
+ */
+ c3cn->dev = dev;
+ c3cn->cdev = cdev;
+ c3cn->tid = cxgb3_alloc_atid(cdev, cdata->client, c3cn);
+ if (c3cn->tid < 0)
+ goto out_err;
+ c3cn->qset = 0;
+ c3cn->l2t = t3_l2t_get(cdev, dst->neighbour, dev);
+ if (!c3cn->l2t)
+ goto free_tid;
+
+ skb = alloc_skb(sizeof(struct cpl_act_open_req),
+ GFP_KERNEL | __GFP_NOFAIL);
+ skb->sk = (struct sock *)c3cn;
+ set_arp_failure_handler(skb, act_open_req_arp_failure);
+
+ c3cn_hold(c3cn);
+
+ init_offload_conn(c3cn, cdev, dst);
+ c3cn->err = 0;
+ c3cn_reset_flag(c3cn, C3CN_DONE);
+
+ mk_act_open_req(c3cn, skb, c3cn->tid, c3cn->l2t);
+ l2t_send(cdev, skb, c3cn->l2t);
+ return 0;
+
+free_tid:
+ free_atid(cdev, c3cn->tid);
+ c3cn->tid = 0;
+out_err:
+ return -1;
+}
+
+/*
+ * Close a connection by sending a CPL_CLOSE_CON_REQ message. Cannot fail
+ * under any circumstances. We take the easy way out and always queue the
+ * message to the write_queue. We can optimize the case where the queue is
+ * already empty though the optimization is probably not worth it.
+ */
+static void mk_close_req(struct s3_conn *c3cn)
+{
+ struct sk_buff *skb;
+ struct cpl_close_con_req *req;
+ unsigned int tid = c3cn->tid;
+
+ c3cn_conn_debug("c3cn 0x%p.\n", c3cn);
+
+ skb = alloc_skb(sizeof(struct cpl_close_con_req),
+ GFP_KERNEL | __GFP_NOFAIL);
+ req = (struct cpl_close_con_req *)__skb_put(skb, sizeof(*req));
+ req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_CLOSE_CON));
+ req->wr.wr_lo = htonl(V_WR_TID(tid));
+ OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_CLOSE_CON_REQ, tid));
+ req->rsvd = htonl(c3cn->write_seq);
+
+ skb_entail(c3cn, skb, C3CB_FLAG_NO_APPEND);
+ if (c3cn->state != C3CN_STATE_SYN_SENT)
+ s3_push_frames(c3cn, 1);
+}
+
+static void skb_entail(struct s3_conn *c3cn, struct sk_buff *skb,
+ int flags)
+{
+ CXGB3_SKB_CB(skb)->seq = c3cn->write_seq;
+ CXGB3_SKB_CB(skb)->flags = flags;
+ __skb_queue_tail(&c3cn->write_queue, skb);
+}
+
+/*
+ * Send RX credits through an RX_DATA_ACK CPL message. If nofail is 0 we are
+ * permitted to return without sending the message in case we cannot allocate
+ * an sk_buff. Returns the number of credits sent.
+ */
+static u32 s3_send_rx_credits(struct s3_conn *c3cn, u32 credits, u32 dack,
+ int nofail)
+{
+ struct sk_buff *skb;
+ struct cpl_rx_data_ack *req;
+
+ skb = (nofail ? alloc_ctrl_skb(c3cn, sizeof(*req))
+ : alloc_skb(sizeof(*req), GFP_ATOMIC));
+ if (!skb)
+ return 0;
+
+ req = (struct cpl_rx_data_ack *)__skb_put(skb, sizeof(*req));
+ req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_FORWARD));
+ OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_RX_DATA_ACK, c3cn->tid));
+ req->credit_dack = htonl(dack | V_RX_CREDITS(credits));
+ skb->priority = CPL_PRIORITY_ACK;
+ cxgb3_ofld_send(c3cn->cdev, skb);
+ return credits;
+}
+
+static void mk_act_open_req(struct s3_conn *c3cn, struct sk_buff *skb,
+ unsigned int atid, const struct l2t_entry *e)
+{
+ struct cpl_act_open_req *req;
+ struct iscsi_ipaddr ii;
+
+ ii.dev = c3cn->dev;
+ c3cn->cdev->ctl(c3cn->cdev, GET_ISCSI_IPADDR, &ii);
+
+ c3cn_conn_debug("c3cn 0x%p, atid 0x%x.\n", c3cn, atid);
+
+ skb->priority = CPL_PRIORITY_SETUP;
+ req = (struct cpl_act_open_req *)__skb_put(skb, sizeof(*req));
+ req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_FORWARD));
+ OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_ACT_OPEN_REQ, atid));
+ req->local_port = c3cn->saddr.sin_port;
+ req->peer_port = c3cn->daddr.sin_port;
+ req->local_ip = ii.ipaddr;
+ req->peer_ip = c3cn->daddr.sin_addr.s_addr;
+ req->opt0h = htonl(calc_opt0h(c3cn) | V_L2T_IDX(e->idx) |
+ V_TX_CHANNEL(e->smt_idx));
+ req->opt0l = htonl(calc_opt0l(c3cn));
+ req->params = 0;
+}
+
+static inline void s3_purge_write_queue(struct s3_conn *c3cn)
+{
+ struct sk_buff *skb;
+
+ while ((skb = __skb_dequeue(&c3cn->write_queue))) {
+ __kfree_skb(skb);
+ }
+}
+
+/*
+ * Definitions and declarations for CPL handler functions.
+ * =======================================================
+ */
+
+/*
+ * Similar to process_cpl_msg() but takes an extra connection reference around
+ * the call to the handler. Should be used if the handler may drop a
+ * connection reference.
+ */
+static inline void process_cpl_msg_ref(void (*fn) (struct s3_conn *,
+ struct sk_buff *),
+ struct s3_conn *c3cn,
+ struct sk_buff *skb)
+{
+ c3cn_hold(c3cn);
+ process_cpl_msg(fn, c3cn, skb);
+ c3cn_put(c3cn);
+}
+
+/*
+ * Return whether a failed active open has allocated a TID
+ */
+static inline int act_open_has_tid(int status)
+{
+ return status != CPL_ERR_TCAM_FULL && status != CPL_ERR_CONN_EXIST &&
+ status != CPL_ERR_ARP_MISS;
+}
+
+/*
+ * Returns true if a connection cannot accept new Rx data.
+ */
+static inline int c3cn_no_receive(const struct s3_conn *c3cn)
+{
+ return (c3cn->shutdown & C3CN_RCV_SHUTDOWN);
+}
+
+/*
+ * A helper function that aborts a connection and increments the given MIB
+ * counter. The supplied skb is used to generate the ABORT_REQ message if
+ * possible. Must be called with softirqs disabled.
+ */
+static inline void abort_conn(struct s3_conn *c3cn,
+ struct sk_buff *skb)
+{
+ struct sk_buff *abort_skb;
+
+ abort_skb = __get_cpl_reply_skb(skb, sizeof(struct cpl_abort_req),
+ GFP_ATOMIC);
+ if (abort_skb)
+ s3_send_reset(c3cn, CPL_ABORT_SEND_RST, abort_skb);
+}
+
+/*
+ * Returns whether an ABORT_REQ_RSS message is a negative advice.
+ */
+static inline int is_neg_adv_abort(unsigned int status)
+{
+ return (status == CPL_ERR_RTX_NEG_ADVICE
+ || status == CPL_ERR_PERSIST_NEG_ADVICE);
+}
+
+/*
+ * CPL handler functions.
+ * ======================
+ */
+
+/*
+ * Process a CPL_ACT_ESTABLISH message.
+ */
+static int do_act_establish(struct t3cdev *cdev, struct sk_buff *skb,
+ void *ctx)
+{
+ struct cpl_act_establish *req = cplhdr(skb);
+ unsigned int tid = GET_TID(req);
+ unsigned int atid = G_PASS_OPEN_TID(ntohl(req->tos_tid));
+ struct s3_conn *c3cn = (struct s3_conn *)ctx;
+ struct cxgb3i_sdev_data *cdata = CXGB3_SDEV_DATA(cdev);
+
+ c3cn_conn_debug("c3cn 0x%p, tid 0x%x.\n", c3cn, tid);
+ /*
+ * It's OK if the TID is currently in use, the owning connection may
+ * have backlogged its last CPL message(s). Just take it away.
+ */
+ c3cn->tid = tid;
+ c3cn_insert_tid(cdata, c3cn, tid);
+ free_atid(cdev, atid);
+
+ c3cn->qset = G_QNUM(ntohl(skb->csum));
+
+ process_cpl_msg(c3cn_act_establish, c3cn, skb);
+ return 0;
+}
+
+/*
+ * Process an ACT_OPEN_RPL CPL message.
+ */
+static int do_act_open_rpl(struct t3cdev *cdev, struct sk_buff *skb, void *ctx)
+{
+ struct s3_conn *c3cn = (struct s3_conn *)ctx;
+ struct cpl_act_open_rpl *rpl = cplhdr(skb);
+
+ c3cn_conn_debug("c3cn 0x%p, status 0x%x.\n", c3cn, rpl->status);
+
+ if (act_open_has_tid(rpl->status))
+ cxgb3_queue_tid_release(cdev, GET_TID(rpl));
+
+ process_cpl_msg_ref(active_open_failed, c3cn, skb);
+ return 0;
+}
+
+/*
+ * Handler RX_ISCSI_HDR CPL messages.
+ */
+static int do_iscsi_hdr(struct t3cdev *t3dev, struct sk_buff *skb, void *ctx)
+{
+ struct s3_conn *c3cn = (struct s3_conn *)ctx;
+ process_cpl_msg(process_rx_iscsi_hdr, c3cn, skb);
+ return 0;
+}
+
+/*
+ * Handler for TX_DATA_ACK CPL messages.
+ */
+static int do_wr_ack(struct t3cdev *cdev, struct sk_buff *skb, void *ctx)
+{
+ struct s3_conn *c3cn = (struct s3_conn *)ctx;
+
+ process_cpl_msg(wr_ack, c3cn, skb);
+ return 0;
+}
+
+/*
+ * Handler for PEER_CLOSE CPL messages.
+ */
+static int do_peer_close(struct t3cdev *cdev, struct sk_buff *skb, void *ctx)
+{
+ struct s3_conn *c3cn = (struct s3_conn *)ctx;
+
+ c3cn_conn_debug("c3cn 0x%p, state 0x%x, flag 0x%lx.\n",
+ c3cn, c3cn->state, c3cn->flags);
+ process_cpl_msg_ref(do_peer_fin, c3cn, skb);
+ return 0;
+}
+
+/*
+ * Handle an ABORT_REQ_RSS CPL message.
+ */
+static int do_abort_req(struct t3cdev *cdev, struct sk_buff *skb, void *ctx)
+{
+ const struct cpl_abort_req_rss *req = cplhdr(skb);
+ struct s3_conn *c3cn = (struct s3_conn *)ctx;
+
+ if (is_neg_adv_abort(req->status)) {
+ __kfree_skb(skb);
+ return 0;
+ }
+
+ c3cn_conn_debug("c3cn 0x%p, state 0x%x, flag 0x%lx.\n",
+ c3cn, c3cn->state, c3cn->flags);
+
+ process_cpl_msg_ref(process_abort_req, c3cn, skb);
+ return 0;
+}
+
+/*
+ * Handle an ABORT_RPL_RSS CPL message.
+ */
+static int do_abort_rpl(struct t3cdev *cdev, struct sk_buff *skb, void *ctx)
+{
+ struct cpl_abort_rpl_rss *rpl = cplhdr(skb);
+ struct s3_conn *c3cn;
+
+ /*
+ * Ignore replies to post-close aborts indicating that the abort was
+ * requested too late. These connections are terminated when we get
+ * PEER_CLOSE or CLOSE_CON_RPL and by the time the abort_rpl_rss
+ * arrives the TID is either no longer used or it has been recycled.
+ */
+ if (rpl->status == CPL_ERR_ABORT_FAILED) {
+discard:
+ __kfree_skb(skb);
+ return 0;
+ }
+
+ c3cn = (struct s3_conn *)ctx;
+ c3cn_conn_debug("c3cn 0x%p, state 0x%x, flag 0x%lx.\n",
+ c3cn, c3cn->state, c3cn->flags);
+
+ /*
+ * Sometimes we've already closed the connection, e.g., a post-close
+ * abort races with ABORT_REQ_RSS, the latter frees the connection
+ * expecting the ABORT_REQ will fail with CPL_ERR_ABORT_FAILED,
+ * but FW turns the ABORT_REQ into a regular one and so we get
+ * ABORT_RPL_RSS with status 0 and no connection. Only on T3A.
+ */
+ if (!c3cn)
+ goto discard;
+
+ process_cpl_msg_ref(process_abort_rpl, c3cn, skb);
+ return 0;
+}
+
+/*
+ * Handler for CLOSE_CON_RPL CPL messages.
+ */
+static int do_close_con_rpl(struct t3cdev *cdev, struct sk_buff *skb,
+ void *ctx)
+{
+ struct s3_conn *c3cn = (struct s3_conn *)ctx;
+
+ c3cn_conn_debug("c3cn 0x%p, state 0x%x, flag 0x%lx.\n",
+ c3cn, c3cn->state, c3cn->flags);
+
+ process_cpl_msg_ref(process_close_con_rpl, c3cn, skb);
+ return 0;
+}
+
+/*
+ * Definitions and declarations for CPL message processing.
+ * ========================================================
+ */
+
+static void make_established(struct s3_conn *, u32, unsigned int);
+static void t3_release_offload_resources(struct s3_conn *);
+static void act_open_retry_timer(unsigned long);
+static void mk_act_open_req(struct s3_conn *, struct sk_buff *,
+ unsigned int, const struct l2t_entry *);
+static int act_open_rpl_status_to_errno(int);
+static void handle_excess_rx(struct s3_conn *, struct sk_buff *);
+static int abort_status_to_errno(struct s3_conn *, int, int *);
+static void send_abort_rpl(struct sk_buff *, struct t3cdev *, int);
+static struct sk_buff *get_cpl_reply_skb(struct sk_buff *, size_t, gfp_t);
+static void t3_defer_reply(struct sk_buff *, struct t3cdev *, defer_handler_t);
+static void send_deferred_abort_rpl(struct t3cdev *, struct sk_buff *);
+
+/*
+ * Dequeue and return the first unacknowledged's WR on a connections's pending
+ * list.
+ */
+static inline struct sk_buff *dequeue_wr(struct s3_conn *c3cn)
+{
+ struct sk_buff *skb = c3cn->wr_pending_head;
+
+ if (likely(skb)) {
+ /* Don't bother clearing the tail */
+ c3cn->wr_pending_head = (struct sk_buff *)skb->sp;
+ skb->sp = NULL;
+ }
+ return skb;
+}
+
+/*
+ * Return the first pending WR without removing it from the list.
+ */
+static inline struct sk_buff *peek_wr(const struct s3_conn *c3cn)
+{
+ return c3cn->wr_pending_head;
+}
+
+static inline void free_wr_skb(struct sk_buff *skb)
+{
+ kfree_skb(skb);
+}
+
+static void purge_wr_queue(struct s3_conn *c3cn)
+{
+ struct sk_buff *skb;
+ while ((skb = dequeue_wr(c3cn)) != NULL)
+ free_wr_skb(skb);
+}
+
+static inline void set_abort_rpl_wr(struct sk_buff *skb, unsigned int tid,
+ int cmd)
+{
+ struct cpl_abort_rpl *rpl = cplhdr(skb);
+
+ rpl->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_HOST_ABORT_CON_RPL));
+ rpl->wr.wr_lo = htonl(V_WR_TID(tid));
+ OPCODE_TID(rpl) = htonl(MK_OPCODE_TID(CPL_ABORT_RPL, tid));
+ rpl->cmd = cmd;
+}
+
+/*
+ * CPL message processing ...
+ * ==========================
+ */
+
+/*
+ * Updates connection state from an active establish CPL message. Runs with
+ * the connection lock held.
+ */
+static void c3cn_act_establish(struct s3_conn *c3cn,
+ struct sk_buff *skb)
+{
+ struct cpl_act_establish *req = cplhdr(skb);
+ u32 rcv_isn = ntohl(req->rcv_isn); /* real RCV_ISN + 1 */
+
+ if (unlikely(c3cn->state != C3CN_STATE_SYN_SENT))
+ printk(KERN_ERR "TID %u expected SYN_SENT, found %d\n",
+ c3cn->tid, c3cn->state);
+
+ c3cn->copied_seq = c3cn->rcv_wup = c3cn->rcv_nxt = rcv_isn;
+ make_established(c3cn, ntohl(req->snd_isn), ntohs(req->tcp_opt));
+
+ __kfree_skb(skb);
+
+ if (s3_push_frames(c3cn, 1))
+ cxgb3i_conn_tx_open(c3cn);
+}
+
+/*
+ * Handle active open failures.
+ */
+static void active_open_failed(struct s3_conn *c3cn,
+ struct sk_buff *skb)
+{
+ struct cpl_act_open_rpl *rpl = cplhdr(skb);
+
+ if (rpl->status == CPL_ERR_CONN_EXIST &&
+ c3cn->retry_timer.function != act_open_retry_timer) {
+ c3cn->retry_timer.function = act_open_retry_timer;
+ c3cn_reset_timer(c3cn, &c3cn->retry_timer,
+ jiffies + HZ / 2);
+ } else
+ fail_act_open(c3cn, act_open_rpl_status_to_errno(rpl->status));
+ __kfree_skb(skb);
+}
+
+/*
+ * Process received pdu for a connection.
+ */
+static void process_rx_iscsi_hdr(struct s3_conn *c3cn,
+ struct sk_buff *skb)
+{
+ struct cpl_iscsi_hdr *hdr_cpl = cplhdr(skb);
+ struct cpl_iscsi_hdr_norss data_cpl;
+ struct cpl_rx_data_ddp_norss ddp_cpl;
+ unsigned int hdr_len, data_len, status;
+ unsigned int len;
+ int err;
+
+ if (unlikely(c3cn_no_receive(c3cn))) {
+ handle_excess_rx(c3cn, skb);
+ return;
+ }
+
+ CXGB3_SKB_CB(skb)->seq = ntohl(hdr_cpl->seq);
+ CXGB3_SKB_CB(skb)->flags = 0;
+
+ skb_reset_transport_header(skb);
+ __skb_pull(skb, sizeof(struct cpl_iscsi_hdr));
+
+ len = hdr_len = ntohs(hdr_cpl->len);
+ /* msg coalesce is off or not enough data received */
+ if (skb->len <= hdr_len) {
+ printk(KERN_ERR "%s: TID %u, ISCSI_HDR, skb len %u < %u.\n",
+ c3cn->cdev->name, c3cn->tid, skb->len, hdr_len);
+ goto abort_conn;
+ }
+
+ err = skb_copy_bits(skb, skb->len - sizeof(ddp_cpl), &ddp_cpl,
+ sizeof(ddp_cpl));
+ if (err < 0)
+ goto abort_conn;
+
+ skb_ulp_mode(skb) = ULP2_FLAG_DATA_READY;
+ skb_ulp_pdulen(skb) = ntohs(ddp_cpl.len);
+ skb_ulp_ddigest(skb) = ntohl(ddp_cpl.ulp_crc);
+ status = ntohl(ddp_cpl.ddp_status);
+
+ if (status & (1 << RX_DDP_STATUS_HCRC_SHIFT))
+ skb_ulp_mode(skb) |= ULP2_FLAG_HCRC_ERROR;
+ if (status & (1 << RX_DDP_STATUS_DCRC_SHIFT))
+ skb_ulp_mode(skb) |= ULP2_FLAG_DCRC_ERROR;
+ if (status & (1 << RX_DDP_STATUS_PAD_SHIFT))
+ skb_ulp_mode(skb) |= ULP2_FLAG_PAD_ERROR;
+
+ if (skb->len > (hdr_len + sizeof(ddp_cpl))) {
+ err = skb_copy_bits(skb, hdr_len, &data_cpl, sizeof(data_cpl));
+ if (err < 0)
+ goto abort_conn;
+ data_len = ntohs(data_cpl.len);
+ len += sizeof(data_cpl) + data_len;
+ } else if (status & (1 << RX_DDP_STATUS_DDP_SHIFT))
+ skb_ulp_mode(skb) |= ULP2_FLAG_DATA_DDPED;
+
+ c3cn->rcv_nxt = ntohl(ddp_cpl.seq) + skb_ulp_pdulen(skb);
+ __pskb_trim(skb, len);
+ __skb_queue_tail(&c3cn->receive_queue, skb);
+ cxgb3i_conn_pdu_ready(c3cn);
+
+ return;
+
+abort_conn:
+ s3_send_reset(c3cn, CPL_ABORT_SEND_RST, NULL);
+ __kfree_skb(skb);
+}
+
+/*
+ * Process an acknowledgment of WR completion. Advance snd_una and send the
+ * next batch of work requests from the write queue.
+ */
+static void wr_ack(struct s3_conn *c3cn, struct sk_buff *skb)
+{
+ struct cpl_wr_ack *hdr = cplhdr(skb);
+ unsigned int credits = ntohs(hdr->credits);
+ u32 snd_una = ntohl(hdr->snd_una);
+
+ c3cn->wr_avail += credits;
+ if (c3cn->wr_unacked > c3cn->wr_max - c3cn->wr_avail)
+ c3cn->wr_unacked = c3cn->wr_max - c3cn->wr_avail;
+
+ while (credits) {
+ struct sk_buff *p = peek_wr(c3cn);
+
+ if (unlikely(!p)) {
+ printk(KERN_ERR "%u WR_ACK credits for TID %u with "
+ "nothing pending, state %u\n",
+ credits, c3cn->tid, c3cn->state);
+ break;
+ }
+ if (unlikely(credits < p->csum)) {
+ p->csum -= credits;
+ break;
+ } else {
+ dequeue_wr(c3cn);
+ credits -= p->csum;
+ free_wr_skb(p);
+ }
+ }
+
+ if (unlikely(before(snd_una, c3cn->snd_una))) {
+ goto out_free;
+ }
+
+ if (c3cn->snd_una != snd_una) {
+ c3cn->snd_una = snd_una;
+ dst_confirm(c3cn->dst_cache);
+ if (c3cn->snd_una == c3cn->snd_nxt)
+ c3cn_reset_flag(c3cn, C3CN_TX_WAIT_IDLE);
+ }
+
+ if (skb_queue_len(&c3cn->write_queue) && s3_push_frames(c3cn, 0))
+ cxgb3i_conn_tx_open(c3cn);
+out_free:
+ __kfree_skb(skb);
+}
+
+/*
+ * Handle a peer FIN.
+ */
+static void do_peer_fin(struct s3_conn *c3cn, struct sk_buff *skb)
+{
+ int keep = 0;
+
+ if (c3cn_flag(c3cn, C3CN_ABORT_RPL_PENDING))
+ goto out;
+
+ c3cn->shutdown |= C3CN_RCV_SHUTDOWN;
+ c3cn_set_flag(c3cn, C3CN_DONE);
+
+ switch (c3cn->state) {
+ case C3CN_STATE_ESTABLISHED:
+ break;
+ case C3CN_STATE_CLOSING:
+ t3_release_offload_resources(c3cn);
+ c3cn_done(c3cn);
+ break;
+ default:
+ printk(KERN_ERR
+ "%s: TID %u received PEER_CLOSE in bad state %d\n",
+ c3cn->cdev->name, c3cn->tid, c3cn->state);
+ }
+
+ cxgb3i_conn_closing(c3cn);
+out:
+ if (!keep)
+ __kfree_skb(skb);
+}
+
+/*
+ * Process abort requests. If we are waiting for an ABORT_RPL we ignore this
+ * request except that we need to reply to it.
+ */
+static void process_abort_req(struct s3_conn *c3cn,
+ struct sk_buff *skb)
+{
+ int rst_status = CPL_ABORT_NO_RST;
+ const struct cpl_abort_req_rss *req = cplhdr(skb);
+
+ if (!c3cn_flag(c3cn, C3CN_ABORT_REQ_RCVD)) {
+ c3cn_set_flag(c3cn, C3CN_ABORT_REQ_RCVD);
+ c3cn_set_flag(c3cn, C3CN_ABORT_SHUTDOWN);
+ __kfree_skb(skb);
+ return;
+ }
+ c3cn_reset_flag(c3cn, C3CN_ABORT_REQ_RCVD);
+
+ /*
+ * Three cases to consider:
+ * a) We haven't sent an abort_req; close the connection.
+ * b) We have sent a post-close abort_req that will get to TP too late
+ * and will generate a CPL_ERR_ABORT_FAILED reply. The reply will
+ * be ignored and the connection should be closed now.
+ * c) We have sent a regular abort_req that will get to TP too late.
+ * That will generate an abort_rpl with status 0, wait for it.
+ */
+ send_abort_rpl(skb, c3cn->cdev, rst_status);
+
+ if (!c3cn_flag(c3cn, C3CN_ABORT_RPL_PENDING)) {
+ c3cn->err =
+ abort_status_to_errno(c3cn, req->status, &rst_status);
+
+ t3_release_offload_resources(c3cn);
+ c3cn_done(c3cn);
+ }
+}
+
+/*
+ * Process abort replies. We only process these messages if we anticipate
+ * them as the coordination between SW and HW in this area is somewhat lacking
+ * and sometimes we get ABORT_RPLs after we are done with the connection that
+ * originated the ABORT_REQ.
+ */
+static void process_abort_rpl(struct s3_conn *c3cn,
+ struct sk_buff *skb)
+{
+ if (c3cn_flag(c3cn, C3CN_ABORT_RPL_PENDING)) {
+ if (!c3cn_flag(c3cn, C3CN_ABORT_RPL_RCVD))
+ c3cn_set_flag(c3cn, C3CN_ABORT_RPL_RCVD);
+ else {
+ c3cn_reset_flag(c3cn, C3CN_ABORT_RPL_RCVD);
+ c3cn_reset_flag(c3cn, C3CN_ABORT_RPL_PENDING);
+ BUG_ON(c3cn_flag(c3cn, C3CN_ABORT_REQ_RCVD));
+ t3_release_offload_resources(c3cn);
+ c3cn_done(c3cn);
+ }
+ }
+ __kfree_skb(skb);
+}
+
+/*
+ * Process a peer ACK to our FIN.
+ */
+static void process_close_con_rpl(struct s3_conn *c3cn,
+ struct sk_buff *skb)
+{
+ struct cpl_close_con_rpl *rpl = cplhdr(skb);
+
+ c3cn->snd_una = ntohl(rpl->snd_nxt) - 1; /* exclude FIN */
+
+ if (c3cn_flag(c3cn, C3CN_ABORT_RPL_PENDING))
+ goto out;
+
+ if (c3cn->state == C3CN_STATE_CLOSING) {
+ t3_release_offload_resources(c3cn);
+ c3cn_done(c3cn);
+ } else
+ printk(KERN_ERR
+ "%s: TID %u received CLOSE_CON_RPL in bad state %d\n",
+ c3cn->cdev->name, c3cn->tid, c3cn->state);
+out:
+ kfree_skb(skb);
+}
+
+/*
+ * Random utility functions for CPL message processing ...
+ * =======================================================
+ */
+
+/**
+ * find_best_mtu - find the entry in the MTU table closest to an MTU
+ * @d: TOM state
+ * @mtu: the target MTU
+ *
+ * Returns the index of the value in the MTU table that is closest to but
+ * does not exceed the target MTU.
+ */
+static unsigned int find_best_mtu(const struct t3c_data *d, unsigned short mtu)
+{
+ int i = 0;
+
+ while (i < d->nmtus - 1 && d->mtus[i + 1] <= mtu)
+ ++i;
+ return i;
+}
+
+static unsigned int select_mss(struct s3_conn *c3cn, unsigned int pmtu)
+{
+ unsigned int idx;
+ struct dst_entry *dst = c3cn->dst_cache;
+ struct t3cdev *cdev = c3cn->cdev;
+ const struct t3c_data *td = T3C_DATA(cdev);
+ u16 advmss = dst_metric(dst, RTAX_ADVMSS);
+
+ if (advmss > pmtu - 40)
+ advmss = pmtu - 40;
+ if (advmss < td->mtus[0] - 40)
+ advmss = td->mtus[0] - 40;
+ idx = find_best_mtu(td, advmss + 40);
+ return idx;
+}
+
+static void fail_act_open(struct s3_conn *c3cn, int errno)
+{
+ c3cn->err = errno;
+ t3_release_offload_resources(c3cn);
+ c3cn_done(c3cn);
+}
+
+/*
+ * Assign offload parameters to some connection fields.
+ */
+static void init_offload_conn(struct s3_conn *c3cn,
+ struct t3cdev *cdev,
+ struct dst_entry *dst)
+{
+ BUG_ON(c3cn->cdev != cdev);
+ c3cn->wr_max = c3cn->wr_avail = T3C_DATA(cdev)->max_wrs;
+ c3cn->wr_unacked = 0;
+ c3cn->mss_idx = select_mss(c3cn, dst_mtu(dst));
+
+ c3cn->ctrl_skb_cache = alloc_skb(CTRL_SKB_LEN, gfp_any());
+ reset_wr_list(c3cn);
+}
+
+static void act_open_retry_timer(unsigned long data)
+{
+ struct sk_buff *skb;
+ struct s3_conn *c3cn = (struct s3_conn *)data;
+
+ spin_lock(&c3cn->lock);
+ skb = alloc_skb(sizeof(struct cpl_act_open_req), GFP_ATOMIC);
+ if (!skb)
+ fail_act_open(c3cn, ENOMEM);
+ else {
+ skb->sk = (struct sock *)c3cn;
+ set_arp_failure_handler(skb, act_open_req_arp_failure);
+ mk_act_open_req(c3cn, skb, c3cn->tid, c3cn->l2t);
+ l2t_send(c3cn->cdev, skb, c3cn->l2t);
+ }
+ spin_unlock(&c3cn->lock);
+ c3cn_put(c3cn);
+}
+
+/*
+ * Convert an ACT_OPEN_RPL status to a Linux errno.
+ */
+static int act_open_rpl_status_to_errno(int status)
+{
+ switch (status) {
+ case CPL_ERR_CONN_RESET:
+ return ECONNREFUSED;
+ case CPL_ERR_ARP_MISS:
+ return EHOSTUNREACH;
+ case CPL_ERR_CONN_TIMEDOUT:
+ return ETIMEDOUT;
+ case CPL_ERR_TCAM_FULL:
+ return ENOMEM;
+ case CPL_ERR_CONN_EXIST:
+ printk(KERN_ERR "ACTIVE_OPEN_RPL: 4-tuple in use\n");
+ return EADDRINUSE;
+ default:
+ return EIO;
+ }
+}
+
+/*
+ * Convert the status code of an ABORT_REQ into a Linux error code. Also
+ * indicate whether RST should be sent in response.
+ */
+static int abort_status_to_errno(struct s3_conn *c3cn,
+ int abort_reason, int *need_rst)
+{
+ switch (abort_reason) {
+ case CPL_ERR_BAD_SYN: /* fall through */
+ case CPL_ERR_CONN_RESET:
+ return c3cn->state == C3CN_STATE_CLOSING ? EPIPE : ECONNRESET;
+ case CPL_ERR_XMIT_TIMEDOUT:
+ case CPL_ERR_PERSIST_TIMEDOUT:
+ case CPL_ERR_FINWAIT2_TIMEDOUT:
+ case CPL_ERR_KEEPALIVE_TIMEDOUT:
+ return ETIMEDOUT;
+ default:
+ return EIO;
+ }
+}
+
+static void send_abort_rpl(struct sk_buff *skb, struct t3cdev *cdev,
+ int rst_status)
+{
+ struct sk_buff *reply_skb;
+ struct cpl_abort_req_rss *req = cplhdr(skb);
+
+ reply_skb = get_cpl_reply_skb(skb, sizeof(struct cpl_abort_rpl),
+ gfp_any());
+ if (!reply_skb) {
+ /* Defer the reply. Stick rst_status into req->cmd. */
+ req->status = rst_status;
+ t3_defer_reply(skb, cdev, send_deferred_abort_rpl);
+ return;
+ }
+
+ reply_skb->priority = CPL_PRIORITY_DATA;
+ set_abort_rpl_wr(reply_skb, GET_TID(req), rst_status);
+ kfree_skb(skb);
+ cxgb3_ofld_send(cdev, reply_skb);
+}
+
+/*
+ * Returns an sk_buff for a reply CPL message of size len. If the input
+ * sk_buff has no other users it is trimmed and reused, otherwise a new buffer
+ * is allocated. The input skb must be of size at least len. Note that this
+ * operation does not destroy the original skb data even if it decides to reuse
+ * the buffer.
+ */
+static struct sk_buff *get_cpl_reply_skb(struct sk_buff *skb, size_t len,
+ gfp_t gfp)
+{
+ if (likely(!skb_cloned(skb))) {
+ BUG_ON(skb->len < len);
+ __skb_trim(skb, len);
+ skb_get(skb);
+ } else {
+ skb = alloc_skb(len, gfp);
+ if (skb)
+ __skb_put(skb, len);
+ }
+ return skb;
+}
+
+/*
+ * Add an skb to the deferred skb queue for processing from process context.
+ */
+static void t3_defer_reply(struct sk_buff *skb, struct t3cdev *cdev,
+ defer_handler_t handler)
+{
+ struct cxgb3i_sdev_data *cdata = CXGB3_SDEV_DATA(cdev);
+
+ DEFERRED_SKB_CB(skb)->handler = handler;
+ spin_lock_bh(&cdata->deferq.lock);
+ __skb_queue_tail(&cdata->deferq, skb);
+ if (skb_queue_len(&cdata->deferq) == 1)
+ schedule_work(&cdata->deferq_task);
+ spin_unlock_bh(&cdata->deferq.lock);
+}
+
+/*
+ * Process the defer queue.
+ */
+static void process_deferq(struct work_struct *task_param)
+{
+ struct sk_buff *skb;
+ struct cxgb3i_sdev_data *cdata = container_of(task_param,
+ struct cxgb3i_sdev_data,
+ deferq_task);
+
+ spin_lock_bh(&cdata->deferq.lock);
+ while ((skb = __skb_dequeue(&cdata->deferq)) != NULL) {
+ spin_unlock_bh(&cdata->deferq.lock);
+ DEFERRED_SKB_CB(skb)->handler(cdata->cdev, skb);
+ spin_lock_bh(&cdata->deferq.lock);
+ }
+ spin_unlock_bh(&cdata->deferq.lock);
+}
+
+static void send_deferred_abort_rpl(struct t3cdev *cdev, struct sk_buff *skb)
+{
+ struct sk_buff *reply_skb;
+ struct cpl_abort_req_rss *req = cplhdr(skb);
+
+ reply_skb = alloc_skb(sizeof(struct cpl_abort_rpl),
+ GFP_KERNEL | __GFP_NOFAIL);
+ reply_skb->priority = CPL_PRIORITY_DATA;
+ __skb_put(reply_skb, sizeof(struct cpl_abort_rpl));
+ set_abort_rpl_wr(reply_skb, GET_TID(req), req->status);
+ cxgb3_ofld_send(cdev, reply_skb);
+ kfree_skb(skb);
+}
+
+/*
+ * Release resources held by an offload connection (TID, L2T entry, etc.)
+ */
+static void t3_release_offload_resources(struct s3_conn *c3cn)
+{
+ struct t3cdev *cdev = c3cn->cdev;
+ unsigned int tid = c3cn->tid;
+
+ if (!cdev)
+ return;
+
+ c3cn->qset = 0;
+
+ kfree_skb(c3cn->ctrl_skb_cache);
+ c3cn->ctrl_skb_cache = NULL;
+
+ if (c3cn->wr_avail != c3cn->wr_max) {
+ purge_wr_queue(c3cn);
+ reset_wr_list(c3cn);
+ }
+
+ if (c3cn->l2t) {
+ l2t_release(L2DATA(cdev), c3cn->l2t);
+ c3cn->l2t = NULL;
+ }
+
+ if (c3cn->state == C3CN_STATE_SYN_SENT) /* we have ATID */
+ free_atid(cdev, tid);
+ else { /* we have TID */
+ cxgb3_remove_tid(cdev, (void *)c3cn, tid);
+ c3cn_put(c3cn);
+ }
+
+ c3cn->cdev = NULL;
+}
+
+/*
+ * Handles Rx data that arrives in a state where the connection isn't
+ * accepting new data.
+ */
+static void handle_excess_rx(struct s3_conn *c3cn, struct sk_buff *skb)
+{
+ if (!c3cn_flag(c3cn, C3CN_ABORT_SHUTDOWN))
+ abort_conn(c3cn, skb);
+
+ kfree_skb(skb);
+}
+
+/*
+ * Like get_cpl_reply_skb() but the returned buffer starts out empty.
+ */
+static struct sk_buff *__get_cpl_reply_skb(struct sk_buff *skb, size_t len,
+ gfp_t gfp)
+{
+ if (likely(!skb_cloned(skb) && !skb->data_len)) {
+ __skb_trim(skb, 0);
+ skb_get(skb);
+ } else
+ skb = alloc_skb(len, gfp);
+ return skb;
+}
+
+/*
+ * Completes some final bits of initialization for just established connections
+ * and changes their state to C3CN_STATE_ESTABLISHED.
+ *
+ * snd_isn here is the ISN after the SYN, i.e., the true ISN + 1.
+ */
+static void make_established(struct s3_conn *c3cn, u32 snd_isn,
+ unsigned int opt)
+{
+ c3cn->write_seq = c3cn->snd_nxt = c3cn->snd_una = snd_isn;
+
+ /*
+ * Causes the first RX_DATA_ACK to supply any Rx credits we couldn't
+ * pass through opt0.
+ */
+ if (rcv_win > (M_RCV_BUFSIZ << 10))
+ c3cn->rcv_wup -= rcv_win - (M_RCV_BUFSIZ << 10);
+
+ dst_confirm(c3cn->dst_cache);
+
+ smp_mb();
+ c3cn_set_state(c3cn, C3CN_STATE_ESTABLISHED);
+}
diff --git a/drivers/scsi/cxgb3i/cxgb3i_offload.h b/drivers/scsi/cxgb3i/cxgb3i_offload.h
new file mode 100644
index 0000000..89217b4
--- /dev/null
+++ b/drivers/scsi/cxgb3i/cxgb3i_offload.h
@@ -0,0 +1,252 @@
+/*
+ * Copyright (C) 2003-2008 Chelsio Communications. All rights reserved.
+ *
+ * Written by Dimitris Michailidis (dm@...lsio.com)
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the LICENSE file included in this
+ * release for licensing terms and conditions.
+ */
+
+#ifndef _CXGB3I_OFFLOAD_H
+#define _CXGB3I_OFFLOAD_H
+
+#include <linux/skbuff.h>
+#include <net/tcp.h>
+
+#include "t3cdev.h"
+#include "cxgb3_offload.h"
+
+#define cxgb3i_log_error(fmt...) printk(KERN_ERR "cxgb3i: ERR! " fmt)
+#define cxgb3i_log_warn(fmt...) printk(KERN_WARNING "cxgb3i: WARN! " fmt)
+#define cxgb3i_log_info(fmt...) printk(KERN_INFO "cxgb3i: " fmt)
+
+#ifdef __DEBUG_CXGB3I__
+#define cxgb3i_log_debug(fmt, args...) \
+ printk(KERN_ERR "cxgb3i: %s - " fmt, __func__ , ## args)
+#else
+#define cxgb3i_log_debug(fmt...)
+#endif
+
+/*
+ * Data structure to keep track of cxgb3 connection.
+ */
+struct s3_conn {
+ struct net_device *dev;
+ struct t3cdev *cdev;
+ unsigned long flags;
+ int tid;
+ int qset;
+ int mss_idx;
+ struct l2t_entry *l2t;
+ int wr_max;
+ int wr_avail;
+ int wr_unacked;
+ struct sk_buff *wr_pending_head;
+ struct sk_buff *wr_pending_tail;
+ struct sk_buff *ctrl_skb_cache;
+
+ spinlock_t lock;
+ atomic_t refcnt;
+ volatile unsigned int state;
+ struct sockaddr_in saddr;
+ struct sockaddr_in daddr;
+ struct dst_entry *dst_cache;
+ unsigned char shutdown;
+ struct sk_buff_head receive_queue;
+ struct sk_buff_head write_queue;
+ struct timer_list retry_timer;
+ int err;
+ rwlock_t callback_lock;
+ void *user_data;
+
+ u32 rcv_nxt; /* What we want to receive next */
+ u32 copied_seq; /* Head of yet unread data */
+ u32 rcv_wup; /* rcv_nxt on last window update sent */
+ u32 snd_nxt; /* Next sequence we send */
+ u32 snd_una; /* First byte we want an ack for */
+
+ u32 write_seq; /* Tail(+1) of data held in send buffer */
+};
+
+/* Flags in c3cn->shutdown */
+#define C3CN_SHUTDOWN_MASK 3
+#define C3CN_RCV_SHUTDOWN 1
+#define C3CN_SEND_SHUTDOWN 2
+
+/*
+ * connection state bitmap
+ */
+#define C3CN_STATE_CLOSE 0x1
+#define C3CN_STATE_SYN_SENT 0x2
+#define C3CN_STATE_ESTABLISHED 0x4
+#define C3CN_STATE_CLOSING 0x8
+#define C3CN_STATE_ABORING 0x10
+
+#define C3CN_STATE_MASK 0xFF
+#define C3CN_NEED_CLOSE 0x100
+
+static inline unsigned int c3cn_in_state(const struct s3_conn *c3cn,
+ unsigned int states)
+{
+ return (states & c3cn->state);
+}
+
+
+/*
+ * Connection flags -- many to track some close related events.
+ */
+enum c3cn_flags {
+ C3CN_ABORT_RPL_RCVD, /* received one ABORT_RPL_RSS message */
+ C3CN_ABORT_REQ_RCVD, /* received one ABORT_REQ_RSS message */
+ C3CN_TX_WAIT_IDLE, /* suspend Tx until in-flight data is ACKed */
+ C3CN_ABORT_SHUTDOWN, /* shouldn't send more abort requests */
+ C3CN_ABORT_RPL_PENDING, /* expecting an abort reply */
+ C3CN_CLOSE_CON_REQUESTED, /* we've sent a close_conn_req */
+ C3CN_TX_DATA_SENT, /* already sent a TX_DATA WR */
+
+ C3CN_DONE,
+};
+
+static inline void c3cn_set_flag(struct s3_conn *c3cn,
+ enum c3cn_flags flag)
+{
+ __set_bit(flag, &c3cn->flags);
+}
+
+static inline void c3cn_reset_flag(struct s3_conn *c3cn,
+ enum c3cn_flags flag)
+{
+ __clear_bit(flag, &c3cn->flags);
+}
+
+static inline int c3cn_flag(struct s3_conn *c3cn, enum c3cn_flags flag)
+{
+ if (c3cn == NULL)
+ return 0;
+ return test_bit(flag, &c3cn->flags);
+}
+
+/*
+ * Per adapter data. Linked off of each Ethernet device port on the adapter.
+ * Also available via the t3cdev structure since we have pointers to our port
+ * net_device's there ...
+ */
+struct cxgb3i_sdev_data {
+ struct list_head list;
+ struct t3cdev *cdev;
+ struct cxgb3_client *client;
+ struct adap_ports *ports;
+ unsigned int rx_page_size;
+ struct sk_buff_head deferq;
+ struct work_struct deferq_task;
+};
+#define NDEV2CDATA(ndev) (*(struct cxgb3i_sdev_data **)&(ndev)->ec_ptr)
+#define CXGB3_SDEV_DATA(cdev) NDEV2CDATA((cdev)->lldev)
+
+static inline void c3cn_hold(struct s3_conn *c3cn)
+{
+ atomic_inc(&c3cn->refcnt);
+}
+
+static inline void c3cn_put(struct s3_conn *c3cn)
+{
+ if (atomic_dec_and_test(&c3cn->refcnt))
+ kfree(c3cn);
+}
+
+void c3cn_close(struct s3_conn *);
+static inline void c3cn_release(struct s3_conn *c3cn)
+{
+ c3cn_close(c3cn);
+ c3cn_put(c3cn);
+}
+
+/*
+ * Primary API routines.
+ */
+
+int cxgb3i_sdev_init(cxgb3_cpl_handler_func *);
+void cxgb3i_sdev_add(struct t3cdev *, struct cxgb3_client *);
+void cxgb3i_sdev_remove(struct t3cdev *);
+
+struct s3_conn *cxgb3i_c3cn_create(void);
+int cxgb3i_c3cn_connect(struct s3_conn *, struct sockaddr_in *);
+void cxgb3i_c3cn_rx_credits(struct s3_conn *, int);
+int cxgb3i_c3cn_send_pdus(struct s3_conn *, struct sk_buff *, int);
+
+/*
+ * Definitions for sk_buff state and ULP mode management.
+ */
+
+struct cxgb3_skb_cb {
+ __u8 flags;
+ __u8 ulp_mode; /* ULP mode/submode of sk_buff */
+ __u32 seq; /* sequence number */
+ __u32 ddigest; /* ULP rx_data_ddp selected field */
+ __u32 pdulen; /* ULP rx_data_ddp selected field */
+ __u8 ulp_data[16]; /* scratch area for ULP */
+};
+
+#define CXGB3_SKB_CB(skb) ((struct cxgb3_skb_cb *)&((skb)->cb[0]))
+
+#define skb_ulp_mode(skb) (CXGB3_SKB_CB(skb)->ulp_mode)
+#define skb_ulp_ddigest(skb) (CXGB3_SKB_CB(skb)->ddigest)
+#define skb_ulp_pdulen(skb) (CXGB3_SKB_CB(skb)->pdulen)
+#define skb_ulp_data(skb) (CXGB3_SKB_CB(skb)->ulp_data)
+
+enum {
+ C3CB_FLAG_NEED_HDR = 1 << 0, /* packet needs a TX_DATA_WR header */
+ C3CB_FLAG_NO_APPEND = 1 << 1, /* don't grow this skb */
+ C3CB_FLAG_BARRIER = 1 << 2, /* set TX_WAIT_IDLE after sending */
+ C3CB_FLAG_COMPL = 1 << 4, /* request WR completion */
+};
+
+/*
+ * Definitions for managing deferred CPL replies from process context.
+ */
+
+typedef void (*defer_handler_t) (struct t3cdev *, struct sk_buff *);
+
+struct deferred_skb_cb {
+ defer_handler_t handler;
+ struct t3cdev *cdev;
+};
+
+#define DEFERRED_SKB_CB(skb) ((struct deferred_skb_cb *)(skb)->cb)
+
+/*
+ * Top-level CPL message processing used by most CPL messages that
+ * pertain to connections.
+ */
+static inline void process_cpl_msg(void (*fn)(struct s3_conn *,
+ struct sk_buff *),
+ struct s3_conn *c3cn,
+ struct sk_buff *skb)
+{
+cxgb3i_log_debug("c3cn 0x%p, locking ....\n", c3cn);
+ spin_lock(&c3cn->lock);
+cxgb3i_log_debug("c3cn 0x%p, locked.\n", c3cn);
+ fn(c3cn, skb);
+cxgb3i_log_debug("c3cn 0x%p, unlock.\n", c3cn);
+ spin_unlock(&c3cn->lock);
+}
+
+/*
+ * Opaque version of structure the SGE stores at skb->head of TX_DATA packets
+ * and for which we must reserve space.
+ */
+struct sge_opaque_hdr {
+ void *dev;
+ dma_addr_t addr[MAX_SKB_FRAGS + 1];
+};
+
+/* for TX: a skb must have a headroom of at least TX_HEADER_LEN bytes */
+#define TX_HEADER_LEN \
+ (sizeof(struct tx_data_wr) + sizeof(struct sge_opaque_hdr))
+
+void *cxgb3i_alloc_big_mem(unsigned int);
+void cxgb3i_free_big_mem(void *);
+
+#endif /* _CXGB3_OFFLOAD_H */
diff --git a/drivers/scsi/cxgb3i/cxgb3i_ulp2.c b/drivers/scsi/cxgb3i/cxgb3i_ulp2.c
new file mode 100644
index 0000000..313bb90
--- /dev/null
+++ b/drivers/scsi/cxgb3i/cxgb3i_ulp2.c
@@ -0,0 +1,692 @@
+/*
+ * cxgb3i_ddp.c: Chelsio S3xx iSCSI driver.
+ *
+ * Copyright (c) 2008 Chelsio Communications, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation.
+ *
+ * Written by: Karen Xie (kxie@...lsio.com)
+ */
+
+#include <linux/skbuff.h>
+#include <scsi/scsi_cmnd.h>
+#include <scsi/scsi_host.h>
+#include <linux/crypto.h>
+#include "../iscsi_tcp.h"
+
+#include "cxgb3i.h"
+#include "cxgb3i_ulp2.h"
+
+#ifdef __DEBUG_CXGB3I_RX__
+#define cxgb3i_rx_debug cxgb3i_log_debug
+#else
+#define cxgb3i_rx_debug(fmt...)
+#endif
+
+#ifdef __DEBUG_CXGB3I_TX__
+#define cxgb3i_tx_debug cxgb3i_log_debug
+#else
+#define cxgb3i_tx_debug(fmt...)
+#endif
+
+#ifdef __DEBUG_CXGB3I_TAG__
+#define cxgb3i_tag_debug cxgb3i_log_debug
+#else
+#define cxgb3i_tag_debug(fmt...)
+#endif
+
+#ifdef __DEBUG_CXGB3I_DDP__
+#define cxgb3i_ddp_debug cxgb3i_log_debug
+#else
+#define cxgb3i_ddp_debug(fmt...)
+#endif
+
+static struct page *pad_page;
+
+#define ULP2_PGIDX_MAX 4
+#define ULP2_4K_PAGE_SHIFT 12
+#define ULP2_4K_PAGE_MASK (~((1UL << ULP2_4K_PAGE_SHIFT) - 1))
+static unsigned char ddp_page_order[ULP2_PGIDX_MAX];
+static unsigned long ddp_page_size[ULP2_PGIDX_MAX];
+static unsigned char ddp_page_shift[ULP2_PGIDX_MAX];
+static unsigned char sw_tag_idx_bits;
+static unsigned char sw_tag_age_bits;
+
+static void cxgb3i_ddp_page_init(void)
+{
+ int i;
+ unsigned long n = PAGE_SIZE >> ULP2_4K_PAGE_SHIFT;
+
+ if (PAGE_SIZE & (~ULP2_4K_PAGE_MASK)) {
+ cxgb3i_log_warn("PAGE_SIZE 0x%lx is not multiple of 4K, "
+ "ddp disabled.\n", PAGE_SIZE);
+ return;
+ }
+ n = __ilog2_u32(n);
+ for (i = 0; i < ULP2_PGIDX_MAX; i++, n++) {
+ ddp_page_order[i] = n;
+ ddp_page_shift[i] = ULP2_4K_PAGE_SHIFT + n;
+ ddp_page_size[i] = 1 << ddp_page_shift[i];
+ cxgb3i_log_debug("%d, order %u, shift %u, size 0x%lx.\n", i,
+ ddp_page_order[i], ddp_page_shift[i],
+ ddp_page_size[i]);
+ }
+
+ sw_tag_idx_bits = (__ilog2_u32(ISCSI_ITT_MASK)) + 1;
+ sw_tag_age_bits = (__ilog2_u32(ISCSI_AGE_MASK)) + 1;
+}
+
+static inline void ulp_mem_io_set_hdr(struct sk_buff *skb, unsigned int addr)
+{
+ struct ulp_mem_io *req = (struct ulp_mem_io *)skb->head;
+ req->wr.wr_lo = 0;
+ req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_BYPASS));
+ req->cmd_lock_addr = htonl(V_ULP_MEMIO_ADDR(addr >> 5) |
+ V_ULPTX_CMD(ULP_MEM_WRITE));
+ req->len = htonl(V_ULP_MEMIO_DATA_LEN(PPOD_SIZE >> 5) |
+ V_ULPTX_NFLITS((PPOD_SIZE >> 3) + 1));
+}
+
+static int set_ddp_map(struct cxgb3i_adapter *snic, struct pagepod_hdr *hdr,
+ unsigned int idx, unsigned int npods,
+ struct scatterlist *sgl, unsigned int sgcnt)
+{
+ struct cxgb3i_ddp_info *ddp = &snic->ddp;
+ struct scatterlist *sg = sgl;
+ unsigned int pm_addr = (idx << PPOD_SIZE_SHIFT) + ddp->llimit;
+ int i;
+
+ for (i = 0; i < npods; i++, pm_addr += PPOD_SIZE) {
+ struct sk_buff *skb;
+ struct pagepod *ppod;
+ int j, k;
+ skb =
+ alloc_skb(sizeof(struct ulp_mem_io) + PPOD_SIZE,
+ GFP_ATOMIC);
+ if (!skb)
+ return -ENOMEM;
+ skb_put(skb, sizeof(struct ulp_mem_io) + PPOD_SIZE);
+
+ ulp_mem_io_set_hdr(skb, pm_addr);
+ ppod =
+ (struct pagepod *)(skb->head + sizeof(struct ulp_mem_io));
+ memcpy(&(ppod->hdr), hdr, sizeof(struct pagepod));
+ for (j = 0, k = i * 4; j < 5; j++, k++) {
+ if (k < sgcnt) {
+ ppod->addr[j] = cpu_to_be64(sg_dma_address(sg));
+ if (j < 4)
+ sg = sg_next(sg);
+ } else
+ ppod->addr[j] = 0UL;
+ }
+
+ skb->priority = CPL_PRIORITY_CONTROL;
+ cxgb3_ofld_send(snic->tdev, skb);
+ }
+ return 0;
+}
+
+static int clear_ddp_map(struct cxgb3i_adapter *snic, unsigned int idx,
+ unsigned int npods)
+{
+ struct cxgb3i_ddp_info *ddp = &snic->ddp;
+ unsigned int pm_addr = (idx << PPOD_SIZE_SHIFT) + ddp->llimit;
+ int i;
+
+ for (i = 0; i < npods; i++, pm_addr += PPOD_SIZE) {
+ struct sk_buff *skb;
+ skb =
+ alloc_skb(sizeof(struct ulp_mem_io) + PPOD_SIZE,
+ GFP_ATOMIC);
+ if (!skb)
+ return -ENOMEM;
+ skb_put(skb, sizeof(struct ulp_mem_io) + PPOD_SIZE);
+ memset((skb->head + sizeof(struct ulp_mem_io)), 0, PPOD_SIZE);
+ ulp_mem_io_set_hdr(skb, pm_addr);
+ skb->priority = CPL_PRIORITY_CONTROL;
+ cxgb3_ofld_send(snic->tdev, skb);
+ }
+ return 0;
+}
+
+static int cxgb3i_ddp_sgl_check(struct scatterlist *sgl, unsigned int sgcnt)
+{
+ struct scatterlist *sg;
+ int i;
+
+ /* make sure the sgl is fit for ddp:
+ * each has the same page size, and
+ * first & last page do not need to be used completely, and
+ * the rest of page must be used completely
+ */
+ for_each_sg(sgl, sg, sgcnt, i) {
+ if ((i && sg->offset) ||
+ ((i != sgcnt - 1) &&
+ (sg->length + sg->offset) != PAGE_SIZE))
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static inline int ddp_find_unused_entries(struct cxgb3i_ddp_info *ddp,
+ int start, int max, int count)
+{
+ unsigned int i, j;
+
+ spin_lock(&ddp->map_lock);
+ for (i = start; i <= max;) {
+ for (j = 0; j < count; j++) {
+ if (ddp->map[i + j])
+ break;
+ }
+ if (j == count) {
+ memset(&ddp->map[i], 1, count);
+ spin_unlock(&ddp->map_lock);
+ return i;
+ }
+ i += j + 1;
+ }
+ spin_unlock(&ddp->map_lock);
+ return -EBUSY;
+}
+
+static inline void ddp_unmark_entries(struct cxgb3i_ddp_info *ddp,
+ int start, int count)
+{
+ spin_lock(&ddp->map_lock);
+ memset(&ddp->map[start], 0, count);
+ spin_unlock(&ddp->map_lock);
+}
+
+u32 cxgb3i_ddp_tag_reserve(struct cxgb3i_adapter *snic, unsigned int tid,
+ u32 sw_tag, unsigned int xferlen,
+ struct scatterlist *sgl, unsigned int sgcnt)
+{
+ struct cxgb3i_ddp_info *ddp = &snic->ddp;
+ struct pagepod_hdr hdr;
+ unsigned int npods;
+ int idx = -1, idx_max;
+ u32 tag;
+ int err;
+
+ if (!ddp || !sgcnt || xferlen < PAGE_SIZE)
+ return RESERVED_ITT;
+
+ err = cxgb3i_ddp_sgl_check(sgl, sgcnt);
+ if (err < 0)
+ return RESERVED_ITT;
+
+ npods = (sgcnt + PPOD_PAGES_MAX - 1) >> PPOD_PAGES_SHIFT;
+ idx_max = ddp->nppods - npods + 1;
+
+ if (ddp->idx_last == ddp->nppods)
+ idx = ddp_find_unused_entries(ddp, 0, idx_max, npods);
+ else {
+ idx = ddp_find_unused_entries(ddp, ddp->idx_last + 1, idx_max,
+ npods);
+ if ((idx < 0) && (ddp->idx_last >= npods))
+ idx = ddp_find_unused_entries(ddp, 0,
+ ddp->idx_last - npods + 1,
+ npods);
+ }
+ if (idx < 0)
+ return RESERVED_ITT;
+
+ if (pci_map_sg(snic->pdev, sgl, sgcnt, PCI_DMA_FROMDEVICE) <= 0)
+ goto unmark_entries;
+
+ tag = sw_tag | (idx << snic->tag_format.rsvd_shift);
+
+ hdr.rsvd = 0;
+ hdr.vld_tid = htonl(F_PPOD_VALID | V_PPOD_TID(tid));
+ hdr.pgsz_tag_clr = htonl(tag);
+ hdr.maxoffset = htonl(xferlen);
+ hdr.pgoffset = htonl(sgl->offset);
+
+ if (set_ddp_map(snic, &hdr, idx, npods, sgl, sgcnt) < 0)
+ goto unmap_sgl;
+
+ ddp->idx_last = idx;
+ cxgb3i_tag_debug("tid 0x%x, xfer %u, 0x%x -> ddp tag 0x%x (%u, %u).\n",
+ tid, xferlen, sw_tag, tag, idx, npods);
+ return tag;
+
+unmap_sgl:
+ pci_unmap_sg(snic->pdev, sgl, sgcnt, PCI_DMA_FROMDEVICE);
+
+unmark_entries:
+ ddp_unmark_entries(ddp, idx, npods);
+ return RESERVED_ITT;
+}
+
+void cxgb3i_ddp_tag_release(struct cxgb3i_adapter *snic, u32 tag,
+ struct scatterlist *sgl, unsigned int sgcnt)
+{
+ u32 idx = (tag >> snic->tag_format.rsvd_shift) &
+ snic->tag_format.rsvd_mask;
+ unsigned int npods = (sgcnt + PPOD_PAGES_MAX - 1) >> PPOD_PAGES_SHIFT;
+
+ if (idx < snic->tag_format.rsvd_mask) {
+ cxgb3i_tag_debug("ddp tag 0x%x, release idx 0x%x, npods %u.\n",
+ tag, idx, npods);
+ clear_ddp_map(snic, idx, npods);
+ ddp_unmark_entries(&snic->ddp, idx, npods);
+ pci_unmap_sg(snic->pdev, sgl, sgcnt, PCI_DMA_FROMDEVICE);
+ }
+}
+
+int cxgb3i_conn_ulp_setup(struct cxgb3i_conn *cconn, int hcrc, int dcrc)
+{
+ struct iscsi_tcp_conn *tcp_conn = cconn->conn->dd_data;
+ struct s3_conn *c3cn = (struct s3_conn *)(tcp_conn->sock);
+ struct sk_buff *skb = alloc_skb(sizeof(struct cpl_set_tcb_field),
+ GFP_KERNEL | __GFP_NOFAIL);
+ struct cpl_set_tcb_field *req;
+ u32 submode = (hcrc ? 1 : 0) | (dcrc ? 2 : 0);
+
+ /* set up ulp submode and page size */
+ req = (struct cpl_set_tcb_field *)skb_put(skb, sizeof(*req));
+ req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_FORWARD));
+ OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_SET_TCB_FIELD, c3cn->tid));
+ req->reply = V_NO_REPLY(1);
+ req->cpu_idx = 0;
+ req->word = htons(31);
+ req->mask = cpu_to_be64(0xFF000000);
+ /* the connection page size is always the same as ddp-pgsz0 */
+ req->val = cpu_to_be64(submode << 24);
+ skb->priority = CPL_PRIORITY_CONTROL;
+
+ cxgb3_ofld_send(c3cn->cdev, skb);
+ return 0;
+}
+
+static int cxgb3i_conn_read_pdu_skb(struct iscsi_conn *conn,
+ struct sk_buff *skb)
+{
+ struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
+ struct iscsi_segment *segment = &tcp_conn->in.segment;
+ struct iscsi_hdr *hdr = (struct iscsi_hdr *)tcp_conn->in.hdr_buf;
+ unsigned char *buf = (unsigned char *)hdr;
+ unsigned int offset = sizeof(struct iscsi_hdr);
+ int err;
+
+ cxgb3i_rx_debug("conn 0x%p, skb 0x%p, len %u, flag 0x%x.\n",
+ conn, skb, skb->len, skb_ulp_mode(skb));
+
+ /* read bhs */
+ err = skb_copy_bits(skb, 0, buf, sizeof(struct iscsi_hdr));
+ if (err < 0)
+ return err;
+ segment->copied = sizeof(struct iscsi_hdr);
+ /* read ahs */
+ if (hdr->hlength) {
+ unsigned int ahslen = hdr->hlength << 2;
+ /* Make sure we don't overflow */
+ if (sizeof(*hdr) + ahslen > sizeof(tcp_conn->in.hdr_buf))
+ return -ISCSI_ERR_AHSLEN;
+ err = skb_copy_bits(skb, offset, buf + offset, ahslen);
+ if (err < 0)
+ return err;
+ offset += ahslen;
+ }
+ /* header digest */
+ if (conn->hdrdgst_en)
+ offset += ISCSI_DIGEST_SIZE;
+
+ /* check header digest */
+ segment->status = (conn->hdrdgst_en &&
+ (skb_ulp_mode(skb) & ULP2_FLAG_HCRC_ERROR)) ?
+ ISCSI_SEGMENT_DGST_ERR : 0;
+
+ hdr->itt = ntohl(hdr->itt);
+ segment->total_copied = segment->total_size;
+ tcp_conn->in.hdr = hdr;
+ err = iscsi_tcp_hdr_dissect(conn, hdr);
+ if (err)
+ return err;
+
+ if (tcp_conn->in.datalen) {
+ segment = &tcp_conn->in.segment;
+ segment->status = (conn->datadgst_en &&
+ (skb_ulp_mode(skb) & ULP2_FLAG_DCRC_ERROR)) ?
+ ISCSI_SEGMENT_DGST_ERR : 0;
+ if (skb_ulp_mode(skb) & ULP2_FLAG_DATA_DDPED) {
+ cxgb3i_ddp_debug("opcode 0x%x, data %u, ddp'ed.\n",
+ hdr->opcode & ISCSI_OPCODE_MASK,
+ tcp_conn->in.datalen);
+ segment->total_copied = segment->total_size;
+ } else {
+ cxgb3i_ddp_debug("opcode 0x%x, data %u, not ddp'ed.\n",
+ hdr->opcode & ISCSI_OPCODE_MASK,
+ tcp_conn->in.datalen);
+ offset += sizeof(struct cpl_iscsi_hdr_norss);
+ }
+ while (segment->total_copied < segment->total_size) {
+ iscsi_tcp_segment_map(segment, 1);
+ err = skb_copy_bits(skb, offset, segment->data,
+ segment->size);
+ iscsi_tcp_segment_unmap(segment);
+ if (err)
+ return err;
+ segment->total_copied += segment->size;
+ offset += segment->size;
+
+ if (segment->total_copied < segment->total_size)
+ iscsi_tcp_segment_init_sg(segment,
+ sg_next(segment->sg),
+ 0);
+ }
+ err = segment->done(tcp_conn, segment);
+ }
+ return err;
+}
+
+static inline void tx_skb_setmode(struct sk_buff *skb, int hcrc, int dcrc)
+{
+ u8 submode = 0;
+ if (hcrc)
+ submode |= 1;
+ if (dcrc)
+ submode |= 2;
+ skb_ulp_mode(skb) = (ULP_MODE_ISCSI << 4) | submode;
+}
+
+int cxgb3i_conn_ulp2_xmit(struct iscsi_conn *conn)
+{
+ struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
+ struct iscsi_segment *hdr_seg = &tcp_conn->out.segment;
+ struct iscsi_segment *data_seg = &tcp_conn->out.data_segment;
+ unsigned int hdrlen = hdr_seg->total_size;
+ unsigned int datalen = data_seg->total_size;
+ unsigned int padlen = iscsi_padding(datalen);
+ unsigned int copymax = SKB_MAX_HEAD(TX_HEADER_LEN);
+ unsigned int copylen;
+ struct sk_buff *skb;
+ unsigned char *dst;
+ int err = -EAGAIN;
+
+ if (conn->suspend_tx)
+ return 0;
+
+ if (data_seg->data && ((datalen + padlen) < copymax))
+ copylen = hdrlen + datalen + padlen;
+ else
+ copylen = hdrlen;
+
+ /* supports max. 16K pdus, so one skb is enough to hold all the data */
+ skb = alloc_skb(TX_HEADER_LEN + copylen, GFP_ATOMIC);
+ if (!skb)
+ return -EAGAIN;
+
+ skb_reserve(skb, TX_HEADER_LEN);
+ skb_put(skb, copylen);
+ dst = skb->data;
+
+ tx_skb_setmode(skb, conn->hdrdgst_en, datalen ? conn->datadgst_en : 0);
+
+ memcpy(dst, hdr_seg->data, hdrlen);
+ dst += hdrlen;
+
+ if (!datalen)
+ goto send_pdu;
+
+ if (data_seg->data) {
+ /* data is in a linear buffer */
+ if (copylen > hdrlen) {
+ /* data fits in the skb's headroom */
+ memcpy(dst, data_seg->data, datalen);
+ dst += datalen;
+ if (padlen)
+ memset(dst, 0, padlen);
+ } else {
+ unsigned int offset = 0;
+ while (datalen) {
+ struct page *page =
+ alloc_pages(GFP_ATOMIC, 0);
+ int idx = skb_shinfo(skb)->nr_frags;
+ skb_frag_t *frag = &skb_shinfo(skb)->frags[idx];
+
+ if (!page)
+ goto free_skb;
+
+ frag->page = page;
+ frag->page_offset = 0;
+ if (datalen > PAGE_SIZE)
+ frag->size = PAGE_SIZE;
+ else
+ frag->size = datalen;
+ memcpy(page_address(page),
+ data_seg->data + offset, frag->size);
+
+ skb_shinfo(skb)->nr_frags++;
+ datalen -= frag->size;
+ offset += frag->size;
+ }
+ }
+ } else {
+ struct scatterlist *sg = data_seg->sg;
+ unsigned int offset = data_seg->sg_offset;
+ while (datalen) {
+ int idx = skb_shinfo(skb)->nr_frags;
+ skb_frag_t *frag = &skb_shinfo(skb)->frags[idx];
+ struct page *pg = sg_page(sg);
+
+ get_page(pg);
+ frag->page = pg;
+ frag->page_offset = offset + sg->offset;
+ frag->size = min(sg->length, datalen);
+
+ offset = 0;
+ skb_shinfo(skb)->nr_frags++;
+ datalen -= frag->size;
+ sg = sg_next(sg);
+ }
+ }
+
+ if (skb_shinfo(skb)->nr_frags) {
+ if (padlen) {
+ int idx = skb_shinfo(skb)->nr_frags;
+ skb_frag_t *frag = &skb_shinfo(skb)->frags[idx];
+ frag->page = pad_page;
+ frag->page_offset = 0;
+ frag->size = padlen;
+ skb_shinfo(skb)->nr_frags++;
+ }
+ datalen = data_seg->total_size + padlen;
+ skb->data_len += datalen;
+ skb->truesize += datalen;
+ skb->len += datalen;
+ }
+
+send_pdu:
+ err = cxgb3i_c3cn_send_pdus((struct s3_conn *)tcp_conn->sock,
+ skb, MSG_DONTWAIT | MSG_NOSIGNAL);
+ if (err > 0) {
+ int pdulen = hdrlen + datalen + padlen;
+ if (conn->hdrdgst_en)
+ pdulen += ISCSI_DIGEST_SIZE;
+ if (datalen && conn->datadgst_en)
+ pdulen += ISCSI_DIGEST_SIZE;
+
+ hdr_seg->total_copied = hdr_seg->total_size;
+ if (datalen)
+ data_seg->total_copied = data_seg->total_size;
+ conn->txdata_octets += pdulen;
+ return pdulen;
+ }
+
+free_skb:
+ kfree_skb(skb);
+ if (err < 0 && err != -EAGAIN) {
+ cxgb3i_log_error("conn 0x%p, xmit err %d.\n", conn, err);
+ iscsi_conn_failure(conn, ISCSI_ERR_CONN_FAILED);
+ return err;
+ }
+ return -EAGAIN;
+}
+
+int cxgb3i_ulp2_init(void)
+{
+ pad_page = alloc_page(GFP_KERNEL);
+ if (!pad_page)
+ return -ENOMEM;
+ memset(page_address(pad_page), 0, PAGE_SIZE);
+ cxgb3i_ddp_page_init();
+ return 0;
+}
+
+void cxgb3i_ulp2_cleanup(void)
+{
+ if (pad_page) {
+ __free_page(pad_page);
+ pad_page = NULL;
+ }
+}
+
+void cxgb3i_conn_pdu_ready(struct s3_conn *c3cn)
+{
+ struct sk_buff *skb;
+ unsigned int read = 0;
+ struct iscsi_conn *conn = c3cn->user_data;
+ int err = 0;
+
+ cxgb3i_rx_debug("cn 0x%p.\n", c3cn);
+
+ read_lock(&c3cn->callback_lock);
+ if (unlikely(!conn || conn->suspend_rx)) {
+ cxgb3i_rx_debug("conn 0x%p, id %d, suspend_rx %d!\n",
+ conn, conn ? conn->id : 0xFF,
+ conn ? conn->suspend_rx : 0xFF);
+ read_unlock(&c3cn->callback_lock);
+ return;
+ }
+ skb = skb_peek(&c3cn->receive_queue);
+ while (!err && skb) {
+ __skb_unlink(skb, &c3cn->receive_queue);
+ read += skb_ulp_pdulen(skb);
+ err = cxgb3i_conn_read_pdu_skb(conn, skb);
+ __kfree_skb(skb);
+ skb = skb_peek(&c3cn->receive_queue);
+ }
+ read_unlock(&c3cn->callback_lock);
+ if (c3cn) {
+ c3cn->copied_seq += read;
+ cxgb3i_c3cn_rx_credits(c3cn, read);
+ }
+ conn->rxdata_octets += read;
+
+ if (err) {
+ cxgb3i_log_info("conn 0x%p rx failed err %d.\n", conn, err);
+ iscsi_conn_failure(conn, ISCSI_ERR_CONN_FAILED);
+ }
+}
+
+void cxgb3i_conn_tx_open(struct s3_conn *c3cn)
+{
+ struct iscsi_conn *conn = (struct iscsi_conn *)c3cn->user_data;
+ struct iscsi_tcp_conn *tcp_conn;
+ cxgb3i_tx_debug("cn 0x%p.\n", c3cn);
+ if (conn) {
+ cxgb3i_tx_debug("cn 0x%p, cid %d.\n", c3cn, conn->id);
+ tcp_conn = conn->dd_data;
+ scsi_queue_work(conn->session->host, &conn->xmitwork);
+ }
+}
+
+void cxgb3i_conn_closing(struct s3_conn *c3cn)
+{
+ struct iscsi_conn *conn;
+ read_lock(&c3cn->callback_lock);
+ conn = (struct iscsi_conn *)c3cn->user_data;
+ if (conn && c3cn->state != C3CN_STATE_ESTABLISHED)
+ iscsi_conn_failure(conn, ISCSI_ERR_CONN_FAILED);
+ read_unlock(&c3cn->callback_lock);
+}
+
+int cxgb3i_adapter_ulp_init(struct cxgb3i_adapter *snic)
+{
+ struct t3cdev *tdev = snic->tdev;
+ struct cxgb3i_ddp_info *ddp = &snic->ddp;
+ struct ulp_iscsi_info uinfo;
+ unsigned int ppmax, bits, max_bits;
+ int i, err;
+
+ spin_lock_init(&ddp->map_lock);
+
+ err = tdev->ctl(tdev, ULP_ISCSI_GET_PARAMS, &uinfo);
+ if (err < 0) {
+ cxgb3i_log_error("%s, failed to get iscsi param err=%d.\n",
+ tdev->name, err);
+ return err;
+ }
+
+ ppmax = (uinfo.ulimit - uinfo.llimit + 1) >> PPOD_SIZE_SHIFT;
+ max_bits = min(PPOD_IDX_MAX_SIZE,
+ (32 - sw_tag_idx_bits - sw_tag_age_bits));
+ bits = __ilog2_u32(ppmax) + 1;
+ if (bits > max_bits)
+ bits = max_bits;
+ ppmax = (1 << bits) - 1;
+
+ snic->tx_max_size = uinfo.max_txsz;
+ snic->rx_max_size = uinfo.max_rxsz;
+ cxgb3i_log_debug("snic tx %u, rx %u.\n", snic->tx_max_size,
+ snic->rx_max_size);
+ snic->tag_format.idx_bits = sw_tag_idx_bits;
+ snic->tag_format.age_bits = sw_tag_age_bits;
+ snic->tag_format.rsvd_bits = bits;
+ snic->tag_format.rsvd_shift = PPOD_IDX_SHIFT;
+ snic->tag_format.rsvd_mask = (1 << snic->tag_format.rsvd_bits) - 1;
+
+ cxgb3i_log_debug("snic nppods %u, rsvd shift %u, bits %u, mask 0x%x.\n",
+ ppmax, snic->tag_format.rsvd_shift,
+ snic->tag_format.rsvd_bits,
+ snic->tag_format.rsvd_mask);
+
+ ddp->map = cxgb3i_alloc_big_mem(ppmax);
+ if (!ddp->map) {
+ cxgb3i_log_warn("snic unable to alloc ddp ppod 0x%u, "
+ "ddp disabled.\n", ppmax);
+ return 0;
+ }
+ ddp->llimit = uinfo.llimit;
+ ddp->ulimit = uinfo.ulimit;
+
+ uinfo.tagmask =
+ snic->tag_format.rsvd_mask << snic->tag_format.rsvd_shift;
+ for (i = 0; i < ULP2_PGIDX_MAX; i++)
+ uinfo.pgsz_factor[i] = ddp_page_order[i];
+
+ err = tdev->ctl(tdev, ULP_ISCSI_SET_PARAMS, &uinfo);
+ if (err < 0) {
+ cxgb3i_log_warn("snic unable to set iscsi param err=%d, "
+ "ddp disabled.\n", err);
+ goto free_ppod_map;
+ }
+
+ ddp->nppods = ppmax;
+ ddp->idx_last = ppmax;
+
+ tdev->ulp_iscsi = ddp;
+
+ return 0;
+
+free_ppod_map:
+ cxgb3i_free_big_mem(ddp->map);
+ return 0;
+}
+
+void cxgb3i_adapter_ulp_cleanup(struct cxgb3i_adapter *snic)
+{
+ u8 *map = snic->ddp.map;
+ if (map) {
+ snic->tdev->ulp_iscsi = NULL;
+ spin_lock(&snic->lock);
+ snic->ddp.map = NULL;
+ spin_unlock(&snic->lock);
+ cxgb3i_free_big_mem(map);
+ }
+}
diff --git a/drivers/scsi/cxgb3i/cxgb3i_ulp2.h b/drivers/scsi/cxgb3i/cxgb3i_ulp2.h
new file mode 100644
index 0000000..e3f46dc
--- /dev/null
+++ b/drivers/scsi/cxgb3i/cxgb3i_ulp2.h
@@ -0,0 +1,106 @@
+/*
+ * cxgb3i_ulp2.h: Chelsio S3xx iSCSI driver.
+ *
+ * Copyright (c) 2008 Chelsio Communications, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation.
+ *
+ * Written by: Karen Xie (kxie@...lsio.com)
+ */
+
+#ifndef __CXGB3I_ULP2_H__
+#define __CXGB3I_ULP2_H__
+
+#define ULP2_PDU_PAYLOAD_DFLT (16224 - ISCSI_PDU_HEADER_MAX)
+#define PPOD_PAGES_MAX 4
+#define PPOD_PAGES_SHIFT 2 /* 4 pages per pod */
+
+struct pagepod_hdr {
+ u32 vld_tid;
+ u32 pgsz_tag_clr;
+ u32 maxoffset;
+ u32 pgoffset;
+ u64 rsvd;
+};
+
+struct pagepod {
+ struct pagepod_hdr hdr;
+ u64 addr[PPOD_PAGES_MAX + 1];
+};
+
+#define PPOD_SIZE sizeof(struct pagepod) /* 64 */
+#define PPOD_SIZE_SHIFT 6
+
+#define PPOD_COLOR_SHIFT 0
+#define PPOD_COLOR_SIZE 6
+#define PPOD_COLOR_MASK ((1 << PPOD_COLOR_SIZE) - 1)
+
+#define PPOD_IDX_SHIFT PPOD_COLOR_SIZE
+#define PPOD_IDX_MAX_SIZE 24
+
+#define S_PPOD_TID 0
+#define M_PPOD_TID 0xFFFFFF
+#define V_PPOD_TID(x) ((x) << S_PPOD_TID)
+
+#define S_PPOD_VALID 24
+#define V_PPOD_VALID(x) ((x) << S_PPOD_VALID)
+#define F_PPOD_VALID V_PPOD_VALID(1U)
+
+#define S_PPOD_COLOR 0
+#define M_PPOD_COLOR 0x3F
+#define V_PPOD_COLOR(x) ((x) << S_PPOD_COLOR)
+
+#define S_PPOD_TAG 6
+#define M_PPOD_TAG 0xFFFFFF
+#define V_PPOD_TAG(x) ((x) << S_PPOD_TAG)
+
+#define S_PPOD_PGSZ 30
+#define M_PPOD_PGSZ 0x3
+#define V_PPOD_PGSZ(x) ((x) << S_PPOD_PGSZ)
+
+struct cpl_iscsi_hdr_norss {
+ union opcode_tid ot;
+ u16 pdu_len_ddp;
+ u16 len;
+ u32 seq;
+ u16 urg;
+ u8 rsvd;
+ u8 status;
+};
+
+struct cpl_rx_data_ddp_norss {
+ union opcode_tid ot;
+ u16 urg;
+ u16 len;
+ u32 seq;
+ u32 nxt_seq;
+ u32 ulp_crc;
+ u32 ddp_status;
+};
+
+#define RX_DDP_STATUS_IPP_SHIFT 27 /* invalid pagepod */
+#define RX_DDP_STATUS_TID_SHIFT 26 /* tid mismatch */
+#define RX_DDP_STATUS_COLOR_SHIFT 25 /* color mismatch */
+#define RX_DDP_STATUS_OFFSET_SHIFT 24 /* offset mismatch */
+#define RX_DDP_STATUS_ULIMIT_SHIFT 23 /* ulimit error */
+#define RX_DDP_STATUS_TAG_SHIFT 22 /* tag mismatch */
+#define RX_DDP_STATUS_DCRC_SHIFT 21 /* dcrc error */
+#define RX_DDP_STATUS_HCRC_SHIFT 20 /* hcrc error */
+#define RX_DDP_STATUS_PAD_SHIFT 19 /* pad error */
+#define RX_DDP_STATUS_PPP_SHIFT 18 /* pagepod parity error */
+#define RX_DDP_STATUS_LLIMIT_SHIFT 17 /* llimit error */
+#define RX_DDP_STATUS_DDP_SHIFT 16 /* ddp'able */
+#define RX_DDP_STATUS_PMM_SHIFT 15 /* pagepod mismatch */
+
+#define ULP2_FLAG_DATA_READY 0x1
+#define ULP2_FLAG_DATA_DDPED 0x2
+#define ULP2_FLAG_HCRC_ERROR 0x10
+#define ULP2_FLAG_DCRC_ERROR 0x20
+#define ULP2_FLAG_PAD_ERROR 0x40
+
+void cxgb3i_conn_closing(struct s3_conn *);
+void cxgb3i_conn_pdu_ready(struct s3_conn *c3cn);
+void cxgb3i_conn_tx_open(struct s3_conn *c3cn);
+#endif
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists