* [RFC-V2 PATCH 0/5] Proposal for iSCSI Network Configuration
@ 2011-04-02 18:34 vikas.chaudhary
2011-04-02 18:34 ` [RFC-V2 PATCH 1/5] iscsi_transport: add support for set_net_config vikas.chaudhary
` (4 more replies)
0 siblings, 5 replies; 24+ messages in thread
From: vikas.chaudhary @ 2011-04-02 18:34 UTC (permalink / raw)
To: James.Bottomley, michaelc
Cc: linux-scsi, open-iscsi, vikas.chaudhary, lalit.chandivade, ravi.anand
From: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
These are RFC-V2 follow-up patches to configuring network setting for
iSCSI adapter (qla4xxx) that use private IP address. Please review
proposal/implementation and let us know the feedback.
Our Initial RFC patches is here:
http://marc.info/?l=linux-scsi&m=129863193520309&w=2
http://groups.google.com/group/open-iscsi/browse_thread/thread/2924e8d871233436
New patches include following changes:
- Take care of comments from Mike Christie about using multiple IO Vector mechanism
to send netconfig parameters from application space to kernel.
- Added support to display IPV6 address and Linklocal address auto_config type in sysfs.
Kernel patches:
Lalit Chandivade (2):
iscsi_transport: add set_net_config
qla4xxx: add support for set_net_config
Vikas Chaudhary (3):
qla4xxx: Added new "struct ipaddress_config"
iscsi_transport: show network configuration in sysfs
qla4xxx: added support to show multiple iface in sysfs
Application side iscsiadm patch:
Lalit Chandivade (1):
iscsiadm: Add netconfig support in iscsiadm and iscsid
Thanks,
Vikas.
^ permalink raw reply [flat|nested] 24+ messages in thread
* [RFC-V2 PATCH 1/5] iscsi_transport: add support for set_net_config
2011-04-02 18:34 [RFC-V2 PATCH 0/5] Proposal for iSCSI Network Configuration vikas.chaudhary
@ 2011-04-02 18:34 ` vikas.chaudhary
2011-04-13 3:41 ` Mike Christie
2011-04-02 18:34 ` [RFC-V2 PATCH 2/5] qla4xxx: " vikas.chaudhary
` (3 subsequent siblings)
4 siblings, 1 reply; 24+ messages in thread
From: vikas.chaudhary @ 2011-04-02 18:34 UTC (permalink / raw)
To: James.Bottomley, michaelc
Cc: linux-scsi, open-iscsi, vikas.chaudhary, lalit.chandivade,
ravi.anand, Harish Zunjarrao
From: Lalit Chandivade <lalit.chandivade@qlogic.com>
Allows user space (iscsiadm) to send down network configuration parameters
for LLD to set private network configuration on the iSCSI adapters.
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: Harish Zunjarrao <harish.zunjarrao@qlogic.com>
Signed-off-by: Lalit Chandivade <lalit.chandivade@qlogic.com>
Reviewed-by: Ravi Anand <ravi.anand@qlogic.com>
---
drivers/scsi/scsi_transport_iscsi.c | 27 ++++++++++++++++++
include/scsi/iscsi_if.h | 51 +++++++++++++++++++++++++++++++++++
include/scsi/scsi_transport_iscsi.h | 1 +
3 files changed, 79 insertions(+), 0 deletions(-)
diff --git a/drivers/scsi/scsi_transport_iscsi.c b/drivers/scsi/scsi_transport_iscsi.c
index 3fd16d7..38fcfc0 100644
--- a/drivers/scsi/scsi_transport_iscsi.c
+++ b/drivers/scsi/scsi_transport_iscsi.c
@@ -1558,6 +1558,30 @@ iscsi_set_path(struct iscsi_transport *transport, struct iscsi_uevent *ev)
}
static int
+iscsi_set_net_config(struct iscsi_transport *transport,
+ struct iscsi_uevent *ev)
+{
+ char *data = (char *)ev + sizeof(*ev);
+ struct Scsi_Host *shost;
+ int err;
+
+ if (!transport->set_net_config)
+ return -ENOSYS;
+
+ shost = scsi_host_lookup(ev->u.set_net_config.host_no);
+ if (!shost) {
+ printk(KERN_ERR "set_net_config could not find host no %u\n",
+ ev->u.set_net_config.host_no);
+ return -ENODEV;
+ }
+
+ err = transport->set_net_config(shost, data,
+ ev->u.set_net_config.count);
+ scsi_host_put(shost);
+ return err;
+}
+
+static int
iscsi_if_recv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, uint32_t *group)
{
int err = 0;
@@ -1696,6 +1720,9 @@ iscsi_if_recv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, uint32_t *group)
case ISCSI_UEVENT_PATH_UPDATE:
err = iscsi_set_path(transport, ev);
break;
+ case ISCSI_UEVENT_SET_NET_CONFIG:
+ err = iscsi_set_net_config(transport, ev);
+ break;
default:
err = -ENOSYS;
break;
diff --git a/include/scsi/iscsi_if.h b/include/scsi/iscsi_if.h
index c3e1cbc..312b495 100644
--- a/include/scsi/iscsi_if.h
+++ b/include/scsi/iscsi_if.h
@@ -59,6 +59,7 @@ enum iscsi_uevent_e {
ISCSI_UEVENT_TRANSPORT_EP_CONNECT_THROUGH_HOST = UEVENT_BASE + 19,
ISCSI_UEVENT_PATH_UPDATE = UEVENT_BASE + 20,
+ ISCSI_UEVENT_SET_NET_CONFIG = UEVENT_BASE + 21,
/* up events */
ISCSI_KEVENT_RECV_PDU = KEVENT_BASE + 1,
@@ -172,6 +173,10 @@ struct iscsi_uevent {
struct msg_set_path {
uint32_t host_no;
} set_path;
+ struct msg_set_net_config {
+ uint32_t host_no;
+ uint32_t count;
+ } set_net_config;
} u;
union {
/* messages k -> u */
@@ -237,6 +242,52 @@ struct iscsi_path {
uint16_t pmtu;
} __attribute__ ((aligned (sizeof(uint64_t))));
+/* ipv6 addr autoconfig type */
+#define ISCSI_IPV6_AUTOCFG_DISABLE 0x01
+#define ISCSI_IPV6_AUTOCFG_ND_ENABLE 0x02
+#define ISCSI_IPV6_AUTOCFG_DHCPV6_ENABLE 0x03
+
+/* ipv6 link local addr type */
+#define ISCSI_IPV6_LINKLOCAL_AUTOCFG_ENABLE 0x01
+#define ISCSI_IPV6_LINKLOCAL_AUTOCFG_DISABLE 0x02
+
+/* ipv6 router addr type */
+#define ISCSI_IPV6_ROUTER_AUTOCFG_ENABLE 0x01
+#define ISCSI_IPV6_ROUTER_AUTOCFG_DISABLE 0x02
+
+/* iSCSI network params */
+enum iscsi_net_param_type {
+ ISCSI_NET_PARAM_IPV4_ADDR = 1,
+ ISCSI_NET_PARAM_IPV4_SUBNET = 2,
+ ISCSI_NET_PARAM_IPV4_GW = 3,
+ ISCSI_NET_PARAM_IPV4_BOOTPROTO = 4,
+ ISCSI_NET_PARAM_VLAN = 5,
+ ISCSI_NET_PARAM_MAC = 6,
+ ISCSI_NET_PARAM_IPV6_LINKLOCAL = 7,
+ ISCSI_NET_PARAM_IPV6_ADDR = 8,
+ ISCSI_NET_PARAM_IPV6_ROUTER = 9,
+ ISCSI_NET_PARAM_IPV6_ADDR_AUTOCFG = 10,
+ ISCSI_NET_PARAM_IPV6_LINKLOCAL_AUTOCFG = 11,
+ ISCSI_NET_PARAM_IPV6_ROUTER_AUTOCFG = 12,
+ ISCSI_NET_PARAM_IFACE_STATE = 13,
+};
+
+/* 20 param per iface * 10 iface per port = 200 params */
+#define ISCSI_MAX_IFACE_PER_HW 10
+#define ISCSI_MAX_PARAM_PER_IFACE 20
+#define ISCSI_MAX_NET_PARAMS (ISCSI_MAX_IFACE_PER_HW * \
+ ISCSI_MAX_PARAM_PER_IFACE)
+
+#define IFACE_TYPE_IPV4 0x01
+#define IFACE_TYPE_IPV6 0x02
+struct iscsi_net_param {
+ uint32_t param_type; /* enum iscsi_net_param */
+ uint32_t iface_type; /* IPv4 or IPv6 */
+ uint32_t iface_num; /* iface number, 0 - n */
+ uint32_t length; /* Actual length of the param */
+ uint8_t value[0]; /* legnth sized value follows */
+} __packed;
+
/*
* Common error codes
*/
diff --git a/include/scsi/scsi_transport_iscsi.h b/include/scsi/scsi_transport_iscsi.h
index bf8f529..45b9df9 100644
--- a/include/scsi/scsi_transport_iscsi.h
+++ b/include/scsi/scsi_transport_iscsi.h
@@ -137,6 +137,7 @@ struct iscsi_transport {
int (*tgt_dscvr) (struct Scsi_Host *shost, enum iscsi_tgt_dscvr type,
uint32_t enable, struct sockaddr *dst_addr);
int (*set_path) (struct Scsi_Host *shost, struct iscsi_path *params);
+ int (*set_net_config) (struct Scsi_Host *shost, char *data, int count);
};
/*
--
1.7.3.2
^ permalink raw reply related [flat|nested] 24+ messages in thread
* [RFC-V2 PATCH 2/5] qla4xxx: add support for set_net_config
2011-04-02 18:34 [RFC-V2 PATCH 0/5] Proposal for iSCSI Network Configuration vikas.chaudhary
2011-04-02 18:34 ` [RFC-V2 PATCH 1/5] iscsi_transport: add support for set_net_config vikas.chaudhary
@ 2011-04-02 18:34 ` vikas.chaudhary
2011-04-13 3:55 ` Mike Christie
2011-04-02 18:34 ` [RFC-V2 PATCH 3/5] qla4xxx: Added new "struct ipaddress_config" vikas.chaudhary
` (2 subsequent siblings)
4 siblings, 1 reply; 24+ messages in thread
From: vikas.chaudhary @ 2011-04-02 18:34 UTC (permalink / raw)
To: James.Bottomley, michaelc
Cc: linux-scsi, open-iscsi, vikas.chaudhary, lalit.chandivade,
ravi.anand, Harish Zunjarrao
From: Lalit Chandivade <lalit.chandivade@qlogic.com>
Allows user space (iscsiadm) to send down network configuration parameters
for LLD to set private network configuration on the iSCSI adapters.
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: Harish Zunjarrao <harish.zunjarrao@qlogic.com>
Signed-off-by: Lalit Chandivade <lalit.chandivade@qlogic.com>
Reviewed-by: Ravi Anand <ravi.anand@qlogic.com>
---
drivers/scsi/qla4xxx/ql4_fw.h | 102 ++++++++++++
drivers/scsi/qla4xxx/ql4_glbl.h | 18 ++-
drivers/scsi/qla4xxx/ql4_mbx.c | 126 ++++++++++++++-
drivers/scsi/qla4xxx/ql4_os.c | 336 +++++++++++++++++++++++++++++++++++++++
4 files changed, 578 insertions(+), 4 deletions(-)
diff --git a/drivers/scsi/qla4xxx/ql4_fw.h b/drivers/scsi/qla4xxx/ql4_fw.h
index 31e2bf9..812254d 100644
--- a/drivers/scsi/qla4xxx/ql4_fw.h
+++ b/drivers/scsi/qla4xxx/ql4_fw.h
@@ -434,6 +434,14 @@ struct qla_flt_region {
#define ACB_STATE_VALID 0x05
#define ACB_STATE_DISABLING 0x06
+/* FLASH offsets */
+#define FLASH_SEGMENT_IFCB 0x04000000
+
+#define FLASH_OPT_RMW_HOLD 0
+#define FLASH_OPT_RMW_INIT 1
+#define FLASH_OPT_COMMIT 2
+#define FLASH_OPT_RMW_COMMIT 3
+
/*************************************************************************/
/* Host Adapter Initialization Control Block (from host) */
@@ -473,6 +481,7 @@ struct addr_ctrl_blk {
uint16_t iscsi_opts; /* 30-31 */
uint16_t ipv4_tcp_opts; /* 32-33 */
+#define TCPOPT_DHCP_ENABLE 0x0200
uint16_t ipv4_ip_opts; /* 34-35 */
#define IPOPT_IPv4_PROTOCOL_ENABLE 0x8000
@@ -574,6 +583,99 @@ struct init_fw_ctrl_blk {
/* struct addr_ctrl_blk sec;*/
};
+struct addr_ctrl_blk_def {
+ uint8_t reserved1[1]; /* 00 */
+#define IFCB_VER_MIN 0x01
+#define IFCB_VER_MAX 0x02
+ uint8_t control; /* 01 */
+
+ uint8_t reserved2[11]; /* 02-0C */
+ uint8_t inst_num; /* 0D */
+ uint8_t reserved3[34]; /* 0E-2F */
+ uint16_t iscsi_opts; /* 30-31 */
+ uint16_t ipv4_tcp_opts; /* 32-33 */
+#define TCPOPT_DHCP_ENABLE 0x0200
+ uint16_t ipv4_ip_opts; /* 34-35 */
+#define IPOPT_IPv4_PROTOCOL_ENABLE 0x8000
+
+ uint16_t iscsi_max_pdu_size; /* 36-37 */
+ uint8_t ipv4_tos; /* 38 */
+ uint8_t ipv4_ttl; /* 39 */
+ uint8_t reserved4[2]; /* 3A-3B */
+ uint16_t def_timeout; /* 3C-3D */
+ uint16_t iscsi_fburst_len; /* 3E-3F */
+ uint8_t reserved5[4]; /* 40-43 */
+ uint16_t iscsi_max_outstnd_r2t; /* 44-45 */
+ uint8_t reserved6[2]; /* 46-47 */
+ uint16_t ipv4_port; /* 48-49 */
+ uint16_t iscsi_max_burst_len; /* 4A-4B */
+ uint8_t reserved7[4]; /* 4C-4F */
+ uint8_t ipv4_addr[4]; /* 50-53 */
+ uint16_t ipv4_vlan_tag; /* 54-55 */
+ uint8_t ipv4_addr_state; /* 56 */
+ uint8_t ipv4_cacheid; /* 57 */
+ uint8_t reserved8[8]; /* 58-5F */
+ uint8_t ipv4_subnet[4]; /* 60-63 */
+ uint8_t reserved9[12]; /* 64-6F */
+ uint8_t ipv4_gw_addr[4]; /* 70-73 */
+ uint8_t reserved10[84]; /* 74-C7 */
+ uint8_t abort_timer; /* C8 */
+ uint8_t ipv4_tcp_wsf; /* C9 */
+ uint8_t reserved11[10]; /* CA-D3 */
+ uint8_t ipv4_dhcp_vid_len; /* D4 */
+ uint8_t ipv4_dhcp_vid[11]; /* D5-DF */
+ uint8_t reserved12[20]; /* E0-F3 */
+ uint8_t ipv4_dhcp_alt_cid_len; /* F4 */
+ uint8_t ipv4_dhcp_alt_cid[11]; /* F5-FF */
+ uint8_t iscsi_name[224]; /* 100-1DF */
+ uint8_t reserved13[32]; /* 1E0-1FF */
+ uint32_t cookie; /* 200-203 */
+ uint16_t ipv6_port; /* 204-205 */
+ uint16_t ipv6_opts; /* 206-207 */
+#define IPV6_OPT_IPV6_PROTOCOL_ENABLE 0x8000
+
+ uint16_t ipv6_addtl_opts; /* 208-209 */
+#define IPV6_ADDOPT_NEIGHBOR_DISCOVERY_ADDR_ENABLE 0x0002 /* Pri ACB
+ Only */
+#define IPV6_ADDOPT_AUTOCONFIG_LINK_LOCAL_ADDR 0x0001
+
+ uint16_t ipv6_tcp_opts; /* 20A-20B */
+ uint8_t ipv6_tcp_wsf; /* 20C */
+ uint16_t ipv6_flow_lbl; /* 20D-20F */
+ uint8_t ipv6_dflt_rtr_addr[16]; /* 210-21F */
+ uint16_t ipv6_vlan_tag; /* 220-221 */
+ uint8_t ipv6_lnk_lcl_addr_state; /* 222 */
+ uint8_t ipv6_addr0_state; /* 223 */
+ uint8_t ipv6_addr1_state; /* 224 */
+#define IP_ADDRSTATE_UNCONFIGURED 0
+#define IP_ADDRSTATE_INVALID 1
+#define IP_ADDRSTATE_ACQUIRING 2
+#define IP_ADDRSTATE_TENTATIVE 3
+#define IP_ADDRSTATE_DEPRICATED 4
+#define IP_ADDRSTATE_PREFERRED 5
+#define IP_ADDRSTATE_DISABLING 6
+
+ uint8_t ipv6_dflt_rtr_state; /* 225 */
+#define IPV6_RTRSTATE_UNKNOWN 0
+#define IPV6_RTRSTATE_MANUAL 1
+#define IPV6_RTRSTATE_ADVERTISED 3
+#define IPV6_RTRSTATE_STALE 4
+
+ uint8_t ipv6_traffic_class; /* 226 */
+ uint8_t ipv6_hop_limit; /* 227 */
+ uint8_t ipv6_if_id[8]; /* 228-22F */
+ uint8_t ipv6_addr0[16]; /* 230-23F */
+ uint8_t ipv6_addr1[16]; /* 240-24F */
+ uint32_t ipv6_nd_reach_time; /* 250-253 */
+ uint32_t ipv6_nd_rexmit_timer; /* 254-257 */
+ uint32_t ipv6_nd_stale_timeout; /* 258-25B */
+ uint8_t ipv6_dup_addr_detect_count; /* 25C */
+ uint8_t ipv6_cache_id; /* 25D */
+ uint8_t reserved14[18]; /* 25E-26F */
+ uint32_t ipv6_gw_advrt_mtu; /* 270-273 */
+ uint8_t reserved15[140]; /* 274-2FF */
+};
+
/*************************************************************************/
struct dev_db_entry {
diff --git a/drivers/scsi/qla4xxx/ql4_glbl.h b/drivers/scsi/qla4xxx/ql4_glbl.h
index cc53e3f..1cadb74 100644
--- a/drivers/scsi/qla4xxx/ql4_glbl.h
+++ b/drivers/scsi/qla4xxx/ql4_glbl.h
@@ -52,7 +52,15 @@ int qla4xxx_get_fwddb_entry(struct scsi_qla_host *ha,
int qla4xxx_set_ddb_entry(struct scsi_qla_host * ha, uint16_t fw_ddb_index,
dma_addr_t fw_ddb_entry_dma);
-
+uint8_t qla4xxx_get_ifcb(struct scsi_qla_host *, uint32_t *, uint32_t *,
+ dma_addr_t);
+int qla4xxx_conn_close_sess_logout(struct scsi_qla_host *,
+ uint16_t, uint16_t, uint16_t);
+int qla4xxx_disable_acb(struct scsi_qla_host *);
+int qla4xxx_set_acb(struct scsi_qla_host *, uint32_t *,
+ uint32_t *, dma_addr_t);
+int qla4xxx_get_acb(struct scsi_qla_host *, uint32_t *,
+ uint32_t *, dma_addr_t);
void qla4xxx_mark_device_missing(struct scsi_qla_host *ha,
struct ddb_entry *ddb_entry);
u16 rd_nvram_word(struct scsi_qla_host *ha, int offset);
@@ -75,7 +83,8 @@ void qla4xxx_dump_buffer(void *b, uint32_t size);
int qla4xxx_send_marker_iocb(struct scsi_qla_host *ha,
struct ddb_entry *ddb_entry, int lun, uint16_t mrkr_mod);
int qla4_is_relogin_allowed(struct scsi_qla_host *ha, uint32_t conn_err);
-
+int qla4xxx_set_flash(struct scsi_qla_host *, dma_addr_t, uint32_t, uint32_t,
+ uint32_t);
int qla4xxx_mailbox_command(struct scsi_qla_host *ha, uint8_t inCount,
uint8_t outCount, uint32_t *mbx_cmd, uint32_t *mbx_sts);
@@ -95,6 +104,11 @@ void qla4xxx_wake_dpc(struct scsi_qla_host *ha);
void qla4xxx_get_conn_event_log(struct scsi_qla_host *ha);
void qla4xxx_mailbox_premature_completion(struct scsi_qla_host *ha);
void qla4xxx_dump_registers(struct scsi_qla_host *ha);
+uint8_t qla4xxx_update_local_ifcb(struct scsi_qla_host *ha,
+ uint32_t *mbox_cmd,
+ uint32_t *mbox_sts,
+ struct addr_ctrl_blk *init_fw_cb,
+ dma_addr_t init_fw_cb_dma);
void qla4_8xxx_pci_config(struct scsi_qla_host *);
int qla4_8xxx_iospace_config(struct scsi_qla_host *ha);
diff --git a/drivers/scsi/qla4xxx/ql4_mbx.c b/drivers/scsi/qla4xxx/ql4_mbx.c
index f9d81c8..ddcbd86 100644
--- a/drivers/scsi/qla4xxx/ql4_mbx.c
+++ b/drivers/scsi/qla4xxx/ql4_mbx.c
@@ -317,7 +317,7 @@ qla4xxx_set_ifcb(struct scsi_qla_host *ha, uint32_t *mbox_cmd,
return QLA_SUCCESS;
}
-static uint8_t
+uint8_t
qla4xxx_get_ifcb(struct scsi_qla_host *ha, uint32_t *mbox_cmd,
uint32_t *mbox_sts, dma_addr_t init_fw_cb_dma)
{
@@ -377,7 +377,7 @@ qla4xxx_update_local_ip(struct scsi_qla_host *ha,
}
}
-static uint8_t
+uint8_t
qla4xxx_update_local_ifcb(struct scsi_qla_host *ha,
uint32_t *mbox_cmd,
uint32_t *mbox_sts,
@@ -1194,3 +1194,125 @@ exit_send_tgts_no_free:
return ret_val;
}
+int qla4xxx_set_flash(struct scsi_qla_host *ha, dma_addr_t dma_addr,
+ uint32_t offset, uint32_t length, uint32_t options)
+{
+ uint32_t mbox_cmd[MBOX_REG_COUNT];
+ uint32_t mbox_sts[MBOX_REG_COUNT];
+
+ memset(&mbox_cmd, 0, sizeof(mbox_cmd));
+ memset(&mbox_sts, 0, sizeof(mbox_sts));
+
+ mbox_cmd[0] = MBOX_CMD_WRITE_FLASH;
+ mbox_cmd[1] = LSDW(dma_addr);
+ mbox_cmd[2] = MSDW(dma_addr);
+ mbox_cmd[3] = offset;
+ mbox_cmd[4] = length;
+ mbox_cmd[5] = options;
+
+ if (qla4xxx_mailbox_command(ha, 6, 2, &mbox_cmd[0], &mbox_sts[0])) {
+ printk(KERN_WARNING "scsi(%ld): %s: MBOX_CMD_WRITE_FLASH"
+ " failed w/ status %04X, mbx1 %04X\n", ha->host_no,
+ __func__, mbox_sts[0], mbox_sts[1]);
+ return QLA_ERROR;
+ }
+ return QLA_SUCCESS;
+}
+
+int qla4xxx_conn_close_sess_logout(struct scsi_qla_host *ha,
+ uint16_t fw_ddb_index,
+ uint16_t connection_id,
+ uint16_t option)
+{
+ uint32_t mbox_cmd[MBOX_REG_COUNT];
+ uint32_t mbox_sts[MBOX_REG_COUNT];
+
+ memset(&mbox_cmd, 0, sizeof(mbox_cmd));
+ memset(&mbox_sts, 0, sizeof(mbox_sts));
+
+ printk(KERN_INFO
+ "%s: fw_ddb_index: %d\n", __func__, fw_ddb_index);
+
+ mbox_cmd[0] = MBOX_CMD_CONN_CLOSE_SESS_LOGOUT;
+ mbox_cmd[1] = fw_ddb_index;
+ mbox_cmd[2] = connection_id;
+ mbox_cmd[3] = option;
+
+ if (qla4xxx_mailbox_command(ha, 4, 2, &mbox_cmd[0],
+ &mbox_sts[0]) != QLA_SUCCESS) {
+ DEBUG2(ql4_printk(KERN_WARNING, ha, "%s: MBOX_CMD_CONN_CLOSE "
+ "option %04x failed sts %04X %04X", __func__, option,
+ mbox_sts[0], mbox_sts[1]));
+ if (mbox_sts[0] == 0x4005) {
+ DEBUG2(printk(KERN_WARNING "%s reason %04X\n",
+ __func__, mbox_sts[1]));
+ } else {
+ DEBUG2(printk(KERN_WARNING "\n"));
+ }
+ return QLA_ERROR;
+ }
+ printk(KERN_INFO "%s: Exit\n", __func__);
+ return QLA_SUCCESS;
+}
+
+int qla4xxx_disable_acb(struct scsi_qla_host *ha)
+{
+ uint32_t mbox_cmd[MBOX_REG_COUNT];
+ uint32_t mbox_sts[MBOX_REG_COUNT];
+
+ memset(&mbox_cmd, 0, sizeof(mbox_cmd));
+ memset(&mbox_sts, 0, sizeof(mbox_sts));
+
+ mbox_cmd[0] = MBOX_CMD_DISABLE_ACB;
+
+ if (qla4xxx_mailbox_command(ha, MBOX_REG_COUNT, 5, &mbox_cmd[0],
+ &mbox_sts[0]) != QLA_SUCCESS) {
+ DEBUG2(ql4_printk(KERN_WARNING, ha, "%s: MBOX_CMD_DISABLE_ACB "
+ "failed sts %04X %04X %04X", __func__,
+ mbox_sts[0], mbox_sts[1], mbox_sts[2]));
+ return QLA_ERROR;
+ }
+ return QLA_SUCCESS;
+}
+
+int qla4xxx_get_acb(struct scsi_qla_host *ha, uint32_t *mbox_cmd,
+ uint32_t *mbox_sts, dma_addr_t acb_dma)
+{
+ memset(mbox_cmd, 0, sizeof(mbox_cmd[0]) * MBOX_REG_COUNT);
+ memset(mbox_sts, 0, sizeof(mbox_sts[0]) * MBOX_REG_COUNT);
+ mbox_cmd[0] = MBOX_CMD_GET_ACB;
+ mbox_cmd[1] = 0; /* Primary ACB */
+ mbox_cmd[2] = LSDW(acb_dma);
+ mbox_cmd[3] = MSDW(acb_dma);
+ mbox_cmd[4] = sizeof(struct addr_ctrl_blk);
+
+ if (qla4xxx_mailbox_command(ha, 5, 5, mbox_cmd, mbox_sts) !=
+ QLA_SUCCESS) {
+ DEBUG2(printk(KERN_WARNING "scsi%ld: %s: "
+ "MBOX_CMD_GET_ACB failed w/ status %04X\n",
+ ha->host_no, __func__, mbox_sts[0]));
+ return QLA_ERROR;
+ }
+ return QLA_SUCCESS;
+}
+
+int qla4xxx_set_acb(struct scsi_qla_host *ha, uint32_t *mbox_cmd,
+ uint32_t *mbox_sts, dma_addr_t acb_dma)
+{
+ memset(mbox_cmd, 0, sizeof(mbox_cmd[0]) * MBOX_REG_COUNT);
+ memset(mbox_sts, 0, sizeof(mbox_sts[0]) * MBOX_REG_COUNT);
+ mbox_cmd[0] = MBOX_CMD_SET_ACB;
+ mbox_cmd[1] = 0; /* Primary ACB */
+ mbox_cmd[2] = LSDW(acb_dma);
+ mbox_cmd[3] = MSDW(acb_dma);
+ mbox_cmd[4] = sizeof(struct addr_ctrl_blk);
+
+ if (qla4xxx_mailbox_command(ha, 5, 5, mbox_cmd, mbox_sts) !=
+ QLA_SUCCESS) {
+ DEBUG2(printk(KERN_WARNING "scsi%ld: %s: "
+ "MBOX_CMD_SET_ACB failed w/ status %04X\n",
+ ha->host_no, __func__, mbox_sts[0]));
+ return QLA_ERROR;
+ }
+ return QLA_SUCCESS;
+}
diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c
index a4acb0d..7b2e237 100644
--- a/drivers/scsi/qla4xxx/ql4_os.c
+++ b/drivers/scsi/qla4xxx/ql4_os.c
@@ -80,6 +80,8 @@ static int qla4xxx_sess_get_param(struct iscsi_cls_session *sess,
enum iscsi_param param, char *buf);
static int qla4xxx_host_get_param(struct Scsi_Host *shost,
enum iscsi_host_param param, char *buf);
+static int qla4xxx_set_net_config(struct Scsi_Host *shost, char *data,
+ int count);
static void qla4xxx_recovery_timedout(struct iscsi_cls_session *session);
static enum blk_eh_timer_return qla4xxx_eh_cmd_timed_out(struct scsi_cmnd *sc);
@@ -141,6 +143,7 @@ static struct iscsi_transport qla4xxx_iscsi_transport = {
.get_conn_param = qla4xxx_conn_get_param,
.get_session_param = qla4xxx_sess_get_param,
.get_host_param = qla4xxx_host_get_param,
+ .set_net_config = qla4xxx_set_net_config,
.session_recovery_timedout = qla4xxx_recovery_timedout,
};
@@ -201,6 +204,339 @@ static int qla4xxx_host_get_param(struct Scsi_Host *shost,
return len;
}
+/* Only for primary ACB, secondary ACB not supported
+ * iface_num = 0: valid -> linklocal, routable, router, autocfg options
+ * iface_num = 1: valid -> routable
+ */
+static inline void qla4xxx_set_ipv6(struct iscsi_net_param *net_param,
+ struct addr_ctrl_blk *init_fw_cb)
+{
+ switch (net_param->param_type) {
+ case ISCSI_NET_PARAM_IPV6_ADDR:
+ if (net_param->iface_num & 0x1) {
+ /* Routable 1 */
+ memcpy(init_fw_cb->ipv6_addr1, net_param->value,
+ sizeof(init_fw_cb->ipv6_addr1));
+ } else {
+ /* Routable 0 */
+ memcpy(init_fw_cb->ipv6_addr0, net_param->value,
+ sizeof(init_fw_cb->ipv6_addr0));
+ }
+ break;
+ case ISCSI_NET_PARAM_IPV6_LINKLOCAL:
+ if (net_param->iface_num & 0x1)
+ break;
+ memcpy(init_fw_cb->ipv6_if_id, &net_param->value[8],
+ sizeof(init_fw_cb->ipv6_if_id));
+
+ break;
+ case ISCSI_NET_PARAM_IPV6_ROUTER:
+ if (net_param->iface_num & 0x1)
+ break;
+ memcpy(init_fw_cb->ipv6_dflt_rtr_addr, net_param->value,
+ sizeof(init_fw_cb->ipv6_dflt_rtr_addr));
+ break;
+ case ISCSI_NET_PARAM_IPV6_ADDR_AUTOCFG:
+ /* Autocfg applies to even interface */
+ if (net_param->iface_num & 0x1)
+ break;
+
+ if (net_param->value[0] & ISCSI_IPV6_AUTOCFG_DISABLE) {
+ DEBUG2(printk(KERN_INFO
+ "%s: User Config IPv6 addr\n", __func__));
+ init_fw_cb->ipv6_addtl_opts &=
+ ~IPV6_ADDOPT_NEIGHBOR_DISCOVERY_ADDR_ENABLE;
+ } else if (net_param->value[0] & ISCSI_IPV6_AUTOCFG_ND_ENABLE) {
+ DEBUG2(printk(KERN_INFO
+ "%s: AUTO Config IPv6 addr\n", __func__));
+ init_fw_cb->ipv6_addtl_opts |=
+ IPV6_ADDOPT_NEIGHBOR_DISCOVERY_ADDR_ENABLE;
+ } else {
+ DEBUG2(printk(KERN_INFO
+ "%s: Invalid setting for IPv6 addr\n", __func__));
+ }
+ break;
+ case ISCSI_NET_PARAM_IPV6_LINKLOCAL_AUTOCFG:
+ /* Autocfg applies to even interface */
+ if (net_param->iface_num & 0x1)
+ break;
+
+ if (net_param->value[0] & ISCSI_IPV6_LINKLOCAL_AUTOCFG_ENABLE) {
+ DEBUG2(printk(KERN_INFO
+ "%s: AUTO Config link local addr\n", __func__));
+ init_fw_cb->ipv6_addtl_opts |=
+ IPV6_ADDOPT_AUTOCONFIG_LINK_LOCAL_ADDR;
+ } else if (net_param->value[0] &
+ ISCSI_IPV6_LINKLOCAL_AUTOCFG_DISABLE) {
+ DEBUG2(printk(KERN_INFO
+ "%s: User Config link local addr\n", __func__));
+ init_fw_cb->ipv6_addtl_opts &=
+ ~IPV6_ADDOPT_AUTOCONFIG_LINK_LOCAL_ADDR;
+ } else {
+ DEBUG2(printk(KERN_INFO
+ "%s: Invalid setting for IPv6 addr\n", __func__));
+ }
+ break;
+ case ISCSI_NET_PARAM_IPV6_ROUTER_AUTOCFG:
+ /* Autocfg applies to even interface */
+ if (net_param->iface_num & 0x1)
+ break;
+
+ if (net_param->value[0] & ISCSI_IPV6_ROUTER_AUTOCFG_ENABLE) {
+ DEBUG2(printk(KERN_INFO
+ "%s: AUTO Config router addr\n", __func__));
+ memset(init_fw_cb->ipv6_dflt_rtr_addr, 0,
+ sizeof(init_fw_cb->ipv6_dflt_rtr_addr));
+ }
+ break;
+ case ISCSI_NET_PARAM_VLAN:
+ if (net_param->length != sizeof(init_fw_cb->ipv6_vlan_tag))
+ break;
+ init_fw_cb->ipv6_vlan_tag = *(uint16_t *)net_param->value;
+ break;
+ case ISCSI_NET_PARAM_IFACE_STATE:
+ if (net_param->value[0])
+ init_fw_cb->ipv6_opts |= IPV6_OPT_IPV6_PROTOCOL_ENABLE;
+ else
+ init_fw_cb->ipv6_opts &= ~IPV6_OPT_IPV6_PROTOCOL_ENABLE;
+ break;
+ default:
+ printk(KERN_INFO
+ "Unknown param = %d\n", net_param->param_type);
+ break;
+ }
+}
+
+static inline void qla4xxx_set_ipv4(struct iscsi_net_param *net_param,
+ struct addr_ctrl_blk *init_fw_cb)
+{
+ switch (net_param->param_type) {
+ case ISCSI_NET_PARAM_IPV4_ADDR:
+ memcpy(init_fw_cb->ipv4_addr, net_param->value,
+ sizeof(init_fw_cb->ipv4_addr));
+ break;
+ case ISCSI_NET_PARAM_IPV4_SUBNET:
+ memcpy(init_fw_cb->ipv4_subnet, net_param->value,
+ sizeof(init_fw_cb->ipv4_subnet));
+ break;
+ case ISCSI_NET_PARAM_IPV4_GW:
+ memcpy(init_fw_cb->ipv4_gw_addr, net_param->value,
+ sizeof(init_fw_cb->ipv4_gw_addr));
+ break;
+ case ISCSI_NET_PARAM_IPV4_BOOTPROTO:
+ if (net_param->value[0])
+ init_fw_cb->ipv4_tcp_opts |= TCPOPT_DHCP_ENABLE;
+ else
+ init_fw_cb->ipv4_tcp_opts &= ~TCPOPT_DHCP_ENABLE;
+ break;
+ case ISCSI_NET_PARAM_VLAN:
+ if (net_param->length != sizeof(init_fw_cb->ipv4_vlan_tag))
+ break;
+ init_fw_cb->ipv4_vlan_tag = *(uint16_t *)net_param->value;
+ /* TODO: Set the bit indicating VLAN enabled */
+ break;
+ case ISCSI_NET_PARAM_IFACE_STATE:
+ if (net_param->value[0])
+ init_fw_cb->ipv4_ip_opts |= IPOPT_IPv4_PROTOCOL_ENABLE;
+ else
+ init_fw_cb->ipv4_ip_opts &= ~IPOPT_IPv4_PROTOCOL_ENABLE;
+ break;
+ default:
+ printk(KERN_ERR
+ "Unknown param = %d\n", net_param->param_type);
+ break;
+ }
+}
+
+static void
+qla4xxx_initcb_to_acb(struct addr_ctrl_blk *init_fw_cb)
+{
+ struct addr_ctrl_blk_def *acb;
+ acb = (struct addr_ctrl_blk_def *)init_fw_cb;
+ memset(acb->reserved1, 0, sizeof(acb->reserved1));
+ memset(acb->reserved2, 0, sizeof(acb->reserved2));
+ memset(acb->reserved3, 0, sizeof(acb->reserved3));
+ memset(acb->reserved4, 0, sizeof(acb->reserved4));
+ memset(acb->reserved5, 0, sizeof(acb->reserved5));
+ memset(acb->reserved6, 0, sizeof(acb->reserved6));
+ memset(acb->reserved7, 0, sizeof(acb->reserved7));
+ memset(acb->reserved8, 0, sizeof(acb->reserved8));
+ memset(acb->reserved9, 0, sizeof(acb->reserved9));
+ memset(acb->reserved10, 0, sizeof(acb->reserved10));
+ memset(acb->reserved11, 0, sizeof(acb->reserved11));
+ memset(acb->reserved12, 0, sizeof(acb->reserved12));
+ memset(acb->reserved13, 0, sizeof(acb->reserved13));
+ memset(acb->reserved14, 0, sizeof(acb->reserved14));
+ memset(acb->reserved15, 0, sizeof(acb->reserved15));
+}
+
+static int
+qla4xxx_set_net_config(struct Scsi_Host *shost, char *data, int count)
+{
+ struct scsi_qla_host *ha = to_qla_host(shost);
+ int rval = 0;
+ struct iscsi_net_param *net_param = NULL;
+ struct addr_ctrl_blk *init_fw_cb = NULL;
+ dma_addr_t init_fw_cb_dma;
+
+ struct addr_ctrl_blk *acb = NULL;
+ dma_addr_t acb_dma;
+ uint32_t ddb_state;
+ struct ddb_entry *ddb_entry, *ddbtemp;
+ unsigned long wait_time;
+
+ uint32_t mbox_cmd[MBOX_REG_COUNT];
+ uint32_t mbox_sts[MBOX_REG_COUNT];
+ uint32_t total_param_count;
+ uint32_t length;
+
+ init_fw_cb = dma_alloc_coherent(&ha->pdev->dev,
+ sizeof(struct addr_ctrl_blk), &init_fw_cb_dma, GFP_KERNEL);
+ if (!init_fw_cb) {
+ printk(KERN_ERR
+ "scsi(%ld): %s: Unable to alloc init_cb\n",
+ ha->host_no, __func__);
+ return -ENOMEM;
+ }
+
+ acb = dma_alloc_coherent(&ha->pdev->dev, sizeof(struct addr_ctrl_blk),
+ &acb_dma, GFP_KERNEL);
+ if (!acb) {
+ printk(KERN_ERR
+ "scsi(%ld): %s: Unable to alloc acb\n",
+ ha->host_no, __func__);
+ rval = -ENOMEM;
+ goto exit_init_fw_cb;
+ }
+ memset(init_fw_cb, 0, sizeof(struct addr_ctrl_blk));
+ memset(acb, 0, sizeof(struct addr_ctrl_blk));
+ memset(&mbox_cmd, 0, sizeof(mbox_cmd));
+ memset(&mbox_sts, 0, sizeof(mbox_sts));
+
+ if (qla4xxx_get_ifcb(ha, &mbox_cmd[0], &mbox_sts[0], init_fw_cb_dma)) {
+ printk(KERN_ERR
+ "scsi(%ld): %s: get ifcb failed\n",
+ ha->host_no, __func__);
+ rval = -EIO;
+ goto exit_init_fw_cb;
+ }
+
+ total_param_count = count;
+ net_param = (struct iscsi_net_param *)data;
+
+ for ( ; total_param_count != 0; total_param_count--) {
+ length = net_param->length;
+
+ switch (net_param->iface_type) {
+ case IFACE_TYPE_IPV4:
+ switch (net_param->iface_num) {
+ case 0:
+ qla4xxx_set_ipv4(net_param, init_fw_cb);
+ break;
+ default:
+ /* Cannot have more than one IPv4 interface */
+ printk(KERN_ERR
+ "IFACE num-%d incorrect for IPv4 setting\n",
+ net_param->iface_num);
+ break;
+ }
+ break;
+ case IFACE_TYPE_IPV6:
+ switch (net_param->iface_num) {
+ case 0:
+ case 1:
+ qla4xxx_set_ipv6(net_param, init_fw_cb);
+ break;
+ default:
+ /* Cannot have more than two IPv6 interface */
+ printk(KERN_ERR
+ "IFACE num-%d incorrect for IPv6 setting\n",
+ net_param->iface_num);
+ break;
+ }
+ break;
+ default:
+ printk(KERN_ERR "Only IPv4 & IPv6 Supported\n");
+ break;
+ }
+
+ net_param = (struct iscsi_net_param *)((uint8_t *)net_param +
+ sizeof(struct iscsi_net_param) + length);
+ }
+
+ init_fw_cb->cookie = 0x11BEAD5A;
+
+ if (qla4xxx_set_flash(ha, init_fw_cb_dma, FLASH_SEGMENT_IFCB,
+ sizeof(struct addr_ctrl_blk), FLASH_OPT_RMW_COMMIT)) {
+ printk(KERN_ERR "scsi(%ld): %s: set flash failed\n",
+ ha->host_no, __func__);
+ rval = -EIO;
+ goto exit_init_fw_cb;
+ }
+
+ /* TODO: See if an option can be added to allow user to do session */
+ /* logout Close all TCP connection */
+ list_for_each_entry_safe(ddb_entry, ddbtemp, &ha->ddb_list, list) {
+ if (ddb_entry->fw_ddb_device_state == DDB_DS_SESSION_ACTIVE)
+ rval = qla4xxx_conn_close_sess_logout(ha,
+ ddb_entry->fw_ddb_index, ddb_entry->connection_id,
+ LOGOUT_OPTION_CLOSE_SESSION);
+ if (rval != QLA_SUCCESS)
+ printk(KERN_ERR
+ "qla4xxx_conn_close_sess_logout failed "
+ "for ddb index = 0x%x\n", ddb_entry->fw_ddb_index);
+ }
+
+ /* Wait for AEN 8014 to come ??? */
+ list_for_each_entry_safe(ddb_entry, ddbtemp, &ha->ddb_list, list) {
+ wait_time = jiffies + (LOGOUT_TOV * HZ);
+ while (!time_after_eq(jiffies, wait_time)) {
+ rval = qla4xxx_get_fwddb_entry(ha,
+ ddb_entry->fw_ddb_index, NULL, 0, NULL, NULL,
+ &ddb_state, NULL, NULL, NULL);
+ if (rval != QLA_SUCCESS)
+ break;
+
+ if ((ddb_state == DDB_DS_UNASSIGNED) ||
+ (ddb_state == DDB_DS_NO_CONNECTION_ACTIVE)) {
+ printk(KERN_INFO
+ "%s: ddb[%d]: conn close\n", __func__,
+ ddb_entry->fw_ddb_index);
+ break;
+ }
+ ssleep(1);
+ }
+ }
+
+ qla4xxx_disable_acb(ha);
+
+ /* Get ACB to check the state of IP address */
+ if (qla4xxx_get_acb(ha, &mbox_cmd[0], &mbox_sts[0], acb_dma)) {
+ printk(KERN_ERR
+ "scsi(%ld): %s: get acb failed\n", ha->host_no, __func__);
+ rval = -EIO;
+ goto exit_acb;
+ }
+
+ qla4xxx_initcb_to_acb(init_fw_cb);
+ qla4xxx_set_acb(ha, &mbox_cmd[0], &mbox_sts[0], init_fw_cb_dma);
+
+ memset(init_fw_cb, 0, sizeof(struct addr_ctrl_blk));
+ qla4xxx_update_local_ifcb(ha, &mbox_cmd[0], &mbox_sts[0], init_fw_cb,
+ init_fw_cb_dma);
+
+ printk(KERN_ERR "%s: cmd=%d, Returning success\n", __func__,
+ rval);
+exit_acb:
+ dma_free_coherent(&ha->pdev->dev, sizeof(struct addr_ctrl_blk),
+ acb, acb_dma);
+exit_init_fw_cb:
+ dma_free_coherent(&ha->pdev->dev, sizeof(struct addr_ctrl_blk),
+ init_fw_cb, init_fw_cb_dma);
+
+ return rval;
+}
+
static int qla4xxx_sess_get_param(struct iscsi_cls_session *sess,
enum iscsi_param param, char *buf)
{
--
1.7.3.2
^ permalink raw reply related [flat|nested] 24+ messages in thread
* [RFC-V2 PATCH 3/5] qla4xxx: Added new "struct ipaddress_config"
2011-04-02 18:34 [RFC-V2 PATCH 0/5] Proposal for iSCSI Network Configuration vikas.chaudhary
2011-04-02 18:34 ` [RFC-V2 PATCH 1/5] iscsi_transport: add support for set_net_config vikas.chaudhary
2011-04-02 18:34 ` [RFC-V2 PATCH 2/5] qla4xxx: " vikas.chaudhary
@ 2011-04-02 18:34 ` vikas.chaudhary
2011-04-02 18:34 ` [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs vikas.chaudhary
2011-04-02 18:34 ` [RFC-V2 PATCH 5/5] qla4xxx: added support to show multiple iface " vikas.chaudhary
4 siblings, 0 replies; 24+ messages in thread
From: vikas.chaudhary @ 2011-04-02 18:34 UTC (permalink / raw)
To: James.Bottomley, michaelc
Cc: linux-scsi, open-iscsi, vikas.chaudhary, lalit.chandivade, ravi.anand
From: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
- Move all ipaddress related param to "struct ipaddress_config"
from "struct scsi_qla_host"
- update function - qla4xxx_update_local_ip()
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
---
drivers/scsi/qla4xxx/ql4_def.h | 47 ++++++++++++++++-------------
drivers/scsi/qla4xxx/ql4_init.c | 51 ++++++++++++++++++-------------
drivers/scsi/qla4xxx/ql4_mbx.c | 64 ++++++++++++++++++++++-----------------
drivers/scsi/qla4xxx/ql4_os.c | 4 +--
4 files changed, 93 insertions(+), 73 deletions(-)
diff --git a/drivers/scsi/qla4xxx/ql4_def.h b/drivers/scsi/qla4xxx/ql4_def.h
index c1f8d1b..74faf54 100644
--- a/drivers/scsi/qla4xxx/ql4_def.h
+++ b/drivers/scsi/qla4xxx/ql4_def.h
@@ -357,6 +357,28 @@ struct isp_operations {
int (*get_sys_info) (struct scsi_qla_host *);
};
+/*qla4xxx ipaddress configuration details */
+struct ipaddress_config {
+ uint16_t ipv4_options;
+ uint16_t tcp_options;
+ uint16_t ipv4_vlan_tag;
+ uint8_t ipv4_addr_state;
+ uint8_t ip_address[IP_ADDR_LEN];
+ uint8_t subnet_mask[IP_ADDR_LEN];
+ uint8_t gateway[IP_ADDR_LEN];
+ uint32_t ipv6_options;
+ uint32_t ipv6_addl_options;
+ uint8_t ipv6_link_local_state;
+ uint8_t ipv6_addr0_state;
+ uint8_t ipv6_addr1_state;
+ uint8_t ipv6_default_router_state;
+ uint16_t ipv6_vlan_tag;
+ struct in6_addr ipv6_link_local_addr;
+ struct in6_addr ipv6_addr0;
+ struct in6_addr ipv6_addr1;
+ struct in6_addr ipv6_default_router_addr;
+};
+
/*
* Linux Host Adapter structure
*/
@@ -452,10 +474,6 @@ struct scsi_qla_host {
/* --- From Init_FW --- */
/* init_cb_t *init_cb; */
uint16_t firmware_options;
- uint16_t tcp_options;
- uint8_t ip_address[IP_ADDR_LEN];
- uint8_t subnet_mask[IP_ADDR_LEN];
- uint8_t gateway[IP_ADDR_LEN];
uint8_t alias[32];
uint8_t name_string[256];
uint8_t heartbeat_interval;
@@ -534,22 +552,7 @@ struct scsi_qla_host {
/* Saved srb for status continuation entry processing */
struct srb *status_srb;
- /* IPv6 support info from InitFW */
uint8_t acb_version;
- uint8_t ipv4_addr_state;
- uint16_t ipv4_options;
-
- uint32_t resvd2;
- uint32_t ipv6_options;
- uint32_t ipv6_addl_options;
- uint8_t ipv6_link_local_state;
- uint8_t ipv6_addr0_state;
- uint8_t ipv6_addr1_state;
- uint8_t ipv6_default_router_state;
- struct in6_addr ipv6_link_local_addr;
- struct in6_addr ipv6_addr0;
- struct in6_addr ipv6_addr1;
- struct in6_addr ipv6_default_router_addr;
/* qla82xx specific fields */
struct device_reg_82xx __iomem *qla4_8xxx_reg; /* Base I/O address */
@@ -584,16 +587,18 @@ struct scsi_qla_host {
uint32_t nx_reset_timeout;
struct completion mbx_intr_comp;
+ struct ipaddress_config ip_config;
};
static inline int is_ipv4_enabled(struct scsi_qla_host *ha)
{
- return ((ha->ipv4_options & IPOPT_IPv4_PROTOCOL_ENABLE) != 0);
+ return ((ha->ip_config.ipv4_options & IPOPT_IPv4_PROTOCOL_ENABLE) != 0);
}
static inline int is_ipv6_enabled(struct scsi_qla_host *ha)
{
- return ((ha->ipv6_options & IPV6_OPT_IPV6_PROTOCOL_ENABLE) != 0);
+ return ((ha->ip_config.ipv6_options &
+ IPV6_OPT_IPV6_PROTOCOL_ENABLE) != 0);
}
static inline int is_qla4010(struct scsi_qla_host *ha)
diff --git a/drivers/scsi/qla4xxx/ql4_init.c b/drivers/scsi/qla4xxx/ql4_init.c
index bbb2e90..464fde7 100644
--- a/drivers/scsi/qla4xxx/ql4_init.c
+++ b/drivers/scsi/qla4xxx/ql4_init.c
@@ -236,38 +236,44 @@ qla4xxx_wait_for_ip_config(struct scsi_qla_host *ha)
FW_ADDSTATE_DHCPv4_LEASE_ACQUIRED) == 0)) {
ipv4_wait = 1;
}
- if (((ha->ipv6_addl_options &
+ if (((ha->ip_config.ipv6_addl_options &
IPV6_ADDOPT_NEIGHBOR_DISCOVERY_ADDR_ENABLE) != 0) &&
- ((ha->ipv6_link_local_state == IP_ADDRSTATE_ACQUIRING) ||
- (ha->ipv6_addr0_state == IP_ADDRSTATE_ACQUIRING) ||
- (ha->ipv6_addr1_state == IP_ADDRSTATE_ACQUIRING))) {
+ ((ha->ip_config.ipv6_link_local_state ==
+ IP_ADDRSTATE_ACQUIRING) ||
+ (ha->ip_config.ipv6_addr0_state ==
+ IP_ADDRSTATE_ACQUIRING) ||
+ (ha->ip_config.ipv6_addr1_state ==
+ IP_ADDRSTATE_ACQUIRING))) {
ipv6_wait = 1;
- if ((ha->ipv6_link_local_state ==
- IP_ADDRSTATE_PREFERRED) ||
- (ha->ipv6_addr0_state == IP_ADDRSTATE_PREFERRED) ||
- (ha->ipv6_addr1_state == IP_ADDRSTATE_PREFERRED)) {
+ if ((ha->ip_config.ipv6_link_local_state ==
+ IP_ADDRSTATE_PREFERRED) ||
+ (ha->ip_config.ipv6_addr0_state ==
+ IP_ADDRSTATE_PREFERRED) ||
+ (ha->ip_config.ipv6_addr1_state ==
+ IP_ADDRSTATE_PREFERRED)) {
DEBUG2(printk(KERN_INFO "scsi%ld: %s: "
"Preferred IP configured."
" Don't wait!\n", ha->host_no,
__func__));
ipv6_wait = 0;
}
- if (memcmp(&ha->ipv6_default_router_addr, ip_address,
- IPv6_ADDR_LEN) == 0) {
+ if (memcmp(&ha->ip_config.ipv6_default_router_addr,
+ ip_address, IPv6_ADDR_LEN) == 0) {
DEBUG2(printk(KERN_INFO "scsi%ld: %s: "
"No Router configured. "
"Don't wait!\n", ha->host_no,
__func__));
ipv6_wait = 0;
}
- if ((ha->ipv6_default_router_state ==
- IPV6_RTRSTATE_MANUAL) &&
- (ha->ipv6_link_local_state ==
- IP_ADDRSTATE_TENTATIVE) &&
- (memcmp(&ha->ipv6_link_local_addr,
- &ha->ipv6_default_router_addr, 4) == 0)) {
+ if ((ha->ip_config.ipv6_default_router_state ==
+ IPV6_RTRSTATE_MANUAL) &&
+ (ha->ip_config.ipv6_link_local_state ==
+ IP_ADDRSTATE_TENTATIVE) &&
+ (memcmp(&ha->ip_config.ipv6_link_local_addr,
+ &ha->ip_config.ipv6_default_router_addr, 4) ==
+ 0)) {
DEBUG2(printk("scsi%ld: %s: LinkLocal Router & "
"IP configured. Don't wait!\n",
ha->host_no, __func__));
@@ -279,11 +285,14 @@ qla4xxx_wait_for_ip_config(struct scsi_qla_host *ha)
"IP(s) \"", ha->host_no, __func__));
if (ipv4_wait)
DEBUG2(printk("IPv4 "));
- if (ha->ipv6_link_local_state == IP_ADDRSTATE_ACQUIRING)
+ if (ha->ip_config.ipv6_link_local_state ==
+ IP_ADDRSTATE_ACQUIRING)
DEBUG2(printk("IPv6LinkLocal "));
- if (ha->ipv6_addr0_state == IP_ADDRSTATE_ACQUIRING)
+ if (ha->ip_config.ipv6_addr0_state ==
+ IP_ADDRSTATE_ACQUIRING)
DEBUG2(printk("IPv6Addr0 "));
- if (ha->ipv6_addr1_state == IP_ADDRSTATE_ACQUIRING)
+ if (ha->ip_config.ipv6_addr1_state ==
+ IP_ADDRSTATE_ACQUIRING)
DEBUG2(printk("IPv6Addr1 "));
DEBUG2(printk("\"\n"));
}
@@ -1297,8 +1306,8 @@ int qla4xxx_initialize_adapter(struct scsi_qla_host *ha,
goto exit_init_online;
/* Skip device discovery if ip and subnet is zero */
- if (memcmp(ha->ip_address, ip_address, IP_ADDR_LEN) == 0 ||
- memcmp(ha->subnet_mask, ip_address, IP_ADDR_LEN) == 0)
+ if (memcmp(ha->ip_config.ip_address, ip_address, IP_ADDR_LEN) == 0 ||
+ memcmp(ha->ip_config.subnet_mask, ip_address, IP_ADDR_LEN) == 0)
goto exit_init_online;
if (renew_ddb_list == PRESERVE_DDB_LIST) {
diff --git a/drivers/scsi/qla4xxx/ql4_mbx.c b/drivers/scsi/qla4xxx/ql4_mbx.c
index ddcbd86..19ce950 100644
--- a/drivers/scsi/qla4xxx/ql4_mbx.c
+++ b/drivers/scsi/qla4xxx/ql4_mbx.c
@@ -343,36 +343,51 @@ static void
qla4xxx_update_local_ip(struct scsi_qla_host *ha,
struct addr_ctrl_blk *init_fw_cb)
{
+ ha->ip_config.tcp_options = le16_to_cpu(init_fw_cb->ipv4_tcp_opts);
+ ha->ip_config.ipv4_options = le16_to_cpu(init_fw_cb->ipv4_ip_opts);
+ ha->ip_config.ipv4_addr_state =
+ le16_to_cpu(init_fw_cb->ipv4_addr_state);
+
+ if (ha->acb_version == ACB_SUPPORTED) {
+ ha->ip_config.ipv6_options = init_fw_cb->ipv6_opts;
+ ha->ip_config.ipv6_addl_options = init_fw_cb->ipv6_addtl_opts;
+ }
+
/* Save IPv4 Address Info */
- memcpy(ha->ip_address, init_fw_cb->ipv4_addr,
- min(sizeof(ha->ip_address), sizeof(init_fw_cb->ipv4_addr)));
- memcpy(ha->subnet_mask, init_fw_cb->ipv4_subnet,
- min(sizeof(ha->subnet_mask), sizeof(init_fw_cb->ipv4_subnet)));
- memcpy(ha->gateway, init_fw_cb->ipv4_gw_addr,
- min(sizeof(ha->gateway), sizeof(init_fw_cb->ipv4_gw_addr)));
+ memcpy(ha->ip_config.ip_address, init_fw_cb->ipv4_addr,
+ min(sizeof(ha->ip_config.ip_address),
+ sizeof(init_fw_cb->ipv4_addr)));
+ memcpy(ha->ip_config.subnet_mask, init_fw_cb->ipv4_subnet,
+ min(sizeof(ha->ip_config.subnet_mask),
+ sizeof(init_fw_cb->ipv4_subnet)));
+ memcpy(ha->ip_config.gateway, init_fw_cb->ipv4_gw_addr,
+ min(sizeof(ha->ip_config.gateway),
+ sizeof(init_fw_cb->ipv4_gw_addr)));
if (is_ipv6_enabled(ha)) {
/* Save IPv6 Address */
- ha->ipv6_link_local_state = init_fw_cb->ipv6_lnk_lcl_addr_state;
- ha->ipv6_addr0_state = init_fw_cb->ipv6_addr0_state;
- ha->ipv6_addr1_state = init_fw_cb->ipv6_addr1_state;
- ha->ipv6_default_router_state = init_fw_cb->ipv6_dflt_rtr_state;
- ha->ipv6_link_local_addr.in6_u.u6_addr8[0] = 0xFE;
- ha->ipv6_link_local_addr.in6_u.u6_addr8[1] = 0x80;
-
- memcpy(&ha->ipv6_link_local_addr.in6_u.u6_addr8[8],
+ ha->ip_config.ipv6_link_local_state =
+ init_fw_cb->ipv6_lnk_lcl_addr_state;
+ ha->ip_config.ipv6_addr0_state = init_fw_cb->ipv6_addr0_state;
+ ha->ip_config.ipv6_addr1_state = init_fw_cb->ipv6_addr1_state;
+ ha->ip_config.ipv6_default_router_state =
+ init_fw_cb->ipv6_dflt_rtr_state;
+ ha->ip_config.ipv6_link_local_addr.in6_u.u6_addr8[0] = 0xFE;
+ ha->ip_config.ipv6_link_local_addr.in6_u.u6_addr8[1] = 0x80;
+
+ memcpy(&ha->ip_config.ipv6_link_local_addr.in6_u.u6_addr8[8],
init_fw_cb->ipv6_if_id,
- min(sizeof(ha->ipv6_link_local_addr)/2,
+ min(sizeof(ha->ip_config.ipv6_link_local_addr)/2,
sizeof(init_fw_cb->ipv6_if_id)));
- memcpy(&ha->ipv6_addr0, init_fw_cb->ipv6_addr0,
- min(sizeof(ha->ipv6_addr0),
+ memcpy(&ha->ip_config.ipv6_addr0, init_fw_cb->ipv6_addr0,
+ min(sizeof(ha->ip_config.ipv6_addr0),
sizeof(init_fw_cb->ipv6_addr0)));
- memcpy(&ha->ipv6_addr1, init_fw_cb->ipv6_addr1,
- min(sizeof(ha->ipv6_addr1),
+ memcpy(&ha->ip_config.ipv6_addr1, init_fw_cb->ipv6_addr1,
+ min(sizeof(ha->ip_config.ipv6_addr1),
sizeof(init_fw_cb->ipv6_addr1)));
- memcpy(&ha->ipv6_default_router_addr,
+ memcpy(&ha->ip_config.ipv6_default_router_addr,
init_fw_cb->ipv6_dflt_rtr_addr,
- min(sizeof(ha->ipv6_default_router_addr),
+ min(sizeof(ha->ip_config.ipv6_default_router_addr),
sizeof(init_fw_cb->ipv6_dflt_rtr_addr)));
}
}
@@ -397,9 +412,6 @@ qla4xxx_update_local_ifcb(struct scsi_qla_host *ha,
/* Save some info in adapter structure. */
ha->acb_version = init_fw_cb->acb_version;
ha->firmware_options = le16_to_cpu(init_fw_cb->fw_options);
- ha->tcp_options = le16_to_cpu(init_fw_cb->ipv4_tcp_opts);
- ha->ipv4_options = le16_to_cpu(init_fw_cb->ipv4_ip_opts);
- ha->ipv4_addr_state = le16_to_cpu(init_fw_cb->ipv4_addr_state);
ha->heartbeat_interval = init_fw_cb->hb_interval;
memcpy(ha->name_string, init_fw_cb->iscsi_name,
min(sizeof(ha->name_string),
@@ -407,10 +419,6 @@ qla4xxx_update_local_ifcb(struct scsi_qla_host *ha,
/*memcpy(ha->alias, init_fw_cb->Alias,
min(sizeof(ha->alias), sizeof(init_fw_cb->Alias)));*/
- if (ha->acb_version == ACB_SUPPORTED) {
- ha->ipv6_options = init_fw_cb->ipv6_opts;
- ha->ipv6_addl_options = init_fw_cb->ipv6_addtl_opts;
- }
qla4xxx_update_local_ip(ha, init_fw_cb);
return QLA_SUCCESS;
diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c
index 7b2e237..3ea19ac 100644
--- a/drivers/scsi/qla4xxx/ql4_os.c
+++ b/drivers/scsi/qla4xxx/ql4_os.c
@@ -190,9 +190,7 @@ static int qla4xxx_host_get_param(struct Scsi_Host *shost,
len = sysfs_format_mac(buf, ha->my_mac, MAC_ADDR_LEN);
break;
case ISCSI_HOST_PARAM_IPADDRESS:
- len = sprintf(buf, "%d.%d.%d.%d\n", ha->ip_address[0],
- ha->ip_address[1], ha->ip_address[2],
- ha->ip_address[3]);
+ len = sprintf(buf, "%pI4\n", &ha->ip_config.ip_address);
break;
case ISCSI_HOST_PARAM_INITIATOR_NAME:
len = sprintf(buf, "%s\n", ha->name_string);
--
1.7.3.2
^ permalink raw reply related [flat|nested] 24+ messages in thread
* [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-02 18:34 [RFC-V2 PATCH 0/5] Proposal for iSCSI Network Configuration vikas.chaudhary
` (2 preceding siblings ...)
2011-04-02 18:34 ` [RFC-V2 PATCH 3/5] qla4xxx: Added new "struct ipaddress_config" vikas.chaudhary
@ 2011-04-02 18:34 ` vikas.chaudhary
2011-04-13 4:23 ` Mike Christie
` (2 more replies)
2011-04-02 18:34 ` [RFC-V2 PATCH 5/5] qla4xxx: added support to show multiple iface " vikas.chaudhary
4 siblings, 3 replies; 24+ messages in thread
From: vikas.chaudhary @ 2011-04-02 18:34 UTC (permalink / raw)
To: James.Bottomley, michaelc
Cc: linux-scsi, open-iscsi, vikas.chaudhary, lalit.chandivade, ravi.anand
From: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
To support multiple network addresses per adapter need to have a new way to
represent network interface (net iface) in sysfs.
Currently only one ipaddress and hwaddress is displayed
\# ls /sys/class/iscsi_host/host18
device hwaddress initiatorname ipaddress power subsystem uevent
In this patch the net iface is presented as a separate class device.
The one that can be added/removed dynamically or statically, based on how
the user configures the multiple net iface on the adapter.
The new sysfs directory would look like this
\# /sys/class/iscsi_iface/
|
|- ipv4-iface-<host_no>-<iface_no>/ <-- for ipv4
|- ipaddress
|- subnet
|- gateway
|- bootproto
|- state
|- ipv6-iface-<host_no>-<iface_no>/ <-- for ipv6
|- ipaddress
|- link_local_addr
|- router_addr
|- ipaddr_autocfg
|- linklocal_autocfg
|- state
We can also get above .iscsi_iface. class from:-
\# /sys/class/iscsi_host/host12/device/iscsi_iface/
With this change, iscsadm would need changes to create iface by getting
hw/ip related data from new sysfs path.
The old path still can be maintained to keep backward compatibility.
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: Lalit Chandivade <lalit.chandivade@qlogic.com>
Reviewed-by: Harish Zunjarrao <harish.zunjarrao@qlogic.com>
---
drivers/scsi/scsi_transport_iscsi.c | 215 ++++++++++++++++++++++++++++++++++-
include/scsi/iscsi_if.h | 17 +++
include/scsi/scsi_transport_iscsi.h | 24 ++++
3 files changed, 255 insertions(+), 1 deletions(-)
diff --git a/drivers/scsi/scsi_transport_iscsi.c b/drivers/scsi/scsi_transport_iscsi.c
index 38fcfc0..35ec15d 100644
--- a/drivers/scsi/scsi_transport_iscsi.c
+++ b/drivers/scsi/scsi_transport_iscsi.c
@@ -23,6 +23,7 @@
#include <linux/module.h>
#include <linux/mutex.h>
#include <linux/slab.h>
+#include <linux/idr.h>
#include <net/tcp.h>
#include <scsi/scsi.h>
#include <scsi/scsi_host.h>
@@ -270,6 +271,190 @@ struct iscsi_endpoint *iscsi_lookup_endpoint(u64 handle)
}
EXPORT_SYMBOL_GPL(iscsi_lookup_endpoint);
+
+/* Inerface to display network param to sysfs */
+
+static void iscsi_iface_release(struct device *dev)
+{
+ struct iscsi_iface *iface = iscsi_dev_to_iface(dev);
+ kfree(iface);
+}
+
+static struct class iscsi_iface_class = {
+ .name = "iscsi_iface",
+ .dev_release = iscsi_iface_release,
+};
+
+#define ISCSI_IFACE_ATTR(_prefix, _name, _mode, _show, _store) \
+struct device_attribute dev_attr_##_prefix##_##_name = \
+ __ATTR(_name, _mode, _show, _store)
+
+/* iface attrs show */
+#define iscsi_iface_attr_show(type, name, param) \
+static ssize_t \
+show_##type##_##name(struct device *dev, struct device_attribute *attr, \
+ char *buf) \
+{ \
+ struct iscsi_iface *iface = iscsi_dev_to_iface(dev); \
+ struct iscsi_transport *t = iface->transport; \
+ return t->get_iface_param(iface, param, buf); \
+} \
+
+#define iscsi_iface_attr(type, name, param) \
+ iscsi_iface_attr_show(type, name, param) \
+static ISCSI_IFACE_ATTR(type, name, S_IRUGO, show_##type##_##name, NULL);
+
+/* generic read only ipvi4 attribute */
+iscsi_iface_attr(ipv4_iface, ipaddress, ISCSI_NET_PARAM_IPV4_ADDR);
+iscsi_iface_attr(ipv4_iface, gateway, ISCSI_NET_PARAM_IPV4_GW);
+iscsi_iface_attr(ipv4_iface, subnet, ISCSI_NET_PARAM_IPV4_SUBNET);
+iscsi_iface_attr(ipv4_iface, state, ISCSI_NET_PARAM_IFACE_STATE);
+iscsi_iface_attr(ipv4_iface, bootproto, ISCSI_NET_PARAM_IPV4_BOOTPROTO);
+
+/* generic read only ipv6 attribute */
+iscsi_iface_attr(ipv6_iface, ipaddress, ISCSI_NET_PARAM_IPV6_ADDR);
+iscsi_iface_attr(ipv6_iface, link_local_addr, ISCSI_NET_PARAM_IPV6_LINKLOCAL);
+iscsi_iface_attr(ipv6_iface, router_addr, ISCSI_NET_PARAM_IPV6_ROUTER);
+iscsi_iface_attr(ipv6_iface, ipaddr_autocfg,
+ ISCSI_NET_PARAM_IPV6_ADDR_AUTOCFG);
+iscsi_iface_attr(ipv6_iface, linklocal_autocfg,
+ ISCSI_NET_PARAM_IPV6_LINKLOCAL_AUTOCFG);
+iscsi_iface_attr(ipv6_iface, state, ISCSI_NET_PARAM_IFACE_STATE);
+
+/* IPv4 Attrs */
+#define ISCSI_IPV4_IFACE_ATTRS 5
+static struct attribute *iscsi_ipv4_iface_attrs[ISCSI_IPV4_IFACE_ATTRS + 1];
+
+#define SETUP_IPV4_IFACE_RD_ATTR(field, param_flag) \
+do { \
+ if (tt->ipv4_iface_param_mask & param_flag) { \
+ iscsi_ipv4_iface_attrs[count] = \
+ &dev_attr_ipv4_iface_##field.attr; \
+ count++; \
+ } \
+} while (0)
+
+static struct attribute_group iscsi_ipv4_iface_group = {
+ .attrs = iscsi_ipv4_iface_attrs,
+};
+
+/* IPv6 Attrs */
+#define ISCSI_IPV6_IFACE_ATTRS 6
+static struct attribute *iscsi_ipv6_iface_attrs[ISCSI_IPV6_IFACE_ATTRS + 1];
+
+#define SETUP_IPV6_IFACE_RD_ATTR(field, param_flag) \
+do { \
+ if (tt->ipv6_iface_param_mask & param_flag) { \
+ iscsi_ipv6_iface_attrs[count] = \
+ &dev_attr_ipv6_iface_##field.attr; \
+ count++; \
+ } \
+} while (0)
+
+static struct attribute_group iscsi_ipv6_iface_group = {
+ .attrs = iscsi_ipv6_iface_attrs,
+};
+
+static DEFINE_IDR(iscsi_iface_idr);
+static DEFINE_SPINLOCK(iscsi_iface_lock);
+
+struct iscsi_iface *
+iscsi_create_iface(struct Scsi_Host *shost, struct iscsi_transport *transport,
+ uint32_t iface_type, uint32_t iface_num, int dd_size)
+{
+ struct iscsi_iface *iface;
+ int id;
+ int err;
+
+ iface = kzalloc(sizeof(*iface) + dd_size, GFP_KERNEL);
+ if (!iface)
+ return NULL;
+
+iface_idr_again:
+ if (!idr_pre_get(&iscsi_iface_idr, GFP_KERNEL))
+ goto free_iface;
+
+ spin_lock(&iscsi_iface_lock);
+ err = idr_get_new(&iscsi_iface_idr, iface, &id);
+ if (err == -EAGAIN) {
+ spin_unlock(&iscsi_iface_lock);
+ goto iface_idr_again;
+ }
+ spin_unlock(&iscsi_iface_lock);
+
+ if (err)
+ goto free_iface;
+
+ iface->id = id;
+ iface->transport = transport;
+ iface->iface_type = iface_type;
+ iface->iface_num = iface_num;
+ iface->dev.class = &iscsi_iface_class;
+ /* parent reference */
+ iface->dev.parent = get_device(&shost->shost_gendev);
+ if (iface_type == IFACE_TYPE_IPV4)
+ dev_set_name(&iface->dev, "ipv4-iface-%u-%u", shost->host_no,
+ iface_num);
+ else if (iface_type == IFACE_TYPE_IPV6)
+ dev_set_name(&iface->dev, "ipv6-iface-%u-%u", shost->host_no,
+ iface_num);
+ else
+ goto free_iface;
+
+ err = device_register(&iface->dev);
+ if (err)
+ goto free_iface;
+
+ if (iface_type == IFACE_TYPE_IPV4)
+ err = sysfs_create_group(&iface->dev.kobj,
+ &iscsi_ipv4_iface_group);
+ else if (iface_type == IFACE_TYPE_IPV6)
+ err = sysfs_create_group(&iface->dev.kobj,
+ &iscsi_ipv6_iface_group);
+
+ if (err)
+ goto iface_unregister_dev;
+
+ if (dd_size)
+ iface->dd_data = &iface[1];
+ return iface;
+
+iface_unregister_dev:
+ idr_remove(&iscsi_iface_idr, id);
+ device_unregister(&iface->dev);
+
+free_iface:
+ kfree(iface);
+ return NULL;
+}
+EXPORT_SYMBOL_GPL(iscsi_create_iface);
+
+void iscsi_destroy_iface(struct iscsi_iface *iface)
+{
+ if (iface->iface_type == IFACE_TYPE_IPV6)
+ sysfs_remove_group(&iface->dev.kobj, &iscsi_ipv6_iface_group);
+ else
+ sysfs_remove_group(&iface->dev.kobj, &iscsi_ipv4_iface_group);
+
+ idr_remove(&iscsi_iface_idr, iface->id);
+ device_unregister(&iface->dev);
+}
+EXPORT_SYMBOL_GPL(iscsi_destroy_iface);
+
+struct iscsi_iface *iscsi_lookup_iface(int handle)
+{
+ struct iscsi_iface *iface;
+
+ iface = idr_find(&iscsi_iface_idr, handle);
+
+ return iface;
+}
+EXPORT_SYMBOL_GPL(iscsi_lookup_iface);
+
+
+/* end */
+
+
static int iscsi_setup_host(struct transport_container *tc, struct device *dev,
struct device *cdev)
{
@@ -2175,6 +2360,27 @@ iscsi_register_transport(struct iscsi_transport *tt)
BUG_ON(count > ISCSI_SESSION_ATTRS);
priv->session_attrs[count] = NULL;
+ count = 0;
+
+ SETUP_IPV4_IFACE_RD_ATTR(ipaddress, ISCSI_NET_IPV4_ADDR);
+ SETUP_IPV4_IFACE_RD_ATTR(gateway, ISCSI_NET_IPV4_ADDR);
+ SETUP_IPV4_IFACE_RD_ATTR(bootproto, ISCSI_NET_IPV4_ADDR);
+ SETUP_IPV4_IFACE_RD_ATTR(subnet, ISCSI_NET_IPV4_ADDR);
+ SETUP_IPV4_IFACE_RD_ATTR(state, ISCSI_NET_IPV4_ADDR);
+ BUG_ON(count > ISCSI_IPV4_IFACE_ATTRS);
+ iscsi_ipv4_iface_attrs[count] = NULL;
+ count = 0;
+
+ SETUP_IPV6_IFACE_RD_ATTR(ipaddress, ISCSI_NET_IPV6_ADDR);
+ SETUP_IPV6_IFACE_RD_ATTR(link_local_addr, ISCSI_NET_IPV6_LINKLOCAL);
+ SETUP_IPV6_IFACE_RD_ATTR(router_addr, ISCSI_NET_IPV6_ROUTER);
+ SETUP_IPV6_IFACE_RD_ATTR(ipaddr_autocfg,
+ ISCSI_NET_IPV6_ADDR_AUTOCFG);
+ SETUP_IPV6_IFACE_RD_ATTR(linklocal_autocfg,
+ ISCSI_NET_IPV6_LINKLOCAL_AUTOCFG);
+ SETUP_IPV6_IFACE_RD_ATTR(state, ISCSI_NET_IFACE_STATE);
+ BUG_ON(count > ISCSI_IPV6_IFACE_ATTRS);
+ iscsi_ipv6_iface_attrs[count] = NULL;
spin_lock_irqsave(&iscsi_transport_lock, flags);
list_add(&priv->list, &iscsi_transports);
@@ -2237,10 +2443,14 @@ static __init int iscsi_transport_init(void)
if (err)
goto unregister_transport_class;
- err = transport_class_register(&iscsi_host_class);
+ err = class_register(&iscsi_iface_class);
if (err)
goto unregister_endpoint_class;
+ err = transport_class_register(&iscsi_host_class);
+ if (err)
+ goto unregister_iface_class;
+
err = transport_class_register(&iscsi_connection_class);
if (err)
goto unregister_host_class;
@@ -2270,6 +2480,8 @@ unregister_conn_class:
transport_class_unregister(&iscsi_connection_class);
unregister_host_class:
transport_class_unregister(&iscsi_host_class);
+unregister_iface_class:
+ class_unregister(&iscsi_iface_class);
unregister_endpoint_class:
class_unregister(&iscsi_endpoint_class);
unregister_transport_class:
@@ -2285,6 +2497,7 @@ static void __exit iscsi_transport_exit(void)
transport_class_unregister(&iscsi_session_class);
transport_class_unregister(&iscsi_host_class);
class_unregister(&iscsi_endpoint_class);
+ class_unregister(&iscsi_iface_class);
class_unregister(&iscsi_transport_class);
}
diff --git a/include/scsi/iscsi_if.h b/include/scsi/iscsi_if.h
index 312b495..a55ee82 100644
--- a/include/scsi/iscsi_if.h
+++ b/include/scsi/iscsi_if.h
@@ -272,6 +272,23 @@ enum iscsi_net_param_type {
ISCSI_NET_PARAM_IFACE_STATE = 13,
};
+#define ISCSI_NET_IPV4_ADDR (1ULL << ISCSI_NET_PARAM_IPV4_ADDR)
+#define ISCSI_NET_IPV4_SUBNET (1ULL << ISCSI_NET_PARAM_IPV4_SUBNET)
+#define ISCSI_NET_IPV4_GW (1ULL << ISCSI_NET_PARAM_IPV4_GW)
+#define ISCSI_NET_IPV4_BOOTPROTO (1ULL << ISCSI_NET_PARAM_IPV4_BOOTPROTO)
+#define ISCSI_NET_VLAN (1ULL << ISCSI_NET_PARAM_VLAN)
+#define ISCSI_NET_MAC (1ULL << ISCSI_NET_PARAM_MAC)
+#define ISCSI_NET_IPV6_LINKLOCAL (1ULL << ISCSI_NET_PARAM_IPV6_LINKLOCAL)
+#define ISCSI_NET_IPV6_ADDR (1ULL << ISCSI_NET_PARAM_IPV6_ADDR)
+#define ISCSI_NET_IPV6_ROUTER (1ULL << ISCSI_NET_PARAM_IPV6_ROUTER)
+#define ISCSI_NET_IPV6_AUTOCFG (1ULL << ISCSI_NET_PARAM_IPV6_AUTOCFG)
+#define ISCSI_NET_IPV6_ADDR_AUTOCFG \
+ (1ULL << ISCSI_NET_PARAM_IPV6_ADDR_AUTOCFG)
+#define ISCSI_NET_IPV6_LINKLOCAL_AUTOCFG \
+ (1ULL << ISCSI_NET_PARAM_IPV6_LINKLOCAL_AUTOCFG)
+#define ISCSI_NET_IFACE_STATE (1ULL << ISCSI_NET_PARAM_IFACE_STATE)
+
+
/* 20 param per iface * 10 iface per port = 200 params */
#define ISCSI_MAX_IFACE_PER_HW 10
#define ISCSI_MAX_PARAM_PER_IFACE 20
diff --git a/include/scsi/scsi_transport_iscsi.h b/include/scsi/scsi_transport_iscsi.h
index 45b9df9..29d314a 100644
--- a/include/scsi/scsi_transport_iscsi.h
+++ b/include/scsi/scsi_transport_iscsi.h
@@ -37,6 +37,7 @@ struct iscsi_cls_conn;
struct iscsi_conn;
struct iscsi_task;
struct sockaddr;
+struct iscsi_iface;
/**
* struct iscsi_transport - iSCSI Transport template
@@ -87,6 +88,8 @@ struct iscsi_transport {
/* LLD sets this to indicate what values it can export to sysfs */
uint64_t param_mask;
uint64_t host_param_mask;
+ uint64_t ipv4_iface_param_mask;
+ uint64_t ipv6_iface_param_mask;
struct iscsi_cls_session *(*create_session) (struct iscsi_endpoint *ep,
uint16_t cmds_max, uint16_t qdepth,
uint32_t sn);
@@ -138,6 +141,8 @@ struct iscsi_transport {
uint32_t enable, struct sockaddr *dst_addr);
int (*set_path) (struct Scsi_Host *shost, struct iscsi_path *params);
int (*set_net_config) (struct Scsi_Host *shost, char *data, int count);
+ int (*get_iface_param) (struct iscsi_iface *iface,
+ enum iscsi_net_param_type param, char *buf);
};
/*
@@ -229,6 +234,20 @@ struct iscsi_endpoint {
struct iscsi_cls_conn *conn;
};
+struct iscsi_iface {
+ struct device dev;
+ struct iscsi_transport *transport;
+ uint64_t id;
+ uint32_t iface_type; /* IPv4 or IPv6 */
+ uint32_t iface_num; /* iface number, 0 - n */
+ void *dd_data; /* LLD private data */
+};
+
+#define iscsi_dev_to_iface(_dev) \
+ container_of(_dev, struct iscsi_iface, dev)
+#define iscsi_iface_to_shost(_iface) \
+ dev_to_shost(_iface->dev.parent)
+
/*
* session and connection functions that can be used by HW iSCSI LLDs
*/
@@ -262,5 +281,10 @@ extern struct iscsi_endpoint *iscsi_create_endpoint(int dd_size);
extern void iscsi_destroy_endpoint(struct iscsi_endpoint *ep);
extern struct iscsi_endpoint *iscsi_lookup_endpoint(u64 handle);
extern int iscsi_block_scsi_eh(struct scsi_cmnd *cmd);
+extern struct iscsi_iface *iscsi_create_iface(struct Scsi_Host *shost,
+ struct iscsi_transport *transport, uint32_t iface_type,
+ uint32_t iface_num, int dd_size);
+extern void iscsi_destroy_iface(struct iscsi_iface *iface);
+extern struct iscsi_iface *iscsi_lookup_iface(int handle);
#endif
--
1.7.3.2
^ permalink raw reply related [flat|nested] 24+ messages in thread
* [RFC-V2 PATCH 5/5] qla4xxx: added support to show multiple iface in sysfs
2011-04-02 18:34 [RFC-V2 PATCH 0/5] Proposal for iSCSI Network Configuration vikas.chaudhary
` (3 preceding siblings ...)
2011-04-02 18:34 ` [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs vikas.chaudhary
@ 2011-04-02 18:34 ` vikas.chaudhary
4 siblings, 0 replies; 24+ messages in thread
From: vikas.chaudhary @ 2011-04-02 18:34 UTC (permalink / raw)
To: James.Bottomley, michaelc
Cc: linux-scsi, open-iscsi, vikas.chaudhary, lalit.chandivade, ravi.anand
From: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Reviewed-by: Lalit Chandivade <lalit.chandivade@qlogic.com>
Reviewed-by: Harish Zunjarrao <harish.zunjarrao@qlogic.com>
---
drivers/scsi/qla4xxx/ql4_def.h | 3 +
drivers/scsi/qla4xxx/ql4_os.c | 112 ++++++++++++++++++++++++++++++++++++++++
2 files changed, 115 insertions(+), 0 deletions(-)
diff --git a/drivers/scsi/qla4xxx/ql4_def.h b/drivers/scsi/qla4xxx/ql4_def.h
index 74faf54..aa3ba03 100644
--- a/drivers/scsi/qla4xxx/ql4_def.h
+++ b/drivers/scsi/qla4xxx/ql4_def.h
@@ -588,6 +588,9 @@ struct scsi_qla_host {
struct completion mbx_intr_comp;
struct ipaddress_config ip_config;
+ struct iscsi_iface *iface_ipv4;
+ struct iscsi_iface *iface_ipv6_1;
+ struct iscsi_iface *iface_ipv6_2;
};
static inline int is_ipv4_enabled(struct scsi_qla_host *ha)
diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c
index 3ea19ac..211fef7 100644
--- a/drivers/scsi/qla4xxx/ql4_os.c
+++ b/drivers/scsi/qla4xxx/ql4_os.c
@@ -82,6 +82,8 @@ static int qla4xxx_host_get_param(struct Scsi_Host *shost,
enum iscsi_host_param param, char *buf);
static int qla4xxx_set_net_config(struct Scsi_Host *shost, char *data,
int count);
+static int qla4xxx_get_iface_param(struct iscsi_iface *iface,
+ enum iscsi_net_param_type param, char *buf);
static void qla4xxx_recovery_timedout(struct iscsi_cls_session *session);
static enum blk_eh_timer_return qla4xxx_eh_cmd_timed_out(struct scsi_cmnd *sc);
@@ -139,16 +141,96 @@ static struct iscsi_transport qla4xxx_iscsi_transport = {
.host_param_mask = ISCSI_HOST_HWADDRESS |
ISCSI_HOST_IPADDRESS |
ISCSI_HOST_INITIATOR_NAME,
+ .ipv4_iface_param_mask = ISCSI_NET_IPV4_ADDR |
+ ISCSI_NET_IPV4_SUBNET |
+ ISCSI_NET_IPV4_GW |
+ ISCSI_NET_IPV4_BOOTPROTO |
+ ISCSI_NET_IFACE_STATE,
+ .ipv6_iface_param_mask = ISCSI_NET_IPV6_LINKLOCAL |
+ ISCSI_NET_IPV6_ADDR |
+ ISCSI_NET_IPV6_ROUTER |
+ ISCSI_NET_IPV6_ADDR_AUTOCFG |
+ ISCSI_NET_IPV6_LINKLOCAL_AUTOCFG |
+ ISCSI_NET_IFACE_STATE,
.tgt_dscvr = qla4xxx_tgt_dscvr,
.get_conn_param = qla4xxx_conn_get_param,
.get_session_param = qla4xxx_sess_get_param,
.get_host_param = qla4xxx_host_get_param,
.set_net_config = qla4xxx_set_net_config,
.session_recovery_timedout = qla4xxx_recovery_timedout,
+ .get_iface_param = qla4xxx_get_iface_param,
};
static struct scsi_transport_template *qla4xxx_scsi_transport;
+static int qla4xxx_get_iface_param(struct iscsi_iface *iface,
+ enum iscsi_net_param_type param, char *buf)
+{
+ struct Scsi_Host *shost = iscsi_iface_to_shost(iface);
+ struct scsi_qla_host *ha = to_qla_host(shost);
+ int len = -ENOSYS;
+
+ switch (param) {
+ case ISCSI_NET_PARAM_IPV4_ADDR:
+ len = sprintf(buf, "%pI4\n", &ha->ip_config.ip_address);
+ break;
+ case ISCSI_NET_PARAM_IPV4_SUBNET:
+ len = sprintf(buf, "%pI4\n", &ha->ip_config.subnet_mask);
+ break;
+ case ISCSI_NET_PARAM_IPV4_GW:
+ len = sprintf(buf, "%pI4\n", &ha->ip_config.gateway);
+ break;
+ case ISCSI_NET_PARAM_IFACE_STATE:
+ if (iface->iface_type == IFACE_TYPE_IPV4)
+ len = sprintf(buf, "%s\n",
+ (ha->ip_config.ipv4_options &
+ IPOPT_IPv4_PROTOCOL_ENABLE) ?
+ "enable" : "disable");
+ else if (iface->iface_type == IFACE_TYPE_IPV6)
+ len = sprintf(buf, "%s\n",
+ (ha->ip_config.ipv6_options &
+ IPV6_OPT_IPV6_PROTOCOL_ENABLE) ?
+ "enable" : "disable");
+ break;
+ case ISCSI_NET_PARAM_IPV4_BOOTPROTO:
+ len = sprintf(buf, "%s\n",
+ (ha->ip_config.tcp_options & TCPOPT_DHCP_ENABLE) ?
+ "dhcp" : "static");
+ break;
+ case ISCSI_NET_PARAM_IPV6_ADDR:
+ if (iface->iface_num == 1)
+ len = sprintf(buf, "%pI6\n", &ha->ip_config.ipv6_addr0);
+ if (iface->iface_num == 2)
+ len = sprintf(buf, "%pI6\n", &ha->ip_config.ipv6_addr1);
+ break;
+ case ISCSI_NET_PARAM_IPV6_LINKLOCAL:
+ len = sprintf(buf, "%pI6\n",
+ &ha->ip_config.ipv6_link_local_addr);
+ break;
+ case ISCSI_NET_PARAM_IPV6_ROUTER:
+ len = sprintf(buf, "%pI6\n",
+ &ha->ip_config.ipv6_default_router_addr);
+ break;
+ case ISCSI_NET_PARAM_IPV6_ADDR_AUTOCFG:
+ len = sprintf(buf, "%s\n",
+ (ha->ip_config.ipv6_addl_options &
+ IPV6_ADDOPT_NEIGHBOR_DISCOVERY_ADDR_ENABLE) ?
+ "nd" : "static");
+ break;
+ case ISCSI_NET_PARAM_IPV6_LINKLOCAL_AUTOCFG:
+ len = sprintf(buf, "%s\n",
+ (ha->ip_config.ipv6_addl_options &
+ IPV6_ADDOPT_AUTOCONFIG_LINK_LOCAL_ADDR) ?
+ "auto" : "static");
+ break;
+ default:
+ len = -ENOSYS;
+ }
+
+ return len;
+}
+
+
static enum blk_eh_timer_return qla4xxx_eh_cmd_timed_out(struct scsi_cmnd *sc)
{
struct iscsi_cls_session *session;
@@ -1933,6 +2015,31 @@ uint16_t qla4_8xxx_rd_shdw_rsp_q_in(struct scsi_qla_host *ha)
return (uint16_t)le32_to_cpu(readl(&ha->qla4_8xxx_reg->rsp_q_in));
}
+static void qla4xxx_create_iface(struct scsi_qla_host *ha)
+{
+ /* IPv4 */
+ ha->iface_ipv4 = iscsi_create_iface(ha->host,
+ &qla4xxx_iscsi_transport, IFACE_TYPE_IPV4, 1, 0);
+
+ /* IPv6 iface-1 */
+ ha->iface_ipv6_1 = iscsi_create_iface(ha->host,
+ &qla4xxx_iscsi_transport, IFACE_TYPE_IPV6, 1, 0);
+
+ /* IPv6 iface-2 */
+ ha->iface_ipv6_2 = iscsi_create_iface(ha->host,
+ &qla4xxx_iscsi_transport, IFACE_TYPE_IPV6, 2, 0);
+}
+
+static void qla4xxx_destroy_iface(struct scsi_qla_host *ha)
+{
+ if (ha->iface_ipv4)
+ iscsi_destroy_iface(ha->iface_ipv4);
+ if (ha->iface_ipv6_1)
+ iscsi_destroy_iface(ha->iface_ipv6_1);
+ if (ha->iface_ipv6_2)
+ iscsi_destroy_iface(ha->iface_ipv6_2);
+}
+
/**
* qla4xxx_probe_adapter - callback function to probe HBA
* @pdev: pointer to pci_dev structure
@@ -2133,6 +2240,8 @@ static int __devinit qla4xxx_probe_adapter(struct pci_dev *pdev,
ha->host_no, ha->firmware_version[0], ha->firmware_version[1],
ha->patch_number, ha->build_number);
scsi_scan_host(host);
+
+ qla4xxx_create_iface(ha);
return 0;
probe_failed:
@@ -2202,6 +2311,9 @@ static void __devexit qla4xxx_remove_adapter(struct pci_dev *pdev)
/* remove devs from iscsi_sessions to scsi_devices */
qla4xxx_free_ddb_list(ha);
+ /* destroy iface from sysfs */
+ qla4xxx_destroy_iface(ha);
+
scsi_remove_host(ha->host);
qla4xxx_free_adapter(ha);
--
1.7.3.2
^ permalink raw reply related [flat|nested] 24+ messages in thread
* Re: [RFC-V2 PATCH 1/5] iscsi_transport: add support for set_net_config
2011-04-02 18:34 ` [RFC-V2 PATCH 1/5] iscsi_transport: add support for set_net_config vikas.chaudhary
@ 2011-04-13 3:41 ` Mike Christie
0 siblings, 0 replies; 24+ messages in thread
From: Mike Christie @ 2011-04-13 3:41 UTC (permalink / raw)
To: open-iscsi
Cc: vikas.chaudhary, James.Bottomley, linux-scsi, lalit.chandivade,
ravi.anand, Harish Zunjarrao
On 04/02/2011 01:34 PM, vikas.chaudhary@qlogic.com wrote:
> +/* 20 param per iface * 10 iface per port = 200 params */
> +#define ISCSI_MAX_IFACE_PER_HW 10
> +#define ISCSI_MAX_PARAM_PER_IFACE 20
> +#define ISCSI_MAX_NET_PARAMS (ISCSI_MAX_IFACE_PER_HW * \
> + ISCSI_MAX_PARAM_PER_IFACE)
> +
This group is not used. Remove.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [RFC-V2 PATCH 2/5] qla4xxx: add support for set_net_config
2011-04-02 18:34 ` [RFC-V2 PATCH 2/5] qla4xxx: " vikas.chaudhary
@ 2011-04-13 3:55 ` Mike Christie
[not found] ` <4DA51ECA.1020507-hcNo3dDEHLuVc3sceRu5cw@public.gmane.org>
0 siblings, 1 reply; 24+ messages in thread
From: Mike Christie @ 2011-04-13 3:55 UTC (permalink / raw)
To: vikas.chaudhary
Cc: James.Bottomley, linux-scsi, open-iscsi, lalit.chandivade,
ravi.anand, Harish Zunjarrao
On 04/02/2011 01:34 PM, vikas.chaudhary@qlogic.com wrote:
>
> +/* Only for primary ACB, secondary ACB not supported
> + * iface_num = 0: valid -> linklocal, routable, router, autocfg options
> + * iface_num = 1: valid -> routable
> + */
> +static inline void qla4xxx_set_ipv6(struct iscsi_net_param *net_param,
> + struct addr_ctrl_blk *init_fw_cb)
> +{
> + switch (net_param->param_type) {
> + case ISCSI_NET_PARAM_IPV6_ADDR:
> + if (net_param->iface_num& 0x1) {
Is this my mailer?
I think you want to check for iface_num == 1. Above in the function
comments and in other places you set the iface_num as a int, but in this
function you check it more like a bitmask.
> + case ISCSI_NET_PARAM_IPV6_LINKLOCAL:
> + if (net_param->iface_num& 0x1)
> + case ISCSI_NET_PARAM_IPV6_ROUTER:
> + if (net_param->iface_num& 0x1)
> + case ISCSI_NET_PARAM_IPV6_ADDR_AUTOCFG:
> + /* Autocfg applies to even interface */
> + if (net_param->iface_num& 0x1)
However, in other places you have that comment about even applying to
autocfg interfaces.
Why does it say even but you only support 2 ipv6 ifaces in the patches?
Can you support more? Are you coding it in preparation for more?
.....
> +
> +static int
> +qla4xxx_set_net_config(struct Scsi_Host *shost, char *data, int count)
> +{
> + struct scsi_qla_host *ha = to_qla_host(shost);
> + int rval = 0;
> + struct iscsi_net_param *net_param = NULL;
> + struct addr_ctrl_blk *init_fw_cb = NULL;
> + dma_addr_t init_fw_cb_dma;
> +
> + struct addr_ctrl_blk *acb = NULL;
> + dma_addr_t acb_dma;
> + uint32_t ddb_state;
> + struct ddb_entry *ddb_entry, *ddbtemp;
> + unsigned long wait_time;
> +
> + uint32_t mbox_cmd[MBOX_REG_COUNT];
> + uint32_t mbox_sts[MBOX_REG_COUNT];
> + uint32_t total_param_count;
> + uint32_t length;
> +
> + init_fw_cb = dma_alloc_coherent(&ha->pdev->dev,
> + sizeof(struct addr_ctrl_blk),&init_fw_cb_dma, GFP_KERNEL);
> + if (!init_fw_cb) {
> + printk(KERN_ERR
> + "scsi(%ld): %s: Unable to alloc init_cb\n",
> + ha->host_no, __func__);
ql4_printk instead of printk. Check in other parts of patches too.
And remember to remove debugging printks or make what looking debug
printks more informative and use ql4_printk.
When we set the net settings we reset the port, right? What is the max
length of time that could take? Like if you set DHCP, do we wait for
that to timeout and what is longest we wait?
It was orignally designed to be singled threaded non blocking, so you
are stopping everything here.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-02 18:34 ` [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs vikas.chaudhary
@ 2011-04-13 4:23 ` Mike Christie
2011-04-13 4:40 ` Mike Christie
2011-04-21 22:10 ` Jayamohan.Kallickal
2011-04-13 17:00 ` Mike Christie
2011-04-19 2:41 ` Mike Christie
2 siblings, 2 replies; 24+ messages in thread
From: Mike Christie @ 2011-04-13 4:23 UTC (permalink / raw)
To: vikas.chaudhary
Cc: James.Bottomley, linux-scsi, open-iscsi, lalit.chandivade,
ravi.anand, Jayamohan.Kallickal
ccing Jay from Emulex to answer question below about lots of ips in his
card.
On 04/02/2011 01:34 PM, vikas.chaudhary@qlogic.com wrote:
> From: Vikas Chaudhary<vikas.chaudhary@qlogic.com>
>
> To support multiple network addresses per adapter need to have a new way to
> represent network interface (net iface) in sysfs.
>
> Currently only one ipaddress and hwaddress is displayed
>
> \# ls /sys/class/iscsi_host/host18
> device hwaddress initiatorname ipaddress power subsystem uevent
>
> In this patch the net iface is presented as a separate class device.
> The one that can be added/removed dynamically or statically, based on how
> the user configures the multiple net iface on the adapter.
>
> The new sysfs directory would look like this
> \# /sys/class/iscsi_iface/
> |
> |- ipv4-iface-<host_no>-<iface_no>/<-- for ipv4
> |- ipaddress
> |- subnet
> |- gateway
> |- bootproto
> |- state
> |- ipv6-iface-<host_no>-<iface_no>/<-- for ipv6
> |- ipaddress
> |- link_local_addr
> |- router_addr
> |- ipaddr_autocfg
> |- linklocal_autocfg
> |- state
>
My only issue with this is that in userpsace we have the iscsi_iface
abstraction and that handles hw, software, scsi, and net abstractions.
This patch adds a iscsi_iface abstraction that seems more related to
just network interfaces.
For example with the userspace iscsi_iface abstraction we can create
virtual iscsi initiator ports by having each iface set its own initiator
name (or we could also set the ISID or both but we do not allow that
right now). And then for the same host you can also define userspace
iscsi ifaces with different net settings like ip addresses, and you can
go crazy and define those with different iscsi initiator port settings
if you wanted to too.
With the kernel iscsi iface it seems we are just supporting net settings.
Do we want to make the kernel iscsi iface more generic or maybe just
rename it to reflect it is just based on net settings?
The issue with making it more generic is that with the current drivers,
we do not care what info like the initiatorname in the kernel code. That
is all handled in userspace because all the session management is there,
so there is no point in adding that complexity here. But will Qlogic
need this? Will we have to tell the firmware to create a virtual iscsi
port, then pass it the initiator name or part of the isid to use? If so
then maybe it would be good to have a kernel interface in place to
set/get those values on per initiator port basis.
I think you want to link the iscsi_session to the iface it is being
accessed through, right?
Jay, for be2iscsi if you support multiple ips per host, can you end up
with a case where a session is running through IP1, but then something
happens to that, and you end up running through IP2? Or do the
interfaces for the different IPs not know anything about each other
(there is not common route table)?
> +struct iscsi_iface *
> +iscsi_create_iface(struct Scsi_Host *shost, struct iscsi_transport *transport,
> + uint32_t iface_type, uint32_t iface_num, int dd_size)
> +{
> + struct iscsi_iface *iface;
> + int id;
> + int err;
> +
> + iface = kzalloc(sizeof(*iface) + dd_size, GFP_KERNEL);
> + if (!iface)
> + return NULL;
> +
> +iface_idr_again:
> + if (!idr_pre_get(&iscsi_iface_idr, GFP_KERNEL))
> + goto free_iface;
> +
> + spin_lock(&iscsi_iface_lock);
> + err = idr_get_new(&iscsi_iface_idr, iface,&id);
> + if (err == -EAGAIN) {
> + spin_unlock(&iscsi_iface_lock);
> + goto iface_idr_again;
> + }
> + spin_unlock(&iscsi_iface_lock);
> +
> + if (err)
> + goto free_iface;
> +
> + iface->id = id;
It looks like we can just kill id, because it is never used.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-13 4:23 ` Mike Christie
@ 2011-04-13 4:40 ` Mike Christie
2011-04-13 14:48 ` Vikas Chaudhary
2011-04-21 22:10 ` Jayamohan.Kallickal
1 sibling, 1 reply; 24+ messages in thread
From: Mike Christie @ 2011-04-13 4:40 UTC (permalink / raw)
To: vikas.chaudhary
Cc: James.Bottomley, linux-scsi, open-iscsi, lalit.chandivade,
ravi.anand, Jayamohan.Kallickal
On 04/12/2011 11:23 PM, Mike Christie wrote:
> Do we want to make the kernel iscsi iface more generic or maybe just
> rename it to reflect it is just based on net settings?
>
> The issue with making it more generic is that with the current drivers,
> we do not care what info like the initiatorname in the kernel code. That
> is all handled in userspace because all the session management is there,
> so there is no point in adding that complexity here.
I take part of this back. We set the iscsi iface name and initiator name
on the session right now. We need that info for relogin type of
situations where iscsid might have died and we need to know the
initiatorname to use. We cannot read it from a disk, because that disk
might be the one are are relogging into.
It might not hurt to make the kernel's iscsi ifaces abstraction more
general.
^ permalink raw reply [flat|nested] 24+ messages in thread
* RE: [RFC-V2 PATCH 2/5] qla4xxx: add support for set_net_config
[not found] ` <4DA51ECA.1020507-hcNo3dDEHLuVc3sceRu5cw@public.gmane.org>
@ 2011-04-13 14:40 ` Vikas Chaudhary
0 siblings, 0 replies; 24+ messages in thread
From: Vikas Chaudhary @ 2011-04-13 14:40 UTC (permalink / raw)
To: open-iscsi-/JYPxA39Uh5TLH3MbocFFw, Mike Christie
Cc: James.Bottomley-l3A5Bk7waGM, linux-scsi-u79uwXL29TY76Z2rM5mHXA,
Lalit Chandivade, Ravi Anand, Harish Zunjarrao
>-----Original Message-----
>From: open-iscsi-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org [mailto:open-iscsi-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org] On
>Behalf Of Mike Christie
>Sent: Wednesday, April 13, 2011 9:26 AM
>To: Vikas Chaudhary
>Cc: James.Bottomley-l3A5Bk7waGM@public.gmane.org; linux-scsi-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; open-
>iscsi-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org; Lalit Chandivade; Ravi Anand; Harish Zunjarrao
>Subject: Re: [RFC-V2 PATCH 2/5] qla4xxx: add support for set_net_config
>
>On 04/02/2011 01:34 PM, vikas.chaudhary-h88ZbnxC6KDQT0dZR+AlfA@public.gmane.org wrote:
>>
>> +/* Only for primary ACB, secondary ACB not supported
>> + * iface_num = 0: valid -> linklocal, routable, router, autocfg options
>> + * iface_num = 1: valid -> routable
>> + */
>> +static inline void qla4xxx_set_ipv6(struct iscsi_net_param *net_param,
>> + struct addr_ctrl_blk *init_fw_cb)
>> +{
>> + switch (net_param->param_type) {
>> + case ISCSI_NET_PARAM_IPV6_ADDR:
>> + if (net_param->iface_num& 0x1) {
>
>Is this my mailer?
>
>I think you want to check for iface_num == 1. Above in the function
>comments and in other places you set the iface_num as a int, but in this
>function you check it more like a bitmask.
>
We want to check even interface here.
I will update code to make iface_num data type to uint8_t.
>
>> + case ISCSI_NET_PARAM_IPV6_LINKLOCAL:
>> + if (net_param->iface_num& 0x1)
>
>> + case ISCSI_NET_PARAM_IPV6_ROUTER:
>> + if (net_param->iface_num& 0x1)
>
>
>> + case ISCSI_NET_PARAM_IPV6_ADDR_AUTOCFG:
>> + /* Autocfg applies to even interface */
>> + if (net_param->iface_num& 0x1)
>
>However, in other places you have that comment about even applying to
>autocfg interfaces.
QLogic adapter supports two IPv6 addresses per port and both address
can be either autocfg or user specified. However autocfg setting is
common for both the addresses, that's reason we update autocfg from
even interface only.
Also both IPv6 addresses will have same linklocal addr and same
router address.
Transport code is written in way that if someone wants to support individual
configuration (autocfg, linklocal addr, etc) for each iface, they can do it.
>
>Why does it say even but you only support 2 ipv6 ifaces in the patches?
>Can you support more? Are you coding it in preparation for more?
>
We are supporting only two IPv6 addresses in our driver for now. In future
we may support more. Transport code is scalable to support up to 10 iface per port.
>
>.....
>
>
>
>> +
>> +static int
>> +qla4xxx_set_net_config(struct Scsi_Host *shost, char *data, int count)
>> +{
>> + struct scsi_qla_host *ha = to_qla_host(shost);
>> + int rval = 0;
>> + struct iscsi_net_param *net_param = NULL;
>> + struct addr_ctrl_blk *init_fw_cb = NULL;
>> + dma_addr_t init_fw_cb_dma;
>> +
>> + struct addr_ctrl_blk *acb = NULL;
>> + dma_addr_t acb_dma;
>> + uint32_t ddb_state;
>> + struct ddb_entry *ddb_entry, *ddbtemp;
>> + unsigned long wait_time;
>> +
>> + uint32_t mbox_cmd[MBOX_REG_COUNT];
>> + uint32_t mbox_sts[MBOX_REG_COUNT];
>> + uint32_t total_param_count;
>> + uint32_t length;
>> +
>> + init_fw_cb = dma_alloc_coherent(&ha->pdev->dev,
>> + sizeof(struct addr_ctrl_blk),&init_fw_cb_dma, GFP_KERNEL);
>> + if (!init_fw_cb) {
>> + printk(KERN_ERR
>> + "scsi(%ld): %s: Unable to alloc init_cb\n",
>> + ha->host_no, __func__);
>
>ql4_printk instead of printk. Check in other parts of patches too.
>
>And remember to remove debugging printks or make what looking debug
>printks more informative and use ql4_printk.
>
Ok.
>
>When we set the net settings we reset the port, right? What is the max
>length of time that could take? Like if you set DHCP, do we wait for
>that to timeout and what is longest we wait?
We do not need to reset the adapter, we just disable and enable FW
Settings (Address control block) which will apply new network
configuration to the port.
But before applying new network configuration we need to logout
all active session so we are waiting up to 10sec for each session
to logout.
>
>It was orignally designed to be singled threaded non blocking, so you
>are stopping everything here.
>
We are waiting for session to logout as I mentioned above.
This message and any attached documents contain information from QLogic Corporation or its wholly-owned subsidiaries that may be confidential. If you are not the intended recipient, you may not read, copy, distribute, or use this information. If you have received this transmission in error, please notify the sender immediately by reply e-mail and then delete this message.
--
You received this message because you are subscribed to the Google Groups "open-iscsi" group.
To post to this group, send email to open-iscsi-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
To unsubscribe from this group, send email to open-iscsi+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/open-iscsi?hl=en.
^ permalink raw reply [flat|nested] 24+ messages in thread
* RE: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-13 4:40 ` Mike Christie
@ 2011-04-13 14:48 ` Vikas Chaudhary
[not found] ` <5E4F49720D0BAD499EE1F01232234BA8728A8CB623-HolNjIBXvBOXx9kJd3VG2h2eb7JE58TQ@public.gmane.org>
0 siblings, 1 reply; 24+ messages in thread
From: Vikas Chaudhary @ 2011-04-13 14:48 UTC (permalink / raw)
To: Mike Christie
Cc: James.Bottomley, linux-scsi, open-iscsi, Lalit Chandivade,
Ravi Anand, Jayamohan.Kallickal
>-----Original Message-----
>From: Mike Christie [mailto:michaelc@cs.wisc.edu]
>Sent: Wednesday, April 13, 2011 10:11 AM
>To: Vikas Chaudhary
>Cc: James.Bottomley@suse.de; linux-scsi@vger.kernel.org; open-
>iscsi@googlegroups.com; Lalit Chandivade; Ravi Anand;
>Jayamohan.Kallickal@Emulex.Com
>Subject: Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration
>in sysfs
>
>On 04/12/2011 11:23 PM, Mike Christie wrote:
>> Do we want to make the kernel iscsi iface more generic or maybe just
>> rename it to reflect it is just based on net settings?
>>
>> The issue with making it more generic is that with the current drivers,
>> we do not care what info like the initiatorname in the kernel code. That
>> is all handled in userspace because all the session management is there,
>> so there is no point in adding that complexity here.
>
>I take part of this back. We set the iscsi iface name and initiator name
>on the session right now. We need that info for relogin type of
>situations where iscsid might have died and we need to know the
>initiatorname to use. We cannot read it from a disk, because that disk
>might be the one are are relogging into.
>
>It might not hurt to make the kernel's iscsi ifaces abstraction more
>general.
Yes and also to reflect correct status of network settings it will be
good to have iscsi_iface in the kernel. This is useful in a case where
user set DHCP settings then we can reflect IP addresses and other
settings in the kernel iface.
This message and any attached documents contain information from QLogic Corporation or its wholly-owned subsidiaries that may be confidential. If you are not the intended recipient, you may not read, copy, distribute, or use this information. If you have received this transmission in error, please notify the sender immediately by reply e-mail and then delete this message.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
[not found] ` <5E4F49720D0BAD499EE1F01232234BA8728A8CB623-HolNjIBXvBOXx9kJd3VG2h2eb7JE58TQ@public.gmane.org>
@ 2011-04-13 16:44 ` Mike Christie
2011-04-14 4:15 ` Vikas Chaudhary
0 siblings, 1 reply; 24+ messages in thread
From: Mike Christie @ 2011-04-13 16:44 UTC (permalink / raw)
To: open-iscsi-/JYPxA39Uh5TLH3MbocFFw
Cc: Vikas Chaudhary, James.Bottomley-l3A5Bk7waGM,
linux-scsi-u79uwXL29TY76Z2rM5mHXA, Lalit Chandivade, Ravi Anand,
Jayamohan.Kallickal-iH1Dq9VlAzfQT0dZR+AlfA
On 04/13/2011 09:48 AM, Vikas Chaudhary wrote:
>
>
>> -----Original Message-----
>> From: Mike Christie [mailto:michaelc-hcNo3dDEHLuVc3sceRu5cw@public.gmane.org]
>> Sent: Wednesday, April 13, 2011 10:11 AM
>> To: Vikas Chaudhary
>> Cc: James.Bottomley-l3A5Bk7waGM@public.gmane.org; linux-scsi-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; open-
>> iscsi-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org; Lalit Chandivade; Ravi Anand;
>> Jayamohan.Kallickal-iH1Dq9VlAzfQT0dZR+AlfA@public.gmane.org
>> Subject: Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration
>> in sysfs
>>
>> On 04/12/2011 11:23 PM, Mike Christie wrote:
>>> Do we want to make the kernel iscsi iface more generic or maybe just
>>> rename it to reflect it is just based on net settings?
>>>
>>> The issue with making it more generic is that with the current drivers,
>>> we do not care what info like the initiatorname in the kernel code. That
>>> is all handled in userspace because all the session management is there,
>>> so there is no point in adding that complexity here.
>>
>> I take part of this back. We set the iscsi iface name and initiator name
>> on the session right now. We need that info for relogin type of
>> situations where iscsid might have died and we need to know the
>> initiatorname to use. We cannot read it from a disk, because that disk
>> might be the one are are relogging into.
>>
>> It might not hurt to make the kernel's iscsi ifaces abstraction more
>> general.
>
> Yes and also to reflect correct status of network settings it will be
> good to have iscsi_iface in the kernel. This is useful in a case where
> user set DHCP settings then we can reflect IP addresses and other
> settings in the kernel iface.
>
How come you did not export some values like the vlan in the iscsi iface?
--
You received this message because you are subscribed to the Google Groups "open-iscsi" group.
To post to this group, send email to open-iscsi-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
To unsubscribe from this group, send email to open-iscsi+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
For more options, visit this group at http://groups.google.com/group/open-iscsi?hl=en.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-02 18:34 ` [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs vikas.chaudhary
2011-04-13 4:23 ` Mike Christie
@ 2011-04-13 17:00 ` Mike Christie
2011-04-13 22:47 ` Shyam_Iyer
2011-04-13 22:53 ` Michael Chan
2011-04-19 2:41 ` Mike Christie
2 siblings, 2 replies; 24+ messages in thread
From: Mike Christie @ 2011-04-13 17:00 UTC (permalink / raw)
To: vikas.chaudhary
Cc: James.Bottomley, linux-scsi, open-iscsi, lalit.chandivade,
ravi.anand, Eddie Wai, Michael Chan
ccing Broadcom devs, Some questions for you guys below
On 04/02/2011 01:34 PM, vikas.chaudhary@qlogic.com wrote:
> From: Vikas Chaudhary<vikas.chaudhary@qlogic.com>
>
> To support multiple network addresses per adapter need to have a new way to
> represent network interface (net iface) in sysfs.
>
> Currently only one ipaddress and hwaddress is displayed
>
> \# ls /sys/class/iscsi_host/host18
> device hwaddress initiatorname ipaddress power subsystem uevent
>
> In this patch the net iface is presented as a separate class device.
> The one that can be added/removed dynamically or statically, based on how
> the user configures the multiple net iface on the adapter.
>
> The new sysfs directory would look like this
> \# /sys/class/iscsi_iface/
> |
> |- ipv4-iface-<host_no>-<iface_no>/<-- for ipv4
> |- ipaddress
> |- subnet
> |- gateway
> |- bootproto
> |- state
> |- ipv6-iface-<host_no>-<iface_no>/<-- for ipv6
> |- ipaddress
> |- link_local_addr
> |- router_addr
> |- ipaddr_autocfg
> |- linklocal_autocfg
> |- state
>
With patch "[RFC-V2 PATCH 1/5] iscsi_transport: add support for
set_net_config" userspace would send down the vlan info.
If we add a vlan sysfs file to the iscsi_iface, to export the info was
bnx2i going call iscsi_create_iface for each vlan? If so I am not sure
what bnx2i will use for the iface_num. It is supposed to be persistent,
right?
For bnx2i, when doing iscsi offload and vlans, do you have to have a
netdev like ethX.Y setup for each vlan or can bnx2i operate without it
(the call to cnic_get_vlan always throws me and I cannot remember if we
were going to still do that or change something in the driver so you did
not need it).
And just to confirm for vlans and bnx2i, when we make a ep and session,
cnic_get_route/cnic_cm_select_dev will do the magic to figure out what
vlan to use?
> +
> +struct iscsi_iface *
> +iscsi_create_iface(struct Scsi_Host *shost, struct iscsi_transport *transport,
> + uint32_t iface_type, uint32_t iface_num, int dd_size)
> +{
> + struct iscsi_iface *iface;
> + int id;
> + int err;
> +
> + iface = kzalloc(sizeof(*iface) + dd_size, GFP_KERNEL);
> + if (!iface)
> + return NULL;
> +
> +iface_idr_again:
> + if (!idr_pre_get(&iscsi_iface_idr, GFP_KERNEL))
> + goto free_iface;
> +
> + spin_lock(&iscsi_iface_lock);
> + err = idr_get_new(&iscsi_iface_idr, iface,&id);
> + if (err == -EAGAIN) {
> + spin_unlock(&iscsi_iface_lock);
> + goto iface_idr_again;
> + }
> + spin_unlock(&iscsi_iface_lock);
> +
> + if (err)
> + goto free_iface;
> +
> + iface->id = id;
> + iface->transport = transport;
> + iface->iface_type = iface_type;
> + iface->iface_num = iface_num;
> + iface->dev.class =&iscsi_iface_class;
> + /* parent reference */
> + iface->dev.parent = get_device(&shost->shost_gendev);
> + if (iface_type == IFACE_TYPE_IPV4)
> + dev_set_name(&iface->dev, "ipv4-iface-%u-%u", shost->host_no,
> + iface_num);
> + else if (iface_type == IFACE_TYPE_IPV6)
> + dev_set_name(&iface->dev, "ipv6-iface-%u-%u", shost->host_no,
> + iface_num);
> + else
> + goto free_iface;
> +
> + err = device_register(&iface->dev);
> + if (err)
> + goto free_iface;
> +
> + if (iface_type == IFACE_TYPE_IPV4)
> + err = sysfs_create_group(&iface->dev.kobj,
> + &iscsi_ipv4_iface_group);
> + else if (iface_type == IFACE_TYPE_IPV6)
> + err = sysfs_create_group(&iface->dev.kobj,
> + &iscsi_ipv6_iface_group);
> +
> + if (err)
> + goto iface_unregister_dev;
> +
> + if (dd_size)
> + iface->dd_data =&iface[1];
> + return iface;
> +
> +iface_unregister_dev:
> + idr_remove(&iscsi_iface_idr, id);
> + device_unregister(&iface->dev);
> +
> +free_iface:
> + kfree(iface);
> + return NULL;
> +}
> +EXPORT_SYMBOL_GPL(iscsi_create_iface);
^ permalink raw reply [flat|nested] 24+ messages in thread
* RE: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-13 17:00 ` Mike Christie
@ 2011-04-13 22:47 ` Shyam_Iyer
2011-04-14 3:07 ` Mike Christie
2011-04-13 22:53 ` Michael Chan
1 sibling, 1 reply; 24+ messages in thread
From: Shyam_Iyer @ 2011-04-13 22:47 UTC (permalink / raw)
To: open-iscsi, vikas.chaudhary
Cc: James.Bottomley, linux-scsi, lalit.chandivade, ravi.anand, waie, mchan
> -----Original Message-----
> From: open-iscsi@googlegroups.com [mailto:open-iscsi@googlegroups.com]
> On Behalf Of Mike Christie
<snip>
> ccing Broadcom devs, Some questions for you guys below
>
>
> On 04/02/2011 01:34 PM, vikas.chaudhary@qlogic.com wrote:
> > From: Vikas Chaudhary<vikas.chaudhary@qlogic.com>
> >
> > To support multiple network addresses per adapter need to have a new
> way to
> > represent network interface (net iface) in sysfs.
> >
> > Currently only one ipaddress and hwaddress is displayed
> >
> > \# ls /sys/class/iscsi_host/host18
> > device hwaddress initiatorname ipaddress power subsystem uevent
> >
> > In this patch the net iface is presented as a separate class device.
> > The one that can be added/removed dynamically or statically, based on
> how
> > the user configures the multiple net iface on the adapter.
> >
> > The new sysfs directory would look like this
> > \# /sys/class/iscsi_iface/
> > |
> > |- ipv4-iface-<host_no>-<iface_no>/<-- for ipv4
> > |- ipaddress
> > |- subnet
> > |- gateway
> > |- bootproto
> > |- state
> > |- ipv6-iface-<host_no>-<iface_no>/<-- for ipv6
> > |- ipaddress
> > |- link_local_addr
> > |- router_addr
> > |- ipaddr_autocfg
> > |- linklocal_autocfg
> > |- state
> >
>
>
> With patch "[RFC-V2 PATCH 1/5] iscsi_transport: add support for
> set_net_config" userspace would send down the vlan info.
How can this converged adapter separate vlans created via ip on the network function from vlans created via iSCSI function vs vlan created on the FCoE function?
>
> If we add a vlan sysfs file to the iscsi_iface, to export the info was
> bnx2i going call iscsi_create_iface for each vlan? If so I am not sure
> what bnx2i will use for the iface_num. It is supposed to be persistent,
> right?
>
Would iscsiadm be used to create the vlan.. This would be introducing another way to create vlans..
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-13 17:00 ` Mike Christie
2011-04-13 22:47 ` Shyam_Iyer
@ 2011-04-13 22:53 ` Michael Chan
2011-04-14 3:24 ` Mike Christie
1 sibling, 1 reply; 24+ messages in thread
From: Michael Chan @ 2011-04-13 22:53 UTC (permalink / raw)
To: Mike Christie
Cc: vikas.chaudhary, James.Bottomley, linux-scsi, open-iscsi,
lalit.chandivade, ravi.anand, Eddie Wai
On Wed, 2011-04-13 at 10:00 -0700, Mike Christie wrote:
> ccing Broadcom devs, Some questions for you guys below
>
>
> On 04/02/2011 01:34 PM, vikas.chaudhary@qlogic.com wrote:
> > From: Vikas Chaudhary<vikas.chaudhary@qlogic.com>
> >
> > To support multiple network addresses per adapter need to have a new way to
> > represent network interface (net iface) in sysfs.
> >
> > Currently only one ipaddress and hwaddress is displayed
> >
> > \# ls /sys/class/iscsi_host/host18
> > device hwaddress initiatorname ipaddress power subsystem uevent
> >
> > In this patch the net iface is presented as a separate class device.
> > The one that can be added/removed dynamically or statically, based on how
> > the user configures the multiple net iface on the adapter.
> >
> > The new sysfs directory would look like this
> > \# /sys/class/iscsi_iface/
> > |
> > |- ipv4-iface-<host_no>-<iface_no>/<-- for ipv4
> > |- ipaddress
> > |- subnet
> > |- gateway
> > |- bootproto
> > |- state
> > |- ipv6-iface-<host_no>-<iface_no>/<-- for ipv6
> > |- ipaddress
> > |- link_local_addr
> > |- router_addr
> > |- ipaddr_autocfg
> > |- linklocal_autocfg
> > |- state
> >
>
>
> With patch "[RFC-V2 PATCH 1/5] iscsi_transport: add support for
> set_net_config" userspace would send down the vlan info.
>
> If we add a vlan sysfs file to the iscsi_iface, to export the info was
> bnx2i going call iscsi_create_iface for each vlan? If so I am not sure
> what bnx2i will use for the iface_num. It is supposed to be persistent,
> right?
bnx2i doesn't need to know the IP addresses and VLANs for each HBA. It
is all handled in userspace by uIP. So if we add this sysfs,
iscsi_transport can handle sysfs creation for devices generically,
right?
>
> For bnx2i, when doing iscsi offload and vlans, do you have to have a
> netdev like ethX.Y setup for each vlan or can bnx2i operate without it
> (the call to cnic_get_vlan always throws me and I cannot remember if we
> were going to still do that or change something in the driver so you did
> not need it).
For now, cnic_get_vlan will try to get a matching VLAN from netdev with
a route to the target. But it is not required. Once we have VLAN
attribute in iface, uIP can override and use the VLAN specified in
iface.
>
> And just to confirm for vlans and bnx2i, when we make a ep and session,
> cnic_get_route/cnic_cm_select_dev will do the magic to figure out what
> vlan to use?
Yes for now. The plan is for uIP to override if a VLAN is specified in
iface.
This is the flow:
-> ep_connect
-> cnic_cm_connect
/* Try to get route/MTU/VLAN from netdev if available */
-> cnic_get_route
-> send nl message to uIP to get next hop
-> recv nl message from uIP
/* at this point we may get a new VLAN ID from uIP */
-> make connection to target
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-13 22:47 ` Shyam_Iyer
@ 2011-04-14 3:07 ` Mike Christie
0 siblings, 0 replies; 24+ messages in thread
From: Mike Christie @ 2011-04-14 3:07 UTC (permalink / raw)
To: Shyam_Iyer
Cc: open-iscsi, vikas.chaudhary, James.Bottomley, linux-scsi,
lalit.chandivade, ravi.anand, waie, mchan
On 04/13/2011 05:47 PM, Shyam_Iyer@Dell.com wrote:
>>
>> If we add a vlan sysfs file to the iscsi_iface, to export the info was
>> bnx2i going call iscsi_create_iface for each vlan? If so I am not sure
>> what bnx2i will use for the iface_num. It is supposed to be persistent,
>> right?
>>
>
> Would iscsiadm be used to create the vlan.. This would be introducing another way to create vlans..
Yes, iscsiadm. We already manage the other net settings for these cards
with iscsiadm. Originally, we wanted to just use the normal net tools,
but that was not allowed, so we do not have very many options.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-13 22:53 ` Michael Chan
@ 2011-04-14 3:24 ` Mike Christie
0 siblings, 0 replies; 24+ messages in thread
From: Mike Christie @ 2011-04-14 3:24 UTC (permalink / raw)
To: Michael Chan
Cc: vikas.chaudhary, James.Bottomley, linux-scsi, open-iscsi,
lalit.chandivade, ravi.anand, Eddie Wai
On 04/13/2011 05:53 PM, Michael Chan wrote:
>
> On Wed, 2011-04-13 at 10:00 -0700, Mike Christie wrote:
>> ccing Broadcom devs, Some questions for you guys below
>>
>>
>> On 04/02/2011 01:34 PM, vikas.chaudhary@qlogic.com wrote:
>>> From: Vikas Chaudhary<vikas.chaudhary@qlogic.com>
>>>
>>> To support multiple network addresses per adapter need to have a new way to
>>> represent network interface (net iface) in sysfs.
>>>
>>> Currently only one ipaddress and hwaddress is displayed
>>>
>>> \# ls /sys/class/iscsi_host/host18
>>> device hwaddress initiatorname ipaddress power subsystem uevent
>>>
>>> In this patch the net iface is presented as a separate class device.
>>> The one that can be added/removed dynamically or statically, based on how
>>> the user configures the multiple net iface on the adapter.
>>>
>>> The new sysfs directory would look like this
>>> \# /sys/class/iscsi_iface/
>>> |
>>> |- ipv4-iface-<host_no>-<iface_no>/<-- for ipv4
>>> |- ipaddress
>>> |- subnet
>>> |- gateway
>>> |- bootproto
>>> |- state
>>> |- ipv6-iface-<host_no>-<iface_no>/<-- for ipv6
>>> |- ipaddress
>>> |- link_local_addr
>>> |- router_addr
>>> |- ipaddr_autocfg
>>> |- linklocal_autocfg
>>> |- state
>>>
>>
>>
>> With patch "[RFC-V2 PATCH 1/5] iscsi_transport: add support for
>> set_net_config" userspace would send down the vlan info.
>>
>> If we add a vlan sysfs file to the iscsi_iface, to export the info was
>> bnx2i going call iscsi_create_iface for each vlan? If so I am not sure
>> what bnx2i will use for the iface_num. It is supposed to be persistent,
>> right?
>
> bnx2i doesn't need to know the IP addresses and VLANs for each HBA. It
> is all handled in userspace by uIP. So if we add this sysfs,
> iscsi_transport can handle sysfs creation for devices generically,
> right?
With these patches the driver need to call iface_create_iface and
implement a callout on the iscsi_transport template so
scsi_transport_iscsi.c can call into the driver to get the values to
display in sysfs like is done for the connection and session settings.
We can add some helpers in libiscsi for this like is done for the
conn/session stuff.
>
>>
>> For bnx2i, when doing iscsi offload and vlans, do you have to have a
>> netdev like ethX.Y setup for each vlan or can bnx2i operate without it
>> (the call to cnic_get_vlan always throws me and I cannot remember if we
>> were going to still do that or change something in the driver so you did
>> not need it).
>
> For now, cnic_get_vlan will try to get a matching VLAN from netdev with
> a route to the target. But it is not required. Once we have VLAN
> attribute in iface, uIP can override and use the VLAN specified in
> iface.
>
Ok.
>
>>
>> And just to confirm for vlans and bnx2i, when we make a ep and session,
>> cnic_get_route/cnic_cm_select_dev will do the magic to figure out what
>> vlan to use?
>
> Yes for now. The plan is for uIP to override if a VLAN is specified in
> iface.
Is the idea to have uIP read the userspace iface config files in
/etc/iscsi/iface or /var/lib/iscsi/ifaces or iscsid (basically something
in userspace) then pass down in a msg_set_path msg?
>
> This is the flow:
>
> -> ep_connect
>
> -> cnic_cm_connect
> /* Try to get route/MTU/VLAN from netdev if available */
> -> cnic_get_route
> -> send nl message to uIP to get next hop
> -> recv nl message from uIP
> /* at this point we may get a new VLAN ID from uIP */
> -> make connection to target
>
>
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 24+ messages in thread
* RE: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-13 16:44 ` Mike Christie
@ 2011-04-14 4:15 ` Vikas Chaudhary
2011-04-14 5:18 ` Mike Christie
0 siblings, 1 reply; 24+ messages in thread
From: Vikas Chaudhary @ 2011-04-14 4:15 UTC (permalink / raw)
To: open-iscsi, Mike Christie
Cc: James.Bottomley, linux-scsi, Lalit Chandivade, Ravi Anand,
Jayamohan.Kallickal
>-----Original Message-----
>From: open-iscsi@googlegroups.com [mailto:open-iscsi@googlegroups.com] On
>Behalf Of Mike Christie
>Sent: Wednesday, April 13, 2011 10:14 PM
>To: open-iscsi@googlegroups.com
>Cc: Vikas Chaudhary; James.Bottomley@suse.de; linux-scsi@vger.kernel.org;
>Lalit Chandivade; Ravi Anand; Jayamohan.Kallickal@Emulex.Com
>Subject: Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration
>in sysfs
>
>On 04/13/2011 09:48 AM, Vikas Chaudhary wrote:
>>
>>
>>> -----Original Message-----
>>> From: Mike Christie [mailto:michaelc@cs.wisc.edu]
>>> Sent: Wednesday, April 13, 2011 10:11 AM
>>> To: Vikas Chaudhary
>>> Cc: James.Bottomley@suse.de; linux-scsi@vger.kernel.org; open-
>>> iscsi@googlegroups.com; Lalit Chandivade; Ravi Anand;
>>> Jayamohan.Kallickal@Emulex.Com
>>> Subject: Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network
>configuration
>>> in sysfs
>>>
>>> On 04/12/2011 11:23 PM, Mike Christie wrote:
>>>> Do we want to make the kernel iscsi iface more generic or maybe just
>>>> rename it to reflect it is just based on net settings?
>>>>
>>>> The issue with making it more generic is that with the current drivers,
>>>> we do not care what info like the initiatorname in the kernel code.
>That
>>>> is all handled in userspace because all the session management is
>there,
>>>> so there is no point in adding that complexity here.
>>>
>>> I take part of this back. We set the iscsi iface name and initiator name
>>> on the session right now. We need that info for relogin type of
>>> situations where iscsid might have died and we need to know the
>>> initiatorname to use. We cannot read it from a disk, because that disk
>>> might be the one are are relogging into.
>>>
>>> It might not hurt to make the kernel's iscsi ifaces abstraction more
>>> general.
>>
>> Yes and also to reflect correct status of network settings it will be
>> good to have iscsi_iface in the kernel. This is useful in a case where
>> user set DHCP settings then we can reflect IP addresses and other
>> settings in the kernel iface.
>>
>
>How come you did not export some values like the vlan in the iscsi iface?
>
Currently we have partial implementation of VLAN. iscsi_transport supports
VLAN set operation but we need to implement this in iscsiadm and export
VLAN to sysfs iface. We will add this support in our next patches.
This message and any attached documents contain information from QLogic Corporation or its wholly-owned subsidiaries that may be confidential. If you are not the intended recipient, you may not read, copy, distribute, or use this information. If you have received this transmission in error, please notify the sender immediately by reply e-mail and then delete this message.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-14 4:15 ` Vikas Chaudhary
@ 2011-04-14 5:18 ` Mike Christie
2011-04-14 5:54 ` Vikas Chaudhary
0 siblings, 1 reply; 24+ messages in thread
From: Mike Christie @ 2011-04-14 5:18 UTC (permalink / raw)
To: open-iscsi
Cc: Vikas Chaudhary, James.Bottomley, linux-scsi, Lalit Chandivade,
Ravi Anand, Jayamohan.Kallickal
On 04/13/2011 11:15 PM, Vikas Chaudhary wrote:
>
> Currently we have partial implementation of VLAN. iscsi_transport supports
> VLAN set operation but we need to implement this in iscsiadm and export
> VLAN to sysfs iface. We will add this support in our next patches.
>
Basically, I just want to know if vlan stuff is going into the
iscsi_iface class or in some new struct?
For the latter would it be something like:
host
|
\-iscsi_iface0
|
\-iscsi_vlan0
|
\-iscsi_vlan1
|
\-iscsi_vlan2
or
host
|
\-iscsi_iface1
|-vlan0
\-iscsi_iface2
|-vlan1
\-iscsi_iface3
|-vlan2
^ permalink raw reply [flat|nested] 24+ messages in thread
* RE: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-14 5:18 ` Mike Christie
@ 2011-04-14 5:54 ` Vikas Chaudhary
0 siblings, 0 replies; 24+ messages in thread
From: Vikas Chaudhary @ 2011-04-14 5:54 UTC (permalink / raw)
To: Mike Christie, open-iscsi
Cc: James.Bottomley, linux-scsi, Lalit Chandivade, Ravi Anand,
Jayamohan.Kallickal
>-----Original Message-----
>From: Mike Christie [mailto:michaelc@cs.wisc.edu]
>Sent: Thursday, April 14, 2011 10:49 AM
>To: open-iscsi@googlegroups.com
>Cc: Vikas Chaudhary; James.Bottomley@suse.de; linux-scsi@vger.kernel.org;
>Lalit Chandivade; Ravi Anand; Jayamohan.Kallickal@Emulex.Com
>Subject: Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration
>in sysfs
>
>On 04/13/2011 11:15 PM, Vikas Chaudhary wrote:
>>
>> Currently we have partial implementation of VLAN. iscsi_transport
>supports
>> VLAN set operation but we need to implement this in iscsiadm and export
>> VLAN to sysfs iface. We will add this support in our next patches.
>>
>
>
>Basically, I just want to know if vlan stuff is going into the
>iscsi_iface class or in some new struct?
>
>For the latter would it be something like:
>
>host
>|
>\-iscsi_iface0
> |
> \-iscsi_vlan0
> |
> \-iscsi_vlan1
> |
> \-iscsi_vlan2
>
>or
>
>host
>|
>\-iscsi_iface1
> |-vlan0
>\-iscsi_iface2
> |-vlan1
>\-iscsi_iface3
> |-vlan2
VLAN will be per iface. So it will be like:
host
|
\-iscsi_iface1
|-vlan0
\-iscsi_iface2
|-vlan1
\-iscsi_iface3
|-vlan2
This message and any attached documents contain information from QLogic Corporation or its wholly-owned subsidiaries that may be confidential. If you are not the intended recipient, you may not read, copy, distribute, or use this information. If you have received this transmission in error, please notify the sender immediately by reply e-mail and then delete this message.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-02 18:34 ` [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs vikas.chaudhary
2011-04-13 4:23 ` Mike Christie
2011-04-13 17:00 ` Mike Christie
@ 2011-04-19 2:41 ` Mike Christie
2011-04-19 13:47 ` Vikas Chaudhary
2 siblings, 1 reply; 24+ messages in thread
From: Mike Christie @ 2011-04-19 2:41 UTC (permalink / raw)
To: vikas.chaudhary
Cc: James.Bottomley, linux-scsi, open-iscsi, lalit.chandivade, ravi.anand
Just noticed I forgot to send some comments.
On 04/02/2011 01:34 PM, vikas.chaudhary@qlogic.com wrote:
> |
> |- ipv4-iface-<host_no>-<iface_no>/<-- for ipv4
> |- ipaddress
> |- subnet
> |- gateway
> |- bootproto
> |- state
What does state mean? I thought it was going to be something like port
up/down/coming up/ or something like that. It instead just indicates if
it is enabled or not, right?
What about
1. renaming to enabled, having it print out True of enabled and False if
not.
2. adding a new file for state that indicates the port's state.
3. rename IPOPT_IPv4_PROTOCOL_ENABLE to IPOPT_IPV4_PROTOCOL_ENABLE to
match ipv6 naming.
^ permalink raw reply [flat|nested] 24+ messages in thread
* RE: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-19 2:41 ` Mike Christie
@ 2011-04-19 13:47 ` Vikas Chaudhary
0 siblings, 0 replies; 24+ messages in thread
From: Vikas Chaudhary @ 2011-04-19 13:47 UTC (permalink / raw)
To: Mike Christie
Cc: James.Bottomley, linux-scsi, open-iscsi, Lalit Chandivade, Ravi Anand
>-----Original Message-----
>From: Mike Christie [mailto:michaelc@cs.wisc.edu]
>Sent: Tuesday, April 19, 2011 8:11 AM
>To: Vikas Chaudhary
>Cc: James.Bottomley@suse.de; linux-scsi@vger.kernel.org; open-
>iscsi@googlegroups.com; Lalit Chandivade; Ravi Anand
>Subject: Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration
>in sysfs
>
>Just noticed I forgot to send some comments.
>
>On 04/02/2011 01:34 PM, vikas.chaudhary@qlogic.com wrote:
>> |
>> |- ipv4-iface-<host_no>-<iface_no>/<-- for ipv4
>> |- ipaddress
>> |- subnet
>> |- gateway
>> |- bootproto
>> |- state
>
>What does state mean? I thought it was going to be something like port
>up/down/coming up/ or something like that. It instead just indicates if
>it is enabled or not, right?
>
Yes, state indicates ipv4/ipv6 enabled or not.
>What about
>
>1. renaming to enabled, having it print out True of enabled and False if
>not.
>2. adding a new file for state that indicates the port's state.
>3. rename IPOPT_IPv4_PROTOCOL_ENABLE to IPOPT_IPV4_PROTOCOL_ENABLE to
>match ipv6 naming.
Ok. I will take care of all three comments in final patch.
This message and any attached documents contain information from QLogic Corporation or its wholly-owned subsidiaries that may be confidential. If you are not the intended recipient, you may not read, copy, distribute, or use this information. If you have received this transmission in error, please notify the sender immediately by reply e-mail and then delete this message.
^ permalink raw reply [flat|nested] 24+ messages in thread
* RE: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
2011-04-13 4:23 ` Mike Christie
2011-04-13 4:40 ` Mike Christie
@ 2011-04-21 22:10 ` Jayamohan.Kallickal
1 sibling, 0 replies; 24+ messages in thread
From: Jayamohan.Kallickal @ 2011-04-21 22:10 UTC (permalink / raw)
To: michaelc, vikas.chaudhary
Cc: James.Bottomley, linux-scsi, open-iscsi, lalit.chandivade, ravi.anand
Pl see inline
-----Original Message-----
From: Mike Christie [mailto:michaelc@cs.wisc.edu]
Sent: Tuesday, April 12, 2011 9:23 PM
To: vikas.chaudhary@qlogic.com
Cc: James.Bottomley@suse.de; linux-scsi@vger.kernel.org; open-iscsi@googlegroups.com; lalit.chandivade@qlogic.com; ravi.anand@qlogic.com; Kallickal, Jayamohan
Subject: Re: [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs
ccing Jay from Emulex to answer question below about lots of ips in his
card.
On 04/02/2011 01:34 PM, vikas.chaudhary@qlogic.com wrote:
> From: Vikas Chaudhary<vikas.chaudhary@qlogic.com>
>
> To support multiple network addresses per adapter need to have a new way to
> represent network interface (net iface) in sysfs.
>
> Currently only one ipaddress and hwaddress is displayed
>
> \# ls /sys/class/iscsi_host/host18
> device hwaddress initiatorname ipaddress power subsystem uevent
>
> In this patch the net iface is presented as a separate class device.
> The one that can be added/removed dynamically or statically, based on how
> the user configures the multiple net iface on the adapter.
>
> The new sysfs directory would look like this
> \# /sys/class/iscsi_iface/
> |
> |- ipv4-iface-<host_no>-<iface_no>/<-- for ipv4
> |- ipaddress
> |- subnet
> |- gateway
> |- bootproto
> |- state
> |- ipv6-iface-<host_no>-<iface_no>/<-- for ipv6
> |- ipaddress
> |- link_local_addr
> |- router_addr
> |- ipaddr_autocfg
> |- linklocal_autocfg
> |- state
>
My only issue with this is that in userpsace we have the iscsi_iface
abstraction and that handles hw, software, scsi, and net abstractions.
This patch adds a iscsi_iface abstraction that seems more related to
just network interfaces.
For example with the userspace iscsi_iface abstraction we can create
virtual iscsi initiator ports by having each iface set its own initiator
name (or we could also set the ISID or both but we do not allow that
right now). And then for the same host you can also define userspace
iscsi ifaces with different net settings like ip addresses, and you can
go crazy and define those with different iscsi initiator port settings
if you wanted to too.
With the kernel iscsi iface it seems we are just supporting net settings.
Do we want to make the kernel iscsi iface more generic or maybe just
rename it to reflect it is just based on net settings?
The issue with making it more generic is that with the current drivers,
we do not care what info like the initiatorname in the kernel code. That
is all handled in userspace because all the session management is there,
so there is no point in adding that complexity here. But will Qlogic
need this? Will we have to tell the firmware to create a virtual iscsi
port, then pass it the initiator name or part of the isid to use? If so
then maybe it would be good to have a kernel interface in place to
set/get those values on per initiator port basis.
I think you want to link the iscsi_session to the iface it is being
accessed through, right?
Jay, for be2iscsi if you support multiple ips per host, can you end up
with a case where a session is running through IP1, but then something
happens to that, and you end up running through IP2? Or do the
interfaces for the different IPs not know anything about each other
(there is not common route table)?
The failover from one IP to another must be done outside the chip/driver .
We need to be able to setup multiple IP's per host and either use that for
Login or find out the best path using internal routing tables in the chip.
> +struct iscsi_iface *
> +iscsi_create_iface(struct Scsi_Host *shost, struct iscsi_transport *transport,
> + uint32_t iface_type, uint32_t iface_num, int dd_size)
> +{
> + struct iscsi_iface *iface;
> + int id;
> + int err;
> +
> + iface = kzalloc(sizeof(*iface) + dd_size, GFP_KERNEL);
> + if (!iface)
> + return NULL;
> +
> +iface_idr_again:
> + if (!idr_pre_get(&iscsi_iface_idr, GFP_KERNEL))
> + goto free_iface;
> +
> + spin_lock(&iscsi_iface_lock);
> + err = idr_get_new(&iscsi_iface_idr, iface,&id);
> + if (err == -EAGAIN) {
> + spin_unlock(&iscsi_iface_lock);
> + goto iface_idr_again;
> + }
> + spin_unlock(&iscsi_iface_lock);
> +
> + if (err)
> + goto free_iface;
> +
> + iface->id = id;
It looks like we can just kill id, because it is never used.
^ permalink raw reply [flat|nested] 24+ messages in thread
end of thread, other threads:[~2011-04-21 22:10 UTC | newest]
Thread overview: 24+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2011-04-02 18:34 [RFC-V2 PATCH 0/5] Proposal for iSCSI Network Configuration vikas.chaudhary
2011-04-02 18:34 ` [RFC-V2 PATCH 1/5] iscsi_transport: add support for set_net_config vikas.chaudhary
2011-04-13 3:41 ` Mike Christie
2011-04-02 18:34 ` [RFC-V2 PATCH 2/5] qla4xxx: " vikas.chaudhary
2011-04-13 3:55 ` Mike Christie
[not found] ` <4DA51ECA.1020507-hcNo3dDEHLuVc3sceRu5cw@public.gmane.org>
2011-04-13 14:40 ` Vikas Chaudhary
2011-04-02 18:34 ` [RFC-V2 PATCH 3/5] qla4xxx: Added new "struct ipaddress_config" vikas.chaudhary
2011-04-02 18:34 ` [RFC-V2 PATCH 4/5] iscsi_transport: show network configuration in sysfs vikas.chaudhary
2011-04-13 4:23 ` Mike Christie
2011-04-13 4:40 ` Mike Christie
2011-04-13 14:48 ` Vikas Chaudhary
[not found] ` <5E4F49720D0BAD499EE1F01232234BA8728A8CB623-HolNjIBXvBOXx9kJd3VG2h2eb7JE58TQ@public.gmane.org>
2011-04-13 16:44 ` Mike Christie
2011-04-14 4:15 ` Vikas Chaudhary
2011-04-14 5:18 ` Mike Christie
2011-04-14 5:54 ` Vikas Chaudhary
2011-04-21 22:10 ` Jayamohan.Kallickal
2011-04-13 17:00 ` Mike Christie
2011-04-13 22:47 ` Shyam_Iyer
2011-04-14 3:07 ` Mike Christie
2011-04-13 22:53 ` Michael Chan
2011-04-14 3:24 ` Mike Christie
2011-04-19 2:41 ` Mike Christie
2011-04-19 13:47 ` Vikas Chaudhary
2011-04-02 18:34 ` [RFC-V2 PATCH 5/5] qla4xxx: added support to show multiple iface " vikas.chaudhary
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.