* [PATCH 000/103] target: Updates for v3.1-rc1
@ 2011-07-21 7:04 Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 001/103] target: Check LUN numbers in transport_get_lun_for_[cmd,tmr] Nicholas A. Bellinger
` (29 more replies)
0 siblings, 30 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Nicholas Bellinger
From: Nicholas Bellinger <nab@hera.kernel.org>
Hello folks,
This is the first set of pending drivers/target/ updates for the v3.1
merge window. It contains updates for drivers/target core and
loopback + tcm_fc modules and get mainline synced up with the latest
lio-core-2.6.git/master v4.1.0-rc1 HEAD as requested by hch.
I'll be squashing these into a proper GIT PULL to Linus shortly, so
please let me know if anything else is missing and/or needs to be
addressed for the first round of v3.1 updates.
The 3.1-pending branch is available directly via git here:
git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending.git 3.1-pending
Thank you!
--nab
Andy Grover (66):
target: use errno values instead of returning -1 for everything
target: Rename transport_calc_sg_num to transport_init_task_sg
target: Fix leak in error path in transport_init_task_sg
target: Remove pscsi_get_sh() in _pscsi.c
target: Make two runtime checks into WARN_ONs
target: Remove hba queue depth and convert to spin_lock_irq usage
target: dev->dev_status_queue_obj is unused
target: Make struct se_queue_req.cmd type struct se_cmd *
target: Remove __transport_get_qr_from_queue()
target: Rename se_dev->g_se_dev_list to se_dev_node
target: Remove struct se_global
target: Simplify scsi mib index table code
target: Make dev_queue_obj a member of se_device instead of a pointer
target: remove extraneous returns at end of void functions
target: Ensure transport_dump_vpd_ident_type returns null-terminated
str
target: Function pointers don't need to use '&' to be assigned
target: Fix comment in __transport_execute_tasks()
target: Misc style cleanups
target: rename struct pr_reservation_template to pr_reservation
target: Remove #defines that just perform indirection
target: Inline transport_get_task_from_execute_queue()
target: Remove unneeded casts to void*
target: Rename get_lun_for_{cmd,tmr} to lookup_{cmd,tmr}_lun
target: Make t_task a member of se_cmd, not a pointer
target: Handle functions returning "-2"
target: Use cmd->se_dev over cmd->se_lun->lun_se_dev
target: Embed qr in struct se_cmd
target: Replace embedded struct se_queue_req with a list_head
target: Rename list_heads that are nodes in struct se_cmd to "*_node"
target: Fold transport_device_setup_cmd() into lookup_{tmr,cmd}_lun()
target: Make t_mem_list and t_mem_list_bidi members of t_task
target: Add comment & cleanup transport_map_sg_to_mem()
target: Remove unneeded checks in transport_free_pages()
target: Replace custom sg<->buf functions with lib funcs
target: Simplify sector limiting code
target: get_cdb should never return NULL
target: Simplify transport_memcpy_se_mem_read_contig
target: Use assignment rather than increment for t_task_cdbs
target: Don't pass dma_size to generic_get_mem
target: Pass sg with type scatterlist in transport_map_sg_to_mem
target: Move task_sg_num next to task_sg in struct se_task
target: inline struct se_transport_task into struct se_cmd
target: Change name & semantics of transport_get_sectors()
target: Remove unused members of se_cmd
target: Rename se_cmd.t_task_cdbs to t_task_list_num
target: Fix some spelling
target: Remove unused var from transport_generic_do_tmr
target: map_sg_to_mem: return sg_count in return value
target/pscsi: Use min_t for sector limits
target/pscsi: Unused param for pscsi_get_bio()
target: Rename get_cdb_count to allocate_tasks
target: Remove ifdeffed code in t_g_process_write
target: Pass 2nd param of transport_split_cdb by value
target: Make all control CDBs scatter-gather
target: Enforce 1 page max for control cdb buffer sizes
target: Remove direct ramdisk code
target: Eliminate usage of struct se_mem
target: Rename task_sg_num to task_sg_nents
target: Remove custom debug macros for pr_debug. Use pr_err().
target: Remove custom debug macros in non-iscsi fabrics
target: Set WSNZ=1 in block limits VPD. Abort if WRITE_SAME sectors =
0
target: Remove transport do_se_mem_map callback
target: Further simplify transport_free_pages
target: Redo task allocation return value handling
target: Remove extra parentheses
target: change alloc_task call to take *cdb, not *cmd
Andy Shevchenko (2):
target: fix typo Assoication -> Association
target: remove custom hex2bin() implementation
Christoph Hellwig (6):
target: remove the always-noop ->new_cmd_failure method
target: remove the unused SCF_SE_DISABLE_ONLINE_CHECK flag
target: remove the unused SCF_CMD_PASSTHROUGH_NOALLOC flag
target: remove the unused SCF_EMULATE_SYNC_UNMAP flag
target: remove the unused SCF_EMULATE_SYNC_CACHE flag
target: merge release_cmd methods
Dan Carpenter (1):
tcm_loop: off by one in tcm_loop_make_naa_tpg()
Fubo Chen (1):
target: Check LUN numbers in transport_get_lun_for_[cmd,tmr]
Jesper Juhl (2):
loopback: Remove duplicate scsi/scsi_tcq.h include
loopback: Fix memory leak in tcm_loop_make_scsi_hba()
Nicholas Bellinger (17):
target: Minor header comment fixes
target: use MAINTENANCE_IN and MAINTENANCE_OUT definitions in scsi.h
target: Remove unnecessary *cdb transport_get_lun_for_cmd parameter
target: Remove unused su_group usage in fabric register/dergister
target: Fix double test of inquiry_prod
target: Fix WRITE_SAME_16 t_task_lba assignment bug
target: Fix WRITE_SAME_[16,32] number of blocks=0 case
tcm_fc: Makefile cleanups
tcm_fc: Convert to wake_up_process and schedule_timeout_interruptible
target: Make transport_generic_new_cmd() available for iscsi-target
target: Remove fabric callback to allocate iovecs
target: Fix transport_generic_new_cmd WRITE comment
target: Add transport_handle_cdb_direct optimization
target: Add SCF_EMULATE_QUEUE_FULL -> transport_handle_queue_full
target/iblock: Use request_queue->nr_request for se_device defaults
target: ->map_task_SG conversion to ->map_control_SG and
->map_data_SG
target: Update QUEUE ALGORITHM MODIFIER control page default
Roland Dreier (7):
target: Make transport_lookup_cmd_lun() locking IRQ-safe
target: Make se_dev_check_online() locking IRQ-safe
target: Make se_tmr_lock IRQ-safe
target/iblock: Remove unused iblock_dev members
target: Allow for built-in target modules
target: Fix reporting of supported VPD pages
target: Implement Block Device Characteristics VPD page
Stefan Hajnoczi (1):
loopback: Only build tcm_loop as module
drivers/target/loopback/Kconfig | 9 +-
drivers/target/loopback/tcm_loop.c | 207 +-
drivers/target/loopback/tcm_loop.h | 6 -
drivers/target/target_core_alua.c | 426 ++--
drivers/target/target_core_cdb.c | 457 ++--
drivers/target/target_core_configfs.c | 667 +++---
drivers/target/target_core_device.c | 819 +++---
drivers/target/target_core_fabric_configfs.c | 122 +-
drivers/target/target_core_fabric_lib.c | 27 +-
drivers/target/target_core_file.c | 149 +-
drivers/target/target_core_file.h | 4 -
drivers/target/target_core_hba.c | 37 +-
drivers/target/target_core_iblock.c | 199 +-
drivers/target/target_core_iblock.h | 9 -
drivers/target/target_core_pr.c | 862 ++++---
drivers/target/target_core_pr.h | 2 +-
drivers/target/target_core_pscsi.c | 316 +--
drivers/target/target_core_pscsi.h | 4 +-
drivers/target/target_core_rd.c | 483 +---
drivers/target/target_core_rd.h | 4 -
drivers/target/target_core_scdb.c | 20 +-
drivers/target/target_core_scdb.h | 10 +-
drivers/target/target_core_stat.c | 112 +-
drivers/target/target_core_tmr.c | 185 +-
drivers/target/target_core_tpg.c | 206 +-
drivers/target/target_core_transport.c | 3688 ++++++++++----------------
drivers/target/target_core_ua.c | 62 +-
drivers/target/tcm_fc/Makefile | 17 +-
drivers/target/tcm_fc/tcm_fc.h | 25 -
drivers/target/tcm_fc/tfc_cmd.c | 114 +-
drivers/target/tcm_fc/tfc_conf.c | 47 +-
drivers/target/tcm_fc/tfc_io.c | 130 +-
drivers/target/tcm_fc/tfc_sess.c | 18 +-
include/target/target_core_base.h | 226 +-
include/target/target_core_device.h | 6 +-
include/target/target_core_fabric_ops.h | 9 +-
include/target/target_core_transport.h | 43 +-
37 files changed, 4089 insertions(+), 5638 deletions(-)
--
1.7.6
^ permalink raw reply [flat|nested] 38+ messages in thread
* [PATCH 001/103] target: Check LUN numbers in transport_get_lun_for_[cmd,tmr]
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 002/103] target: Minor header comment fixes Nicholas A. Bellinger
` (28 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Fubo Chen, Nicholas A. Bellinger
From: Fubo Chen <fubo.chen@gmail.com>
This patch checks the passed 'unpacked_lun' against TRANSPORT_MAX_LUNS_PER_TPG
before reading from struct se_node_acl->device_list[].
Signed-off-by: Fubo Chen <fubo.chen@gmail.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_device.c | 12 ++++++++++++
1 files changed, 12 insertions(+), 0 deletions(-)
diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c
index ba698ea..f5516a1 100644
--- a/drivers/target/target_core_device.c
+++ b/drivers/target/target_core_device.c
@@ -65,6 +65,12 @@ int transport_get_lun_for_cmd(
unsigned long flags;
int read_only = 0;
+ if (unpacked_lun >= TRANSPORT_MAX_LUNS_PER_TPG) {
+ se_cmd->scsi_sense_reason = TCM_NON_EXISTENT_LUN;
+ se_cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
+ return -1;
+ }
+
spin_lock_irq(&SE_NODE_ACL(se_sess)->device_list_lock);
deve = se_cmd->se_deve =
&SE_NODE_ACL(se_sess)->device_list[unpacked_lun];
@@ -187,6 +193,12 @@ int transport_get_lun_for_tmr(
struct se_session *se_sess = SE_SESS(se_cmd);
struct se_tmr_req *se_tmr = se_cmd->se_tmr_req;
+ if (unpacked_lun >= TRANSPORT_MAX_LUNS_PER_TPG) {
+ se_cmd->scsi_sense_reason = TCM_NON_EXISTENT_LUN;
+ se_cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
+ return -1;
+ }
+
spin_lock_irq(&SE_NODE_ACL(se_sess)->device_list_lock);
deve = se_cmd->se_deve =
&SE_NODE_ACL(se_sess)->device_list[unpacked_lun];
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 002/103] target: Minor header comment fixes
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 001/103] target: Check LUN numbers in transport_get_lun_for_[cmd,tmr] Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 003/103] target: use MAINTENANCE_IN and MAINTENANCE_OUT definitions in scsi.h Nicholas A. Bellinger
` (27 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Nicholas Bellinger
From: Nicholas Bellinger <nab@linux-iscsi.org>
Reported-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_device.c | 2 +-
drivers/target/target_core_hba.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c
index f5516a1..1b5d78f 100644
--- a/drivers/target/target_core_device.c
+++ b/drivers/target/target_core_device.c
@@ -1,7 +1,7 @@
/*******************************************************************************
* Filename: target_core_device.c (based on iscsi_target_device.c)
*
- * This file contains the iSCSI Virtual Device and Disk Transport
+ * This file contains the TCM Virtual Device and Disk Transport
* agnostic related functions.
*
* Copyright (c) 2003, 2004, 2005 PyX Technologies, Inc.
diff --git a/drivers/target/target_core_hba.c b/drivers/target/target_core_hba.c
index 0b8f8da..ddf5ae0 100644
--- a/drivers/target/target_core_hba.c
+++ b/drivers/target/target_core_hba.c
@@ -1,7 +1,7 @@
/*******************************************************************************
* Filename: target_core_hba.c
*
- * This file copntains the iSCSI HBA Transport related functions.
+ * This file contains the TCM HBA Transport related functions.
*
* Copyright (c) 2003, 2004, 2005 PyX Technologies, Inc.
* Copyright (c) 2005, 2006, 2007 SBE, Inc.
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 003/103] target: use MAINTENANCE_IN and MAINTENANCE_OUT definitions in scsi.h
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 001/103] target: Check LUN numbers in transport_get_lun_for_[cmd,tmr] Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 002/103] target: Minor header comment fixes Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 004/103] target: Remove unnecessary *cdb transport_get_lun_for_cmd parameter Nicholas A. Bellinger
` (26 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Nicholas Bellinger
From: Nicholas Bellinger <nab@linux-iscsi.org>
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_transport.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 4b9b716..1882251 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -3325,7 +3325,7 @@ static int transport_generic_cmd_sequencer(
goto out_unsupported_cdb;
}
break;
- case 0xa3:
+ case MAINTENANCE_IN:
if (TRANSPORT(dev)->get_device_type(dev) != TYPE_ROM) {
/* MAINTENANCE_IN from SCC-2 */
/*
@@ -3395,7 +3395,7 @@ static int transport_generic_cmd_sequencer(
size = READ_POSITION_LEN;
cmd->se_cmd_flags |= SCF_SCSI_CONTROL_NONSG_IO_CDB;
break;
- case 0xa4:
+ case MAINTENANCE_OUT:
if (TRANSPORT(dev)->get_device_type(dev) != TYPE_ROM) {
/* MAINTENANCE_OUT from SCC-2
*
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 004/103] target: Remove unnecessary *cdb transport_get_lun_for_cmd parameter
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (2 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 003/103] target: use MAINTENANCE_IN and MAINTENANCE_OUT definitions in scsi.h Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 005/103] loopback: Only build tcm_loop as module Nicholas A. Bellinger
` (25 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Nicholas Bellinger
From: Nicholas Bellinger <nab@linux-iscsi.org>
This patch removes the now unnecessary 'unsigned char *cdb' function
parameter from transport_get_lun_for_cmd(). This also includes updating
lio-target, tcm_loop and tcm_fc usage of transport_get_lun_for_cmd().
Reported-by: Fubo Chen <fubo.chen@gmail.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/loopback/tcm_loop.c | 2 +-
drivers/target/target_core_device.c | 1 -
drivers/target/tcm_fc/tfc_cmd.c | 2 +-
include/target/target_core_device.h | 2 +-
4 files changed, 3 insertions(+), 4 deletions(-)
diff --git a/drivers/target/loopback/tcm_loop.c b/drivers/target/loopback/tcm_loop.c
index 70c2e7f..2d0f22a 100644
--- a/drivers/target/loopback/tcm_loop.c
+++ b/drivers/target/loopback/tcm_loop.c
@@ -122,7 +122,7 @@ static struct se_cmd *tcm_loop_allocate_core_cmd(
/*
* Locate the struct se_lun pointer and attach it to struct se_cmd
*/
- if (transport_get_lun_for_cmd(se_cmd, NULL, tl_cmd->sc->device->lun) < 0) {
+ if (transport_get_lun_for_cmd(se_cmd, tl_cmd->sc->device->lun) < 0) {
kmem_cache_free(tcm_loop_cmd_cache, tl_cmd);
set_host_byte(sc, DID_NO_CONNECT);
return NULL;
diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c
index 1b5d78f..b0ebe9a 100644
--- a/drivers/target/target_core_device.c
+++ b/drivers/target/target_core_device.c
@@ -56,7 +56,6 @@ static void se_dev_stop(struct se_device *dev);
int transport_get_lun_for_cmd(
struct se_cmd *se_cmd,
- unsigned char *cdb,
u32 unpacked_lun)
{
struct se_dev_entry *deve;
diff --git a/drivers/target/tcm_fc/tfc_cmd.c b/drivers/target/tcm_fc/tfc_cmd.c
index b2a1067..92a449a 100644
--- a/drivers/target/tcm_fc/tfc_cmd.c
+++ b/drivers/target/tcm_fc/tfc_cmd.c
@@ -637,7 +637,7 @@ static void ft_send_cmd(struct ft_cmd *cmd)
fc_seq_exch(cmd->seq)->lp->tt.seq_set_resp(cmd->seq, ft_recv_seq, cmd);
cmd->lun = scsilun_to_int((struct scsi_lun *)fcp->fc_lun);
- ret = transport_get_lun_for_cmd(&cmd->se_cmd, NULL, cmd->lun);
+ ret = transport_get_lun_for_cmd(&cmd->se_cmd, cmd->lun);
if (ret < 0) {
ft_dump_cmd(cmd, __func__);
transport_send_check_condition_and_sense(&cmd->se_cmd,
diff --git a/include/target/target_core_device.h b/include/target/target_core_device.h
index 52b18a5..d9745bf 100644
--- a/include/target/target_core_device.h
+++ b/include/target/target_core_device.h
@@ -1,7 +1,7 @@
#ifndef TARGET_CORE_DEVICE_H
#define TARGET_CORE_DEVICE_H
-extern int transport_get_lun_for_cmd(struct se_cmd *, unsigned char *, u32);
+extern int transport_get_lun_for_cmd(struct se_cmd *, u32);
extern int transport_get_lun_for_tmr(struct se_cmd *, u32);
extern struct se_dev_entry *core_get_se_deve_from_rtpi(
struct se_node_acl *, u16);
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 005/103] loopback: Only build tcm_loop as module
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (3 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 004/103] target: Remove unnecessary *cdb transport_get_lun_for_cmd parameter Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 21:18 ` Roland Dreier
2011-07-21 7:04 ` [PATCH 006/103] target: Remove unused su_group usage in fabric register/dergister Nicholas A. Bellinger
` (24 subsequent siblings)
29 siblings, 1 reply; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Stefan Hajnoczi
From: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
The tcm_loop must be built as a module. It cannot be built in.
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Nicholas A. Bellinger <nab@risingtidesystems.com>
---
drivers/target/loopback/Kconfig | 3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/drivers/target/loopback/Kconfig b/drivers/target/loopback/Kconfig
index 57dcbc2..1bd6a65 100644
--- a/drivers/target/loopback/Kconfig
+++ b/drivers/target/loopback/Kconfig
@@ -1,7 +1,8 @@
config LOOPBACK_TARGET
tristate "TCM Virtual SAS target and Linux/SCSI LDD fabric loopback module"
+ depends on m
help
- Say Y here to enable the TCM Virtual SAS target and Linux/SCSI LLD
+ Say M here to enable the TCM Virtual SAS target and Linux/SCSI LLD
fabric loopback module.
config LOOPBACK_TARGET_CDB_DEBUG
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 006/103] target: Remove unused su_group usage in fabric register/dergister
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (4 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 005/103] loopback: Only build tcm_loop as module Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 007/103] target: use errno values instead of returning -1 for everything Nicholas A. Bellinger
` (23 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Nicholas Bellinger
From: Nicholas Bellinger <nab@linux-iscsi.org>
This patch removes two instances of left over v3.x code performing local
scope access to struct target_core_fabric_ops->tf_subsys->su_group in
target_fabric_configfs_register() and target_fabric_configfs_deregister().
Reported-by: Christophe Fergeau <cfergeau@gmail.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_configfs.c | 15 ---------------
1 files changed, 0 insertions(+), 15 deletions(-)
diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c
index 25c1f49..c614000 100644
--- a/drivers/target/target_core_configfs.c
+++ b/drivers/target/target_core_configfs.c
@@ -533,7 +533,6 @@ static int target_fabric_tf_ops_check(
int target_fabric_configfs_register(
struct target_fabric_configfs *tf)
{
- struct config_group *su_group;
int ret;
if (!(tf)) {
@@ -546,12 +545,6 @@ int target_fabric_configfs_register(
" pointer\n");
return -EINVAL;
}
- su_group = &tf->tf_subsys->su_group;
- if (!(su_group)) {
- printk(KERN_ERR "Unable to locate target struct config_group"
- " pointer\n");
- return -EINVAL;
- }
ret = target_fabric_tf_ops_check(tf);
if (ret < 0)
return ret;
@@ -565,7 +558,6 @@ EXPORT_SYMBOL(target_fabric_configfs_register);
void target_fabric_configfs_deregister(
struct target_fabric_configfs *tf)
{
- struct config_group *su_group;
struct configfs_subsystem *su;
if (!(tf)) {
@@ -579,13 +571,6 @@ void target_fabric_configfs_deregister(
" pointer\n");
return;
}
- su_group = &tf->tf_subsys->su_group;
- if (!(su_group)) {
- printk(KERN_ERR "Unable to locate target struct config_group"
- " pointer\n");
- return;
- }
-
printk(KERN_INFO "<<<<<<<<<<<<<<<<<<<<<< BEGIN FABRIC API >>>>>>>>>>"
">>>>>>>>>>>>\n");
mutex_lock(&g_tf_lock);
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 007/103] target: use errno values instead of returning -1 for everything
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (5 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 006/103] target: Remove unused su_group usage in fabric register/dergister Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 008/103] target: Rename transport_calc_sg_num to transport_init_task_sg Nicholas A. Bellinger
` (22 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover, Nicholas A. Bellinger
From: Andy Grover <agrover@redhat.com>
Return errno values instead of -1. If checking the errno of a function,
return what that function returned instead of -1, as well.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/loopback/tcm_loop.c | 4 +-
drivers/target/target_core_alua.c | 24 ++++----
drivers/target/target_core_cdb.c | 12 ++--
drivers/target/target_core_configfs.c | 12 ++--
drivers/target/target_core_device.c | 98 ++++++++++++++++----------------
drivers/target/target_core_file.c | 14 ++--
drivers/target/target_core_iblock.c | 2 +-
drivers/target/target_core_pr.c | 36 ++++++------
drivers/target/target_core_pscsi.c | 8 +-
drivers/target/target_core_rd.c | 24 ++++----
drivers/target/target_core_tpg.c | 11 ++-
drivers/target/target_core_transport.c | 51 +++++++++--------
drivers/target/target_core_ua.c | 16 +++---
drivers/target/tcm_fc/tfc_conf.c | 4 +-
14 files changed, 161 insertions(+), 155 deletions(-)
diff --git a/drivers/target/loopback/tcm_loop.c b/drivers/target/loopback/tcm_loop.c
index 2d0f22a..3274fec 100644
--- a/drivers/target/loopback/tcm_loop.c
+++ b/drivers/target/loopback/tcm_loop.c
@@ -1402,9 +1402,9 @@ static int tcm_loop_register_configfs(void)
* Register the top level struct config_item_type with TCM core
*/
fabric = target_fabric_configfs_init(THIS_MODULE, "loopback");
- if (!fabric) {
+ if (IS_ERR(fabric)) {
printk(KERN_ERR "tcm_loop_register_configfs() failed!\n");
- return -1;
+ return PTR_ERR(fabric);
}
/*
* Setup the fabric API of function pointers used by target_core_mod
diff --git a/drivers/target/target_core_alua.c b/drivers/target/target_core_alua.c
index 47abb42..a993f3c 100644
--- a/drivers/target/target_core_alua.c
+++ b/drivers/target/target_core_alua.c
@@ -522,7 +522,7 @@ static int core_alua_state_check(
default:
printk(KERN_ERR "Unknown ALUA access state: 0x%02x\n",
out_alua_state);
- return -1;
+ return -EINVAL;
}
return 0;
@@ -553,7 +553,7 @@ static int core_alua_check_transition(int state, int *primary)
break;
default:
printk(KERN_ERR "Unknown ALUA access state: 0x%02x\n", state);
- return -1;
+ return -EINVAL;
}
return 0;
@@ -977,7 +977,7 @@ static int core_alua_set_tg_pt_secondary_state(
spin_unlock(&tg_pt_gp_mem->tg_pt_gp_mem_lock);
printk(KERN_ERR "Unable to complete secondary state"
" transition\n");
- return -1;
+ return -EINVAL;
}
trans_delay_msecs = tg_pt_gp->tg_pt_gp_trans_delay_msecs;
/*
@@ -1015,7 +1015,7 @@ static int core_alua_set_tg_pt_secondary_state(
if (!(md_buf)) {
printk(KERN_ERR "Unable to allocate md_buf for"
" secondary ALUA access metadata\n");
- return -1;
+ return -ENOMEM;
}
mutex_lock(&port->sep_tg_pt_md_mutex);
core_alua_update_tpg_secondary_metadata(tg_pt_gp_mem, port,
@@ -1062,7 +1062,7 @@ int core_alua_set_lu_gp_id(struct t10_alua_lu_gp *lu_gp, u16 lu_gp_id)
if (lu_gp->lu_gp_valid_id) {
printk(KERN_WARNING "ALUA LU Group already has a valid ID,"
" ignoring request\n");
- return -1;
+ return -EINVAL;
}
spin_lock(&se_global->lu_gps_lock);
@@ -1071,7 +1071,7 @@ int core_alua_set_lu_gp_id(struct t10_alua_lu_gp *lu_gp, u16 lu_gp_id)
" 0x0000ffff reached\n");
spin_unlock(&se_global->lu_gps_lock);
kmem_cache_free(t10_alua_lu_gp_cache, lu_gp);
- return -1;
+ return -ENOSPC;
}
again:
lu_gp_id_tmp = (lu_gp_id != 0) ? lu_gp_id :
@@ -1086,7 +1086,7 @@ again:
" already exists, ignoring request\n",
lu_gp_id);
spin_unlock(&se_global->lu_gps_lock);
- return -1;
+ return -EINVAL;
}
}
@@ -1330,7 +1330,7 @@ int core_alua_set_tg_pt_gp_id(
if (tg_pt_gp->tg_pt_gp_valid_id) {
printk(KERN_WARNING "ALUA TG PT Group already has a valid ID,"
" ignoring request\n");
- return -1;
+ return -EINVAL;
}
spin_lock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
@@ -1339,7 +1339,7 @@ int core_alua_set_tg_pt_gp_id(
" 0x0000ffff reached\n");
spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
kmem_cache_free(t10_alua_tg_pt_gp_cache, tg_pt_gp);
- return -1;
+ return -ENOSPC;
}
again:
tg_pt_gp_id_tmp = (tg_pt_gp_id != 0) ? tg_pt_gp_id :
@@ -1354,7 +1354,7 @@ again:
printk(KERN_ERR "ALUA Target Port Group ID: %hu already"
" exists, ignoring request\n", tg_pt_gp_id);
spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
- return -1;
+ return -EINVAL;
}
}
@@ -1967,8 +1967,8 @@ int core_setup_alua(struct se_device *dev, int force_pt)
* LUN Group.
*/
lu_gp_mem = core_alua_allocate_lu_gp_mem(dev);
- if (IS_ERR(lu_gp_mem) || !lu_gp_mem)
- return -1;
+ if (IS_ERR(lu_gp_mem))
+ return PTR_ERR(lu_gp_mem);
alua->alua_type = SPC3_ALUA_EMULATED;
alua->alua_state_check = &core_alua_state_check;
diff --git a/drivers/target/target_core_cdb.c b/drivers/target/target_core_cdb.c
index 7f19c8b..b39ffad 100644
--- a/drivers/target/target_core_cdb.c
+++ b/drivers/target/target_core_cdb.c
@@ -75,7 +75,7 @@ target_emulate_inquiry_std(struct se_cmd *cmd)
if (cmd->data_length < 6) {
printk(KERN_ERR "SCSI Inquiry payload length: %u"
" too small for EVPD=0\n", cmd->data_length);
- return -1;
+ return -EINVAL;
}
buf[0] = dev->transport->get_device_type(dev);
@@ -501,7 +501,7 @@ target_emulate_evpd_b0(struct se_cmd *cmd, unsigned char *buf)
printk(KERN_INFO "Received data_length: %u"
" too small for EVPD 0xb0\n",
cmd->data_length);
- return -1;
+ return -EINVAL;
}
if (have_tp && cmd->data_length < (0x3c + 4)) {
@@ -637,7 +637,7 @@ target_emulate_inquiry(struct se_cmd *cmd)
if (cmd->data_length < 4) {
printk(KERN_ERR "SCSI Inquiry payload length: %u"
" too small for EVPD=1\n", cmd->data_length);
- return -1;
+ return -EINVAL;
}
buf[0] = dev->transport->get_device_type(dev);
@@ -656,7 +656,7 @@ target_emulate_inquiry(struct se_cmd *cmd)
return target_emulate_evpd_b2(cmd, buf);
default:
printk(KERN_ERR "Unknown VPD Code: 0x%02x\n", cdb[2]);
- return -1;
+ return -EINVAL;
}
return 0;
@@ -991,7 +991,7 @@ target_emulate_unmap(struct se_task *task)
if (ret < 0) {
printk(KERN_ERR "blkdev_issue_discard() failed: %d\n",
ret);
- return -1;
+ return ret;
}
ptr += 16;
@@ -1024,7 +1024,7 @@ target_emulate_write_same(struct se_task *task)
ret = dev->transport->do_discard(dev, lba, range);
if (ret < 0) {
printk(KERN_INFO "blkdev_issue_discard() failed for WRITE_SAME\n");
- return -1;
+ return ret;
}
task->task_scsi_status = GOOD;
diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c
index c614000..d140258 100644
--- a/drivers/target/target_core_configfs.c
+++ b/drivers/target/target_core_configfs.c
@@ -298,21 +298,21 @@ struct target_fabric_configfs *target_fabric_configfs_init(
if (!(fabric_mod)) {
printk(KERN_ERR "Missing struct module *fabric_mod pointer\n");
- return NULL;
+ return ERR_PTR(-EINVAL);
}
if (!(name)) {
printk(KERN_ERR "Unable to locate passed fabric name\n");
- return NULL;
+ return ERR_PTR(-EINVAL);
}
if (strlen(name) >= TARGET_FABRIC_NAME_SIZE) {
printk(KERN_ERR "Passed name: %s exceeds TARGET_FABRIC"
"_NAME_SIZE\n", name);
- return NULL;
+ return ERR_PTR(-EINVAL);
}
tf = kzalloc(sizeof(struct target_fabric_configfs), GFP_KERNEL);
if (!(tf))
- return NULL;
+ return ERR_PTR(-ENOMEM);
INIT_LIST_HEAD(&tf->tf_list);
atomic_set(&tf->tf_access_cnt, 0);
@@ -3133,7 +3133,7 @@ static int __init target_core_init_configfs(void)
init_scsi_index_table();
ret = init_se_global();
if (ret < 0)
- return -1;
+ return ret;
/*
* Create $CONFIGFS/target/core default group for HBA <-> Storage Object
* and ALUA Logical Unit Group and Target Port Group infrastructure.
@@ -3241,7 +3241,7 @@ out_global:
kfree(hba_cg->default_groups);
kfree(target_cg->default_groups);
release_se_global();
- return -1;
+ return ret;
}
static void __exit target_core_exit_configfs(void)
diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c
index b0ebe9a..818c65b 100644
--- a/drivers/target/target_core_device.c
+++ b/drivers/target/target_core_device.c
@@ -67,7 +67,7 @@ int transport_get_lun_for_cmd(
if (unpacked_lun >= TRANSPORT_MAX_LUNS_PER_TPG) {
se_cmd->scsi_sense_reason = TCM_NON_EXISTENT_LUN;
se_cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
- return -1;
+ return -ENODEV;
}
spin_lock_irq(&SE_NODE_ACL(se_sess)->device_list_lock);
@@ -109,7 +109,7 @@ out:
" Access for 0x%08x\n",
CMD_TFO(se_cmd)->get_fabric_name(),
unpacked_lun);
- return -1;
+ return -EACCES;
} else {
/*
* Use the se_portal_group->tpg_virt_lun0 to allow for
@@ -123,7 +123,7 @@ out:
" Access for 0x%08x\n",
CMD_TFO(se_cmd)->get_fabric_name(),
unpacked_lun);
- return -1;
+ return -ENODEV;
}
/*
* Force WRITE PROTECT for virtual LUN 0
@@ -132,7 +132,7 @@ out:
(se_cmd->data_direction != DMA_NONE)) {
se_cmd->scsi_sense_reason = TCM_WRITE_PROTECTED;
se_cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
- return -1;
+ return -EACCES;
}
#if 0
printk("TARGET_CORE[%s]: Using virtual LUN0! :-)\n",
@@ -151,7 +151,7 @@ out:
if (se_dev_check_online(se_lun->lun_se_dev) != 0) {
se_cmd->scsi_sense_reason = TCM_NON_EXISTENT_LUN;
se_cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
- return -1;
+ return -ENODEV;
}
{
@@ -195,7 +195,7 @@ int transport_get_lun_for_tmr(
if (unpacked_lun >= TRANSPORT_MAX_LUNS_PER_TPG) {
se_cmd->scsi_sense_reason = TCM_NON_EXISTENT_LUN;
se_cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
- return -1;
+ return -ENODEV;
}
spin_lock_irq(&SE_NODE_ACL(se_sess)->device_list_lock);
@@ -217,7 +217,7 @@ int transport_get_lun_for_tmr(
CMD_TFO(se_cmd)->get_fabric_name(),
unpacked_lun);
se_cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
- return -1;
+ return -ENODEV;
}
/*
* Determine if the struct se_lun is online.
@@ -225,7 +225,7 @@ int transport_get_lun_for_tmr(
/* #warning FIXME: Check for LUN_RESET + UNIT Attention */
if (se_dev_check_online(se_lun->lun_se_dev) != 0) {
se_cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
- return -1;
+ return -ENODEV;
}
se_tmr->tmr_dev = dev;
@@ -408,14 +408,14 @@ int core_update_device_list_for_node(
" already set for demo mode -> explict"
" LUN ACL transition\n");
spin_unlock_irq(&nacl->device_list_lock);
- return -1;
+ return -EINVAL;
}
if (deve->se_lun != lun) {
printk(KERN_ERR "struct se_dev_entry->se_lun does"
" match passed struct se_lun for demo mode"
" -> explict LUN ACL transition\n");
spin_unlock_irq(&nacl->device_list_lock);
- return -1;
+ return -EINVAL;
}
deve->se_lun_acl = lun_acl;
trans = 1;
@@ -514,7 +514,7 @@ static struct se_port *core_alloc_port(struct se_device *dev)
port = kzalloc(sizeof(struct se_port), GFP_KERNEL);
if (!(port)) {
printk(KERN_ERR "Unable to allocate struct se_port\n");
- return NULL;
+ return ERR_PTR(-ENOMEM);
}
INIT_LIST_HEAD(&port->sep_alua_list);
INIT_LIST_HEAD(&port->sep_list);
@@ -527,7 +527,7 @@ static struct se_port *core_alloc_port(struct se_device *dev)
printk(KERN_WARNING "Reached dev->dev_port_count =="
" 0x0000ffff\n");
spin_unlock(&dev->se_port_lock);
- return NULL;
+ return ERR_PTR(-ENOSPC);
}
again:
/*
@@ -630,8 +630,8 @@ int core_dev_export(
struct se_port *port;
port = core_alloc_port(dev);
- if (!(port))
- return -1;
+ if (IS_ERR(port))
+ return PTR_ERR(port);
lun->lun_se_dev = dev;
se_dev_start(dev);
@@ -909,7 +909,7 @@ int se_dev_set_task_timeout(struct se_device *dev, u32 task_timeout)
if (task_timeout > DA_TASK_TIMEOUT_MAX) {
printk(KERN_ERR "dev[%p]: Passed task_timeout: %u larger then"
" DA_TASK_TIMEOUT_MAX\n", dev, task_timeout);
- return -1;
+ return -EINVAL;
} else {
DEV_ATTRIB(dev)->task_timeout = task_timeout;
printk(KERN_INFO "dev[%p]: Set SE Device task_timeout: %u\n",
@@ -963,15 +963,15 @@ int se_dev_set_emulate_dpo(struct se_device *dev, int flag)
{
if ((flag != 0) && (flag != 1)) {
printk(KERN_ERR "Illegal value %d\n", flag);
- return -1;
+ return -EINVAL;
}
if (TRANSPORT(dev)->dpo_emulated == NULL) {
printk(KERN_ERR "TRANSPORT(dev)->dpo_emulated is NULL\n");
- return -1;
+ return -EINVAL;
}
if (TRANSPORT(dev)->dpo_emulated(dev) == 0) {
printk(KERN_ERR "TRANSPORT(dev)->dpo_emulated not supported\n");
- return -1;
+ return -EINVAL;
}
DEV_ATTRIB(dev)->emulate_dpo = flag;
printk(KERN_INFO "dev[%p]: SE Device Page Out (DPO) Emulation"
@@ -983,15 +983,15 @@ int se_dev_set_emulate_fua_write(struct se_device *dev, int flag)
{
if ((flag != 0) && (flag != 1)) {
printk(KERN_ERR "Illegal value %d\n", flag);
- return -1;
+ return -EINVAL;
}
if (TRANSPORT(dev)->fua_write_emulated == NULL) {
printk(KERN_ERR "TRANSPORT(dev)->fua_write_emulated is NULL\n");
- return -1;
+ return -EINVAL;
}
if (TRANSPORT(dev)->fua_write_emulated(dev) == 0) {
printk(KERN_ERR "TRANSPORT(dev)->fua_write_emulated not supported\n");
- return -1;
+ return -EINVAL;
}
DEV_ATTRIB(dev)->emulate_fua_write = flag;
printk(KERN_INFO "dev[%p]: SE Device Forced Unit Access WRITEs: %d\n",
@@ -1003,15 +1003,15 @@ int se_dev_set_emulate_fua_read(struct se_device *dev, int flag)
{
if ((flag != 0) && (flag != 1)) {
printk(KERN_ERR "Illegal value %d\n", flag);
- return -1;
+ return -EINVAL;
}
if (TRANSPORT(dev)->fua_read_emulated == NULL) {
printk(KERN_ERR "TRANSPORT(dev)->fua_read_emulated is NULL\n");
- return -1;
+ return -EINVAL;
}
if (TRANSPORT(dev)->fua_read_emulated(dev) == 0) {
printk(KERN_ERR "TRANSPORT(dev)->fua_read_emulated not supported\n");
- return -1;
+ return -EINVAL;
}
DEV_ATTRIB(dev)->emulate_fua_read = flag;
printk(KERN_INFO "dev[%p]: SE Device Forced Unit Access READs: %d\n",
@@ -1023,15 +1023,15 @@ int se_dev_set_emulate_write_cache(struct se_device *dev, int flag)
{
if ((flag != 0) && (flag != 1)) {
printk(KERN_ERR "Illegal value %d\n", flag);
- return -1;
+ return -EINVAL;
}
if (TRANSPORT(dev)->write_cache_emulated == NULL) {
printk(KERN_ERR "TRANSPORT(dev)->write_cache_emulated is NULL\n");
- return -1;
+ return -EINVAL;
}
if (TRANSPORT(dev)->write_cache_emulated(dev) == 0) {
printk(KERN_ERR "TRANSPORT(dev)->write_cache_emulated not supported\n");
- return -1;
+ return -EINVAL;
}
DEV_ATTRIB(dev)->emulate_write_cache = flag;
printk(KERN_INFO "dev[%p]: SE Device WRITE_CACHE_EMULATION flag: %d\n",
@@ -1043,7 +1043,7 @@ int se_dev_set_emulate_ua_intlck_ctrl(struct se_device *dev, int flag)
{
if ((flag != 0) && (flag != 1) && (flag != 2)) {
printk(KERN_ERR "Illegal value %d\n", flag);
- return -1;
+ return -EINVAL;
}
if (atomic_read(&dev->dev_export_obj.obj_access_count)) {
@@ -1051,7 +1051,7 @@ int se_dev_set_emulate_ua_intlck_ctrl(struct se_device *dev, int flag)
" UA_INTRLCK_CTRL while dev_export_obj: %d count"
" exists\n", dev,
atomic_read(&dev->dev_export_obj.obj_access_count));
- return -1;
+ return -EINVAL;
}
DEV_ATTRIB(dev)->emulate_ua_intlck_ctrl = flag;
printk(KERN_INFO "dev[%p]: SE Device UA_INTRLCK_CTRL flag: %d\n",
@@ -1064,14 +1064,14 @@ int se_dev_set_emulate_tas(struct se_device *dev, int flag)
{
if ((flag != 0) && (flag != 1)) {
printk(KERN_ERR "Illegal value %d\n", flag);
- return -1;
+ return -EINVAL;
}
if (atomic_read(&dev->dev_export_obj.obj_access_count)) {
printk(KERN_ERR "dev[%p]: Unable to change SE Device TAS while"
" dev_export_obj: %d count exists\n", dev,
atomic_read(&dev->dev_export_obj.obj_access_count));
- return -1;
+ return -EINVAL;
}
DEV_ATTRIB(dev)->emulate_tas = flag;
printk(KERN_INFO "dev[%p]: SE Device TASK_ABORTED status bit: %s\n",
@@ -1084,7 +1084,7 @@ int se_dev_set_emulate_tpu(struct se_device *dev, int flag)
{
if ((flag != 0) && (flag != 1)) {
printk(KERN_ERR "Illegal value %d\n", flag);
- return -1;
+ return -EINVAL;
}
/*
* We expect this value to be non-zero when generic Block Layer
@@ -1105,7 +1105,7 @@ int se_dev_set_emulate_tpws(struct se_device *dev, int flag)
{
if ((flag != 0) && (flag != 1)) {
printk(KERN_ERR "Illegal value %d\n", flag);
- return -1;
+ return -EINVAL;
}
/*
* We expect this value to be non-zero when generic Block Layer
@@ -1126,7 +1126,7 @@ int se_dev_set_enforce_pr_isids(struct se_device *dev, int flag)
{
if ((flag != 0) && (flag != 1)) {
printk(KERN_ERR "Illegal value %d\n", flag);
- return -1;
+ return -EINVAL;
}
DEV_ATTRIB(dev)->enforce_pr_isids = flag;
printk(KERN_INFO "dev[%p]: SE Device enforce_pr_isids bit: %s\n", dev,
@@ -1145,12 +1145,12 @@ int se_dev_set_queue_depth(struct se_device *dev, u32 queue_depth)
printk(KERN_ERR "dev[%p]: Unable to change SE Device TCQ while"
" dev_export_obj: %d count exists\n", dev,
atomic_read(&dev->dev_export_obj.obj_access_count));
- return -1;
+ return -EINVAL;
}
if (!(queue_depth)) {
printk(KERN_ERR "dev[%p]: Illegal ZERO value for queue"
"_depth\n", dev);
- return -1;
+ return -EINVAL;
}
if (TRANSPORT(dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
@@ -1159,7 +1159,7 @@ int se_dev_set_queue_depth(struct se_device *dev, u32 queue_depth)
" exceeds TCM/SE_Device TCQ: %u\n",
dev, queue_depth,
DEV_ATTRIB(dev)->hw_queue_depth);
- return -1;
+ return -EINVAL;
}
} else {
if (queue_depth > DEV_ATTRIB(dev)->queue_depth) {
@@ -1168,7 +1168,7 @@ int se_dev_set_queue_depth(struct se_device *dev, u32 queue_depth)
" %u exceeds TCM/SE_Device MAX"
" TCQ: %u\n", dev, queue_depth,
DEV_ATTRIB(dev)->hw_queue_depth);
- return -1;
+ return -EINVAL;
}
}
}
@@ -1192,18 +1192,18 @@ int se_dev_set_max_sectors(struct se_device *dev, u32 max_sectors)
printk(KERN_ERR "dev[%p]: Unable to change SE Device"
" max_sectors while dev_export_obj: %d count exists\n",
dev, atomic_read(&dev->dev_export_obj.obj_access_count));
- return -1;
+ return -EINVAL;
}
if (!(max_sectors)) {
printk(KERN_ERR "dev[%p]: Illegal ZERO value for"
" max_sectors\n", dev);
- return -1;
+ return -EINVAL;
}
if (max_sectors < DA_STATUS_MAX_SECTORS_MIN) {
printk(KERN_ERR "dev[%p]: Passed max_sectors: %u less than"
" DA_STATUS_MAX_SECTORS_MIN: %u\n", dev, max_sectors,
DA_STATUS_MAX_SECTORS_MIN);
- return -1;
+ return -EINVAL;
}
if (TRANSPORT(dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
if (max_sectors > DEV_ATTRIB(dev)->hw_max_sectors) {
@@ -1211,7 +1211,7 @@ int se_dev_set_max_sectors(struct se_device *dev, u32 max_sectors)
" greater than TCM/SE_Device max_sectors:"
" %u\n", dev, max_sectors,
DEV_ATTRIB(dev)->hw_max_sectors);
- return -1;
+ return -EINVAL;
}
} else {
if (!(force) && (max_sectors >
@@ -1220,14 +1220,14 @@ int se_dev_set_max_sectors(struct se_device *dev, u32 max_sectors)
" greater than TCM/SE_Device max_sectors"
": %u, use force=1 to override.\n", dev,
max_sectors, DEV_ATTRIB(dev)->hw_max_sectors);
- return -1;
+ return -EINVAL;
}
if (max_sectors > DA_STATUS_MAX_SECTORS_MAX) {
printk(KERN_ERR "dev[%p]: Passed max_sectors: %u"
" greater than DA_STATUS_MAX_SECTORS_MAX:"
" %u\n", dev, max_sectors,
DA_STATUS_MAX_SECTORS_MAX);
- return -1;
+ return -EINVAL;
}
}
@@ -1269,7 +1269,7 @@ int se_dev_set_block_size(struct se_device *dev, u32 block_size)
printk(KERN_ERR "dev[%p]: Unable to change SE Device block_size"
" while dev_export_obj: %d count exists\n", dev,
atomic_read(&dev->dev_export_obj.obj_access_count));
- return -1;
+ return -EINVAL;
}
if ((block_size != 512) &&
@@ -1279,14 +1279,14 @@ int se_dev_set_block_size(struct se_device *dev, u32 block_size)
printk(KERN_ERR "dev[%p]: Illegal value for block_device: %u"
" for SE device, must be 512, 1024, 2048 or 4096\n",
dev, block_size);
- return -1;
+ return -EINVAL;
}
if (TRANSPORT(dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
printk(KERN_ERR "dev[%p]: Not allowed to change block_size for"
" Physical Device, use for Linux/SCSI to change"
" block_size for underlying hardware\n", dev);
- return -1;
+ return -EINVAL;
}
DEV_ATTRIB(dev)->block_size = block_size;
@@ -1620,8 +1620,8 @@ int core_dev_setup_virtual_lun0(void)
t->set_configfs_dev_params(hba, se_dev, buf, sizeof(buf));
dev = t->create_virtdevice(hba, se_dev, se_dev->se_dev_su_ptr);
- if (!(dev) || IS_ERR(dev)) {
- ret = -ENOMEM;
+ if (IS_ERR(dev)) {
+ ret = PTR_ERR(dev);
goto out;
}
se_dev->se_dev_ptr = dev;
diff --git a/drivers/target/target_core_file.c b/drivers/target/target_core_file.c
index 150c430..40386ec 100644
--- a/drivers/target/target_core_file.c
+++ b/drivers/target/target_core_file.c
@@ -67,7 +67,7 @@ static int fd_attach_hba(struct se_hba *hba, u32 host_id)
fd_host = kzalloc(sizeof(struct fd_host), GFP_KERNEL);
if (!(fd_host)) {
printk(KERN_ERR "Unable to allocate memory for struct fd_host\n");
- return -1;
+ return -ENOMEM;
}
fd_host->fd_host_id = host_id;
@@ -300,7 +300,7 @@ static int fd_do_readv(struct se_task *task)
iov = kzalloc(sizeof(struct iovec) * task->task_sg_num, GFP_KERNEL);
if (!(iov)) {
printk(KERN_ERR "Unable to allocate fd_do_readv iov[]\n");
- return -1;
+ return -ENOMEM;
}
for (i = 0; i < task->task_sg_num; i++) {
@@ -324,13 +324,13 @@ static int fd_do_readv(struct se_task *task)
printk(KERN_ERR "vfs_readv() returned %d,"
" expecting %d for S_ISBLK\n", ret,
(int)task->task_size);
- return -1;
+ return (ret < 0 ? ret : -EINVAL);
}
} else {
if (ret < 0) {
printk(KERN_ERR "vfs_readv() returned %d for non"
" S_ISBLK\n", ret);
- return -1;
+ return ret;
}
}
@@ -350,7 +350,7 @@ static int fd_do_writev(struct se_task *task)
iov = kzalloc(sizeof(struct iovec) * task->task_sg_num, GFP_KERNEL);
if (!(iov)) {
printk(KERN_ERR "Unable to allocate fd_do_writev iov[]\n");
- return -1;
+ return -ENOMEM;
}
for (i = 0; i < task->task_sg_num; i++) {
@@ -367,7 +367,7 @@ static int fd_do_writev(struct se_task *task)
if (ret < 0 || ret != task->task_size) {
printk(KERN_ERR "vfs_writev() returned %d\n", ret);
- return -1;
+ return (ret < 0 ? ret : -EINVAL);
}
return 1;
@@ -599,7 +599,7 @@ static ssize_t fd_check_configfs_dev_params(struct se_hba *hba, struct se_subsys
if (!(fd_dev->fbd_flags & FBDF_HAS_PATH)) {
printk(KERN_ERR "Missing fd_dev_name=\n");
- return -1;
+ return -EINVAL;
}
return 0;
diff --git a/drivers/target/target_core_iblock.c b/drivers/target/target_core_iblock.c
index 8663900..5d5bd6e 100644
--- a/drivers/target/target_core_iblock.c
+++ b/drivers/target/target_core_iblock.c
@@ -527,7 +527,7 @@ static ssize_t iblock_check_configfs_dev_params(
if (!(ibd->ibd_flags & IBDF_HAS_UDEV_PATH)) {
printk(KERN_ERR "Missing udev_path= parameters for IBLOCK\n");
- return -1;
+ return -EINVAL;
}
return 0;
diff --git a/drivers/target/target_core_pr.c b/drivers/target/target_core_pr.c
index b662db3..99b97db 100644
--- a/drivers/target/target_core_pr.c
+++ b/drivers/target/target_core_pr.c
@@ -105,13 +105,13 @@ static int core_scsi2_reservation_check(struct se_cmd *cmd, u32 *pr_reg_type)
}
if (dev->dev_reserved_node_acl != sess->se_node_acl) {
spin_unlock(&dev->dev_reservation_lock);
- return -1;
+ return -EINVAL;
}
if (!(dev->dev_flags & DF_SPC2_RESERVATIONS_WITH_ISID)) {
spin_unlock(&dev->dev_reservation_lock);
return 0;
}
- ret = (dev->dev_res_bin_isid == sess->sess_bin_isid) ? 0 : -1;
+ ret = (dev->dev_res_bin_isid == sess->sess_bin_isid) ? 0 : -EINVAL;
spin_unlock(&dev->dev_reservation_lock);
return ret;
@@ -362,7 +362,7 @@ static int core_scsi3_pr_seq_non_holder(
registered_nexus = 1;
break;
default:
- return -1;
+ return -EINVAL;
}
/*
* Referenced from spc4r17 table 45 for *NON* PR holder access
@@ -414,7 +414,7 @@ static int core_scsi3_pr_seq_non_holder(
default:
printk(KERN_ERR "Unknown PERSISTENT_RESERVE_OUT service"
" action: 0x%02x\n", cdb[1] & 0x1f);
- return -1;
+ return -EINVAL;
}
break;
case RELEASE:
@@ -461,7 +461,7 @@ static int core_scsi3_pr_seq_non_holder(
default:
printk(KERN_ERR "Unknown MI Service Action: 0x%02x\n",
(cdb[1] & 0x1f));
- return -1;
+ return -EINVAL;
}
break;
case ACCESS_CONTROL_IN:
@@ -592,14 +592,14 @@ static int core_scsi3_pr_reservation_check(
cmd->pr_res_key = dev->dev_pr_res_holder->pr_res_key;
if (dev->dev_pr_res_holder->pr_reg_nacl != sess->se_node_acl) {
spin_unlock(&dev->dev_reservation_lock);
- return -1;
+ return -EINVAL;
}
if (!(dev->dev_pr_res_holder->isid_present_at_reg)) {
spin_unlock(&dev->dev_reservation_lock);
return 0;
}
ret = (dev->dev_pr_res_holder->pr_reg_bin_isid ==
- sess->sess_bin_isid) ? 0 : -1;
+ sess->sess_bin_isid) ? 0 : -EINVAL;
/*
* Use bit in *pr_reg_type to notify ISID mismatch in
* core_scsi3_pr_seq_non_holder().
@@ -819,13 +819,13 @@ int core_scsi3_alloc_aptpl_registration(
if (!(i_port) || !(t_port) || !(sa_res_key)) {
printk(KERN_ERR "Illegal parameters for APTPL registration\n");
- return -1;
+ return -EINVAL;
}
pr_reg = kmem_cache_zalloc(t10_pr_reg_cache, GFP_KERNEL);
if (!(pr_reg)) {
printk(KERN_ERR "Unable to allocate struct t10_pr_registration\n");
- return -1;
+ return -ENOMEM;
}
pr_reg->pr_aptpl_buf = kzalloc(pr_tmpl->pr_aptpl_buf_len, GFP_KERNEL);
@@ -1107,7 +1107,7 @@ static int core_scsi3_alloc_registration(
pr_reg = __core_scsi3_alloc_registration(dev, nacl, deve, isid,
sa_res_key, all_tg_pt, aptpl);
if (!(pr_reg))
- return -1;
+ return -EPERM;
__core_scsi3_add_registration(dev, nacl, pr_reg,
register_type, register_move);
@@ -1240,7 +1240,7 @@ static int core_scsi3_check_implict_release(
" UNREGISTER while existing reservation with matching"
" key 0x%016Lx is present from another SCSI Initiator"
" Port\n", pr_reg->pr_res_key);
- ret = -1;
+ ret = -EPERM;
}
spin_unlock(&dev->dev_reservation_lock);
@@ -1920,7 +1920,7 @@ static int __core_scsi3_update_aptpl_buf(
printk(KERN_ERR "Unable to update renaming"
" APTPL metadata\n");
spin_unlock(&T10_RES(su_dev)->registration_lock);
- return -1;
+ return -EMSGSIZE;
}
len += sprintf(buf+len, "%s", tmp);
@@ -1938,7 +1938,7 @@ static int __core_scsi3_update_aptpl_buf(
printk(KERN_ERR "Unable to update renaming"
" APTPL metadata\n");
spin_unlock(&T10_RES(su_dev)->registration_lock);
- return -1;
+ return -EMSGSIZE;
}
len += sprintf(buf+len, "%s", tmp);
reg_count++;
@@ -1989,7 +1989,7 @@ static int __core_scsi3_write_aptpl_to_file(
if (strlen(&wwn->unit_serial[0]) >= 512) {
printk(KERN_ERR "WWN value for struct se_device does not fit"
" into path buffer\n");
- return -1;
+ return -EMSGSIZE;
}
snprintf(path, 512, "/var/target/pr/aptpl_%s", &wwn->unit_serial[0]);
@@ -1997,7 +1997,7 @@ static int __core_scsi3_write_aptpl_to_file(
if (IS_ERR(file) || !file || !file->f_dentry) {
printk(KERN_ERR "filp_open(%s) for APTPL metadata"
" failed\n", path);
- return -1;
+ return (PTR_ERR(file) < 0 ? PTR_ERR(file) : -ENOENT);
}
iov[0].iov_base = &buf[0];
@@ -2014,7 +2014,7 @@ static int __core_scsi3_write_aptpl_to_file(
if (ret < 0) {
printk("Error writing APTPL metadata file: %s\n", path);
filp_close(file, NULL);
- return -1;
+ return -EIO;
}
filp_close(file, NULL);
@@ -2049,14 +2049,14 @@ static int core_scsi3_update_and_write_aptpl(
ret = core_scsi3_update_aptpl_buf(dev, buf, pr_aptpl_buf_len,
clear_aptpl_metadata);
if (ret != 0)
- return -1;
+ return ret;
/*
* __core_scsi3_write_aptpl_to_file() will call strlen()
* on the passed buf to determine pr_aptpl_buf_len.
*/
ret = __core_scsi3_write_aptpl_to_file(dev, buf, 0);
if (ret != 0)
- return -1;
+ return ret;
return ret;
}
diff --git a/drivers/target/target_core_pscsi.c b/drivers/target/target_core_pscsi.c
index 331d423..fb2e8cb 100644
--- a/drivers/target/target_core_pscsi.c
+++ b/drivers/target/target_core_pscsi.c
@@ -86,7 +86,7 @@ static int pscsi_attach_hba(struct se_hba *hba, u32 host_id)
phv = kzalloc(sizeof(struct pscsi_hba_virt), GFP_KERNEL);
if (!(phv)) {
printk(KERN_ERR "Unable to allocate struct pscsi_hba_virt\n");
- return -1;
+ return -ENOMEM;
}
phv->phv_host_id = host_id;
phv->phv_mode = PHV_VIRUTAL_HOST_ID;
@@ -236,7 +236,7 @@ pscsi_get_inquiry_vpd_serial(struct scsi_device *sdev, struct t10_wwn *wwn)
buf = kzalloc(INQUIRY_VPD_SERIAL_LEN, GFP_KERNEL);
if (!buf)
- return -1;
+ return -ENOMEM;
memset(cdb, 0, MAX_COMMAND_SIZE);
cdb[0] = INQUIRY;
@@ -259,7 +259,7 @@ pscsi_get_inquiry_vpd_serial(struct scsi_device *sdev, struct t10_wwn *wwn)
out_free:
kfree(buf);
- return -1;
+ return -EPERM;
}
static void
@@ -1030,7 +1030,7 @@ static ssize_t pscsi_check_configfs_dev_params(
!(pdv->pdv_flags & PDF_HAS_LUN_ID)) {
printk(KERN_ERR "Missing scsi_channel_id=, scsi_target_id= and"
" scsi_lun_id= parameters\n");
- return -1;
+ return -EINVAL;
}
return 0;
diff --git a/drivers/target/target_core_rd.c b/drivers/target/target_core_rd.c
index 7837dd3..793a94b 100644
--- a/drivers/target/target_core_rd.c
+++ b/drivers/target/target_core_rd.c
@@ -383,7 +383,7 @@ static int rd_MEMCPY_read(struct rd_request *req)
table = rd_get_sg_table(dev, req->rd_page);
if (!(table))
- return -1;
+ return -EINVAL;
table_sg_end = (table->page_end_offset - req->rd_page);
sg_d = task->task_sg;
@@ -481,7 +481,7 @@ static int rd_MEMCPY_read(struct rd_request *req)
#endif
table = rd_get_sg_table(dev, req->rd_page);
if (!(table))
- return -1;
+ return -EINVAL;
sg_s = &table->sg_table[j = 0];
}
@@ -506,7 +506,7 @@ static int rd_MEMCPY_write(struct rd_request *req)
table = rd_get_sg_table(dev, req->rd_page);
if (!(table))
- return -1;
+ return -EINVAL;
table_sg_end = (table->page_end_offset - req->rd_page);
sg_d = &table->sg_table[req->rd_page - table->page_start_offset];
@@ -604,7 +604,7 @@ static int rd_MEMCPY_write(struct rd_request *req)
#endif
table = rd_get_sg_table(dev, req->rd_page);
if (!(table))
- return -1;
+ return -EINVAL;
sg_d = &table->sg_table[j = 0];
}
@@ -664,7 +664,7 @@ static int rd_DIRECT_with_offset(
table = rd_get_sg_table(dev, req->rd_page);
if (!(table))
- return -1;
+ return -EINVAL;
table_sg_end = (table->page_end_offset - req->rd_page);
sg_s = &table->sg_table[req->rd_page - table->page_start_offset];
@@ -678,7 +678,7 @@ static int rd_DIRECT_with_offset(
se_mem = kmem_cache_zalloc(se_mem_cache, GFP_KERNEL);
if (!(se_mem)) {
printk(KERN_ERR "Unable to allocate struct se_mem\n");
- return -1;
+ return -ENOMEM;
}
INIT_LIST_HEAD(&se_mem->se_list);
@@ -734,7 +734,7 @@ check_eot:
#endif
table = rd_get_sg_table(dev, req->rd_page);
if (!(table))
- return -1;
+ return -EINVAL;
sg_s = &table->sg_table[j = 0];
}
@@ -767,7 +767,7 @@ static int rd_DIRECT_without_offset(
table = rd_get_sg_table(dev, req->rd_page);
if (!(table))
- return -1;
+ return -EINVAL;
sg_s = &table->sg_table[req->rd_page - table->page_start_offset];
#ifdef DEBUG_RAMDISK_DR
@@ -780,7 +780,7 @@ static int rd_DIRECT_without_offset(
se_mem = kmem_cache_zalloc(se_mem_cache, GFP_KERNEL);
if (!(se_mem)) {
printk(KERN_ERR "Unable to allocate struct se_mem\n");
- return -1;
+ return -ENOMEM;
}
INIT_LIST_HEAD(&se_mem->se_list);
@@ -816,7 +816,7 @@ static int rd_DIRECT_without_offset(
#endif
table = rd_get_sg_table(dev, req->rd_page);
if (!(table))
- return -1;
+ return -EINVAL;
sg_s = &table->sg_table[j = 0];
}
@@ -876,7 +876,7 @@ static int rd_DIRECT_do_se_mem_map(
if (cmd->data_direction == DMA_TO_DEVICE) {
printk(KERN_ERR "DMA_TO_DEVICE not supported for"
" RAMDISK_DR with task_sg_chaining=1\n");
- return -1;
+ return -ENOSYS;
}
/*
* Special case for if task_sg_chaining is enabled, then
@@ -975,7 +975,7 @@ static ssize_t rd_check_configfs_dev_params(struct se_hba *hba, struct se_subsys
if (!(rd_dev->rd_flags & RDF_HAS_PAGE_COUNT)) {
printk(KERN_INFO "Missing rd_pages= parameter\n");
- return -1;
+ return -EINVAL;
}
return 0;
diff --git a/drivers/target/target_core_tpg.c b/drivers/target/target_core_tpg.c
index 5ec745f..dc024c1 100644
--- a/drivers/target/target_core_tpg.c
+++ b/drivers/target/target_core_tpg.c
@@ -233,7 +233,7 @@ static int core_create_device_list_for_node(struct se_node_acl *nacl)
if (!(nacl->device_list)) {
printk(KERN_ERR "Unable to allocate memory for"
" struct se_node_acl->device_list\n");
- return -1;
+ return -ENOMEM;
}
for (i = 0; i < TRANSPORT_MAX_LUNS_PER_TPG; i++) {
deve = &nacl->device_list[i];
@@ -614,7 +614,7 @@ static int core_tpg_setup_virtual_lun0(struct se_portal_group *se_tpg)
ret = core_tpg_post_addlun(se_tpg, lun, lun_access, dev);
if (ret < 0)
- return -1;
+ return ret;
return 0;
}
@@ -772,8 +772,11 @@ int core_tpg_post_addlun(
u32 lun_access,
void *lun_ptr)
{
- if (core_dev_export(lun_ptr, tpg, lun) < 0)
- return -1;
+ int ret;
+
+ ret = core_dev_export(lun_ptr, tpg, lun);
+ if (ret < 0)
+ return ret;
spin_lock(&tpg->tpg_lun_lock);
lun->lun_access = lun_access;
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 1882251..66c4093 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -234,7 +234,7 @@ int init_se_global(void)
global = kzalloc(sizeof(struct se_global), GFP_KERNEL);
if (!(global)) {
printk(KERN_ERR "Unable to allocate memory for struct se_global\n");
- return -1;
+ return -ENOMEM;
}
INIT_LIST_HEAD(&global->g_lu_gps_list);
@@ -350,7 +350,7 @@ out:
if (t10_alua_tg_pt_gp_mem_cache)
kmem_cache_destroy(t10_alua_tg_pt_gp_mem_cache);
kfree(global);
- return -1;
+ return -ENOMEM;
}
void release_se_global(void)
@@ -444,13 +444,16 @@ static int transport_subsystem_reqmods(void)
int transport_subsystem_check_init(void)
{
+ int ret;
+
if (se_global->g_sub_api_initialized)
return 0;
/*
* Request the loading of known TCM subsystem plugins..
*/
- if (transport_subsystem_reqmods() < 0)
- return -1;
+ ret = transport_subsystem_reqmods();
+ if (ret < 0)
+ return ret;
se_global->g_sub_api_initialized = 1;
return 0;
@@ -810,7 +813,7 @@ static int transport_add_cmd_to_queue(
if (!(qr)) {
printk(KERN_ERR "Unable to allocate memory for"
" struct se_queue_req\n");
- return -1;
+ return -ENOMEM;
}
INIT_LIST_HEAD(&qr->qr_list);
@@ -1404,7 +1407,7 @@ int transport_dump_vpd_assoc(
break;
default:
sprintf(buf+len, "Unknown 0x%02x\n", vpd->association);
- ret = -1;
+ ret = -EINVAL;
break;
}
@@ -1461,7 +1464,7 @@ int transport_dump_vpd_ident_type(
default:
sprintf(buf+len, "Unsupported: 0x%02x\n",
vpd->device_identifier_type);
- ret = -1;
+ ret = -EINVAL;
break;
}
@@ -1511,7 +1514,7 @@ int transport_dump_vpd_ident(
default:
sprintf(buf, "T10 VPD Device Identifier encoding unsupported:"
" 0x%02x", vpd->device_identifier_code_set);
- ret = -1;
+ ret = -EINVAL;
break;
}
@@ -1876,7 +1879,7 @@ static int transport_check_alloc_task_attr(struct se_cmd *cmd)
if (cmd->sam_task_attr == MSG_ACA_TAG) {
DEBUG_STA("SAM Task Attribute ACA"
" emulation is not supported\n");
- return -1;
+ return -EINVAL;
}
/*
* Used to determine when ORDERED commands should go from
@@ -1931,7 +1934,7 @@ int transport_generic_allocate_tasks(
printk(KERN_ERR "Received SCSI CDB with command_size: %d that"
" exceeds SCSI_MAX_VARLEN_CDB_SIZE: %d\n",
scsi_command_size(cdb), SCSI_MAX_VARLEN_CDB_SIZE);
- return -1;
+ return -EINVAL;
}
/*
* If the received CDB is larger than TCM_MAX_COMMAND_SIZE,
@@ -1946,7 +1949,7 @@ int transport_generic_allocate_tasks(
" %u > sizeof(T_TASK(cmd)->__t_task_cdb): %lu ops\n",
scsi_command_size(cdb),
(unsigned long)sizeof(T_TASK(cmd)->__t_task_cdb));
- return -1;
+ return -ENOMEM;
}
} else
T_TASK(cmd)->t_task_cdb = &T_TASK(cmd)->__t_task_cdb[0];
@@ -1989,7 +1992,7 @@ int transport_generic_handle_cdb(
if (!SE_LUN(cmd)) {
dump_stack();
printk(KERN_ERR "SE_LUN(cmd) is NULL\n");
- return -1;
+ return -EINVAL;
}
transport_add_cmd_to_queue(cmd, TRANSPORT_NEW_CMD);
@@ -2008,7 +2011,7 @@ int transport_generic_handle_cdb_map(
if (!SE_LUN(cmd)) {
dump_stack();
printk(KERN_ERR "SE_LUN(cmd) is NULL\n");
- return -1;
+ return -EINVAL;
}
transport_add_cmd_to_queue(cmd, TRANSPORT_NEW_CMD_MAP);
@@ -2030,7 +2033,7 @@ int transport_generic_handle_data(
* in interrupt code, the signal_pending() check is skipped.
*/
if (!in_interrupt() && signal_pending(current))
- return -1;
+ return -EPERM;
/*
* If the received CDB has aleady been ABORTED by the generic
* target engine, we now call transport_check_aborted_status()
@@ -2318,7 +2321,7 @@ transport_generic_allocate_buf(struct se_cmd *cmd, u32 data_length)
buf = kzalloc(data_length, GFP_KERNEL);
if (!(buf)) {
printk(KERN_ERR "Unable to allocate memory for buffer\n");
- return -1;
+ return -ENOMEM;
}
T_TASK(cmd)->t_tasks_se_num = 0;
@@ -2801,7 +2804,7 @@ static inline u32 transport_get_sectors_10(
* XXX_10 is not defined in SSC, throw an exception
*/
if (TRANSPORT(dev)->get_device_type(dev) == TYPE_TAPE) {
- *ret = -1;
+ *ret = -EINVAL;
return 0;
}
@@ -2831,7 +2834,7 @@ static inline u32 transport_get_sectors_12(
* XXX_12 is not defined in SSC, throw an exception
*/
if (TRANSPORT(dev)->get_device_type(dev) == TYPE_TAPE) {
- *ret = -1;
+ *ret = -EINVAL;
return 0;
}
@@ -4405,7 +4408,7 @@ out:
if (se_mem)
__free_pages(se_mem->se_page, 0);
kmem_cache_free(se_mem_cache, se_mem);
- return -1;
+ return -ENOMEM;
}
u32 transport_calc_sg_num(
@@ -4484,7 +4487,7 @@ next:
if (!(task->task_sg)) {
printk(KERN_ERR "Unable to allocate memory for"
" task->task_sg\n");
- return 0;
+ return -ENOMEM;
}
sg_init_table(&task->task_sg[0], task_sg_num_padded);
/*
@@ -4498,7 +4501,7 @@ next:
if (!(task->task_sg_bidi)) {
printk(KERN_ERR "Unable to allocate memory for"
" task->task_sg_bidi\n");
- return 0;
+ return -ENOMEM;
}
sg_init_table(&task->task_sg_bidi[0], task_sg_num_padded);
}
@@ -4600,7 +4603,7 @@ static int transport_map_sg_to_mem(
se_mem = kmem_cache_zalloc(se_mem_cache, GFP_KERNEL);
if (!(se_mem)) {
printk(KERN_ERR "Unable to allocate struct se_mem\n");
- return -1;
+ return -ENOMEM;
}
INIT_LIST_HEAD(&se_mem->se_list);
DEBUG_MEM("sg_to_mem: Starting loop with cmd_size: %u"
@@ -4658,7 +4661,7 @@ int transport_map_mem_to_sg(
if (!sg) {
printk(KERN_ERR "Unable to locate valid struct"
" scatterlist pointer\n");
- return -1;
+ return -EINVAL;
}
while (task_size != 0) {
@@ -5350,7 +5353,7 @@ static int transport_lun_wait_for_tasks(struct se_cmd *cmd, struct se_lun *lun)
" TRUE, skipping\n", CMD_TFO(cmd)->get_task_tag(cmd));
spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
transport_cmd_check_stop(cmd, 1, 0);
- return -1;
+ return -EPERM;
}
atomic_set(&T_TASK(cmd)->transport_lun_fe_stop, 1);
spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
@@ -5506,7 +5509,7 @@ int transport_clear_lun_from_sessions(struct se_lun *lun)
"tcm_cl_%u", lun->unpacked_lun);
if (IS_ERR(kt)) {
printk(KERN_ERR "Unable to start clear_lun thread\n");
- return -1;
+ return PTR_ERR(kt);
}
wait_for_completion(&lun->lun_shutdown_comp);
diff --git a/drivers/target/target_core_ua.c b/drivers/target/target_core_ua.c
index df35517..4f754ca 100644
--- a/drivers/target/target_core_ua.c
+++ b/drivers/target/target_core_ua.c
@@ -80,10 +80,10 @@ int core_scsi3_ua_check(
case REQUEST_SENSE:
return 0;
default:
- return -1;
+ return -EINVAL;
}
- return -1;
+ return -EINVAL;
}
int core_scsi3_ua_allocate(
@@ -98,12 +98,12 @@ int core_scsi3_ua_allocate(
* PASSTHROUGH OPS
*/
if (!(nacl))
- return -1;
+ return -EINVAL;
ua = kmem_cache_zalloc(se_ua_cache, GFP_ATOMIC);
if (!(ua)) {
printk(KERN_ERR "Unable to allocate struct se_ua\n");
- return -1;
+ return -ENOMEM;
}
INIT_LIST_HEAD(&ua->ua_dev_list);
INIT_LIST_HEAD(&ua->ua_nacl_list);
@@ -285,17 +285,17 @@ int core_scsi3_ua_clear_for_request_sense(
int head = 1;
if (!(sess))
- return -1;
+ return -EINVAL;
nacl = sess->se_node_acl;
if (!(nacl))
- return -1;
+ return -EINVAL;
spin_lock_irq(&nacl->device_list_lock);
deve = &nacl->device_list[cmd->orig_fe_lun];
if (!(atomic_read(&deve->ua_count))) {
spin_unlock_irq(&nacl->device_list_lock);
- return -1;
+ return -EPERM;
}
/*
* The highest priority Unit Attentions are placed at the head of the
@@ -328,5 +328,5 @@ int core_scsi3_ua_clear_for_request_sense(
" ASCQ: 0x%02x\n", TPG_TFO(nacl->se_tpg)->get_fabric_name(),
cmd->orig_fe_lun, *asc, *ascq);
- return (head) ? -1 : 0;
+ return (head) ? -EPERM : 0;
}
diff --git a/drivers/target/tcm_fc/tfc_conf.c b/drivers/target/tcm_fc/tfc_conf.c
index 84e868c..8c5067c 100644
--- a/drivers/target/tcm_fc/tfc_conf.c
+++ b/drivers/target/tcm_fc/tfc_conf.c
@@ -582,10 +582,10 @@ int ft_register_configfs(void)
* Register the top level struct config_item_type with TCM core
*/
fabric = target_fabric_configfs_init(THIS_MODULE, "fc");
- if (!fabric) {
+ if (IS_ERR(fabric)) {
printk(KERN_INFO "%s: target_fabric_configfs_init() failed!\n",
__func__);
- return -1;
+ return PTR_ERR(fabric);
}
fabric->tf_ops = ft_fabric_ops;
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 008/103] target: Rename transport_calc_sg_num to transport_init_task_sg
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (6 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 007/103] target: use errno values instead of returning -1 for everything Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 009/103] target: Fix leak in error path in transport_init_task_sg Nicholas A. Bellinger
` (21 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover, Nicholas A. Bellinger
From: Andy Grover <agrover@redhat.com>
The function doesn't just calculate how many sgs there are, it also
allocates and initializes the task_sg.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_rd.c | 7 ++++---
drivers/target/target_core_transport.c | 15 ++++++++-------
include/target/target_core_transport.h | 2 +-
3 files changed, 13 insertions(+), 11 deletions(-)
diff --git a/drivers/target/target_core_rd.c b/drivers/target/target_core_rd.c
index 793a94b..2a00f70 100644
--- a/drivers/target/target_core_rd.c
+++ b/drivers/target/target_core_rd.c
@@ -884,11 +884,12 @@ static int rd_DIRECT_do_se_mem_map(
* transport_do_task_sg_chain() for creating chainged SGLs
* across multiple struct se_task->task_sg[].
*/
- if (!(transport_calc_sg_num(task,
+ ret = transport_init_task_sg(task,
list_entry(T_TASK(cmd)->t_mem_list->next,
struct se_mem, se_list),
- task_offset)))
- return -1;
+ task_offset);
+ if (ret <= 0)
+ return ret;
return transport_map_mem_to_sg(task, se_mem_list, task->task_sg,
list_entry(T_TASK(cmd)->t_mem_list->next,
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 66c4093..bae584d 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -4411,7 +4411,7 @@ out:
return -ENOMEM;
}
-u32 transport_calc_sg_num(
+int transport_init_task_sg(
struct se_task *task,
struct se_mem *in_se_mem,
u32 task_offset)
@@ -4786,7 +4786,7 @@ void transport_do_task_sg_chain(struct se_cmd *cmd)
&T_TASK(cmd)->t_task_list))) {
/*
* Clear existing SGL termination bit set in
- * transport_calc_sg_num(), see sg_mark_end()
+ * transport_init_task_sg(), see sg_mark_end()
*/
sg_end_cur = &task->task_sg[task->task_sg_num - 1];
sg_end_cur->page_link &= ~0x02;
@@ -4812,7 +4812,7 @@ void transport_do_task_sg_chain(struct se_cmd *cmd)
if (!(list_is_last(&task->t_list, &T_TASK(cmd)->t_task_list))) {
/*
* Clear existing SGL termination bit set in
- * transport_calc_sg_num(), see sg_mark_end()
+ * transport_init_task_sg(), see sg_mark_end()
*/
sg_end = &task->task_sg[task->task_sg_num - 1];
sg_end->page_link &= ~0x02;
@@ -4878,7 +4878,7 @@ static int transport_do_se_mem_map(
* This is the normal path for all normal non BIDI and BIDI-COMMAND
* WRITE payloads.. If we need to do BIDI READ passthrough for
* TCM/pSCSI the first call to transport_do_se_mem_map ->
- * transport_calc_sg_num() -> transport_map_mem_to_sg() will do the
+ * transport_init_task_sg() -> transport_map_mem_to_sg() will do the
* allocation for task->task_sg_bidi, and the subsequent call to
* transport_do_se_mem_map() from transport_generic_get_cdb_count()
*/
@@ -4887,8 +4887,9 @@ static int transport_do_se_mem_map(
* Assume default that transport plugin speaks preallocated
* scatterlists.
*/
- if (!(transport_calc_sg_num(task, in_se_mem, task_offset)))
- return -1;
+ ret = transport_init_task_sg(task, in_se_mem, task_offset);
+ if (ret <= 0)
+ return ret;
/*
* struct se_task->task_sg now contains the struct scatterlist array.
*/
@@ -4986,7 +4987,7 @@ static u32 transport_generic_get_cdb_count(
*
* Note that the first call to transport_do_se_mem_map() above will
* allocate struct se_task->task_sg_bidi in transport_do_se_mem_map()
- * -> transport_calc_sg_num(), and the second here will do the
+ * -> transport_init_task_sg(), and the second here will do the
* mapping for SCSI READ for BIDI-COMMAND passthrough with TCM/pSCSI.
*/
if (task->task_sg_bidi != NULL) {
diff --git a/include/target/target_core_transport.h b/include/target/target_core_transport.h
index 24a1c6c..199a404 100644
--- a/include/target/target_core_transport.h
+++ b/include/target/target_core_transport.h
@@ -184,7 +184,7 @@ extern void transport_send_task_abort(struct se_cmd *);
extern void transport_release_cmd_to_pool(struct se_cmd *);
extern void transport_generic_free_cmd(struct se_cmd *, int, int, int);
extern void transport_generic_wait_for_cmds(struct se_cmd *, int);
-extern u32 transport_calc_sg_num(struct se_task *, struct se_mem *, u32);
+extern int transport_init_task_sg(struct se_task *, struct se_mem *, u32);
extern int transport_map_mem_to_sg(struct se_task *, struct list_head *,
void *, struct se_mem *,
struct se_mem **, u32 *, u32 *);
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 009/103] target: Fix leak in error path in transport_init_task_sg
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (7 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 008/103] target: Rename transport_calc_sg_num to transport_init_task_sg Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 010/103] target: Remove pscsi_get_sh() in _pscsi.c Nicholas A. Bellinger
` (20 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover, Nicholas A. Bellinger
From: Andy Grover <agrover@redhat.com>
If the function fails to allocate mem for task_sg_bidi, it would
return with an error, and the task_sg would be allocated but only
partially initialized. This patch frees the task_sg in that scenario.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_transport.c | 2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index bae584d..edf24b4 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -4499,6 +4499,8 @@ next:
task->task_sg_bidi = kzalloc(task_sg_num_padded *
sizeof(struct scatterlist), GFP_KERNEL);
if (!(task->task_sg_bidi)) {
+ kfree(task->task_sg);
+ task->task_sg = NULL;
printk(KERN_ERR "Unable to allocate memory for"
" task->task_sg_bidi\n");
return -ENOMEM;
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 010/103] target: Remove pscsi_get_sh() in _pscsi.c
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (8 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 009/103] target: Fix leak in error path in transport_init_task_sg Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 011/103] target: Make two runtime checks into WARN_ONs Nicholas A. Bellinger
` (19 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover, Nicholas A. Bellinger
From: Andy Grover <agrover@redhat.com>
It only wrapped scsi_host_lookup, which can just be used directly
without too much trouble.
(nab: Fix pdv_host_id usage)
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_pscsi.c | 30 ++++++------------------------
1 files changed, 6 insertions(+), 24 deletions(-)
diff --git a/drivers/target/target_core_pscsi.c b/drivers/target/target_core_pscsi.c
index fb2e8cb..d39226f 100644
--- a/drivers/target/target_core_pscsi.c
+++ b/drivers/target/target_core_pscsi.c
@@ -55,24 +55,6 @@ static struct se_subsystem_api pscsi_template;
static void pscsi_req_done(struct request *, int);
-/* pscsi_get_sh():
- *
- *
- */
-static struct Scsi_Host *pscsi_get_sh(u32 host_no)
-{
- struct Scsi_Host *sh = NULL;
-
- sh = scsi_host_lookup(host_no);
- if (IS_ERR(sh)) {
- printk(KERN_ERR "Unable to locate SCSI HBA with Host ID:"
- " %u\n", host_no);
- return NULL;
- }
-
- return sh;
-}
-
/* pscsi_attach_hba():
*
* pscsi_get_sh() used scsi_host_lookup() to locate struct Scsi_Host.
@@ -154,11 +136,11 @@ static int pscsi_pmode_enable_hba(struct se_hba *hba, unsigned long mode_flag)
* Otherwise, locate struct Scsi_Host from the original passed
* pSCSI Host ID and enable for phba mode
*/
- sh = pscsi_get_sh(phv->phv_host_id);
- if (!(sh)) {
+ sh = scsi_host_lookup(phv->phv_host_id);
+ if (IS_ERR(sh)) {
printk(KERN_ERR "pSCSI: Unable to locate SCSI Host for"
" phv_host_id: %d\n", phv->phv_host_id);
- return -1;
+ return PTR_ERR(sh);
}
/*
* Usually the SCSI LLD will use the hostt->can_queue value to define
@@ -601,11 +583,11 @@ static struct se_device *pscsi_create_virtdevice(
hba->hba_flags |= HBA_FLAGS_PSCSI_MODE;
sh = phv->phv_lld_host;
} else {
- sh = pscsi_get_sh(pdv->pdv_host_id);
- if (!(sh)) {
+ sh = scsi_host_lookup(pdv->pdv_host_id);
+ if (IS_ERR(sh)) {
printk(KERN_ERR "pSCSI: Unable to locate"
" pdv_host_id: %d\n", pdv->pdv_host_id);
- return ERR_PTR(-ENODEV);
+ return (struct se_device *) sh;
}
}
} else {
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 011/103] target: Make two runtime checks into WARN_ONs
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (9 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 010/103] target: Remove pscsi_get_sh() in _pscsi.c Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 012/103] tcm_loop: off by one in tcm_loop_make_naa_tpg() Nicholas A. Bellinger
` (18 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover, Nicholas A. Bellinger
From: Andy Grover <agrover@redhat.com>
The params are checked in the calling functions, so these ones in
transport_get_sense_data and transport_map_sg_to_mem should be unnecessary,
unless calling code becomes buggy.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_transport.c | 12 ++++--------
1 files changed, 4 insertions(+), 8 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index edf24b4..99b9e33 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -2997,10 +2997,8 @@ static int transport_get_sense_data(struct se_cmd *cmd)
unsigned long flags;
u32 offset = 0;
- if (!SE_LUN(cmd)) {
- printk(KERN_ERR "SE_LUN(cmd) is NULL\n");
- return -1;
- }
+ WARN_ON(!SE_LUN(cmd));
+
spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
if (cmd->se_cmd_flags & SCF_SENT_CHECK_CONDITION) {
spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
@@ -4595,10 +4593,8 @@ static int transport_map_sg_to_mem(
struct scatterlist *sg;
u32 sg_count = 1, cmd_size = cmd->data_length;
- if (!in_mem) {
- printk(KERN_ERR "No source scatterlist\n");
- return -1;
- }
+ WARN_ON(!in_mem);
+
sg = (struct scatterlist *)in_mem;
while (cmd_size) {
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 012/103] tcm_loop: off by one in tcm_loop_make_naa_tpg()
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (10 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 011/103] target: Make two runtime checks into WARN_ONs Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 013/103] target: Remove hba queue depth and convert to spin_lock_irq usage Nicholas A. Bellinger
` (17 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Dan Carpenter, Nicholas A. Bellinger
From: Dan Carpenter <error27@gmail.com>
This is an off by one 'tgpt' check in tcm_loop_make_naa_tpg() that could result
in memory corruption.
Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/loopback/tcm_loop.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/target/loopback/tcm_loop.c b/drivers/target/loopback/tcm_loop.c
index 3274fec..26e3681 100644
--- a/drivers/target/loopback/tcm_loop.c
+++ b/drivers/target/loopback/tcm_loop.c
@@ -1230,7 +1230,7 @@ struct se_portal_group *tcm_loop_make_naa_tpg(
tpgt_str += 5; /* Skip ahead of "tpgt_" */
tpgt = (unsigned short int) simple_strtoul(tpgt_str, &end_ptr, 0);
- if (tpgt > TL_TPGS_PER_HBA) {
+ if (tpgt >= TL_TPGS_PER_HBA) {
printk(KERN_ERR "Passed tpgt: %hu exceeds TL_TPGS_PER_HBA:"
" %u\n", tpgt, TL_TPGS_PER_HBA);
return ERR_PTR(-EINVAL);
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 013/103] target: Remove hba queue depth and convert to spin_lock_irq usage
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (11 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 012/103] tcm_loop: off by one in tcm_loop_make_naa_tpg() Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 014/103] target: dev->dev_status_queue_obj is unused Nicholas A. Bellinger
` (16 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover, Nicholas A. Bellinger
From: Andy Grover <agrover@redhat.com>
The queue depth in the virtual hba is no longer needed, we can just
rely on the per-device queue depth limits.
Remove hba_queue_lock, since updating a single atomic no longer needs
protection. This patch also converts se_dev->execute_task_lock usage in
__transport_execute_tasks to use execute_task_lock with spin_lock_irq
Remove load_balance_queue, unused.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_file.c | 7 +----
drivers/target/target_core_file.h | 2 -
drivers/target/target_core_hba.c | 4 ---
drivers/target/target_core_iblock.c | 7 +----
drivers/target/target_core_iblock.h | 1 -
drivers/target/target_core_pscsi.c | 22 +------------------
drivers/target/target_core_pscsi.h | 1 -
drivers/target/target_core_rd.c | 7 +----
drivers/target/target_core_rd.h | 2 -
drivers/target/target_core_transport.c | 35 ++++++-------------------------
include/target/target_core_base.h | 5 ----
11 files changed, 15 insertions(+), 78 deletions(-)
diff --git a/drivers/target/target_core_file.c b/drivers/target/target_core_file.c
index 40386ec..5ca1f5b 100644
--- a/drivers/target/target_core_file.c
+++ b/drivers/target/target_core_file.c
@@ -72,17 +72,14 @@ static int fd_attach_hba(struct se_hba *hba, u32 host_id)
fd_host->fd_host_id = host_id;
- atomic_set(&hba->left_queue_depth, FD_HBA_QUEUE_DEPTH);
- atomic_set(&hba->max_queue_depth, FD_HBA_QUEUE_DEPTH);
hba->hba_ptr = (void *) fd_host;
printk(KERN_INFO "CORE_HBA[%d] - TCM FILEIO HBA Driver %s on Generic"
" Target Core Stack %s\n", hba->hba_id, FD_VERSION,
TARGET_CORE_MOD_VERSION);
printk(KERN_INFO "CORE_HBA[%d] - Attached FILEIO HBA: %u to Generic"
- " Target Core with TCQ Depth: %d MaxSectors: %u\n",
- hba->hba_id, fd_host->fd_host_id,
- atomic_read(&hba->max_queue_depth), FD_MAX_SECTORS);
+ " MaxSectors: %u\n",
+ hba->hba_id, fd_host->fd_host_id, FD_MAX_SECTORS);
return 0;
}
diff --git a/drivers/target/target_core_file.h b/drivers/target/target_core_file.h
index ef4de2b..6386d3f 100644
--- a/drivers/target/target_core_file.h
+++ b/drivers/target/target_core_file.h
@@ -4,8 +4,6 @@
#define FD_VERSION "4.0"
#define FD_MAX_DEV_NAME 256
-/* Maximum queuedepth for the FILEIO HBA */
-#define FD_HBA_QUEUE_DEPTH 256
#define FD_DEVICE_QUEUE_DEPTH 32
#define FD_MAX_DEVICE_QUEUE_DEPTH 128
#define FD_BLOCKSIZE 512
diff --git a/drivers/target/target_core_hba.c b/drivers/target/target_core_hba.c
index ddf5ae0..8f3a0c4 100644
--- a/drivers/target/target_core_hba.c
+++ b/drivers/target/target_core_hba.c
@@ -110,15 +110,11 @@ core_alloc_hba(const char *plugin_name, u32 plugin_dep_id, u32 hba_flags)
INIT_LIST_HEAD(&hba->hba_dev_list);
spin_lock_init(&hba->device_lock);
- spin_lock_init(&hba->hba_queue_lock);
mutex_init(&hba->hba_access_mutex);
hba->hba_index = scsi_get_new_index(SCSI_INST_INDEX);
hba->hba_flags |= hba_flags;
- atomic_set(&hba->max_queue_depth, 0);
- atomic_set(&hba->left_queue_depth, 0);
-
hba->transport = core_get_backend(plugin_name);
if (!hba->transport) {
ret = -EINVAL;
diff --git a/drivers/target/target_core_iblock.c b/drivers/target/target_core_iblock.c
index 5d5bd6e..5bd4f1c 100644
--- a/drivers/target/target_core_iblock.c
+++ b/drivers/target/target_core_iblock.c
@@ -74,17 +74,14 @@ static int iblock_attach_hba(struct se_hba *hba, u32 host_id)
ib_host->iblock_host_id = host_id;
- atomic_set(&hba->left_queue_depth, IBLOCK_HBA_QUEUE_DEPTH);
- atomic_set(&hba->max_queue_depth, IBLOCK_HBA_QUEUE_DEPTH);
hba->hba_ptr = (void *) ib_host;
printk(KERN_INFO "CORE_HBA[%d] - TCM iBlock HBA Driver %s on"
" Generic Target Core Stack %s\n", hba->hba_id,
IBLOCK_VERSION, TARGET_CORE_MOD_VERSION);
- printk(KERN_INFO "CORE_HBA[%d] - Attached iBlock HBA: %u to Generic"
- " Target Core TCQ Depth: %d\n", hba->hba_id,
- ib_host->iblock_host_id, atomic_read(&hba->max_queue_depth));
+ printk(KERN_INFO "CORE_HBA[%d] - Attached iBlock HBA: %u to Generic\n",
+ hba->hba_id, ib_host->iblock_host_id);
return 0;
}
diff --git a/drivers/target/target_core_iblock.h b/drivers/target/target_core_iblock.h
index 64c1f4d..6b6d17b 100644
--- a/drivers/target/target_core_iblock.h
+++ b/drivers/target/target_core_iblock.h
@@ -3,7 +3,6 @@
#define IBLOCK_VERSION "4.0"
-#define IBLOCK_HBA_QUEUE_DEPTH 512
#define IBLOCK_DEVICE_QUEUE_DEPTH 32
#define IBLOCK_MAX_DEVICE_QUEUE_DEPTH 128
#define IBLOCK_MAX_CDBS 16
diff --git a/drivers/target/target_core_pscsi.c b/drivers/target/target_core_pscsi.c
index d39226f..e221dea 100644
--- a/drivers/target/target_core_pscsi.c
+++ b/drivers/target/target_core_pscsi.c
@@ -62,7 +62,6 @@ static void pscsi_req_done(struct request *, int);
*/
static int pscsi_attach_hba(struct se_hba *hba, u32 host_id)
{
- int hba_depth;
struct pscsi_hba_virt *phv;
phv = kzalloc(sizeof(struct pscsi_hba_virt), GFP_KERNEL);
@@ -72,18 +71,14 @@ static int pscsi_attach_hba(struct se_hba *hba, u32 host_id)
}
phv->phv_host_id = host_id;
phv->phv_mode = PHV_VIRUTAL_HOST_ID;
- hba_depth = PSCSI_VIRTUAL_HBA_DEPTH;
- atomic_set(&hba->left_queue_depth, hba_depth);
- atomic_set(&hba->max_queue_depth, hba_depth);
hba->hba_ptr = (void *)phv;
printk(KERN_INFO "CORE_HBA[%d] - TCM SCSI HBA Driver %s on"
" Generic Target Core Stack %s\n", hba->hba_id,
PSCSI_VERSION, TARGET_CORE_MOD_VERSION);
- printk(KERN_INFO "CORE_HBA[%d] - Attached SCSI HBA to Generic"
- " Target Core with TCQ Depth: %d\n", hba->hba_id,
- atomic_read(&hba->max_queue_depth));
+ printk(KERN_INFO "CORE_HBA[%d] - Attached SCSI HBA to Generic\n",
+ hba->hba_id);
return 0;
}
@@ -112,7 +107,6 @@ static int pscsi_pmode_enable_hba(struct se_hba *hba, unsigned long mode_flag)
{
struct pscsi_hba_virt *phv = (struct pscsi_hba_virt *)hba->hba_ptr;
struct Scsi_Host *sh = phv->phv_lld_host;
- int hba_depth = PSCSI_VIRTUAL_HBA_DEPTH;
/*
* Release the struct Scsi_Host
*/
@@ -122,8 +116,6 @@ static int pscsi_pmode_enable_hba(struct se_hba *hba, unsigned long mode_flag)
phv->phv_lld_host = NULL;
phv->phv_mode = PHV_VIRUTAL_HOST_ID;
- atomic_set(&hba->left_queue_depth, hba_depth);
- atomic_set(&hba->max_queue_depth, hba_depth);
printk(KERN_INFO "CORE_HBA[%d] - Disabled pSCSI HBA Passthrough"
" %s\n", hba->hba_id, (sh->hostt->name) ?
@@ -142,16 +134,6 @@ static int pscsi_pmode_enable_hba(struct se_hba *hba, unsigned long mode_flag)
" phv_host_id: %d\n", phv->phv_host_id);
return PTR_ERR(sh);
}
- /*
- * Usually the SCSI LLD will use the hostt->can_queue value to define
- * its HBA TCQ depth. Some other drivers (like 2.6 megaraid) don't set
- * this at all and set sh->can_queue at runtime.
- */
- hba_depth = (sh->hostt->can_queue > sh->can_queue) ?
- sh->hostt->can_queue : sh->can_queue;
-
- atomic_set(&hba->left_queue_depth, hba_depth);
- atomic_set(&hba->max_queue_depth, hba_depth);
phv->phv_lld_host = sh;
phv->phv_mode = PHV_LLD_SCSI_HOST_NO;
diff --git a/drivers/target/target_core_pscsi.h b/drivers/target/target_core_pscsi.h
index a4cd5d3..280b689 100644
--- a/drivers/target/target_core_pscsi.h
+++ b/drivers/target/target_core_pscsi.h
@@ -2,7 +2,6 @@
#define TARGET_CORE_PSCSI_H
#define PSCSI_VERSION "v4.0"
-#define PSCSI_VIRTUAL_HBA_DEPTH 2048
/* used in pscsi_find_alloc_len() */
#ifndef INQUIRY_DATA_SIZE
diff --git a/drivers/target/target_core_rd.c b/drivers/target/target_core_rd.c
index 2a00f70..0a424bb 100644
--- a/drivers/target/target_core_rd.c
+++ b/drivers/target/target_core_rd.c
@@ -66,17 +66,14 @@ static int rd_attach_hba(struct se_hba *hba, u32 host_id)
rd_host->rd_host_id = host_id;
- atomic_set(&hba->left_queue_depth, RD_HBA_QUEUE_DEPTH);
- atomic_set(&hba->max_queue_depth, RD_HBA_QUEUE_DEPTH);
hba->hba_ptr = (void *) rd_host;
printk(KERN_INFO "CORE_HBA[%d] - TCM Ramdisk HBA Driver %s on"
" Generic Target Core Stack %s\n", hba->hba_id,
RD_HBA_VERSION, TARGET_CORE_MOD_VERSION);
printk(KERN_INFO "CORE_HBA[%d] - Attached Ramdisk HBA: %u to Generic"
- " Target Core TCQ Depth: %d MaxSectors: %u\n", hba->hba_id,
- rd_host->rd_host_id, atomic_read(&hba->max_queue_depth),
- RD_MAX_SECTORS);
+ " MaxSectors: %u\n", hba->hba_id,
+ rd_host->rd_host_id, RD_MAX_SECTORS);
return 0;
}
diff --git a/drivers/target/target_core_rd.h b/drivers/target/target_core_rd.h
index 3ea19e2..bab9302 100644
--- a/drivers/target/target_core_rd.h
+++ b/drivers/target/target_core_rd.h
@@ -7,8 +7,6 @@
/* Largest piece of memory kmalloc can allocate */
#define RD_MAX_ALLOCATION_SIZE 65536
-/* Maximum queuedepth for the Ramdisk HBA */
-#define RD_HBA_QUEUE_DEPTH 256
#define RD_DEVICE_QUEUE_DEPTH 32
#define RD_MAX_DEVICE_QUEUE_DEPTH 128
#define RD_BLOCKSIZE 512
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 99b9e33..ae153bf 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -959,12 +959,8 @@ void transport_complete_task(struct se_task *task, int success)
printk(KERN_INFO "task: %p CDB: 0x%02x obj_ptr: %p\n", task,
T_TASK(cmd)->t_task_cdb[0], dev);
#endif
- if (dev) {
- spin_lock_irqsave(&SE_HBA(dev)->hba_queue_lock, flags);
+ if (dev)
atomic_inc(&dev->depth_left);
- atomic_inc(&SE_HBA(dev)->left_queue_depth);
- spin_unlock_irqrestore(&SE_HBA(dev)->hba_queue_lock, flags);
- }
spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
atomic_set(&task->task_active, 0);
@@ -2140,16 +2136,6 @@ static int transport_stop_tasks_for_cmd(struct se_cmd *cmd)
return ret;
}
-static void transport_failure_reset_queue_depth(struct se_device *dev)
-{
- unsigned long flags;
-
- spin_lock_irqsave(&SE_HBA(dev)->hba_queue_lock, flags);
- atomic_inc(&dev->depth_left);
- atomic_inc(&SE_HBA(dev)->left_queue_depth);
- spin_unlock_irqrestore(&SE_HBA(dev)->hba_queue_lock, flags);
-}
-
/*
* Handle SAM-esque emulation for generic transport request failures.
*/
@@ -2181,7 +2167,7 @@ static void transport_generic_request_failure(
transport_stop_all_task_timers(cmd);
if (dev)
- transport_failure_reset_queue_depth(dev);
+ atomic_inc(&dev->depth_left);
/*
* For SAM Task Attribute emulation for failed struct se_cmd
*/
@@ -2650,26 +2636,19 @@ static int __transport_execute_tasks(struct se_device *dev)
* struct se_transport_task's to the selected transport.
*/
check_depth:
- spin_lock_irqsave(&SE_HBA(dev)->hba_queue_lock, flags);
- if (!(atomic_read(&dev->depth_left)) ||
- !(atomic_read(&SE_HBA(dev)->left_queue_depth))) {
- spin_unlock_irqrestore(&SE_HBA(dev)->hba_queue_lock, flags);
+ if (!atomic_read(&dev->depth_left))
return transport_tcq_window_closed(dev);
- }
+
dev->dev_tcq_window_closed = 0;
- spin_lock(&dev->execute_task_lock);
+ spin_lock_irq(&dev->execute_task_lock);
task = transport_get_task_from_execute_queue(dev);
- spin_unlock(&dev->execute_task_lock);
+ spin_unlock_irq(&dev->execute_task_lock);
- if (!task) {
- spin_unlock_irqrestore(&SE_HBA(dev)->hba_queue_lock, flags);
+ if (!task)
return 0;
- }
atomic_dec(&dev->depth_left);
- atomic_dec(&SE_HBA(dev)->left_queue_depth);
- spin_unlock_irqrestore(&SE_HBA(dev)->hba_queue_lock, flags);
cmd = TASK_CMD(task);
diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h
index 561ac99..e35e9e4 100644
--- a/include/target/target_core_base.h
+++ b/include/target/target_core_base.h
@@ -837,17 +837,12 @@ struct se_hba {
/* Virtual iSCSI devices attached. */
u32 dev_count;
u32 hba_index;
- atomic_t load_balance_queue;
- atomic_t left_queue_depth;
- /* Maximum queue depth the HBA can handle. */
- atomic_t max_queue_depth;
/* Pointer to transport specific host structure. */
void *hba_ptr;
/* Linked list for struct se_device */
struct list_head hba_dev_list;
struct list_head hba_list;
spinlock_t device_lock;
- spinlock_t hba_queue_lock;
struct config_group hba_group;
struct mutex hba_access_mutex;
struct se_subsystem_api *transport;
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 014/103] target: dev->dev_status_queue_obj is unused
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (12 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 013/103] target: Remove hba queue depth and convert to spin_lock_irq usage Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 015/103] target: Make struct se_queue_req.cmd type struct se_cmd * Nicholas A. Bellinger
` (15 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover, Nicholas A. Bellinger
From: Andy Grover <agrover@redhat.com>
Looks safe to remove.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_device.c | 1 -
drivers/target/target_core_transport.c | 12 ------------
include/target/target_core_base.h | 1 -
3 files changed, 0 insertions(+), 14 deletions(-)
diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c
index 818c65b..f03e0cc 100644
--- a/drivers/target/target_core_device.c
+++ b/drivers/target/target_core_device.c
@@ -755,7 +755,6 @@ void se_release_device_for_hba(struct se_device *dev)
core_scsi3_free_all_registrations(dev);
se_release_vpd_for_dev(dev);
- kfree(dev->dev_status_queue_obj);
kfree(dev->dev_queue_obj);
kfree(dev);
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index ae153bf..fffa905 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -1642,17 +1642,6 @@ struct se_device *transport_add_device_to_core_hba(
}
transport_init_queue_obj(dev->dev_queue_obj);
- dev->dev_status_queue_obj = kzalloc(sizeof(struct se_queue_obj),
- GFP_KERNEL);
- if (!(dev->dev_status_queue_obj)) {
- printk(KERN_ERR "Unable to allocate memory for"
- " dev->dev_status_queue_obj\n");
- kfree(dev->dev_queue_obj);
- kfree(dev);
- return NULL;
- }
- transport_init_queue_obj(dev->dev_status_queue_obj);
-
dev->dev_flags = device_flags;
dev->dev_status |= TRANSPORT_DEVICE_DEACTIVATED;
dev->dev_ptr = (void *) transport_dev;
@@ -1753,7 +1742,6 @@ out:
se_release_vpd_for_dev(dev);
- kfree(dev->dev_status_queue_obj);
kfree(dev->dev_queue_obj);
kfree(dev);
diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h
index e35e9e4..b2b4cf9 100644
--- a/include/target/target_core_base.h
+++ b/include/target/target_core_base.h
@@ -784,7 +784,6 @@ struct se_device {
struct se_obj dev_access_obj;
struct se_obj dev_export_obj;
struct se_queue_obj *dev_queue_obj;
- struct se_queue_obj *dev_status_queue_obj;
spinlock_t delayed_cmd_lock;
spinlock_t ordered_cmd_lock;
spinlock_t execute_task_lock;
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 015/103] target: Make struct se_queue_req.cmd type struct se_cmd *
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (13 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 014/103] target: dev->dev_status_queue_obj is unused Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 016/103] target: Remove __transport_get_qr_from_queue() Nicholas A. Bellinger
` (14 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover, Nicholas A. Bellinger
From: Andy Grover <agrover@redhat.com>
Doesn't appear to ever be used as anything but a se_cmd*, so
it's safer to make it a pointer to a type.
Simplifies a few functions.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_transport.c | 30 ++++++++++++------------------
include/target/target_core_base.h | 2 +-
2 files changed, 13 insertions(+), 19 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index fffa905..002ea6f 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -817,7 +817,7 @@ static int transport_add_cmd_to_queue(
}
INIT_LIST_HEAD(&qr->qr_list);
- qr->cmd = (void *)cmd;
+ qr->cmd = cmd;
qr->state = t_state;
if (t_state) {
@@ -843,7 +843,6 @@ static int transport_add_cmd_to_queue(
static struct se_queue_req *
__transport_get_qr_from_queue(struct se_queue_obj *qobj)
{
- struct se_cmd *cmd;
struct se_queue_req *qr = NULL;
if (list_empty(&qobj->qobj_list))
@@ -852,10 +851,9 @@ __transport_get_qr_from_queue(struct se_queue_obj *qobj)
list_for_each_entry(qr, &qobj->qobj_list, qr_list)
break;
- if (qr->cmd) {
- cmd = (struct se_cmd *)qr->cmd;
- atomic_dec(&T_TASK(cmd)->t_transport_queue_active);
- }
+ if (qr->cmd)
+ atomic_dec(&T_TASK(qr->cmd)->t_transport_queue_active);
+
list_del(&qr->qr_list);
atomic_dec(&qobj->queue_cnt);
@@ -865,7 +863,6 @@ __transport_get_qr_from_queue(struct se_queue_obj *qobj)
static struct se_queue_req *
transport_get_qr_from_queue(struct se_queue_obj *qobj)
{
- struct se_cmd *cmd;
struct se_queue_req *qr;
unsigned long flags;
@@ -878,10 +875,9 @@ transport_get_qr_from_queue(struct se_queue_obj *qobj)
list_for_each_entry(qr, &qobj->qobj_list, qr_list)
break;
- if (qr->cmd) {
- cmd = (struct se_cmd *)qr->cmd;
- atomic_dec(&T_TASK(cmd)->t_transport_queue_active);
- }
+ if (qr->cmd)
+ atomic_dec(&T_TASK(qr->cmd)->t_transport_queue_active);
+
list_del(&qr->qr_list);
atomic_dec(&qobj->queue_cnt);
spin_unlock_irqrestore(&qobj->cmd_queue_lock, flags);
@@ -892,7 +888,6 @@ transport_get_qr_from_queue(struct se_queue_obj *qobj)
static void transport_remove_cmd_from_queue(struct se_cmd *cmd,
struct se_queue_obj *qobj)
{
- struct se_cmd *q_cmd;
struct se_queue_req *qr = NULL, *qr_p = NULL;
unsigned long flags;
@@ -903,11 +898,10 @@ static void transport_remove_cmd_from_queue(struct se_cmd *cmd,
}
list_for_each_entry_safe(qr, qr_p, &qobj->qobj_list, qr_list) {
- q_cmd = (struct se_cmd *)qr->cmd;
- if (q_cmd != cmd)
+ if (qr->cmd != cmd)
continue;
- atomic_dec(&T_TASK(q_cmd)->t_transport_queue_active);
+ atomic_dec(&T_TASK(qr->cmd)->t_transport_queue_active);
atomic_dec(&qobj->queue_cnt);
list_del(&qr->qr_list);
kfree(qr);
@@ -1287,7 +1281,7 @@ static void transport_release_all_cmds(struct se_device *dev)
list_for_each_entry_safe(qr, qr_p, &dev->dev_queue_obj->qobj_list,
qr_list) {
- cmd = (struct se_cmd *)qr->cmd;
+ cmd = qr->cmd;
t_state = qr->state;
list_del(&qr->qr_list);
kfree(qr);
@@ -6023,7 +6017,7 @@ static void transport_processing_shutdown(struct se_device *dev)
while ((qr = __transport_get_qr_from_queue(dev->dev_queue_obj))) {
spin_unlock_irqrestore(
&dev->dev_queue_obj->cmd_queue_lock, flags);
- cmd = (struct se_cmd *)qr->cmd;
+ cmd = qr->cmd;
state = qr->state;
kfree(qr);
@@ -6081,7 +6075,7 @@ get_cmd:
if (!(qr))
continue;
- cmd = (struct se_cmd *)qr->cmd;
+ cmd = qr->cmd;
t_state = qr->state;
kfree(qr);
diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h
index b2b4cf9..e35338d 100644
--- a/include/target/target_core_base.h
+++ b/include/target/target_core_base.h
@@ -397,7 +397,7 @@ struct t10_reservation_template {
struct se_queue_req {
int state;
- void *cmd;
+ struct se_cmd *cmd;
struct list_head qr_list;
} ____cacheline_aligned;
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 016/103] target: Remove __transport_get_qr_from_queue()
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (14 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 015/103] target: Make struct se_queue_req.cmd type struct se_cmd * Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 017/103] target: Rename se_dev->g_se_dev_list to se_dev_node Nicholas A. Bellinger
` (13 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover, Nicholas A. Bellinger
From: Andy Grover <agrover@redhat.com>
The only use of it seems to be wrapped by a lock, so it should just use
the locked version.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_transport.c | 27 +--------------------------
1 files changed, 1 insertions(+), 26 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 002ea6f..56249cf 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -841,26 +841,6 @@ static int transport_add_cmd_to_queue(
* Called with struct se_queue_obj->cmd_queue_lock held.
*/
static struct se_queue_req *
-__transport_get_qr_from_queue(struct se_queue_obj *qobj)
-{
- struct se_queue_req *qr = NULL;
-
- if (list_empty(&qobj->qobj_list))
- return NULL;
-
- list_for_each_entry(qr, &qobj->qobj_list, qr_list)
- break;
-
- if (qr->cmd)
- atomic_dec(&T_TASK(qr->cmd)->t_transport_queue_active);
-
- list_del(&qr->qr_list);
- atomic_dec(&qobj->queue_cnt);
-
- return qr;
-}
-
-static struct se_queue_req *
transport_get_qr_from_queue(struct se_queue_obj *qobj)
{
struct se_queue_req *qr;
@@ -6013,10 +5993,7 @@ static void transport_processing_shutdown(struct se_device *dev)
/*
* Empty the struct se_device's struct se_cmd list.
*/
- spin_lock_irqsave(&dev->dev_queue_obj->cmd_queue_lock, flags);
- while ((qr = __transport_get_qr_from_queue(dev->dev_queue_obj))) {
- spin_unlock_irqrestore(
- &dev->dev_queue_obj->cmd_queue_lock, flags);
+ while ((qr = transport_get_qr_from_queue(dev->dev_queue_obj))) {
cmd = qr->cmd;
state = qr->state;
kfree(qr);
@@ -6035,9 +6012,7 @@ static void transport_processing_shutdown(struct se_device *dev)
if (transport_cmd_check_stop(cmd, 1, 0))
transport_generic_remove(cmd, 0, 0);
}
- spin_lock_irqsave(&dev->dev_queue_obj->cmd_queue_lock, flags);
}
- spin_unlock_irqrestore(&dev->dev_queue_obj->cmd_queue_lock, flags);
}
/* transport_processing_thread():
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 017/103] target: Rename se_dev->g_se_dev_list to se_dev_node
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (15 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 016/103] target: Remove __transport_get_qr_from_queue() Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 018/103] target: Remove struct se_global Nicholas A. Bellinger
` (12 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover, Nicholas A. Bellinger
From: Andy Grover <agrover@redhat.com>
The g_ annotation might lead one to believe this is a global variable,
when it is per-subsystem. Calling it se_dev_node also connotes the member
is for being put on a list, and it is not the head of a list itself.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_configfs.c | 6 +++---
drivers/target/target_core_device.c | 2 +-
include/target/target_core_base.h | 2 +-
3 files changed, 5 insertions(+), 5 deletions(-)
diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c
index d140258..03cda0e 100644
--- a/drivers/target/target_core_configfs.c
+++ b/drivers/target/target_core_configfs.c
@@ -2748,7 +2748,7 @@ static struct config_group *target_core_make_subdev(
" struct se_subsystem_dev\n");
goto unlock;
}
- INIT_LIST_HEAD(&se_dev->g_se_dev_list);
+ INIT_LIST_HEAD(&se_dev->se_dev_node);
INIT_LIST_HEAD(&se_dev->t10_wwn.t10_vpd_list);
spin_lock_init(&se_dev->t10_wwn.t10_vpd_lock);
INIT_LIST_HEAD(&se_dev->t10_reservation.registration_list);
@@ -2785,7 +2785,7 @@ static struct config_group *target_core_make_subdev(
goto out;
}
spin_lock(&se_global->g_device_lock);
- list_add_tail(&se_dev->g_se_dev_list, &se_global->g_se_dev_list);
+ list_add_tail(&se_dev->se_dev_node, &se_global->g_se_dev_list);
spin_unlock(&se_global->g_device_lock);
config_group_init_type_name(&se_dev->se_dev_group, name,
@@ -2882,7 +2882,7 @@ static void target_core_drop_subdev(
t = hba->transport;
spin_lock(&se_global->g_device_lock);
- list_del(&se_dev->g_se_dev_list);
+ list_del(&se_dev->se_dev_node);
spin_unlock(&se_global->g_device_lock);
dev_stat_grp = &DEV_STAT_GRP(se_dev)->stat_group;
diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c
index f03e0cc..4d8da32 100644
--- a/drivers/target/target_core_device.c
+++ b/drivers/target/target_core_device.c
@@ -1589,7 +1589,7 @@ int core_dev_setup_virtual_lun0(void)
ret = -ENOMEM;
goto out;
}
- INIT_LIST_HEAD(&se_dev->g_se_dev_list);
+ INIT_LIST_HEAD(&se_dev->se_dev_node);
INIT_LIST_HEAD(&se_dev->t10_wwn.t10_vpd_list);
spin_lock_init(&se_dev->t10_wwn.t10_vpd_lock);
INIT_LIST_HEAD(&se_dev->t10_reservation.registration_list);
diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h
index e35338d..8467a8d 100644
--- a/include/target/target_core_base.h
+++ b/include/target/target_core_base.h
@@ -730,7 +730,7 @@ struct se_subsystem_dev {
struct t10_reservation_template t10_reservation;
spinlock_t se_dev_lock;
void *se_dev_su_ptr;
- struct list_head g_se_dev_list;
+ struct list_head se_dev_node;
struct config_group se_dev_group;
/* For T10 Reservations */
struct config_group se_dev_pr_group;
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 018/103] target: Remove struct se_global
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (16 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 017/103] target: Rename se_dev->g_se_dev_list to se_dev_node Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-23 8:38 ` Fubo Chen
2011-07-23 8:40 ` Fubo Chen
2011-07-21 7:04 ` [PATCH 019/103] target: Simplify scsi mib index table code Nicholas A. Bellinger
` (11 subsequent siblings)
29 siblings, 2 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover, Nicholas A. Bellinger
From: Andy Grover <agrover@redhat.com>
Much of what was in se_global was actually only used by one source file,
so those variables could be taken out of se_global and made static in that
file.
Removed unneeded g_ prefixes from variables affected.
Renamed struct list_heads that were there to be put on another list head
from *_list to *_node, which helps to differentiate things a little better.
from list_heads that are actually the heads of lists.
Rename init/release_se_global() to *_kmem_caches(), to reflect the changes.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_alua.c | 60 +++++++++++++++++-------------
drivers/target/target_core_configfs.c | 63 ++++++++++++++++++--------------
drivers/target/target_core_device.c | 27 ++++++++------
drivers/target/target_core_hba.c | 19 ++++++----
drivers/target/target_core_tpg.c | 22 +++++++----
drivers/target/target_core_transport.c | 40 +++------------------
include/target/target_core_base.h | 30 ++--------------
include/target/target_core_transport.h | 6 +--
8 files changed, 122 insertions(+), 145 deletions(-)
diff --git a/drivers/target/target_core_alua.c b/drivers/target/target_core_alua.c
index a993f3c..62522d6 100644
--- a/drivers/target/target_core_alua.c
+++ b/drivers/target/target_core_alua.c
@@ -46,6 +46,14 @@ static int core_alua_set_tg_pt_secondary_state(
struct t10_alua_tg_pt_gp_member *tg_pt_gp_mem,
struct se_port *port, int explict, int offline);
+static u16 alua_lu_gps_counter;
+static u32 alua_lu_gps_count;
+
+static DEFINE_SPINLOCK(lu_gps_lock);
+static LIST_HEAD(lu_gps_list);
+
+struct t10_alua_lu_gp *default_lu_gp;
+
/*
* REPORT_TARGET_PORT_GROUPS
*
@@ -1038,15 +1046,15 @@ core_alua_allocate_lu_gp(const char *name, int def_group)
printk(KERN_ERR "Unable to allocate struct t10_alua_lu_gp\n");
return ERR_PTR(-ENOMEM);
}
- INIT_LIST_HEAD(&lu_gp->lu_gp_list);
+ INIT_LIST_HEAD(&lu_gp->lu_gp_node);
INIT_LIST_HEAD(&lu_gp->lu_gp_mem_list);
spin_lock_init(&lu_gp->lu_gp_lock);
atomic_set(&lu_gp->lu_gp_ref_cnt, 0);
if (def_group) {
- lu_gp->lu_gp_id = se_global->alua_lu_gps_counter++;
+ lu_gp->lu_gp_id = alua_lu_gps_counter++;
lu_gp->lu_gp_valid_id = 1;
- se_global->alua_lu_gps_count++;
+ alua_lu_gps_count++;
}
return lu_gp;
@@ -1065,19 +1073,19 @@ int core_alua_set_lu_gp_id(struct t10_alua_lu_gp *lu_gp, u16 lu_gp_id)
return -EINVAL;
}
- spin_lock(&se_global->lu_gps_lock);
- if (se_global->alua_lu_gps_count == 0x0000ffff) {
- printk(KERN_ERR "Maximum ALUA se_global->alua_lu_gps_count:"
+ spin_lock(&lu_gps_lock);
+ if (alua_lu_gps_count == 0x0000ffff) {
+ printk(KERN_ERR "Maximum ALUA alua_lu_gps_count:"
" 0x0000ffff reached\n");
- spin_unlock(&se_global->lu_gps_lock);
+ spin_unlock(&lu_gps_lock);
kmem_cache_free(t10_alua_lu_gp_cache, lu_gp);
return -ENOSPC;
}
again:
lu_gp_id_tmp = (lu_gp_id != 0) ? lu_gp_id :
- se_global->alua_lu_gps_counter++;
+ alua_lu_gps_counter++;
- list_for_each_entry(lu_gp_tmp, &se_global->g_lu_gps_list, lu_gp_list) {
+ list_for_each_entry(lu_gp_tmp, &lu_gps_list, lu_gp_node) {
if (lu_gp_tmp->lu_gp_id == lu_gp_id_tmp) {
if (!(lu_gp_id))
goto again;
@@ -1085,16 +1093,16 @@ again:
printk(KERN_WARNING "ALUA Logical Unit Group ID: %hu"
" already exists, ignoring request\n",
lu_gp_id);
- spin_unlock(&se_global->lu_gps_lock);
+ spin_unlock(&lu_gps_lock);
return -EINVAL;
}
}
lu_gp->lu_gp_id = lu_gp_id_tmp;
lu_gp->lu_gp_valid_id = 1;
- list_add_tail(&lu_gp->lu_gp_list, &se_global->g_lu_gps_list);
- se_global->alua_lu_gps_count++;
- spin_unlock(&se_global->lu_gps_lock);
+ list_add_tail(&lu_gp->lu_gp_node, &lu_gps_list);
+ alua_lu_gps_count++;
+ spin_unlock(&lu_gps_lock);
return 0;
}
@@ -1130,11 +1138,11 @@ void core_alua_free_lu_gp(struct t10_alua_lu_gp *lu_gp)
* no associations can be made while we are releasing
* struct t10_alua_lu_gp.
*/
- spin_lock(&se_global->lu_gps_lock);
+ spin_lock(&lu_gps_lock);
atomic_set(&lu_gp->lu_gp_shutdown, 1);
- list_del(&lu_gp->lu_gp_list);
- se_global->alua_lu_gps_count--;
- spin_unlock(&se_global->lu_gps_lock);
+ list_del(&lu_gp->lu_gp_node);
+ alua_lu_gps_count--;
+ spin_unlock(&lu_gps_lock);
/*
* Allow struct t10_alua_lu_gp * referenced by core_alua_get_lu_gp_by_name()
* in target_core_configfs.c:target_core_store_alua_lu_gp() to be
@@ -1165,9 +1173,9 @@ void core_alua_free_lu_gp(struct t10_alua_lu_gp *lu_gp)
* we want to re-assocate a given lu_gp_mem with default_lu_gp.
*/
spin_lock(&lu_gp_mem->lu_gp_mem_lock);
- if (lu_gp != se_global->default_lu_gp)
+ if (lu_gp != default_lu_gp)
__core_alua_attach_lu_gp_mem(lu_gp_mem,
- se_global->default_lu_gp);
+ default_lu_gp);
else
lu_gp_mem->lu_gp = NULL;
spin_unlock(&lu_gp_mem->lu_gp_mem_lock);
@@ -1218,27 +1226,27 @@ struct t10_alua_lu_gp *core_alua_get_lu_gp_by_name(const char *name)
struct t10_alua_lu_gp *lu_gp;
struct config_item *ci;
- spin_lock(&se_global->lu_gps_lock);
- list_for_each_entry(lu_gp, &se_global->g_lu_gps_list, lu_gp_list) {
+ spin_lock(&lu_gps_lock);
+ list_for_each_entry(lu_gp, &lu_gps_list, lu_gp_node) {
if (!(lu_gp->lu_gp_valid_id))
continue;
ci = &lu_gp->lu_gp_group.cg_item;
if (!(strcmp(config_item_name(ci), name))) {
atomic_inc(&lu_gp->lu_gp_ref_cnt);
- spin_unlock(&se_global->lu_gps_lock);
+ spin_unlock(&lu_gps_lock);
return lu_gp;
}
}
- spin_unlock(&se_global->lu_gps_lock);
+ spin_unlock(&lu_gps_lock);
return NULL;
}
void core_alua_put_lu_gp_from_name(struct t10_alua_lu_gp *lu_gp)
{
- spin_lock(&se_global->lu_gps_lock);
+ spin_lock(&lu_gps_lock);
atomic_dec(&lu_gp->lu_gp_ref_cnt);
- spin_unlock(&se_global->lu_gps_lock);
+ spin_unlock(&lu_gps_lock);
}
/*
@@ -1974,7 +1982,7 @@ int core_setup_alua(struct se_device *dev, int force_pt)
alua->alua_state_check = &core_alua_state_check;
spin_lock(&lu_gp_mem->lu_gp_mem_lock);
__core_alua_attach_lu_gp_mem(lu_gp_mem,
- se_global->default_lu_gp);
+ default_lu_gp);
spin_unlock(&lu_gp_mem->lu_gp_mem_lock);
printk(KERN_INFO "%s: Adding to default ALUA LU Group:"
diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c
index 03cda0e..82c8e9a 100644
--- a/drivers/target/target_core_configfs.c
+++ b/drivers/target/target_core_configfs.c
@@ -37,6 +37,7 @@
#include <linux/parser.h>
#include <linux/syscalls.h>
#include <linux/configfs.h>
+#include <linux/spinlock.h>
#include <target/target_core_base.h>
#include <target/target_core_device.h>
@@ -52,6 +53,8 @@
#include "target_core_rd.h"
#include "target_core_stat.h"
+extern struct t10_alua_lu_gp *default_lu_gp;
+
static struct list_head g_tf_list;
static struct mutex g_tf_lock;
@@ -61,6 +64,13 @@ struct target_core_configfs_attribute {
ssize_t (*store)(void *, const char *, size_t);
};
+static struct config_group target_core_hbagroup;
+static struct config_group alua_group;
+static struct config_group alua_lu_gps_group;
+
+static DEFINE_SPINLOCK(se_device_lock);
+static LIST_HEAD(se_dev_list);
+
static inline struct se_hba *
item_to_hba(struct config_item *item)
{
@@ -2784,9 +2794,9 @@ static struct config_group *target_core_make_subdev(
" from allocate_virtdevice()\n");
goto out;
}
- spin_lock(&se_global->g_device_lock);
- list_add_tail(&se_dev->se_dev_node, &se_global->g_se_dev_list);
- spin_unlock(&se_global->g_device_lock);
+ spin_lock(&se_device_lock);
+ list_add_tail(&se_dev->se_dev_node, &se_dev_list);
+ spin_unlock(&se_device_lock);
config_group_init_type_name(&se_dev->se_dev_group, name,
&target_core_dev_cit);
@@ -2881,9 +2891,9 @@ static void target_core_drop_subdev(
mutex_lock(&hba->hba_access_mutex);
t = hba->transport;
- spin_lock(&se_global->g_device_lock);
+ spin_lock(&se_device_lock);
list_del(&se_dev->se_dev_node);
- spin_unlock(&se_global->g_device_lock);
+ spin_unlock(&se_device_lock);
dev_stat_grp = &DEV_STAT_GRP(se_dev)->stat_group;
for (i = 0; dev_stat_grp->default_groups[i]; i++) {
@@ -3131,7 +3141,7 @@ static int __init target_core_init_configfs(void)
INIT_LIST_HEAD(&g_tf_list);
mutex_init(&g_tf_lock);
init_scsi_index_table();
- ret = init_se_global();
+ ret = init_se_kmem_caches();
if (ret < 0)
return ret;
/*
@@ -3146,29 +3156,29 @@ static int __init target_core_init_configfs(void)
goto out_global;
}
- config_group_init_type_name(&se_global->target_core_hbagroup,
+ config_group_init_type_name(&target_core_hbagroup,
"core", &target_core_cit);
- target_cg->default_groups[0] = &se_global->target_core_hbagroup;
+ target_cg->default_groups[0] = &target_core_hbagroup;
target_cg->default_groups[1] = NULL;
/*
* Create ALUA infrastructure under /sys/kernel/config/target/core/alua/
*/
- hba_cg = &se_global->target_core_hbagroup;
+ hba_cg = &target_core_hbagroup;
hba_cg->default_groups = kzalloc(sizeof(struct config_group) * 2,
GFP_KERNEL);
if (!(hba_cg->default_groups)) {
printk(KERN_ERR "Unable to allocate hba_cg->default_groups\n");
goto out_global;
}
- config_group_init_type_name(&se_global->alua_group,
+ config_group_init_type_name(&alua_group,
"alua", &target_core_alua_cit);
- hba_cg->default_groups[0] = &se_global->alua_group;
+ hba_cg->default_groups[0] = &alua_group;
hba_cg->default_groups[1] = NULL;
/*
* Add ALUA Logical Unit Group and Target Port Group ConfigFS
* groups under /sys/kernel/config/target/core/alua/
*/
- alua_cg = &se_global->alua_group;
+ alua_cg = &alua_group;
alua_cg->default_groups = kzalloc(sizeof(struct config_group) * 2,
GFP_KERNEL);
if (!(alua_cg->default_groups)) {
@@ -3176,9 +3186,9 @@ static int __init target_core_init_configfs(void)
goto out_global;
}
- config_group_init_type_name(&se_global->alua_lu_gps_group,
+ config_group_init_type_name(&alua_lu_gps_group,
"lu_gps", &target_core_alua_lu_gps_cit);
- alua_cg->default_groups[0] = &se_global->alua_lu_gps_group;
+ alua_cg->default_groups[0] = &alua_lu_gps_group;
alua_cg->default_groups[1] = NULL;
/*
* Add core/alua/lu_gps/default_lu_gp
@@ -3187,7 +3197,7 @@ static int __init target_core_init_configfs(void)
if (IS_ERR(lu_gp))
goto out_global;
- lu_gp_cg = &se_global->alua_lu_gps_group;
+ lu_gp_cg = &alua_lu_gps_group;
lu_gp_cg->default_groups = kzalloc(sizeof(struct config_group) * 2,
GFP_KERNEL);
if (!(lu_gp_cg->default_groups)) {
@@ -3199,7 +3209,7 @@ static int __init target_core_init_configfs(void)
&target_core_alua_lu_gp_cit);
lu_gp_cg->default_groups[0] = &lu_gp->lu_gp_group;
lu_gp_cg->default_groups[1] = NULL;
- se_global->default_lu_gp = lu_gp;
+ default_lu_gp = lu_gp;
/*
* Register the target_core_mod subsystem with configfs.
*/
@@ -3229,9 +3239,9 @@ out:
core_dev_release_virtual_lun0();
rd_module_exit();
out_global:
- if (se_global->default_lu_gp) {
- core_alua_free_lu_gp(se_global->default_lu_gp);
- se_global->default_lu_gp = NULL;
+ if (default_lu_gp) {
+ core_alua_free_lu_gp(default_lu_gp);
+ default_lu_gp = NULL;
}
if (lu_gp_cg)
kfree(lu_gp_cg->default_groups);
@@ -3240,7 +3250,7 @@ out_global:
if (hba_cg)
kfree(hba_cg->default_groups);
kfree(target_cg->default_groups);
- release_se_global();
+ release_se_kmem_caches();
return ret;
}
@@ -3251,10 +3261,9 @@ static void __exit target_core_exit_configfs(void)
struct config_item *item;
int i;
- se_global->in_shutdown = 1;
subsys = target_core_subsystem[0];
- lu_gp_cg = &se_global->alua_lu_gps_group;
+ lu_gp_cg = &alua_lu_gps_group;
for (i = 0; lu_gp_cg->default_groups[i]; i++) {
item = &lu_gp_cg->default_groups[i]->cg_item;
lu_gp_cg->default_groups[i] = NULL;
@@ -3263,7 +3272,7 @@ static void __exit target_core_exit_configfs(void)
kfree(lu_gp_cg->default_groups);
lu_gp_cg->default_groups = NULL;
- alua_cg = &se_global->alua_group;
+ alua_cg = &alua_group;
for (i = 0; alua_cg->default_groups[i]; i++) {
item = &alua_cg->default_groups[i]->cg_item;
alua_cg->default_groups[i] = NULL;
@@ -3272,7 +3281,7 @@ static void __exit target_core_exit_configfs(void)
kfree(alua_cg->default_groups);
alua_cg->default_groups = NULL;
- hba_cg = &se_global->target_core_hbagroup;
+ hba_cg = &target_core_hbagroup;
for (i = 0; hba_cg->default_groups[i]; i++) {
item = &hba_cg->default_groups[i]->cg_item;
hba_cg->default_groups[i] = NULL;
@@ -3287,15 +3296,15 @@ static void __exit target_core_exit_configfs(void)
configfs_unregister_subsystem(subsys);
kfree(subsys->su_group.default_groups);
- core_alua_free_lu_gp(se_global->default_lu_gp);
- se_global->default_lu_gp = NULL;
+ core_alua_free_lu_gp(default_lu_gp);
+ default_lu_gp = NULL;
printk(KERN_INFO "TARGET_CORE[0]: Released ConfigFS Fabric"
" Infrastructure\n");
core_dev_release_virtual_lun0();
rd_module_exit();
- release_se_global();
+ release_se_kmem_caches();
return;
}
diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c
index 4d8da32..c9a3f73 100644
--- a/drivers/target/target_core_device.c
+++ b/drivers/target/target_core_device.c
@@ -54,6 +54,11 @@
static void se_dev_start(struct se_device *dev);
static void se_dev_stop(struct se_device *dev);
+static struct se_hba *lun0_hba;
+static struct se_subsystem_dev *lun0_su_dev;
+/* not static, needed by tpg.c */
+struct se_device *g_lun0_dev;
+
int transport_get_lun_for_cmd(
struct se_cmd *se_cmd,
u32 unpacked_lun)
@@ -1579,7 +1584,7 @@ int core_dev_setup_virtual_lun0(void)
if (IS_ERR(hba))
return PTR_ERR(hba);
- se_global->g_lun0_hba = hba;
+ lun0_hba = hba;
t = hba->transport;
se_dev = kzalloc(sizeof(struct se_subsystem_dev), GFP_KERNEL);
@@ -1612,7 +1617,7 @@ int core_dev_setup_virtual_lun0(void)
ret = -ENOMEM;
goto out;
}
- se_global->g_lun0_su_dev = se_dev;
+ lun0_su_dev = se_dev;
memset(buf, 0, 16);
sprintf(buf, "rd_pages=8");
@@ -1624,15 +1629,15 @@ int core_dev_setup_virtual_lun0(void)
goto out;
}
se_dev->se_dev_ptr = dev;
- se_global->g_lun0_dev = dev;
+ g_lun0_dev = dev;
return 0;
out:
- se_global->g_lun0_su_dev = NULL;
+ lun0_su_dev = NULL;
kfree(se_dev);
- if (se_global->g_lun0_hba) {
- core_delete_hba(se_global->g_lun0_hba);
- se_global->g_lun0_hba = NULL;
+ if (lun0_hba) {
+ core_delete_hba(lun0_hba);
+ lun0_hba = NULL;
}
return ret;
}
@@ -1640,14 +1645,14 @@ out:
void core_dev_release_virtual_lun0(void)
{
- struct se_hba *hba = se_global->g_lun0_hba;
- struct se_subsystem_dev *su_dev = se_global->g_lun0_su_dev;
+ struct se_hba *hba = lun0_hba;
+ struct se_subsystem_dev *su_dev = lun0_su_dev;
if (!(hba))
return;
- if (se_global->g_lun0_dev)
- se_free_virtual_device(se_global->g_lun0_dev, hba);
+ if (g_lun0_dev)
+ se_free_virtual_device(g_lun0_dev, hba);
kfree(su_dev);
core_delete_hba(hba);
diff --git a/drivers/target/target_core_hba.c b/drivers/target/target_core_hba.c
index 8f3a0c4..bd9da25 100644
--- a/drivers/target/target_core_hba.c
+++ b/drivers/target/target_core_hba.c
@@ -45,6 +45,11 @@
static LIST_HEAD(subsystem_list);
static DEFINE_MUTEX(subsystem_mutex);
+static u32 hba_id_counter;
+
+static DEFINE_SPINLOCK(hba_lock);
+static LIST_HEAD(hba_list);
+
int transport_subsystem_register(struct se_subsystem_api *sub_api)
{
struct se_subsystem_api *s;
@@ -125,10 +130,10 @@ core_alloc_hba(const char *plugin_name, u32 plugin_dep_id, u32 hba_flags)
if (ret < 0)
goto out_module_put;
- spin_lock(&se_global->hba_lock);
- hba->hba_id = se_global->g_hba_id_counter++;
- list_add_tail(&hba->hba_list, &se_global->g_hba_list);
- spin_unlock(&se_global->hba_lock);
+ spin_lock(&hba_lock);
+ hba->hba_id = hba_id_counter++;
+ list_add_tail(&hba->hba_node, &hba_list);
+ spin_unlock(&hba_lock);
printk(KERN_INFO "CORE_HBA[%d] - Attached HBA to Generic Target"
" Core\n", hba->hba_id);
@@ -152,9 +157,9 @@ core_delete_hba(struct se_hba *hba)
hba->transport->detach_hba(hba);
- spin_lock(&se_global->hba_lock);
- list_del(&hba->hba_list);
- spin_unlock(&se_global->hba_lock);
+ spin_lock(&hba_lock);
+ list_del(&hba->hba_node);
+ spin_unlock(&hba_lock);
printk(KERN_INFO "CORE_HBA[%d] - Detached HBA from Generic Target"
" Core\n", hba->hba_id);
diff --git a/drivers/target/target_core_tpg.c b/drivers/target/target_core_tpg.c
index dc024c1..4e2bee9 100644
--- a/drivers/target/target_core_tpg.c
+++ b/drivers/target/target_core_tpg.c
@@ -44,6 +44,12 @@
#include <target/target_core_fabric_ops.h>
#include "target_core_hba.h"
+#include "target_core_stat.h"
+
+extern struct se_device *g_lun0_dev;
+
+static DEFINE_SPINLOCK(tpg_lock);
+static LIST_HEAD(tpg_list);
/* core_clear_initiator_node_from_tpg():
*
@@ -597,7 +603,7 @@ EXPORT_SYMBOL(core_tpg_set_initiator_node_queue_depth);
static int core_tpg_setup_virtual_lun0(struct se_portal_group *se_tpg)
{
/* Set in core_dev_setup_virtual_lun0() */
- struct se_device *dev = se_global->g_lun0_dev;
+ struct se_device *dev = g_lun0_dev;
struct se_lun *lun = &se_tpg->tpg_virt_lun0;
u32 lun_access = TRANSPORT_LUNFLAGS_READ_ONLY;
int ret;
@@ -663,7 +669,7 @@ int core_tpg_register(
se_tpg->se_tpg_wwn = se_wwn;
atomic_set(&se_tpg->tpg_pr_ref_count, 0);
INIT_LIST_HEAD(&se_tpg->acl_node_list);
- INIT_LIST_HEAD(&se_tpg->se_tpg_list);
+ INIT_LIST_HEAD(&se_tpg->se_tpg_node);
INIT_LIST_HEAD(&se_tpg->tpg_sess_list);
spin_lock_init(&se_tpg->acl_node_lock);
spin_lock_init(&se_tpg->session_lock);
@@ -676,9 +682,9 @@ int core_tpg_register(
}
}
- spin_lock_bh(&se_global->se_tpg_lock);
- list_add_tail(&se_tpg->se_tpg_list, &se_global->g_se_tpg_list);
- spin_unlock_bh(&se_global->se_tpg_lock);
+ spin_lock_bh(&tpg_lock);
+ list_add_tail(&se_tpg->se_tpg_node, &tpg_list);
+ spin_unlock_bh(&tpg_lock);
printk(KERN_INFO "TARGET_CORE[%s]: Allocated %s struct se_portal_group for"
" endpoint: %s, Portal Tag: %u\n", tfo->get_fabric_name(),
@@ -701,9 +707,9 @@ int core_tpg_deregister(struct se_portal_group *se_tpg)
TPG_TFO(se_tpg)->tpg_get_wwn(se_tpg),
TPG_TFO(se_tpg)->tpg_get_tag(se_tpg));
- spin_lock_bh(&se_global->se_tpg_lock);
- list_del(&se_tpg->se_tpg_list);
- spin_unlock_bh(&se_global->se_tpg_lock);
+ spin_lock_bh(&tpg_lock);
+ list_del(&se_tpg->se_tpg_node);
+ spin_unlock_bh(&tpg_lock);
while (atomic_read(&se_tpg->tpg_pr_ref_count) != 0)
cpu_relax();
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 56249cf..0f084a1 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -184,7 +184,7 @@
#define DEBUG_STA(x...)
#endif
-struct se_global *se_global;
+static int sub_api_initialized;
static struct kmem_cache *se_cmd_cache;
static struct kmem_cache *se_sess_cache;
@@ -227,26 +227,8 @@ static void transport_remove_cmd_from_queue(struct se_cmd *cmd,
static int transport_set_sense_codes(struct se_cmd *cmd, u8 asc, u8 ascq);
static void transport_stop_all_task_timers(struct se_cmd *cmd);
-int init_se_global(void)
+int init_se_kmem_caches(void)
{
- struct se_global *global;
-
- global = kzalloc(sizeof(struct se_global), GFP_KERNEL);
- if (!(global)) {
- printk(KERN_ERR "Unable to allocate memory for struct se_global\n");
- return -ENOMEM;
- }
-
- INIT_LIST_HEAD(&global->g_lu_gps_list);
- INIT_LIST_HEAD(&global->g_se_tpg_list);
- INIT_LIST_HEAD(&global->g_hba_list);
- INIT_LIST_HEAD(&global->g_se_dev_list);
- spin_lock_init(&global->g_device_lock);
- spin_lock_init(&global->hba_lock);
- spin_lock_init(&global->se_tpg_lock);
- spin_lock_init(&global->lu_gps_lock);
- spin_lock_init(&global->plugin_class_lock);
-
se_cmd_cache = kmem_cache_create("se_cmd_cache",
sizeof(struct se_cmd), __alignof__(struct se_cmd), 0, NULL);
if (!(se_cmd_cache)) {
@@ -325,8 +307,6 @@ int init_se_global(void)
goto out;
}
- se_global = global;
-
return 0;
out:
if (se_cmd_cache)
@@ -349,18 +329,11 @@ out:
kmem_cache_destroy(t10_alua_tg_pt_gp_cache);
if (t10_alua_tg_pt_gp_mem_cache)
kmem_cache_destroy(t10_alua_tg_pt_gp_mem_cache);
- kfree(global);
return -ENOMEM;
}
-void release_se_global(void)
+void release_se_kmem_caches(void)
{
- struct se_global *global;
-
- global = se_global;
- if (!(global))
- return;
-
kmem_cache_destroy(se_cmd_cache);
kmem_cache_destroy(se_tmr_req_cache);
kmem_cache_destroy(se_sess_cache);
@@ -371,9 +344,6 @@ void release_se_global(void)
kmem_cache_destroy(t10_alua_lu_gp_mem_cache);
kmem_cache_destroy(t10_alua_tg_pt_gp_cache);
kmem_cache_destroy(t10_alua_tg_pt_gp_mem_cache);
- kfree(global);
-
- se_global = NULL;
}
/* SCSI statistics table index */
@@ -446,7 +416,7 @@ int transport_subsystem_check_init(void)
{
int ret;
- if (se_global->g_sub_api_initialized)
+ if (sub_api_initialized)
return 0;
/*
* Request the loading of known TCM subsystem plugins..
@@ -455,7 +425,7 @@ int transport_subsystem_check_init(void)
if (ret < 0)
return ret;
- se_global->g_sub_api_initialized = 1;
+ sub_api_initialized = 1;
return 0;
}
diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h
index 8467a8d..de7bec6 100644
--- a/include/target/target_core_base.h
+++ b/include/target/target_core_base.h
@@ -235,7 +235,7 @@ struct t10_alua_lu_gp {
atomic_t lu_gp_ref_cnt;
spinlock_t lu_gp_lock;
struct config_group lu_gp_group;
- struct list_head lu_gp_list;
+ struct list_head lu_gp_node;
struct list_head lu_gp_mem_list;
} ____cacheline_aligned;
@@ -840,7 +840,7 @@ struct se_hba {
void *hba_ptr;
/* Linked list for struct se_device */
struct list_head hba_dev_list;
- struct list_head hba_list;
+ struct list_head hba_node;
spinlock_t device_lock;
struct config_group hba_group;
struct mutex hba_access_mutex;
@@ -924,7 +924,7 @@ struct se_portal_group {
spinlock_t tpg_lun_lock;
/* Pointer to $FABRIC_MOD portal group */
void *se_tpg_fabric_ptr;
- struct list_head se_tpg_list;
+ struct list_head se_tpg_node;
/* linked list for initiator ACL list */
struct list_head acl_node_list;
struct se_lun *tpg_lun_list;
@@ -952,28 +952,4 @@ struct se_wwn {
struct config_group fabric_stat_group;
} ____cacheline_aligned;
-struct se_global {
- u16 alua_lu_gps_counter;
- int g_sub_api_initialized;
- u32 in_shutdown;
- u32 alua_lu_gps_count;
- u32 g_hba_id_counter;
- struct config_group target_core_hbagroup;
- struct config_group alua_group;
- struct config_group alua_lu_gps_group;
- struct list_head g_lu_gps_list;
- struct list_head g_se_tpg_list;
- struct list_head g_hba_list;
- struct list_head g_se_dev_list;
- struct se_hba *g_lun0_hba;
- struct se_subsystem_dev *g_lun0_su_dev;
- struct se_device *g_lun0_dev;
- struct t10_alua_lu_gp *default_lu_gp;
- spinlock_t g_device_lock;
- spinlock_t hba_lock;
- spinlock_t se_tpg_lock;
- spinlock_t lu_gps_lock;
- spinlock_t plugin_class_lock;
-} ____cacheline_aligned;
-
#endif /* TARGET_CORE_BASE_H */
diff --git a/include/target/target_core_transport.h b/include/target/target_core_transport.h
index 199a404..06b8900 100644
--- a/include/target/target_core_transport.h
+++ b/include/target/target_core_transport.h
@@ -111,8 +111,8 @@ struct se_subsystem_api;
extern struct kmem_cache *se_mem_cache;
-extern int init_se_global(void);
-extern void release_se_global(void);
+extern int init_se_kmem_caches(void);
+extern void release_se_kmem_caches(void);
extern void init_scsi_index_table(void);
extern u32 scsi_get_new_index(scsi_index_t);
extern void transport_init_queue_obj(struct se_queue_obj *);
@@ -355,6 +355,4 @@ struct se_subsystem_api {
#define TRANSPORT(dev) ((dev)->transport)
#define HBA_TRANSPORT(hba) ((hba)->transport)
-extern struct se_global *se_global;
-
#endif /* TARGET_CORE_TRANSPORT_H */
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 019/103] target: Simplify scsi mib index table code
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (17 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 018/103] target: Remove struct se_global Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 020/103] target: Make dev_queue_obj a member of se_device instead of a pointer Nicholas A. Bellinger
` (10 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover, Nicholas A. Bellinger
From: Andy Grover <agrover@redhat.com>
Taking scsi_mib_index and its lock out of the struct makes it possible to
initialize both without code.
Replace type check with BUG_ON, since only bugs will cause type to be invalid.
Nobody checks retval, in any case.
Remove wraparound protection. Will anyone really ever hit 4 billion mib
index values?? If this is a concern, then the solution should probably be
to make the index a u64 -- this interface wants to hand out unique values,
and on wrap they would no longer meet that criteria, even if we skip 0.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_configfs.c | 1 -
drivers/target/target_core_transport.c | 28 +++++++---------------------
include/target/target_core_base.h | 5 -----
include/target/target_core_transport.h | 1 -
4 files changed, 7 insertions(+), 28 deletions(-)
diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c
index 82c8e9a..27effbc 100644
--- a/drivers/target/target_core_configfs.c
+++ b/drivers/target/target_core_configfs.c
@@ -3140,7 +3140,6 @@ static int __init target_core_init_configfs(void)
INIT_LIST_HEAD(&g_tf_list);
mutex_init(&g_tf_lock);
- init_scsi_index_table();
ret = init_se_kmem_caches();
if (ret < 0)
return ret;
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 0f084a1..4bfa61f 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -346,18 +346,9 @@ void release_se_kmem_caches(void)
kmem_cache_destroy(t10_alua_tg_pt_gp_mem_cache);
}
-/* SCSI statistics table index */
-static struct scsi_index_table scsi_index_table;
-
-/*
- * Initialize the index table for allocating unique row indexes to various mib
- * tables.
- */
-void init_scsi_index_table(void)
-{
- memset(&scsi_index_table, 0, sizeof(struct scsi_index_table));
- spin_lock_init(&scsi_index_table.lock);
-}
+/* This code ensures unique mib indexes are handed out. */
+static DEFINE_SPINLOCK(scsi_mib_index_lock);
+static u32 scsi_mib_index[SCSI_INDEX_TYPE_MAX];
/*
* Allocate a new row index for the entry type specified
@@ -366,16 +357,11 @@ u32 scsi_get_new_index(scsi_index_t type)
{
u32 new_index;
- if ((type < 0) || (type >= SCSI_INDEX_TYPE_MAX)) {
- printk(KERN_ERR "Invalid index type %d\n", type);
- return -EINVAL;
- }
+ BUG_ON((type < 0) || (type >= SCSI_INDEX_TYPE_MAX));
- spin_lock(&scsi_index_table.lock);
- new_index = ++scsi_index_table.scsi_mib_index[type];
- if (new_index == 0)
- new_index = ++scsi_index_table.scsi_mib_index[type];
- spin_unlock(&scsi_index_table.lock);
+ spin_lock(&scsi_mib_index_lock);
+ new_index = ++scsi_mib_index[type];
+ spin_unlock(&scsi_mib_index_lock);
return new_index;
}
diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h
index de7bec6..23dd6b7 100644
--- a/include/target/target_core_base.h
+++ b/include/target/target_core_base.h
@@ -205,11 +205,6 @@ typedef enum {
SCSI_INDEX_TYPE_MAX
} scsi_index_t;
-struct scsi_index_table {
- spinlock_t lock;
- u32 scsi_mib_index[SCSI_INDEX_TYPE_MAX];
-} ____cacheline_aligned;
-
struct se_cmd;
struct t10_alua {
diff --git a/include/target/target_core_transport.h b/include/target/target_core_transport.h
index 06b8900..da982e9 100644
--- a/include/target/target_core_transport.h
+++ b/include/target/target_core_transport.h
@@ -113,7 +113,6 @@ extern struct kmem_cache *se_mem_cache;
extern int init_se_kmem_caches(void);
extern void release_se_kmem_caches(void);
-extern void init_scsi_index_table(void);
extern u32 scsi_get_new_index(scsi_index_t);
extern void transport_init_queue_obj(struct se_queue_obj *);
extern int transport_subsystem_check_init(void);
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 020/103] target: Make dev_queue_obj a member of se_device instead of a pointer
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (18 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 019/103] target: Simplify scsi mib index table code Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 021/103] target: remove extraneous returns at end of void functions Nicholas A. Bellinger
` (9 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover
From: Andy Grover <agrover@redhat.com>
This saves us having to alloc/free memory for it separately from se_device.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_device.c | 1 -
drivers/target/target_core_tmr.c | 2 +-
drivers/target/target_core_transport.c | 52 +++++++++++++------------------
include/target/target_core_base.h | 2 +-
4 files changed, 24 insertions(+), 33 deletions(-)
diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c
index c9a3f73..fe5474c 100644
--- a/drivers/target/target_core_device.c
+++ b/drivers/target/target_core_device.c
@@ -760,7 +760,6 @@ void se_release_device_for_hba(struct se_device *dev)
core_scsi3_free_all_registrations(dev);
se_release_vpd_for_dev(dev);
- kfree(dev->dev_queue_obj);
kfree(dev);
return;
diff --git a/drivers/target/target_core_tmr.c b/drivers/target/target_core_tmr.c
index 179063d..7cd8e29 100644
--- a/drivers/target/target_core_tmr.c
+++ b/drivers/target/target_core_tmr.c
@@ -117,7 +117,7 @@ int core_tmr_lun_reset(
struct se_queue_req *qr, *qr_tmp;
struct se_node_acl *tmr_nacl = NULL;
struct se_portal_group *tmr_tpg = NULL;
- struct se_queue_obj *qobj = dev->dev_queue_obj;
+ struct se_queue_obj *qobj = &dev->dev_queue_obj;
struct se_tmr_req *tmr_p, *tmr_pp;
struct se_task *task, *task_tmp;
unsigned long flags;
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 4bfa61f..7e0d69a 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -737,7 +737,7 @@ check_lun:
void transport_cmd_finish_abort(struct se_cmd *cmd, int remove)
{
- transport_remove_cmd_from_queue(cmd, SE_DEV(cmd)->dev_queue_obj);
+ transport_remove_cmd_from_queue(cmd, &SE_DEV(cmd)->dev_queue_obj);
transport_lun_remove_cmd(cmd);
if (transport_cmd_check_stop_to_fabric(cmd))
@@ -748,7 +748,7 @@ void transport_cmd_finish_abort(struct se_cmd *cmd, int remove)
void transport_cmd_finish_abort_tmr(struct se_cmd *cmd)
{
- transport_remove_cmd_from_queue(cmd, SE_DEV(cmd)->dev_queue_obj);
+ transport_remove_cmd_from_queue(cmd, &SE_DEV(cmd)->dev_queue_obj);
if (transport_cmd_check_stop_to_fabric(cmd))
return;
@@ -761,7 +761,7 @@ static int transport_add_cmd_to_queue(
int t_state)
{
struct se_device *dev = cmd->se_dev;
- struct se_queue_obj *qobj = dev->dev_queue_obj;
+ struct se_queue_obj *qobj = &dev->dev_queue_obj;
struct se_queue_req *qr;
unsigned long flags;
@@ -1213,15 +1213,15 @@ static void transport_release_all_cmds(struct se_device *dev)
int bug_out = 0, t_state;
unsigned long flags;
- spin_lock_irqsave(&dev->dev_queue_obj->cmd_queue_lock, flags);
- list_for_each_entry_safe(qr, qr_p, &dev->dev_queue_obj->qobj_list,
+ spin_lock_irqsave(&dev->dev_queue_obj.cmd_queue_lock, flags);
+ list_for_each_entry_safe(qr, qr_p, &dev->dev_queue_obj.qobj_list,
qr_list) {
cmd = qr->cmd;
t_state = qr->state;
list_del(&qr->qr_list);
kfree(qr);
- spin_unlock_irqrestore(&dev->dev_queue_obj->cmd_queue_lock,
+ spin_unlock_irqrestore(&dev->dev_queue_obj.cmd_queue_lock,
flags);
printk(KERN_ERR "Releasing ITT: 0x%08x, i_state: %u,"
@@ -1232,9 +1232,9 @@ static void transport_release_all_cmds(struct se_device *dev)
transport_release_fe_cmd(cmd);
bug_out = 1;
- spin_lock_irqsave(&dev->dev_queue_obj->cmd_queue_lock, flags);
+ spin_lock_irqsave(&dev->dev_queue_obj.cmd_queue_lock, flags);
}
- spin_unlock_irqrestore(&dev->dev_queue_obj->cmd_queue_lock, flags);
+ spin_unlock_irqrestore(&dev->dev_queue_obj.cmd_queue_lock, flags);
#if 0
if (bug_out)
BUG();
@@ -1563,15 +1563,8 @@ struct se_device *transport_add_device_to_core_hba(
printk(KERN_ERR "Unable to allocate memory for se_dev_t\n");
return NULL;
}
- dev->dev_queue_obj = kzalloc(sizeof(struct se_queue_obj), GFP_KERNEL);
- if (!(dev->dev_queue_obj)) {
- printk(KERN_ERR "Unable to allocate memory for"
- " dev->dev_queue_obj\n");
- kfree(dev);
- return NULL;
- }
- transport_init_queue_obj(dev->dev_queue_obj);
+ transport_init_queue_obj(&dev->dev_queue_obj);
dev->dev_flags = device_flags;
dev->dev_status |= TRANSPORT_DEVICE_DEACTIVATED;
dev->dev_ptr = (void *) transport_dev;
@@ -1672,7 +1665,6 @@ out:
se_release_vpd_for_dev(dev);
- kfree(dev->dev_queue_obj);
kfree(dev);
return NULL;
@@ -2405,7 +2397,7 @@ static inline int transport_tcq_window_closed(struct se_device *dev)
} else
msleep(PYX_TRANSPORT_WINDOW_CLOSED_WAIT_LONG);
- wake_up_interruptible(&dev->dev_queue_obj->thread_wq);
+ wake_up_interruptible(&dev->dev_queue_obj.thread_wq);
return 0;
}
@@ -3744,7 +3736,7 @@ static void transport_complete_task_attr(struct se_cmd *cmd)
* to do the processing of the Active tasks.
*/
if (new_active_tasks != 0)
- wake_up_interruptible(&dev->dev_queue_obj->thread_wq);
+ wake_up_interruptible(&dev->dev_queue_obj.thread_wq);
}
static void transport_generic_complete_ok(struct se_cmd *cmd)
@@ -5254,7 +5246,7 @@ static int transport_lun_wait_for_tasks(struct se_cmd *cmd, struct se_lun *lun)
atomic_set(&T_TASK(cmd)->transport_lun_fe_stop, 1);
spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
- wake_up_interruptible(&SE_DEV(cmd)->dev_queue_obj->thread_wq);
+ wake_up_interruptible(&SE_DEV(cmd)->dev_queue_obj.thread_wq);
ret = transport_stop_tasks_for_cmd(cmd);
@@ -5267,7 +5259,7 @@ static int transport_lun_wait_for_tasks(struct se_cmd *cmd, struct se_lun *lun)
DEBUG_TRANSPORT_S("ConfigFS: ITT[0x%08x] - stopped cmd....\n",
CMD_TFO(cmd)->get_task_tag(cmd));
}
- transport_remove_cmd_from_queue(cmd, SE_DEV(cmd)->dev_queue_obj);
+ transport_remove_cmd_from_queue(cmd, &SE_DEV(cmd)->dev_queue_obj);
return 0;
}
@@ -5480,7 +5472,7 @@ static void transport_generic_wait_for_tasks(
spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
- wake_up_interruptible(&SE_DEV(cmd)->dev_queue_obj->thread_wq);
+ wake_up_interruptible(&SE_DEV(cmd)->dev_queue_obj.thread_wq);
wait_for_completion(&T_TASK(cmd)->t_transport_stop_comp);
@@ -5898,7 +5890,7 @@ static void transport_processing_shutdown(struct se_device *dev)
cmd, TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE,
0);
transport_remove_cmd_from_queue(cmd,
- SE_DEV(cmd)->dev_queue_obj);
+ &SE_DEV(cmd)->dev_queue_obj);
transport_lun_remove_cmd(cmd);
transport_cmd_check_stop(cmd, 1, 0);
@@ -5907,7 +5899,7 @@ static void transport_processing_shutdown(struct se_device *dev)
&T_TASK(cmd)->t_state_lock, flags);
transport_remove_cmd_from_queue(cmd,
- SE_DEV(cmd)->dev_queue_obj);
+ &SE_DEV(cmd)->dev_queue_obj);
transport_lun_remove_cmd(cmd);
@@ -5927,7 +5919,7 @@ static void transport_processing_shutdown(struct se_device *dev)
transport_send_check_condition_and_sense(cmd,
TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE, 0);
transport_remove_cmd_from_queue(cmd,
- SE_DEV(cmd)->dev_queue_obj);
+ &SE_DEV(cmd)->dev_queue_obj);
transport_lun_remove_cmd(cmd);
transport_cmd_check_stop(cmd, 1, 0);
@@ -5936,7 +5928,7 @@ static void transport_processing_shutdown(struct se_device *dev)
&T_TASK(cmd)->t_state_lock, flags);
transport_remove_cmd_from_queue(cmd,
- SE_DEV(cmd)->dev_queue_obj);
+ &SE_DEV(cmd)->dev_queue_obj);
transport_lun_remove_cmd(cmd);
if (transport_cmd_check_stop(cmd, 1, 0))
@@ -5949,7 +5941,7 @@ static void transport_processing_shutdown(struct se_device *dev)
/*
* Empty the struct se_device's struct se_cmd list.
*/
- while ((qr = transport_get_qr_from_queue(dev->dev_queue_obj))) {
+ while ((qr = transport_get_qr_from_queue(&dev->dev_queue_obj))) {
cmd = qr->cmd;
state = qr->state;
kfree(qr);
@@ -5985,8 +5977,8 @@ static int transport_processing_thread(void *param)
set_user_nice(current, -20);
while (!kthread_should_stop()) {
- ret = wait_event_interruptible(dev->dev_queue_obj->thread_wq,
- atomic_read(&dev->dev_queue_obj->queue_cnt) ||
+ ret = wait_event_interruptible(dev->dev_queue_obj.thread_wq,
+ atomic_read(&dev->dev_queue_obj.queue_cnt) ||
kthread_should_stop());
if (ret < 0)
goto out;
@@ -6002,7 +5994,7 @@ static int transport_processing_thread(void *param)
get_cmd:
__transport_execute_tasks(dev);
- qr = transport_get_qr_from_queue(dev->dev_queue_obj);
+ qr = transport_get_qr_from_queue(&dev->dev_queue_obj);
if (!(qr))
continue;
diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h
index 23dd6b7..112eb38 100644
--- a/include/target/target_core_base.h
+++ b/include/target/target_core_base.h
@@ -778,7 +778,7 @@ struct se_device {
struct se_obj dev_obj;
struct se_obj dev_access_obj;
struct se_obj dev_export_obj;
- struct se_queue_obj *dev_queue_obj;
+ struct se_queue_obj dev_queue_obj;
spinlock_t delayed_cmd_lock;
spinlock_t ordered_cmd_lock;
spinlock_t execute_task_lock;
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 021/103] target: remove extraneous returns at end of void functions
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (19 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 020/103] target: Make dev_queue_obj a member of se_device instead of a pointer Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 022/103] target: Ensure transport_dump_vpd_ident_type returns null-terminated str Nicholas A. Bellinger
` (8 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover
From: Andy Grover <agrover@redhat.com>
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_configfs.c | 3 ---
drivers/target/target_core_device.c | 12 ------------
drivers/target/target_core_pscsi.c | 2 --
drivers/target/target_core_transport.c | 2 --
drivers/target/tcm_fc/tfc_cmd.c | 1 -
5 files changed, 0 insertions(+), 20 deletions(-)
diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c
index 27effbc..8196b72 100644
--- a/drivers/target/target_core_configfs.c
+++ b/drivers/target/target_core_configfs.c
@@ -601,7 +601,6 @@ void target_fabric_configfs_deregister(
printk("<<<<<<<<<<<<<<<<<<<<<< END FABRIC API >>>>>>>>>>>>>>>>>"
">>>>>\n");
- return;
}
EXPORT_SYMBOL(target_fabric_configfs_deregister);
@@ -3304,8 +3303,6 @@ static void __exit target_core_exit_configfs(void)
core_dev_release_virtual_lun0();
rd_module_exit();
release_se_kmem_caches();
-
- return;
}
MODULE_DESCRIPTION("Target_Core_Mod/ConfigFS");
diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c
index fe5474c..ff57891 100644
--- a/drivers/target/target_core_device.c
+++ b/drivers/target/target_core_device.c
@@ -339,8 +339,6 @@ void core_dec_lacl_count(struct se_node_acl *se_nacl, struct se_cmd *se_cmd)
deve = &se_nacl->device_list[se_cmd->orig_fe_lun];
deve->deve_cmds--;
spin_unlock_irq(&se_nacl->device_list_lock);
-
- return;
}
void core_update_device_list_access(
@@ -360,8 +358,6 @@ void core_update_device_list_access(
deve->lun_flags |= TRANSPORT_LUNFLAGS_READ_ONLY;
}
spin_unlock_irq(&nacl->device_list_lock);
-
- return;
}
/* core_update_device_list_for_node():
@@ -508,8 +504,6 @@ void core_clear_lun_from_tpg(struct se_lun *lun, struct se_portal_group *tpg)
spin_lock_bh(&tpg->acl_node_lock);
}
spin_unlock_bh(&tpg->acl_node_lock);
-
- return;
}
static struct se_port *core_alloc_port(struct se_device *dev)
@@ -623,8 +617,6 @@ static void core_release_port(struct se_device *dev, struct se_port *port)
list_del(&port->sep_list);
dev->dev_port_count--;
kfree(port);
-
- return;
}
int core_dev_export(
@@ -761,8 +753,6 @@ void se_release_device_for_hba(struct se_device *dev)
se_release_vpd_for_dev(dev);
kfree(dev);
-
- return;
}
void se_release_vpd_for_dev(struct se_device *dev)
@@ -776,8 +766,6 @@ void se_release_vpd_for_dev(struct se_device *dev)
kfree(vpd);
}
spin_unlock(&DEV_T10_WWN(dev)->t10_vpd_lock);
-
- return;
}
/* se_free_virtual_device():
diff --git a/drivers/target/target_core_pscsi.c b/drivers/target/target_core_pscsi.c
index e221dea..afe5f57 100644
--- a/drivers/target/target_core_pscsi.c
+++ b/drivers/target/target_core_pscsi.c
@@ -1369,8 +1369,6 @@ static inline void pscsi_process_SAM_status(
transport_complete_task(task, 0);
break;
}
-
- return;
}
static void pscsi_req_done(struct request *req, int uptodate)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 7e0d69a..473032e 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -1103,8 +1103,6 @@ static void transport_add_tasks_from_cmd(struct se_cmd *cmd)
task_prev = task;
}
spin_unlock_irqrestore(&dev->execute_task_lock, flags);
-
- return;
}
/* transport_get_task_from_execute_queue():
diff --git a/drivers/target/tcm_fc/tfc_cmd.c b/drivers/target/tcm_fc/tfc_cmd.c
index 92a449a..5823ccf 100644
--- a/drivers/target/tcm_fc/tfc_cmd.c
+++ b/drivers/target/tcm_fc/tfc_cmd.c
@@ -670,7 +670,6 @@ static void ft_send_cmd(struct ft_cmd *cmd)
err:
ft_send_resp_code(cmd, FCP_CMND_FIELDS_INVALID);
- return;
}
/*
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 022/103] target: Ensure transport_dump_vpd_ident_type returns null-terminated str
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (20 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 021/103] target: remove extraneous returns at end of void functions Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 023/103] target: Function pointers don't need to use '&' to be assigned Nicholas A. Bellinger
` (7 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover
From: Andy Grover <agrover@redhat.com>
strncpy() may leave the string not NULL-terminated if the buffer length
is shorter than the string. Return -EINVAL if too short, instead.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_transport.c | 7 +++++--
1 files changed, 5 insertions(+), 2 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 473032e..72e4726 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -1392,10 +1392,13 @@ int transport_dump_vpd_ident_type(
break;
}
- if (p_buf)
+ if (p_buf) {
+ if (p_buf_len < strlen(buf)+1)
+ return -EINVAL;
strncpy(p_buf, buf, p_buf_len);
- else
+ } else {
printk("%s", buf);
+ }
return ret;
}
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 023/103] target: Function pointers don't need to use '&' to be assigned
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (21 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 022/103] target: Ensure transport_dump_vpd_ident_type returns null-terminated str Nicholas A. Bellinger
@ 2011-07-21 7:04 ` Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 024/103] target: Fix comment in __transport_execute_tasks() Nicholas A. Bellinger
` (6 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:04 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover
From: Andy Grover <agrover@redhat.com>
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_transport.c | 12 ++++++------
1 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 72e4726..acb5ef7 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -3226,7 +3226,7 @@ static int transport_generic_cmd_sequencer(
cmd->transport_emulate_cdb =
(T10_ALUA(su_dev)->alua_type ==
SPC3_ALUA_EMULATED) ?
- &core_emulate_report_target_port_groups :
+ core_emulate_report_target_port_groups :
NULL;
}
size = (cdb[6] << 24) | (cdb[7] << 16) |
@@ -3273,7 +3273,7 @@ static int transport_generic_cmd_sequencer(
cmd->transport_emulate_cdb =
(T10_RES(su_dev)->res_type ==
SPC3_PERSISTENT_RESERVATIONS) ?
- &core_scsi3_emulate_pr : NULL;
+ core_scsi3_emulate_pr : NULL;
size = (cdb[7] << 8) + cdb[8];
cmd->se_cmd_flags |= SCF_SCSI_CONTROL_NONSG_IO_CDB;
break;
@@ -3296,7 +3296,7 @@ static int transport_generic_cmd_sequencer(
cmd->transport_emulate_cdb =
(T10_ALUA(su_dev)->alua_type ==
SPC3_ALUA_EMULATED) ?
- &core_emulate_set_target_port_groups :
+ core_emulate_set_target_port_groups :
NULL;
}
@@ -3393,7 +3393,7 @@ static int transport_generic_cmd_sequencer(
cmd->transport_emulate_cdb =
(T10_RES(su_dev)->res_type !=
SPC_PASSTHROUGH) ?
- &core_scsi2_emulate_crh : NULL;
+ core_scsi2_emulate_crh : NULL;
cmd->se_cmd_flags |= SCF_SCSI_NON_DATA_CDB;
break;
case RELEASE:
@@ -3410,7 +3410,7 @@ static int transport_generic_cmd_sequencer(
cmd->transport_emulate_cdb =
(T10_RES(su_dev)->res_type !=
SPC_PASSTHROUGH) ?
- &core_scsi2_emulate_crh : NULL;
+ core_scsi2_emulate_crh : NULL;
cmd->se_cmd_flags |= SCF_SCSI_NON_DATA_CDB;
break;
case SYNCHRONIZE_CACHE:
@@ -3516,7 +3516,7 @@ static int transport_generic_cmd_sequencer(
break;
case REPORT_LUNS:
cmd->transport_emulate_cdb =
- &transport_core_report_lun_response;
+ transport_core_report_lun_response;
size = (cdb[6] << 24) | (cdb[7] << 16) | (cdb[8] << 8) | cdb[9];
/*
* Do implict HEAD_OF_QUEUE processing for REPORT_LUNS
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 024/103] target: Fix comment in __transport_execute_tasks()
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (22 preceding siblings ...)
2011-07-21 7:04 ` [PATCH 023/103] target: Function pointers don't need to use '&' to be assigned Nicholas A. Bellinger
@ 2011-07-21 7:05 ` Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 025/103] target: Misc style cleanups Nicholas A. Bellinger
` (5 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:05 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover
From: Andy Grover <agrover@redhat.com>
cmd->transport_emulate_cdb is used for a variety of purposes beyond
REPORT_LUNS currently, so update the comment explaining its use to
reflect this.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_transport.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index acb5ef7..a224010 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -2576,7 +2576,7 @@ check_depth:
spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
/*
* The struct se_cmd->transport_emulate_cdb() function pointer is used
- * to grab REPORT_LUNS CDBs before they hit the
+ * to grab REPORT_LUNS and other CDBs we want to handle before they hit the
* struct se_subsystem_api->do_task() caller below.
*/
if (cmd->transport_emulate_cdb) {
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 025/103] target: Misc style cleanups
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (23 preceding siblings ...)
2011-07-21 7:05 ` [PATCH 024/103] target: Fix comment in __transport_execute_tasks() Nicholas A. Bellinger
@ 2011-07-21 7:05 ` Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 026/103] target: rename struct pr_reservation_template to pr_reservation Nicholas A. Bellinger
` (4 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:05 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover
From: Andy Grover <agrover@redhat.com>
Typo
Local variables should be declared on separate lines (simplifies diffs)
Prefer (!foo) over (foo == 0)
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_transport.c | 10 ++++++----
1 files changed, 6 insertions(+), 4 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index a224010..101bc20 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -644,7 +644,7 @@ static int transport_cmd_check_stop(
}
/*
* Determine if frontend context caller is requesting the stopping of
- * this command for frontend excpections.
+ * this command for frontend exceptions.
*/
if (atomic_read(&T_TASK(cmd)->t_transport_stop)) {
DEBUG_CS("%s:%d atomic_read(&T_TASK(cmd)->t_transport_stop) =="
@@ -1314,7 +1314,8 @@ int transport_dump_vpd_assoc(
int p_buf_len)
{
unsigned char buf[VPD_TMP_BUF_SIZE];
- int ret = 0, len;
+ int ret = 0;
+ int len;
memset(buf, 0, VPD_TMP_BUF_SIZE);
len = sprintf(buf, "T10 VPD Identifier Association: ");
@@ -1361,7 +1362,8 @@ int transport_dump_vpd_ident_type(
int p_buf_len)
{
unsigned char buf[VPD_TMP_BUF_SIZE];
- int ret = 0, len;
+ int ret = 0;
+ int len;
memset(buf, 0, VPD_TMP_BUF_SIZE);
len = sprintf(buf, "T10 VPD Identifier Type: ");
@@ -2510,7 +2512,7 @@ static int transport_execute_tasks(struct se_cmd *cmd)
* attribute for the tasks of the received struct se_cmd CDB
*/
add_tasks = transport_execute_task_attr(cmd);
- if (add_tasks == 0)
+ if (!add_tasks)
goto execute_tasks;
/*
* This calls transport_add_tasks_from_cmd() to handle
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 026/103] target: rename struct pr_reservation_template to pr_reservation
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (24 preceding siblings ...)
2011-07-21 7:05 ` [PATCH 025/103] target: Misc style cleanups Nicholas A. Bellinger
@ 2011-07-21 7:05 ` Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 027/103] target: Remove #defines that just perform indirection Nicholas A. Bellinger
` (3 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:05 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover
From: Andy Grover <agrover@redhat.com>
"_template" in the name doesn't seem (to me) to be worth the added
verbosity.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_pr.c | 4 ++--
drivers/target/target_core_pr.h | 2 +-
include/target/target_core_base.h | 6 +++---
3 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/drivers/target/target_core_pr.c b/drivers/target/target_core_pr.c
index 99b97db..2ddbe0d 100644
--- a/drivers/target/target_core_pr.c
+++ b/drivers/target/target_core_pr.c
@@ -4019,7 +4019,7 @@ static int core_scsi3_pri_read_full_status(struct se_cmd *cmd)
struct se_subsystem_dev *su_dev = SU_DEV(se_dev);
struct se_portal_group *se_tpg;
struct t10_pr_registration *pr_reg, *pr_reg_tmp;
- struct t10_reservation_template *pr_tmpl = &SU_DEV(se_dev)->t10_reservation;
+ struct t10_reservation *pr_tmpl = &SU_DEV(se_dev)->t10_reservation;
unsigned char *buf = (unsigned char *)T_TASK(cmd)->t_task_buf;
u32 add_desc_len = 0, add_len = 0, desc_len, exp_desc_len;
u32 off = 8; /* off into first Full Status descriptor */
@@ -4213,7 +4213,7 @@ static int core_pt_seq_non_holder(
int core_setup_reservations(struct se_device *dev, int force_pt)
{
struct se_subsystem_dev *su_dev = dev->se_sub_dev;
- struct t10_reservation_template *rest = &su_dev->t10_reservation;
+ struct t10_reservation *rest = &su_dev->t10_reservation;
/*
* If this device is from Target_Core_Mod/pSCSI, use the reservations
* of the Underlying SCSI hardware. In Linux/SCSI terms, this can
diff --git a/drivers/target/target_core_pr.h b/drivers/target/target_core_pr.h
index 5603bcf..c8f47d0 100644
--- a/drivers/target/target_core_pr.h
+++ b/drivers/target/target_core_pr.h
@@ -49,7 +49,7 @@ extern int core_pr_dump_initiator_port(struct t10_pr_registration *,
char *, u32);
extern int core_scsi2_emulate_crh(struct se_cmd *);
extern int core_scsi3_alloc_aptpl_registration(
- struct t10_reservation_template *, u64,
+ struct t10_reservation *, u64,
unsigned char *, unsigned char *, u32,
unsigned char *, u16, u32, int, int, u8);
extern int core_scsi3_check_aptpl_registration(struct se_device *,
diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h
index 112eb38..2b989b3 100644
--- a/include/target/target_core_base.h
+++ b/include/target/target_core_base.h
@@ -361,13 +361,13 @@ struct t10_reservation_ops {
int (*t10_pr_clear)(struct se_cmd *);
};
-struct t10_reservation_template {
+struct t10_reservation {
/* Reservation effects all target ports */
int pr_all_tg_pt;
/* Activate Persistence across Target Power Loss enabled
* for SCSI device */
int pr_aptpl_active;
- /* Used by struct t10_reservation_template->pr_aptpl_buf_len */
+ /* Used by struct t10_reservation->pr_aptpl_buf_len */
#define PR_APTPL_BUF_LEN 8192
u32 pr_aptpl_buf_len;
u32 pr_generation;
@@ -722,7 +722,7 @@ struct se_subsystem_dev {
/* T10 Inquiry and VPD WWN Information */
struct t10_wwn t10_wwn;
/* T10 SPC-2 + SPC-3 Reservations */
- struct t10_reservation_template t10_reservation;
+ struct t10_reservation t10_reservation;
spinlock_t se_dev_lock;
void *se_dev_su_ptr;
struct list_head se_dev_node;
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 027/103] target: Remove #defines that just perform indirection
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (25 preceding siblings ...)
2011-07-21 7:05 ` [PATCH 026/103] target: rename struct pr_reservation_template to pr_reservation Nicholas A. Bellinger
@ 2011-07-21 7:05 ` Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 028/103] target: Inline transport_get_task_from_execute_queue() Nicholas A. Bellinger
` (2 subsequent siblings)
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:05 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover
From: Andy Grover <agrover@redhat.com>
Generally, the kernel style prefers to use such macros for longer
conversions, container_of, that sort of thing. If it's just to refer to a
member variable, then for clarity it can just be used directly.
Usually the length difference is small, but one macro does shorten a
struct member name: se_subsystem_dev.t10_reservation. We shorten
t10_reservation member name (but not type) to t10_pr.
Finally, there are some bare numbers in _stat.c that this patch fixes
by using sizeof(), and we also remove extraneous memcpy()s.
Finally, make string fields in struct t10_wwn chars, to cut down on
casting.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
---
drivers/target/loopback/tcm_loop.c | 4 +-
drivers/target/target_core_alua.c | 138 ++--
drivers/target/target_core_cdb.c | 134 ++--
drivers/target/target_core_configfs.c | 79 +-
drivers/target/target_core_device.c | 297 ++++----
drivers/target/target_core_fabric_configfs.c | 12 +-
drivers/target/target_core_file.c | 24 +-
drivers/target/target_core_iblock.c | 42 +-
drivers/target/target_core_pr.c | 328 ++++----
drivers/target/target_core_pscsi.c | 19 +-
drivers/target/target_core_rd.c | 28 +-
drivers/target/target_core_stat.c | 112 ++--
drivers/target/target_core_tmr.c | 82 +-
drivers/target/target_core_tpg.c | 106 ++--
drivers/target/target_core_transport.c | 1196 +++++++++++++-------------
drivers/target/target_core_ua.c | 16 +-
drivers/target/tcm_fc/tfc_cmd.c | 6 +-
drivers/target/tcm_fc/tfc_io.c | 4 +-
include/target/target_core_base.h | 38 +-
include/target/target_core_transport.h | 3 -
20 files changed, 1318 insertions(+), 1350 deletions(-)
diff --git a/drivers/target/loopback/tcm_loop.c b/drivers/target/loopback/tcm_loop.c
index 26e3681..1f1695a 100644
--- a/drivers/target/loopback/tcm_loop.c
+++ b/drivers/target/loopback/tcm_loop.c
@@ -118,7 +118,7 @@ static struct se_cmd *tcm_loop_allocate_core_cmd(
* Signal BIDI usage with T_TASK(cmd)->t_tasks_bidi
*/
if (scsi_bidi_cmnd(sc))
- T_TASK(se_cmd)->t_tasks_bidi = 1;
+ se_cmd->t_task->t_tasks_bidi = 1;
/*
* Locate the struct se_lun pointer and attach it to struct se_cmd
*/
@@ -176,7 +176,7 @@ static int tcm_loop_new_cmd_map(struct se_cmd *se_cmd)
* For BIDI commands, pass in the extra READ buffer
* to transport_generic_map_mem_to_cmd() below..
*/
- if (T_TASK(se_cmd)->t_tasks_bidi) {
+ if (se_cmd->t_task->t_tasks_bidi) {
struct scsi_data_buffer *sdb = scsi_in(sc);
mem_bidi_ptr = (void *)sdb->table.sgl;
diff --git a/drivers/target/target_core_alua.c b/drivers/target/target_core_alua.c
index 62522d6..bfc42ad 100644
--- a/drivers/target/target_core_alua.c
+++ b/drivers/target/target_core_alua.c
@@ -61,16 +61,16 @@ struct t10_alua_lu_gp *default_lu_gp;
*/
int core_emulate_report_target_port_groups(struct se_cmd *cmd)
{
- struct se_subsystem_dev *su_dev = SE_DEV(cmd)->se_sub_dev;
+ struct se_subsystem_dev *su_dev = cmd->se_lun->lun_se_dev->se_sub_dev;
struct se_port *port;
struct t10_alua_tg_pt_gp *tg_pt_gp;
struct t10_alua_tg_pt_gp_member *tg_pt_gp_mem;
- unsigned char *buf = (unsigned char *)T_TASK(cmd)->t_task_buf;
+ unsigned char *buf = (unsigned char *)cmd->t_task->t_task_buf;
u32 rd_len = 0, off = 4; /* Skip over RESERVED area to first
Target port group descriptor */
- spin_lock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
- list_for_each_entry(tg_pt_gp, &T10_ALUA(su_dev)->tg_pt_gps_list,
+ spin_lock(&su_dev->t10_alua.tg_pt_gps_lock);
+ list_for_each_entry(tg_pt_gp, &su_dev->t10_alua.tg_pt_gps_list,
tg_pt_gp_list) {
/*
* PREF: Preferred target port bit, determine if this
@@ -132,7 +132,7 @@ int core_emulate_report_target_port_groups(struct se_cmd *cmd)
}
spin_unlock(&tg_pt_gp->tg_pt_gp_lock);
}
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
/*
* Set the RETURN DATA LENGTH set in the header of the DataIN Payload
*/
@@ -151,13 +151,13 @@ int core_emulate_report_target_port_groups(struct se_cmd *cmd)
*/
int core_emulate_set_target_port_groups(struct se_cmd *cmd)
{
- struct se_device *dev = SE_DEV(cmd);
- struct se_subsystem_dev *su_dev = SE_DEV(cmd)->se_sub_dev;
- struct se_port *port, *l_port = SE_LUN(cmd)->lun_sep;
- struct se_node_acl *nacl = SE_SESS(cmd)->se_node_acl;
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
+ struct se_subsystem_dev *su_dev = dev->se_sub_dev;
+ struct se_port *port, *l_port = cmd->se_lun->lun_sep;
+ struct se_node_acl *nacl = cmd->se_sess->se_node_acl;
struct t10_alua_tg_pt_gp *tg_pt_gp = NULL, *l_tg_pt_gp;
struct t10_alua_tg_pt_gp_member *tg_pt_gp_mem, *l_tg_pt_gp_mem;
- unsigned char *buf = (unsigned char *)T_TASK(cmd)->t_task_buf;
+ unsigned char *buf = (unsigned char *)cmd->t_task->t_task_buf;
unsigned char *ptr = &buf[4]; /* Skip over RESERVED area in header */
u32 len = 4; /* Skip over RESERVED area in header */
int alua_access_state, primary = 0, rc;
@@ -232,9 +232,9 @@ int core_emulate_set_target_port_groups(struct se_cmd *cmd)
* Locate the matching target port group ID from
* the global tg_pt_gp list
*/
- spin_lock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_lock(&su_dev->t10_alua.tg_pt_gps_lock);
list_for_each_entry(tg_pt_gp,
- &T10_ALUA(su_dev)->tg_pt_gps_list,
+ &su_dev->t10_alua.tg_pt_gps_list,
tg_pt_gp_list) {
if (!(tg_pt_gp->tg_pt_gp_valid_id))
continue;
@@ -244,18 +244,18 @@ int core_emulate_set_target_port_groups(struct se_cmd *cmd)
atomic_inc(&tg_pt_gp->tg_pt_gp_ref_cnt);
smp_mb__after_atomic_inc();
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
rc = core_alua_do_port_transition(tg_pt_gp,
dev, l_port, nacl,
alua_access_state, 1);
- spin_lock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_lock(&su_dev->t10_alua.tg_pt_gps_lock);
atomic_dec(&tg_pt_gp->tg_pt_gp_ref_cnt);
smp_mb__after_atomic_dec();
break;
}
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
/*
* If not matching target port group ID can be located
* throw an exception with ASCQ: INVALID_PARAMETER_LIST
@@ -472,7 +472,7 @@ static int core_alua_state_check(
unsigned char *cdb,
u8 *alua_ascq)
{
- struct se_lun *lun = SE_LUN(cmd);
+ struct se_lun *lun = cmd->se_lun;
struct se_port *port = lun->lun_sep;
struct t10_alua_tg_pt_gp *tg_pt_gp;
struct t10_alua_tg_pt_gp_member *tg_pt_gp_mem;
@@ -874,9 +874,9 @@ int core_alua_do_port_transition(
smp_mb__after_atomic_inc();
spin_unlock(&lu_gp->lu_gp_lock);
- spin_lock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_lock(&su_dev->t10_alua.tg_pt_gps_lock);
list_for_each_entry(tg_pt_gp,
- &T10_ALUA(su_dev)->tg_pt_gps_list,
+ &su_dev->t10_alua.tg_pt_gps_list,
tg_pt_gp_list) {
if (!(tg_pt_gp->tg_pt_gp_valid_id))
@@ -901,7 +901,7 @@ int core_alua_do_port_transition(
}
atomic_inc(&tg_pt_gp->tg_pt_gp_ref_cnt);
smp_mb__after_atomic_inc();
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
/*
* core_alua_do_transition_tg_pt() will always return
* success.
@@ -909,11 +909,11 @@ int core_alua_do_port_transition(
core_alua_do_transition_tg_pt(tg_pt_gp, port,
nacl, md_buf, new_state, explict);
- spin_lock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_lock(&su_dev->t10_alua.tg_pt_gps_lock);
atomic_dec(&tg_pt_gp->tg_pt_gp_ref_cnt);
smp_mb__after_atomic_dec();
}
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
spin_lock(&lu_gp->lu_gp_lock);
atomic_dec(&lu_gp_mem->lu_gp_mem_ref_cnt);
@@ -950,11 +950,11 @@ static int core_alua_update_tpg_secondary_metadata(
memset(wwn, 0, ALUA_SECONDARY_METADATA_WWN_LEN);
len = snprintf(wwn, ALUA_SECONDARY_METADATA_WWN_LEN, "%s",
- TPG_TFO(se_tpg)->tpg_get_wwn(se_tpg));
+ se_tpg->se_tpg_tfo->tpg_get_wwn(se_tpg));
- if (TPG_TFO(se_tpg)->tpg_get_tag != NULL)
+ if (se_tpg->se_tpg_tfo->tpg_get_tag != NULL)
snprintf(wwn+len, ALUA_SECONDARY_METADATA_WWN_LEN-len, "+%hu",
- TPG_TFO(se_tpg)->tpg_get_tag(se_tpg));
+ se_tpg->se_tpg_tfo->tpg_get_tag(se_tpg));
len = snprintf(md_buf, md_buf_len, "alua_tg_pt_offline=%d\n"
"alua_tg_pt_status=0x%02x\n",
@@ -962,7 +962,7 @@ static int core_alua_update_tpg_secondary_metadata(
port->sep_tg_pt_secondary_stat);
snprintf(path, ALUA_METADATA_PATH_LEN, "/var/target/alua/%s/%s/lun_%u",
- TPG_TFO(se_tpg)->get_fabric_name(), wwn,
+ se_tpg->se_tpg_tfo->get_fabric_name(), wwn,
port->sep_lun->unpacked_lun);
return core_alua_write_tpg_metadata(path, md_buf, len);
@@ -1190,7 +1190,7 @@ void core_alua_free_lu_gp(struct t10_alua_lu_gp *lu_gp)
void core_alua_free_lu_gp_mem(struct se_device *dev)
{
struct se_subsystem_dev *su_dev = dev->se_sub_dev;
- struct t10_alua *alua = T10_ALUA(su_dev);
+ struct t10_alua *alua = &su_dev->t10_alua;
struct t10_alua_lu_gp *lu_gp;
struct t10_alua_lu_gp_member *lu_gp_mem;
@@ -1312,14 +1312,14 @@ struct t10_alua_tg_pt_gp *core_alua_allocate_tg_pt_gp(
tg_pt_gp->tg_pt_gp_trans_delay_msecs = ALUA_DEFAULT_TRANS_DELAY_MSECS;
if (def_group) {
- spin_lock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_lock(&su_dev->t10_alua.tg_pt_gps_lock);
tg_pt_gp->tg_pt_gp_id =
- T10_ALUA(su_dev)->alua_tg_pt_gps_counter++;
+ su_dev->t10_alua.alua_tg_pt_gps_counter++;
tg_pt_gp->tg_pt_gp_valid_id = 1;
- T10_ALUA(su_dev)->alua_tg_pt_gps_count++;
+ su_dev->t10_alua.alua_tg_pt_gps_count++;
list_add_tail(&tg_pt_gp->tg_pt_gp_list,
- &T10_ALUA(su_dev)->tg_pt_gps_list);
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ &su_dev->t10_alua.tg_pt_gps_list);
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
}
return tg_pt_gp;
@@ -1341,19 +1341,19 @@ int core_alua_set_tg_pt_gp_id(
return -EINVAL;
}
- spin_lock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
- if (T10_ALUA(su_dev)->alua_tg_pt_gps_count == 0x0000ffff) {
+ spin_lock(&su_dev->t10_alua.tg_pt_gps_lock);
+ if (su_dev->t10_alua.alua_tg_pt_gps_count == 0x0000ffff) {
printk(KERN_ERR "Maximum ALUA alua_tg_pt_gps_count:"
" 0x0000ffff reached\n");
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
kmem_cache_free(t10_alua_tg_pt_gp_cache, tg_pt_gp);
return -ENOSPC;
}
again:
tg_pt_gp_id_tmp = (tg_pt_gp_id != 0) ? tg_pt_gp_id :
- T10_ALUA(su_dev)->alua_tg_pt_gps_counter++;
+ su_dev->t10_alua.alua_tg_pt_gps_counter++;
- list_for_each_entry(tg_pt_gp_tmp, &T10_ALUA(su_dev)->tg_pt_gps_list,
+ list_for_each_entry(tg_pt_gp_tmp, &su_dev->t10_alua.tg_pt_gps_list,
tg_pt_gp_list) {
if (tg_pt_gp_tmp->tg_pt_gp_id == tg_pt_gp_id_tmp) {
if (!(tg_pt_gp_id))
@@ -1361,7 +1361,7 @@ again:
printk(KERN_ERR "ALUA Target Port Group ID: %hu already"
" exists, ignoring request\n", tg_pt_gp_id);
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
return -EINVAL;
}
}
@@ -1369,9 +1369,9 @@ again:
tg_pt_gp->tg_pt_gp_id = tg_pt_gp_id_tmp;
tg_pt_gp->tg_pt_gp_valid_id = 1;
list_add_tail(&tg_pt_gp->tg_pt_gp_list,
- &T10_ALUA(su_dev)->tg_pt_gps_list);
- T10_ALUA(su_dev)->alua_tg_pt_gps_count++;
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ &su_dev->t10_alua.tg_pt_gps_list);
+ su_dev->t10_alua.alua_tg_pt_gps_count++;
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
return 0;
}
@@ -1411,10 +1411,10 @@ void core_alua_free_tg_pt_gp(
* no assications *OR* explict ALUA via SET_TARGET_PORT_GROUPS
* can be made while we are releasing struct t10_alua_tg_pt_gp.
*/
- spin_lock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_lock(&su_dev->t10_alua.tg_pt_gps_lock);
list_del(&tg_pt_gp->tg_pt_gp_list);
- T10_ALUA(su_dev)->alua_tg_pt_gps_counter--;
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ su_dev->t10_alua.alua_tg_pt_gps_counter--;
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
/*
* Allow a struct t10_alua_tg_pt_gp_member * referenced by
* core_alua_get_tg_pt_gp_by_name() in
@@ -1446,9 +1446,9 @@ void core_alua_free_tg_pt_gp(
* default_tg_pt_gp.
*/
spin_lock(&tg_pt_gp_mem->tg_pt_gp_mem_lock);
- if (tg_pt_gp != T10_ALUA(su_dev)->default_tg_pt_gp) {
+ if (tg_pt_gp != su_dev->t10_alua.default_tg_pt_gp) {
__core_alua_attach_tg_pt_gp_mem(tg_pt_gp_mem,
- T10_ALUA(su_dev)->default_tg_pt_gp);
+ su_dev->t10_alua.default_tg_pt_gp);
} else
tg_pt_gp_mem->tg_pt_gp = NULL;
spin_unlock(&tg_pt_gp_mem->tg_pt_gp_mem_lock);
@@ -1463,7 +1463,7 @@ void core_alua_free_tg_pt_gp(
void core_alua_free_tg_pt_gp_mem(struct se_port *port)
{
struct se_subsystem_dev *su_dev = port->sep_lun->lun_se_dev->se_sub_dev;
- struct t10_alua *alua = T10_ALUA(su_dev);
+ struct t10_alua *alua = &su_dev->t10_alua;
struct t10_alua_tg_pt_gp *tg_pt_gp;
struct t10_alua_tg_pt_gp_member *tg_pt_gp_mem;
@@ -1501,19 +1501,19 @@ static struct t10_alua_tg_pt_gp *core_alua_get_tg_pt_gp_by_name(
struct t10_alua_tg_pt_gp *tg_pt_gp;
struct config_item *ci;
- spin_lock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
- list_for_each_entry(tg_pt_gp, &T10_ALUA(su_dev)->tg_pt_gps_list,
+ spin_lock(&su_dev->t10_alua.tg_pt_gps_lock);
+ list_for_each_entry(tg_pt_gp, &su_dev->t10_alua.tg_pt_gps_list,
tg_pt_gp_list) {
if (!(tg_pt_gp->tg_pt_gp_valid_id))
continue;
ci = &tg_pt_gp->tg_pt_gp_group.cg_item;
if (!(strcmp(config_item_name(ci), name))) {
atomic_inc(&tg_pt_gp->tg_pt_gp_ref_cnt);
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
return tg_pt_gp;
}
}
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
return NULL;
}
@@ -1523,9 +1523,9 @@ static void core_alua_put_tg_pt_gp_from_name(
{
struct se_subsystem_dev *su_dev = tg_pt_gp->tg_pt_gp_su_dev;
- spin_lock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_lock(&su_dev->t10_alua.tg_pt_gps_lock);
atomic_dec(&tg_pt_gp->tg_pt_gp_ref_cnt);
- spin_unlock(&T10_ALUA(su_dev)->tg_pt_gps_lock);
+ spin_unlock(&su_dev->t10_alua.tg_pt_gps_lock);
}
/*
@@ -1563,7 +1563,7 @@ ssize_t core_alua_show_tg_pt_gp_info(struct se_port *port, char *page)
{
struct se_subsystem_dev *su_dev = port->sep_lun->lun_se_dev->se_sub_dev;
struct config_item *tg_pt_ci;
- struct t10_alua *alua = T10_ALUA(su_dev);
+ struct t10_alua *alua = &su_dev->t10_alua;
struct t10_alua_tg_pt_gp *tg_pt_gp;
struct t10_alua_tg_pt_gp_member *tg_pt_gp_mem;
ssize_t len = 0;
@@ -1613,10 +1613,10 @@ ssize_t core_alua_store_tg_pt_gp_info(
tpg = port->sep_tpg;
lun = port->sep_lun;
- if (T10_ALUA(su_dev)->alua_type != SPC3_ALUA_EMULATED) {
+ if (su_dev->t10_alua.alua_type != SPC3_ALUA_EMULATED) {
printk(KERN_WARNING "SPC3_ALUA_EMULATED not enabled for"
- " %s/tpgt_%hu/%s\n", TPG_TFO(tpg)->tpg_get_wwn(tpg),
- TPG_TFO(tpg)->tpg_get_tag(tpg),
+ " %s/tpgt_%hu/%s\n", tpg->se_tpg_tfo->tpg_get_wwn(tpg),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg),
config_item_name(&lun->lun_group.cg_item));
return -EINVAL;
}
@@ -1662,8 +1662,8 @@ ssize_t core_alua_store_tg_pt_gp_info(
" %s/tpgt_%hu/%s from ALUA Target Port Group:"
" alua/%s, ID: %hu back to"
" default_tg_pt_gp\n",
- TPG_TFO(tpg)->tpg_get_wwn(tpg),
- TPG_TFO(tpg)->tpg_get_tag(tpg),
+ tpg->se_tpg_tfo->tpg_get_wwn(tpg),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg),
config_item_name(&lun->lun_group.cg_item),
config_item_name(
&tg_pt_gp->tg_pt_gp_group.cg_item),
@@ -1671,7 +1671,7 @@ ssize_t core_alua_store_tg_pt_gp_info(
__core_alua_drop_tg_pt_gp_mem(tg_pt_gp_mem, tg_pt_gp);
__core_alua_attach_tg_pt_gp_mem(tg_pt_gp_mem,
- T10_ALUA(su_dev)->default_tg_pt_gp);
+ su_dev->t10_alua.default_tg_pt_gp);
spin_unlock(&tg_pt_gp_mem->tg_pt_gp_mem_lock);
return count;
@@ -1689,8 +1689,8 @@ ssize_t core_alua_store_tg_pt_gp_info(
spin_unlock(&tg_pt_gp_mem->tg_pt_gp_mem_lock);
printk(KERN_INFO "Target_Core_ConfigFS: %s %s/tpgt_%hu/%s to ALUA"
" Target Port Group: alua/%s, ID: %hu\n", (move) ?
- "Moving" : "Adding", TPG_TFO(tpg)->tpg_get_wwn(tpg),
- TPG_TFO(tpg)->tpg_get_tag(tpg),
+ "Moving" : "Adding", tpg->se_tpg_tfo->tpg_get_wwn(tpg),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg),
config_item_name(&lun->lun_group.cg_item),
config_item_name(&tg_pt_gp_new->tg_pt_gp_group.cg_item),
tg_pt_gp_new->tg_pt_gp_id);
@@ -1947,7 +1947,7 @@ ssize_t core_alua_store_secondary_write_metadata(
int core_setup_alua(struct se_device *dev, int force_pt)
{
struct se_subsystem_dev *su_dev = dev->se_sub_dev;
- struct t10_alua *alua = T10_ALUA(su_dev);
+ struct t10_alua *alua = &su_dev->t10_alua;
struct t10_alua_lu_gp_member *lu_gp_mem;
/*
* If this device is from Target_Core_Mod/pSCSI, use the ALUA logic
@@ -1955,21 +1955,21 @@ int core_setup_alua(struct se_device *dev, int force_pt)
* cause a problem because libata and some SATA RAID HBAs appear
* under Linux/SCSI, but emulate SCSI logic themselves.
*/
- if (((TRANSPORT(dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) &&
- !(DEV_ATTRIB(dev)->emulate_alua)) || force_pt) {
+ if (((dev->transport->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) &&
+ !(dev->se_sub_dev->se_dev_attrib.emulate_alua)) || force_pt) {
alua->alua_type = SPC_ALUA_PASSTHROUGH;
alua->alua_state_check = &core_alua_state_check_nop;
printk(KERN_INFO "%s: Using SPC_ALUA_PASSTHROUGH, no ALUA"
- " emulation\n", TRANSPORT(dev)->name);
+ " emulation\n", dev->transport->name);
return 0;
}
/*
* If SPC-3 or above is reported by real or emulated struct se_device,
* use emulated ALUA.
*/
- if (TRANSPORT(dev)->get_device_rev(dev) >= SCSI_3) {
+ if (dev->transport->get_device_rev(dev) >= SCSI_3) {
printk(KERN_INFO "%s: Enabling ALUA Emulation for SPC-3"
- " device\n", TRANSPORT(dev)->name);
+ " device\n", dev->transport->name);
/*
* Associate this struct se_device with the default ALUA
* LUN Group.
@@ -1987,12 +1987,12 @@ int core_setup_alua(struct se_device *dev, int force_pt)
printk(KERN_INFO "%s: Adding to default ALUA LU Group:"
" core/alua/lu_gps/default_lu_gp\n",
- TRANSPORT(dev)->name);
+ dev->transport->name);
} else {
alua->alua_type = SPC2_ALUA_DISABLED;
alua->alua_state_check = &core_alua_state_check_nop;
printk(KERN_INFO "%s: Disabling ALUA Emulation for SPC-2"
- " device\n", TRANSPORT(dev)->name);
+ " device\n", dev->transport->name);
}
return 0;
diff --git a/drivers/target/target_core_cdb.c b/drivers/target/target_core_cdb.c
index b39ffad..7d9ccf3 100644
--- a/drivers/target/target_core_cdb.c
+++ b/drivers/target/target_core_cdb.c
@@ -64,8 +64,8 @@ target_fill_alua_data(struct se_port *port, unsigned char *buf)
static int
target_emulate_inquiry_std(struct se_cmd *cmd)
{
- struct se_lun *lun = SE_LUN(cmd);
- struct se_device *dev = SE_DEV(cmd);
+ struct se_lun *lun = cmd->se_lun;
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
unsigned char *buf = cmd->t_task->t_task_buf;
/*
@@ -86,7 +86,7 @@ target_emulate_inquiry_std(struct se_cmd *cmd)
/*
* Enable SCCS and TPGS fields for Emulated ALUA
*/
- if (T10_ALUA(dev->se_sub_dev)->alua_type == SPC3_ALUA_EMULATED)
+ if (dev->se_sub_dev->t10_alua.alua_type == SPC3_ALUA_EMULATED)
target_fill_alua_data(lun->lun_sep, buf);
if (cmd->data_length < 8) {
@@ -107,9 +107,9 @@ target_emulate_inquiry_std(struct se_cmd *cmd)
snprintf((unsigned char *)&buf[8], 8, "LIO-ORG");
snprintf((unsigned char *)&buf[16], 16, "%s",
- &DEV_T10_WWN(dev)->model[0]);
+ &dev->se_sub_dev->t10_wwn.model[0]);
snprintf((unsigned char *)&buf[32], 4, "%s",
- &DEV_T10_WWN(dev)->revision[0]);
+ &dev->se_sub_dev->t10_wwn.revision[0]);
buf[4] = 31; /* Set additional length to 31 */
return 0;
}
@@ -128,7 +128,7 @@ target_emulate_evpd_00(struct se_cmd *cmd, unsigned char *buf)
* Registered Extended LUN WWN has been set via ConfigFS
* during device creation/restart.
*/
- if (SE_DEV(cmd)->se_sub_dev->su_dev_flags &
+ if (cmd->se_lun->lun_se_dev->se_sub_dev->su_dev_flags &
SDF_EMULATED_VPD_UNIT_SERIAL) {
buf[3] = 3;
buf[5] = 0x80;
@@ -143,7 +143,7 @@ target_emulate_evpd_00(struct se_cmd *cmd, unsigned char *buf)
static int
target_emulate_evpd_80(struct se_cmd *cmd, unsigned char *buf)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
u16 len = 0;
buf[1] = 0x80;
@@ -152,7 +152,7 @@ target_emulate_evpd_80(struct se_cmd *cmd, unsigned char *buf)
u32 unit_serial_len;
unit_serial_len =
- strlen(&DEV_T10_WWN(dev)->unit_serial[0]);
+ strlen(&dev->se_sub_dev->t10_wwn.unit_serial[0]);
unit_serial_len++; /* For NULL Terminator */
if (((len + 4) + unit_serial_len) > cmd->data_length) {
@@ -162,7 +162,7 @@ target_emulate_evpd_80(struct se_cmd *cmd, unsigned char *buf)
return 0;
}
len += sprintf((unsigned char *)&buf[4], "%s",
- &DEV_T10_WWN(dev)->unit_serial[0]);
+ &dev->se_sub_dev->t10_wwn.unit_serial[0]);
len++; /* Extra Byte for NULL Terminator */
buf[3] = len;
}
@@ -176,15 +176,15 @@ target_emulate_evpd_80(struct se_cmd *cmd, unsigned char *buf)
static int
target_emulate_evpd_83(struct se_cmd *cmd, unsigned char *buf)
{
- struct se_device *dev = SE_DEV(cmd);
- struct se_lun *lun = SE_LUN(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
+ struct se_lun *lun = cmd->se_lun;
struct se_port *port = NULL;
struct se_portal_group *tpg = NULL;
struct t10_alua_lu_gp_member *lu_gp_mem;
struct t10_alua_tg_pt_gp *tg_pt_gp;
struct t10_alua_tg_pt_gp_member *tg_pt_gp_mem;
unsigned char binary, binary_new;
- unsigned char *prod = &DEV_T10_WWN(dev)->model[0];
+ unsigned char *prod = &dev->se_sub_dev->t10_wwn.model[0];
u32 prod_len;
u32 unit_serial_len, off = 0;
int i;
@@ -238,11 +238,11 @@ target_emulate_evpd_83(struct se_cmd *cmd, unsigned char *buf)
* VENDOR_SPECIFIC_IDENTIFIER_EXTENTION
*/
binary = transport_asciihex_to_binaryhex(
- &DEV_T10_WWN(dev)->unit_serial[0]);
+ &dev->se_sub_dev->t10_wwn.unit_serial[0]);
buf[off++] |= (binary & 0xf0) >> 4;
for (i = 0; i < 24; i += 2) {
binary_new = transport_asciihex_to_binaryhex(
- &DEV_T10_WWN(dev)->unit_serial[i+2]);
+ &dev->se_sub_dev->t10_wwn.unit_serial[i+2]);
buf[off] = (binary & 0x0f) << 4;
buf[off++] |= (binary_new & 0xf0) >> 4;
binary = binary_new;
@@ -263,7 +263,7 @@ check_t10_vend_desc:
if (dev->se_sub_dev->su_dev_flags &
SDF_EMULATED_VPD_UNIT_SERIAL) {
unit_serial_len =
- strlen(&DEV_T10_WWN(dev)->unit_serial[0]);
+ strlen(&dev->se_sub_dev->t10_wwn.unit_serial[0]);
unit_serial_len++; /* For NULL Terminator */
if ((len + (id_len + 4) +
@@ -274,7 +274,7 @@ check_t10_vend_desc:
}
id_len += sprintf((unsigned char *)&buf[off+12],
"%s:%s", prod,
- &DEV_T10_WWN(dev)->unit_serial[0]);
+ &dev->se_sub_dev->t10_wwn.unit_serial[0]);
}
buf[off] = 0x2; /* ASCII */
buf[off+1] = 0x1; /* T10 Vendor ID */
@@ -312,7 +312,7 @@ check_port:
goto check_tpgi;
}
buf[off] =
- (TPG_TFO(tpg)->get_fabric_proto_ident(tpg) << 4);
+ (tpg->se_tpg_tfo->get_fabric_proto_ident(tpg) << 4);
buf[off++] |= 0x1; /* CODE SET == Binary */
buf[off] = 0x80; /* Set PIV=1 */
/* Set ASSOICATION == target port: 01b */
@@ -335,7 +335,7 @@ check_port:
* section 7.5.1 Table 362
*/
check_tpgi:
- if (T10_ALUA(dev->se_sub_dev)->alua_type !=
+ if (dev->se_sub_dev->t10_alua.alua_type !=
SPC3_ALUA_EMULATED)
goto check_scsi_name;
@@ -357,7 +357,7 @@ check_tpgi:
spin_unlock(&tg_pt_gp_mem->tg_pt_gp_mem_lock);
buf[off] =
- (TPG_TFO(tpg)->get_fabric_proto_ident(tpg) << 4);
+ (tpg->se_tpg_tfo->get_fabric_proto_ident(tpg) << 4);
buf[off++] |= 0x1; /* CODE SET == Binary */
buf[off] = 0x80; /* Set PIV=1 */
/* Set ASSOICATION == target port: 01b */
@@ -409,7 +409,7 @@ check_lu_gp:
* section 7.5.1 Table 362
*/
check_scsi_name:
- scsi_name_len = strlen(TPG_TFO(tpg)->tpg_get_wwn(tpg));
+ scsi_name_len = strlen(tpg->se_tpg_tfo->tpg_get_wwn(tpg));
/* UTF-8 ",t,0x<16-bit TPGT>" + NULL Terminator */
scsi_name_len += 10;
/* Check for 4-byte padding */
@@ -424,7 +424,7 @@ check_scsi_name:
goto set_len;
}
buf[off] =
- (TPG_TFO(tpg)->get_fabric_proto_ident(tpg) << 4);
+ (tpg->se_tpg_tfo->get_fabric_proto_ident(tpg) << 4);
buf[off++] |= 0x3; /* CODE SET == UTF-8 */
buf[off] = 0x80; /* Set PIV=1 */
/* Set ASSOICATION == target port: 01b */
@@ -438,9 +438,9 @@ check_scsi_name:
* Target Port, this means "<iSCSI name>,t,0x<TPGT> in
* UTF-8 encoding.
*/
- tpgt = TPG_TFO(tpg)->tpg_get_tag(tpg);
+ tpgt = tpg->se_tpg_tfo->tpg_get_tag(tpg);
scsi_name_len = sprintf(&buf[off], "%s,t,0x%04x",
- TPG_TFO(tpg)->tpg_get_wwn(tpg), tpgt);
+ tpg->se_tpg_tfo->tpg_get_wwn(tpg), tpgt);
scsi_name_len += 1 /* Include NULL terminator */;
/*
* The null-terminated, null-padded (see 4.4.2) SCSI
@@ -477,7 +477,7 @@ target_emulate_evpd_86(struct se_cmd *cmd, unsigned char *buf)
buf[5] = 0x07;
/* If WriteCache emulation is enabled, set V_SUP */
- if (DEV_ATTRIB(SE_DEV(cmd))->emulate_write_cache > 0)
+ if (cmd->se_lun->lun_se_dev->se_sub_dev->se_dev_attrib.emulate_write_cache > 0)
buf[6] = 0x01;
return 0;
}
@@ -486,7 +486,7 @@ target_emulate_evpd_86(struct se_cmd *cmd, unsigned char *buf)
static int
target_emulate_evpd_b0(struct se_cmd *cmd, unsigned char *buf)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
int have_tp = 0;
/*
@@ -494,7 +494,7 @@ target_emulate_evpd_b0(struct se_cmd *cmd, unsigned char *buf)
* emulate_tpu=1 or emulate_tpws=1 we will be expect a
* different page length for Thin Provisioning.
*/
- if (DEV_ATTRIB(dev)->emulate_tpu || DEV_ATTRIB(dev)->emulate_tpws)
+ if (dev->se_sub_dev->se_dev_attrib.emulate_tpu || dev->se_sub_dev->se_dev_attrib.emulate_tpws)
have_tp = 1;
if (cmd->data_length < (0x10 + 4)) {
@@ -523,12 +523,12 @@ target_emulate_evpd_b0(struct se_cmd *cmd, unsigned char *buf)
/*
* Set MAXIMUM TRANSFER LENGTH
*/
- put_unaligned_be32(DEV_ATTRIB(dev)->max_sectors, &buf[8]);
+ put_unaligned_be32(dev->se_sub_dev->se_dev_attrib.max_sectors, &buf[8]);
/*
* Set OPTIMAL TRANSFER LENGTH
*/
- put_unaligned_be32(DEV_ATTRIB(dev)->optimal_sectors, &buf[12]);
+ put_unaligned_be32(dev->se_sub_dev->se_dev_attrib.optimal_sectors, &buf[12]);
/*
* Exit now if we don't support TP or the initiator sent a too
@@ -540,25 +540,25 @@ target_emulate_evpd_b0(struct se_cmd *cmd, unsigned char *buf)
/*
* Set MAXIMUM UNMAP LBA COUNT
*/
- put_unaligned_be32(DEV_ATTRIB(dev)->max_unmap_lba_count, &buf[20]);
+ put_unaligned_be32(dev->se_sub_dev->se_dev_attrib.max_unmap_lba_count, &buf[20]);
/*
* Set MAXIMUM UNMAP BLOCK DESCRIPTOR COUNT
*/
- put_unaligned_be32(DEV_ATTRIB(dev)->max_unmap_block_desc_count,
+ put_unaligned_be32(dev->se_sub_dev->se_dev_attrib.max_unmap_block_desc_count,
&buf[24]);
/*
* Set OPTIMAL UNMAP GRANULARITY
*/
- put_unaligned_be32(DEV_ATTRIB(dev)->unmap_granularity, &buf[28]);
+ put_unaligned_be32(dev->se_sub_dev->se_dev_attrib.unmap_granularity, &buf[28]);
/*
* UNMAP GRANULARITY ALIGNMENT
*/
- put_unaligned_be32(DEV_ATTRIB(dev)->unmap_granularity_alignment,
+ put_unaligned_be32(dev->se_sub_dev->se_dev_attrib.unmap_granularity_alignment,
&buf[32]);
- if (DEV_ATTRIB(dev)->unmap_granularity_alignment != 0)
+ if (dev->se_sub_dev->se_dev_attrib.unmap_granularity_alignment != 0)
buf[32] |= 0x80; /* Set the UGAVALID bit */
return 0;
@@ -568,7 +568,7 @@ target_emulate_evpd_b0(struct se_cmd *cmd, unsigned char *buf)
static int
target_emulate_evpd_b2(struct se_cmd *cmd, unsigned char *buf)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
/*
* From sbc3r22 section 6.5.4 Thin Provisioning VPD page:
@@ -602,7 +602,7 @@ target_emulate_evpd_b2(struct se_cmd *cmd, unsigned char *buf)
* the UNMAP command (see 5.25). A TPU bit set to zero indicates
* that the device server does not support the UNMAP command.
*/
- if (DEV_ATTRIB(dev)->emulate_tpu != 0)
+ if (dev->se_sub_dev->se_dev_attrib.emulate_tpu != 0)
buf[5] = 0x80;
/*
@@ -611,7 +611,7 @@ target_emulate_evpd_b2(struct se_cmd *cmd, unsigned char *buf)
* A TPWS bit set to zero indicates that the device server does not
* support the use of the WRITE SAME (16) command to unmap LBAs.
*/
- if (DEV_ATTRIB(dev)->emulate_tpws != 0)
+ if (dev->se_sub_dev->se_dev_attrib.emulate_tpws != 0)
buf[5] |= 0x40;
return 0;
@@ -620,7 +620,7 @@ target_emulate_evpd_b2(struct se_cmd *cmd, unsigned char *buf)
static int
target_emulate_inquiry(struct se_cmd *cmd)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
unsigned char *buf = cmd->t_task->t_task_buf;
unsigned char *cdb = cmd->t_task->t_task_cdb;
@@ -665,7 +665,7 @@ target_emulate_inquiry(struct se_cmd *cmd)
static int
target_emulate_readcapacity(struct se_cmd *cmd)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
unsigned char *buf = cmd->t_task->t_task_buf;
unsigned long long blocks_long = dev->transport->get_blocks(dev);
u32 blocks;
@@ -679,14 +679,14 @@ target_emulate_readcapacity(struct se_cmd *cmd)
buf[1] = (blocks >> 16) & 0xff;
buf[2] = (blocks >> 8) & 0xff;
buf[3] = blocks & 0xff;
- buf[4] = (DEV_ATTRIB(dev)->block_size >> 24) & 0xff;
- buf[5] = (DEV_ATTRIB(dev)->block_size >> 16) & 0xff;
- buf[6] = (DEV_ATTRIB(dev)->block_size >> 8) & 0xff;
- buf[7] = DEV_ATTRIB(dev)->block_size & 0xff;
+ buf[4] = (dev->se_sub_dev->se_dev_attrib.block_size >> 24) & 0xff;
+ buf[5] = (dev->se_sub_dev->se_dev_attrib.block_size >> 16) & 0xff;
+ buf[6] = (dev->se_sub_dev->se_dev_attrib.block_size >> 8) & 0xff;
+ buf[7] = dev->se_sub_dev->se_dev_attrib.block_size & 0xff;
/*
* Set max 32-bit blocks to signal SERVICE ACTION READ_CAPACITY_16
*/
- if (DEV_ATTRIB(dev)->emulate_tpu || DEV_ATTRIB(dev)->emulate_tpws)
+ if (dev->se_sub_dev->se_dev_attrib.emulate_tpu || dev->se_sub_dev->se_dev_attrib.emulate_tpws)
put_unaligned_be32(0xFFFFFFFF, &buf[0]);
return 0;
@@ -695,7 +695,7 @@ target_emulate_readcapacity(struct se_cmd *cmd)
static int
target_emulate_readcapacity_16(struct se_cmd *cmd)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
unsigned char *buf = cmd->t_task->t_task_buf;
unsigned long long blocks = dev->transport->get_blocks(dev);
@@ -707,15 +707,15 @@ target_emulate_readcapacity_16(struct se_cmd *cmd)
buf[5] = (blocks >> 16) & 0xff;
buf[6] = (blocks >> 8) & 0xff;
buf[7] = blocks & 0xff;
- buf[8] = (DEV_ATTRIB(dev)->block_size >> 24) & 0xff;
- buf[9] = (DEV_ATTRIB(dev)->block_size >> 16) & 0xff;
- buf[10] = (DEV_ATTRIB(dev)->block_size >> 8) & 0xff;
- buf[11] = DEV_ATTRIB(dev)->block_size & 0xff;
+ buf[8] = (dev->se_sub_dev->se_dev_attrib.block_size >> 24) & 0xff;
+ buf[9] = (dev->se_sub_dev->se_dev_attrib.block_size >> 16) & 0xff;
+ buf[10] = (dev->se_sub_dev->se_dev_attrib.block_size >> 8) & 0xff;
+ buf[11] = dev->se_sub_dev->se_dev_attrib.block_size & 0xff;
/*
* Set Thin Provisioning Enable bit following sbc3r22 in section
* READ CAPACITY (16) byte 14 if emulate_tpu or emulate_tpws is enabled.
*/
- if (DEV_ATTRIB(dev)->emulate_tpu || DEV_ATTRIB(dev)->emulate_tpws)
+ if (dev->se_sub_dev->se_dev_attrib.emulate_tpu || dev->se_sub_dev->se_dev_attrib.emulate_tpws)
buf[14] = 0x80;
return 0;
@@ -765,8 +765,8 @@ target_modesense_control(struct se_device *dev, unsigned char *p)
* for a BUSY, TASK SET FULL, or RESERVATION CONFLICT status regardless
* to the number of commands completed with one of those status codes.
*/
- p[4] = (DEV_ATTRIB(dev)->emulate_ua_intlck_ctrl == 2) ? 0x30 :
- (DEV_ATTRIB(dev)->emulate_ua_intlck_ctrl == 1) ? 0x20 : 0x00;
+ p[4] = (dev->se_sub_dev->se_dev_attrib.emulate_ua_intlck_ctrl == 2) ? 0x30 :
+ (dev->se_sub_dev->se_dev_attrib.emulate_ua_intlck_ctrl == 1) ? 0x20 : 0x00;
/*
* From spc4r17, section 7.4.6 Control mode Page
*
@@ -779,7 +779,7 @@ target_modesense_control(struct se_device *dev, unsigned char *p)
* which the command was received shall be completed with TASK ABORTED
* status (see SAM-4).
*/
- p[5] = (DEV_ATTRIB(dev)->emulate_tas) ? 0x40 : 0x00;
+ p[5] = (dev->se_sub_dev->se_dev_attrib.emulate_tas) ? 0x40 : 0x00;
p[8] = 0xff;
p[9] = 0xff;
p[11] = 30;
@@ -792,7 +792,7 @@ target_modesense_caching(struct se_device *dev, unsigned char *p)
{
p[0] = 0x08;
p[1] = 0x12;
- if (DEV_ATTRIB(dev)->emulate_write_cache > 0)
+ if (dev->se_sub_dev->se_dev_attrib.emulate_write_cache > 0)
p[2] = 0x04; /* Write Cache Enable */
p[12] = 0x20; /* Disabled Read Ahead */
@@ -830,7 +830,7 @@ target_modesense_dpofua(unsigned char *buf, int type)
static int
target_emulate_modesense(struct se_cmd *cmd, int ten)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
char *cdb = cmd->t_task->t_task_cdb;
unsigned char *rbuf = cmd->t_task->t_task_buf;
int type = dev->transport->get_device_type(dev);
@@ -867,13 +867,13 @@ target_emulate_modesense(struct se_cmd *cmd, int ten)
buf[0] = (offset >> 8) & 0xff;
buf[1] = offset & 0xff;
- if ((SE_LUN(cmd)->lun_access & TRANSPORT_LUNFLAGS_READ_ONLY) ||
+ if ((cmd->se_lun->lun_access & TRANSPORT_LUNFLAGS_READ_ONLY) ||
(cmd->se_deve &&
(cmd->se_deve->lun_flags & TRANSPORT_LUNFLAGS_READ_ONLY)))
target_modesense_write_protect(&buf[3], type);
- if ((DEV_ATTRIB(dev)->emulate_write_cache > 0) &&
- (DEV_ATTRIB(dev)->emulate_fua_write > 0))
+ if ((dev->se_sub_dev->se_dev_attrib.emulate_write_cache > 0) &&
+ (dev->se_sub_dev->se_dev_attrib.emulate_fua_write > 0))
target_modesense_dpofua(&buf[3], type);
if ((offset + 2) > cmd->data_length)
@@ -883,13 +883,13 @@ target_emulate_modesense(struct se_cmd *cmd, int ten)
offset -= 1;
buf[0] = offset & 0xff;
- if ((SE_LUN(cmd)->lun_access & TRANSPORT_LUNFLAGS_READ_ONLY) ||
+ if ((cmd->se_lun->lun_access & TRANSPORT_LUNFLAGS_READ_ONLY) ||
(cmd->se_deve &&
(cmd->se_deve->lun_flags & TRANSPORT_LUNFLAGS_READ_ONLY)))
target_modesense_write_protect(&buf[2], type);
- if ((DEV_ATTRIB(dev)->emulate_write_cache > 0) &&
- (DEV_ATTRIB(dev)->emulate_fua_write > 0))
+ if ((dev->se_sub_dev->se_dev_attrib.emulate_write_cache > 0) &&
+ (dev->se_sub_dev->se_dev_attrib.emulate_fua_write > 0))
target_modesense_dpofua(&buf[2], type);
if ((offset + 1) > cmd->data_length)
@@ -963,8 +963,8 @@ target_emulate_request_sense(struct se_cmd *cmd)
static int
target_emulate_unmap(struct se_task *task)
{
- struct se_cmd *cmd = TASK_CMD(task);
- struct se_device *dev = SE_DEV(cmd);
+ struct se_cmd *cmd = task->task_se_cmd;
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
unsigned char *buf = cmd->t_task->t_task_buf, *ptr = NULL;
unsigned char *cdb = &cmd->t_task->t_task_cdb[0];
sector_t lba;
@@ -1010,13 +1010,13 @@ target_emulate_unmap(struct se_task *task)
static int
target_emulate_write_same(struct se_task *task)
{
- struct se_cmd *cmd = TASK_CMD(task);
- struct se_device *dev = SE_DEV(cmd);
+ struct se_cmd *cmd = task->task_se_cmd;
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
sector_t lba = cmd->t_task->t_task_lba;
unsigned int range;
int ret;
- range = (cmd->data_length / DEV_ATTRIB(dev)->block_size);
+ range = (cmd->data_length / dev->se_sub_dev->se_dev_attrib.block_size);
printk(KERN_INFO "WRITE_SAME UNMAP: LBA: %llu Range: %u\n",
(unsigned long long)lba, range);
@@ -1035,8 +1035,8 @@ target_emulate_write_same(struct se_task *task)
int
transport_emulate_control_cdb(struct se_task *task)
{
- struct se_cmd *cmd = TASK_CMD(task);
- struct se_device *dev = SE_DEV(cmd);
+ struct se_cmd *cmd = task->task_se_cmd;
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
unsigned short service_action;
int ret = 0;
diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c
index 8196b72..64418ef 100644
--- a/drivers/target/target_core_configfs.c
+++ b/drivers/target/target_core_configfs.c
@@ -625,7 +625,8 @@ static ssize_t target_core_dev_show_attr_##_name( \
spin_unlock(&se_dev->se_dev_lock); \
return -ENODEV; \
} \
- rb = snprintf(page, PAGE_SIZE, "%u\n", (u32)DEV_ATTRIB(dev)->_name); \
+ rb = snprintf(page, PAGE_SIZE, "%u\n", \
+ (u32)dev->se_sub_dev->se_dev_attrib._name); \
spin_unlock(&se_dev->se_dev_lock); \
\
return rb; \
@@ -1087,7 +1088,7 @@ static ssize_t target_core_dev_pr_show_spc3_res(
PR_REG_ISID_ID_LEN);
*len += sprintf(page + *len, "SPC-3 Reservation: %s Initiator: %s%s\n",
- TPG_TFO(se_nacl->se_tpg)->get_fabric_name(),
+ se_nacl->se_tpg->se_tpg_tfo->get_fabric_name(),
se_nacl->initiatorname, (prf_isid) ? &i_buf[0] : "");
spin_unlock(&dev->dev_reservation_lock);
@@ -1109,7 +1110,7 @@ static ssize_t target_core_dev_pr_show_spc2_res(
return *len;
}
*len += sprintf(page + *len, "SPC-2 Reservation: %s Initiator: %s\n",
- TPG_TFO(se_nacl->se_tpg)->get_fabric_name(),
+ se_nacl->se_tpg->se_tpg_tfo->get_fabric_name(),
se_nacl->initiatorname);
spin_unlock(&dev->dev_reservation_lock);
@@ -1125,7 +1126,7 @@ static ssize_t target_core_dev_pr_show_attr_res_holder(
if (!(su_dev->se_dev_ptr))
return -ENODEV;
- switch (T10_RES(su_dev)->res_type) {
+ switch (su_dev->t10_pr.res_type) {
case SPC3_PERSISTENT_RESERVATIONS:
target_core_dev_pr_show_spc3_res(su_dev->se_dev_ptr,
page, &len);
@@ -1162,7 +1163,7 @@ static ssize_t target_core_dev_pr_show_attr_res_pr_all_tgt_pts(
if (!(dev))
return -ENODEV;
- if (T10_RES(su_dev)->res_type != SPC3_PERSISTENT_RESERVATIONS)
+ if (su_dev->t10_pr.res_type != SPC3_PERSISTENT_RESERVATIONS)
return len;
spin_lock(&dev->dev_reservation_lock);
@@ -1199,10 +1200,10 @@ static ssize_t target_core_dev_pr_show_attr_res_pr_generation(
if (!(su_dev->se_dev_ptr))
return -ENODEV;
- if (T10_RES(su_dev)->res_type != SPC3_PERSISTENT_RESERVATIONS)
+ if (su_dev->t10_pr.res_type != SPC3_PERSISTENT_RESERVATIONS)
return 0;
- return sprintf(page, "0x%08x\n", T10_RES(su_dev)->pr_generation);
+ return sprintf(page, "0x%08x\n", su_dev->t10_pr.pr_generation);
}
SE_DEV_PR_ATTR_RO(res_pr_generation);
@@ -1226,7 +1227,7 @@ static ssize_t target_core_dev_pr_show_attr_res_pr_holder_tg_port(
if (!(dev))
return -ENODEV;
- if (T10_RES(su_dev)->res_type != SPC3_PERSISTENT_RESERVATIONS)
+ if (su_dev->t10_pr.res_type != SPC3_PERSISTENT_RESERVATIONS)
return len;
spin_lock(&dev->dev_reservation_lock);
@@ -1239,7 +1240,7 @@ static ssize_t target_core_dev_pr_show_attr_res_pr_holder_tg_port(
se_nacl = pr_reg->pr_reg_nacl;
se_tpg = se_nacl->se_tpg;
lun = pr_reg->pr_reg_tg_pt_lun;
- tfo = TPG_TFO(se_tpg);
+ tfo = se_tpg->se_tpg_tfo;
len += sprintf(page+len, "SPC-3 Reservation: %s"
" Target Node Endpoint: %s\n", tfo->get_fabric_name(),
@@ -1273,13 +1274,13 @@ static ssize_t target_core_dev_pr_show_attr_res_pr_registered_i_pts(
if (!(su_dev->se_dev_ptr))
return -ENODEV;
- if (T10_RES(su_dev)->res_type != SPC3_PERSISTENT_RESERVATIONS)
+ if (su_dev->t10_pr.res_type != SPC3_PERSISTENT_RESERVATIONS)
return len;
len += sprintf(page+len, "SPC-3 PR Registrations:\n");
- spin_lock(&T10_RES(su_dev)->registration_lock);
- list_for_each_entry(pr_reg, &T10_RES(su_dev)->registration_list,
+ spin_lock(&su_dev->t10_pr.registration_lock);
+ list_for_each_entry(pr_reg, &su_dev->t10_pr.registration_list,
pr_reg_list) {
memset(buf, 0, 384);
@@ -1299,7 +1300,7 @@ static ssize_t target_core_dev_pr_show_attr_res_pr_registered_i_pts(
len += sprintf(page+len, "%s", buf);
reg_count++;
}
- spin_unlock(&T10_RES(su_dev)->registration_lock);
+ spin_unlock(&su_dev->t10_pr.registration_lock);
if (!(reg_count))
len += sprintf(page+len, "None\n");
@@ -1324,7 +1325,7 @@ static ssize_t target_core_dev_pr_show_attr_res_pr_type(
if (!(dev))
return -ENODEV;
- if (T10_RES(su_dev)->res_type != SPC3_PERSISTENT_RESERVATIONS)
+ if (su_dev->t10_pr.res_type != SPC3_PERSISTENT_RESERVATIONS)
return len;
spin_lock(&dev->dev_reservation_lock);
@@ -1355,7 +1356,7 @@ static ssize_t target_core_dev_pr_show_attr_res_type(
if (!(su_dev->se_dev_ptr))
return -ENODEV;
- switch (T10_RES(su_dev)->res_type) {
+ switch (su_dev->t10_pr.res_type) {
case SPC3_PERSISTENT_RESERVATIONS:
len = sprintf(page, "SPC3_PERSISTENT_RESERVATIONS\n");
break;
@@ -1386,11 +1387,11 @@ static ssize_t target_core_dev_pr_show_attr_res_aptpl_active(
if (!(su_dev->se_dev_ptr))
return -ENODEV;
- if (T10_RES(su_dev)->res_type != SPC3_PERSISTENT_RESERVATIONS)
+ if (su_dev->t10_pr.res_type != SPC3_PERSISTENT_RESERVATIONS)
return 0;
return sprintf(page, "APTPL Bit Status: %s\n",
- (T10_RES(su_dev)->pr_aptpl_active) ? "Activated" : "Disabled");
+ (su_dev->t10_pr.pr_aptpl_active) ? "Activated" : "Disabled");
}
SE_DEV_PR_ATTR_RO(res_aptpl_active);
@@ -1405,7 +1406,7 @@ static ssize_t target_core_dev_pr_show_attr_res_aptpl_metadata(
if (!(su_dev->se_dev_ptr))
return -ENODEV;
- if (T10_RES(su_dev)->res_type != SPC3_PERSISTENT_RESERVATIONS)
+ if (su_dev->t10_pr.res_type != SPC3_PERSISTENT_RESERVATIONS)
return 0;
return sprintf(page, "Ready to process PR APTPL metadata..\n");
@@ -1457,7 +1458,7 @@ static ssize_t target_core_dev_pr_store_attr_res_aptpl_metadata(
if (!(dev))
return -ENODEV;
- if (T10_RES(su_dev)->res_type != SPC3_PERSISTENT_RESERVATIONS)
+ if (su_dev->t10_pr.res_type != SPC3_PERSISTENT_RESERVATIONS)
return 0;
if (atomic_read(&dev->dev_export_obj.obj_access_count)) {
@@ -1603,7 +1604,7 @@ static ssize_t target_core_dev_pr_store_attr_res_aptpl_metadata(
goto out;
}
- ret = core_scsi3_alloc_aptpl_registration(T10_RES(su_dev), sa_res_key,
+ ret = core_scsi3_alloc_aptpl_registration(&su_dev->t10_pr, sa_res_key,
i_port, isid, mapped_lun, t_port, tpgt, target_lun,
res_holder, all_tg_pt, type);
out:
@@ -1851,7 +1852,7 @@ static ssize_t target_core_show_alua_lu_gp(void *p, char *page)
if (!(dev))
return -ENODEV;
- if (T10_ALUA(su_dev)->alua_type != SPC3_ALUA_EMULATED)
+ if (su_dev->t10_alua.alua_type != SPC3_ALUA_EMULATED)
return len;
lu_gp_mem = dev->dev_alua_lu_gp_mem;
@@ -1890,7 +1891,7 @@ static ssize_t target_core_store_alua_lu_gp(
if (!(dev))
return -ENODEV;
- if (T10_ALUA(su_dev)->alua_type != SPC3_ALUA_EMULATED) {
+ if (su_dev->t10_alua.alua_type != SPC3_ALUA_EMULATED) {
printk(KERN_WARNING "SPC3_ALUA_EMULATED not enabled for %s/%s\n",
config_item_name(&hba->hba_group.cg_item),
config_item_name(&su_dev->se_dev_group.cg_item));
@@ -2566,9 +2567,9 @@ static ssize_t target_core_alua_tg_pt_gp_show_attr_members(
lun = port->sep_lun;
cur_len = snprintf(buf, TG_PT_GROUP_NAME_BUF, "%s/%s/tpgt_%hu"
- "/%s\n", TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_wwn(tpg),
- TPG_TFO(tpg)->tpg_get_tag(tpg),
+ "/%s\n", tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_wwn(tpg),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg),
config_item_name(&lun->lun_group.cg_item));
cur_len++; /* Extra byte for NULL terminator */
@@ -2760,14 +2761,14 @@ static struct config_group *target_core_make_subdev(
INIT_LIST_HEAD(&se_dev->se_dev_node);
INIT_LIST_HEAD(&se_dev->t10_wwn.t10_vpd_list);
spin_lock_init(&se_dev->t10_wwn.t10_vpd_lock);
- INIT_LIST_HEAD(&se_dev->t10_reservation.registration_list);
- INIT_LIST_HEAD(&se_dev->t10_reservation.aptpl_reg_list);
- spin_lock_init(&se_dev->t10_reservation.registration_lock);
- spin_lock_init(&se_dev->t10_reservation.aptpl_reg_lock);
+ INIT_LIST_HEAD(&se_dev->t10_pr.registration_list);
+ INIT_LIST_HEAD(&se_dev->t10_pr.aptpl_reg_list);
+ spin_lock_init(&se_dev->t10_pr.registration_lock);
+ spin_lock_init(&se_dev->t10_pr.aptpl_reg_lock);
INIT_LIST_HEAD(&se_dev->t10_alua.tg_pt_gps_list);
spin_lock_init(&se_dev->t10_alua.tg_pt_gps_lock);
spin_lock_init(&se_dev->se_dev_lock);
- se_dev->t10_reservation.pr_aptpl_buf_len = PR_APTPL_BUF_LEN;
+ se_dev->t10_pr.pr_aptpl_buf_len = PR_APTPL_BUF_LEN;
se_dev->t10_wwn.t10_sub_dev = se_dev;
se_dev->t10_alua.t10_sub_dev = se_dev;
se_dev->se_dev_attrib.da_sub_dev = se_dev;
@@ -2823,7 +2824,7 @@ static struct config_group *target_core_make_subdev(
if (!(tg_pt_gp))
goto out;
- tg_pt_gp_cg = &T10_ALUA(se_dev)->alua_tg_pt_gps_group;
+ tg_pt_gp_cg = &se_dev->t10_alua.alua_tg_pt_gps_group;
tg_pt_gp_cg->default_groups = kzalloc(sizeof(struct config_group) * 2,
GFP_KERNEL);
if (!(tg_pt_gp_cg->default_groups)) {
@@ -2836,11 +2837,11 @@ static struct config_group *target_core_make_subdev(
"default_tg_pt_gp", &target_core_alua_tg_pt_gp_cit);
tg_pt_gp_cg->default_groups[0] = &tg_pt_gp->tg_pt_gp_group;
tg_pt_gp_cg->default_groups[1] = NULL;
- T10_ALUA(se_dev)->default_tg_pt_gp = tg_pt_gp;
+ se_dev->t10_alua.default_tg_pt_gp = tg_pt_gp;
/*
* Add core/$HBA/$DEV/statistics/ default groups
*/
- dev_stat_grp = &DEV_STAT_GRP(se_dev)->stat_group;
+ dev_stat_grp = &se_dev->dev_stat_grps.stat_group;
dev_stat_grp->default_groups = kzalloc(sizeof(struct config_group) * 4,
GFP_KERNEL);
if (!dev_stat_grp->default_groups) {
@@ -2855,9 +2856,9 @@ static struct config_group *target_core_make_subdev(
mutex_unlock(&hba->hba_access_mutex);
return &se_dev->se_dev_group;
out:
- if (T10_ALUA(se_dev)->default_tg_pt_gp) {
- core_alua_free_tg_pt_gp(T10_ALUA(se_dev)->default_tg_pt_gp);
- T10_ALUA(se_dev)->default_tg_pt_gp = NULL;
+ if (se_dev->t10_alua.default_tg_pt_gp) {
+ core_alua_free_tg_pt_gp(se_dev->t10_alua.default_tg_pt_gp);
+ se_dev->t10_alua.default_tg_pt_gp = NULL;
}
if (dev_stat_grp)
kfree(dev_stat_grp->default_groups);
@@ -2894,7 +2895,7 @@ static void target_core_drop_subdev(
list_del(&se_dev->se_dev_node);
spin_unlock(&se_device_lock);
- dev_stat_grp = &DEV_STAT_GRP(se_dev)->stat_group;
+ dev_stat_grp = &se_dev->dev_stat_grps.stat_group;
for (i = 0; dev_stat_grp->default_groups[i]; i++) {
df_item = &dev_stat_grp->default_groups[i]->cg_item;
dev_stat_grp->default_groups[i] = NULL;
@@ -2902,7 +2903,7 @@ static void target_core_drop_subdev(
}
kfree(dev_stat_grp->default_groups);
- tg_pt_gp_cg = &T10_ALUA(se_dev)->alua_tg_pt_gps_group;
+ tg_pt_gp_cg = &se_dev->t10_alua.alua_tg_pt_gps_group;
for (i = 0; tg_pt_gp_cg->default_groups[i]; i++) {
df_item = &tg_pt_gp_cg->default_groups[i]->cg_item;
tg_pt_gp_cg->default_groups[i] = NULL;
@@ -2913,7 +2914,7 @@ static void target_core_drop_subdev(
* core_alua_free_tg_pt_gp() is called from ->default_tg_pt_gp
* directly from target_core_alua_tg_pt_gp_release().
*/
- T10_ALUA(se_dev)->default_tg_pt_gp = NULL;
+ se_dev->t10_alua.default_tg_pt_gp = NULL;
dev_cg = &se_dev->se_dev_group;
for (i = 0; dev_cg->default_groups[i]; i++) {
diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c
index ff57891..fd92385 100644
--- a/drivers/target/target_core_device.c
+++ b/drivers/target/target_core_device.c
@@ -65,7 +65,7 @@ int transport_get_lun_for_cmd(
{
struct se_dev_entry *deve;
struct se_lun *se_lun = NULL;
- struct se_session *se_sess = SE_SESS(se_cmd);
+ struct se_session *se_sess = se_cmd->se_sess;
unsigned long flags;
int read_only = 0;
@@ -75,9 +75,9 @@ int transport_get_lun_for_cmd(
return -ENODEV;
}
- spin_lock_irq(&SE_NODE_ACL(se_sess)->device_list_lock);
+ spin_lock_irq(&se_sess->se_node_acl->device_list_lock);
deve = se_cmd->se_deve =
- &SE_NODE_ACL(se_sess)->device_list[unpacked_lun];
+ &se_sess->se_node_acl->device_list[unpacked_lun];
if (deve->lun_flags & TRANSPORT_LUNFLAGS_INITIATOR_ACCESS) {
if (se_cmd) {
deve->total_cmds++;
@@ -100,11 +100,11 @@ int transport_get_lun_for_cmd(
se_lun = se_cmd->se_lun = deve->se_lun;
se_cmd->pr_res_key = deve->pr_res_key;
se_cmd->orig_fe_lun = unpacked_lun;
- se_cmd->se_orig_obj_ptr = SE_LUN(se_cmd)->lun_se_dev;
+ se_cmd->se_orig_obj_ptr = se_cmd->se_lun->lun_se_dev;
se_cmd->se_cmd_flags |= SCF_SE_LUN_CMD;
}
out:
- spin_unlock_irq(&SE_NODE_ACL(se_sess)->device_list_lock);
+ spin_unlock_irq(&se_sess->se_node_acl->device_list_lock);
if (!se_lun) {
if (read_only) {
@@ -112,7 +112,7 @@ out:
se_cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
printk("TARGET_CORE[%s]: Detected WRITE_PROTECTED LUN"
" Access for 0x%08x\n",
- CMD_TFO(se_cmd)->get_fabric_name(),
+ se_cmd->se_tfo->get_fabric_name(),
unpacked_lun);
return -EACCES;
} else {
@@ -126,7 +126,7 @@ out:
se_cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
printk("TARGET_CORE[%s]: Detected NON_EXISTENT_LUN"
" Access for 0x%08x\n",
- CMD_TFO(se_cmd)->get_fabric_name(),
+ se_cmd->se_tfo->get_fabric_name(),
unpacked_lun);
return -ENODEV;
}
@@ -141,11 +141,11 @@ out:
}
#if 0
printk("TARGET_CORE[%s]: Using virtual LUN0! :-)\n",
- CMD_TFO(se_cmd)->get_fabric_name());
+ se_cmd->se_tfo->get_fabric_name());
#endif
se_lun = se_cmd->se_lun = &se_sess->se_tpg->tpg_virt_lun0;
se_cmd->orig_fe_lun = 0;
- se_cmd->se_orig_obj_ptr = SE_LUN(se_cmd)->lun_se_dev;
+ se_cmd->se_orig_obj_ptr = se_cmd->se_lun->lun_se_dev;
se_cmd->se_cmd_flags |= SCF_SE_LUN_CMD;
}
}
@@ -176,10 +176,10 @@ out:
*/
spin_lock_irqsave(&se_lun->lun_cmd_lock, flags);
list_add_tail(&se_cmd->se_lun_list, &se_lun->lun_cmd_list);
- atomic_set(&T_TASK(se_cmd)->transport_lun_active, 1);
+ atomic_set(&se_cmd->t_task->transport_lun_active, 1);
#if 0
printk(KERN_INFO "Adding ITT: 0x%08x to LUN LIST[%d]\n",
- CMD_TFO(se_cmd)->get_task_tag(se_cmd), se_lun->unpacked_lun);
+ se_cmd->se_tfo->get_task_tag(se_cmd), se_lun->unpacked_lun);
#endif
spin_unlock_irqrestore(&se_lun->lun_cmd_lock, flags);
@@ -194,7 +194,7 @@ int transport_get_lun_for_tmr(
struct se_device *dev = NULL;
struct se_dev_entry *deve;
struct se_lun *se_lun = NULL;
- struct se_session *se_sess = SE_SESS(se_cmd);
+ struct se_session *se_sess = se_cmd->se_sess;
struct se_tmr_req *se_tmr = se_cmd->se_tmr_req;
if (unpacked_lun >= TRANSPORT_MAX_LUNS_PER_TPG) {
@@ -203,23 +203,23 @@ int transport_get_lun_for_tmr(
return -ENODEV;
}
- spin_lock_irq(&SE_NODE_ACL(se_sess)->device_list_lock);
+ spin_lock_irq(&se_sess->se_node_acl->device_list_lock);
deve = se_cmd->se_deve =
- &SE_NODE_ACL(se_sess)->device_list[unpacked_lun];
+ &se_sess->se_node_acl->device_list[unpacked_lun];
if (deve->lun_flags & TRANSPORT_LUNFLAGS_INITIATOR_ACCESS) {
se_lun = se_cmd->se_lun = se_tmr->tmr_lun = deve->se_lun;
dev = se_lun->lun_se_dev;
se_cmd->pr_res_key = deve->pr_res_key;
se_cmd->orig_fe_lun = unpacked_lun;
- se_cmd->se_orig_obj_ptr = SE_LUN(se_cmd)->lun_se_dev;
+ se_cmd->se_orig_obj_ptr = se_cmd->se_lun->lun_se_dev;
/* se_cmd->se_cmd_flags |= SCF_SE_LUN_CMD; */
}
- spin_unlock_irq(&SE_NODE_ACL(se_sess)->device_list_lock);
+ spin_unlock_irq(&se_sess->se_node_acl->device_list_lock);
if (!se_lun) {
printk(KERN_INFO "TARGET_CORE[%s]: Detected NON_EXISTENT_LUN"
" Access for 0x%08x\n",
- CMD_TFO(se_cmd)->get_fabric_name(),
+ se_cmd->se_tfo->get_fabric_name(),
unpacked_lun);
se_cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
return -ENODEV;
@@ -268,14 +268,14 @@ struct se_dev_entry *core_get_se_deve_from_rtpi(
if (!(lun)) {
printk(KERN_ERR "%s device entries device pointer is"
" NULL, but Initiator has access.\n",
- TPG_TFO(tpg)->get_fabric_name());
+ tpg->se_tpg_tfo->get_fabric_name());
continue;
}
port = lun->lun_sep;
if (!(port)) {
printk(KERN_ERR "%s device entries device pointer is"
" NULL, but Initiator has access.\n",
- TPG_TFO(tpg)->get_fabric_name());
+ tpg->se_tpg_tfo->get_fabric_name());
continue;
}
if (port->sep_rtpi != rtpi)
@@ -313,7 +313,7 @@ int core_free_device_list_for_node(
if (!deve->se_lun) {
printk(KERN_ERR "%s device entries device pointer is"
" NULL, but Initiator has access.\n",
- TPG_TFO(tpg)->get_fabric_name());
+ tpg->se_tpg_tfo->get_fabric_name());
continue;
}
lun = deve->se_lun;
@@ -564,7 +564,7 @@ static void core_export_port(
struct se_port *port,
struct se_lun *lun)
{
- struct se_subsystem_dev *su_dev = SU_DEV(dev);
+ struct se_subsystem_dev *su_dev = dev->se_sub_dev;
struct t10_alua_tg_pt_gp_member *tg_pt_gp_mem = NULL;
spin_lock(&dev->se_port_lock);
@@ -577,7 +577,7 @@ static void core_export_port(
list_add_tail(&port->sep_list, &dev->dev_sep_list);
spin_unlock(&dev->se_port_lock);
- if (T10_ALUA(su_dev)->alua_type == SPC3_ALUA_EMULATED) {
+ if (su_dev->t10_alua.alua_type == SPC3_ALUA_EMULATED) {
tg_pt_gp_mem = core_alua_allocate_tg_pt_gp_mem(port);
if (IS_ERR(tg_pt_gp_mem) || !tg_pt_gp_mem) {
printk(KERN_ERR "Unable to allocate t10_alua_tg_pt"
@@ -586,11 +586,11 @@ static void core_export_port(
}
spin_lock(&tg_pt_gp_mem->tg_pt_gp_mem_lock);
__core_alua_attach_tg_pt_gp_mem(tg_pt_gp_mem,
- T10_ALUA(su_dev)->default_tg_pt_gp);
+ su_dev->t10_alua.default_tg_pt_gp);
spin_unlock(&tg_pt_gp_mem->tg_pt_gp_mem_lock);
printk(KERN_INFO "%s/%s: Adding to default ALUA Target Port"
" Group: alua/default_tg_pt_gp\n",
- TRANSPORT(dev)->name, TPG_TFO(tpg)->get_fabric_name());
+ dev->transport->name, tpg->se_tpg_tfo->get_fabric_name());
}
dev->dev_port_count++;
@@ -665,12 +665,12 @@ int transport_core_report_lun_response(struct se_cmd *se_cmd)
{
struct se_dev_entry *deve;
struct se_lun *se_lun;
- struct se_session *se_sess = SE_SESS(se_cmd);
+ struct se_session *se_sess = se_cmd->se_sess;
struct se_task *se_task;
- unsigned char *buf = (unsigned char *)T_TASK(se_cmd)->t_task_buf;
+ unsigned char *buf = se_cmd->t_task->t_task_buf;
u32 cdb_offset = 0, lun_count = 0, offset = 8, i;
- list_for_each_entry(se_task, &T_TASK(se_cmd)->t_task_list, t_list)
+ list_for_each_entry(se_task, &se_cmd->t_task->t_task_list, t_list)
break;
if (!(se_task)) {
@@ -689,9 +689,9 @@ int transport_core_report_lun_response(struct se_cmd *se_cmd)
goto done;
}
- spin_lock_irq(&SE_NODE_ACL(se_sess)->device_list_lock);
+ spin_lock_irq(&se_sess->se_node_acl->device_list_lock);
for (i = 0; i < TRANSPORT_MAX_LUNS_PER_TPG; i++) {
- deve = &SE_NODE_ACL(se_sess)->device_list[i];
+ deve = &se_sess->se_node_acl->device_list[i];
if (!(deve->lun_flags & TRANSPORT_LUNFLAGS_INITIATOR_ACCESS))
continue;
se_lun = deve->se_lun;
@@ -708,7 +708,7 @@ int transport_core_report_lun_response(struct se_cmd *se_cmd)
offset += 8;
cdb_offset += 8;
}
- spin_unlock_irq(&SE_NODE_ACL(se_sess)->device_list_lock);
+ spin_unlock_irq(&se_sess->se_node_acl->device_list_lock);
/*
* See SPC3 r07, page 159.
@@ -759,13 +759,13 @@ void se_release_vpd_for_dev(struct se_device *dev)
{
struct t10_vpd *vpd, *vpd_tmp;
- spin_lock(&DEV_T10_WWN(dev)->t10_vpd_lock);
+ spin_lock(&dev->se_sub_dev->t10_wwn.t10_vpd_lock);
list_for_each_entry_safe(vpd, vpd_tmp,
- &DEV_T10_WWN(dev)->t10_vpd_list, vpd_list) {
+ &dev->se_sub_dev->t10_wwn.t10_vpd_list, vpd_list) {
list_del(&vpd->vpd_list);
kfree(vpd);
}
- spin_unlock(&DEV_T10_WWN(dev)->t10_vpd_lock);
+ spin_unlock(&dev->se_sub_dev->t10_wwn.t10_vpd_lock);
}
/* se_free_virtual_device():
@@ -851,48 +851,48 @@ void se_dev_set_default_attribs(
{
struct queue_limits *limits = &dev_limits->limits;
- DEV_ATTRIB(dev)->emulate_dpo = DA_EMULATE_DPO;
- DEV_ATTRIB(dev)->emulate_fua_write = DA_EMULATE_FUA_WRITE;
- DEV_ATTRIB(dev)->emulate_fua_read = DA_EMULATE_FUA_READ;
- DEV_ATTRIB(dev)->emulate_write_cache = DA_EMULATE_WRITE_CACHE;
- DEV_ATTRIB(dev)->emulate_ua_intlck_ctrl = DA_EMULATE_UA_INTLLCK_CTRL;
- DEV_ATTRIB(dev)->emulate_tas = DA_EMULATE_TAS;
- DEV_ATTRIB(dev)->emulate_tpu = DA_EMULATE_TPU;
- DEV_ATTRIB(dev)->emulate_tpws = DA_EMULATE_TPWS;
- DEV_ATTRIB(dev)->emulate_reservations = DA_EMULATE_RESERVATIONS;
- DEV_ATTRIB(dev)->emulate_alua = DA_EMULATE_ALUA;
- DEV_ATTRIB(dev)->enforce_pr_isids = DA_ENFORCE_PR_ISIDS;
+ dev->se_sub_dev->se_dev_attrib.emulate_dpo = DA_EMULATE_DPO;
+ dev->se_sub_dev->se_dev_attrib.emulate_fua_write = DA_EMULATE_FUA_WRITE;
+ dev->se_sub_dev->se_dev_attrib.emulate_fua_read = DA_EMULATE_FUA_READ;
+ dev->se_sub_dev->se_dev_attrib.emulate_write_cache = DA_EMULATE_WRITE_CACHE;
+ dev->se_sub_dev->se_dev_attrib.emulate_ua_intlck_ctrl = DA_EMULATE_UA_INTLLCK_CTRL;
+ dev->se_sub_dev->se_dev_attrib.emulate_tas = DA_EMULATE_TAS;
+ dev->se_sub_dev->se_dev_attrib.emulate_tpu = DA_EMULATE_TPU;
+ dev->se_sub_dev->se_dev_attrib.emulate_tpws = DA_EMULATE_TPWS;
+ dev->se_sub_dev->se_dev_attrib.emulate_reservations = DA_EMULATE_RESERVATIONS;
+ dev->se_sub_dev->se_dev_attrib.emulate_alua = DA_EMULATE_ALUA;
+ dev->se_sub_dev->se_dev_attrib.enforce_pr_isids = DA_ENFORCE_PR_ISIDS;
/*
* The TPU=1 and TPWS=1 settings will be set in TCM/IBLOCK
* iblock_create_virtdevice() from struct queue_limits values
* if blk_queue_discard()==1
*/
- DEV_ATTRIB(dev)->max_unmap_lba_count = DA_MAX_UNMAP_LBA_COUNT;
- DEV_ATTRIB(dev)->max_unmap_block_desc_count =
- DA_MAX_UNMAP_BLOCK_DESC_COUNT;
- DEV_ATTRIB(dev)->unmap_granularity = DA_UNMAP_GRANULARITY_DEFAULT;
- DEV_ATTRIB(dev)->unmap_granularity_alignment =
+ dev->se_sub_dev->se_dev_attrib.max_unmap_lba_count = DA_MAX_UNMAP_LBA_COUNT;
+ dev->se_sub_dev->se_dev_attrib.max_unmap_block_desc_count =
+ DA_MAX_UNMAP_BLOCK_DESC_COUNT;
+ dev->se_sub_dev->se_dev_attrib.unmap_granularity = DA_UNMAP_GRANULARITY_DEFAULT;
+ dev->se_sub_dev->se_dev_attrib.unmap_granularity_alignment =
DA_UNMAP_GRANULARITY_ALIGNMENT_DEFAULT;
/*
* block_size is based on subsystem plugin dependent requirements.
*/
- DEV_ATTRIB(dev)->hw_block_size = limits->logical_block_size;
- DEV_ATTRIB(dev)->block_size = limits->logical_block_size;
+ dev->se_sub_dev->se_dev_attrib.hw_block_size = limits->logical_block_size;
+ dev->se_sub_dev->se_dev_attrib.block_size = limits->logical_block_size;
/*
* max_sectors is based on subsystem plugin dependent requirements.
*/
- DEV_ATTRIB(dev)->hw_max_sectors = limits->max_hw_sectors;
- DEV_ATTRIB(dev)->max_sectors = limits->max_sectors;
+ dev->se_sub_dev->se_dev_attrib.hw_max_sectors = limits->max_hw_sectors;
+ dev->se_sub_dev->se_dev_attrib.max_sectors = limits->max_sectors;
/*
* Set optimal_sectors from max_sectors, which can be lowered via
* configfs.
*/
- DEV_ATTRIB(dev)->optimal_sectors = limits->max_sectors;
+ dev->se_sub_dev->se_dev_attrib.optimal_sectors = limits->max_sectors;
/*
* queue_depth is based on subsystem plugin dependent requirements.
*/
- DEV_ATTRIB(dev)->hw_queue_depth = dev_limits->hw_queue_depth;
- DEV_ATTRIB(dev)->queue_depth = dev_limits->queue_depth;
+ dev->se_sub_dev->se_dev_attrib.hw_queue_depth = dev_limits->hw_queue_depth;
+ dev->se_sub_dev->se_dev_attrib.queue_depth = dev_limits->queue_depth;
}
int se_dev_set_task_timeout(struct se_device *dev, u32 task_timeout)
@@ -902,7 +902,7 @@ int se_dev_set_task_timeout(struct se_device *dev, u32 task_timeout)
" DA_TASK_TIMEOUT_MAX\n", dev, task_timeout);
return -EINVAL;
} else {
- DEV_ATTRIB(dev)->task_timeout = task_timeout;
+ dev->se_sub_dev->se_dev_attrib.task_timeout = task_timeout;
printk(KERN_INFO "dev[%p]: Set SE Device task_timeout: %u\n",
dev, task_timeout);
}
@@ -914,9 +914,9 @@ int se_dev_set_max_unmap_lba_count(
struct se_device *dev,
u32 max_unmap_lba_count)
{
- DEV_ATTRIB(dev)->max_unmap_lba_count = max_unmap_lba_count;
+ dev->se_sub_dev->se_dev_attrib.max_unmap_lba_count = max_unmap_lba_count;
printk(KERN_INFO "dev[%p]: Set max_unmap_lba_count: %u\n",
- dev, DEV_ATTRIB(dev)->max_unmap_lba_count);
+ dev, dev->se_sub_dev->se_dev_attrib.max_unmap_lba_count);
return 0;
}
@@ -924,9 +924,10 @@ int se_dev_set_max_unmap_block_desc_count(
struct se_device *dev,
u32 max_unmap_block_desc_count)
{
- DEV_ATTRIB(dev)->max_unmap_block_desc_count = max_unmap_block_desc_count;
+ dev->se_sub_dev->se_dev_attrib.max_unmap_block_desc_count =
+ max_unmap_block_desc_count;
printk(KERN_INFO "dev[%p]: Set max_unmap_block_desc_count: %u\n",
- dev, DEV_ATTRIB(dev)->max_unmap_block_desc_count);
+ dev, dev->se_sub_dev->se_dev_attrib.max_unmap_block_desc_count);
return 0;
}
@@ -934,9 +935,9 @@ int se_dev_set_unmap_granularity(
struct se_device *dev,
u32 unmap_granularity)
{
- DEV_ATTRIB(dev)->unmap_granularity = unmap_granularity;
+ dev->se_sub_dev->se_dev_attrib.unmap_granularity = unmap_granularity;
printk(KERN_INFO "dev[%p]: Set unmap_granularity: %u\n",
- dev, DEV_ATTRIB(dev)->unmap_granularity);
+ dev, dev->se_sub_dev->se_dev_attrib.unmap_granularity);
return 0;
}
@@ -944,9 +945,9 @@ int se_dev_set_unmap_granularity_alignment(
struct se_device *dev,
u32 unmap_granularity_alignment)
{
- DEV_ATTRIB(dev)->unmap_granularity_alignment = unmap_granularity_alignment;
+ dev->se_sub_dev->se_dev_attrib.unmap_granularity_alignment = unmap_granularity_alignment;
printk(KERN_INFO "dev[%p]: Set unmap_granularity_alignment: %u\n",
- dev, DEV_ATTRIB(dev)->unmap_granularity_alignment);
+ dev, dev->se_sub_dev->se_dev_attrib.unmap_granularity_alignment);
return 0;
}
@@ -956,17 +957,17 @@ int se_dev_set_emulate_dpo(struct se_device *dev, int flag)
printk(KERN_ERR "Illegal value %d\n", flag);
return -EINVAL;
}
- if (TRANSPORT(dev)->dpo_emulated == NULL) {
- printk(KERN_ERR "TRANSPORT(dev)->dpo_emulated is NULL\n");
+ if (dev->transport->dpo_emulated == NULL) {
+ printk(KERN_ERR "dev->transport->dpo_emulated is NULL\n");
return -EINVAL;
}
- if (TRANSPORT(dev)->dpo_emulated(dev) == 0) {
- printk(KERN_ERR "TRANSPORT(dev)->dpo_emulated not supported\n");
+ if (dev->transport->dpo_emulated(dev) == 0) {
+ printk(KERN_ERR "dev->transport->dpo_emulated not supported\n");
return -EINVAL;
}
- DEV_ATTRIB(dev)->emulate_dpo = flag;
+ dev->se_sub_dev->se_dev_attrib.emulate_dpo = flag;
printk(KERN_INFO "dev[%p]: SE Device Page Out (DPO) Emulation"
- " bit: %d\n", dev, DEV_ATTRIB(dev)->emulate_dpo);
+ " bit: %d\n", dev, dev->se_sub_dev->se_dev_attrib.emulate_dpo);
return 0;
}
@@ -976,17 +977,17 @@ int se_dev_set_emulate_fua_write(struct se_device *dev, int flag)
printk(KERN_ERR "Illegal value %d\n", flag);
return -EINVAL;
}
- if (TRANSPORT(dev)->fua_write_emulated == NULL) {
- printk(KERN_ERR "TRANSPORT(dev)->fua_write_emulated is NULL\n");
+ if (dev->transport->fua_write_emulated == NULL) {
+ printk(KERN_ERR "dev->transport->fua_write_emulated is NULL\n");
return -EINVAL;
}
- if (TRANSPORT(dev)->fua_write_emulated(dev) == 0) {
- printk(KERN_ERR "TRANSPORT(dev)->fua_write_emulated not supported\n");
+ if (dev->transport->fua_write_emulated(dev) == 0) {
+ printk(KERN_ERR "dev->transport->fua_write_emulated not supported\n");
return -EINVAL;
}
- DEV_ATTRIB(dev)->emulate_fua_write = flag;
+ dev->se_sub_dev->se_dev_attrib.emulate_fua_write = flag;
printk(KERN_INFO "dev[%p]: SE Device Forced Unit Access WRITEs: %d\n",
- dev, DEV_ATTRIB(dev)->emulate_fua_write);
+ dev, dev->se_sub_dev->se_dev_attrib.emulate_fua_write);
return 0;
}
@@ -996,17 +997,17 @@ int se_dev_set_emulate_fua_read(struct se_device *dev, int flag)
printk(KERN_ERR "Illegal value %d\n", flag);
return -EINVAL;
}
- if (TRANSPORT(dev)->fua_read_emulated == NULL) {
- printk(KERN_ERR "TRANSPORT(dev)->fua_read_emulated is NULL\n");
+ if (dev->transport->fua_read_emulated == NULL) {
+ printk(KERN_ERR "dev->transport->fua_read_emulated is NULL\n");
return -EINVAL;
}
- if (TRANSPORT(dev)->fua_read_emulated(dev) == 0) {
- printk(KERN_ERR "TRANSPORT(dev)->fua_read_emulated not supported\n");
+ if (dev->transport->fua_read_emulated(dev) == 0) {
+ printk(KERN_ERR "dev->transport->fua_read_emulated not supported\n");
return -EINVAL;
}
- DEV_ATTRIB(dev)->emulate_fua_read = flag;
+ dev->se_sub_dev->se_dev_attrib.emulate_fua_read = flag;
printk(KERN_INFO "dev[%p]: SE Device Forced Unit Access READs: %d\n",
- dev, DEV_ATTRIB(dev)->emulate_fua_read);
+ dev, dev->se_sub_dev->se_dev_attrib.emulate_fua_read);
return 0;
}
@@ -1016,17 +1017,17 @@ int se_dev_set_emulate_write_cache(struct se_device *dev, int flag)
printk(KERN_ERR "Illegal value %d\n", flag);
return -EINVAL;
}
- if (TRANSPORT(dev)->write_cache_emulated == NULL) {
- printk(KERN_ERR "TRANSPORT(dev)->write_cache_emulated is NULL\n");
+ if (dev->transport->write_cache_emulated == NULL) {
+ printk(KERN_ERR "dev->transport->write_cache_emulated is NULL\n");
return -EINVAL;
}
- if (TRANSPORT(dev)->write_cache_emulated(dev) == 0) {
- printk(KERN_ERR "TRANSPORT(dev)->write_cache_emulated not supported\n");
+ if (dev->transport->write_cache_emulated(dev) == 0) {
+ printk(KERN_ERR "dev->transport->write_cache_emulated not supported\n");
return -EINVAL;
}
- DEV_ATTRIB(dev)->emulate_write_cache = flag;
+ dev->se_sub_dev->se_dev_attrib.emulate_write_cache = flag;
printk(KERN_INFO "dev[%p]: SE Device WRITE_CACHE_EMULATION flag: %d\n",
- dev, DEV_ATTRIB(dev)->emulate_write_cache);
+ dev, dev->se_sub_dev->se_dev_attrib.emulate_write_cache);
return 0;
}
@@ -1044,9 +1045,9 @@ int se_dev_set_emulate_ua_intlck_ctrl(struct se_device *dev, int flag)
atomic_read(&dev->dev_export_obj.obj_access_count));
return -EINVAL;
}
- DEV_ATTRIB(dev)->emulate_ua_intlck_ctrl = flag;
+ dev->se_sub_dev->se_dev_attrib.emulate_ua_intlck_ctrl = flag;
printk(KERN_INFO "dev[%p]: SE Device UA_INTRLCK_CTRL flag: %d\n",
- dev, DEV_ATTRIB(dev)->emulate_ua_intlck_ctrl);
+ dev, dev->se_sub_dev->se_dev_attrib.emulate_ua_intlck_ctrl);
return 0;
}
@@ -1064,9 +1065,9 @@ int se_dev_set_emulate_tas(struct se_device *dev, int flag)
atomic_read(&dev->dev_export_obj.obj_access_count));
return -EINVAL;
}
- DEV_ATTRIB(dev)->emulate_tas = flag;
+ dev->se_sub_dev->se_dev_attrib.emulate_tas = flag;
printk(KERN_INFO "dev[%p]: SE Device TASK_ABORTED status bit: %s\n",
- dev, (DEV_ATTRIB(dev)->emulate_tas) ? "Enabled" : "Disabled");
+ dev, (dev->se_sub_dev->se_dev_attrib.emulate_tas) ? "Enabled" : "Disabled");
return 0;
}
@@ -1081,12 +1082,12 @@ int se_dev_set_emulate_tpu(struct se_device *dev, int flag)
* We expect this value to be non-zero when generic Block Layer
* Discard supported is detected iblock_create_virtdevice().
*/
- if (!(DEV_ATTRIB(dev)->max_unmap_block_desc_count)) {
+ if (!(dev->se_sub_dev->se_dev_attrib.max_unmap_block_desc_count)) {
printk(KERN_ERR "Generic Block Discard not supported\n");
return -ENOSYS;
}
- DEV_ATTRIB(dev)->emulate_tpu = flag;
+ dev->se_sub_dev->se_dev_attrib.emulate_tpu = flag;
printk(KERN_INFO "dev[%p]: SE Device Thin Provisioning UNMAP bit: %d\n",
dev, flag);
return 0;
@@ -1102,12 +1103,12 @@ int se_dev_set_emulate_tpws(struct se_device *dev, int flag)
* We expect this value to be non-zero when generic Block Layer
* Discard supported is detected iblock_create_virtdevice().
*/
- if (!(DEV_ATTRIB(dev)->max_unmap_block_desc_count)) {
+ if (!(dev->se_sub_dev->se_dev_attrib.max_unmap_block_desc_count)) {
printk(KERN_ERR "Generic Block Discard not supported\n");
return -ENOSYS;
}
- DEV_ATTRIB(dev)->emulate_tpws = flag;
+ dev->se_sub_dev->se_dev_attrib.emulate_tpws = flag;
printk(KERN_INFO "dev[%p]: SE Device Thin Provisioning WRITE_SAME: %d\n",
dev, flag);
return 0;
@@ -1119,9 +1120,9 @@ int se_dev_set_enforce_pr_isids(struct se_device *dev, int flag)
printk(KERN_ERR "Illegal value %d\n", flag);
return -EINVAL;
}
- DEV_ATTRIB(dev)->enforce_pr_isids = flag;
+ dev->se_sub_dev->se_dev_attrib.enforce_pr_isids = flag;
printk(KERN_INFO "dev[%p]: SE Device enforce_pr_isids bit: %s\n", dev,
- (DEV_ATTRIB(dev)->enforce_pr_isids) ? "Enabled" : "Disabled");
+ (dev->se_sub_dev->se_dev_attrib.enforce_pr_isids) ? "Enabled" : "Disabled");
return 0;
}
@@ -1144,27 +1145,27 @@ int se_dev_set_queue_depth(struct se_device *dev, u32 queue_depth)
return -EINVAL;
}
- if (TRANSPORT(dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
- if (queue_depth > DEV_ATTRIB(dev)->hw_queue_depth) {
+ if (dev->transport->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
+ if (queue_depth > dev->se_sub_dev->se_dev_attrib.hw_queue_depth) {
printk(KERN_ERR "dev[%p]: Passed queue_depth: %u"
" exceeds TCM/SE_Device TCQ: %u\n",
dev, queue_depth,
- DEV_ATTRIB(dev)->hw_queue_depth);
+ dev->se_sub_dev->se_dev_attrib.hw_queue_depth);
return -EINVAL;
}
} else {
- if (queue_depth > DEV_ATTRIB(dev)->queue_depth) {
- if (queue_depth > DEV_ATTRIB(dev)->hw_queue_depth) {
+ if (queue_depth > dev->se_sub_dev->se_dev_attrib.queue_depth) {
+ if (queue_depth > dev->se_sub_dev->se_dev_attrib.hw_queue_depth) {
printk(KERN_ERR "dev[%p]: Passed queue_depth:"
" %u exceeds TCM/SE_Device MAX"
" TCQ: %u\n", dev, queue_depth,
- DEV_ATTRIB(dev)->hw_queue_depth);
+ dev->se_sub_dev->se_dev_attrib.hw_queue_depth);
return -EINVAL;
}
}
}
- DEV_ATTRIB(dev)->queue_depth = dev->queue_depth = queue_depth;
+ dev->se_sub_dev->se_dev_attrib.queue_depth = dev->queue_depth = queue_depth;
if (queue_depth > orig_queue_depth)
atomic_add(queue_depth - orig_queue_depth, &dev->depth_left);
else if (queue_depth < orig_queue_depth)
@@ -1196,21 +1197,21 @@ int se_dev_set_max_sectors(struct se_device *dev, u32 max_sectors)
DA_STATUS_MAX_SECTORS_MIN);
return -EINVAL;
}
- if (TRANSPORT(dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
- if (max_sectors > DEV_ATTRIB(dev)->hw_max_sectors) {
+ if (dev->transport->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
+ if (max_sectors > dev->se_sub_dev->se_dev_attrib.hw_max_sectors) {
printk(KERN_ERR "dev[%p]: Passed max_sectors: %u"
" greater than TCM/SE_Device max_sectors:"
" %u\n", dev, max_sectors,
- DEV_ATTRIB(dev)->hw_max_sectors);
+ dev->se_sub_dev->se_dev_attrib.hw_max_sectors);
return -EINVAL;
}
} else {
if (!(force) && (max_sectors >
- DEV_ATTRIB(dev)->hw_max_sectors)) {
+ dev->se_sub_dev->se_dev_attrib.hw_max_sectors)) {
printk(KERN_ERR "dev[%p]: Passed max_sectors: %u"
" greater than TCM/SE_Device max_sectors"
": %u, use force=1 to override.\n", dev,
- max_sectors, DEV_ATTRIB(dev)->hw_max_sectors);
+ max_sectors, dev->se_sub_dev->se_dev_attrib.hw_max_sectors);
return -EINVAL;
}
if (max_sectors > DA_STATUS_MAX_SECTORS_MAX) {
@@ -1222,7 +1223,7 @@ int se_dev_set_max_sectors(struct se_device *dev, u32 max_sectors)
}
}
- DEV_ATTRIB(dev)->max_sectors = max_sectors;
+ dev->se_sub_dev->se_dev_attrib.max_sectors = max_sectors;
printk("dev[%p]: SE Device max_sectors changed to %u\n",
dev, max_sectors);
return 0;
@@ -1236,19 +1237,19 @@ int se_dev_set_optimal_sectors(struct se_device *dev, u32 optimal_sectors)
dev, atomic_read(&dev->dev_export_obj.obj_access_count));
return -EINVAL;
}
- if (TRANSPORT(dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
+ if (dev->transport->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
printk(KERN_ERR "dev[%p]: Passed optimal_sectors cannot be"
" changed for TCM/pSCSI\n", dev);
return -EINVAL;
}
- if (optimal_sectors > DEV_ATTRIB(dev)->max_sectors) {
+ if (optimal_sectors > dev->se_sub_dev->se_dev_attrib.max_sectors) {
printk(KERN_ERR "dev[%p]: Passed optimal_sectors %u cannot be"
" greater than max_sectors: %u\n", dev,
- optimal_sectors, DEV_ATTRIB(dev)->max_sectors);
+ optimal_sectors, dev->se_sub_dev->se_dev_attrib.max_sectors);
return -EINVAL;
}
- DEV_ATTRIB(dev)->optimal_sectors = optimal_sectors;
+ dev->se_sub_dev->se_dev_attrib.optimal_sectors = optimal_sectors;
printk(KERN_INFO "dev[%p]: SE Device optimal_sectors changed to %u\n",
dev, optimal_sectors);
return 0;
@@ -1273,14 +1274,14 @@ int se_dev_set_block_size(struct se_device *dev, u32 block_size)
return -EINVAL;
}
- if (TRANSPORT(dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
+ if (dev->transport->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
printk(KERN_ERR "dev[%p]: Not allowed to change block_size for"
" Physical Device, use for Linux/SCSI to change"
" block_size for underlying hardware\n", dev);
return -EINVAL;
}
- DEV_ATTRIB(dev)->block_size = block_size;
+ dev->se_sub_dev->se_dev_attrib.block_size = block_size;
printk(KERN_INFO "dev[%p]: SE Device block_size changed to %u\n",
dev, block_size);
return 0;
@@ -1314,14 +1315,14 @@ struct se_lun *core_dev_add_lun(
return NULL;
printk(KERN_INFO "%s_TPG[%u]_LUN[%u] - Activated %s Logical Unit from"
- " CORE HBA: %u\n", TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_tag(tpg), lun_p->unpacked_lun,
- TPG_TFO(tpg)->get_fabric_name(), hba->hba_id);
+ " CORE HBA: %u\n", tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg), lun_p->unpacked_lun,
+ tpg->se_tpg_tfo->get_fabric_name(), hba->hba_id);
/*
* Update LUN maps for dynamically added initiators when
* generate_node_acl is enabled.
*/
- if (TPG_TFO(tpg)->tpg_check_demo_mode(tpg)) {
+ if (tpg->se_tpg_tfo->tpg_check_demo_mode(tpg)) {
struct se_node_acl *acl;
spin_lock_bh(&tpg->acl_node_lock);
list_for_each_entry(acl, &tpg->acl_node_list, acl_list) {
@@ -1355,9 +1356,9 @@ int core_dev_del_lun(
core_tpg_post_dellun(tpg, lun);
printk(KERN_INFO "%s_TPG[%u]_LUN[%u] - Deactivated %s Logical Unit from"
- " device object\n", TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_tag(tpg), unpacked_lun,
- TPG_TFO(tpg)->get_fabric_name());
+ " device object\n", tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg), unpacked_lun,
+ tpg->se_tpg_tfo->get_fabric_name());
return 0;
}
@@ -1370,9 +1371,9 @@ struct se_lun *core_get_lun_from_tpg(struct se_portal_group *tpg, u32 unpacked_l
if (unpacked_lun > (TRANSPORT_MAX_LUNS_PER_TPG-1)) {
printk(KERN_ERR "%s LUN: %u exceeds TRANSPORT_MAX_LUNS"
"_PER_TPG-1: %u for Target Portal Group: %hu\n",
- TPG_TFO(tpg)->get_fabric_name(), unpacked_lun,
+ tpg->se_tpg_tfo->get_fabric_name(), unpacked_lun,
TRANSPORT_MAX_LUNS_PER_TPG-1,
- TPG_TFO(tpg)->tpg_get_tag(tpg));
+ tpg->se_tpg_tfo->tpg_get_tag(tpg));
spin_unlock(&tpg->tpg_lun_lock);
return NULL;
}
@@ -1381,8 +1382,8 @@ struct se_lun *core_get_lun_from_tpg(struct se_portal_group *tpg, u32 unpacked_l
if (lun->lun_status != TRANSPORT_LUN_STATUS_FREE) {
printk(KERN_ERR "%s Logical Unit Number: %u is not free on"
" Target Portal Group: %hu, ignoring request.\n",
- TPG_TFO(tpg)->get_fabric_name(), unpacked_lun,
- TPG_TFO(tpg)->tpg_get_tag(tpg));
+ tpg->se_tpg_tfo->get_fabric_name(), unpacked_lun,
+ tpg->se_tpg_tfo->tpg_get_tag(tpg));
spin_unlock(&tpg->tpg_lun_lock);
return NULL;
}
@@ -1403,9 +1404,9 @@ static struct se_lun *core_dev_get_lun(struct se_portal_group *tpg, u32 unpacked
if (unpacked_lun > (TRANSPORT_MAX_LUNS_PER_TPG-1)) {
printk(KERN_ERR "%s LUN: %u exceeds TRANSPORT_MAX_LUNS_PER"
"_TPG-1: %u for Target Portal Group: %hu\n",
- TPG_TFO(tpg)->get_fabric_name(), unpacked_lun,
+ tpg->se_tpg_tfo->get_fabric_name(), unpacked_lun,
TRANSPORT_MAX_LUNS_PER_TPG-1,
- TPG_TFO(tpg)->tpg_get_tag(tpg));
+ tpg->se_tpg_tfo->tpg_get_tag(tpg));
spin_unlock(&tpg->tpg_lun_lock);
return NULL;
}
@@ -1414,8 +1415,8 @@ static struct se_lun *core_dev_get_lun(struct se_portal_group *tpg, u32 unpacked
if (lun->lun_status != TRANSPORT_LUN_STATUS_ACTIVE) {
printk(KERN_ERR "%s Logical Unit Number: %u is not active on"
" Target Portal Group: %hu, ignoring request.\n",
- TPG_TFO(tpg)->get_fabric_name(), unpacked_lun,
- TPG_TFO(tpg)->tpg_get_tag(tpg));
+ tpg->se_tpg_tfo->get_fabric_name(), unpacked_lun,
+ tpg->se_tpg_tfo->tpg_get_tag(tpg));
spin_unlock(&tpg->tpg_lun_lock);
return NULL;
}
@@ -1435,7 +1436,7 @@ struct se_lun_acl *core_dev_init_initiator_node_lun_acl(
if (strlen(initiatorname) >= TRANSPORT_IQN_LEN) {
printk(KERN_ERR "%s InitiatorName exceeds maximum size.\n",
- TPG_TFO(tpg)->get_fabric_name());
+ tpg->se_tpg_tfo->get_fabric_name());
*ret = -EOVERFLOW;
return NULL;
}
@@ -1472,8 +1473,8 @@ int core_dev_add_initiator_node_lun_acl(
if (!(lun)) {
printk(KERN_ERR "%s Logical Unit Number: %u is not active on"
" Target Portal Group: %hu, ignoring request.\n",
- TPG_TFO(tpg)->get_fabric_name(), unpacked_lun,
- TPG_TFO(tpg)->tpg_get_tag(tpg));
+ tpg->se_tpg_tfo->get_fabric_name(), unpacked_lun,
+ tpg->se_tpg_tfo->tpg_get_tag(tpg));
return -EINVAL;
}
@@ -1498,8 +1499,8 @@ int core_dev_add_initiator_node_lun_acl(
spin_unlock(&lun->lun_acl_lock);
printk(KERN_INFO "%s_TPG[%hu]_LUN[%u->%u] - Added %s ACL for "
- " InitiatorNode: %s\n", TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_tag(tpg), unpacked_lun, lacl->mapped_lun,
+ " InitiatorNode: %s\n", tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg), unpacked_lun, lacl->mapped_lun,
(lun_access & TRANSPORT_LUNFLAGS_READ_WRITE) ? "RW" : "RO",
lacl->initiatorname);
/*
@@ -1538,8 +1539,8 @@ int core_dev_del_initiator_node_lun_acl(
printk(KERN_INFO "%s_TPG[%hu]_LUN[%u] - Removed ACL for"
" InitiatorNode: %s Mapped LUN: %u\n",
- TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_tag(tpg), lun->unpacked_lun,
+ tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg), lun->unpacked_lun,
lacl->initiatorname, lacl->mapped_lun);
return 0;
@@ -1550,9 +1551,9 @@ void core_dev_free_initiator_node_lun_acl(
struct se_lun_acl *lacl)
{
printk("%s_TPG[%hu] - Freeing ACL for %s InitiatorNode: %s"
- " Mapped LUN: %u\n", TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_tag(tpg),
- TPG_TFO(tpg)->get_fabric_name(),
+ " Mapped LUN: %u\n", tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg),
+ tpg->se_tpg_tfo->get_fabric_name(),
lacl->initiatorname, lacl->mapped_lun);
kfree(lacl);
@@ -1584,14 +1585,14 @@ int core_dev_setup_virtual_lun0(void)
INIT_LIST_HEAD(&se_dev->se_dev_node);
INIT_LIST_HEAD(&se_dev->t10_wwn.t10_vpd_list);
spin_lock_init(&se_dev->t10_wwn.t10_vpd_lock);
- INIT_LIST_HEAD(&se_dev->t10_reservation.registration_list);
- INIT_LIST_HEAD(&se_dev->t10_reservation.aptpl_reg_list);
- spin_lock_init(&se_dev->t10_reservation.registration_lock);
- spin_lock_init(&se_dev->t10_reservation.aptpl_reg_lock);
+ INIT_LIST_HEAD(&se_dev->t10_pr.registration_list);
+ INIT_LIST_HEAD(&se_dev->t10_pr.aptpl_reg_list);
+ spin_lock_init(&se_dev->t10_pr.registration_lock);
+ spin_lock_init(&se_dev->t10_pr.aptpl_reg_lock);
INIT_LIST_HEAD(&se_dev->t10_alua.tg_pt_gps_list);
spin_lock_init(&se_dev->t10_alua.tg_pt_gps_lock);
spin_lock_init(&se_dev->se_dev_lock);
- se_dev->t10_reservation.pr_aptpl_buf_len = PR_APTPL_BUF_LEN;
+ se_dev->t10_pr.pr_aptpl_buf_len = PR_APTPL_BUF_LEN;
se_dev->t10_wwn.t10_sub_dev = se_dev;
se_dev->t10_alua.t10_sub_dev = se_dev;
se_dev->se_dev_attrib.da_sub_dev = se_dev;
diff --git a/drivers/target/target_core_fabric_configfs.c b/drivers/target/target_core_fabric_configfs.c
index 07ab5a3..0b1659d 100644
--- a/drivers/target/target_core_fabric_configfs.c
+++ b/drivers/target/target_core_fabric_configfs.c
@@ -118,7 +118,7 @@ static int target_fabric_mappedlun_link(
lun_access = deve->lun_flags;
else
lun_access =
- (TPG_TFO(se_tpg)->tpg_check_prod_mode_write_protect(
+ (se_tpg->se_tpg_tfo->tpg_check_prod_mode_write_protect(
se_tpg)) ? TRANSPORT_LUNFLAGS_READ_ONLY :
TRANSPORT_LUNFLAGS_READ_WRITE;
spin_unlock_irq(&lacl->se_lun_nacl->device_list_lock);
@@ -204,7 +204,7 @@ static ssize_t target_fabric_mappedlun_store_write_protect(
printk(KERN_INFO "%s_ConfigFS: Changed Initiator ACL: %s"
" Mapped LUN: %u Write Protect bit to %s\n",
- TPG_TFO(se_tpg)->get_fabric_name(),
+ se_tpg->se_tpg_tfo->get_fabric_name(),
lacl->initiatorname, lacl->mapped_lun, (op) ? "ON" : "OFF");
return count;
@@ -379,7 +379,7 @@ static struct config_group *target_fabric_make_mappedlun(
lacl_cg->default_groups[0] = &lacl->ml_stat_grps.stat_group;
lacl_cg->default_groups[1] = NULL;
- ml_stat_grp = &ML_STAT_GRPS(lacl)->stat_group;
+ ml_stat_grp = &lacl->ml_stat_grps.stat_group;
ml_stat_grp->default_groups = kzalloc(sizeof(struct config_group) * 3,
GFP_KERNEL);
if (!ml_stat_grp->default_groups) {
@@ -408,7 +408,7 @@ static void target_fabric_drop_mappedlun(
struct config_group *lacl_cg = NULL, *ml_stat_grp = NULL;
int i;
- ml_stat_grp = &ML_STAT_GRPS(lacl)->stat_group;
+ ml_stat_grp = &lacl->ml_stat_grps.stat_group;
for (i = 0; ml_stat_grp->default_groups[i]; i++) {
df_item = &ml_stat_grp->default_groups[i]->cg_item;
ml_stat_grp->default_groups[i] = NULL;
@@ -914,7 +914,7 @@ static struct config_group *target_fabric_make_lun(
lun_cg->default_groups[0] = &lun->port_stat_grps.stat_group;
lun_cg->default_groups[1] = NULL;
- port_stat_grp = &PORT_STAT_GRP(lun)->stat_group;
+ port_stat_grp = &lun->port_stat_grps.stat_group;
port_stat_grp->default_groups = kzalloc(sizeof(struct config_group) * 3,
GFP_KERNEL);
if (!port_stat_grp->default_groups) {
@@ -941,7 +941,7 @@ static void target_fabric_drop_lun(
struct config_group *lun_cg, *port_stat_grp;
int i;
- port_stat_grp = &PORT_STAT_GRP(lun)->stat_group;
+ port_stat_grp = &lun->port_stat_grps.stat_group;
for (i = 0; port_stat_grp->default_groups[i]; i++) {
df_item = &port_stat_grp->default_groups[i]->cg_item;
port_stat_grp->default_groups[i] = NULL;
diff --git a/drivers/target/target_core_file.c b/drivers/target/target_core_file.c
index 5ca1f5b..0c44bc0 100644
--- a/drivers/target/target_core_file.c
+++ b/drivers/target/target_core_file.c
@@ -72,7 +72,7 @@ static int fd_attach_hba(struct se_hba *hba, u32 host_id)
fd_host->fd_host_id = host_id;
- hba->hba_ptr = (void *) fd_host;
+ hba->hba_ptr = fd_host;
printk(KERN_INFO "CORE_HBA[%d] - TCM FILEIO HBA Driver %s on Generic"
" Target Core Stack %s\n", hba->hba_id, FD_VERSION,
@@ -279,7 +279,7 @@ fd_alloc_task(struct se_cmd *cmd)
return NULL;
}
- fd_req->fd_dev = SE_DEV(cmd)->dev_ptr;
+ fd_req->fd_dev = cmd->se_lun->lun_se_dev->dev_ptr;
return &fd_req->fd_task;
}
@@ -291,7 +291,8 @@ static int fd_do_readv(struct se_task *task)
struct scatterlist *sg = task->task_sg;
struct iovec *iov;
mm_segment_t old_fs;
- loff_t pos = (task->task_lba * DEV_ATTRIB(task->se_dev)->block_size);
+ loff_t pos = (task->task_lba *
+ task->se_dev->se_sub_dev->se_dev_attrib.block_size);
int ret = 0, i;
iov = kzalloc(sizeof(struct iovec) * task->task_sg_num, GFP_KERNEL);
@@ -341,7 +342,8 @@ static int fd_do_writev(struct se_task *task)
struct scatterlist *sg = task->task_sg;
struct iovec *iov;
mm_segment_t old_fs;
- loff_t pos = (task->task_lba * DEV_ATTRIB(task->se_dev)->block_size);
+ loff_t pos = (task->task_lba *
+ task->se_dev->se_sub_dev->se_dev_attrib.block_size);
int ret, i = 0;
iov = kzalloc(sizeof(struct iovec) * task->task_sg_num, GFP_KERNEL);
@@ -372,7 +374,7 @@ static int fd_do_writev(struct se_task *task)
static void fd_emulate_sync_cache(struct se_task *task)
{
- struct se_cmd *cmd = TASK_CMD(task);
+ struct se_cmd *cmd = task->task_se_cmd;
struct se_device *dev = cmd->se_dev;
struct fd_dev *fd_dev = dev->dev_ptr;
int immed = (cmd->t_task->t_task_cdb[1] & 0x2);
@@ -393,7 +395,7 @@ static void fd_emulate_sync_cache(struct se_task *task)
start = 0;
end = LLONG_MAX;
} else {
- start = cmd->t_task->t_task_lba * DEV_ATTRIB(dev)->block_size;
+ start = cmd->t_task->t_task_lba * dev->se_sub_dev->se_dev_attrib.block_size;
if (cmd->data_length)
end = start + cmd->data_length;
else
@@ -443,7 +445,7 @@ static void fd_emulate_write_fua(struct se_cmd *cmd, struct se_task *task)
{
struct se_device *dev = cmd->se_dev;
struct fd_dev *fd_dev = dev->dev_ptr;
- loff_t start = task->task_lba * DEV_ATTRIB(dev)->block_size;
+ loff_t start = task->task_lba * dev->se_sub_dev->se_dev_attrib.block_size;
loff_t end = start + task->task_size;
int ret;
@@ -471,9 +473,9 @@ static int fd_do_task(struct se_task *task)
ret = fd_do_writev(task);
if (ret > 0 &&
- DEV_ATTRIB(dev)->emulate_write_cache > 0 &&
- DEV_ATTRIB(dev)->emulate_fua_write > 0 &&
- T_TASK(cmd)->t_tasks_fua) {
+ dev->se_sub_dev->se_dev_attrib.emulate_write_cache > 0 &&
+ dev->se_sub_dev->se_dev_attrib.emulate_fua_write > 0 &&
+ cmd->t_task->t_tasks_fua) {
/*
* We might need to be a bit smarter here
* and return some sense data to let the initiator
@@ -651,7 +653,7 @@ static sector_t fd_get_blocks(struct se_device *dev)
{
struct fd_dev *fd_dev = dev->dev_ptr;
unsigned long long blocks_long = div_u64(fd_dev->fd_dev_size,
- DEV_ATTRIB(dev)->block_size);
+ dev->se_sub_dev->se_dev_attrib.block_size);
return blocks_long;
}
diff --git a/drivers/target/target_core_iblock.c b/drivers/target/target_core_iblock.c
index 5bd4f1c..fb15987 100644
--- a/drivers/target/target_core_iblock.c
+++ b/drivers/target/target_core_iblock.c
@@ -185,15 +185,15 @@ static struct se_device *iblock_create_virtdevice(
* in ATA and we need to set TPE=1
*/
if (blk_queue_discard(q)) {
- DEV_ATTRIB(dev)->max_unmap_lba_count =
+ dev->se_sub_dev->se_dev_attrib.max_unmap_lba_count =
q->limits.max_discard_sectors;
/*
* Currently hardcoded to 1 in Linux/SCSI code..
*/
- DEV_ATTRIB(dev)->max_unmap_block_desc_count = 1;
- DEV_ATTRIB(dev)->unmap_granularity =
+ dev->se_sub_dev->se_dev_attrib.max_unmap_block_desc_count = 1;
+ dev->se_sub_dev->se_dev_attrib.unmap_granularity =
q->limits.discard_granularity;
- DEV_ATTRIB(dev)->unmap_granularity_alignment =
+ dev->se_sub_dev->se_dev_attrib.unmap_granularity_alignment =
q->limits.discard_alignment;
printk(KERN_INFO "IBLOCK: BLOCK Discard support available,"
@@ -240,7 +240,7 @@ iblock_alloc_task(struct se_cmd *cmd)
return NULL;
}
- ib_req->ib_dev = SE_DEV(cmd)->dev_ptr;
+ ib_req->ib_dev = cmd->se_lun->lun_se_dev->dev_ptr;
atomic_set(&ib_req->ib_bio_cnt, 0);
return &ib_req->ib_task;
}
@@ -254,12 +254,12 @@ static unsigned long long iblock_emulate_read_cap_with_block_size(
bdev_logical_block_size(bd)) - 1);
u32 block_size = bdev_logical_block_size(bd);
- if (block_size == DEV_ATTRIB(dev)->block_size)
+ if (block_size == dev->se_sub_dev->se_dev_attrib.block_size)
return blocks_long;
switch (block_size) {
case 4096:
- switch (DEV_ATTRIB(dev)->block_size) {
+ switch (dev->se_sub_dev->se_dev_attrib.block_size) {
case 2048:
blocks_long <<= 1;
break;
@@ -273,7 +273,7 @@ static unsigned long long iblock_emulate_read_cap_with_block_size(
}
break;
case 2048:
- switch (DEV_ATTRIB(dev)->block_size) {
+ switch (dev->se_sub_dev->se_dev_attrib.block_size) {
case 4096:
blocks_long >>= 1;
break;
@@ -288,7 +288,7 @@ static unsigned long long iblock_emulate_read_cap_with_block_size(
}
break;
case 1024:
- switch (DEV_ATTRIB(dev)->block_size) {
+ switch (dev->se_sub_dev->se_dev_attrib.block_size) {
case 4096:
blocks_long >>= 2;
break;
@@ -303,7 +303,7 @@ static unsigned long long iblock_emulate_read_cap_with_block_size(
}
break;
case 512:
- switch (DEV_ATTRIB(dev)->block_size) {
+ switch (dev->se_sub_dev->se_dev_attrib.block_size) {
case 4096:
blocks_long >>= 3;
break;
@@ -329,9 +329,9 @@ static unsigned long long iblock_emulate_read_cap_with_block_size(
*/
static void iblock_emulate_sync_cache(struct se_task *task)
{
- struct se_cmd *cmd = TASK_CMD(task);
+ struct se_cmd *cmd = task->task_se_cmd;
struct iblock_dev *ib_dev = cmd->se_dev->dev_ptr;
- int immed = (T_TASK(cmd)->t_task_cdb[1] & 0x2);
+ int immed = (cmd->t_task->t_task_cdb[1] & 0x2);
sector_t error_sector;
int ret;
@@ -398,9 +398,9 @@ static int iblock_do_task(struct se_task *task)
* Force data to disk if we pretend to not have a volatile
* write cache, or the initiator set the Force Unit Access bit.
*/
- if (DEV_ATTRIB(dev)->emulate_write_cache == 0 ||
- (DEV_ATTRIB(dev)->emulate_fua_write > 0 &&
- T_TASK(task->task_se_cmd)->t_tasks_fua))
+ if (dev->se_sub_dev->se_dev_attrib.emulate_write_cache == 0 ||
+ (dev->se_sub_dev->se_dev_attrib.emulate_fua_write > 0 &&
+ task->task_se_cmd->t_task->t_tasks_fua))
rw = WRITE_FUA;
else
rw = WRITE;
@@ -608,7 +608,7 @@ static struct bio *iblock_get_bio(
static int iblock_map_task_SG(struct se_task *task)
{
struct se_cmd *cmd = task->task_se_cmd;
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
struct iblock_dev *ib_dev = task->se_dev->dev_ptr;
struct iblock_req *ib_req = IBLOCK_REQ(task);
struct bio *bio = NULL, *hbio = NULL, *tbio = NULL;
@@ -620,17 +620,17 @@ static int iblock_map_task_SG(struct se_task *task)
* Do starting conversion up from non 512-byte blocksize with
* struct se_task SCSI blocksize into Linux/Block 512 units for BIO.
*/
- if (DEV_ATTRIB(dev)->block_size == 4096)
+ if (dev->se_sub_dev->se_dev_attrib.block_size == 4096)
block_lba = (task->task_lba << 3);
- else if (DEV_ATTRIB(dev)->block_size == 2048)
+ else if (dev->se_sub_dev->se_dev_attrib.block_size == 2048)
block_lba = (task->task_lba << 2);
- else if (DEV_ATTRIB(dev)->block_size == 1024)
+ else if (dev->se_sub_dev->se_dev_attrib.block_size == 1024)
block_lba = (task->task_lba << 1);
- else if (DEV_ATTRIB(dev)->block_size == 512)
+ else if (dev->se_sub_dev->se_dev_attrib.block_size == 512)
block_lba = task->task_lba;
else {
printk(KERN_ERR "Unsupported SCSI -> BLOCK LBA conversion:"
- " %u\n", DEV_ATTRIB(dev)->block_size);
+ " %u\n", dev->se_sub_dev->se_dev_attrib.block_size);
return PYX_TRANSPORT_LU_COMM_FAILURE;
}
diff --git a/drivers/target/target_core_pr.c b/drivers/target/target_core_pr.c
index 2ddbe0d..27a7525 100644
--- a/drivers/target/target_core_pr.c
+++ b/drivers/target/target_core_pr.c
@@ -143,8 +143,8 @@ static int core_scsi2_reservation_release(struct se_cmd *cmd)
dev->dev_flags &= ~DF_SPC2_RESERVATIONS_WITH_ISID;
}
printk(KERN_INFO "SCSI-2 Released reservation for %s LUN: %u ->"
- " MAPPED LUN: %u for %s\n", TPG_TFO(tpg)->get_fabric_name(),
- SE_LUN(cmd)->unpacked_lun, cmd->se_deve->mapped_lun,
+ " MAPPED LUN: %u for %s\n", tpg->se_tpg_tfo->get_fabric_name(),
+ cmd->se_lun->unpacked_lun, cmd->se_deve->mapped_lun,
sess->se_node_acl->initiatorname);
spin_unlock(&dev->dev_reservation_lock);
@@ -157,8 +157,8 @@ static int core_scsi2_reservation_reserve(struct se_cmd *cmd)
struct se_session *sess = cmd->se_sess;
struct se_portal_group *tpg = sess->se_tpg;
- if ((T_TASK(cmd)->t_task_cdb[1] & 0x01) &&
- (T_TASK(cmd)->t_task_cdb[1] & 0x02)) {
+ if ((cmd->t_task->t_task_cdb[1] & 0x01) &&
+ (cmd->t_task->t_task_cdb[1] & 0x02)) {
printk(KERN_ERR "LongIO and Obselete Bits set, returning"
" ILLEGAL_REQUEST\n");
return PYX_TRANSPORT_ILLEGAL_REQUEST;
@@ -174,12 +174,12 @@ static int core_scsi2_reservation_reserve(struct se_cmd *cmd)
if (dev->dev_reserved_node_acl &&
(dev->dev_reserved_node_acl != sess->se_node_acl)) {
printk(KERN_ERR "SCSI-2 RESERVATION CONFLIFT for %s fabric\n",
- TPG_TFO(tpg)->get_fabric_name());
+ tpg->se_tpg_tfo->get_fabric_name());
printk(KERN_ERR "Original reserver LUN: %u %s\n",
- SE_LUN(cmd)->unpacked_lun,
+ cmd->se_lun->unpacked_lun,
dev->dev_reserved_node_acl->initiatorname);
printk(KERN_ERR "Current attempt - LUN: %u -> MAPPED LUN: %u"
- " from %s \n", SE_LUN(cmd)->unpacked_lun,
+ " from %s \n", cmd->se_lun->unpacked_lun,
cmd->se_deve->mapped_lun,
sess->se_node_acl->initiatorname);
spin_unlock(&dev->dev_reservation_lock);
@@ -193,8 +193,8 @@ static int core_scsi2_reservation_reserve(struct se_cmd *cmd)
dev->dev_flags |= DF_SPC2_RESERVATIONS_WITH_ISID;
}
printk(KERN_INFO "SCSI-2 Reserved %s LUN: %u -> MAPPED LUN: %u"
- " for %s\n", TPG_TFO(tpg)->get_fabric_name(),
- SE_LUN(cmd)->unpacked_lun, cmd->se_deve->mapped_lun,
+ " for %s\n", tpg->se_tpg_tfo->get_fabric_name(),
+ cmd->se_lun->unpacked_lun, cmd->se_deve->mapped_lun,
sess->se_node_acl->initiatorname);
spin_unlock(&dev->dev_reservation_lock);
@@ -215,9 +215,9 @@ int core_scsi2_emulate_crh(struct se_cmd *cmd)
struct se_session *se_sess = cmd->se_sess;
struct se_subsystem_dev *su_dev = cmd->se_dev->se_sub_dev;
struct t10_pr_registration *pr_reg;
- struct t10_reservation_template *pr_tmpl = &su_dev->t10_reservation;
- unsigned char *cdb = &T_TASK(cmd)->t_task_cdb[0];
- int crh = (T10_RES(su_dev)->res_type == SPC3_PERSISTENT_RESERVATIONS);
+ struct t10_reservation *pr_tmpl = &su_dev->t10_pr;
+ unsigned char *cdb = &cmd->t_task->t_task_cdb[0];
+ int crh = (su_dev->t10_pr.res_type == SPC3_PERSISTENT_RESERVATIONS);
int conflict = 0;
if (!(se_sess))
@@ -307,7 +307,7 @@ static int core_scsi3_pr_seq_non_holder(
u32 pr_reg_type)
{
struct se_dev_entry *se_deve;
- struct se_session *se_sess = SE_SESS(cmd);
+ struct se_session *se_sess = cmd->se_sess;
int other_cdb = 0, ignore_reg;
int registered_nexus = 0, ret = 1; /* Conflict by default */
int all_reg = 0, reg_only = 0; /* ALL_REG, REG_ONLY */
@@ -549,7 +549,7 @@ static int core_scsi3_pr_seq_non_holder(
static u32 core_scsi3_pr_generation(struct se_device *dev)
{
- struct se_subsystem_dev *su_dev = SU_DEV(dev);
+ struct se_subsystem_dev *su_dev = dev->se_sub_dev;
u32 prg;
/*
* PRGeneration field shall contain the value of a 32-bit wrapping
@@ -561,7 +561,7 @@ static u32 core_scsi3_pr_generation(struct se_device *dev)
* See spc4r17 section 6.3.12 READ_KEYS service action
*/
spin_lock(&dev->dev_reservation_lock);
- prg = T10_RES(su_dev)->pr_generation++;
+ prg = su_dev->t10_pr.pr_generation++;
spin_unlock(&dev->dev_reservation_lock);
return prg;
@@ -620,7 +620,7 @@ static struct t10_pr_registration *__core_scsi3_do_alloc_registration(
int all_tg_pt,
int aptpl)
{
- struct se_subsystem_dev *su_dev = SU_DEV(dev);
+ struct se_subsystem_dev *su_dev = dev->se_sub_dev;
struct t10_pr_registration *pr_reg;
pr_reg = kmem_cache_zalloc(t10_pr_reg_cache, GFP_ATOMIC);
@@ -629,7 +629,7 @@ static struct t10_pr_registration *__core_scsi3_do_alloc_registration(
return NULL;
}
- pr_reg->pr_aptpl_buf = kzalloc(T10_RES(su_dev)->pr_aptpl_buf_len,
+ pr_reg->pr_aptpl_buf = kzalloc(su_dev->t10_pr.pr_aptpl_buf_len,
GFP_ATOMIC);
if (!(pr_reg->pr_aptpl_buf)) {
printk(KERN_ERR "Unable to allocate pr_reg->pr_aptpl_buf\n");
@@ -803,7 +803,7 @@ out:
}
int core_scsi3_alloc_aptpl_registration(
- struct t10_reservation_template *pr_tmpl,
+ struct t10_reservation *pr_tmpl,
u64 sa_res_key,
unsigned char *i_port,
unsigned char *isid,
@@ -893,11 +893,11 @@ static void core_scsi3_aptpl_reserve(
printk(KERN_INFO "SPC-3 PR [%s] Service Action: APTPL RESERVE created"
" new reservation holder TYPE: %s ALL_TG_PT: %d\n",
- TPG_TFO(tpg)->get_fabric_name(),
+ tpg->se_tpg_tfo->get_fabric_name(),
core_scsi3_pr_dump_type(pr_reg->pr_res_type),
(pr_reg->pr_reg_all_tg_pt) ? 1 : 0);
printk(KERN_INFO "SPC-3 PR [%s] RESERVE Node: %s%s\n",
- TPG_TFO(tpg)->get_fabric_name(), node_acl->initiatorname,
+ tpg->se_tpg_tfo->get_fabric_name(), node_acl->initiatorname,
(prf_isid) ? &i_buf[0] : "");
}
@@ -913,7 +913,7 @@ static int __core_scsi3_check_aptpl_registration(
struct se_dev_entry *deve)
{
struct t10_pr_registration *pr_reg, *pr_reg_tmp;
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
unsigned char i_port[PR_APTPL_MAX_IPORT_LEN];
unsigned char t_port[PR_APTPL_MAX_TPORT_LEN];
u16 tpgt;
@@ -925,8 +925,8 @@ static int __core_scsi3_check_aptpl_registration(
*/
snprintf(i_port, PR_APTPL_MAX_IPORT_LEN, "%s", nacl->initiatorname);
snprintf(t_port, PR_APTPL_MAX_TPORT_LEN, "%s",
- TPG_TFO(tpg)->tpg_get_wwn(tpg));
- tpgt = TPG_TFO(tpg)->tpg_get_tag(tpg);
+ tpg->se_tpg_tfo->tpg_get_wwn(tpg));
+ tpgt = tpg->se_tpg_tfo->tpg_get_tag(tpg);
/*
* Look for the matching registrations+reservation from those
* created from APTPL metadata. Note that multiple registrations
@@ -980,11 +980,11 @@ int core_scsi3_check_aptpl_registration(
struct se_lun *lun,
struct se_lun_acl *lun_acl)
{
- struct se_subsystem_dev *su_dev = SU_DEV(dev);
+ struct se_subsystem_dev *su_dev = dev->se_sub_dev;
struct se_node_acl *nacl = lun_acl->se_lun_nacl;
struct se_dev_entry *deve = &nacl->device_list[lun_acl->mapped_lun];
- if (T10_RES(su_dev)->res_type != SPC3_PERSISTENT_RESERVATIONS)
+ if (su_dev->t10_pr.res_type != SPC3_PERSISTENT_RESERVATIONS)
return 0;
return __core_scsi3_check_aptpl_registration(dev, tpg, lun,
@@ -1017,7 +1017,7 @@ static void __core_scsi3_dump_registration(
printk(KERN_INFO "SPC-3 PR [%s] for %s TCM Subsystem %s Object Target"
" Port(s)\n", tfo->get_fabric_name(),
(pr_reg->pr_reg_all_tg_pt) ? "ALL" : "SINGLE",
- TRANSPORT(dev)->name);
+ dev->transport->name);
printk(KERN_INFO "SPC-3 PR [%s] SA Res Key: 0x%016Lx PRgeneration:"
" 0x%08x APTPL: %d\n", tfo->get_fabric_name(),
pr_reg->pr_res_key, pr_reg->pr_res_generation,
@@ -1035,10 +1035,10 @@ static void __core_scsi3_add_registration(
int register_type,
int register_move)
{
- struct se_subsystem_dev *su_dev = SU_DEV(dev);
+ struct se_subsystem_dev *su_dev = dev->se_sub_dev;
struct target_core_fabric_ops *tfo = nacl->se_tpg->se_tpg_tfo;
struct t10_pr_registration *pr_reg_tmp, *pr_reg_tmp_safe;
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
/*
* Increment PRgeneration counter for struct se_device upon a successful
@@ -1050,7 +1050,7 @@ static void __core_scsi3_add_registration(
* for the REGISTER.
*/
pr_reg->pr_res_generation = (register_move) ?
- T10_RES(su_dev)->pr_generation++ :
+ su_dev->t10_pr.pr_generation++ :
core_scsi3_pr_generation(dev);
spin_lock(&pr_tmpl->registration_lock);
@@ -1119,7 +1119,7 @@ static struct t10_pr_registration *__core_scsi3_locate_pr_reg(
struct se_node_acl *nacl,
unsigned char *isid)
{
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
struct t10_pr_registration *pr_reg, *pr_reg_tmp;
struct se_portal_group *tpg;
@@ -1143,8 +1143,8 @@ static struct t10_pr_registration *__core_scsi3_locate_pr_reg(
* SCSI Intiatior TransportID w/ ISIDs is enforced
* for fabric modules (iSCSI) requiring them.
*/
- if (TPG_TFO(tpg)->sess_get_initiator_sid != NULL) {
- if (DEV_ATTRIB(dev)->enforce_pr_isids)
+ if (tpg->se_tpg_tfo->sess_get_initiator_sid != NULL) {
+ if (dev->se_sub_dev->se_dev_attrib.enforce_pr_isids)
continue;
}
atomic_inc(&pr_reg->pr_res_holders);
@@ -1180,9 +1180,9 @@ static struct t10_pr_registration *core_scsi3_locate_pr_reg(
struct se_portal_group *tpg = nacl->se_tpg;
unsigned char buf[PR_REG_ISID_LEN], *isid_ptr = NULL;
- if (TPG_TFO(tpg)->sess_get_initiator_sid != NULL) {
+ if (tpg->se_tpg_tfo->sess_get_initiator_sid != NULL) {
memset(&buf[0], 0, PR_REG_ISID_LEN);
- TPG_TFO(tpg)->sess_get_initiator_sid(sess, &buf[0],
+ tpg->se_tpg_tfo->sess_get_initiator_sid(sess, &buf[0],
PR_REG_ISID_LEN);
isid_ptr = &buf[0];
}
@@ -1248,7 +1248,7 @@ static int core_scsi3_check_implict_release(
}
/*
- * Called with struct t10_reservation_template->registration_lock held.
+ * Called with struct t10_reservation->registration_lock held.
*/
static void __core_scsi3_free_registration(
struct se_device *dev,
@@ -1258,7 +1258,7 @@ static void __core_scsi3_free_registration(
{
struct target_core_fabric_ops *tfo =
pr_reg->pr_reg_nacl->se_tpg->se_tpg_tfo;
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
char i_buf[PR_REG_ISID_ID_LEN];
int prf_isid;
@@ -1296,7 +1296,7 @@ static void __core_scsi3_free_registration(
printk(KERN_INFO "SPC-3 PR [%s] for %s TCM Subsystem %s Object Target"
" Port(s)\n", tfo->get_fabric_name(),
(pr_reg->pr_reg_all_tg_pt) ? "ALL" : "SINGLE",
- TRANSPORT(dev)->name);
+ dev->transport->name);
printk(KERN_INFO "SPC-3 PR [%s] SA Res Key: 0x%016Lx PRgeneration:"
" 0x%08x\n", tfo->get_fabric_name(), pr_reg->pr_res_key,
pr_reg->pr_res_generation);
@@ -1319,7 +1319,7 @@ void core_scsi3_free_pr_reg_from_nacl(
struct se_device *dev,
struct se_node_acl *nacl)
{
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
struct t10_pr_registration *pr_reg, *pr_reg_tmp, *pr_res_holder;
/*
* If the passed se_node_acl matches the reservation holder,
@@ -1349,7 +1349,7 @@ void core_scsi3_free_pr_reg_from_nacl(
void core_scsi3_free_all_registrations(
struct se_device *dev)
{
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
struct t10_pr_registration *pr_reg, *pr_reg_tmp, *pr_res_holder;
spin_lock(&dev->dev_reservation_lock);
@@ -1381,13 +1381,13 @@ void core_scsi3_free_all_registrations(
static int core_scsi3_tpg_depend_item(struct se_portal_group *tpg)
{
- return configfs_depend_item(TPG_TFO(tpg)->tf_subsys,
+ return configfs_depend_item(tpg->se_tpg_tfo->tf_subsys,
&tpg->tpg_group.cg_item);
}
static void core_scsi3_tpg_undepend_item(struct se_portal_group *tpg)
{
- configfs_undepend_item(TPG_TFO(tpg)->tf_subsys,
+ configfs_undepend_item(tpg->se_tpg_tfo->tf_subsys,
&tpg->tpg_group.cg_item);
atomic_dec(&tpg->tpg_pr_ref_count);
@@ -1401,7 +1401,7 @@ static int core_scsi3_nodeacl_depend_item(struct se_node_acl *nacl)
if (nacl->dynamic_node_acl)
return 0;
- return configfs_depend_item(TPG_TFO(tpg)->tf_subsys,
+ return configfs_depend_item(tpg->se_tpg_tfo->tf_subsys,
&nacl->acl_group.cg_item);
}
@@ -1415,7 +1415,7 @@ static void core_scsi3_nodeacl_undepend_item(struct se_node_acl *nacl)
return;
}
- configfs_undepend_item(TPG_TFO(tpg)->tf_subsys,
+ configfs_undepend_item(tpg->se_tpg_tfo->tf_subsys,
&nacl->acl_group.cg_item);
atomic_dec(&nacl->acl_pr_ref_count);
@@ -1436,7 +1436,7 @@ static int core_scsi3_lunacl_depend_item(struct se_dev_entry *se_deve)
nacl = lun_acl->se_lun_nacl;
tpg = nacl->se_tpg;
- return configfs_depend_item(TPG_TFO(tpg)->tf_subsys,
+ return configfs_depend_item(tpg->se_tpg_tfo->tf_subsys,
&lun_acl->se_lun_group.cg_item);
}
@@ -1456,7 +1456,7 @@ static void core_scsi3_lunacl_undepend_item(struct se_dev_entry *se_deve)
nacl = lun_acl->se_lun_nacl;
tpg = nacl->se_tpg;
- configfs_undepend_item(TPG_TFO(tpg)->tf_subsys,
+ configfs_undepend_item(tpg->se_tpg_tfo->tf_subsys,
&lun_acl->se_lun_group.cg_item);
atomic_dec(&se_deve->pr_ref_count);
@@ -1471,10 +1471,10 @@ static int core_scsi3_decode_spec_i_port(
int all_tg_pt,
int aptpl)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
struct se_port *tmp_port;
struct se_portal_group *dest_tpg = NULL, *tmp_tpg;
- struct se_session *se_sess = SE_SESS(cmd);
+ struct se_session *se_sess = cmd->se_sess;
struct se_node_acl *dest_node_acl = NULL;
struct se_dev_entry *dest_se_deve = NULL, *local_se_deve;
struct t10_pr_registration *dest_pr_reg, *local_pr_reg, *pr_reg_e;
@@ -1482,7 +1482,7 @@ static int core_scsi3_decode_spec_i_port(
struct list_head tid_dest_list;
struct pr_transport_id_holder *tidh_new, *tidh, *tidh_tmp;
struct target_core_fabric_ops *tmp_tf_ops;
- unsigned char *buf = (unsigned char *)T_TASK(cmd)->t_task_buf;
+ unsigned char *buf = (unsigned char *)cmd->t_task->t_task_buf;
unsigned char *ptr, *i_str = NULL, proto_ident, tmp_proto_ident;
char *iport_ptr = NULL, dest_iport[64], i_buf[PR_REG_ISID_ID_LEN];
u32 tpdl, tid_len = 0;
@@ -1509,7 +1509,7 @@ static int core_scsi3_decode_spec_i_port(
tidh_new->dest_node_acl = se_sess->se_node_acl;
tidh_new->dest_se_deve = local_se_deve;
- local_pr_reg = __core_scsi3_alloc_registration(SE_DEV(cmd),
+ local_pr_reg = __core_scsi3_alloc_registration(cmd->se_lun->lun_se_dev,
se_sess->se_node_acl, local_se_deve, l_isid,
sa_res_key, all_tg_pt, aptpl);
if (!(local_pr_reg)) {
@@ -1557,7 +1557,7 @@ static int core_scsi3_decode_spec_i_port(
tmp_tpg = tmp_port->sep_tpg;
if (!(tmp_tpg))
continue;
- tmp_tf_ops = TPG_TFO(tmp_tpg);
+ tmp_tf_ops = tmp_tpg->se_tpg_tfo;
if (!(tmp_tf_ops))
continue;
if (!(tmp_tf_ops->get_fabric_proto_ident) ||
@@ -1625,7 +1625,7 @@ static int core_scsi3_decode_spec_i_port(
dest_tpg = tmp_tpg;
printk(KERN_INFO "SPC-3 PR SPEC_I_PT: Located %s Node:"
" %s Port RTPI: %hu\n",
- TPG_TFO(dest_tpg)->get_fabric_name(),
+ dest_tpg->se_tpg_tfo->get_fabric_name(),
dest_node_acl->initiatorname, dest_rtpi);
spin_lock(&dev->se_port_lock);
@@ -1642,7 +1642,7 @@ static int core_scsi3_decode_spec_i_port(
#if 0
printk("SPC-3 PR SPEC_I_PT: Got %s data_length: %u tpdl: %u"
" tid_len: %d for %s + %s\n",
- TPG_TFO(dest_tpg)->get_fabric_name(), cmd->data_length,
+ dest_tpg->se_tpg_tfo->get_fabric_name(), cmd->data_length,
tpdl, tid_len, i_str, iport_ptr);
#endif
if (tid_len > tpdl) {
@@ -1663,7 +1663,7 @@ static int core_scsi3_decode_spec_i_port(
if (!(dest_se_deve)) {
printk(KERN_ERR "Unable to locate %s dest_se_deve"
" from destination RTPI: %hu\n",
- TPG_TFO(dest_tpg)->get_fabric_name(),
+ dest_tpg->se_tpg_tfo->get_fabric_name(),
dest_rtpi);
core_scsi3_nodeacl_undepend_item(dest_node_acl);
@@ -1686,7 +1686,7 @@ static int core_scsi3_decode_spec_i_port(
#if 0
printk(KERN_INFO "SPC-3 PR SPEC_I_PT: Located %s Node: %s"
" dest_se_deve mapped_lun: %u\n",
- TPG_TFO(dest_tpg)->get_fabric_name(),
+ dest_tpg->se_tpg_tfo->get_fabric_name(),
dest_node_acl->initiatorname, dest_se_deve->mapped_lun);
#endif
/*
@@ -1741,7 +1741,7 @@ static int core_scsi3_decode_spec_i_port(
* and then call __core_scsi3_add_registration() in the
* 2nd loop which will never fail.
*/
- dest_pr_reg = __core_scsi3_alloc_registration(SE_DEV(cmd),
+ dest_pr_reg = __core_scsi3_alloc_registration(cmd->se_lun->lun_se_dev,
dest_node_acl, dest_se_deve, iport_ptr,
sa_res_key, all_tg_pt, aptpl);
if (!(dest_pr_reg)) {
@@ -1787,12 +1787,12 @@ static int core_scsi3_decode_spec_i_port(
prf_isid = core_pr_dump_initiator_port(dest_pr_reg, &i_buf[0],
PR_REG_ISID_ID_LEN);
- __core_scsi3_add_registration(SE_DEV(cmd), dest_node_acl,
+ __core_scsi3_add_registration(cmd->se_lun->lun_se_dev, dest_node_acl,
dest_pr_reg, 0, 0);
printk(KERN_INFO "SPC-3 PR [%s] SPEC_I_PT: Successfully"
" registered Transport ID for Node: %s%s Mapped LUN:"
- " %u\n", TPG_TFO(dest_tpg)->get_fabric_name(),
+ " %u\n", dest_tpg->se_tpg_tfo->get_fabric_name(),
dest_node_acl->initiatorname, (prf_isid) ?
&i_buf[0] : "", dest_se_deve->mapped_lun);
@@ -1855,7 +1855,7 @@ static int __core_scsi3_update_aptpl_buf(
{
struct se_lun *lun;
struct se_portal_group *tpg;
- struct se_subsystem_dev *su_dev = SU_DEV(dev);
+ struct se_subsystem_dev *su_dev = dev->se_sub_dev;
struct t10_pr_registration *pr_reg;
unsigned char tmp[512], isid_buf[32];
ssize_t len = 0;
@@ -1873,8 +1873,8 @@ static int __core_scsi3_update_aptpl_buf(
/*
* Walk the registration list..
*/
- spin_lock(&T10_RES(su_dev)->registration_lock);
- list_for_each_entry(pr_reg, &T10_RES(su_dev)->registration_list,
+ spin_lock(&su_dev->t10_pr.registration_lock);
+ list_for_each_entry(pr_reg, &su_dev->t10_pr.registration_list,
pr_reg_list) {
tmp[0] = '\0';
@@ -1900,7 +1900,7 @@ static int __core_scsi3_update_aptpl_buf(
"res_holder=1\nres_type=%02x\n"
"res_scope=%02x\nres_all_tg_pt=%d\n"
"mapped_lun=%u\n", reg_count,
- TPG_TFO(tpg)->get_fabric_name(),
+ tpg->se_tpg_tfo->get_fabric_name(),
pr_reg->pr_reg_nacl->initiatorname, isid_buf,
pr_reg->pr_res_key, pr_reg->pr_res_type,
pr_reg->pr_res_scope, pr_reg->pr_reg_all_tg_pt,
@@ -1910,7 +1910,7 @@ static int __core_scsi3_update_aptpl_buf(
"initiator_fabric=%s\ninitiator_node=%s\n%s"
"sa_res_key=%llu\nres_holder=0\n"
"res_all_tg_pt=%d\nmapped_lun=%u\n",
- reg_count, TPG_TFO(tpg)->get_fabric_name(),
+ reg_count, tpg->se_tpg_tfo->get_fabric_name(),
pr_reg->pr_reg_nacl->initiatorname, isid_buf,
pr_reg->pr_res_key, pr_reg->pr_reg_all_tg_pt,
pr_reg->pr_res_mapped_lun);
@@ -1919,7 +1919,7 @@ static int __core_scsi3_update_aptpl_buf(
if ((len + strlen(tmp) >= pr_aptpl_buf_len)) {
printk(KERN_ERR "Unable to update renaming"
" APTPL metadata\n");
- spin_unlock(&T10_RES(su_dev)->registration_lock);
+ spin_unlock(&su_dev->t10_pr.registration_lock);
return -EMSGSIZE;
}
len += sprintf(buf+len, "%s", tmp);
@@ -1929,21 +1929,21 @@ static int __core_scsi3_update_aptpl_buf(
*/
snprintf(tmp, 512, "target_fabric=%s\ntarget_node=%s\n"
"tpgt=%hu\nport_rtpi=%hu\ntarget_lun=%u\nPR_REG_END:"
- " %d\n", TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_wwn(tpg),
- TPG_TFO(tpg)->tpg_get_tag(tpg),
+ " %d\n", tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_wwn(tpg),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg),
lun->lun_sep->sep_rtpi, lun->unpacked_lun, reg_count);
if ((len + strlen(tmp) >= pr_aptpl_buf_len)) {
printk(KERN_ERR "Unable to update renaming"
" APTPL metadata\n");
- spin_unlock(&T10_RES(su_dev)->registration_lock);
+ spin_unlock(&su_dev->t10_pr.registration_lock);
return -EMSGSIZE;
}
len += sprintf(buf+len, "%s", tmp);
reg_count++;
}
- spin_unlock(&T10_RES(su_dev)->registration_lock);
+ spin_unlock(&su_dev->t10_pr.registration_lock);
if (!(reg_count))
len += sprintf(buf+len, "No Registrations or Reservations");
@@ -1975,7 +1975,7 @@ static int __core_scsi3_write_aptpl_to_file(
unsigned char *buf,
u32 pr_aptpl_buf_len)
{
- struct t10_wwn *wwn = &SU_DEV(dev)->t10_wwn;
+ struct t10_wwn *wwn = &dev->se_sub_dev->t10_wwn;
struct file *file;
struct iovec iov[1];
mm_segment_t old_fs;
@@ -2070,13 +2070,13 @@ static int core_scsi3_emulate_pro_register(
int spec_i_pt,
int ignore_key)
{
- struct se_session *se_sess = SE_SESS(cmd);
- struct se_device *dev = SE_DEV(cmd);
+ struct se_session *se_sess = cmd->se_sess;
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
struct se_dev_entry *se_deve;
- struct se_lun *se_lun = SE_LUN(cmd);
+ struct se_lun *se_lun = cmd->se_lun;
struct se_portal_group *se_tpg;
struct t10_pr_registration *pr_reg, *pr_reg_p, *pr_reg_tmp, *pr_reg_e;
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
/* Used for APTPL metadata w/ UNREGISTER */
unsigned char *pr_aptpl_buf = NULL;
unsigned char isid_buf[PR_REG_ISID_LEN], *isid_ptr = NULL;
@@ -2089,9 +2089,9 @@ static int core_scsi3_emulate_pro_register(
se_tpg = se_sess->se_tpg;
se_deve = &se_sess->se_node_acl->device_list[cmd->orig_fe_lun];
- if (TPG_TFO(se_tpg)->sess_get_initiator_sid != NULL) {
+ if (se_tpg->se_tpg_tfo->sess_get_initiator_sid) {
memset(&isid_buf[0], 0, PR_REG_ISID_LEN);
- TPG_TFO(se_tpg)->sess_get_initiator_sid(se_sess, &isid_buf[0],
+ se_tpg->se_tpg_tfo->sess_get_initiator_sid(se_sess, &isid_buf[0],
PR_REG_ISID_LEN);
isid_ptr = &isid_buf[0];
}
@@ -2117,7 +2117,7 @@ static int core_scsi3_emulate_pro_register(
* Port Endpoint that the PRO was received from on the
* Logical Unit of the SCSI device server.
*/
- ret = core_scsi3_alloc_registration(SE_DEV(cmd),
+ ret = core_scsi3_alloc_registration(cmd->se_lun->lun_se_dev,
se_sess->se_node_acl, se_deve, isid_ptr,
sa_res_key, all_tg_pt, aptpl,
ignore_key, 0);
@@ -2145,7 +2145,7 @@ static int core_scsi3_emulate_pro_register(
*/
if (!(aptpl)) {
pr_tmpl->pr_aptpl_active = 0;
- core_scsi3_update_and_write_aptpl(SE_DEV(cmd), NULL, 0);
+ core_scsi3_update_and_write_aptpl(cmd->se_lun->lun_se_dev, NULL, 0);
printk("SPC-3 PR: Set APTPL Bit Deactivated for"
" REGISTER\n");
return 0;
@@ -2155,10 +2155,10 @@ static int core_scsi3_emulate_pro_register(
* update the APTPL metadata information using its
* preallocated *pr_reg->pr_aptpl_buf.
*/
- pr_reg = core_scsi3_locate_pr_reg(SE_DEV(cmd),
+ pr_reg = core_scsi3_locate_pr_reg(cmd->se_lun->lun_se_dev,
se_sess->se_node_acl, se_sess);
- ret = core_scsi3_update_and_write_aptpl(SE_DEV(cmd),
+ ret = core_scsi3_update_and_write_aptpl(cmd->se_lun->lun_se_dev,
&pr_reg->pr_aptpl_buf[0],
pr_tmpl->pr_aptpl_buf_len);
if (!(ret)) {
@@ -2223,7 +2223,7 @@ static int core_scsi3_emulate_pro_register(
*/
if (!(sa_res_key)) {
pr_holder = core_scsi3_check_implict_release(
- SE_DEV(cmd), pr_reg);
+ cmd->se_lun->lun_se_dev, pr_reg);
if (pr_holder < 0) {
kfree(pr_aptpl_buf);
core_scsi3_put_pr_reg(pr_reg);
@@ -2260,7 +2260,7 @@ static int core_scsi3_emulate_pro_register(
/*
* Release the calling I_T Nexus registration now..
*/
- __core_scsi3_free_registration(SE_DEV(cmd), pr_reg,
+ __core_scsi3_free_registration(cmd->se_lun->lun_se_dev, pr_reg,
NULL, 1);
/*
* From spc4r17, section 5.7.11.3 Unregistering
@@ -2315,11 +2315,11 @@ static int core_scsi3_emulate_pro_register(
* READ_KEYS service action.
*/
pr_reg->pr_res_generation = core_scsi3_pr_generation(
- SE_DEV(cmd));
+ cmd->se_lun->lun_se_dev);
pr_reg->pr_res_key = sa_res_key;
printk("SPC-3 PR [%s] REGISTER%s: Changed Reservation"
" Key for %s to: 0x%016Lx PRgeneration:"
- " 0x%08x\n", CMD_TFO(cmd)->get_fabric_name(),
+ " 0x%08x\n", cmd->se_tfo->get_fabric_name(),
(ignore_key) ? "_AND_IGNORE_EXISTING_KEY" : "",
pr_reg->pr_reg_nacl->initiatorname,
pr_reg->pr_res_key, pr_reg->pr_res_generation);
@@ -2378,12 +2378,12 @@ static int core_scsi3_pro_reserve(
int scope,
u64 res_key)
{
- struct se_session *se_sess = SE_SESS(cmd);
+ struct se_session *se_sess = cmd->se_sess;
struct se_dev_entry *se_deve;
- struct se_lun *se_lun = SE_LUN(cmd);
+ struct se_lun *se_lun = cmd->se_lun;
struct se_portal_group *se_tpg;
struct t10_pr_registration *pr_reg, *pr_res_holder;
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
char i_buf[PR_REG_ISID_ID_LEN];
int ret, prf_isid;
@@ -2398,7 +2398,7 @@ static int core_scsi3_pro_reserve(
/*
* Locate the existing *pr_reg via struct se_node_acl pointers
*/
- pr_reg = core_scsi3_locate_pr_reg(SE_DEV(cmd), se_sess->se_node_acl,
+ pr_reg = core_scsi3_locate_pr_reg(cmd->se_lun->lun_se_dev, se_sess->se_node_acl,
se_sess);
if (!(pr_reg)) {
printk(KERN_ERR "SPC-3 PR: Unable to locate"
@@ -2459,9 +2459,9 @@ static int core_scsi3_pro_reserve(
printk(KERN_ERR "SPC-3 PR: Attempted RESERVE from"
" [%s]: %s while reservation already held by"
" [%s]: %s, returning RESERVATION_CONFLICT\n",
- CMD_TFO(cmd)->get_fabric_name(),
+ cmd->se_tfo->get_fabric_name(),
se_sess->se_node_acl->initiatorname,
- TPG_TFO(pr_res_nacl->se_tpg)->get_fabric_name(),
+ pr_res_nacl->se_tpg->se_tpg_tfo->get_fabric_name(),
pr_res_holder->pr_reg_nacl->initiatorname);
spin_unlock(&dev->dev_reservation_lock);
@@ -2482,9 +2482,9 @@ static int core_scsi3_pro_reserve(
" [%s]: %s trying to change TYPE and/or SCOPE,"
" while reservation already held by [%s]: %s,"
" returning RESERVATION_CONFLICT\n",
- CMD_TFO(cmd)->get_fabric_name(),
+ cmd->se_tfo->get_fabric_name(),
se_sess->se_node_acl->initiatorname,
- TPG_TFO(pr_res_nacl->se_tpg)->get_fabric_name(),
+ pr_res_nacl->se_tpg->se_tpg_tfo->get_fabric_name(),
pr_res_holder->pr_reg_nacl->initiatorname);
spin_unlock(&dev->dev_reservation_lock);
@@ -2518,16 +2518,16 @@ static int core_scsi3_pro_reserve(
printk(KERN_INFO "SPC-3 PR [%s] Service Action: RESERVE created new"
" reservation holder TYPE: %s ALL_TG_PT: %d\n",
- CMD_TFO(cmd)->get_fabric_name(), core_scsi3_pr_dump_type(type),
+ cmd->se_tfo->get_fabric_name(), core_scsi3_pr_dump_type(type),
(pr_reg->pr_reg_all_tg_pt) ? 1 : 0);
printk(KERN_INFO "SPC-3 PR [%s] RESERVE Node: %s%s\n",
- CMD_TFO(cmd)->get_fabric_name(),
+ cmd->se_tfo->get_fabric_name(),
se_sess->se_node_acl->initiatorname,
(prf_isid) ? &i_buf[0] : "");
spin_unlock(&dev->dev_reservation_lock);
if (pr_tmpl->pr_aptpl_active) {
- ret = core_scsi3_update_and_write_aptpl(SE_DEV(cmd),
+ ret = core_scsi3_update_and_write_aptpl(cmd->se_lun->lun_se_dev,
&pr_reg->pr_aptpl_buf[0],
pr_tmpl->pr_aptpl_buf_len);
if (!(ret))
@@ -2608,10 +2608,10 @@ static int core_scsi3_emulate_pro_release(
u64 res_key)
{
struct se_device *dev = cmd->se_dev;
- struct se_session *se_sess = SE_SESS(cmd);
- struct se_lun *se_lun = SE_LUN(cmd);
+ struct se_session *se_sess = cmd->se_sess;
+ struct se_lun *se_lun = cmd->se_lun;
struct t10_pr_registration *pr_reg, *pr_reg_p, *pr_res_holder;
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
int ret, all_reg = 0;
if (!(se_sess) || !(se_lun)) {
@@ -2698,9 +2698,9 @@ static int core_scsi3_emulate_pro_release(
" reservation from [%s]: %s with different TYPE "
"and/or SCOPE while reservation already held by"
" [%s]: %s, returning RESERVATION_CONFLICT\n",
- CMD_TFO(cmd)->get_fabric_name(),
+ cmd->se_tfo->get_fabric_name(),
se_sess->se_node_acl->initiatorname,
- TPG_TFO(pr_res_nacl->se_tpg)->get_fabric_name(),
+ pr_res_nacl->se_tpg->se_tpg_tfo->get_fabric_name(),
pr_res_holder->pr_reg_nacl->initiatorname);
spin_unlock(&dev->dev_reservation_lock);
@@ -2758,7 +2758,7 @@ static int core_scsi3_emulate_pro_release(
write_aptpl:
if (pr_tmpl->pr_aptpl_active) {
- ret = core_scsi3_update_and_write_aptpl(SE_DEV(cmd),
+ ret = core_scsi3_update_and_write_aptpl(cmd->se_lun->lun_se_dev,
&pr_reg->pr_aptpl_buf[0],
pr_tmpl->pr_aptpl_buf_len);
if (!(ret))
@@ -2775,15 +2775,15 @@ static int core_scsi3_emulate_pro_clear(
{
struct se_device *dev = cmd->se_dev;
struct se_node_acl *pr_reg_nacl;
- struct se_session *se_sess = SE_SESS(cmd);
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
+ struct se_session *se_sess = cmd->se_sess;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
struct t10_pr_registration *pr_reg, *pr_reg_tmp, *pr_reg_n, *pr_res_holder;
u32 pr_res_mapped_lun = 0;
int calling_it_nexus = 0;
/*
* Locate the existing *pr_reg via struct se_node_acl pointers
*/
- pr_reg_n = core_scsi3_locate_pr_reg(SE_DEV(cmd),
+ pr_reg_n = core_scsi3_locate_pr_reg(cmd->se_lun->lun_se_dev,
se_sess->se_node_acl, se_sess);
if (!(pr_reg_n)) {
printk(KERN_ERR "SPC-3 PR: Unable to locate"
@@ -2846,10 +2846,10 @@ static int core_scsi3_emulate_pro_clear(
spin_unlock(&pr_tmpl->registration_lock);
printk(KERN_INFO "SPC-3 PR [%s] Service Action: CLEAR complete\n",
- CMD_TFO(cmd)->get_fabric_name());
+ cmd->se_tfo->get_fabric_name());
if (pr_tmpl->pr_aptpl_active) {
- core_scsi3_update_and_write_aptpl(SE_DEV(cmd), NULL, 0);
+ core_scsi3_update_and_write_aptpl(cmd->se_lun->lun_se_dev, NULL, 0);
printk(KERN_INFO "SPC-3 PR: Updated APTPL metadata"
" for CLEAR\n");
}
@@ -2954,13 +2954,13 @@ static int core_scsi3_pro_preempt(
u64 sa_res_key,
int abort)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
struct se_dev_entry *se_deve;
struct se_node_acl *pr_reg_nacl;
- struct se_session *se_sess = SE_SESS(cmd);
+ struct se_session *se_sess = cmd->se_sess;
struct list_head preempt_and_abort_list;
struct t10_pr_registration *pr_reg, *pr_reg_tmp, *pr_reg_n, *pr_res_holder;
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
u32 pr_res_mapped_lun = 0;
int all_reg = 0, calling_it_nexus = 0, released_regs = 0;
int prh_type = 0, prh_scope = 0, ret;
@@ -2969,7 +2969,7 @@ static int core_scsi3_pro_preempt(
return PYX_TRANSPORT_LU_COMM_FAILURE;
se_deve = &se_sess->se_node_acl->device_list[cmd->orig_fe_lun];
- pr_reg_n = core_scsi3_locate_pr_reg(SE_DEV(cmd), se_sess->se_node_acl,
+ pr_reg_n = core_scsi3_locate_pr_reg(cmd->se_lun->lun_se_dev, se_sess->se_node_acl,
se_sess);
if (!(pr_reg_n)) {
printk(KERN_ERR "SPC-3 PR: Unable to locate"
@@ -3111,7 +3111,7 @@ static int core_scsi3_pro_preempt(
spin_unlock(&dev->dev_reservation_lock);
if (pr_tmpl->pr_aptpl_active) {
- ret = core_scsi3_update_and_write_aptpl(SE_DEV(cmd),
+ ret = core_scsi3_update_and_write_aptpl(cmd->se_lun->lun_se_dev,
&pr_reg_n->pr_aptpl_buf[0],
pr_tmpl->pr_aptpl_buf_len);
if (!(ret))
@@ -3121,7 +3121,7 @@ static int core_scsi3_pro_preempt(
}
core_scsi3_put_pr_reg(pr_reg_n);
- core_scsi3_pr_generation(SE_DEV(cmd));
+ core_scsi3_pr_generation(cmd->se_lun->lun_se_dev);
return 0;
}
/*
@@ -3247,7 +3247,7 @@ static int core_scsi3_pro_preempt(
}
if (pr_tmpl->pr_aptpl_active) {
- ret = core_scsi3_update_and_write_aptpl(SE_DEV(cmd),
+ ret = core_scsi3_update_and_write_aptpl(cmd->se_lun->lun_se_dev,
&pr_reg_n->pr_aptpl_buf[0],
pr_tmpl->pr_aptpl_buf_len);
if (!(ret))
@@ -3256,7 +3256,7 @@ static int core_scsi3_pro_preempt(
}
core_scsi3_put_pr_reg(pr_reg_n);
- core_scsi3_pr_generation(SE_DEV(cmd));
+ core_scsi3_pr_generation(cmd->se_lun->lun_se_dev);
return 0;
}
@@ -3297,17 +3297,17 @@ static int core_scsi3_emulate_pro_register_and_move(
int aptpl,
int unreg)
{
- struct se_session *se_sess = SE_SESS(cmd);
- struct se_device *dev = SE_DEV(cmd);
+ struct se_session *se_sess = cmd->se_sess;
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
struct se_dev_entry *se_deve, *dest_se_deve = NULL;
- struct se_lun *se_lun = SE_LUN(cmd);
+ struct se_lun *se_lun = cmd->se_lun;
struct se_node_acl *pr_res_nacl, *pr_reg_nacl, *dest_node_acl = NULL;
struct se_port *se_port;
struct se_portal_group *se_tpg, *dest_se_tpg = NULL;
struct target_core_fabric_ops *dest_tf_ops = NULL, *tf_ops;
struct t10_pr_registration *pr_reg, *pr_res_holder, *dest_pr_reg;
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
- unsigned char *buf = (unsigned char *)T_TASK(cmd)->t_task_buf;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
+ unsigned char *buf = (unsigned char *)cmd->t_task->t_task_buf;
unsigned char *initiator_str;
char *iport_ptr = NULL, dest_iport[64], i_buf[PR_REG_ISID_ID_LEN];
u32 tid_len, tmp_tid_len;
@@ -3322,7 +3322,7 @@ static int core_scsi3_emulate_pro_register_and_move(
memset(dest_iport, 0, 64);
memset(i_buf, 0, PR_REG_ISID_ID_LEN);
se_tpg = se_sess->se_tpg;
- tf_ops = TPG_TFO(se_tpg);
+ tf_ops = se_tpg->se_tpg_tfo;
se_deve = &se_sess->se_node_acl->device_list[cmd->orig_fe_lun];
/*
* Follow logic from spc4r17 Section 5.7.8, Table 50 --
@@ -3330,7 +3330,7 @@ static int core_scsi3_emulate_pro_register_and_move(
*
* Locate the existing *pr_reg via struct se_node_acl pointers
*/
- pr_reg = core_scsi3_locate_pr_reg(SE_DEV(cmd), se_sess->se_node_acl,
+ pr_reg = core_scsi3_locate_pr_reg(cmd->se_lun->lun_se_dev, se_sess->se_node_acl,
se_sess);
if (!(pr_reg)) {
printk(KERN_ERR "SPC-3 PR: Unable to locate PR_REGISTERED"
@@ -3384,7 +3384,7 @@ static int core_scsi3_emulate_pro_register_and_move(
dest_se_tpg = se_port->sep_tpg;
if (!(dest_se_tpg))
continue;
- dest_tf_ops = TPG_TFO(dest_se_tpg);
+ dest_tf_ops = dest_se_tpg->se_tpg_tfo;
if (!(dest_tf_ops))
continue;
@@ -3612,7 +3612,7 @@ after_iport_check:
dest_pr_reg = __core_scsi3_locate_pr_reg(dev, dest_node_acl,
iport_ptr);
if (!(dest_pr_reg)) {
- ret = core_scsi3_alloc_registration(SE_DEV(cmd),
+ ret = core_scsi3_alloc_registration(cmd->se_lun->lun_se_dev,
dest_node_acl, dest_se_deve, iport_ptr,
sa_res_key, 0, aptpl, 2, 1);
if (ret != 0) {
@@ -3683,12 +3683,12 @@ after_iport_check:
*/
if (!(aptpl)) {
pr_tmpl->pr_aptpl_active = 0;
- core_scsi3_update_and_write_aptpl(SE_DEV(cmd), NULL, 0);
+ core_scsi3_update_and_write_aptpl(cmd->se_lun->lun_se_dev, NULL, 0);
printk("SPC-3 PR: Set APTPL Bit Deactivated for"
" REGISTER_AND_MOVE\n");
} else {
pr_tmpl->pr_aptpl_active = 1;
- ret = core_scsi3_update_and_write_aptpl(SE_DEV(cmd),
+ ret = core_scsi3_update_and_write_aptpl(cmd->se_lun->lun_se_dev,
&dest_pr_reg->pr_aptpl_buf[0],
pr_tmpl->pr_aptpl_buf_len);
if (!(ret))
@@ -3723,7 +3723,7 @@ static unsigned long long core_scsi3_extract_reservation_key(unsigned char *cdb)
*/
static int core_scsi3_emulate_pr_out(struct se_cmd *cmd, unsigned char *cdb)
{
- unsigned char *buf = (unsigned char *)T_TASK(cmd)->t_task_buf;
+ unsigned char *buf = (unsigned char *)cmd->t_task->t_task_buf;
u64 res_key, sa_res_key;
int sa, scope, type, aptpl;
int spec_i_pt = 0, all_tg_pt = 0, unreg = 0;
@@ -3731,7 +3731,7 @@ static int core_scsi3_emulate_pr_out(struct se_cmd *cmd, unsigned char *cdb)
* FIXME: A NULL struct se_session pointer means an this is not coming from
* a $FABRIC_MOD's nexus, but from internal passthrough ops.
*/
- if (!(SE_SESS(cmd)))
+ if (!(cmd->se_sess))
return PYX_TRANSPORT_LU_COMM_FAILURE;
if (cmd->data_length < 24) {
@@ -3827,10 +3827,10 @@ static int core_scsi3_emulate_pr_out(struct se_cmd *cmd, unsigned char *cdb)
*/
static int core_scsi3_pri_read_keys(struct se_cmd *cmd)
{
- struct se_device *se_dev = SE_DEV(cmd);
- struct se_subsystem_dev *su_dev = SU_DEV(se_dev);
+ struct se_device *se_dev = cmd->se_lun->lun_se_dev;
+ struct se_subsystem_dev *su_dev = se_dev->se_sub_dev;
struct t10_pr_registration *pr_reg;
- unsigned char *buf = (unsigned char *)T_TASK(cmd)->t_task_buf;
+ unsigned char *buf = (unsigned char *)cmd->t_task->t_task_buf;
u32 add_len = 0, off = 8;
if (cmd->data_length < 8) {
@@ -3839,13 +3839,13 @@ static int core_scsi3_pri_read_keys(struct se_cmd *cmd)
return PYX_TRANSPORT_INVALID_CDB_FIELD;
}
- buf[0] = ((T10_RES(su_dev)->pr_generation >> 24) & 0xff);
- buf[1] = ((T10_RES(su_dev)->pr_generation >> 16) & 0xff);
- buf[2] = ((T10_RES(su_dev)->pr_generation >> 8) & 0xff);
- buf[3] = (T10_RES(su_dev)->pr_generation & 0xff);
+ buf[0] = ((su_dev->t10_pr.pr_generation >> 24) & 0xff);
+ buf[1] = ((su_dev->t10_pr.pr_generation >> 16) & 0xff);
+ buf[2] = ((su_dev->t10_pr.pr_generation >> 8) & 0xff);
+ buf[3] = (su_dev->t10_pr.pr_generation & 0xff);
- spin_lock(&T10_RES(su_dev)->registration_lock);
- list_for_each_entry(pr_reg, &T10_RES(su_dev)->registration_list,
+ spin_lock(&su_dev->t10_pr.registration_lock);
+ list_for_each_entry(pr_reg, &su_dev->t10_pr.registration_list,
pr_reg_list) {
/*
* Check for overflow of 8byte PRI READ_KEYS payload and
@@ -3865,7 +3865,7 @@ static int core_scsi3_pri_read_keys(struct se_cmd *cmd)
add_len += 8;
}
- spin_unlock(&T10_RES(su_dev)->registration_lock);
+ spin_unlock(&su_dev->t10_pr.registration_lock);
buf[4] = ((add_len >> 24) & 0xff);
buf[5] = ((add_len >> 16) & 0xff);
@@ -3882,10 +3882,10 @@ static int core_scsi3_pri_read_keys(struct se_cmd *cmd)
*/
static int core_scsi3_pri_read_reservation(struct se_cmd *cmd)
{
- struct se_device *se_dev = SE_DEV(cmd);
- struct se_subsystem_dev *su_dev = SU_DEV(se_dev);
+ struct se_device *se_dev = cmd->se_lun->lun_se_dev;
+ struct se_subsystem_dev *su_dev = se_dev->se_sub_dev;
struct t10_pr_registration *pr_reg;
- unsigned char *buf = (unsigned char *)T_TASK(cmd)->t_task_buf;
+ unsigned char *buf = (unsigned char *)cmd->t_task->t_task_buf;
u64 pr_res_key;
u32 add_len = 16; /* Hardcoded to 16 when a reservation is held. */
@@ -3895,10 +3895,10 @@ static int core_scsi3_pri_read_reservation(struct se_cmd *cmd)
return PYX_TRANSPORT_INVALID_CDB_FIELD;
}
- buf[0] = ((T10_RES(su_dev)->pr_generation >> 24) & 0xff);
- buf[1] = ((T10_RES(su_dev)->pr_generation >> 16) & 0xff);
- buf[2] = ((T10_RES(su_dev)->pr_generation >> 8) & 0xff);
- buf[3] = (T10_RES(su_dev)->pr_generation & 0xff);
+ buf[0] = ((su_dev->t10_pr.pr_generation >> 24) & 0xff);
+ buf[1] = ((su_dev->t10_pr.pr_generation >> 16) & 0xff);
+ buf[2] = ((su_dev->t10_pr.pr_generation >> 8) & 0xff);
+ buf[3] = (su_dev->t10_pr.pr_generation & 0xff);
spin_lock(&se_dev->dev_reservation_lock);
pr_reg = se_dev->dev_pr_res_holder;
@@ -3963,9 +3963,9 @@ static int core_scsi3_pri_read_reservation(struct se_cmd *cmd)
*/
static int core_scsi3_pri_report_capabilities(struct se_cmd *cmd)
{
- struct se_device *dev = SE_DEV(cmd);
- struct t10_reservation_template *pr_tmpl = &SU_DEV(dev)->t10_reservation;
- unsigned char *buf = (unsigned char *)T_TASK(cmd)->t_task_buf;
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
+ struct t10_reservation *pr_tmpl = &dev->se_sub_dev->t10_pr;
+ unsigned char *buf = (unsigned char *)cmd->t_task->t_task_buf;
u16 add_len = 8; /* Hardcoded to 8. */
if (cmd->data_length < 6) {
@@ -4014,13 +4014,13 @@ static int core_scsi3_pri_report_capabilities(struct se_cmd *cmd)
*/
static int core_scsi3_pri_read_full_status(struct se_cmd *cmd)
{
- struct se_device *se_dev = SE_DEV(cmd);
+ struct se_device *se_dev = cmd->se_lun->lun_se_dev;
struct se_node_acl *se_nacl;
- struct se_subsystem_dev *su_dev = SU_DEV(se_dev);
+ struct se_subsystem_dev *su_dev = se_dev->se_sub_dev;
struct se_portal_group *se_tpg;
struct t10_pr_registration *pr_reg, *pr_reg_tmp;
- struct t10_reservation *pr_tmpl = &SU_DEV(se_dev)->t10_reservation;
- unsigned char *buf = (unsigned char *)T_TASK(cmd)->t_task_buf;
+ struct t10_reservation *pr_tmpl = &se_dev->se_sub_dev->t10_pr;
+ unsigned char *buf = (unsigned char *)cmd->t_task->t_task_buf;
u32 add_desc_len = 0, add_len = 0, desc_len, exp_desc_len;
u32 off = 8; /* off into first Full Status descriptor */
int format_code = 0;
@@ -4031,10 +4031,10 @@ static int core_scsi3_pri_read_full_status(struct se_cmd *cmd)
return PYX_TRANSPORT_INVALID_CDB_FIELD;
}
- buf[0] = ((T10_RES(su_dev)->pr_generation >> 24) & 0xff);
- buf[1] = ((T10_RES(su_dev)->pr_generation >> 16) & 0xff);
- buf[2] = ((T10_RES(su_dev)->pr_generation >> 8) & 0xff);
- buf[3] = (T10_RES(su_dev)->pr_generation & 0xff);
+ buf[0] = ((su_dev->t10_pr.pr_generation >> 24) & 0xff);
+ buf[1] = ((su_dev->t10_pr.pr_generation >> 16) & 0xff);
+ buf[2] = ((su_dev->t10_pr.pr_generation >> 8) & 0xff);
+ buf[3] = (su_dev->t10_pr.pr_generation & 0xff);
spin_lock(&pr_tmpl->registration_lock);
list_for_each_entry_safe(pr_reg, pr_reg_tmp,
@@ -4051,7 +4051,7 @@ static int core_scsi3_pri_read_full_status(struct se_cmd *cmd)
* Determine expected length of $FABRIC_MOD specific
* TransportID full status descriptor..
*/
- exp_desc_len = TPG_TFO(se_tpg)->tpg_get_pr_transport_id_len(
+ exp_desc_len = se_tpg->se_tpg_tfo->tpg_get_pr_transport_id_len(
se_tpg, se_nacl, pr_reg, &format_code);
if ((exp_desc_len + add_len) > cmd->data_length) {
@@ -4116,7 +4116,7 @@ static int core_scsi3_pri_read_full_status(struct se_cmd *cmd)
/*
* Now, have the $FABRIC_MOD fill in the protocol identifier
*/
- desc_len = TPG_TFO(se_tpg)->tpg_get_pr_transport_id(se_tpg,
+ desc_len = se_tpg->se_tpg_tfo->tpg_get_pr_transport_id(se_tpg,
se_nacl, pr_reg, &format_code, &buf[off+4]);
spin_lock(&pr_tmpl->registration_lock);
@@ -4174,7 +4174,7 @@ static int core_scsi3_emulate_pr_in(struct se_cmd *cmd, unsigned char *cdb)
int core_scsi3_emulate_pr(struct se_cmd *cmd)
{
- unsigned char *cdb = &T_TASK(cmd)->t_task_cdb[0];
+ unsigned char *cdb = &cmd->t_task->t_task_cdb[0];
struct se_device *dev = cmd->se_dev;
/*
* Following spc2r20 5.5.1 Reservations overview:
@@ -4213,39 +4213,39 @@ static int core_pt_seq_non_holder(
int core_setup_reservations(struct se_device *dev, int force_pt)
{
struct se_subsystem_dev *su_dev = dev->se_sub_dev;
- struct t10_reservation *rest = &su_dev->t10_reservation;
+ struct t10_reservation *rest = &su_dev->t10_pr;
/*
* If this device is from Target_Core_Mod/pSCSI, use the reservations
* of the Underlying SCSI hardware. In Linux/SCSI terms, this can
* cause a problem because libata and some SATA RAID HBAs appear
* under Linux/SCSI, but to emulate reservations themselves.
*/
- if (((TRANSPORT(dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) &&
- !(DEV_ATTRIB(dev)->emulate_reservations)) || force_pt) {
+ if (((dev->transport->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) &&
+ !(dev->se_sub_dev->se_dev_attrib.emulate_reservations)) || force_pt) {
rest->res_type = SPC_PASSTHROUGH;
rest->pr_ops.t10_reservation_check = &core_pt_reservation_check;
rest->pr_ops.t10_seq_non_holder = &core_pt_seq_non_holder;
printk(KERN_INFO "%s: Using SPC_PASSTHROUGH, no reservation"
- " emulation\n", TRANSPORT(dev)->name);
+ " emulation\n", dev->transport->name);
return 0;
}
/*
* If SPC-3 or above is reported by real or emulated struct se_device,
* use emulated Persistent Reservations.
*/
- if (TRANSPORT(dev)->get_device_rev(dev) >= SCSI_3) {
+ if (dev->transport->get_device_rev(dev) >= SCSI_3) {
rest->res_type = SPC3_PERSISTENT_RESERVATIONS;
rest->pr_ops.t10_reservation_check = &core_scsi3_pr_reservation_check;
rest->pr_ops.t10_seq_non_holder = &core_scsi3_pr_seq_non_holder;
printk(KERN_INFO "%s: Using SPC3_PERSISTENT_RESERVATIONS"
- " emulation\n", TRANSPORT(dev)->name);
+ " emulation\n", dev->transport->name);
} else {
rest->res_type = SPC2_RESERVATIONS;
rest->pr_ops.t10_reservation_check = &core_scsi2_reservation_check;
rest->pr_ops.t10_seq_non_holder =
&core_scsi2_reservation_seq_non_holder;
printk(KERN_INFO "%s: Using SPC2_RESERVATIONS emulation\n",
- TRANSPORT(dev)->name);
+ dev->transport->name);
}
return 0;
diff --git a/drivers/target/target_core_pscsi.c b/drivers/target/target_core_pscsi.c
index afe5f57..44a79a5 100644
--- a/drivers/target/target_core_pscsi.c
+++ b/drivers/target/target_core_pscsi.c
@@ -692,13 +692,12 @@ static int pscsi_transport_complete(struct se_task *task)
*/
if (((cdb[0] == MODE_SENSE) || (cdb[0] == MODE_SENSE_10)) &&
(status_byte(result) << 1) == SAM_STAT_GOOD) {
- if (!TASK_CMD(task)->se_deve)
+ if (!task->task_se_cmd->se_deve)
goto after_mode_sense;
- if (TASK_CMD(task)->se_deve->lun_flags &
+ if (task->task_se_cmd->se_deve->lun_flags &
TRANSPORT_LUNFLAGS_READ_ONLY) {
- unsigned char *buf = (unsigned char *)
- T_TASK(task->task_se_cmd)->t_task_buf;
+ unsigned char *buf = task->task_se_cmd->t_task->t_task_buf;
if (cdb[0] == MODE_SENSE_10) {
if (!(buf[3] & 0x80))
@@ -764,7 +763,7 @@ static struct se_task *
pscsi_alloc_task(struct se_cmd *cmd)
{
struct pscsi_plugin_task *pt;
- unsigned char *cdb = T_TASK(cmd)->t_task_cdb;
+ unsigned char *cdb = cmd->t_task->t_task_cdb;
pt = kzalloc(sizeof(struct pscsi_plugin_task), GFP_KERNEL);
if (!pt) {
@@ -777,7 +776,7 @@ pscsi_alloc_task(struct se_cmd *cmd)
* allocate the extended CDB buffer for per struct se_task context
* pt->pscsi_cdb now.
*/
- if (T_TASK(cmd)->t_task_cdb != T_TASK(cmd)->__t_task_cdb) {
+ if (cmd->t_task->t_task_cdb != cmd->t_task->__t_task_cdb) {
pt->pscsi_cdb = kzalloc(scsi_command_size(cdb), GFP_KERNEL);
if (!(pt->pscsi_cdb)) {
@@ -890,7 +889,7 @@ static void pscsi_free_task(struct se_task *task)
* Release the extended CDB allocation from pscsi_alloc_task()
* if one exists.
*/
- if (T_TASK(cmd)->t_task_cdb != T_TASK(cmd)->__t_task_cdb)
+ if (cmd->t_task->t_task_cdb != cmd->t_task->__t_task_cdb)
kfree(pt->pscsi_cdb);
/*
* We do not release the bio(s) here associated with this task, as
@@ -1255,7 +1254,7 @@ static int pscsi_map_task_SG(struct se_task *task)
*/
static int pscsi_map_task_non_SG(struct se_task *task)
{
- struct se_cmd *cmd = TASK_CMD(task);
+ struct se_cmd *cmd = task->task_se_cmd;
struct pscsi_plugin_task *pt = PSCSI_TASK(task);
struct pscsi_dev_virt *pdv = task->se_dev->dev_ptr;
int ret = 0;
@@ -1267,7 +1266,7 @@ static int pscsi_map_task_non_SG(struct se_task *task)
return 0;
ret = blk_rq_map_kern(pdv->pdv_sd->request_queue,
- pt->pscsi_req, T_TASK(cmd)->t_task_buf,
+ pt->pscsi_req, cmd->t_task->t_task_buf,
task->task_size, GFP_KERNEL);
if (ret < 0) {
printk(KERN_ERR "PSCSI: blk_rq_map_kern() failed: %d\n", ret);
@@ -1364,7 +1363,7 @@ static inline void pscsi_process_SAM_status(
pt->pscsi_result);
task->task_scsi_status = SAM_STAT_CHECK_CONDITION;
task->task_error_status = PYX_TRANSPORT_UNKNOWN_SAM_OPCODE;
- TASK_CMD(task)->transport_error_status =
+ task->task_se_cmd->transport_error_status =
PYX_TRANSPORT_UNKNOWN_SAM_OPCODE;
transport_complete_task(task, 0);
break;
diff --git a/drivers/target/target_core_rd.c b/drivers/target/target_core_rd.c
index 0a424bb..fbf06c3 100644
--- a/drivers/target/target_core_rd.c
+++ b/drivers/target/target_core_rd.c
@@ -336,7 +336,7 @@ rd_alloc_task(struct se_cmd *cmd)
printk(KERN_ERR "Unable to allocate struct rd_request\n");
return NULL;
}
- rd_req->rd_dev = SE_DEV(cmd)->dev_ptr;
+ rd_req->rd_dev = cmd->se_lun->lun_se_dev->dev_ptr;
return &rd_req->rd_task;
}
@@ -620,11 +620,11 @@ static int rd_MEMCPY_do_task(struct se_task *task)
unsigned long long lba;
int ret;
- req->rd_page = (task->task_lba * DEV_ATTRIB(dev)->block_size) / PAGE_SIZE;
+ req->rd_page = (task->task_lba * dev->se_sub_dev->se_dev_attrib.block_size) / PAGE_SIZE;
lba = task->task_lba;
req->rd_offset = (do_div(lba,
- (PAGE_SIZE / DEV_ATTRIB(dev)->block_size))) *
- DEV_ATTRIB(dev)->block_size;
+ (PAGE_SIZE / dev->se_sub_dev->se_dev_attrib.block_size))) *
+ dev->se_sub_dev->se_dev_attrib.block_size;
req->rd_size = task->task_size;
if (task->task_data_direction == DMA_FROM_DEVICE)
@@ -737,7 +737,7 @@ check_eot:
}
out:
- T_TASK(task->task_se_cmd)->t_tasks_se_num += *se_mem_cnt;
+ task->task_se_cmd->t_task->t_tasks_se_num += *se_mem_cnt;
#ifdef DEBUG_RAMDISK_DR
printk(KERN_INFO "RD_DR - Allocated %u struct se_mem segments for task\n",
*se_mem_cnt);
@@ -819,7 +819,7 @@ static int rd_DIRECT_without_offset(
}
out:
- T_TASK(task->task_se_cmd)->t_tasks_se_num += *se_mem_cnt;
+ task->task_se_cmd->t_task->t_tasks_se_num += *se_mem_cnt;
#ifdef DEBUG_RAMDISK_DR
printk(KERN_INFO "RD_DR - Allocated %u struct se_mem segments for task\n",
*se_mem_cnt);
@@ -845,13 +845,11 @@ static int rd_DIRECT_do_se_mem_map(
u32 task_offset = *task_offset_in;
unsigned long long lba;
int ret;
+ int block_size = task->se_dev->se_sub_dev->se_dev_attrib.block_size;
- req->rd_page = ((task->task_lba * DEV_ATTRIB(task->se_dev)->block_size) /
- PAGE_SIZE);
lba = task->task_lba;
- req->rd_offset = (do_div(lba,
- (PAGE_SIZE / DEV_ATTRIB(task->se_dev)->block_size))) *
- DEV_ATTRIB(task->se_dev)->block_size;
+ req->rd_page = ((task->task_lba * block_size) / PAGE_SIZE);
+ req->rd_offset = (do_div(lba, (PAGE_SIZE / block_size))) * block_size;
req->rd_size = task->task_size;
if (req->rd_offset)
@@ -864,7 +862,7 @@ static int rd_DIRECT_do_se_mem_map(
if (ret < 0)
return ret;
- if (CMD_TFO(cmd)->task_sg_chaining == 0)
+ if (cmd->se_tfo->task_sg_chaining == 0)
return 0;
/*
* Currently prevent writers from multiple HW fabrics doing
@@ -882,14 +880,14 @@ static int rd_DIRECT_do_se_mem_map(
* across multiple struct se_task->task_sg[].
*/
ret = transport_init_task_sg(task,
- list_entry(T_TASK(cmd)->t_mem_list->next,
+ list_entry(cmd->t_task->t_mem_list->next,
struct se_mem, se_list),
task_offset);
if (ret <= 0)
return ret;
return transport_map_mem_to_sg(task, se_mem_list, task->task_sg,
- list_entry(T_TASK(cmd)->t_mem_list->next,
+ list_entry(cmd->t_task->t_mem_list->next,
struct se_mem, se_list),
out_se_mem, se_mem_cnt, task_offset_in);
}
@@ -1019,7 +1017,7 @@ static sector_t rd_get_blocks(struct se_device *dev)
{
struct rd_dev *rd_dev = dev->dev_ptr;
unsigned long long blocks_long = ((rd_dev->rd_page_count * PAGE_SIZE) /
- DEV_ATTRIB(dev)->block_size) - 1;
+ dev->se_sub_dev->se_dev_attrib.block_size) - 1;
return blocks_long;
}
diff --git a/drivers/target/target_core_stat.c b/drivers/target/target_core_stat.c
index 5e3a067..a8d6e1d 100644
--- a/drivers/target/target_core_stat.c
+++ b/drivers/target/target_core_stat.c
@@ -402,8 +402,8 @@ static ssize_t target_stat_scsi_lu_show_attr_lu_name(
return -ENODEV;
/* scsiLuWwnName */
return snprintf(page, PAGE_SIZE, "%s\n",
- (strlen(DEV_T10_WWN(dev)->unit_serial)) ?
- (char *)&DEV_T10_WWN(dev)->unit_serial[0] : "None");
+ (strlen(dev->se_sub_dev->t10_wwn.unit_serial)) ?
+ dev->se_sub_dev->t10_wwn.unit_serial : "None");
}
DEV_STAT_SCSI_LU_ATTR_RO(lu_name);
@@ -413,17 +413,17 @@ static ssize_t target_stat_scsi_lu_show_attr_vend(
struct se_subsystem_dev *se_subdev = container_of(sgrps,
struct se_subsystem_dev, dev_stat_grps);
struct se_device *dev = se_subdev->se_dev_ptr;
- int j;
- char str[28];
+ int i;
+ char str[sizeof(dev->se_sub_dev->t10_wwn.vendor)+1];
if (!dev)
return -ENODEV;
+
/* scsiLuVendorId */
- memcpy(&str[0], (void *)DEV_T10_WWN(dev), 28);
- for (j = 0; j < 8; j++)
- str[j] = ISPRINT(DEV_T10_WWN(dev)->vendor[j]) ?
- DEV_T10_WWN(dev)->vendor[j] : 0x20;
- str[8] = 0;
+ for (i = 0; i < sizeof(dev->se_sub_dev->t10_wwn.vendor); i++)
+ str[i] = ISPRINT(dev->se_sub_dev->t10_wwn.vendor[i]) ?
+ dev->se_sub_dev->t10_wwn.vendor[i] : ' ';
+ str[i] = '\0';
return snprintf(page, PAGE_SIZE, "%s\n", str);
}
DEV_STAT_SCSI_LU_ATTR_RO(vend);
@@ -434,18 +434,17 @@ static ssize_t target_stat_scsi_lu_show_attr_prod(
struct se_subsystem_dev *se_subdev = container_of(sgrps,
struct se_subsystem_dev, dev_stat_grps);
struct se_device *dev = se_subdev->se_dev_ptr;
- int j;
- char str[28];
+ int i;
+ char str[sizeof(dev->se_sub_dev->t10_wwn.model)+1];
if (!dev)
return -ENODEV;
/* scsiLuProductId */
- memcpy(&str[0], (void *)DEV_T10_WWN(dev), 28);
- for (j = 0; j < 16; j++)
- str[j] = ISPRINT(DEV_T10_WWN(dev)->model[j]) ?
- DEV_T10_WWN(dev)->model[j] : 0x20;
- str[16] = 0;
+ for (i = 0; i < sizeof(dev->se_sub_dev->t10_wwn.vendor); i++)
+ str[i] = ISPRINT(dev->se_sub_dev->t10_wwn.model[i]) ?
+ dev->se_sub_dev->t10_wwn.model[i] : ' ';
+ str[i] = '\0';
return snprintf(page, PAGE_SIZE, "%s\n", str);
}
DEV_STAT_SCSI_LU_ATTR_RO(prod);
@@ -456,18 +455,17 @@ static ssize_t target_stat_scsi_lu_show_attr_rev(
struct se_subsystem_dev *se_subdev = container_of(sgrps,
struct se_subsystem_dev, dev_stat_grps);
struct se_device *dev = se_subdev->se_dev_ptr;
- int j;
- char str[28];
+ int i;
+ char str[sizeof(dev->se_sub_dev->t10_wwn.revision)+1];
if (!dev)
return -ENODEV;
/* scsiLuRevisionId */
- memcpy(&str[0], (void *)DEV_T10_WWN(dev), 28);
- for (j = 0; j < 4; j++)
- str[j] = ISPRINT(DEV_T10_WWN(dev)->revision[j]) ?
- DEV_T10_WWN(dev)->revision[j] : 0x20;
- str[4] = 0;
+ for (i = 0; i < sizeof(dev->se_sub_dev->t10_wwn.revision); i++)
+ str[i] = ISPRINT(dev->se_sub_dev->t10_wwn.revision[i]) ?
+ dev->se_sub_dev->t10_wwn.revision[i] : ' ';
+ str[i] = '\0';
return snprintf(page, PAGE_SIZE, "%s\n", str);
}
DEV_STAT_SCSI_LU_ATTR_RO(rev);
@@ -484,7 +482,7 @@ static ssize_t target_stat_scsi_lu_show_attr_dev_type(
/* scsiLuPeripheralType */
return snprintf(page, PAGE_SIZE, "%u\n",
- TRANSPORT(dev)->get_device_type(dev));
+ dev->transport->get_device_type(dev));
}
DEV_STAT_SCSI_LU_ATTR_RO(dev_type);
@@ -668,18 +666,18 @@ static struct config_item_type target_stat_scsi_lu_cit = {
*/
void target_stat_setup_dev_default_groups(struct se_subsystem_dev *se_subdev)
{
- struct config_group *dev_stat_grp = &DEV_STAT_GRP(se_subdev)->stat_group;
+ struct config_group *dev_stat_grp = &se_subdev->dev_stat_grps.stat_group;
- config_group_init_type_name(&DEV_STAT_GRP(se_subdev)->scsi_dev_group,
+ config_group_init_type_name(&se_subdev->dev_stat_grps.scsi_dev_group,
"scsi_dev", &target_stat_scsi_dev_cit);
- config_group_init_type_name(&DEV_STAT_GRP(se_subdev)->scsi_tgt_dev_group,
+ config_group_init_type_name(&se_subdev->dev_stat_grps.scsi_tgt_dev_group,
"scsi_tgt_dev", &target_stat_scsi_tgt_dev_cit);
- config_group_init_type_name(&DEV_STAT_GRP(se_subdev)->scsi_lu_group,
+ config_group_init_type_name(&se_subdev->dev_stat_grps.scsi_lu_group,
"scsi_lu", &target_stat_scsi_lu_cit);
- dev_stat_grp->default_groups[0] = &DEV_STAT_GRP(se_subdev)->scsi_dev_group;
- dev_stat_grp->default_groups[1] = &DEV_STAT_GRP(se_subdev)->scsi_tgt_dev_group;
- dev_stat_grp->default_groups[2] = &DEV_STAT_GRP(se_subdev)->scsi_lu_group;
+ dev_stat_grp->default_groups[0] = &se_subdev->dev_stat_grps.scsi_dev_group;
+ dev_stat_grp->default_groups[1] = &se_subdev->dev_stat_grps.scsi_tgt_dev_group;
+ dev_stat_grp->default_groups[2] = &se_subdev->dev_stat_grps.scsi_lu_group;
dev_stat_grp->default_groups[3] = NULL;
}
@@ -922,7 +920,7 @@ static ssize_t target_stat_scsi_tgt_port_show_attr_name(
tpg = sep->sep_tpg;
ret = snprintf(page, PAGE_SIZE, "%sPort#%u\n",
- TPG_TFO(tpg)->get_fabric_name(), sep->sep_index);
+ tpg->se_tpg_tfo->get_fabric_name(), sep->sep_index);
spin_unlock(&lun->lun_sep_lock);
return ret;
}
@@ -945,8 +943,8 @@ static ssize_t target_stat_scsi_tgt_port_show_attr_port_index(
tpg = sep->sep_tpg;
ret = snprintf(page, PAGE_SIZE, "%s%s%d\n",
- TPG_TFO(tpg)->tpg_get_wwn(tpg), "+t+",
- TPG_TFO(tpg)->tpg_get_tag(tpg));
+ tpg->se_tpg_tfo->tpg_get_wwn(tpg), "+t+",
+ tpg->se_tpg_tfo->tpg_get_tag(tpg));
spin_unlock(&lun->lun_sep_lock);
return ret;
}
@@ -1128,7 +1126,7 @@ static ssize_t target_stat_scsi_transport_show_attr_device(
tpg = sep->sep_tpg;
/* scsiTransportType */
ret = snprintf(page, PAGE_SIZE, "scsiTransport%s\n",
- TPG_TFO(tpg)->get_fabric_name());
+ tpg->se_tpg_tfo->get_fabric_name());
spin_unlock(&lun->lun_sep_lock);
return ret;
}
@@ -1150,7 +1148,7 @@ static ssize_t target_stat_scsi_transport_show_attr_indx(
}
tpg = sep->sep_tpg;
ret = snprintf(page, PAGE_SIZE, "%u\n",
- TPG_TFO(tpg)->tpg_get_inst_index(tpg));
+ tpg->se_tpg_tfo->tpg_get_inst_index(tpg));
spin_unlock(&lun->lun_sep_lock);
return ret;
}
@@ -1173,10 +1171,10 @@ static ssize_t target_stat_scsi_transport_show_attr_dev_name(
return -ENODEV;
}
tpg = sep->sep_tpg;
- wwn = DEV_T10_WWN(dev);
+ wwn = &dev->se_sub_dev->t10_wwn;
/* scsiTransportDevName */
ret = snprintf(page, PAGE_SIZE, "%s+%s\n",
- TPG_TFO(tpg)->tpg_get_wwn(tpg),
+ tpg->se_tpg_tfo->tpg_get_wwn(tpg),
(strlen(wwn->unit_serial)) ? wwn->unit_serial :
wwn->vendor);
spin_unlock(&lun->lun_sep_lock);
@@ -1212,18 +1210,18 @@ static struct config_item_type target_stat_scsi_transport_cit = {
*/
void target_stat_setup_port_default_groups(struct se_lun *lun)
{
- struct config_group *port_stat_grp = &PORT_STAT_GRP(lun)->stat_group;
+ struct config_group *port_stat_grp = &lun->port_stat_grps.stat_group;
- config_group_init_type_name(&PORT_STAT_GRP(lun)->scsi_port_group,
+ config_group_init_type_name(&lun->port_stat_grps.scsi_port_group,
"scsi_port", &target_stat_scsi_port_cit);
- config_group_init_type_name(&PORT_STAT_GRP(lun)->scsi_tgt_port_group,
+ config_group_init_type_name(&lun->port_stat_grps.scsi_tgt_port_group,
"scsi_tgt_port", &target_stat_scsi_tgt_port_cit);
- config_group_init_type_name(&PORT_STAT_GRP(lun)->scsi_transport_group,
+ config_group_init_type_name(&lun->port_stat_grps.scsi_transport_group,
"scsi_transport", &target_stat_scsi_transport_cit);
- port_stat_grp->default_groups[0] = &PORT_STAT_GRP(lun)->scsi_port_group;
- port_stat_grp->default_groups[1] = &PORT_STAT_GRP(lun)->scsi_tgt_port_group;
- port_stat_grp->default_groups[2] = &PORT_STAT_GRP(lun)->scsi_transport_group;
+ port_stat_grp->default_groups[0] = &lun->port_stat_grps.scsi_port_group;
+ port_stat_grp->default_groups[1] = &lun->port_stat_grps.scsi_tgt_port_group;
+ port_stat_grp->default_groups[2] = &lun->port_stat_grps.scsi_transport_group;
port_stat_grp->default_groups[3] = NULL;
}
@@ -1264,7 +1262,7 @@ static ssize_t target_stat_scsi_auth_intr_show_attr_inst(
tpg = nacl->se_tpg;
/* scsiInstIndex */
ret = snprintf(page, PAGE_SIZE, "%u\n",
- TPG_TFO(tpg)->tpg_get_inst_index(tpg));
+ tpg->se_tpg_tfo->tpg_get_inst_index(tpg));
spin_unlock_irq(&nacl->device_list_lock);
return ret;
}
@@ -1314,7 +1312,7 @@ static ssize_t target_stat_scsi_auth_intr_show_attr_port(
}
tpg = nacl->se_tpg;
/* scsiAuthIntrTgtPortIndex */
- ret = snprintf(page, PAGE_SIZE, "%u\n", TPG_TFO(tpg)->tpg_get_tag(tpg));
+ ret = snprintf(page, PAGE_SIZE, "%u\n", tpg->se_tpg_tfo->tpg_get_tag(tpg));
spin_unlock_irq(&nacl->device_list_lock);
return ret;
}
@@ -1632,7 +1630,7 @@ static ssize_t target_stat_scsi_att_intr_port_show_attr_inst(
tpg = nacl->se_tpg;
/* scsiInstIndex */
ret = snprintf(page, PAGE_SIZE, "%u\n",
- TPG_TFO(tpg)->tpg_get_inst_index(tpg));
+ tpg->se_tpg_tfo->tpg_get_inst_index(tpg));
spin_unlock_irq(&nacl->device_list_lock);
return ret;
}
@@ -1682,7 +1680,7 @@ static ssize_t target_stat_scsi_att_intr_port_show_attr_port(
}
tpg = nacl->se_tpg;
/* scsiPortIndex */
- ret = snprintf(page, PAGE_SIZE, "%u\n", TPG_TFO(tpg)->tpg_get_tag(tpg));
+ ret = snprintf(page, PAGE_SIZE, "%u\n", tpg->se_tpg_tfo->tpg_get_tag(tpg));
spin_unlock_irq(&nacl->device_list_lock);
return ret;
}
@@ -1708,7 +1706,7 @@ static ssize_t target_stat_scsi_att_intr_port_show_attr_indx(
tpg = nacl->se_tpg;
/* scsiAttIntrPortIndex */
ret = snprintf(page, PAGE_SIZE, "%u\n",
- TPG_TFO(tpg)->sess_get_index(se_sess));
+ tpg->se_tpg_tfo->sess_get_index(se_sess));
spin_unlock_irq(&nacl->nacl_sess_lock);
return ret;
}
@@ -1757,8 +1755,8 @@ static ssize_t target_stat_scsi_att_intr_port_show_attr_port_ident(
tpg = nacl->se_tpg;
/* scsiAttIntrPortName+scsiAttIntrPortIdentifier */
memset(buf, 0, 64);
- if (TPG_TFO(tpg)->sess_get_initiator_sid != NULL)
- TPG_TFO(tpg)->sess_get_initiator_sid(se_sess,
+ if (tpg->se_tpg_tfo->sess_get_initiator_sid != NULL)
+ tpg->se_tpg_tfo->sess_get_initiator_sid(se_sess,
(unsigned char *)&buf[0], 64);
ret = snprintf(page, PAGE_SIZE, "%s+i+%s\n", nacl->initiatorname, buf);
@@ -1797,14 +1795,14 @@ static struct config_item_type target_stat_scsi_att_intr_port_cit = {
*/
void target_stat_setup_mappedlun_default_groups(struct se_lun_acl *lacl)
{
- struct config_group *ml_stat_grp = &ML_STAT_GRPS(lacl)->stat_group;
+ struct config_group *ml_stat_grp = &lacl->ml_stat_grps.stat_group;
- config_group_init_type_name(&ML_STAT_GRPS(lacl)->scsi_auth_intr_group,
+ config_group_init_type_name(&lacl->ml_stat_grps.scsi_auth_intr_group,
"scsi_auth_intr", &target_stat_scsi_auth_intr_cit);
- config_group_init_type_name(&ML_STAT_GRPS(lacl)->scsi_att_intr_port_group,
+ config_group_init_type_name(&lacl->ml_stat_grps.scsi_att_intr_port_group,
"scsi_att_intr_port", &target_stat_scsi_att_intr_port_cit);
- ml_stat_grp->default_groups[0] = &ML_STAT_GRPS(lacl)->scsi_auth_intr_group;
- ml_stat_grp->default_groups[1] = &ML_STAT_GRPS(lacl)->scsi_att_intr_port_group;
+ ml_stat_grp->default_groups[0] = &lacl->ml_stat_grps.scsi_auth_intr_group;
+ ml_stat_grp->default_groups[1] = &lacl->ml_stat_grps.scsi_att_intr_port_group;
ml_stat_grp->default_groups[2] = NULL;
}
diff --git a/drivers/target/target_core_tmr.c b/drivers/target/target_core_tmr.c
index 7cd8e29..2f73749 100644
--- a/drivers/target/target_core_tmr.c
+++ b/drivers/target/target_core_tmr.c
@@ -133,7 +133,7 @@ int core_tmr_lun_reset(
* which the command was received shall be completed with TASK ABORTED
* status (see SAM-4).
*/
- tas = DEV_ATTRIB(dev)->emulate_tas;
+ tas = dev->se_sub_dev->se_dev_attrib.emulate_tas;
/*
* Determine if this se_tmr is coming from a $FABRIC_MOD
* or struct se_device passthrough..
@@ -144,13 +144,13 @@ int core_tmr_lun_reset(
if (tmr_nacl && tmr_tpg) {
DEBUG_LR("LUN_RESET: TMR caller fabric: %s"
" initiator port %s\n",
- TPG_TFO(tmr_tpg)->get_fabric_name(),
+ tmr_tpg->se_tpg_tfo->get_fabric_name(),
tmr_nacl->initiatorname);
}
}
DEBUG_LR("LUN_RESET: %s starting for [%s], tas: %d\n",
(preempt_and_abort_list) ? "Preempt" : "TMR",
- TRANSPORT(dev)->name, tas);
+ dev->transport->name, tas);
/*
* Release all pending and outgoing TMRs aside from the received
* LUN_RESET tmr..
@@ -179,14 +179,14 @@ int core_tmr_lun_reset(
continue;
spin_unlock(&dev->se_tmr_lock);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- if (!(atomic_read(&T_TASK(cmd)->t_transport_active))) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ if (!(atomic_read(&cmd->t_task->t_transport_active))) {
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
spin_lock(&dev->se_tmr_lock);
continue;
}
if (cmd->t_state == TRANSPORT_ISTATE_PROCESSING) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
spin_lock(&dev->se_tmr_lock);
continue;
}
@@ -194,7 +194,7 @@ int core_tmr_lun_reset(
" Response: 0x%02x, t_state: %d\n",
(preempt_and_abort_list) ? "Preempt" : "", tmr_p,
tmr_p->function, tmr_p->response, cmd->t_state);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
transport_cmd_finish_abort_tmr(cmd);
spin_lock(&dev->se_tmr_lock);
@@ -224,16 +224,16 @@ int core_tmr_lun_reset(
spin_lock_irqsave(&dev->execute_task_lock, flags);
list_for_each_entry_safe(task, task_tmp, &dev->state_task_list,
t_state_list) {
- if (!(TASK_CMD(task))) {
- printk(KERN_ERR "TASK_CMD(task) is NULL!\n");
+ if (!task->task_se_cmd) {
+ printk(KERN_ERR "task->task_se_cmd is NULL!\n");
continue;
}
- cmd = TASK_CMD(task);
+ cmd = task->task_se_cmd;
- if (!T_TASK(cmd)) {
- printk(KERN_ERR "T_TASK(cmd) is NULL for task: %p cmd:"
+ if (!cmd->t_task) {
+ printk(KERN_ERR "cmd->t_task is NULL for task: %p cmd:"
" %p ITT: 0x%08x\n", task, cmd,
- CMD_TFO(cmd)->get_task_tag(cmd));
+ cmd->se_tfo->get_task_tag(cmd));
continue;
}
/*
@@ -254,38 +254,38 @@ int core_tmr_lun_reset(
atomic_set(&task->task_state_active, 0);
spin_unlock_irqrestore(&dev->execute_task_lock, flags);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
DEBUG_LR("LUN_RESET: %s cmd: %p task: %p"
" ITT/CmdSN: 0x%08x/0x%08x, i_state: %d, t_state/"
"def_t_state: %d/%d cdb: 0x%02x\n",
(preempt_and_abort_list) ? "Preempt" : "", cmd, task,
- CMD_TFO(cmd)->get_task_tag(cmd), 0,
- CMD_TFO(cmd)->get_cmd_state(cmd), cmd->t_state,
- cmd->deferred_t_state, T_TASK(cmd)->t_task_cdb[0]);
+ cmd->se_tfo->get_task_tag(cmd), 0,
+ cmd->se_tfo->get_cmd_state(cmd), cmd->t_state,
+ cmd->deferred_t_state, cmd->t_task->t_task_cdb[0]);
DEBUG_LR("LUN_RESET: ITT[0x%08x] - pr_res_key: 0x%016Lx"
" t_task_cdbs: %d t_task_cdbs_left: %d"
" t_task_cdbs_sent: %d -- t_transport_active: %d"
" t_transport_stop: %d t_transport_sent: %d\n",
- CMD_TFO(cmd)->get_task_tag(cmd), cmd->pr_res_key,
- T_TASK(cmd)->t_task_cdbs,
- atomic_read(&T_TASK(cmd)->t_task_cdbs_left),
- atomic_read(&T_TASK(cmd)->t_task_cdbs_sent),
- atomic_read(&T_TASK(cmd)->t_transport_active),
- atomic_read(&T_TASK(cmd)->t_transport_stop),
- atomic_read(&T_TASK(cmd)->t_transport_sent));
+ cmd->se_tfo->get_task_tag(cmd), cmd->pr_res_key,
+ cmd->t_task->t_task_cdbs,
+ atomic_read(&cmd->t_task->t_task_cdbs_left),
+ atomic_read(&cmd->t_task->t_task_cdbs_sent),
+ atomic_read(&cmd->t_task->t_transport_active),
+ atomic_read(&cmd->t_task->t_transport_stop),
+ atomic_read(&cmd->t_task->t_transport_sent));
if (atomic_read(&task->task_active)) {
atomic_set(&task->task_stop, 1);
spin_unlock_irqrestore(
- &T_TASK(cmd)->t_state_lock, flags);
+ &cmd->t_task->t_state_lock, flags);
DEBUG_LR("LUN_RESET: Waiting for task: %p to shutdown"
" for dev: %p\n", task, dev);
wait_for_completion(&task->task_stop_comp);
DEBUG_LR("LUN_RESET Completed task: %p shutdown for"
" dev: %p\n", task, dev);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- atomic_dec(&T_TASK(cmd)->t_task_cdbs_left);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ atomic_dec(&cmd->t_task->t_task_cdbs_left);
atomic_set(&task->task_active, 0);
atomic_set(&task->task_stop, 0);
@@ -295,24 +295,24 @@ int core_tmr_lun_reset(
}
__transport_stop_task_timer(task, &flags);
- if (!(atomic_dec_and_test(&T_TASK(cmd)->t_task_cdbs_ex_left))) {
+ if (!(atomic_dec_and_test(&cmd->t_task->t_task_cdbs_ex_left))) {
spin_unlock_irqrestore(
- &T_TASK(cmd)->t_state_lock, flags);
+ &cmd->t_task->t_state_lock, flags);
DEBUG_LR("LUN_RESET: Skipping task: %p, dev: %p for"
" t_task_cdbs_ex_left: %d\n", task, dev,
- atomic_read(&T_TASK(cmd)->t_task_cdbs_ex_left));
+ atomic_read(&cmd->t_task->t_task_cdbs_ex_left));
spin_lock_irqsave(&dev->execute_task_lock, flags);
continue;
}
- fe_count = atomic_read(&T_TASK(cmd)->t_fe_count);
+ fe_count = atomic_read(&cmd->t_task->t_fe_count);
- if (atomic_read(&T_TASK(cmd)->t_transport_active)) {
+ if (atomic_read(&cmd->t_task->t_transport_active)) {
DEBUG_LR("LUN_RESET: got t_transport_active = 1 for"
" task: %p, t_fe_count: %d dev: %p\n", task,
fe_count, dev);
- atomic_set(&T_TASK(cmd)->t_transport_aborted, 1);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock,
+ atomic_set(&cmd->t_task->t_transport_aborted, 1);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock,
flags);
core_tmr_handle_tas_abort(tmr_nacl, cmd, tas, fe_count);
@@ -321,8 +321,8 @@ int core_tmr_lun_reset(
}
DEBUG_LR("LUN_RESET: Got t_transport_active = 0 for task: %p,"
" t_fe_count: %d dev: %p\n", task, fe_count, dev);
- atomic_set(&T_TASK(cmd)->t_transport_aborted, 1);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ atomic_set(&cmd->t_task->t_transport_aborted, 1);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
core_tmr_handle_tas_abort(tmr_nacl, cmd, tas, fe_count);
spin_lock_irqsave(&dev->execute_task_lock, flags);
@@ -365,7 +365,7 @@ int core_tmr_lun_reset(
if (prout_cmd == cmd)
continue;
- atomic_dec(&T_TASK(cmd)->t_transport_queue_active);
+ atomic_dec(&cmd->t_task->t_transport_queue_active);
atomic_dec(&qobj->queue_cnt);
list_del(&qr->qr_list);
spin_unlock_irqrestore(&qobj->cmd_queue_lock, flags);
@@ -376,7 +376,7 @@ int core_tmr_lun_reset(
DEBUG_LR("LUN_RESET: %s from Device Queue: cmd: %p t_state:"
" %d t_fe_count: %d\n", (preempt_and_abort_list) ?
"Preempt" : "", cmd, state,
- atomic_read(&T_TASK(cmd)->t_fe_count));
+ atomic_read(&cmd->t_task->t_fe_count));
/*
* Signal that the command has failed via cmd->se_cmd_flags,
* and call TFO->new_cmd_failure() to wakeup any fabric
@@ -388,7 +388,7 @@ int core_tmr_lun_reset(
transport_new_cmd_failure(cmd);
core_tmr_handle_tas_abort(tmr_nacl, cmd, tas,
- atomic_read(&T_TASK(cmd)->t_fe_count));
+ atomic_read(&cmd->t_task->t_fe_count));
spin_lock_irqsave(&qobj->cmd_queue_lock, flags);
}
spin_unlock_irqrestore(&qobj->cmd_queue_lock, flags);
@@ -411,6 +411,6 @@ int core_tmr_lun_reset(
DEBUG_LR("LUN_RESET: %s for [%s] Complete\n",
(preempt_and_abort_list) ? "Preempt" : "TMR",
- TRANSPORT(dev)->name);
+ dev->transport->name);
return 0;
}
diff --git a/drivers/target/target_core_tpg.c b/drivers/target/target_core_tpg.c
index 4e2bee9..448129f 100644
--- a/drivers/target/target_core_tpg.c
+++ b/drivers/target/target_core_tpg.c
@@ -74,7 +74,7 @@ static void core_clear_initiator_node_from_tpg(
if (!deve->se_lun) {
printk(KERN_ERR "%s device entries device pointer is"
" NULL, but Initiator has access.\n",
- TPG_TFO(tpg)->get_fabric_name());
+ tpg->se_tpg_tfo->get_fabric_name());
continue;
}
@@ -177,7 +177,7 @@ void core_tpg_add_node_to_devs(
* By default in LIO-Target $FABRIC_MOD,
* demo_mode_write_protect is ON, or READ_ONLY;
*/
- if (!(TPG_TFO(tpg)->tpg_check_demo_mode_write_protect(tpg))) {
+ if (!(tpg->se_tpg_tfo->tpg_check_demo_mode_write_protect(tpg))) {
if (dev->dev_flags & DF_READ_ONLY)
lun_access = TRANSPORT_LUNFLAGS_READ_ONLY;
else
@@ -187,7 +187,7 @@ void core_tpg_add_node_to_devs(
* Allow only optical drives to issue R/W in default RO
* demo mode.
*/
- if (TRANSPORT(dev)->get_device_type(dev) == TYPE_DISK)
+ if (dev->transport->get_device_type(dev) == TYPE_DISK)
lun_access = TRANSPORT_LUNFLAGS_READ_ONLY;
else
lun_access = TRANSPORT_LUNFLAGS_READ_WRITE;
@@ -195,8 +195,8 @@ void core_tpg_add_node_to_devs(
printk(KERN_INFO "TARGET_CORE[%s]->TPG[%u]_LUN[%u] - Adding %s"
" access for LUN in Demo Mode\n",
- TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_tag(tpg), lun->unpacked_lun,
+ tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg), lun->unpacked_lun,
(lun_access == TRANSPORT_LUNFLAGS_READ_WRITE) ?
"READ-WRITE" : "READ-ONLY");
@@ -217,7 +217,7 @@ static int core_set_queue_depth_for_node(
{
if (!acl->queue_depth) {
printk(KERN_ERR "Queue depth for %s Initiator Node: %s is 0,"
- "defaulting to 1.\n", TPG_TFO(tpg)->get_fabric_name(),
+ "defaulting to 1.\n", tpg->se_tpg_tfo->get_fabric_name(),
acl->initiatorname);
acl->queue_depth = 1;
}
@@ -268,10 +268,10 @@ struct se_node_acl *core_tpg_check_initiator_node_acl(
if ((acl))
return acl;
- if (!(TPG_TFO(tpg)->tpg_check_demo_mode(tpg)))
+ if (!(tpg->se_tpg_tfo->tpg_check_demo_mode(tpg)))
return NULL;
- acl = TPG_TFO(tpg)->tpg_alloc_fabric_acl(tpg);
+ acl = tpg->se_tpg_tfo->tpg_alloc_fabric_acl(tpg);
if (!(acl))
return NULL;
@@ -280,23 +280,23 @@ struct se_node_acl *core_tpg_check_initiator_node_acl(
spin_lock_init(&acl->device_list_lock);
spin_lock_init(&acl->nacl_sess_lock);
atomic_set(&acl->acl_pr_ref_count, 0);
- acl->queue_depth = TPG_TFO(tpg)->tpg_get_default_depth(tpg);
+ acl->queue_depth = tpg->se_tpg_tfo->tpg_get_default_depth(tpg);
snprintf(acl->initiatorname, TRANSPORT_IQN_LEN, "%s", initiatorname);
acl->se_tpg = tpg;
acl->acl_index = scsi_get_new_index(SCSI_AUTH_INTR_INDEX);
spin_lock_init(&acl->stats_lock);
acl->dynamic_node_acl = 1;
- TPG_TFO(tpg)->set_default_node_attributes(acl);
+ tpg->se_tpg_tfo->set_default_node_attributes(acl);
if (core_create_device_list_for_node(acl) < 0) {
- TPG_TFO(tpg)->tpg_release_fabric_acl(tpg, acl);
+ tpg->se_tpg_tfo->tpg_release_fabric_acl(tpg, acl);
return NULL;
}
if (core_set_queue_depth_for_node(tpg, acl) < 0) {
core_free_device_list_for_node(acl, tpg);
- TPG_TFO(tpg)->tpg_release_fabric_acl(tpg, acl);
+ tpg->se_tpg_tfo->tpg_release_fabric_acl(tpg, acl);
return NULL;
}
@@ -308,9 +308,9 @@ struct se_node_acl *core_tpg_check_initiator_node_acl(
spin_unlock_bh(&tpg->acl_node_lock);
printk("%s_TPG[%u] - Added DYNAMIC ACL with TCQ Depth: %d for %s"
- " Initiator Node: %s\n", TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_tag(tpg), acl->queue_depth,
- TPG_TFO(tpg)->get_fabric_name(), initiatorname);
+ " Initiator Node: %s\n", tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg), acl->queue_depth,
+ tpg->se_tpg_tfo->get_fabric_name(), initiatorname);
return acl;
}
@@ -361,8 +361,8 @@ struct se_node_acl *core_tpg_add_initiator_node_acl(
if (acl->dynamic_node_acl) {
acl->dynamic_node_acl = 0;
printk(KERN_INFO "%s_TPG[%u] - Replacing dynamic ACL"
- " for %s\n", TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_tag(tpg), initiatorname);
+ " for %s\n", tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg), initiatorname);
spin_unlock_bh(&tpg->acl_node_lock);
/*
* Release the locally allocated struct se_node_acl
@@ -370,15 +370,15 @@ struct se_node_acl *core_tpg_add_initiator_node_acl(
* a pointer to an existing demo mode node ACL.
*/
if (se_nacl)
- TPG_TFO(tpg)->tpg_release_fabric_acl(tpg,
+ tpg->se_tpg_tfo->tpg_release_fabric_acl(tpg,
se_nacl);
goto done;
}
printk(KERN_ERR "ACL entry for %s Initiator"
" Node %s already exists for TPG %u, ignoring"
- " request.\n", TPG_TFO(tpg)->get_fabric_name(),
- initiatorname, TPG_TFO(tpg)->tpg_get_tag(tpg));
+ " request.\n", tpg->se_tpg_tfo->get_fabric_name(),
+ initiatorname, tpg->se_tpg_tfo->tpg_get_tag(tpg));
spin_unlock_bh(&tpg->acl_node_lock);
return ERR_PTR(-EEXIST);
}
@@ -406,16 +406,16 @@ struct se_node_acl *core_tpg_add_initiator_node_acl(
acl->acl_index = scsi_get_new_index(SCSI_AUTH_INTR_INDEX);
spin_lock_init(&acl->stats_lock);
- TPG_TFO(tpg)->set_default_node_attributes(acl);
+ tpg->se_tpg_tfo->set_default_node_attributes(acl);
if (core_create_device_list_for_node(acl) < 0) {
- TPG_TFO(tpg)->tpg_release_fabric_acl(tpg, acl);
+ tpg->se_tpg_tfo->tpg_release_fabric_acl(tpg, acl);
return ERR_PTR(-ENOMEM);
}
if (core_set_queue_depth_for_node(tpg, acl) < 0) {
core_free_device_list_for_node(acl, tpg);
- TPG_TFO(tpg)->tpg_release_fabric_acl(tpg, acl);
+ tpg->se_tpg_tfo->tpg_release_fabric_acl(tpg, acl);
return ERR_PTR(-EINVAL);
}
@@ -426,9 +426,9 @@ struct se_node_acl *core_tpg_add_initiator_node_acl(
done:
printk(KERN_INFO "%s_TPG[%hu] - Added ACL with TCQ Depth: %d for %s"
- " Initiator Node: %s\n", TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_tag(tpg), acl->queue_depth,
- TPG_TFO(tpg)->get_fabric_name(), initiatorname);
+ " Initiator Node: %s\n", tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg), acl->queue_depth,
+ tpg->se_tpg_tfo->get_fabric_name(), initiatorname);
return acl;
}
@@ -463,7 +463,7 @@ int core_tpg_del_initiator_node_acl(
/*
* Determine if the session needs to be closed by our context.
*/
- if (!(TPG_TFO(tpg)->shutdown_session(sess)))
+ if (!(tpg->se_tpg_tfo->shutdown_session(sess)))
continue;
spin_unlock_bh(&tpg->session_lock);
@@ -471,7 +471,7 @@ int core_tpg_del_initiator_node_acl(
* If the $FABRIC_MOD session for the Initiator Node ACL exists,
* forcefully shutdown the $FABRIC_MOD session/nexus.
*/
- TPG_TFO(tpg)->close_session(sess);
+ tpg->se_tpg_tfo->close_session(sess);
spin_lock_bh(&tpg->session_lock);
}
@@ -482,9 +482,9 @@ int core_tpg_del_initiator_node_acl(
core_free_device_list_for_node(acl, tpg);
printk(KERN_INFO "%s_TPG[%hu] - Deleted ACL with TCQ Depth: %d for %s"
- " Initiator Node: %s\n", TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_tag(tpg), acl->queue_depth,
- TPG_TFO(tpg)->get_fabric_name(), acl->initiatorname);
+ " Initiator Node: %s\n", tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg), acl->queue_depth,
+ tpg->se_tpg_tfo->get_fabric_name(), acl->initiatorname);
return 0;
}
@@ -509,8 +509,8 @@ int core_tpg_set_initiator_node_queue_depth(
if (!(acl)) {
printk(KERN_ERR "Access Control List entry for %s Initiator"
" Node %s does not exists for TPG %hu, ignoring"
- " request.\n", TPG_TFO(tpg)->get_fabric_name(),
- initiatorname, TPG_TFO(tpg)->tpg_get_tag(tpg));
+ " request.\n", tpg->se_tpg_tfo->get_fabric_name(),
+ initiatorname, tpg->se_tpg_tfo->tpg_get_tag(tpg));
spin_unlock_bh(&tpg->acl_node_lock);
return -ENODEV;
}
@@ -531,7 +531,7 @@ int core_tpg_set_initiator_node_queue_depth(
" operational. To forcefully change the queue"
" depth and force session reinstatement"
" use the \"force=1\" parameter.\n",
- TPG_TFO(tpg)->get_fabric_name(), initiatorname);
+ tpg->se_tpg_tfo->get_fabric_name(), initiatorname);
spin_unlock_bh(&tpg->session_lock);
spin_lock_bh(&tpg->acl_node_lock);
@@ -543,7 +543,7 @@ int core_tpg_set_initiator_node_queue_depth(
/*
* Determine if the session needs to be closed by our context.
*/
- if (!(TPG_TFO(tpg)->shutdown_session(sess)))
+ if (!(tpg->se_tpg_tfo->shutdown_session(sess)))
continue;
init_sess = sess;
@@ -555,7 +555,7 @@ int core_tpg_set_initiator_node_queue_depth(
* Change the value in the Node's struct se_node_acl, and call
* core_set_queue_depth_for_node() to add the requested queue depth.
*
- * Finally call TPG_TFO(tpg)->close_session() to force session
+ * Finally call tpg->se_tpg_tfo->close_session() to force session
* reinstatement to occur if there is an active session for the
* $FABRIC_MOD Initiator Node in question.
*/
@@ -567,10 +567,10 @@ int core_tpg_set_initiator_node_queue_depth(
* Force session reinstatement if
* core_set_queue_depth_for_node() failed, because we assume
* the $FABRIC_MOD has already the set session reinstatement
- * bit from TPG_TFO(tpg)->shutdown_session() called above.
+ * bit from tpg->se_tpg_tfo->shutdown_session() called above.
*/
if (init_sess)
- TPG_TFO(tpg)->close_session(init_sess);
+ tpg->se_tpg_tfo->close_session(init_sess);
spin_lock_bh(&tpg->acl_node_lock);
if (dynamic_acl)
@@ -584,12 +584,12 @@ int core_tpg_set_initiator_node_queue_depth(
* forcefully shutdown the $FABRIC_MOD session/nexus.
*/
if (init_sess)
- TPG_TFO(tpg)->close_session(init_sess);
+ tpg->se_tpg_tfo->close_session(init_sess);
printk(KERN_INFO "Successfuly changed queue depth to: %d for Initiator"
" Node: %s on %s Target Portal Group: %u\n", queue_depth,
- initiatorname, TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_tag(tpg));
+ initiatorname, tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg));
spin_lock_bh(&tpg->acl_node_lock);
if (dynamic_acl)
@@ -703,9 +703,9 @@ int core_tpg_deregister(struct se_portal_group *se_tpg)
printk(KERN_INFO "TARGET_CORE[%s]: Deallocating %s struct se_portal_group"
" for endpoint: %s Portal Tag %u\n",
(se_tpg->se_tpg_type == TRANSPORT_TPG_TYPE_NORMAL) ?
- "Normal" : "Discovery", TPG_TFO(se_tpg)->get_fabric_name(),
- TPG_TFO(se_tpg)->tpg_get_wwn(se_tpg),
- TPG_TFO(se_tpg)->tpg_get_tag(se_tpg));
+ "Normal" : "Discovery", se_tpg->se_tpg_tfo->get_fabric_name(),
+ se_tpg->se_tpg_tfo->tpg_get_wwn(se_tpg),
+ se_tpg->se_tpg_tfo->tpg_get_tag(se_tpg));
spin_lock_bh(&tpg_lock);
list_del(&se_tpg->se_tpg_node);
@@ -727,7 +727,7 @@ int core_tpg_deregister(struct se_portal_group *se_tpg)
core_tpg_wait_for_nacl_pr_ref(nacl);
core_free_device_list_for_node(nacl, se_tpg);
- TPG_TFO(se_tpg)->tpg_release_fabric_acl(se_tpg, nacl);
+ se_tpg->se_tpg_tfo->tpg_release_fabric_acl(se_tpg, nacl);
spin_lock_bh(&se_tpg->acl_node_lock);
}
@@ -751,9 +751,9 @@ struct se_lun *core_tpg_pre_addlun(
if (unpacked_lun > (TRANSPORT_MAX_LUNS_PER_TPG-1)) {
printk(KERN_ERR "%s LUN: %u exceeds TRANSPORT_MAX_LUNS_PER_TPG"
"-1: %u for Target Portal Group: %u\n",
- TPG_TFO(tpg)->get_fabric_name(),
+ tpg->se_tpg_tfo->get_fabric_name(),
unpacked_lun, TRANSPORT_MAX_LUNS_PER_TPG-1,
- TPG_TFO(tpg)->tpg_get_tag(tpg));
+ tpg->se_tpg_tfo->tpg_get_tag(tpg));
return ERR_PTR(-EOVERFLOW);
}
@@ -762,8 +762,8 @@ struct se_lun *core_tpg_pre_addlun(
if (lun->lun_status == TRANSPORT_LUN_STATUS_ACTIVE) {
printk(KERN_ERR "TPG Logical Unit Number: %u is already active"
" on %s Target Portal Group: %u, ignoring request.\n",
- unpacked_lun, TPG_TFO(tpg)->get_fabric_name(),
- TPG_TFO(tpg)->tpg_get_tag(tpg));
+ unpacked_lun, tpg->se_tpg_tfo->get_fabric_name(),
+ tpg->se_tpg_tfo->tpg_get_tag(tpg));
spin_unlock(&tpg->tpg_lun_lock);
return ERR_PTR(-EINVAL);
}
@@ -810,9 +810,9 @@ struct se_lun *core_tpg_pre_dellun(
if (unpacked_lun > (TRANSPORT_MAX_LUNS_PER_TPG-1)) {
printk(KERN_ERR "%s LUN: %u exceeds TRANSPORT_MAX_LUNS_PER_TPG"
"-1: %u for Target Portal Group: %u\n",
- TPG_TFO(tpg)->get_fabric_name(), unpacked_lun,
+ tpg->se_tpg_tfo->get_fabric_name(), unpacked_lun,
TRANSPORT_MAX_LUNS_PER_TPG-1,
- TPG_TFO(tpg)->tpg_get_tag(tpg));
+ tpg->se_tpg_tfo->tpg_get_tag(tpg));
return ERR_PTR(-EOVERFLOW);
}
@@ -821,8 +821,8 @@ struct se_lun *core_tpg_pre_dellun(
if (lun->lun_status != TRANSPORT_LUN_STATUS_ACTIVE) {
printk(KERN_ERR "%s Logical Unit Number: %u is not active on"
" Target Portal Group: %u, ignoring request.\n",
- TPG_TFO(tpg)->get_fabric_name(), unpacked_lun,
- TPG_TFO(tpg)->tpg_get_tag(tpg));
+ tpg->se_tpg_tfo->get_fabric_name(), unpacked_lun,
+ tpg->se_tpg_tfo->tpg_get_tag(tpg));
spin_unlock(&tpg->tpg_lun_lock);
return ERR_PTR(-ENODEV);
}
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 101bc20..9c8641ea 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -456,9 +456,9 @@ void __transport_register_session(
* If the fabric module supports an ISID based TransportID,
* save this value in binary from the fabric I_T Nexus now.
*/
- if (TPG_TFO(se_tpg)->sess_get_initiator_sid != NULL) {
+ if (se_tpg->se_tpg_tfo->sess_get_initiator_sid != NULL) {
memset(&buf[0], 0, PR_REG_ISID_LEN);
- TPG_TFO(se_tpg)->sess_get_initiator_sid(se_sess,
+ se_tpg->se_tpg_tfo->sess_get_initiator_sid(se_sess,
&buf[0], PR_REG_ISID_LEN);
se_sess->sess_bin_isid = get_unaligned_be64(&buf[0]);
}
@@ -476,7 +476,7 @@ void __transport_register_session(
list_add_tail(&se_sess->sess_list, &se_tpg->tpg_sess_list);
printk(KERN_INFO "TARGET_CORE[%s]: Registered fabric_sess_ptr: %p\n",
- TPG_TFO(se_tpg)->get_fabric_name(), se_sess->fabric_sess_ptr);
+ se_tpg->se_tpg_tfo->get_fabric_name(), se_sess->fabric_sess_ptr);
}
EXPORT_SYMBOL(__transport_register_session);
@@ -550,7 +550,7 @@ void transport_deregister_session(struct se_session *se_sess)
if ((se_nacl)) {
spin_lock_bh(&se_tpg->acl_node_lock);
if (se_nacl->dynamic_node_acl) {
- if (!(TPG_TFO(se_tpg)->tpg_check_demo_mode_cache(
+ if (!(se_tpg->se_tpg_tfo->tpg_check_demo_mode_cache(
se_tpg))) {
list_del(&se_nacl->acl_list);
se_tpg->num_node_acls--;
@@ -558,7 +558,7 @@ void transport_deregister_session(struct se_session *se_sess)
core_tpg_wait_for_nacl_pr_ref(se_nacl);
core_free_device_list_for_node(se_nacl, se_tpg);
- TPG_TFO(se_tpg)->tpg_release_fabric_acl(se_tpg,
+ se_tpg->se_tpg_tfo->tpg_release_fabric_acl(se_tpg,
se_nacl);
spin_lock_bh(&se_tpg->acl_node_lock);
}
@@ -569,12 +569,12 @@ void transport_deregister_session(struct se_session *se_sess)
transport_free_session(se_sess);
printk(KERN_INFO "TARGET_CORE[%s]: Deregistered fabric_sess\n",
- TPG_TFO(se_tpg)->get_fabric_name());
+ se_tpg->se_tpg_tfo->get_fabric_name());
}
EXPORT_SYMBOL(transport_deregister_session);
/*
- * Called with T_TASK(cmd)->t_state_lock held.
+ * Called with cmd->t_task->t_state_lock held.
*/
static void transport_all_task_dev_remove_state(struct se_cmd *cmd)
{
@@ -582,10 +582,10 @@ static void transport_all_task_dev_remove_state(struct se_cmd *cmd)
struct se_task *task;
unsigned long flags;
- if (!T_TASK(cmd))
+ if (!cmd->t_task)
return;
- list_for_each_entry(task, &T_TASK(cmd)->t_task_list, t_list) {
+ list_for_each_entry(task, &cmd->t_task->t_task_list, t_list) {
dev = task->se_dev;
if (!(dev))
continue;
@@ -599,11 +599,11 @@ static void transport_all_task_dev_remove_state(struct se_cmd *cmd)
spin_lock_irqsave(&dev->execute_task_lock, flags);
list_del(&task->t_state_list);
DEBUG_TSTATE("Removed ITT: 0x%08x dev: %p task[%p]\n",
- CMD_TFO(cmd)->tfo_get_task_tag(cmd), dev, task);
+ cmd->se_tfo->tfo_get_task_tag(cmd), dev, task);
spin_unlock_irqrestore(&dev->execute_task_lock, flags);
atomic_set(&task->task_state_active, 0);
- atomic_dec(&T_TASK(cmd)->t_task_cdbs_ex_left);
+ atomic_dec(&cmd->t_task->t_task_cdbs_ex_left);
}
}
@@ -622,34 +622,34 @@ static int transport_cmd_check_stop(
{
unsigned long flags;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
/*
* Determine if IOCTL context caller in requesting the stopping of this
* command for LUN shutdown purposes.
*/
- if (atomic_read(&T_TASK(cmd)->transport_lun_stop)) {
- DEBUG_CS("%s:%d atomic_read(&T_TASK(cmd)->transport_lun_stop)"
+ if (atomic_read(&cmd->t_task->transport_lun_stop)) {
+ DEBUG_CS("%s:%d atomic_read(&cmd->t_task->transport_lun_stop)"
" == TRUE for ITT: 0x%08x\n", __func__, __LINE__,
- CMD_TFO(cmd)->get_task_tag(cmd));
+ cmd->se_tfo->get_task_tag(cmd));
cmd->deferred_t_state = cmd->t_state;
cmd->t_state = TRANSPORT_DEFERRED_CMD;
- atomic_set(&T_TASK(cmd)->t_transport_active, 0);
+ atomic_set(&cmd->t_task->t_transport_active, 0);
if (transport_off == 2)
transport_all_task_dev_remove_state(cmd);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
- complete(&T_TASK(cmd)->transport_lun_stop_comp);
+ complete(&cmd->t_task->transport_lun_stop_comp);
return 1;
}
/*
* Determine if frontend context caller is requesting the stopping of
* this command for frontend exceptions.
*/
- if (atomic_read(&T_TASK(cmd)->t_transport_stop)) {
- DEBUG_CS("%s:%d atomic_read(&T_TASK(cmd)->t_transport_stop) =="
+ if (atomic_read(&cmd->t_task->t_transport_stop)) {
+ DEBUG_CS("%s:%d atomic_read(&cmd->t_task->t_transport_stop) =="
" TRUE for ITT: 0x%08x\n", __func__, __LINE__,
- CMD_TFO(cmd)->get_task_tag(cmd));
+ cmd->se_tfo->get_task_tag(cmd));
cmd->deferred_t_state = cmd->t_state;
cmd->t_state = TRANSPORT_DEFERRED_CMD;
@@ -662,13 +662,13 @@ static int transport_cmd_check_stop(
*/
if (transport_off == 2)
cmd->se_lun = NULL;
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
- complete(&T_TASK(cmd)->t_transport_stop_comp);
+ complete(&cmd->t_task->t_transport_stop_comp);
return 1;
}
if (transport_off) {
- atomic_set(&T_TASK(cmd)->t_transport_active, 0);
+ atomic_set(&cmd->t_task->t_transport_active, 0);
if (transport_off == 2) {
transport_all_task_dev_remove_state(cmd);
/*
@@ -681,20 +681,20 @@ static int transport_cmd_check_stop(
* their internally allocated I/O reference now and
* struct se_cmd now.
*/
- if (CMD_TFO(cmd)->check_stop_free != NULL) {
+ if (cmd->se_tfo->check_stop_free != NULL) {
spin_unlock_irqrestore(
- &T_TASK(cmd)->t_state_lock, flags);
+ &cmd->t_task->t_state_lock, flags);
- CMD_TFO(cmd)->check_stop_free(cmd);
+ cmd->se_tfo->check_stop_free(cmd);
return 1;
}
}
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return 0;
} else if (t_state)
cmd->t_state = t_state;
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return 0;
}
@@ -706,30 +706,30 @@ static int transport_cmd_check_stop_to_fabric(struct se_cmd *cmd)
static void transport_lun_remove_cmd(struct se_cmd *cmd)
{
- struct se_lun *lun = SE_LUN(cmd);
+ struct se_lun *lun = cmd->se_lun;
unsigned long flags;
if (!lun)
return;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- if (!(atomic_read(&T_TASK(cmd)->transport_dev_active))) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ if (!(atomic_read(&cmd->t_task->transport_dev_active))) {
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
goto check_lun;
}
- atomic_set(&T_TASK(cmd)->transport_dev_active, 0);
+ atomic_set(&cmd->t_task->transport_dev_active, 0);
transport_all_task_dev_remove_state(cmd);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
check_lun:
spin_lock_irqsave(&lun->lun_cmd_lock, flags);
- if (atomic_read(&T_TASK(cmd)->transport_lun_active)) {
+ if (atomic_read(&cmd->t_task->transport_lun_active)) {
list_del(&cmd->se_lun_list);
- atomic_set(&T_TASK(cmd)->transport_lun_active, 0);
+ atomic_set(&cmd->t_task->transport_lun_active, 0);
#if 0
printk(KERN_INFO "Removed ITT: 0x%08x from LUN LIST[%d]\n"
- CMD_TFO(cmd)->get_task_tag(cmd), lun->unpacked_lun);
+ cmd->se_tfo->get_task_tag(cmd), lun->unpacked_lun);
#endif
}
spin_unlock_irqrestore(&lun->lun_cmd_lock, flags);
@@ -737,7 +737,7 @@ check_lun:
void transport_cmd_finish_abort(struct se_cmd *cmd, int remove)
{
- transport_remove_cmd_from_queue(cmd, &SE_DEV(cmd)->dev_queue_obj);
+ transport_remove_cmd_from_queue(cmd, &cmd->se_lun->lun_se_dev->dev_queue_obj);
transport_lun_remove_cmd(cmd);
if (transport_cmd_check_stop_to_fabric(cmd))
@@ -748,7 +748,7 @@ void transport_cmd_finish_abort(struct se_cmd *cmd, int remove)
void transport_cmd_finish_abort_tmr(struct se_cmd *cmd)
{
- transport_remove_cmd_from_queue(cmd, &SE_DEV(cmd)->dev_queue_obj);
+ transport_remove_cmd_from_queue(cmd, &cmd->se_lun->lun_se_dev->dev_queue_obj);
if (transport_cmd_check_stop_to_fabric(cmd))
return;
@@ -777,15 +777,15 @@ static int transport_add_cmd_to_queue(
qr->state = t_state;
if (t_state) {
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
cmd->t_state = t_state;
- atomic_set(&T_TASK(cmd)->t_transport_active, 1);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ atomic_set(&cmd->t_task->t_transport_active, 1);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
}
spin_lock_irqsave(&qobj->cmd_queue_lock, flags);
list_add_tail(&qr->qr_list, &qobj->qobj_list);
- atomic_inc(&T_TASK(cmd)->t_transport_queue_active);
+ atomic_inc(&cmd->t_task->t_transport_queue_active);
spin_unlock_irqrestore(&qobj->cmd_queue_lock, flags);
atomic_inc(&qobj->queue_cnt);
@@ -812,7 +812,7 @@ transport_get_qr_from_queue(struct se_queue_obj *qobj)
break;
if (qr->cmd)
- atomic_dec(&T_TASK(qr->cmd)->t_transport_queue_active);
+ atomic_dec(&qr->cmd->t_task->t_transport_queue_active);
list_del(&qr->qr_list);
atomic_dec(&qobj->queue_cnt);
@@ -828,7 +828,7 @@ static void transport_remove_cmd_from_queue(struct se_cmd *cmd,
unsigned long flags;
spin_lock_irqsave(&qobj->cmd_queue_lock, flags);
- if (!(atomic_read(&T_TASK(cmd)->t_transport_queue_active))) {
+ if (!(atomic_read(&cmd->t_task->t_transport_queue_active))) {
spin_unlock_irqrestore(&qobj->cmd_queue_lock, flags);
return;
}
@@ -837,17 +837,17 @@ static void transport_remove_cmd_from_queue(struct se_cmd *cmd,
if (qr->cmd != cmd)
continue;
- atomic_dec(&T_TASK(qr->cmd)->t_transport_queue_active);
+ atomic_dec(&qr->cmd->t_task->t_transport_queue_active);
atomic_dec(&qobj->queue_cnt);
list_del(&qr->qr_list);
kfree(qr);
}
spin_unlock_irqrestore(&qobj->cmd_queue_lock, flags);
- if (atomic_read(&T_TASK(cmd)->t_transport_queue_active)) {
+ if (atomic_read(&cmd->t_task->t_transport_queue_active)) {
printk(KERN_ERR "ITT: 0x%08x t_transport_queue_active: %d\n",
- CMD_TFO(cmd)->get_task_tag(cmd),
- atomic_read(&T_TASK(cmd)->t_transport_queue_active));
+ cmd->se_tfo->get_task_tag(cmd),
+ atomic_read(&cmd->t_task->t_transport_queue_active));
}
}
@@ -857,7 +857,7 @@ static void transport_remove_cmd_from_queue(struct se_cmd *cmd,
*/
void transport_complete_sync_cache(struct se_cmd *cmd, int good)
{
- struct se_task *task = list_entry(T_TASK(cmd)->t_task_list.next,
+ struct se_task *task = list_entry(cmd->t_task->t_task_list.next,
struct se_task, t_list);
if (good) {
@@ -866,7 +866,7 @@ void transport_complete_sync_cache(struct se_cmd *cmd, int good)
} else {
task->task_scsi_status = SAM_STAT_CHECK_CONDITION;
task->task_error_status = PYX_TRANSPORT_ILLEGAL_REQUEST;
- TASK_CMD(task)->transport_error_status =
+ task->task_se_cmd->transport_error_status =
PYX_TRANSPORT_ILLEGAL_REQUEST;
}
@@ -881,18 +881,18 @@ EXPORT_SYMBOL(transport_complete_sync_cache);
*/
void transport_complete_task(struct se_task *task, int success)
{
- struct se_cmd *cmd = TASK_CMD(task);
+ struct se_cmd *cmd = task->task_se_cmd;
struct se_device *dev = task->se_dev;
int t_state;
unsigned long flags;
#if 0
printk(KERN_INFO "task: %p CDB: 0x%02x obj_ptr: %p\n", task,
- T_TASK(cmd)->t_task_cdb[0], dev);
+ cmd->t_task->t_task_cdb[0], dev);
#endif
if (dev)
atomic_inc(&dev->depth_left);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
atomic_set(&task->task_active, 0);
/*
@@ -914,14 +914,14 @@ void transport_complete_task(struct se_task *task, int success)
*/
if (atomic_read(&task->task_stop)) {
/*
- * Decrement T_TASK(cmd)->t_se_count if this task had
+ * Decrement cmd->t_task->t_se_count if this task had
* previously thrown its timeout exception handler.
*/
if (atomic_read(&task->task_timeout)) {
- atomic_dec(&T_TASK(cmd)->t_se_count);
+ atomic_dec(&cmd->t_task->t_se_count);
atomic_set(&task->task_timeout, 0);
}
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
complete(&task->task_stop_comp);
return;
@@ -933,33 +933,33 @@ void transport_complete_task(struct se_task *task, int success)
*/
if (atomic_read(&task->task_timeout)) {
if (!(atomic_dec_and_test(
- &T_TASK(cmd)->t_task_cdbs_timeout_left))) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock,
+ &cmd->t_task->t_task_cdbs_timeout_left))) {
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock,
flags);
return;
}
t_state = TRANSPORT_COMPLETE_TIMEOUT;
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
transport_add_cmd_to_queue(cmd, t_state);
return;
}
- atomic_dec(&T_TASK(cmd)->t_task_cdbs_timeout_left);
+ atomic_dec(&cmd->t_task->t_task_cdbs_timeout_left);
/*
* Decrement the outstanding t_task_cdbs_left count. The last
* struct se_task from struct se_cmd will complete itself into the
* device queue depending upon int success.
*/
- if (!(atomic_dec_and_test(&T_TASK(cmd)->t_task_cdbs_left))) {
+ if (!(atomic_dec_and_test(&cmd->t_task->t_task_cdbs_left))) {
if (!success)
- T_TASK(cmd)->t_tasks_failed = 1;
+ cmd->t_task->t_tasks_failed = 1;
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return;
}
- if (!success || T_TASK(cmd)->t_tasks_failed) {
+ if (!success || cmd->t_task->t_tasks_failed) {
t_state = TRANSPORT_COMPLETE_FAILURE;
if (!task->task_error_status) {
task->task_error_status =
@@ -968,10 +968,10 @@ void transport_complete_task(struct se_task *task, int success)
PYX_TRANSPORT_UNKNOWN_SAM_OPCODE;
}
} else {
- atomic_set(&T_TASK(cmd)->t_transport_complete, 1);
+ atomic_set(&cmd->t_task->t_transport_complete, 1);
t_state = TRANSPORT_COMPLETE_OK;
}
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
transport_add_cmd_to_queue(cmd, t_state);
}
@@ -1054,7 +1054,7 @@ static void __transport_add_task_to_execute_queue(
atomic_set(&task->task_state_active, 1);
DEBUG_TSTATE("Added ITT: 0x%08x task[%p] to dev: %p\n",
- CMD_TFO(task->task_se_cmd)->get_task_tag(task->task_se_cmd),
+ task->task_se_cmd->se_tfo->get_task_tag(task->task_se_cmd),
task, dev);
}
@@ -1064,8 +1064,8 @@ static void transport_add_tasks_to_state_queue(struct se_cmd *cmd)
struct se_task *task;
unsigned long flags;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- list_for_each_entry(task, &T_TASK(cmd)->t_task_list, t_list) {
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ list_for_each_entry(task, &cmd->t_task->t_task_list, t_list) {
dev = task->se_dev;
if (atomic_read(&task->task_state_active))
@@ -1076,22 +1076,22 @@ static void transport_add_tasks_to_state_queue(struct se_cmd *cmd)
atomic_set(&task->task_state_active, 1);
DEBUG_TSTATE("Added ITT: 0x%08x task[%p] to dev: %p\n",
- CMD_TFO(task->task_se_cmd)->get_task_tag(
+ task->se_cmd->se_tfo->get_task_tag(
task->task_se_cmd), task, dev);
spin_unlock(&dev->execute_task_lock);
}
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
}
static void transport_add_tasks_from_cmd(struct se_cmd *cmd)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
struct se_task *task, *task_prev = NULL;
unsigned long flags;
spin_lock_irqsave(&dev->execute_task_lock, flags);
- list_for_each_entry(task, &T_TASK(cmd)->t_task_list, t_list) {
+ list_for_each_entry(task, &cmd->t_task->t_task_list, t_list) {
if (atomic_read(&task->task_execute_queue))
continue;
/*
@@ -1196,7 +1196,7 @@ void transport_dump_dev_state(
atomic_read(&dev->execute_tasks), atomic_read(&dev->depth_left),
dev->queue_depth);
*bl += sprintf(b + *bl, " SectorSize: %u MaxSectors: %u\n",
- DEV_ATTRIB(dev)->block_size, DEV_ATTRIB(dev)->max_sectors);
+ dev->se_sub_dev->se_dev_attrib.block_size, dev->se_sub_dev->se_dev_attrib.max_sectors);
*bl += sprintf(b + *bl, " ");
}
@@ -1224,8 +1224,8 @@ static void transport_release_all_cmds(struct se_device *dev)
printk(KERN_ERR "Releasing ITT: 0x%08x, i_state: %u,"
" t_state: %u directly\n",
- CMD_TFO(cmd)->get_task_tag(cmd),
- CMD_TFO(cmd)->get_cmd_state(cmd), t_state);
+ cmd->se_tfo->get_task_tag(cmd),
+ cmd->se_tfo->get_cmd_state(cmd), t_state);
transport_release_fe_cmd(cmd);
bug_out = 1;
@@ -1501,20 +1501,20 @@ static void core_setup_task_attr_emulation(struct se_device *dev)
* This is currently not available in upsream Linux/SCSI Target
* mode code, and is assumed to be disabled while using TCM/pSCSI.
*/
- if (TRANSPORT(dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
+ if (dev->transport->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV) {
dev->dev_task_attr_type = SAM_TASK_ATTR_PASSTHROUGH;
return;
}
dev->dev_task_attr_type = SAM_TASK_ATTR_EMULATED;
DEBUG_STA("%s: Using SAM_TASK_ATTR_EMULATED for SPC: 0x%02x"
- " device\n", TRANSPORT(dev)->name,
- TRANSPORT(dev)->get_device_rev(dev));
+ " device\n", dev->transport->name,
+ dev->transport->get_device_rev(dev));
}
static void scsi_dump_inquiry(struct se_device *dev)
{
- struct t10_wwn *wwn = DEV_T10_WWN(dev);
+ struct t10_wwn *wwn = &dev->se_sub_dev->t10_wwn;
int i, device_type;
/*
* Print Linux/SCSI style INQUIRY formatting to the kernel ring buffer
@@ -1542,10 +1542,10 @@ static void scsi_dump_inquiry(struct se_device *dev)
printk("\n");
- device_type = TRANSPORT(dev)->get_device_type(dev);
+ device_type = dev->transport->get_device_type(dev);
printk(" Type: %s ", scsi_device_type(device_type));
printk(" ANSI SCSI revision: %02x\n",
- TRANSPORT(dev)->get_device_rev(dev));
+ dev->transport->get_device_rev(dev));
}
struct se_device *transport_add_device_to_core_hba(
@@ -1629,10 +1629,10 @@ struct se_device *transport_add_device_to_core_hba(
* Startup the struct se_device processing thread
*/
dev->process_thread = kthread_run(transport_processing_thread, dev,
- "LIO_%s", TRANSPORT(dev)->name);
+ "LIO_%s", dev->transport->name);
if (IS_ERR(dev->process_thread)) {
printk(KERN_ERR "Unable to create kthread: LIO_%s\n",
- TRANSPORT(dev)->name);
+ dev->transport->name);
goto out;
}
@@ -1644,16 +1644,16 @@ struct se_device *transport_add_device_to_core_hba(
* originals once back into DEV_T10_WWN(dev) for the virtual device
* setup.
*/
- if (TRANSPORT(dev)->transport_type != TRANSPORT_PLUGIN_PHBA_PDEV) {
+ if (dev->transport->transport_type != TRANSPORT_PLUGIN_PHBA_PDEV) {
if (!(inquiry_prod) || !(inquiry_prod)) {
printk(KERN_ERR "All non TCM/pSCSI plugins require"
" INQUIRY consts\n");
goto out;
}
- strncpy(&DEV_T10_WWN(dev)->vendor[0], "LIO-ORG", 8);
- strncpy(&DEV_T10_WWN(dev)->model[0], inquiry_prod, 16);
- strncpy(&DEV_T10_WWN(dev)->revision[0], inquiry_rev, 4);
+ strncpy(&dev->se_sub_dev->t10_wwn.vendor[0], "LIO-ORG", 8);
+ strncpy(&dev->se_sub_dev->t10_wwn.model[0], inquiry_prod, 16);
+ strncpy(&dev->se_sub_dev->t10_wwn.revision[0], inquiry_rev, 4);
}
scsi_dump_inquiry(dev);
@@ -1706,7 +1706,7 @@ transport_generic_get_task(struct se_cmd *cmd,
enum dma_data_direction data_direction)
{
struct se_task *task;
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
unsigned long flags;
task = dev->transport->alloc_task(cmd);
@@ -1719,14 +1719,14 @@ transport_generic_get_task(struct se_cmd *cmd,
INIT_LIST_HEAD(&task->t_execute_list);
INIT_LIST_HEAD(&task->t_state_list);
init_completion(&task->task_stop_comp);
- task->task_no = T_TASK(cmd)->t_tasks_no++;
+ task->task_no = cmd->t_task->t_tasks_no++;
task->task_se_cmd = cmd;
task->se_dev = dev;
task->task_data_direction = data_direction;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- list_add_tail(&task->t_list, &T_TASK(cmd)->t_task_list);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ list_add_tail(&task->t_list, &cmd->t_task->t_task_list);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return task;
}
@@ -1735,7 +1735,7 @@ static int transport_generic_cmd_sequencer(struct se_cmd *, unsigned char *);
void transport_device_setup_cmd(struct se_cmd *cmd)
{
- cmd->se_dev = SE_LUN(cmd)->lun_se_dev;
+ cmd->se_dev = cmd->se_lun->lun_se_dev;
}
EXPORT_SYMBOL(transport_device_setup_cmd);
@@ -1760,12 +1760,12 @@ void transport_init_se_cmd(
*/
cmd->t_task = &cmd->t_task_backstore;
- INIT_LIST_HEAD(&T_TASK(cmd)->t_task_list);
- init_completion(&T_TASK(cmd)->transport_lun_fe_stop_comp);
- init_completion(&T_TASK(cmd)->transport_lun_stop_comp);
- init_completion(&T_TASK(cmd)->t_transport_stop_comp);
- spin_lock_init(&T_TASK(cmd)->t_state_lock);
- atomic_set(&T_TASK(cmd)->transport_dev_active, 1);
+ INIT_LIST_HEAD(&cmd->t_task->t_task_list);
+ init_completion(&cmd->t_task->transport_lun_fe_stop_comp);
+ init_completion(&cmd->t_task->transport_lun_stop_comp);
+ init_completion(&cmd->t_task->t_transport_stop_comp);
+ spin_lock_init(&cmd->t_task->t_state_lock);
+ atomic_set(&cmd->t_task->transport_dev_active, 1);
cmd->se_tfo = tfo;
cmd->se_sess = se_sess;
@@ -1782,7 +1782,7 @@ static int transport_check_alloc_task_attr(struct se_cmd *cmd)
* Check if SAM Task Attribute emulation is enabled for this
* struct se_device storage object
*/
- if (SE_DEV(cmd)->dev_task_attr_type != SAM_TASK_ATTR_EMULATED)
+ if (cmd->se_lun->lun_se_dev->dev_task_attr_type != SAM_TASK_ATTR_EMULATED)
return 0;
if (cmd->sam_task_attr == MSG_ACA_TAG) {
@@ -1794,7 +1794,7 @@ static int transport_check_alloc_task_attr(struct se_cmd *cmd)
* Used to determine when ORDERED commands should go from
* Dormant to Active status.
*/
- cmd->se_ordered_id = atomic_inc_return(&SE_DEV(cmd)->dev_ordered_id);
+ cmd->se_ordered_id = atomic_inc_return(&cmd->se_lun->lun_se_dev->dev_ordered_id);
smp_mb__after_atomic_inc();
DEBUG_STA("Allocated se_ordered_id: %u for Task Attr: 0x%02x on %s\n",
cmd->se_ordered_id, cmd->sam_task_attr,
@@ -1810,8 +1810,8 @@ void transport_free_se_cmd(
/*
* Check and free any extended CDB buffer that was allocated
*/
- if (T_TASK(se_cmd)->t_task_cdb != T_TASK(se_cmd)->__t_task_cdb)
- kfree(T_TASK(se_cmd)->t_task_cdb);
+ if (se_cmd->t_task->t_task_cdb != se_cmd->t_task->__t_task_cdb)
+ kfree(se_cmd->t_task->t_task_cdb);
}
EXPORT_SYMBOL(transport_free_se_cmd);
@@ -1850,26 +1850,26 @@ int transport_generic_allocate_tasks(
* allocate the additional extended CDB buffer now.. Otherwise
* setup the pointer from __t_task_cdb to t_task_cdb.
*/
- if (scsi_command_size(cdb) > sizeof(T_TASK(cmd)->__t_task_cdb)) {
- T_TASK(cmd)->t_task_cdb = kzalloc(scsi_command_size(cdb),
+ if (scsi_command_size(cdb) > sizeof(cmd->t_task->__t_task_cdb)) {
+ cmd->t_task->t_task_cdb = kzalloc(scsi_command_size(cdb),
GFP_KERNEL);
- if (!(T_TASK(cmd)->t_task_cdb)) {
- printk(KERN_ERR "Unable to allocate T_TASK(cmd)->t_task_cdb"
- " %u > sizeof(T_TASK(cmd)->__t_task_cdb): %lu ops\n",
+ if (!(cmd->t_task->t_task_cdb)) {
+ printk(KERN_ERR "Unable to allocate cmd->t_task->t_task_cdb"
+ " %u > sizeof(cmd->t_task->__t_task_cdb): %lu ops\n",
scsi_command_size(cdb),
- (unsigned long)sizeof(T_TASK(cmd)->__t_task_cdb));
+ (unsigned long)sizeof(cmd->t_task->__t_task_cdb));
return -ENOMEM;
}
} else
- T_TASK(cmd)->t_task_cdb = &T_TASK(cmd)->__t_task_cdb[0];
+ cmd->t_task->t_task_cdb = &cmd->t_task->__t_task_cdb[0];
/*
- * Copy the original CDB into T_TASK(cmd).
+ * Copy the original CDB into cmd->t_task.
*/
- memcpy(T_TASK(cmd)->t_task_cdb, cdb, scsi_command_size(cdb));
+ memcpy(cmd->t_task->t_task_cdb, cdb, scsi_command_size(cdb));
/*
* Setup the received CDB based on SCSI defined opcodes and
* perform unit attention, persistent reservations and ALUA
- * checks for virtual device backends. The T_TASK(cmd)->t_task_cdb
+ * checks for virtual device backends. The cmd->t_task->t_task_cdb
* pointer is expected to be setup before we reach this point.
*/
ret = transport_generic_cmd_sequencer(cmd, cdb);
@@ -1898,9 +1898,9 @@ EXPORT_SYMBOL(transport_generic_allocate_tasks);
int transport_generic_handle_cdb(
struct se_cmd *cmd)
{
- if (!SE_LUN(cmd)) {
+ if (!cmd->se_lun) {
dump_stack();
- printk(KERN_ERR "SE_LUN(cmd) is NULL\n");
+ printk(KERN_ERR "cmd->se_lun is NULL\n");
return -EINVAL;
}
@@ -1917,9 +1917,9 @@ EXPORT_SYMBOL(transport_generic_handle_cdb);
int transport_generic_handle_cdb_map(
struct se_cmd *cmd)
{
- if (!SE_LUN(cmd)) {
+ if (!cmd->se_lun) {
dump_stack();
- printk(KERN_ERR "SE_LUN(cmd) is NULL\n");
+ printk(KERN_ERR "cmd->se_lun is NULL\n");
return -EINVAL;
}
@@ -1990,14 +1990,14 @@ static int transport_stop_tasks_for_cmd(struct se_cmd *cmd)
int ret = 0;
DEBUG_TS("ITT[0x%08x] - Stopping tasks\n",
- CMD_TFO(cmd)->get_task_tag(cmd));
+ cmd->se_tfo->get_task_tag(cmd));
/*
* No tasks remain in the execution queue
*/
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
list_for_each_entry_safe(task, task_tmp,
- &T_TASK(cmd)->t_task_list, t_list) {
+ &cmd->t_task->t_task_list, t_list) {
DEBUG_TS("task_no[%d] - Processing task %p\n",
task->task_no, task);
/*
@@ -2006,14 +2006,14 @@ static int transport_stop_tasks_for_cmd(struct se_cmd *cmd)
*/
if (!atomic_read(&task->task_sent) &&
!atomic_read(&task->task_active)) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock,
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock,
flags);
transport_remove_task_from_execute_queue(task,
task->se_dev);
DEBUG_TS("task_no[%d] - Removed from execute queue\n",
task->task_no);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
continue;
}
@@ -2023,7 +2023,7 @@ static int transport_stop_tasks_for_cmd(struct se_cmd *cmd)
*/
if (atomic_read(&task->task_active)) {
atomic_set(&task->task_stop, 1);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock,
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock,
flags);
DEBUG_TS("task_no[%d] - Waiting to complete\n",
@@ -2032,8 +2032,8 @@ static int transport_stop_tasks_for_cmd(struct se_cmd *cmd)
DEBUG_TS("task_no[%d] - Stopped successfully\n",
task->task_no);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- atomic_dec(&T_TASK(cmd)->t_task_cdbs_left);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ atomic_dec(&cmd->t_task->t_task_cdbs_left);
atomic_set(&task->task_active, 0);
atomic_set(&task->task_stop, 0);
@@ -2044,7 +2044,7 @@ static int transport_stop_tasks_for_cmd(struct se_cmd *cmd)
__transport_stop_task_timer(task, &flags);
}
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return ret;
}
@@ -2059,23 +2059,23 @@ static void transport_generic_request_failure(
int sc)
{
DEBUG_GRF("-----[ Storage Engine Exception for cmd: %p ITT: 0x%08x"
- " CDB: 0x%02x\n", cmd, CMD_TFO(cmd)->get_task_tag(cmd),
- T_TASK(cmd)->t_task_cdb[0]);
+ " CDB: 0x%02x\n", cmd, cmd->se_tfo->get_task_tag(cmd),
+ cmd->t_task->t_task_cdb[0]);
DEBUG_GRF("-----[ i_state: %d t_state/def_t_state:"
" %d/%d transport_error_status: %d\n",
- CMD_TFO(cmd)->get_cmd_state(cmd),
+ cmd->se_tfo->get_cmd_state(cmd),
cmd->t_state, cmd->deferred_t_state,
cmd->transport_error_status);
DEBUG_GRF("-----[ t_task_cdbs: %d t_task_cdbs_left: %d"
" t_task_cdbs_sent: %d t_task_cdbs_ex_left: %d --"
" t_transport_active: %d t_transport_stop: %d"
- " t_transport_sent: %d\n", T_TASK(cmd)->t_task_cdbs,
- atomic_read(&T_TASK(cmd)->t_task_cdbs_left),
- atomic_read(&T_TASK(cmd)->t_task_cdbs_sent),
- atomic_read(&T_TASK(cmd)->t_task_cdbs_ex_left),
- atomic_read(&T_TASK(cmd)->t_transport_active),
- atomic_read(&T_TASK(cmd)->t_transport_stop),
- atomic_read(&T_TASK(cmd)->t_transport_sent));
+ " t_transport_sent: %d\n", cmd->t_task->t_task_cdbs,
+ atomic_read(&cmd->t_task->t_task_cdbs_left),
+ atomic_read(&cmd->t_task->t_task_cdbs_sent),
+ atomic_read(&cmd->t_task->t_task_cdbs_ex_left),
+ atomic_read(&cmd->t_task->t_transport_active),
+ atomic_read(&cmd->t_task->t_transport_stop),
+ atomic_read(&cmd->t_task->t_transport_sent));
transport_stop_all_task_timers(cmd);
@@ -2113,8 +2113,8 @@ static void transport_generic_request_failure(
* we force this session to fall back to session
* recovery.
*/
- CMD_TFO(cmd)->fall_back_to_erl0(cmd->se_sess);
- CMD_TFO(cmd)->stop_session(cmd->se_sess, 0, 0);
+ cmd->se_tfo->fall_back_to_erl0(cmd->se_sess);
+ cmd->se_tfo->stop_session(cmd->se_sess, 0, 0);
goto check_stop;
case PYX_TRANSPORT_LU_COMM_FAILURE:
@@ -2142,13 +2142,13 @@ static void transport_generic_request_failure(
*
* See spc4r17, section 7.4.6 Control Mode Page, Table 349
*/
- if (SE_SESS(cmd) &&
- DEV_ATTRIB(cmd->se_dev)->emulate_ua_intlck_ctrl == 2)
- core_scsi3_ua_allocate(SE_SESS(cmd)->se_node_acl,
+ if (cmd->se_sess &&
+ cmd->se_dev->se_sub_dev->se_dev_attrib.emulate_ua_intlck_ctrl == 2)
+ core_scsi3_ua_allocate(cmd->se_sess->se_node_acl,
cmd->orig_fe_lun, 0x2C,
ASCQ_2CH_PREVIOUS_RESERVATION_CONFLICT_STATUS);
- CMD_TFO(cmd)->queue_status(cmd);
+ cmd->se_tfo->queue_status(cmd);
goto check_stop;
case PYX_TRANSPORT_USE_SENSE_REASON:
/*
@@ -2157,7 +2157,7 @@ static void transport_generic_request_failure(
break;
default:
printk(KERN_ERR "Unknown transport error for CDB 0x%02x: %d\n",
- T_TASK(cmd)->t_task_cdb[0],
+ cmd->t_task->t_task_cdb[0],
cmd->transport_error_status);
cmd->scsi_sense_reason = TCM_UNSUPPORTED_SCSI_OPCODE;
break;
@@ -2178,19 +2178,19 @@ static void transport_direct_request_timeout(struct se_cmd *cmd)
{
unsigned long flags;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- if (!(atomic_read(&T_TASK(cmd)->t_transport_timeout))) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ if (!(atomic_read(&cmd->t_task->t_transport_timeout))) {
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return;
}
- if (atomic_read(&T_TASK(cmd)->t_task_cdbs_timeout_left)) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ if (atomic_read(&cmd->t_task->t_task_cdbs_timeout_left)) {
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return;
}
- atomic_sub(atomic_read(&T_TASK(cmd)->t_transport_timeout),
- &T_TASK(cmd)->t_se_count);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ atomic_sub(atomic_read(&cmd->t_task->t_transport_timeout),
+ &cmd->t_task->t_se_count);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
}
static void transport_generic_request_timeout(struct se_cmd *cmd)
@@ -2198,16 +2198,16 @@ static void transport_generic_request_timeout(struct se_cmd *cmd)
unsigned long flags;
/*
- * Reset T_TASK(cmd)->t_se_count to allow transport_generic_remove()
+ * Reset cmd->t_task->t_se_count to allow transport_generic_remove()
* to allow last call to free memory resources.
*/
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- if (atomic_read(&T_TASK(cmd)->t_transport_timeout) > 1) {
- int tmp = (atomic_read(&T_TASK(cmd)->t_transport_timeout) - 1);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ if (atomic_read(&cmd->t_task->t_transport_timeout) > 1) {
+ int tmp = (atomic_read(&cmd->t_task->t_transport_timeout) - 1);
- atomic_sub(tmp, &T_TASK(cmd)->t_se_count);
+ atomic_sub(tmp, &cmd->t_task->t_se_count);
}
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
transport_generic_remove(cmd, 0, 0);
}
@@ -2223,8 +2223,8 @@ transport_generic_allocate_buf(struct se_cmd *cmd, u32 data_length)
return -ENOMEM;
}
- T_TASK(cmd)->t_tasks_se_num = 0;
- T_TASK(cmd)->t_task_buf = buf;
+ cmd->t_task->t_tasks_se_num = 0;
+ cmd->t_task->t_task_buf = buf;
return 0;
}
@@ -2266,9 +2266,9 @@ static void transport_set_supported_SAM_opcode(struct se_cmd *se_cmd)
{
unsigned long flags;
- spin_lock_irqsave(&T_TASK(se_cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&se_cmd->t_task->t_state_lock, flags);
se_cmd->se_cmd_flags |= SCF_SUPPORTED_SAM_OPCODE;
- spin_unlock_irqrestore(&T_TASK(se_cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&se_cmd->t_task->t_state_lock, flags);
}
/*
@@ -2277,14 +2277,14 @@ static void transport_set_supported_SAM_opcode(struct se_cmd *se_cmd)
static void transport_task_timeout_handler(unsigned long data)
{
struct se_task *task = (struct se_task *)data;
- struct se_cmd *cmd = TASK_CMD(task);
+ struct se_cmd *cmd = task->task_se_cmd;
unsigned long flags;
DEBUG_TT("transport task timeout fired! task: %p cmd: %p\n", task, cmd);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
if (task->task_flags & TF_STOP) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return;
}
task->task_flags &= ~TF_RUNNING;
@@ -2295,13 +2295,13 @@ static void transport_task_timeout_handler(unsigned long data)
if (!(atomic_read(&task->task_active))) {
DEBUG_TT("transport task: %p cmd: %p timeout task_active"
" == 0\n", task, cmd);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return;
}
- atomic_inc(&T_TASK(cmd)->t_se_count);
- atomic_inc(&T_TASK(cmd)->t_transport_timeout);
- T_TASK(cmd)->t_tasks_failed = 1;
+ atomic_inc(&cmd->t_task->t_se_count);
+ atomic_inc(&cmd->t_task->t_transport_timeout);
+ cmd->t_task->t_tasks_failed = 1;
atomic_set(&task->task_timeout, 1);
task->task_error_status = PYX_TRANSPORT_TASK_TIMEOUT;
@@ -2310,28 +2310,28 @@ static void transport_task_timeout_handler(unsigned long data)
if (atomic_read(&task->task_stop)) {
DEBUG_TT("transport task: %p cmd: %p timeout task_stop"
" == 1\n", task, cmd);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
complete(&task->task_stop_comp);
return;
}
- if (!(atomic_dec_and_test(&T_TASK(cmd)->t_task_cdbs_left))) {
+ if (!(atomic_dec_and_test(&cmd->t_task->t_task_cdbs_left))) {
DEBUG_TT("transport task: %p cmd: %p timeout non zero"
" t_task_cdbs_left\n", task, cmd);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return;
}
DEBUG_TT("transport task: %p cmd: %p timeout ZERO t_task_cdbs_left\n",
task, cmd);
cmd->t_state = TRANSPORT_COMPLETE_FAILURE;
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
transport_add_cmd_to_queue(cmd, TRANSPORT_COMPLETE_FAILURE);
}
/*
- * Called with T_TASK(cmd)->t_state_lock held.
+ * Called with cmd->t_task->t_state_lock held.
*/
static void transport_start_task_timer(struct se_task *task)
{
@@ -2343,7 +2343,7 @@ static void transport_start_task_timer(struct se_task *task)
/*
* If the task_timeout is disabled, exit now.
*/
- timeout = DEV_ATTRIB(dev)->task_timeout;
+ timeout = dev->se_sub_dev->se_dev_attrib.task_timeout;
if (!(timeout))
return;
@@ -2361,21 +2361,21 @@ static void transport_start_task_timer(struct se_task *task)
}
/*
- * Called with spin_lock_irq(&T_TASK(cmd)->t_state_lock) held.
+ * Called with spin_lock_irq(&cmd->t_task->t_state_lock) held.
*/
void __transport_stop_task_timer(struct se_task *task, unsigned long *flags)
{
- struct se_cmd *cmd = TASK_CMD(task);
+ struct se_cmd *cmd = task->task_se_cmd;
if (!(task->task_flags & TF_RUNNING))
return;
task->task_flags |= TF_STOP;
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, *flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, *flags);
del_timer_sync(&task->task_timer);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, *flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, *flags);
task->task_flags &= ~TF_RUNNING;
task->task_flags &= ~TF_STOP;
}
@@ -2385,11 +2385,11 @@ static void transport_stop_all_task_timers(struct se_cmd *cmd)
struct se_task *task = NULL, *task_tmp;
unsigned long flags;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
list_for_each_entry_safe(task, task_tmp,
- &T_TASK(cmd)->t_task_list, t_list)
+ &cmd->t_task->t_task_list, t_list)
__transport_stop_task_timer(task, &flags);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
}
static inline int transport_tcq_window_closed(struct se_device *dev)
@@ -2413,45 +2413,45 @@ static inline int transport_tcq_window_closed(struct se_device *dev)
*/
static inline int transport_execute_task_attr(struct se_cmd *cmd)
{
- if (SE_DEV(cmd)->dev_task_attr_type != SAM_TASK_ATTR_EMULATED)
+ if (cmd->se_lun->lun_se_dev->dev_task_attr_type != SAM_TASK_ATTR_EMULATED)
return 1;
/*
* Check for the existence of HEAD_OF_QUEUE, and if true return 1
* to allow the passed struct se_cmd list of tasks to the front of the list.
*/
if (cmd->sam_task_attr == MSG_HEAD_TAG) {
- atomic_inc(&SE_DEV(cmd)->dev_hoq_count);
+ atomic_inc(&cmd->se_lun->lun_se_dev->dev_hoq_count);
smp_mb__after_atomic_inc();
DEBUG_STA("Added HEAD_OF_QUEUE for CDB:"
" 0x%02x, se_ordered_id: %u\n",
- T_TASK(cmd)->t_task_cdb[0],
+ cmd->t_task->t_task_cdb[0],
cmd->se_ordered_id);
return 1;
} else if (cmd->sam_task_attr == MSG_ORDERED_TAG) {
- spin_lock(&SE_DEV(cmd)->ordered_cmd_lock);
+ spin_lock(&cmd->se_lun->lun_se_dev->ordered_cmd_lock);
list_add_tail(&cmd->se_ordered_list,
- &SE_DEV(cmd)->ordered_cmd_list);
- spin_unlock(&SE_DEV(cmd)->ordered_cmd_lock);
+ &cmd->se_lun->lun_se_dev->ordered_cmd_list);
+ spin_unlock(&cmd->se_lun->lun_se_dev->ordered_cmd_lock);
- atomic_inc(&SE_DEV(cmd)->dev_ordered_sync);
+ atomic_inc(&cmd->se_lun->lun_se_dev->dev_ordered_sync);
smp_mb__after_atomic_inc();
DEBUG_STA("Added ORDERED for CDB: 0x%02x to ordered"
" list, se_ordered_id: %u\n",
- T_TASK(cmd)->t_task_cdb[0],
+ cmd->t_task->t_task_cdb[0],
cmd->se_ordered_id);
/*
* Add ORDERED command to tail of execution queue if
* no other older commands exist that need to be
* completed first.
*/
- if (!(atomic_read(&SE_DEV(cmd)->simple_cmds)))
+ if (!(atomic_read(&cmd->se_lun->lun_se_dev->simple_cmds)))
return 1;
} else {
/*
* For SIMPLE and UNTAGGED Task Attribute commands
*/
- atomic_inc(&SE_DEV(cmd)->simple_cmds);
+ atomic_inc(&cmd->se_lun->lun_se_dev->simple_cmds);
smp_mb__after_atomic_inc();
}
/*
@@ -2459,20 +2459,20 @@ static inline int transport_execute_task_attr(struct se_cmd *cmd)
* add the dormant task(s) built for the passed struct se_cmd to the
* execution queue and become in Active state for this struct se_device.
*/
- if (atomic_read(&SE_DEV(cmd)->dev_ordered_sync) != 0) {
+ if (atomic_read(&cmd->se_lun->lun_se_dev->dev_ordered_sync) != 0) {
/*
* Otherwise, add cmd w/ tasks to delayed cmd queue that
* will be drained upon completion of HEAD_OF_QUEUE task.
*/
- spin_lock(&SE_DEV(cmd)->delayed_cmd_lock);
+ spin_lock(&cmd->se_lun->lun_se_dev->delayed_cmd_lock);
cmd->se_cmd_flags |= SCF_DELAYED_CMD_FROM_SAM_ATTR;
list_add_tail(&cmd->se_delayed_list,
- &SE_DEV(cmd)->delayed_cmd_list);
- spin_unlock(&SE_DEV(cmd)->delayed_cmd_lock);
+ &cmd->se_lun->lun_se_dev->delayed_cmd_list);
+ spin_unlock(&cmd->se_lun->lun_se_dev->delayed_cmd_lock);
DEBUG_STA("Added CDB: 0x%02x Task Attr: 0x%02x to"
" delayed CMD list, se_ordered_id: %u\n",
- T_TASK(cmd)->t_task_cdb[0], cmd->sam_task_attr,
+ cmd->t_task->t_task_cdb[0], cmd->sam_task_attr,
cmd->se_ordered_id);
/*
* Return zero to let transport_execute_tasks() know
@@ -2527,7 +2527,7 @@ static int transport_execute_tasks(struct se_cmd *cmd)
* storage object.
*/
execute_tasks:
- __transport_execute_tasks(SE_DEV(cmd));
+ __transport_execute_tasks(cmd->se_lun->lun_se_dev);
return 0;
}
@@ -2563,19 +2563,19 @@ check_depth:
atomic_dec(&dev->depth_left);
- cmd = TASK_CMD(task);
+ cmd = task->task_se_cmd;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
atomic_set(&task->task_active, 1);
atomic_set(&task->task_sent, 1);
- atomic_inc(&T_TASK(cmd)->t_task_cdbs_sent);
+ atomic_inc(&cmd->t_task->t_task_cdbs_sent);
- if (atomic_read(&T_TASK(cmd)->t_task_cdbs_sent) ==
- T_TASK(cmd)->t_task_cdbs)
+ if (atomic_read(&cmd->t_task->t_task_cdbs_sent) ==
+ cmd->t_task->t_task_cdbs)
atomic_set(&cmd->transport_sent, 1);
transport_start_task_timer(task);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
/*
* The struct se_cmd->transport_emulate_cdb() function pointer is used
* to grab REPORT_LUNS and other CDBs we want to handle before they hit the
@@ -2613,11 +2613,11 @@ check_depth:
* call ->do_task() directly and let the underlying TCM subsystem plugin
* code handle the CDB emulation.
*/
- if ((TRANSPORT(dev)->transport_type != TRANSPORT_PLUGIN_PHBA_PDEV) &&
- (!(TASK_CMD(task)->se_cmd_flags & SCF_SCSI_DATA_SG_IO_CDB)))
+ if ((dev->transport->transport_type != TRANSPORT_PLUGIN_PHBA_PDEV) &&
+ (!(task->task_se_cmd->se_cmd_flags & SCF_SCSI_DATA_SG_IO_CDB)))
error = transport_emulate_control_cdb(task);
else
- error = TRANSPORT(dev)->do_task(task);
+ error = dev->transport->do_task(task);
if (error != 0) {
cmd->transport_error_status = error;
@@ -2640,12 +2640,12 @@ void transport_new_cmd_failure(struct se_cmd *se_cmd)
* Any unsolicited data will get dumped for failed command inside of
* the fabric plugin
*/
- spin_lock_irqsave(&T_TASK(se_cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&se_cmd->t_task->t_state_lock, flags);
se_cmd->se_cmd_flags |= SCF_SE_CMD_FAILED;
se_cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
- spin_unlock_irqrestore(&T_TASK(se_cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&se_cmd->t_task->t_state_lock, flags);
- CMD_TFO(se_cmd)->new_cmd_failure(se_cmd);
+ se_cmd->se_tfo->new_cmd_failure(se_cmd);
}
static void transport_nop_wait_for_tasks(struct se_cmd *, int, int);
@@ -2655,7 +2655,7 @@ static inline u32 transport_get_sectors_6(
struct se_cmd *cmd,
int *ret)
{
- struct se_device *dev = SE_LUN(cmd)->lun_se_dev;
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
/*
* Assume TYPE_DISK for non struct se_device objects.
@@ -2667,7 +2667,7 @@ static inline u32 transport_get_sectors_6(
/*
* Use 24-bit allocation length for TYPE_TAPE.
*/
- if (TRANSPORT(dev)->get_device_type(dev) == TYPE_TAPE)
+ if (dev->transport->get_device_type(dev) == TYPE_TAPE)
return (u32)(cdb[2] << 16) + (cdb[3] << 8) + cdb[4];
/*
@@ -2683,7 +2683,7 @@ static inline u32 transport_get_sectors_10(
struct se_cmd *cmd,
int *ret)
{
- struct se_device *dev = SE_LUN(cmd)->lun_se_dev;
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
/*
* Assume TYPE_DISK for non struct se_device objects.
@@ -2695,7 +2695,7 @@ static inline u32 transport_get_sectors_10(
/*
* XXX_10 is not defined in SSC, throw an exception
*/
- if (TRANSPORT(dev)->get_device_type(dev) == TYPE_TAPE) {
+ if (dev->transport->get_device_type(dev) == TYPE_TAPE) {
*ret = -EINVAL;
return 0;
}
@@ -2713,7 +2713,7 @@ static inline u32 transport_get_sectors_12(
struct se_cmd *cmd,
int *ret)
{
- struct se_device *dev = SE_LUN(cmd)->lun_se_dev;
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
/*
* Assume TYPE_DISK for non struct se_device objects.
@@ -2725,7 +2725,7 @@ static inline u32 transport_get_sectors_12(
/*
* XXX_12 is not defined in SSC, throw an exception
*/
- if (TRANSPORT(dev)->get_device_type(dev) == TYPE_TAPE) {
+ if (dev->transport->get_device_type(dev) == TYPE_TAPE) {
*ret = -EINVAL;
return 0;
}
@@ -2743,7 +2743,7 @@ static inline u32 transport_get_sectors_16(
struct se_cmd *cmd,
int *ret)
{
- struct se_device *dev = SE_LUN(cmd)->lun_se_dev;
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
/*
* Assume TYPE_DISK for non struct se_device objects.
@@ -2755,7 +2755,7 @@ static inline u32 transport_get_sectors_16(
/*
* Use 24-bit allocation length for TYPE_TAPE.
*/
- if (TRANSPORT(dev)->get_device_type(dev) == TYPE_TAPE)
+ if (dev->transport->get_device_type(dev) == TYPE_TAPE)
return (u32)(cdb[12] << 16) + (cdb[13] << 8) + cdb[14];
type_disk:
@@ -2785,21 +2785,21 @@ static inline u32 transport_get_size(
unsigned char *cdb,
struct se_cmd *cmd)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
- if (TRANSPORT(dev)->get_device_type(dev) == TYPE_TAPE) {
+ if (dev->transport->get_device_type(dev) == TYPE_TAPE) {
if (cdb[1] & 1) { /* sectors */
- return DEV_ATTRIB(dev)->block_size * sectors;
+ return dev->se_sub_dev->se_dev_attrib.block_size * sectors;
} else /* bytes */
return sectors;
}
#if 0
printk(KERN_INFO "Returning block_size: %u, sectors: %u == %u for"
- " %s object\n", DEV_ATTRIB(dev)->block_size, sectors,
- DEV_ATTRIB(dev)->block_size * sectors,
- TRANSPORT(dev)->name);
+ " %s object\n", dev->se_sub_dev->se_dev_attrib.block_size, sectors,
+ dev->se_sub_dev->se_dev_attrib.block_size * sectors,
+ dev->transport->name);
#endif
- return DEV_ATTRIB(dev)->block_size * sectors;
+ return dev->se_sub_dev->se_dev_attrib.block_size * sectors;
}
unsigned char transport_asciihex_to_binaryhex(unsigned char val[2])
@@ -2853,17 +2853,17 @@ static void transport_xor_callback(struct se_cmd *cmd)
return;
}
/*
- * Copy the scatterlist WRITE buffer located at T_TASK(cmd)->t_mem_list
+ * Copy the scatterlist WRITE buffer located at cmd->t_task->t_mem_list
* into the locally allocated *buf
*/
- transport_memcpy_se_mem_read_contig(cmd, buf, T_TASK(cmd)->t_mem_list);
+ transport_memcpy_se_mem_read_contig(cmd, buf, cmd->t_task->t_mem_list);
/*
* Now perform the XOR against the BIDI read memory located at
- * T_TASK(cmd)->t_mem_bidi_list
+ * cmd->t_task->t_mem_bidi_list
*/
offset = 0;
- list_for_each_entry(se_mem, T_TASK(cmd)->t_mem_bidi_list, se_list) {
+ list_for_each_entry(se_mem, cmd->t_task->t_mem_bidi_list, se_list) {
addr = (unsigned char *)kmap_atomic(se_mem->se_page, KM_USER0);
if (!(addr))
goto out;
@@ -2889,16 +2889,16 @@ static int transport_get_sense_data(struct se_cmd *cmd)
unsigned long flags;
u32 offset = 0;
- WARN_ON(!SE_LUN(cmd));
+ WARN_ON(!cmd->se_lun);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
if (cmd->se_cmd_flags & SCF_SENT_CHECK_CONDITION) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return 0;
}
list_for_each_entry_safe(task, task_tmp,
- &T_TASK(cmd)->t_task_list, t_list) {
+ &cmd->t_task->t_task_list, t_list) {
if (!task->task_sense)
continue;
@@ -2907,22 +2907,22 @@ static int transport_get_sense_data(struct se_cmd *cmd)
if (!(dev))
continue;
- if (!TRANSPORT(dev)->get_sense_buffer) {
- printk(KERN_ERR "TRANSPORT(dev)->get_sense_buffer"
+ if (!dev->transport->get_sense_buffer) {
+ printk(KERN_ERR "dev->transport->get_sense_buffer"
" is NULL\n");
continue;
}
- sense_buffer = TRANSPORT(dev)->get_sense_buffer(task);
+ sense_buffer = dev->transport->get_sense_buffer(task);
if (!(sense_buffer)) {
printk(KERN_ERR "ITT[0x%08x]_TASK[%d]: Unable to locate"
" sense buffer for task with sense\n",
- CMD_TFO(cmd)->get_task_tag(cmd), task->task_no);
+ cmd->se_tfo->get_task_tag(cmd), task->task_no);
continue;
}
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
- offset = CMD_TFO(cmd)->set_fabric_sense_len(cmd,
+ offset = cmd->se_tfo->set_fabric_sense_len(cmd,
TRANSPORT_SENSE_BUFFER);
memcpy((void *)&buffer[offset], (void *)sense_buffer,
@@ -2934,11 +2934,11 @@ static int transport_get_sense_data(struct se_cmd *cmd)
printk(KERN_INFO "HBA_[%u]_PLUG[%s]: Set SAM STATUS: 0x%02x"
" and sense\n",
- dev->se_hba->hba_id, TRANSPORT(dev)->name,
+ dev->se_hba->hba_id, dev->transport->name,
cmd->scsi_status);
return 0;
}
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return -1;
}
@@ -2970,9 +2970,9 @@ transport_handle_reservation_conflict(struct se_cmd *cmd)
*
* See spc4r17, section 7.4.6 Control Mode Page, Table 349
*/
- if (SE_SESS(cmd) &&
- DEV_ATTRIB(cmd->se_dev)->emulate_ua_intlck_ctrl == 2)
- core_scsi3_ua_allocate(SE_SESS(cmd)->se_node_acl,
+ if (cmd->se_sess &&
+ cmd->se_dev->se_sub_dev->se_dev_attrib.emulate_ua_intlck_ctrl == 2)
+ core_scsi3_ua_allocate(cmd->se_sess->se_node_acl,
cmd->orig_fe_lun, 0x2C,
ASCQ_2CH_PREVIOUS_RESERVATION_CONFLICT_STATUS);
return -2;
@@ -2992,7 +2992,7 @@ static int transport_generic_cmd_sequencer(
struct se_cmd *cmd,
unsigned char *cdb)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
struct se_subsystem_dev *su_dev = dev->se_sub_dev;
int ret = 0, sector_ret = 0, passthrough;
u32 sectors = 0, size = 0, pr_reg_type = 0;
@@ -3011,7 +3011,7 @@ static int transport_generic_cmd_sequencer(
/*
* Check status of Asymmetric Logical Unit Assignment port
*/
- ret = T10_ALUA(su_dev)->alua_state_check(cmd, cdb, &alua_ascq);
+ ret = su_dev->t10_alua.alua_state_check(cmd, cdb, &alua_ascq);
if (ret != 0) {
cmd->transport_wait_for_tasks = &transport_nop_wait_for_tasks;
/*
@@ -3023,7 +3023,7 @@ static int transport_generic_cmd_sequencer(
#if 0
printk(KERN_INFO "[%s]: ALUA TG Port not available,"
" SenseKey: NOT_READY, ASC/ASCQ: 0x04/0x%02x\n",
- CMD_TFO(cmd)->get_fabric_name(), alua_ascq);
+ cmd->se_tfo->get_fabric_name(), alua_ascq);
#endif
transport_set_sense_codes(cmd, 0x04, alua_ascq);
cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
@@ -3035,8 +3035,8 @@ static int transport_generic_cmd_sequencer(
/*
* Check status for SPC-3 Persistent Reservations
*/
- if (T10_PR_OPS(su_dev)->t10_reservation_check(cmd, &pr_reg_type) != 0) {
- if (T10_PR_OPS(su_dev)->t10_seq_non_holder(
+ if (su_dev->t10_pr.pr_ops.t10_reservation_check(cmd, &pr_reg_type) != 0) {
+ if (su_dev->t10_pr.pr_ops.t10_seq_non_holder(
cmd, cdb, pr_reg_type) != 0)
return transport_handle_reservation_conflict(cmd);
/*
@@ -3053,7 +3053,7 @@ static int transport_generic_cmd_sequencer(
goto out_unsupported_cdb;
size = transport_get_size(sectors, cdb, cmd);
cmd->transport_split_cdb = &split_cdb_XX_6;
- T_TASK(cmd)->t_task_lba = transport_lba_21(cdb);
+ cmd->t_task->t_task_lba = transport_lba_21(cdb);
cmd->se_cmd_flags |= SCF_SCSI_DATA_SG_IO_CDB;
break;
case READ_10:
@@ -3062,7 +3062,7 @@ static int transport_generic_cmd_sequencer(
goto out_unsupported_cdb;
size = transport_get_size(sectors, cdb, cmd);
cmd->transport_split_cdb = &split_cdb_XX_10;
- T_TASK(cmd)->t_task_lba = transport_lba_32(cdb);
+ cmd->t_task->t_task_lba = transport_lba_32(cdb);
cmd->se_cmd_flags |= SCF_SCSI_DATA_SG_IO_CDB;
break;
case READ_12:
@@ -3071,7 +3071,7 @@ static int transport_generic_cmd_sequencer(
goto out_unsupported_cdb;
size = transport_get_size(sectors, cdb, cmd);
cmd->transport_split_cdb = &split_cdb_XX_12;
- T_TASK(cmd)->t_task_lba = transport_lba_32(cdb);
+ cmd->t_task->t_task_lba = transport_lba_32(cdb);
cmd->se_cmd_flags |= SCF_SCSI_DATA_SG_IO_CDB;
break;
case READ_16:
@@ -3080,7 +3080,7 @@ static int transport_generic_cmd_sequencer(
goto out_unsupported_cdb;
size = transport_get_size(sectors, cdb, cmd);
cmd->transport_split_cdb = &split_cdb_XX_16;
- T_TASK(cmd)->t_task_lba = transport_lba_64(cdb);
+ cmd->t_task->t_task_lba = transport_lba_64(cdb);
cmd->se_cmd_flags |= SCF_SCSI_DATA_SG_IO_CDB;
break;
case WRITE_6:
@@ -3089,7 +3089,7 @@ static int transport_generic_cmd_sequencer(
goto out_unsupported_cdb;
size = transport_get_size(sectors, cdb, cmd);
cmd->transport_split_cdb = &split_cdb_XX_6;
- T_TASK(cmd)->t_task_lba = transport_lba_21(cdb);
+ cmd->t_task->t_task_lba = transport_lba_21(cdb);
cmd->se_cmd_flags |= SCF_SCSI_DATA_SG_IO_CDB;
break;
case WRITE_10:
@@ -3098,8 +3098,8 @@ static int transport_generic_cmd_sequencer(
goto out_unsupported_cdb;
size = transport_get_size(sectors, cdb, cmd);
cmd->transport_split_cdb = &split_cdb_XX_10;
- T_TASK(cmd)->t_task_lba = transport_lba_32(cdb);
- T_TASK(cmd)->t_tasks_fua = (cdb[1] & 0x8);
+ cmd->t_task->t_task_lba = transport_lba_32(cdb);
+ cmd->t_task->t_tasks_fua = (cdb[1] & 0x8);
cmd->se_cmd_flags |= SCF_SCSI_DATA_SG_IO_CDB;
break;
case WRITE_12:
@@ -3108,8 +3108,8 @@ static int transport_generic_cmd_sequencer(
goto out_unsupported_cdb;
size = transport_get_size(sectors, cdb, cmd);
cmd->transport_split_cdb = &split_cdb_XX_12;
- T_TASK(cmd)->t_task_lba = transport_lba_32(cdb);
- T_TASK(cmd)->t_tasks_fua = (cdb[1] & 0x8);
+ cmd->t_task->t_task_lba = transport_lba_32(cdb);
+ cmd->t_task->t_tasks_fua = (cdb[1] & 0x8);
cmd->se_cmd_flags |= SCF_SCSI_DATA_SG_IO_CDB;
break;
case WRITE_16:
@@ -3118,22 +3118,22 @@ static int transport_generic_cmd_sequencer(
goto out_unsupported_cdb;
size = transport_get_size(sectors, cdb, cmd);
cmd->transport_split_cdb = &split_cdb_XX_16;
- T_TASK(cmd)->t_task_lba = transport_lba_64(cdb);
- T_TASK(cmd)->t_tasks_fua = (cdb[1] & 0x8);
+ cmd->t_task->t_task_lba = transport_lba_64(cdb);
+ cmd->t_task->t_tasks_fua = (cdb[1] & 0x8);
cmd->se_cmd_flags |= SCF_SCSI_DATA_SG_IO_CDB;
break;
case XDWRITEREAD_10:
if ((cmd->data_direction != DMA_TO_DEVICE) ||
- !(T_TASK(cmd)->t_tasks_bidi))
+ !(cmd->t_task->t_tasks_bidi))
goto out_invalid_cdb_field;
sectors = transport_get_sectors_10(cdb, cmd, §or_ret);
if (sector_ret)
goto out_unsupported_cdb;
size = transport_get_size(sectors, cdb, cmd);
cmd->transport_split_cdb = &split_cdb_XX_10;
- T_TASK(cmd)->t_task_lba = transport_lba_32(cdb);
+ cmd->t_task->t_task_lba = transport_lba_32(cdb);
cmd->se_cmd_flags |= SCF_SCSI_DATA_SG_IO_CDB;
- passthrough = (TRANSPORT(dev)->transport_type ==
+ passthrough = (dev->transport->transport_type ==
TRANSPORT_PLUGIN_PHBA_PDEV);
/*
* Skip the remaining assignments for TCM/PSCSI passthrough
@@ -3144,7 +3144,7 @@ static int transport_generic_cmd_sequencer(
* Setup BIDI XOR callback to be run during transport_generic_complete_ok()
*/
cmd->transport_complete_callback = &transport_xor_callback;
- T_TASK(cmd)->t_tasks_fua = (cdb[1] & 0x8);
+ cmd->t_task->t_tasks_fua = (cdb[1] & 0x8);
break;
case VARIABLE_LENGTH_CMD:
service_action = get_unaligned_be16(&cdb[8]);
@@ -3152,7 +3152,7 @@ static int transport_generic_cmd_sequencer(
* Determine if this is TCM/PSCSI device and we should disable
* internal emulation for this CDB.
*/
- passthrough = (TRANSPORT(dev)->transport_type ==
+ passthrough = (dev->transport->transport_type ==
TRANSPORT_PLUGIN_PHBA_PDEV);
switch (service_action) {
@@ -3166,7 +3166,7 @@ static int transport_generic_cmd_sequencer(
* XDWRITE_READ_32 logic.
*/
cmd->transport_split_cdb = &split_cdb_XX_32;
- T_TASK(cmd)->t_task_lba = transport_lba_64_ext(cdb);
+ cmd->t_task->t_task_lba = transport_lba_64_ext(cdb);
cmd->se_cmd_flags |= SCF_SCSI_DATA_SG_IO_CDB;
/*
@@ -3180,14 +3180,14 @@ static int transport_generic_cmd_sequencer(
* transport_generic_complete_ok()
*/
cmd->transport_complete_callback = &transport_xor_callback;
- T_TASK(cmd)->t_tasks_fua = (cdb[10] & 0x8);
+ cmd->t_task->t_tasks_fua = (cdb[10] & 0x8);
break;
case WRITE_SAME_32:
sectors = transport_get_sectors_32(cdb, cmd, §or_ret);
if (sector_ret)
goto out_unsupported_cdb;
size = transport_get_size(sectors, cdb, cmd);
- T_TASK(cmd)->t_task_lba = get_unaligned_be64(&cdb[12]);
+ cmd->t_task->t_task_lba = get_unaligned_be64(&cdb[12]);
cmd->se_cmd_flags |= SCF_SCSI_CONTROL_SG_IO_CDB;
/*
@@ -3219,14 +3219,14 @@ static int transport_generic_cmd_sequencer(
}
break;
case MAINTENANCE_IN:
- if (TRANSPORT(dev)->get_device_type(dev) != TYPE_ROM) {
+ if (dev->transport->get_device_type(dev) != TYPE_ROM) {
/* MAINTENANCE_IN from SCC-2 */
/*
* Check for emulated MI_REPORT_TARGET_PGS.
*/
if (cdb[1] == MI_REPORT_TARGET_PGS) {
cmd->transport_emulate_cdb =
- (T10_ALUA(su_dev)->alua_type ==
+ (su_dev->t10_alua.alua_type ==
SPC3_ALUA_EMULATED) ?
core_emulate_report_target_port_groups :
NULL;
@@ -3273,7 +3273,7 @@ static int transport_generic_cmd_sequencer(
case PERSISTENT_RESERVE_IN:
case PERSISTENT_RESERVE_OUT:
cmd->transport_emulate_cdb =
- (T10_RES(su_dev)->res_type ==
+ (su_dev->t10_pr.res_type ==
SPC3_PERSISTENT_RESERVATIONS) ?
core_scsi3_emulate_pr : NULL;
size = (cdb[7] << 8) + cdb[8];
@@ -3289,14 +3289,14 @@ static int transport_generic_cmd_sequencer(
cmd->se_cmd_flags |= SCF_SCSI_CONTROL_NONSG_IO_CDB;
break;
case MAINTENANCE_OUT:
- if (TRANSPORT(dev)->get_device_type(dev) != TYPE_ROM) {
+ if (dev->transport->get_device_type(dev) != TYPE_ROM) {
/* MAINTENANCE_OUT from SCC-2
*
* Check for emulated MO_SET_TARGET_PGS.
*/
if (cdb[1] == MO_SET_TARGET_PGS) {
cmd->transport_emulate_cdb =
- (T10_ALUA(su_dev)->alua_type ==
+ (su_dev->t10_alua.alua_type ==
SPC3_ALUA_EMULATED) ?
core_emulate_set_target_port_groups :
NULL;
@@ -3316,7 +3316,7 @@ static int transport_generic_cmd_sequencer(
* Do implict HEAD_OF_QUEUE processing for INQUIRY.
* See spc4r17 section 5.3
*/
- if (SE_DEV(cmd)->dev_task_attr_type == SAM_TASK_ATTR_EMULATED)
+ if (cmd->se_lun->lun_se_dev->dev_task_attr_type == SAM_TASK_ATTR_EMULATED)
cmd->sam_task_attr = MSG_HEAD_TAG;
cmd->se_cmd_flags |= SCF_SCSI_CONTROL_NONSG_IO_CDB;
break;
@@ -3393,7 +3393,7 @@ static int transport_generic_cmd_sequencer(
* emulation disabled.
*/
cmd->transport_emulate_cdb =
- (T10_RES(su_dev)->res_type !=
+ (su_dev->t10_pr.res_type !=
SPC_PASSTHROUGH) ?
core_scsi2_emulate_crh : NULL;
cmd->se_cmd_flags |= SCF_SCSI_NON_DATA_CDB;
@@ -3410,7 +3410,7 @@ static int transport_generic_cmd_sequencer(
size = cmd->data_length;
cmd->transport_emulate_cdb =
- (T10_RES(su_dev)->res_type !=
+ (su_dev->t10_pr.res_type !=
SPC_PASSTHROUGH) ?
core_scsi2_emulate_crh : NULL;
cmd->se_cmd_flags |= SCF_SCSI_NON_DATA_CDB;
@@ -3422,10 +3422,10 @@ static int transport_generic_cmd_sequencer(
*/
if (cdb[0] == SYNCHRONIZE_CACHE) {
sectors = transport_get_sectors_10(cdb, cmd, §or_ret);
- T_TASK(cmd)->t_task_lba = transport_lba_32(cdb);
+ cmd->t_task->t_task_lba = transport_lba_32(cdb);
} else {
sectors = transport_get_sectors_16(cdb, cmd, §or_ret);
- T_TASK(cmd)->t_task_lba = transport_lba_64(cdb);
+ cmd->t_task->t_task_lba = transport_lba_64(cdb);
}
if (sector_ret)
goto out_unsupported_cdb;
@@ -3436,7 +3436,7 @@ static int transport_generic_cmd_sequencer(
/*
* For TCM/pSCSI passthrough, skip cmd->transport_emulate_cdb()
*/
- if (TRANSPORT(dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV)
+ if (dev->transport->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV)
break;
/*
* Set SCF_EMULATE_CDB_ASYNC to ensure asynchronous operation
@@ -3452,7 +3452,7 @@ static int transport_generic_cmd_sequencer(
break;
case UNMAP:
size = get_unaligned_be16(&cdb[7]);
- passthrough = (TRANSPORT(dev)->transport_type ==
+ passthrough = (dev->transport->transport_type ==
TRANSPORT_PLUGIN_PHBA_PDEV);
/*
* Determine if the received UNMAP used to for direct passthrough
@@ -3471,8 +3471,8 @@ static int transport_generic_cmd_sequencer(
if (sector_ret)
goto out_unsupported_cdb;
size = transport_get_size(sectors, cdb, cmd);
- T_TASK(cmd)->t_task_lba = get_unaligned_be16(&cdb[2]);
- passthrough = (TRANSPORT(dev)->transport_type ==
+ cmd->t_task->t_task_lba = get_unaligned_be16(&cdb[2]);
+ passthrough = (dev->transport->transport_type ==
TRANSPORT_PLUGIN_PHBA_PDEV);
/*
* Determine if the received WRITE_SAME_16 is used to for direct
@@ -3524,14 +3524,14 @@ static int transport_generic_cmd_sequencer(
* Do implict HEAD_OF_QUEUE processing for REPORT_LUNS
* See spc4r17 section 5.3
*/
- if (SE_DEV(cmd)->dev_task_attr_type == SAM_TASK_ATTR_EMULATED)
+ if (cmd->se_lun->lun_se_dev->dev_task_attr_type == SAM_TASK_ATTR_EMULATED)
cmd->sam_task_attr = MSG_HEAD_TAG;
cmd->se_cmd_flags |= SCF_SCSI_CONTROL_NONSG_IO_CDB;
break;
default:
printk(KERN_WARNING "TARGET_CORE[%s]: Unsupported SCSI Opcode"
" 0x%02x, sending CHECK_CONDITION.\n",
- CMD_TFO(cmd)->get_fabric_name(), cdb[0]);
+ cmd->se_tfo->get_fabric_name(), cdb[0]);
cmd->transport_wait_for_tasks = &transport_nop_wait_for_tasks;
goto out_unsupported_cdb;
}
@@ -3539,7 +3539,7 @@ static int transport_generic_cmd_sequencer(
if (size != cmd->data_length) {
printk(KERN_WARNING "TARGET_CORE[%s]: Expected Transfer Length:"
" %u does not match SCSI CDB Length: %u for SAM Opcode:"
- " 0x%02x\n", CMD_TFO(cmd)->get_fabric_name(),
+ " 0x%02x\n", cmd->se_tfo->get_fabric_name(),
cmd->data_length, size, cdb[0]);
cmd->cmd_spdtl = size;
@@ -3553,10 +3553,10 @@ static int transport_generic_cmd_sequencer(
* Reject READ_* or WRITE_* with overflow/underflow for
* type SCF_SCSI_DATA_SG_IO_CDB.
*/
- if (!(ret) && (DEV_ATTRIB(dev)->block_size != 512)) {
+ if (!(ret) && (dev->se_sub_dev->se_dev_attrib.block_size != 512)) {
printk(KERN_ERR "Failing OVERFLOW/UNDERFLOW for LBA op"
" CDB on non 512-byte sector setup subsystem"
- " plugin: %s\n", TRANSPORT(dev)->name);
+ " plugin: %s\n", dev->transport->name);
/* Returns CHECK_CONDITION + INVALID_CDB_FIELD */
goto out_invalid_cdb_field;
}
@@ -3679,7 +3679,7 @@ static void transport_memcpy_se_mem_read_contig(
*/
static void transport_complete_task_attr(struct se_cmd *cmd)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
struct se_cmd *cmd_p, *cmd_tmp;
int new_active_tasks = 0;
@@ -3750,7 +3750,7 @@ static void transport_generic_complete_ok(struct se_cmd *cmd)
* delayed execution list after a HEAD_OF_QUEUE or ORDERED Task
* Attribute.
*/
- if (SE_DEV(cmd)->dev_task_attr_type == SAM_TASK_ATTR_EMULATED)
+ if (cmd->se_lun->lun_se_dev->dev_task_attr_type == SAM_TASK_ATTR_EMULATED)
transport_complete_task_attr(cmd);
/*
* Check if we need to retrieve a sense buffer from
@@ -3782,8 +3782,8 @@ static void transport_generic_complete_ok(struct se_cmd *cmd)
switch (cmd->data_direction) {
case DMA_FROM_DEVICE:
spin_lock(&cmd->se_lun->lun_sep_lock);
- if (SE_LUN(cmd)->lun_sep) {
- SE_LUN(cmd)->lun_sep->sep_stats.tx_data_octets +=
+ if (cmd->se_lun->lun_sep) {
+ cmd->se_lun->lun_sep->sep_stats.tx_data_octets +=
cmd->data_length;
}
spin_unlock(&cmd->se_lun->lun_sep_lock);
@@ -3794,34 +3794,34 @@ static void transport_generic_complete_ok(struct se_cmd *cmd)
*/
if (cmd->se_cmd_flags & SCF_PASSTHROUGH_CONTIG_TO_SG)
transport_memcpy_write_contig(cmd,
- T_TASK(cmd)->t_task_pt_sgl,
- T_TASK(cmd)->t_task_buf);
+ cmd->t_task->t_task_pt_sgl,
+ cmd->t_task->t_task_buf);
- CMD_TFO(cmd)->queue_data_in(cmd);
+ cmd->se_tfo->queue_data_in(cmd);
break;
case DMA_TO_DEVICE:
spin_lock(&cmd->se_lun->lun_sep_lock);
- if (SE_LUN(cmd)->lun_sep) {
- SE_LUN(cmd)->lun_sep->sep_stats.rx_data_octets +=
+ if (cmd->se_lun->lun_sep) {
+ cmd->se_lun->lun_sep->sep_stats.rx_data_octets +=
cmd->data_length;
}
spin_unlock(&cmd->se_lun->lun_sep_lock);
/*
* Check if we need to send READ payload for BIDI-COMMAND
*/
- if (T_TASK(cmd)->t_mem_bidi_list != NULL) {
+ if (cmd->t_task->t_mem_bidi_list != NULL) {
spin_lock(&cmd->se_lun->lun_sep_lock);
- if (SE_LUN(cmd)->lun_sep) {
- SE_LUN(cmd)->lun_sep->sep_stats.tx_data_octets +=
+ if (cmd->se_lun->lun_sep) {
+ cmd->se_lun->lun_sep->sep_stats.tx_data_octets +=
cmd->data_length;
}
spin_unlock(&cmd->se_lun->lun_sep_lock);
- CMD_TFO(cmd)->queue_data_in(cmd);
+ cmd->se_tfo->queue_data_in(cmd);
break;
}
/* Fall through for DMA_TO_DEVICE */
case DMA_NONE:
- CMD_TFO(cmd)->queue_status(cmd);
+ cmd->se_tfo->queue_status(cmd);
break;
default:
break;
@@ -3836,9 +3836,9 @@ static void transport_free_dev_tasks(struct se_cmd *cmd)
struct se_task *task, *task_tmp;
unsigned long flags;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
list_for_each_entry_safe(task, task_tmp,
- &T_TASK(cmd)->t_task_list, t_list) {
+ &cmd->t_task->t_task_list, t_list) {
if (atomic_read(&task->task_active))
continue;
@@ -3847,15 +3847,15 @@ static void transport_free_dev_tasks(struct se_cmd *cmd)
list_del(&task->t_list);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
if (task->se_dev)
- TRANSPORT(task->se_dev)->free_task(task);
+ task->se_dev->transport->free_task(task);
else
printk(KERN_ERR "task[%u] - task->se_dev is NULL\n",
task->task_no);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
}
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
}
static inline void transport_free_pages(struct se_cmd *cmd)
@@ -3868,9 +3868,9 @@ static inline void transport_free_pages(struct se_cmd *cmd)
if (cmd->se_dev->transport->do_se_mem_map)
free_page = 0;
- if (T_TASK(cmd)->t_task_buf) {
- kfree(T_TASK(cmd)->t_task_buf);
- T_TASK(cmd)->t_task_buf = NULL;
+ if (cmd->t_task->t_task_buf) {
+ kfree(cmd->t_task->t_task_buf);
+ cmd->t_task->t_task_buf = NULL;
return;
}
@@ -3880,11 +3880,11 @@ static inline void transport_free_pages(struct se_cmd *cmd)
if (cmd->se_cmd_flags & SCF_CMD_PASSTHROUGH_NOALLOC)
return;
- if (!(T_TASK(cmd)->t_tasks_se_num))
+ if (!(cmd->t_task->t_tasks_se_num))
return;
list_for_each_entry_safe(se_mem, se_mem_tmp,
- T_TASK(cmd)->t_mem_list, se_list) {
+ cmd->t_task->t_mem_list, se_list) {
/*
* We only release call __free_page(struct se_mem->se_page) when
* SCF_PASSTHROUGH_SG_TO_MEM_NOALLOC is NOT in use,
@@ -3896,9 +3896,9 @@ static inline void transport_free_pages(struct se_cmd *cmd)
kmem_cache_free(se_mem_cache, se_mem);
}
- if (T_TASK(cmd)->t_mem_bidi_list && T_TASK(cmd)->t_tasks_se_bidi_num) {
+ if (cmd->t_task->t_mem_bidi_list && cmd->t_task->t_tasks_se_bidi_num) {
list_for_each_entry_safe(se_mem, se_mem_tmp,
- T_TASK(cmd)->t_mem_bidi_list, se_list) {
+ cmd->t_task->t_mem_bidi_list, se_list) {
/*
* We only release call __free_page(struct se_mem->se_page) when
* SCF_PASSTHROUGH_SG_TO_MEM_NOALLOC is NOT in use,
@@ -3911,11 +3911,11 @@ static inline void transport_free_pages(struct se_cmd *cmd)
}
}
- kfree(T_TASK(cmd)->t_mem_bidi_list);
- T_TASK(cmd)->t_mem_bidi_list = NULL;
- kfree(T_TASK(cmd)->t_mem_list);
- T_TASK(cmd)->t_mem_list = NULL;
- T_TASK(cmd)->t_tasks_se_num = 0;
+ kfree(cmd->t_task->t_mem_bidi_list);
+ cmd->t_task->t_mem_bidi_list = NULL;
+ kfree(cmd->t_task->t_mem_list);
+ cmd->t_task->t_mem_list = NULL;
+ cmd->t_task->t_tasks_se_num = 0;
}
static inline void transport_release_tasks(struct se_cmd *cmd)
@@ -3927,23 +3927,23 @@ static inline int transport_dec_and_check(struct se_cmd *cmd)
{
unsigned long flags;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- if (atomic_read(&T_TASK(cmd)->t_fe_count)) {
- if (!(atomic_dec_and_test(&T_TASK(cmd)->t_fe_count))) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock,
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ if (atomic_read(&cmd->t_task->t_fe_count)) {
+ if (!(atomic_dec_and_test(&cmd->t_task->t_fe_count))) {
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock,
flags);
return 1;
}
}
- if (atomic_read(&T_TASK(cmd)->t_se_count)) {
- if (!(atomic_dec_and_test(&T_TASK(cmd)->t_se_count))) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock,
+ if (atomic_read(&cmd->t_task->t_se_count)) {
+ if (!(atomic_dec_and_test(&cmd->t_task->t_se_count))) {
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock,
flags);
return 1;
}
}
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return 0;
}
@@ -3955,20 +3955,20 @@ static void transport_release_fe_cmd(struct se_cmd *cmd)
if (transport_dec_and_check(cmd))
return;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- if (!(atomic_read(&T_TASK(cmd)->transport_dev_active))) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ if (!(atomic_read(&cmd->t_task->transport_dev_active))) {
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
goto free_pages;
}
- atomic_set(&T_TASK(cmd)->transport_dev_active, 0);
+ atomic_set(&cmd->t_task->transport_dev_active, 0);
transport_all_task_dev_remove_state(cmd);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
transport_release_tasks(cmd);
free_pages:
transport_free_pages(cmd);
transport_free_se_cmd(cmd);
- CMD_TFO(cmd)->release_cmd_direct(cmd);
+ cmd->se_tfo->release_cmd_direct(cmd);
}
static int transport_generic_remove(
@@ -3978,27 +3978,27 @@ static int transport_generic_remove(
{
unsigned long flags;
- if (!(T_TASK(cmd)))
+ if (!(cmd->t_task))
goto release_cmd;
if (transport_dec_and_check(cmd)) {
if (session_reinstatement) {
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
transport_all_task_dev_remove_state(cmd);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock,
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock,
flags);
}
return 1;
}
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- if (!(atomic_read(&T_TASK(cmd)->transport_dev_active))) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ if (!(atomic_read(&cmd->t_task->transport_dev_active))) {
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
goto free_pages;
}
- atomic_set(&T_TASK(cmd)->transport_dev_active, 0);
+ atomic_set(&cmd->t_task->transport_dev_active, 0);
transport_all_task_dev_remove_state(cmd);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
transport_release_tasks(cmd);
free_pages:
@@ -4009,7 +4009,7 @@ release_cmd:
transport_release_cmd_to_pool(cmd);
} else {
transport_free_se_cmd(cmd);
- CMD_TFO(cmd)->release_cmd_direct(cmd);
+ cmd->se_tfo->release_cmd_direct(cmd);
}
return 0;
@@ -4049,8 +4049,8 @@ int transport_generic_map_mem_to_cmd(
return -ENOSYS;
}
- T_TASK(cmd)->t_mem_list = (struct list_head *)mem;
- T_TASK(cmd)->t_tasks_se_num = sg_mem_num;
+ cmd->t_task->t_mem_list = (struct list_head *)mem;
+ cmd->t_task->t_tasks_se_num = sg_mem_num;
cmd->se_cmd_flags |= SCF_CMD_PASSTHROUGH_NOALLOC;
return 0;
}
@@ -4065,36 +4065,36 @@ int transport_generic_map_mem_to_cmd(
* processed into a TCM struct se_subsystem_dev, we do the mapping
* from the passed physical memory to struct se_mem->se_page here.
*/
- T_TASK(cmd)->t_mem_list = transport_init_se_mem_list();
- if (!(T_TASK(cmd)->t_mem_list))
+ cmd->t_task->t_mem_list = transport_init_se_mem_list();
+ if (!(cmd->t_task->t_mem_list))
return -ENOMEM;
ret = transport_map_sg_to_mem(cmd,
- T_TASK(cmd)->t_mem_list, mem, &se_mem_cnt_out);
+ cmd->t_task->t_mem_list, mem, &se_mem_cnt_out);
if (ret < 0)
return -ENOMEM;
- T_TASK(cmd)->t_tasks_se_num = se_mem_cnt_out;
+ cmd->t_task->t_tasks_se_num = se_mem_cnt_out;
/*
* Setup BIDI READ list of struct se_mem elements
*/
if ((mem_bidi_in) && (sg_mem_bidi_num)) {
- T_TASK(cmd)->t_mem_bidi_list = transport_init_se_mem_list();
- if (!(T_TASK(cmd)->t_mem_bidi_list)) {
- kfree(T_TASK(cmd)->t_mem_list);
+ cmd->t_task->t_mem_bidi_list = transport_init_se_mem_list();
+ if (!(cmd->t_task->t_mem_bidi_list)) {
+ kfree(cmd->t_task->t_mem_list);
return -ENOMEM;
}
se_mem_cnt_out = 0;
ret = transport_map_sg_to_mem(cmd,
- T_TASK(cmd)->t_mem_bidi_list, mem_bidi_in,
+ cmd->t_task->t_mem_bidi_list, mem_bidi_in,
&se_mem_cnt_out);
if (ret < 0) {
- kfree(T_TASK(cmd)->t_mem_list);
+ kfree(cmd->t_task->t_mem_list);
return -ENOMEM;
}
- T_TASK(cmd)->t_tasks_se_bidi_num = se_mem_cnt_out;
+ cmd->t_task->t_tasks_se_bidi_num = se_mem_cnt_out;
}
cmd->se_cmd_flags |= SCF_PASSTHROUGH_SG_TO_MEM_NOALLOC;
@@ -4114,7 +4114,7 @@ int transport_generic_map_mem_to_cmd(
* struct scatterlist format.
*/
cmd->se_cmd_flags |= SCF_PASSTHROUGH_CONTIG_TO_SG;
- T_TASK(cmd)->t_task_pt_sgl = mem;
+ cmd->t_task->t_task_pt_sgl = mem;
}
return 0;
@@ -4129,21 +4129,21 @@ static inline long long transport_dev_end_lba(struct se_device *dev)
static int transport_get_sectors(struct se_cmd *cmd)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
- T_TASK(cmd)->t_tasks_sectors =
- (cmd->data_length / DEV_ATTRIB(dev)->block_size);
- if (!(T_TASK(cmd)->t_tasks_sectors))
- T_TASK(cmd)->t_tasks_sectors = 1;
+ cmd->t_task->t_tasks_sectors =
+ (cmd->data_length / dev->se_sub_dev->se_dev_attrib.block_size);
+ if (!(cmd->t_task->t_tasks_sectors))
+ cmd->t_task->t_tasks_sectors = 1;
- if (TRANSPORT(dev)->get_device_type(dev) != TYPE_DISK)
+ if (dev->transport->get_device_type(dev) != TYPE_DISK)
return 0;
- if ((T_TASK(cmd)->t_task_lba + T_TASK(cmd)->t_tasks_sectors) >
+ if ((cmd->t_task->t_task_lba + cmd->t_task->t_tasks_sectors) >
transport_dev_end_lba(dev)) {
printk(KERN_ERR "LBA: %llu Sectors: %u exceeds"
" transport_dev_end_lba(): %llu\n",
- T_TASK(cmd)->t_task_lba, T_TASK(cmd)->t_tasks_sectors,
+ cmd->t_task->t_task_lba, cmd->t_task->t_tasks_sectors,
transport_dev_end_lba(dev));
cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
cmd->scsi_sense_reason = TCM_SECTOR_COUNT_TOO_MANY;
@@ -4155,26 +4155,26 @@ static int transport_get_sectors(struct se_cmd *cmd)
static int transport_new_cmd_obj(struct se_cmd *cmd)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
u32 task_cdbs = 0, rc;
if (!(cmd->se_cmd_flags & SCF_SCSI_DATA_SG_IO_CDB)) {
task_cdbs++;
- T_TASK(cmd)->t_task_cdbs++;
+ cmd->t_task->t_task_cdbs++;
} else {
int set_counts = 1;
/*
* Setup any BIDI READ tasks and memory from
- * T_TASK(cmd)->t_mem_bidi_list so the READ struct se_tasks
+ * cmd->t_task->t_mem_bidi_list so the READ struct se_tasks
* are queued first for the non pSCSI passthrough case.
*/
- if ((T_TASK(cmd)->t_mem_bidi_list != NULL) &&
- (TRANSPORT(dev)->transport_type != TRANSPORT_PLUGIN_PHBA_PDEV)) {
+ if ((cmd->t_task->t_mem_bidi_list != NULL) &&
+ (dev->transport->transport_type != TRANSPORT_PLUGIN_PHBA_PDEV)) {
rc = transport_generic_get_cdb_count(cmd,
- T_TASK(cmd)->t_task_lba,
- T_TASK(cmd)->t_tasks_sectors,
- DMA_FROM_DEVICE, T_TASK(cmd)->t_mem_bidi_list,
+ cmd->t_task->t_task_lba,
+ cmd->t_task->t_tasks_sectors,
+ DMA_FROM_DEVICE, cmd->t_task->t_mem_bidi_list,
set_counts);
if (!(rc)) {
cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
@@ -4185,13 +4185,13 @@ static int transport_new_cmd_obj(struct se_cmd *cmd)
set_counts = 0;
}
/*
- * Setup the tasks and memory from T_TASK(cmd)->t_mem_list
+ * Setup the tasks and memory from cmd->t_task->t_mem_list
* Note for BIDI transfers this will contain the WRITE payload
*/
task_cdbs = transport_generic_get_cdb_count(cmd,
- T_TASK(cmd)->t_task_lba,
- T_TASK(cmd)->t_tasks_sectors,
- cmd->data_direction, T_TASK(cmd)->t_mem_list,
+ cmd->t_task->t_task_lba,
+ cmd->t_task->t_tasks_sectors,
+ cmd->data_direction, cmd->t_task->t_mem_list,
set_counts);
if (!(task_cdbs)) {
cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
@@ -4199,19 +4199,19 @@ static int transport_new_cmd_obj(struct se_cmd *cmd)
TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
return PYX_TRANSPORT_LU_COMM_FAILURE;
}
- T_TASK(cmd)->t_task_cdbs += task_cdbs;
+ cmd->t_task->t_task_cdbs += task_cdbs;
#if 0
printk(KERN_INFO "data_length: %u, LBA: %llu t_tasks_sectors:"
" %u, t_task_cdbs: %u\n", obj_ptr, cmd->data_length,
- T_TASK(cmd)->t_task_lba, T_TASK(cmd)->t_tasks_sectors,
- T_TASK(cmd)->t_task_cdbs);
+ cmd->t_task->t_task_lba, cmd->t_task->t_tasks_sectors,
+ cmd->t_task->t_task_cdbs);
#endif
}
- atomic_set(&T_TASK(cmd)->t_task_cdbs_left, task_cdbs);
- atomic_set(&T_TASK(cmd)->t_task_cdbs_ex_left, task_cdbs);
- atomic_set(&T_TASK(cmd)->t_task_cdbs_timeout_left, task_cdbs);
+ atomic_set(&cmd->t_task->t_task_cdbs_left, task_cdbs);
+ atomic_set(&cmd->t_task->t_task_cdbs_ex_left, task_cdbs);
+ atomic_set(&cmd->t_task->t_task_cdbs_timeout_left, task_cdbs);
return 0;
}
@@ -4235,8 +4235,8 @@ transport_generic_get_mem(struct se_cmd *cmd, u32 length, u32 dma_size)
unsigned char *buf;
struct se_mem *se_mem;
- T_TASK(cmd)->t_mem_list = transport_init_se_mem_list();
- if (!(T_TASK(cmd)->t_mem_list))
+ cmd->t_task->t_mem_list = transport_init_se_mem_list();
+ if (!(cmd->t_task->t_mem_list))
return -ENOMEM;
/*
@@ -4248,10 +4248,10 @@ transport_generic_get_mem(struct se_cmd *cmd, u32 length, u32 dma_size)
/*
* Setup BIDI-COMMAND READ list of struct se_mem elements
*/
- if (T_TASK(cmd)->t_tasks_bidi) {
- T_TASK(cmd)->t_mem_bidi_list = transport_init_se_mem_list();
- if (!(T_TASK(cmd)->t_mem_bidi_list)) {
- kfree(T_TASK(cmd)->t_mem_list);
+ if (cmd->t_task->t_tasks_bidi) {
+ cmd->t_task->t_mem_bidi_list = transport_init_se_mem_list();
+ if (!(cmd->t_task->t_mem_bidi_list)) {
+ kfree(cmd->t_task->t_mem_list);
return -ENOMEM;
}
}
@@ -4280,8 +4280,8 @@ transport_generic_get_mem(struct se_cmd *cmd, u32 length, u32 dma_size)
memset(buf, 0, se_mem->se_len);
kunmap_atomic(buf, KM_IRQ0);
- list_add_tail(&se_mem->se_list, T_TASK(cmd)->t_mem_list);
- T_TASK(cmd)->t_tasks_se_num++;
+ list_add_tail(&se_mem->se_list, cmd->t_task->t_mem_list);
+ cmd->t_task->t_tasks_se_num++;
DEBUG_MEM("Allocated struct se_mem page(%p) Length(%u)"
" Offset(%u)\n", se_mem->se_page, se_mem->se_len,
@@ -4291,7 +4291,7 @@ transport_generic_get_mem(struct se_cmd *cmd, u32 length, u32 dma_size)
}
DEBUG_MEM("Allocated total struct se_mem elements(%u)\n",
- T_TASK(cmd)->t_tasks_se_num);
+ cmd->t_task->t_tasks_se_num);
return 0;
out:
@@ -4307,9 +4307,9 @@ int transport_init_task_sg(
u32 task_offset)
{
struct se_cmd *se_cmd = task->task_se_cmd;
- struct se_device *se_dev = SE_DEV(se_cmd);
+ struct se_device *se_dev = se_cmd->se_lun->lun_se_dev;
struct se_mem *se_mem = in_se_mem;
- struct target_core_fabric_ops *tfo = CMD_TFO(se_cmd);
+ struct target_core_fabric_ops *tfo = se_cmd->se_tfo;
u32 sg_length, task_size = task->task_size, task_sg_num_padded;
while (task_size != 0) {
@@ -4323,7 +4323,7 @@ int transport_init_task_sg(
sg_length = se_mem->se_len;
if (!(list_is_last(&se_mem->se_list,
- T_TASK(se_cmd)->t_mem_list)))
+ se_cmd->t_task->t_mem_list)))
se_mem = list_entry(se_mem->se_list.next,
struct se_mem, se_list);
} else {
@@ -4343,7 +4343,7 @@ int transport_init_task_sg(
sg_length = (se_mem->se_len - task_offset);
if (!(list_is_last(&se_mem->se_list,
- T_TASK(se_cmd)->t_mem_list)))
+ se_cmd->t_task->t_mem_list)))
se_mem = list_entry(se_mem->se_list.next,
struct se_mem, se_list);
}
@@ -4384,8 +4384,8 @@ next:
* Setup task->task_sg_bidi for SCSI READ payload for
* TCM/pSCSI passthrough if present for BIDI-COMMAND
*/
- if ((T_TASK(se_cmd)->t_mem_bidi_list != NULL) &&
- (TRANSPORT(se_dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV)) {
+ if ((se_cmd->t_task->t_mem_bidi_list != NULL) &&
+ (se_dev->transport->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV)) {
task->task_sg_bidi = kzalloc(task_sg_num_padded *
sizeof(struct scatterlist), GFP_KERNEL);
if (!(task->task_sg_bidi)) {
@@ -4430,13 +4430,13 @@ static inline int transport_set_tasks_sectors_disk(
if ((lba + sectors) > transport_dev_end_lba(dev)) {
task->task_sectors = ((transport_dev_end_lba(dev) - lba) + 1);
- if (task->task_sectors > DEV_ATTRIB(dev)->max_sectors) {
- task->task_sectors = DEV_ATTRIB(dev)->max_sectors;
+ if (task->task_sectors > dev->se_sub_dev->se_dev_attrib.max_sectors) {
+ task->task_sectors = dev->se_sub_dev->se_dev_attrib.max_sectors;
*max_sectors_set = 1;
}
} else {
- if (sectors > DEV_ATTRIB(dev)->max_sectors) {
- task->task_sectors = DEV_ATTRIB(dev)->max_sectors;
+ if (sectors > dev->se_sub_dev->se_dev_attrib.max_sectors) {
+ task->task_sectors = dev->se_sub_dev->se_dev_attrib.max_sectors;
*max_sectors_set = 1;
} else
task->task_sectors = sectors;
@@ -4452,8 +4452,8 @@ static inline int transport_set_tasks_sectors_non_disk(
u32 sectors,
int *max_sectors_set)
{
- if (sectors > DEV_ATTRIB(dev)->max_sectors) {
- task->task_sectors = DEV_ATTRIB(dev)->max_sectors;
+ if (sectors > dev->se_sub_dev->se_dev_attrib.max_sectors) {
+ task->task_sectors = dev->se_sub_dev->se_dev_attrib.max_sectors;
*max_sectors_set = 1;
} else
task->task_sectors = sectors;
@@ -4468,7 +4468,7 @@ static inline int transport_set_tasks_sectors(
u32 sectors,
int *max_sectors_set)
{
- return (TRANSPORT(dev)->get_device_type(dev) == TYPE_DISK) ?
+ return (dev->transport->get_device_type(dev) == TYPE_DISK) ?
transport_set_tasks_sectors_disk(task, dev, lba, sectors,
max_sectors_set) :
transport_set_tasks_sectors_non_disk(task, dev, lba, sectors,
@@ -4568,7 +4568,7 @@ int transport_map_mem_to_sg(
sg->length = se_mem->se_len;
if (!(list_is_last(&se_mem->se_list,
- T_TASK(se_cmd)->t_mem_list))) {
+ se_cmd->t_task->t_mem_list))) {
se_mem = list_entry(se_mem->se_list.next,
struct se_mem, se_list);
(*se_mem_cnt)++;
@@ -4604,7 +4604,7 @@ int transport_map_mem_to_sg(
sg->length = (se_mem->se_len - *task_offset);
if (!(list_is_last(&se_mem->se_list,
- T_TASK(se_cmd)->t_mem_list))) {
+ se_cmd->t_task->t_mem_list))) {
se_mem = list_entry(se_mem->se_list.next,
struct se_mem, se_list);
(*se_mem_cnt)++;
@@ -4648,7 +4648,7 @@ void transport_do_task_sg_chain(struct se_cmd *cmd)
struct scatterlist *sg_head_cur = NULL, *sg_link_cur = NULL;
struct scatterlist *sg, *sg_end = NULL, *sg_end_cur = NULL;
struct se_task *task;
- struct target_core_fabric_ops *tfo = CMD_TFO(cmd);
+ struct target_core_fabric_ops *tfo = cmd->se_tfo;
u32 task_sg_num = 0, sg_count = 0;
int i;
@@ -4662,7 +4662,7 @@ void transport_do_task_sg_chain(struct se_cmd *cmd)
* Walk the struct se_task list and setup scatterlist chains
* for each contiguosly allocated struct se_task->task_sg[].
*/
- list_for_each_entry(task, &T_TASK(cmd)->t_task_list, t_list) {
+ list_for_each_entry(task, &cmd->t_task->t_task_list, t_list) {
if (!(task->task_sg) || !(task->task_padded_sg))
continue;
@@ -4673,7 +4673,7 @@ void transport_do_task_sg_chain(struct se_cmd *cmd)
* Either add chain or mark end of scatterlist
*/
if (!(list_is_last(&task->t_list,
- &T_TASK(cmd)->t_task_list))) {
+ &cmd->t_task->t_task_list))) {
/*
* Clear existing SGL termination bit set in
* transport_init_task_sg(), see sg_mark_end()
@@ -4699,7 +4699,7 @@ void transport_do_task_sg_chain(struct se_cmd *cmd)
/*
* Check for single task..
*/
- if (!(list_is_last(&task->t_list, &T_TASK(cmd)->t_task_list))) {
+ if (!(list_is_last(&task->t_list, &cmd->t_task->t_task_list))) {
/*
* Clear existing SGL termination bit set in
* transport_init_task_sg(), see sg_mark_end()
@@ -4717,15 +4717,15 @@ void transport_do_task_sg_chain(struct se_cmd *cmd)
* Setup the starting pointer and total t_tasks_sg_linked_no including
* padding SGs for linking and to mark the end.
*/
- T_TASK(cmd)->t_tasks_sg_chained = sg_first;
- T_TASK(cmd)->t_tasks_sg_chained_no = sg_count;
+ cmd->t_task->t_tasks_sg_chained = sg_first;
+ cmd->t_task->t_tasks_sg_chained_no = sg_count;
- DEBUG_CMD_M("Setup cmd: %p T_TASK(cmd)->t_tasks_sg_chained: %p and"
- " t_tasks_sg_chained_no: %u\n", cmd, T_TASK(cmd)->t_tasks_sg_chained,
- T_TASK(cmd)->t_tasks_sg_chained_no);
+ DEBUG_CMD_M("Setup cmd: %p cmd->t_task->t_tasks_sg_chained: %p and"
+ " t_tasks_sg_chained_no: %u\n", cmd, cmd->t_task->t_tasks_sg_chained,
+ cmd->t_task->t_tasks_sg_chained_no);
- for_each_sg(T_TASK(cmd)->t_tasks_sg_chained, sg,
- T_TASK(cmd)->t_tasks_sg_chained_no, i) {
+ for_each_sg(cmd->t_task->t_tasks_sg_chained, sg,
+ cmd->t_task->t_tasks_sg_chained_no, i) {
DEBUG_CMD_M("SG[%d]: %p page: %p length: %d offset: %d, magic: 0x%08x\n",
i, sg, sg_page(sg), sg->length, sg->offset, sg->sg_magic);
@@ -4753,12 +4753,12 @@ static int transport_do_se_mem_map(
* se_subsystem_api_t->do_se_mem_map is used when internal allocation
* has been done by the transport plugin.
*/
- if (TRANSPORT(dev)->do_se_mem_map) {
- ret = TRANSPORT(dev)->do_se_mem_map(task, se_mem_list,
+ if (dev->transport->do_se_mem_map) {
+ ret = dev->transport->do_se_mem_map(task, se_mem_list,
in_mem, in_se_mem, out_se_mem, se_mem_cnt,
task_offset_in);
if (ret == 0)
- T_TASK(task->task_se_cmd)->t_tasks_se_num += *se_mem_cnt;
+ task->task_se_cmd->t_task->t_tasks_se_num += *se_mem_cnt;
return ret;
}
@@ -4808,7 +4808,7 @@ static u32 transport_generic_get_cdb_count(
struct se_task *task;
struct se_mem *se_mem = NULL, *se_mem_lout = NULL;
struct se_mem *se_mem_bidi = NULL, *se_mem_bidi_lout = NULL;
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
int max_sectors_set = 0, ret;
u32 task_offset_in = 0, se_mem_cnt = 0, se_mem_bidi_cnt = 0, task_cdbs = 0;
@@ -4827,15 +4827,15 @@ static u32 transport_generic_get_cdb_count(
* Check for extra se_mem_bidi mapping for BIDI-COMMANDs to
* struct se_task->task_sg_bidi for TCM/pSCSI passthrough operation
*/
- if ((T_TASK(cmd)->t_mem_bidi_list != NULL) &&
- !(list_empty(T_TASK(cmd)->t_mem_bidi_list)) &&
- (TRANSPORT(dev)->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV))
- se_mem_bidi = list_entry(T_TASK(cmd)->t_mem_bidi_list->next,
+ if ((cmd->t_task->t_mem_bidi_list != NULL) &&
+ !(list_empty(cmd->t_task->t_mem_bidi_list)) &&
+ (dev->transport->transport_type == TRANSPORT_PLUGIN_PHBA_PDEV))
+ se_mem_bidi = list_entry(cmd->t_task->t_mem_bidi_list->next,
struct se_mem, se_list);
while (sectors) {
DEBUG_VOL("ITT[0x%08x] LBA(%llu) SectorsLeft(%u) EOBJ(%llu)\n",
- CMD_TFO(cmd)->get_task_tag(cmd), lba, sectors,
+ cmd->se_tfo->get_task_tag(cmd), lba, sectors,
transport_dev_end_lba(dev));
task = transport_generic_get_task(cmd, data_direction);
@@ -4849,19 +4849,19 @@ static u32 transport_generic_get_cdb_count(
lba += task->task_sectors;
sectors -= task->task_sectors;
task->task_size = (task->task_sectors *
- DEV_ATTRIB(dev)->block_size);
+ dev->se_sub_dev->se_dev_attrib.block_size);
- cdb = TRANSPORT(dev)->get_cdb(task);
+ cdb = dev->transport->get_cdb(task);
if ((cdb)) {
- memcpy(cdb, T_TASK(cmd)->t_task_cdb,
- scsi_command_size(T_TASK(cmd)->t_task_cdb));
+ memcpy(cdb, cmd->t_task->t_task_cdb,
+ scsi_command_size(cmd->t_task->t_task_cdb));
cmd->transport_split_cdb(task->task_lba,
&task->task_sectors, cdb);
}
/*
* Perform the SE OBJ plugin and/or Transport plugin specific
- * mapping for T_TASK(cmd)->t_mem_list. And setup the
+ * mapping for cmd->t_task->t_mem_list. And setup the
* task->task_sg and if necessary task->task_sg_bidi
*/
ret = transport_do_se_mem_map(dev, task, mem_list,
@@ -4872,7 +4872,7 @@ static u32 transport_generic_get_cdb_count(
se_mem = se_mem_lout;
/*
- * Setup the T_TASK(cmd)->t_mem_bidi_list -> task->task_sg_bidi
+ * Setup the cmd->t_task->t_mem_bidi_list -> task->task_sg_bidi
* mapping for SCSI READ for BIDI-COMMAND passthrough with TCM/pSCSI
*
* Note that the first call to transport_do_se_mem_map() above will
@@ -4882,7 +4882,7 @@ static u32 transport_generic_get_cdb_count(
*/
if (task->task_sg_bidi != NULL) {
ret = transport_do_se_mem_map(dev, task,
- T_TASK(cmd)->t_mem_bidi_list, NULL,
+ cmd->t_task->t_mem_bidi_list, NULL,
se_mem_bidi, &se_mem_bidi_lout, &se_mem_bidi_cnt,
&task_offset_in);
if (ret < 0)
@@ -4905,12 +4905,12 @@ static u32 transport_generic_get_cdb_count(
}
if (set_counts) {
- atomic_inc(&T_TASK(cmd)->t_fe_count);
- atomic_inc(&T_TASK(cmd)->t_se_count);
+ atomic_inc(&cmd->t_task->t_fe_count);
+ atomic_inc(&cmd->t_task->t_se_count);
}
DEBUG_VOL("ITT[0x%08x] total %s cdbs(%u)\n",
- CMD_TFO(cmd)->get_task_tag(cmd), (data_direction == DMA_TO_DEVICE)
+ cmd->se_tfo->get_task_tag(cmd), (data_direction == DMA_TO_DEVICE)
? "DMA_TO_DEVICE" : "DMA_FROM_DEVICE", task_cdbs);
return task_cdbs;
@@ -4921,7 +4921,7 @@ out:
static int
transport_map_control_cmd_to_task(struct se_cmd *cmd)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
unsigned char *cdb;
struct se_task *task;
int ret;
@@ -4930,7 +4930,7 @@ transport_map_control_cmd_to_task(struct se_cmd *cmd)
if (!task)
return PYX_TRANSPORT_OUT_OF_MEMORY_RESOURCES;
- cdb = TRANSPORT(dev)->get_cdb(task);
+ cdb = dev->transport->get_cdb(task);
if (cdb)
memcpy(cdb, cmd->t_task->t_task_cdb,
scsi_command_size(cmd->t_task->t_task_cdb));
@@ -4946,8 +4946,8 @@ transport_map_control_cmd_to_task(struct se_cmd *cmd)
struct se_mem *se_mem = NULL, *se_mem_lout = NULL;
u32 se_mem_cnt = 0, task_offset = 0;
- if (!list_empty(T_TASK(cmd)->t_mem_list))
- se_mem = list_entry(T_TASK(cmd)->t_mem_list->next,
+ if (!list_empty(cmd->t_task->t_mem_list))
+ se_mem = list_entry(cmd->t_task->t_mem_list->next,
struct se_mem, se_list);
ret = transport_do_se_mem_map(dev, task,
@@ -4986,14 +4986,14 @@ static int transport_generic_new_cmd(struct se_cmd *cmd)
{
struct se_portal_group *se_tpg;
struct se_task *task;
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
int ret = 0;
/*
* Determine is the TCM fabric module has already allocated physical
* memory, and is directly calling transport_generic_map_mem_to_cmd()
* to setup beforehand the linked list of physical memory at
- * T_TASK(cmd)->t_mem_list of struct se_mem->se_page
+ * cmd->t_task->t_mem_list of struct se_mem->se_page
*/
if (!(cmd->se_cmd_flags & SCF_PASSTHROUGH_SG_TO_MEM_NOALLOC)) {
ret = transport_allocate_resources(cmd);
@@ -5014,15 +5014,15 @@ static int transport_generic_new_cmd(struct se_cmd *cmd)
* Linux/NET via kernel sockets and needs to allocate a
* struct iovec array to complete the struct se_cmd
*/
- se_tpg = SE_LUN(cmd)->lun_sep->sep_tpg;
- if (TPG_TFO(se_tpg)->alloc_cmd_iovecs != NULL) {
- ret = TPG_TFO(se_tpg)->alloc_cmd_iovecs(cmd);
+ se_tpg = cmd->se_lun->lun_sep->sep_tpg;
+ if (se_tpg->se_tpg_tfo->alloc_cmd_iovecs != NULL) {
+ ret = se_tpg->se_tpg_tfo->alloc_cmd_iovecs(cmd);
if (ret < 0)
return PYX_TRANSPORT_OUT_OF_MEMORY_RESOURCES;
}
if (cmd->se_cmd_flags & SCF_SCSI_DATA_SG_IO_CDB) {
- list_for_each_entry(task, &T_TASK(cmd)->t_task_list, t_list) {
+ list_for_each_entry(task, &cmd->t_task->t_task_list, t_list) {
if (atomic_read(&task->task_sent))
continue;
if (!dev->transport->map_task_SG)
@@ -5069,9 +5069,9 @@ void transport_generic_process_write(struct se_cmd *cmd)
* original EDTL
*/
if (cmd->se_cmd_flags & SCF_UNDERFLOW_BIT) {
- if (!T_TASK(cmd)->t_tasks_se_num) {
+ if (!cmd->t_task->t_tasks_se_num) {
unsigned char *dst, *buf =
- (unsigned char *)T_TASK(cmd)->t_task_buf;
+ (unsigned char *)cmd->t_task->t_task_buf;
dst = kzalloc(cmd->cmd_spdtl), GFP_KERNEL);
if (!(dst)) {
@@ -5083,15 +5083,15 @@ void transport_generic_process_write(struct se_cmd *cmd)
}
memcpy(dst, buf, cmd->cmd_spdtl);
- kfree(T_TASK(cmd)->t_task_buf);
- T_TASK(cmd)->t_task_buf = dst;
+ kfree(cmd->t_task->t_task_buf);
+ cmd->t_task->t_task_buf = dst;
} else {
struct scatterlist *sg =
- (struct scatterlist *sg)T_TASK(cmd)->t_task_buf;
+ (struct scatterlist *sg)cmd->t_task->t_task_buf;
struct scatterlist *orig_sg;
orig_sg = kzalloc(sizeof(struct scatterlist) *
- T_TASK(cmd)->t_tasks_se_num,
+ cmd->t_task->t_tasks_se_num,
GFP_KERNEL))) {
if (!(orig_sg)) {
printk(KERN_ERR "Unable to allocate memory"
@@ -5101,9 +5101,9 @@ void transport_generic_process_write(struct se_cmd *cmd)
return;
}
- memcpy(orig_sg, T_TASK(cmd)->t_task_buf,
+ memcpy(orig_sg, cmd->t_task->t_task_buf,
sizeof(struct scatterlist) *
- T_TASK(cmd)->t_tasks_se_num);
+ cmd->t_task->t_tasks_se_num);
cmd->data_length = cmd->cmd_spdtl;
/*
@@ -5134,24 +5134,24 @@ static int transport_generic_write_pending(struct se_cmd *cmd)
unsigned long flags;
int ret;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
cmd->t_state = TRANSPORT_WRITE_PENDING;
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
/*
* For the TCM control CDBs using a contiguous buffer, do the memcpy
* from the passed Linux/SCSI struct scatterlist located at
- * T_TASK(se_cmd)->t_task_pt_buf to the contiguous buffer at
- * T_TASK(se_cmd)->t_task_buf.
+ * se_cmd->t_task->t_task_pt_buf to the contiguous buffer at
+ * se_cmd->t_task->t_task_buf.
*/
if (cmd->se_cmd_flags & SCF_PASSTHROUGH_CONTIG_TO_SG)
transport_memcpy_read_contig(cmd,
- T_TASK(cmd)->t_task_buf,
- T_TASK(cmd)->t_task_pt_sgl);
+ cmd->t_task->t_task_buf,
+ cmd->t_task->t_task_pt_sgl);
/*
* Clear the se_cmd for WRITE_PENDING status in order to set
- * T_TASK(cmd)->t_transport_active=0 so that transport_generic_handle_data
+ * cmd->t_task->t_transport_active=0 so that transport_generic_handle_data
* can be called from HW target mode interrupt code. This is safe
- * to be called with transport_off=1 before the CMD_TFO(cmd)->write_pending
+ * to be called with transport_off=1 before the cmd->se_tfo->write_pending
* because the se_cmd->se_lun pointer is not being cleared.
*/
transport_cmd_check_stop(cmd, 1, 0);
@@ -5160,7 +5160,7 @@ static int transport_generic_write_pending(struct se_cmd *cmd)
* Call the fabric write_pending function here to let the
* frontend know that WRITE buffers are ready.
*/
- ret = CMD_TFO(cmd)->write_pending(cmd);
+ ret = cmd->se_tfo->write_pending(cmd);
if (ret < 0)
return ret;
@@ -5173,11 +5173,11 @@ static int transport_generic_write_pending(struct se_cmd *cmd)
*/
void transport_release_cmd_to_pool(struct se_cmd *cmd)
{
- BUG_ON(!T_TASK(cmd));
- BUG_ON(!CMD_TFO(cmd));
+ BUG_ON(!cmd->t_task);
+ BUG_ON(!cmd->se_tfo);
transport_free_se_cmd(cmd);
- CMD_TFO(cmd)->release_cmd_to_pool(cmd);
+ cmd->se_tfo->release_cmd_to_pool(cmd);
}
EXPORT_SYMBOL(transport_release_cmd_to_pool);
@@ -5191,16 +5191,16 @@ void transport_generic_free_cmd(
int release_to_pool,
int session_reinstatement)
{
- if (!(cmd->se_cmd_flags & SCF_SE_LUN_CMD) || !T_TASK(cmd))
+ if (!(cmd->se_cmd_flags & SCF_SE_LUN_CMD) || !cmd->t_task)
transport_release_cmd_to_pool(cmd);
else {
core_dec_lacl_count(cmd->se_sess->se_node_acl, cmd);
- if (SE_LUN(cmd)) {
+ if (cmd->se_lun) {
#if 0
printk(KERN_INFO "cmd: %p ITT: 0x%08x contains"
- " SE_LUN(cmd)\n", cmd,
- CMD_TFO(cmd)->get_task_tag(cmd));
+ " cmd->se_lun\n", cmd,
+ cmd->se_tfo->get_task_tag(cmd));
#endif
transport_lun_remove_cmd(cmd);
}
@@ -5237,32 +5237,32 @@ static int transport_lun_wait_for_tasks(struct se_cmd *cmd, struct se_lun *lun)
* If the frontend has already requested this struct se_cmd to
* be stopped, we can safely ignore this struct se_cmd.
*/
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- if (atomic_read(&T_TASK(cmd)->t_transport_stop)) {
- atomic_set(&T_TASK(cmd)->transport_lun_stop, 0);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ if (atomic_read(&cmd->t_task->t_transport_stop)) {
+ atomic_set(&cmd->t_task->transport_lun_stop, 0);
DEBUG_TRANSPORT_S("ConfigFS ITT[0x%08x] - t_transport_stop =="
- " TRUE, skipping\n", CMD_TFO(cmd)->get_task_tag(cmd));
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ " TRUE, skipping\n", cmd->se_tfo->get_task_tag(cmd));
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
transport_cmd_check_stop(cmd, 1, 0);
return -EPERM;
}
- atomic_set(&T_TASK(cmd)->transport_lun_fe_stop, 1);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ atomic_set(&cmd->t_task->transport_lun_fe_stop, 1);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
- wake_up_interruptible(&SE_DEV(cmd)->dev_queue_obj.thread_wq);
+ wake_up_interruptible(&cmd->se_lun->lun_se_dev->dev_queue_obj.thread_wq);
ret = transport_stop_tasks_for_cmd(cmd);
DEBUG_TRANSPORT_S("ConfigFS: cmd: %p t_task_cdbs: %d stop tasks ret:"
- " %d\n", cmd, T_TASK(cmd)->t_task_cdbs, ret);
+ " %d\n", cmd, cmd->t_task->t_task_cdbs, ret);
if (!ret) {
DEBUG_TRANSPORT_S("ConfigFS: ITT[0x%08x] - stopping cmd....\n",
- CMD_TFO(cmd)->get_task_tag(cmd));
- wait_for_completion(&T_TASK(cmd)->transport_lun_stop_comp);
+ cmd->se_tfo->get_task_tag(cmd));
+ wait_for_completion(&cmd->t_task->transport_lun_stop_comp);
DEBUG_TRANSPORT_S("ConfigFS: ITT[0x%08x] - stopped cmd....\n",
- CMD_TFO(cmd)->get_task_tag(cmd));
+ cmd->se_tfo->get_task_tag(cmd));
}
- transport_remove_cmd_from_queue(cmd, &SE_DEV(cmd)->dev_queue_obj);
+ transport_remove_cmd_from_queue(cmd, &cmd->se_lun->lun_se_dev->dev_queue_obj);
return 0;
}
@@ -5288,33 +5288,33 @@ static void __transport_clear_lun_from_sessions(struct se_lun *lun)
struct se_cmd, se_lun_list);
list_del(&cmd->se_lun_list);
- if (!(T_TASK(cmd))) {
- printk(KERN_ERR "ITT: 0x%08x, T_TASK(cmd) = NULL"
+ if (!(cmd->t_task)) {
+ printk(KERN_ERR "ITT: 0x%08x, cmd->t_task = NULL"
"[i,t]_state: %u/%u\n",
- CMD_TFO(cmd)->get_task_tag(cmd),
- CMD_TFO(cmd)->get_cmd_state(cmd), cmd->t_state);
+ cmd->se_tfo->get_task_tag(cmd),
+ cmd->se_tfo->get_cmd_state(cmd), cmd->t_state);
BUG();
}
- atomic_set(&T_TASK(cmd)->transport_lun_active, 0);
+ atomic_set(&cmd->t_task->transport_lun_active, 0);
/*
* This will notify iscsi_target_transport.c:
* transport_cmd_check_stop() that a LUN shutdown is in
* progress for the iscsi_cmd_t.
*/
- spin_lock(&T_TASK(cmd)->t_state_lock);
- DEBUG_CLEAR_L("SE_LUN[%d] - Setting T_TASK(cmd)->transport"
+ spin_lock(&cmd->t_task->t_state_lock);
+ DEBUG_CLEAR_L("SE_LUN[%d] - Setting cmd->t_task->transport"
"_lun_stop for ITT: 0x%08x\n",
- SE_LUN(cmd)->unpacked_lun,
- CMD_TFO(cmd)->get_task_tag(cmd));
- atomic_set(&T_TASK(cmd)->transport_lun_stop, 1);
- spin_unlock(&T_TASK(cmd)->t_state_lock);
+ cmd->se_lun->unpacked_lun,
+ cmd->se_tfo->get_task_tag(cmd));
+ atomic_set(&cmd->t_task->transport_lun_stop, 1);
+ spin_unlock(&cmd->t_task->t_state_lock);
spin_unlock_irqrestore(&lun->lun_cmd_lock, lun_flags);
- if (!(SE_LUN(cmd))) {
+ if (!(cmd->se_lun)) {
printk(KERN_ERR "ITT: 0x%08x, [i,t]_state: %u/%u\n",
- CMD_TFO(cmd)->get_task_tag(cmd),
- CMD_TFO(cmd)->get_cmd_state(cmd), cmd->t_state);
+ cmd->se_tfo->get_task_tag(cmd),
+ cmd->se_tfo->get_cmd_state(cmd), cmd->t_state);
BUG();
}
/*
@@ -5322,27 +5322,27 @@ static void __transport_clear_lun_from_sessions(struct se_lun *lun)
* and/or stop its context.
*/
DEBUG_CLEAR_L("SE_LUN[%d] - ITT: 0x%08x before transport"
- "_lun_wait_for_tasks()\n", SE_LUN(cmd)->unpacked_lun,
- CMD_TFO(cmd)->get_task_tag(cmd));
+ "_lun_wait_for_tasks()\n", cmd->se_lun->unpacked_lun,
+ cmd->se_tfo->get_task_tag(cmd));
- if (transport_lun_wait_for_tasks(cmd, SE_LUN(cmd)) < 0) {
+ if (transport_lun_wait_for_tasks(cmd, cmd->se_lun) < 0) {
spin_lock_irqsave(&lun->lun_cmd_lock, lun_flags);
continue;
}
DEBUG_CLEAR_L("SE_LUN[%d] - ITT: 0x%08x after transport_lun"
"_wait_for_tasks(): SUCCESS\n",
- SE_LUN(cmd)->unpacked_lun,
- CMD_TFO(cmd)->get_task_tag(cmd));
+ cmd->se_lun->unpacked_lun,
+ cmd->se_tfo->get_task_tag(cmd));
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, cmd_flags);
- if (!(atomic_read(&T_TASK(cmd)->transport_dev_active))) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, cmd_flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, cmd_flags);
+ if (!(atomic_read(&cmd->t_task->transport_dev_active))) {
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, cmd_flags);
goto check_cond;
}
- atomic_set(&T_TASK(cmd)->transport_dev_active, 0);
+ atomic_set(&cmd->t_task->transport_dev_active, 0);
transport_all_task_dev_remove_state(cmd);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, cmd_flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, cmd_flags);
transport_free_dev_tasks(cmd);
/*
@@ -5359,24 +5359,24 @@ check_cond:
* be released, notify the waiting thread now that LU has
* finished accessing it.
*/
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, cmd_flags);
- if (atomic_read(&T_TASK(cmd)->transport_lun_fe_stop)) {
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, cmd_flags);
+ if (atomic_read(&cmd->t_task->transport_lun_fe_stop)) {
DEBUG_CLEAR_L("SE_LUN[%d] - Detected FE stop for"
" struct se_cmd: %p ITT: 0x%08x\n",
lun->unpacked_lun,
- cmd, CMD_TFO(cmd)->get_task_tag(cmd));
+ cmd, cmd->se_tfo->get_task_tag(cmd));
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock,
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock,
cmd_flags);
transport_cmd_check_stop(cmd, 1, 0);
- complete(&T_TASK(cmd)->transport_lun_fe_stop_comp);
+ complete(&cmd->t_task->transport_lun_fe_stop_comp);
spin_lock_irqsave(&lun->lun_cmd_lock, lun_flags);
continue;
}
DEBUG_CLEAR_L("SE_LUN[%d] - ITT: 0x%08x finished processing\n",
- lun->unpacked_lun, CMD_TFO(cmd)->get_task_tag(cmd));
+ lun->unpacked_lun, cmd->se_tfo->get_task_tag(cmd));
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, cmd_flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, cmd_flags);
spin_lock_irqsave(&lun->lun_cmd_lock, lun_flags);
}
spin_unlock_irqrestore(&lun->lun_cmd_lock, lun_flags);
@@ -5422,20 +5422,20 @@ static void transport_generic_wait_for_tasks(
if (!(cmd->se_cmd_flags & SCF_SE_LUN_CMD) && !(cmd->se_tmr_req))
return;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
/*
* If we are already stopped due to an external event (ie: LUN shutdown)
* sleep until the connection can have the passed struct se_cmd back.
- * The T_TASK(cmd)->transport_lun_stopped_sem will be upped by
+ * The cmd->t_task->transport_lun_stopped_sem will be upped by
* transport_clear_lun_from_sessions() once the ConfigFS context caller
* has completed its operation on the struct se_cmd.
*/
- if (atomic_read(&T_TASK(cmd)->transport_lun_stop)) {
+ if (atomic_read(&cmd->t_task->transport_lun_stop)) {
DEBUG_TRANSPORT_S("wait_for_tasks: Stopping"
- " wait_for_completion(&T_TASK(cmd)transport_lun_fe"
+ " wait_for_completion(&cmd->t_tasktransport_lun_fe"
"_stop_comp); for ITT: 0x%08x\n",
- CMD_TFO(cmd)->get_task_tag(cmd));
+ cmd->se_tfo->get_task_tag(cmd));
/*
* There is a special case for WRITES where a FE exception +
* LUN shutdown means ConfigFS context is still sleeping on
@@ -5443,10 +5443,10 @@ static void transport_generic_wait_for_tasks(
* We go ahead and up transport_lun_stop_comp just to be sure
* here.
*/
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
- complete(&T_TASK(cmd)->transport_lun_stop_comp);
- wait_for_completion(&T_TASK(cmd)->transport_lun_fe_stop_comp);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
+ complete(&cmd->t_task->transport_lun_stop_comp);
+ wait_for_completion(&cmd->t_task->transport_lun_fe_stop_comp);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
transport_all_task_dev_remove_state(cmd);
/*
@@ -5455,39 +5455,39 @@ static void transport_generic_wait_for_tasks(
* normal means below.
*/
DEBUG_TRANSPORT_S("wait_for_tasks: Stopped"
- " wait_for_completion(&T_TASK(cmd)transport_lun_fe_"
+ " wait_for_completion(&cmd->t_tasktransport_lun_fe_"
"stop_comp); for ITT: 0x%08x\n",
- CMD_TFO(cmd)->get_task_tag(cmd));
+ cmd->se_tfo->get_task_tag(cmd));
- atomic_set(&T_TASK(cmd)->transport_lun_stop, 0);
+ atomic_set(&cmd->t_task->transport_lun_stop, 0);
}
- if (!atomic_read(&T_TASK(cmd)->t_transport_active) ||
- atomic_read(&T_TASK(cmd)->t_transport_aborted))
+ if (!atomic_read(&cmd->t_task->t_transport_active) ||
+ atomic_read(&cmd->t_task->t_transport_aborted))
goto remove;
- atomic_set(&T_TASK(cmd)->t_transport_stop, 1);
+ atomic_set(&cmd->t_task->t_transport_stop, 1);
DEBUG_TRANSPORT_S("wait_for_tasks: Stopping %p ITT: 0x%08x"
" i_state: %d, t_state/def_t_state: %d/%d, t_transport_stop"
- " = TRUE\n", cmd, CMD_TFO(cmd)->get_task_tag(cmd),
- CMD_TFO(cmd)->get_cmd_state(cmd), cmd->t_state,
+ " = TRUE\n", cmd, cmd->se_tfo->get_task_tag(cmd),
+ cmd->se_tfo->get_cmd_state(cmd), cmd->t_state,
cmd->deferred_t_state);
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
- wake_up_interruptible(&SE_DEV(cmd)->dev_queue_obj.thread_wq);
+ wake_up_interruptible(&cmd->se_lun->lun_se_dev->dev_queue_obj.thread_wq);
- wait_for_completion(&T_TASK(cmd)->t_transport_stop_comp);
+ wait_for_completion(&cmd->t_task->t_transport_stop_comp);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- atomic_set(&T_TASK(cmd)->t_transport_active, 0);
- atomic_set(&T_TASK(cmd)->t_transport_stop, 0);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ atomic_set(&cmd->t_task->t_transport_active, 0);
+ atomic_set(&cmd->t_task->t_transport_stop, 0);
DEBUG_TRANSPORT_S("wait_for_tasks: Stopped wait_for_compltion("
- "&T_TASK(cmd)->t_transport_stop_comp) for ITT: 0x%08x\n",
- CMD_TFO(cmd)->get_task_tag(cmd));
+ "&cmd->t_task->t_transport_stop_comp) for ITT: 0x%08x\n",
+ cmd->se_tfo->get_task_tag(cmd));
remove:
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
if (!remove_cmd)
return;
@@ -5526,13 +5526,13 @@ int transport_send_check_condition_and_sense(
int offset;
u8 asc = 0, ascq = 0;
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
if (cmd->se_cmd_flags & SCF_SENT_CHECK_CONDITION) {
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
return 0;
}
cmd->se_cmd_flags |= SCF_SENT_CHECK_CONDITION;
- spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags);
+ spin_unlock_irqrestore(&cmd->t_task->t_state_lock, flags);
if (!reason && from_transport)
goto after_reason;
@@ -5545,7 +5545,7 @@ int transport_send_check_condition_and_sense(
* TRANSPORT_SENSE_BUFFER is now set to SCSI_SENSE_BUFFERSIZE
* from include/scsi/scsi_cmnd.h
*/
- offset = CMD_TFO(cmd)->set_fabric_sense_len(cmd,
+ offset = cmd->se_tfo->set_fabric_sense_len(cmd,
TRANSPORT_SENSE_BUFFER);
/*
* Actual SENSE DATA, see SPC-3 7.23.2 SPC_SENSE_KEY_OFFSET uses
@@ -5682,7 +5682,7 @@ int transport_send_check_condition_and_sense(
cmd->scsi_sense_length = TRANSPORT_SENSE_BUFFER + offset;
after_reason:
- CMD_TFO(cmd)->queue_status(cmd);
+ cmd->se_tfo->queue_status(cmd);
return 0;
}
EXPORT_SYMBOL(transport_send_check_condition_and_sense);
@@ -5691,18 +5691,18 @@ int transport_check_aborted_status(struct se_cmd *cmd, int send_status)
{
int ret = 0;
- if (atomic_read(&T_TASK(cmd)->t_transport_aborted) != 0) {
+ if (atomic_read(&cmd->t_task->t_transport_aborted) != 0) {
if (!(send_status) ||
(cmd->se_cmd_flags & SCF_SENT_DELAYED_TAS))
return 1;
#if 0
printk(KERN_INFO "Sending delayed SAM_STAT_TASK_ABORTED"
" status for CDB: 0x%02x ITT: 0x%08x\n",
- T_TASK(cmd)->t_task_cdb[0],
- CMD_TFO(cmd)->get_task_tag(cmd));
+ cmd->t_task->t_task_cdb[0],
+ cmd->se_tfo->get_task_tag(cmd));
#endif
cmd->se_cmd_flags |= SCF_SENT_DELAYED_TAS;
- CMD_TFO(cmd)->queue_status(cmd);
+ cmd->se_tfo->queue_status(cmd);
ret = 1;
}
return ret;
@@ -5718,8 +5718,8 @@ void transport_send_task_abort(struct se_cmd *cmd)
* queued back to fabric module by transport_check_aborted_status().
*/
if (cmd->data_direction == DMA_TO_DEVICE) {
- if (CMD_TFO(cmd)->write_pending_status(cmd) != 0) {
- atomic_inc(&T_TASK(cmd)->t_transport_aborted);
+ if (cmd->se_tfo->write_pending_status(cmd) != 0) {
+ atomic_inc(&cmd->t_task->t_transport_aborted);
smp_mb__after_atomic_inc();
cmd->scsi_status = SAM_STAT_TASK_ABORTED;
transport_new_cmd_failure(cmd);
@@ -5729,10 +5729,10 @@ void transport_send_task_abort(struct se_cmd *cmd)
cmd->scsi_status = SAM_STAT_TASK_ABORTED;
#if 0
printk(KERN_INFO "Setting SAM_STAT_TASK_ABORTED status for CDB: 0x%02x,"
- " ITT: 0x%08x\n", T_TASK(cmd)->t_task_cdb[0],
- CMD_TFO(cmd)->get_task_tag(cmd));
+ " ITT: 0x%08x\n", cmd->t_task->t_task_cdb[0],
+ cmd->se_tfo->get_task_tag(cmd));
#endif
- CMD_TFO(cmd)->queue_status(cmd);
+ cmd->se_tfo->queue_status(cmd);
}
/* transport_generic_do_tmr():
@@ -5742,7 +5742,7 @@ void transport_send_task_abort(struct se_cmd *cmd)
int transport_generic_do_tmr(struct se_cmd *cmd)
{
struct se_cmd *ref_cmd;
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
struct se_tmr_req *tmr = cmd->se_tmr_req;
int ret;
@@ -5775,7 +5775,7 @@ int transport_generic_do_tmr(struct se_cmd *cmd)
}
cmd->t_state = TRANSPORT_ISTATE_PROCESSING;
- CMD_TFO(cmd)->queue_tm_rsp(cmd);
+ cmd->se_tfo->queue_tm_rsp(cmd);
transport_cmd_check_stop(cmd, 2, 0);
return 0;
@@ -5814,44 +5814,44 @@ static void transport_processing_shutdown(struct se_device *dev)
*/
spin_lock_irqsave(&dev->execute_task_lock, flags);
while ((task = transport_get_task_from_state_list(dev))) {
- if (!(TASK_CMD(task))) {
- printk(KERN_ERR "TASK_CMD(task) is NULL!\n");
+ if (!task->task_se_cmd) {
+ printk(KERN_ERR "task->task_se_cmd is NULL!\n");
continue;
}
- cmd = TASK_CMD(task);
+ cmd = task->task_se_cmd;
- if (!T_TASK(cmd)) {
- printk(KERN_ERR "T_TASK(cmd) is NULL for task: %p cmd:"
+ if (!cmd->t_task) {
+ printk(KERN_ERR "cmd->t_task is NULL for task: %p cmd:"
" %p ITT: 0x%08x\n", task, cmd,
- CMD_TFO(cmd)->get_task_tag(cmd));
+ cmd->se_tfo->get_task_tag(cmd));
continue;
}
spin_unlock_irqrestore(&dev->execute_task_lock, flags);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
DEBUG_DO("PT: cmd: %p task: %p ITT/CmdSN: 0x%08x/0x%08x,"
" i_state/def_i_state: %d/%d, t_state/def_t_state:"
" %d/%d cdb: 0x%02x\n", cmd, task,
- CMD_TFO(cmd)->get_task_tag(cmd), cmd->cmd_sn,
- CMD_TFO(cmd)->get_cmd_state(cmd), cmd->deferred_i_state,
+ cmd->se_tfo->get_task_tag(cmd), cmd->cmd_sn,
+ cmd->se_tfo->get_cmd_state(cmd), cmd->deferred_i_state,
cmd->t_state, cmd->deferred_t_state,
- T_TASK(cmd)->t_task_cdb[0]);
+ cmd->t_task->t_task_cdb[0]);
DEBUG_DO("PT: ITT[0x%08x] - t_task_cdbs: %d t_task_cdbs_left:"
" %d t_task_cdbs_sent: %d -- t_transport_active: %d"
" t_transport_stop: %d t_transport_sent: %d\n",
- CMD_TFO(cmd)->get_task_tag(cmd),
- T_TASK(cmd)->t_task_cdbs,
- atomic_read(&T_TASK(cmd)->t_task_cdbs_left),
- atomic_read(&T_TASK(cmd)->t_task_cdbs_sent),
- atomic_read(&T_TASK(cmd)->t_transport_active),
- atomic_read(&T_TASK(cmd)->t_transport_stop),
- atomic_read(&T_TASK(cmd)->t_transport_sent));
+ cmd->se_tfo->get_task_tag(cmd),
+ cmd->t_task->t_task_cdbs,
+ atomic_read(&cmd->t_task->t_task_cdbs_left),
+ atomic_read(&cmd->t_task->t_task_cdbs_sent),
+ atomic_read(&cmd->t_task->t_transport_active),
+ atomic_read(&cmd->t_task->t_transport_stop),
+ atomic_read(&cmd->t_task->t_transport_sent));
if (atomic_read(&task->task_active)) {
atomic_set(&task->task_stop, 1);
spin_unlock_irqrestore(
- &T_TASK(cmd)->t_state_lock, flags);
+ &cmd->t_task->t_state_lock, flags);
DEBUG_DO("Waiting for task: %p to shutdown for dev:"
" %p\n", task, dev);
@@ -5859,8 +5859,8 @@ static void transport_processing_shutdown(struct se_device *dev)
DEBUG_DO("Completed task: %p shutdown for dev: %p\n",
task, dev);
- spin_lock_irqsave(&T_TASK(cmd)->t_state_lock, flags);
- atomic_dec(&T_TASK(cmd)->t_task_cdbs_left);
+ spin_lock_irqsave(&cmd->t_task->t_state_lock, flags);
+ atomic_dec(&cmd->t_task->t_task_cdbs_left);
atomic_set(&task->task_active, 0);
atomic_set(&task->task_stop, 0);
@@ -5870,39 +5870,39 @@ static void transport_processing_shutdown(struct se_device *dev)
}
__transport_stop_task_timer(task, &flags);
- if (!(atomic_dec_and_test(&T_TASK(cmd)->t_task_cdbs_ex_left))) {
+ if (!(atomic_dec_and_test(&cmd->t_task->t_task_cdbs_ex_left))) {
spin_unlock_irqrestore(
- &T_TASK(cmd)->t_state_lock, flags);
+ &cmd->t_task->t_state_lock, flags);
DEBUG_DO("Skipping task: %p, dev: %p for"
" t_task_cdbs_ex_left: %d\n", task, dev,
- atomic_read(&T_TASK(cmd)->t_task_cdbs_ex_left));
+ atomic_read(&cmd->t_task->t_task_cdbs_ex_left));
spin_lock_irqsave(&dev->execute_task_lock, flags);
continue;
}
- if (atomic_read(&T_TASK(cmd)->t_transport_active)) {
+ if (atomic_read(&cmd->t_task->t_transport_active)) {
DEBUG_DO("got t_transport_active = 1 for task: %p, dev:"
" %p\n", task, dev);
- if (atomic_read(&T_TASK(cmd)->t_fe_count)) {
+ if (atomic_read(&cmd->t_task->t_fe_count)) {
spin_unlock_irqrestore(
- &T_TASK(cmd)->t_state_lock, flags);
+ &cmd->t_task->t_state_lock, flags);
transport_send_check_condition_and_sense(
cmd, TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE,
0);
transport_remove_cmd_from_queue(cmd,
- &SE_DEV(cmd)->dev_queue_obj);
+ &cmd->se_lun->lun_se_dev->dev_queue_obj);
transport_lun_remove_cmd(cmd);
transport_cmd_check_stop(cmd, 1, 0);
} else {
spin_unlock_irqrestore(
- &T_TASK(cmd)->t_state_lock, flags);
+ &cmd->t_task->t_state_lock, flags);
transport_remove_cmd_from_queue(cmd,
- &SE_DEV(cmd)->dev_queue_obj);
+ &cmd->se_lun->lun_se_dev->dev_queue_obj);
transport_lun_remove_cmd(cmd);
@@ -5916,22 +5916,22 @@ static void transport_processing_shutdown(struct se_device *dev)
DEBUG_DO("Got t_transport_active = 0 for task: %p, dev: %p\n",
task, dev);
- if (atomic_read(&T_TASK(cmd)->t_fe_count)) {
+ if (atomic_read(&cmd->t_task->t_fe_count)) {
spin_unlock_irqrestore(
- &T_TASK(cmd)->t_state_lock, flags);
+ &cmd->t_task->t_state_lock, flags);
transport_send_check_condition_and_sense(cmd,
TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE, 0);
transport_remove_cmd_from_queue(cmd,
- &SE_DEV(cmd)->dev_queue_obj);
+ &cmd->se_lun->lun_se_dev->dev_queue_obj);
transport_lun_remove_cmd(cmd);
transport_cmd_check_stop(cmd, 1, 0);
} else {
spin_unlock_irqrestore(
- &T_TASK(cmd)->t_state_lock, flags);
+ &cmd->t_task->t_state_lock, flags);
transport_remove_cmd_from_queue(cmd,
- &SE_DEV(cmd)->dev_queue_obj);
+ &cmd->se_lun->lun_se_dev->dev_queue_obj);
transport_lun_remove_cmd(cmd);
if (transport_cmd_check_stop(cmd, 1, 0))
@@ -5952,7 +5952,7 @@ static void transport_processing_shutdown(struct se_device *dev)
DEBUG_DO("From Device Queue: cmd: %p t_state: %d\n",
cmd, state);
- if (atomic_read(&T_TASK(cmd)->t_fe_count)) {
+ if (atomic_read(&cmd->t_task->t_fe_count)) {
transport_send_check_condition_and_sense(cmd,
TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE, 0);
@@ -6007,12 +6007,12 @@ get_cmd:
switch (t_state) {
case TRANSPORT_NEW_CMD_MAP:
- if (!(CMD_TFO(cmd)->new_cmd_map)) {
- printk(KERN_ERR "CMD_TFO(cmd)->new_cmd_map is"
+ if (!(cmd->se_tfo->new_cmd_map)) {
+ printk(KERN_ERR "cmd->se_tfo->new_cmd_map is"
" NULL for TRANSPORT_NEW_CMD_MAP\n");
BUG();
}
- ret = CMD_TFO(cmd)->new_cmd_map(cmd);
+ ret = cmd->se_tfo->new_cmd_map(cmd);
if (ret < 0) {
cmd->transport_error_status = ret;
transport_generic_request_failure(cmd, NULL,
@@ -6057,9 +6057,9 @@ get_cmd:
printk(KERN_ERR "Unknown t_state: %d deferred_t_state:"
" %d for ITT: 0x%08x i_state: %d on SE LUN:"
" %u\n", t_state, cmd->deferred_t_state,
- CMD_TFO(cmd)->get_task_tag(cmd),
- CMD_TFO(cmd)->get_cmd_state(cmd),
- SE_LUN(cmd)->unpacked_lun);
+ cmd->se_tfo->get_task_tag(cmd),
+ cmd->se_tfo->get_cmd_state(cmd),
+ cmd->se_lun->unpacked_lun);
BUG();
}
diff --git a/drivers/target/target_core_ua.c b/drivers/target/target_core_ua.c
index 4f754ca..16f41d1 100644
--- a/drivers/target/target_core_ua.c
+++ b/drivers/target/target_core_ua.c
@@ -179,7 +179,7 @@ int core_scsi3_ua_allocate(
printk(KERN_INFO "[%s]: Allocated UNIT ATTENTION, mapped LUN: %u, ASC:"
" 0x%02x, ASCQ: 0x%02x\n",
- TPG_TFO(nacl->se_tpg)->get_fabric_name(), unpacked_lun,
+ nacl->se_tpg->se_tpg_tfo->get_fabric_name(), unpacked_lun,
asc, ascq);
atomic_inc(&deve->ua_count);
@@ -208,7 +208,7 @@ void core_scsi3_ua_for_check_condition(
u8 *asc,
u8 *ascq)
{
- struct se_device *dev = SE_DEV(cmd);
+ struct se_device *dev = cmd->se_lun->lun_se_dev;
struct se_dev_entry *deve;
struct se_session *sess = cmd->se_sess;
struct se_node_acl *nacl;
@@ -240,7 +240,7 @@ void core_scsi3_ua_for_check_condition(
* highest priority UNIT_ATTENTION and ASC/ASCQ without
* clearing it.
*/
- if (DEV_ATTRIB(dev)->emulate_ua_intlck_ctrl != 0) {
+ if (dev->se_sub_dev->se_dev_attrib.emulate_ua_intlck_ctrl != 0) {
*asc = ua->ua_asc;
*ascq = ua->ua_ascq;
break;
@@ -267,10 +267,10 @@ void core_scsi3_ua_for_check_condition(
printk(KERN_INFO "[%s]: %s UNIT ATTENTION condition with"
" INTLCK_CTRL: %d, mapped LUN: %u, got CDB: 0x%02x"
" reported ASC: 0x%02x, ASCQ: 0x%02x\n",
- TPG_TFO(nacl->se_tpg)->get_fabric_name(),
- (DEV_ATTRIB(dev)->emulate_ua_intlck_ctrl != 0) ? "Reporting" :
- "Releasing", DEV_ATTRIB(dev)->emulate_ua_intlck_ctrl,
- cmd->orig_fe_lun, T_TASK(cmd)->t_task_cdb[0], *asc, *ascq);
+ nacl->se_tpg->se_tpg_tfo->get_fabric_name(),
+ (dev->se_sub_dev->se_dev_attrib.emulate_ua_intlck_ctrl != 0) ? "Reporting" :
+ "Releasing", dev->se_sub_dev->se_dev_attrib.emulate_ua_intlck_ctrl,
+ cmd->orig_fe_lun, cmd->t_task->t_task_cdb[0], *asc, *ascq);
}
int core_scsi3_ua_clear_for_request_sense(
@@ -325,7 +325,7 @@ int core_scsi3_ua_clear_for_request_sense(
printk(KERN_INFO "[%s]: Released UNIT ATTENTION condition, mapped"
" LUN: %u, got REQUEST_SENSE reported ASC: 0x%02x,"
- " ASCQ: 0x%02x\n", TPG_TFO(nacl->se_tpg)->get_fabric_name(),
+ " ASCQ: 0x%02x\n", nacl->se_tpg->se_tpg_tfo->get_fabric_name(),
cmd->orig_fe_lun, *asc, *ascq);
return (head) ? -EPERM : 0;
diff --git a/drivers/target/tcm_fc/tfc_cmd.c b/drivers/target/tcm_fc/tfc_cmd.c
index 5823ccf..19b2b99 100644
--- a/drivers/target/tcm_fc/tfc_cmd.c
+++ b/drivers/target/tcm_fc/tfc_cmd.c
@@ -72,7 +72,7 @@ void ft_dump_cmd(struct ft_cmd *cmd, const char *caller)
caller, cmd, cmd->cdb);
printk(KERN_INFO "%s: cmd %p lun %d\n", caller, cmd, cmd->lun);
- task = T_TASK(se_cmd);
+ task = se_cmd->t_task;
printk(KERN_INFO "%s: cmd %p task %p se_num %u buf %p len %u se_cmd_flags <0x%x>\n",
caller, cmd, task, task->t_tasks_se_num,
task->t_task_buf, se_cmd->data_length, se_cmd->se_cmd_flags);
@@ -262,9 +262,9 @@ int ft_write_pending(struct se_cmd *se_cmd)
* TCM/LIO target
*/
transport_do_task_sg_chain(se_cmd);
- cmd->sg = T_TASK(se_cmd)->t_tasks_sg_chained;
+ cmd->sg = se_cmd->t_task->t_tasks_sg_chained;
cmd->sg_cnt =
- T_TASK(se_cmd)->t_tasks_sg_chained_no;
+ se_cmd->t_task->t_tasks_sg_chained_no;
}
if (cmd->sg && lport->tt.ddp_setup(lport, ep->xid,
cmd->sg, cmd->sg_cnt))
diff --git a/drivers/target/tcm_fc/tfc_io.c b/drivers/target/tcm_fc/tfc_io.c
index 8c4a240..47efcfb 100644
--- a/drivers/target/tcm_fc/tfc_io.c
+++ b/drivers/target/tcm_fc/tfc_io.c
@@ -90,7 +90,7 @@ int ft_queue_data_in(struct se_cmd *se_cmd)
lport = ep->lp;
cmd->seq = lport->tt.seq_start_next(cmd->seq);
- task = T_TASK(se_cmd);
+ task = se_cmd->t_task;
BUG_ON(!task);
remaining = se_cmd->data_length;
@@ -236,7 +236,7 @@ void ft_recv_write_data(struct ft_cmd *cmd, struct fc_frame *fp)
u32 f_ctl;
void *buf;
- task = T_TASK(se_cmd);
+ task = se_cmd->t_task;
BUG_ON(!task);
fh = fc_frame_header_get(fp);
diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h
index 2b989b3..b0b83ed 100644
--- a/include/target/target_core_base.h
+++ b/include/target/target_core_base.h
@@ -286,10 +286,10 @@ struct t10_vpd {
} ____cacheline_aligned;
struct t10_wwn {
- unsigned char vendor[8];
- unsigned char model[16];
- unsigned char revision[4];
- unsigned char unit_serial[INQUIRY_VPD_SERIAL_LEN];
+ char vendor[8];
+ char model[16];
+ char revision[4];
+ char unit_serial[INQUIRY_VPD_SERIAL_LEN];
spinlock_t t10_vpd_lock;
struct se_subsystem_dev *t10_sub_dev;
struct config_group t10_wwn_group;
@@ -490,9 +490,6 @@ struct se_task {
struct list_head t_state_list;
} ____cacheline_aligned;
-#define TASK_CMD(task) ((task)->task_se_cmd)
-#define TASK_DEV(task) ((task)->se_dev)
-
struct se_cmd {
/* SAM response code being sent to initiator */
u8 scsi_status;
@@ -547,9 +544,6 @@ struct se_cmd {
void (*transport_complete_callback)(struct se_cmd *);
} ____cacheline_aligned;
-#define T_TASK(cmd) ((cmd)->t_task)
-#define CMD_TFO(cmd) ((cmd)->se_tfo)
-
struct se_tmr_req {
/* Task Management function to be preformed */
u8 function;
@@ -612,9 +606,6 @@ struct se_session {
struct list_head sess_acl_list;
} ____cacheline_aligned;
-#define SE_SESS(cmd) ((cmd)->se_sess)
-#define SE_NODE_ACL(sess) ((sess)->se_node_acl)
-
struct se_device;
struct se_transform_info;
struct scatterlist;
@@ -635,8 +626,6 @@ struct se_lun_acl {
struct se_ml_stat_grps ml_stat_grps;
} ____cacheline_aligned;
-#define ML_STAT_GRPS(lacl) (&(lacl)->ml_stat_grps)
-
struct se_dev_entry {
bool def_pr_registered;
/* See transport_lunflags_table */
@@ -722,7 +711,7 @@ struct se_subsystem_dev {
/* T10 Inquiry and VPD WWN Information */
struct t10_wwn t10_wwn;
/* T10 SPC-2 + SPC-3 Reservations */
- struct t10_reservation t10_reservation;
+ struct t10_reservation t10_pr;
spinlock_t se_dev_lock;
void *se_dev_su_ptr;
struct list_head se_dev_node;
@@ -733,11 +722,6 @@ struct se_subsystem_dev {
struct se_dev_stat_grps dev_stat_grps;
} ____cacheline_aligned;
-#define T10_ALUA(su_dev) (&(su_dev)->t10_alua)
-#define T10_RES(su_dev) (&(su_dev)->t10_reservation)
-#define T10_PR_OPS(su_dev) (&(su_dev)->t10_reservation.pr_ops)
-#define DEV_STAT_GRP(dev) (&(dev)->dev_stat_grps)
-
struct se_device {
/* Set to 1 if thread is NOT sleeping on thread_sem */
u8 thread_active;
@@ -818,11 +802,6 @@ struct se_device {
struct list_head g_se_dev_list;
} ____cacheline_aligned;
-#define SE_DEV(cmd) ((cmd)->se_lun->lun_se_dev)
-#define SU_DEV(dev) ((dev)->se_sub_dev)
-#define DEV_ATTRIB(dev) (&(dev)->se_sub_dev->se_dev_attrib)
-#define DEV_T10_WWN(dev) (&(dev)->se_sub_dev->t10_wwn)
-
struct se_hba {
u16 hba_tpgt;
u32 hba_id;
@@ -842,8 +821,6 @@ struct se_hba {
struct se_subsystem_api *transport;
} ____cacheline_aligned;
-#define SE_HBA(dev) ((dev)->se_hba)
-
struct se_port_stat_grps {
struct config_group stat_group;
struct config_group scsi_port_group;
@@ -870,9 +847,6 @@ struct se_lun {
struct se_port_stat_grps port_stat_grps;
} ____cacheline_aligned;
-#define SE_LUN(cmd) ((cmd)->se_lun)
-#define PORT_STAT_GRP(lun) (&(lun)->port_stat_grps)
-
struct scsi_port_stats {
u64 cmd_pdus;
u64 tx_data_octets;
@@ -938,8 +912,6 @@ struct se_portal_group {
struct config_group tpg_param_group;
} ____cacheline_aligned;
-#define TPG_TFO(se_tpg) ((se_tpg)->se_tpg_tfo)
-
struct se_wwn {
struct target_fabric_configfs *wwn_tf;
struct config_group wwn_group;
diff --git a/include/target/target_core_transport.h b/include/target/target_core_transport.h
index da982e9..1dd4d18 100644
--- a/include/target/target_core_transport.h
+++ b/include/target/target_core_transport.h
@@ -351,7 +351,4 @@ struct se_subsystem_api {
unsigned char *(*get_sense_buffer)(struct se_task *);
} ____cacheline_aligned;
-#define TRANSPORT(dev) ((dev)->transport)
-#define HBA_TRANSPORT(hba) ((hba)->transport)
-
#endif /* TARGET_CORE_TRANSPORT_H */
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 028/103] target: Inline transport_get_task_from_execute_queue()
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (26 preceding siblings ...)
2011-07-21 7:05 ` [PATCH 027/103] target: Remove #defines that just perform indirection Nicholas A. Bellinger
@ 2011-07-21 7:05 ` Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 029/103] target: Fix double test of inquiry_prod Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 030/103] target: Remove unneeded casts to void* Nicholas A. Bellinger
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:05 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover
From: Andy Grover <agrover@redhat.com>
Only used in one place, and inline code is pretty short.
(nab: Fix __transport_execute_tasks() bugs from Andy's
original patch)
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_transport.c | 37 ++++++++-----------------------
1 files changed, 10 insertions(+), 27 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 9c8641ea..e4406c9 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -1105,28 +1105,6 @@ static void transport_add_tasks_from_cmd(struct se_cmd *cmd)
spin_unlock_irqrestore(&dev->execute_task_lock, flags);
}
-/* transport_get_task_from_execute_queue():
- *
- * Called with dev->execute_task_lock held.
- */
-static struct se_task *
-transport_get_task_from_execute_queue(struct se_device *dev)
-{
- struct se_task *task;
-
- if (list_empty(&dev->execute_task_list))
- return NULL;
-
- list_for_each_entry(task, &dev->execute_task_list, t_execute_list)
- break;
-
- list_del(&task->t_execute_list);
- atomic_set(&task->task_execute_queue, 0);
- atomic_dec(&dev->execute_tasks);
-
- return task;
-}
-
/* transport_remove_task_from_execute_queue():
*
*
@@ -2541,7 +2519,7 @@ static int __transport_execute_tasks(struct se_device *dev)
{
int error;
struct se_cmd *cmd = NULL;
- struct se_task *task;
+ struct se_task *task = NULL;
unsigned long flags;
/*
@@ -2555,11 +2533,16 @@ check_depth:
dev->dev_tcq_window_closed = 0;
spin_lock_irq(&dev->execute_task_lock);
- task = transport_get_task_from_execute_queue(dev);
- spin_unlock_irq(&dev->execute_task_lock);
-
- if (!task)
+ if (list_empty(&dev->execute_task_list)) {
+ spin_unlock_irq(&dev->execute_task_lock);
return 0;
+ }
+ task = list_first_entry(&dev->execute_task_list,
+ struct se_task, t_execute_list);
+ list_del(&task->t_execute_list);
+ atomic_set(&task->task_execute_queue, 0);
+ atomic_dec(&dev->execute_tasks);
+ spin_unlock_irq(&dev->execute_task_lock);
atomic_dec(&dev->depth_left);
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 029/103] target: Fix double test of inquiry_prod
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (27 preceding siblings ...)
2011-07-21 7:05 ` [PATCH 028/103] target: Inline transport_get_task_from_execute_queue() Nicholas A. Bellinger
@ 2011-07-21 7:05 ` Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 030/103] target: Remove unneeded casts to void* Nicholas A. Bellinger
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:05 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Nicholas Bellinger
From: Nicholas Bellinger <nab@linux-iscsi.org>
The code in transport_add_device_to_core_hba() really intends to make sure
that neither inquiry_prod nor inquiry_rev is NULL.
Signed-off-by: Roland Dreier <roland@purestorage.com>
Signed-off-by: Nicholas A. Bellinger <nab@linux-iscsi.org>
---
drivers/target/target_core_transport.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index e4406c9..6f5d4df 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -1623,7 +1623,7 @@ struct se_device *transport_add_device_to_core_hba(
* setup.
*/
if (dev->transport->transport_type != TRANSPORT_PLUGIN_PHBA_PDEV) {
- if (!(inquiry_prod) || !(inquiry_prod)) {
+ if (!inquiry_prod || !inquiry_rev) {
printk(KERN_ERR "All non TCM/pSCSI plugins require"
" INQUIRY consts\n");
goto out;
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 030/103] target: Remove unneeded casts to void*
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
` (28 preceding siblings ...)
2011-07-21 7:05 ` [PATCH 029/103] target: Fix double test of inquiry_prod Nicholas A. Bellinger
@ 2011-07-21 7:05 ` Nicholas A. Bellinger
29 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-21 7:05 UTC (permalink / raw)
To: target-devel; +Cc: linux-scsi, Andy Grover
From: Andy Grover <agrover@redhat.com>
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
---
drivers/target/loopback/tcm_loop.c | 12 ++++++------
drivers/target/target_core_configfs.c | 4 ++--
drivers/target/target_core_file.c | 2 +-
drivers/target/target_core_iblock.c | 6 +++---
drivers/target/target_core_pscsi.c | 10 +++++-----
drivers/target/target_core_rd.c | 4 ++--
drivers/target/target_core_transport.c | 8 ++++----
drivers/target/tcm_fc/tfc_conf.c | 2 +-
8 files changed, 24 insertions(+), 24 deletions(-)
diff --git a/drivers/target/loopback/tcm_loop.c b/drivers/target/loopback/tcm_loop.c
index 1f1695a..e1d17b9 100644
--- a/drivers/target/loopback/tcm_loop.c
+++ b/drivers/target/loopback/tcm_loop.c
@@ -171,7 +171,7 @@ static int tcm_loop_new_cmd_map(struct se_cmd *se_cmd)
*/
if (scsi_sg_count(sc)) {
se_cmd->se_cmd_flags |= SCF_PASSTHROUGH_SG_TO_MEM;
- mem_ptr = (void *)scsi_sglist(sc);
+ mem_ptr = scsi_sglist(sc);
/*
* For BIDI commands, pass in the extra READ buffer
* to transport_generic_map_mem_to_cmd() below..
@@ -179,7 +179,7 @@ static int tcm_loop_new_cmd_map(struct se_cmd *se_cmd)
if (se_cmd->t_task->t_tasks_bidi) {
struct scsi_data_buffer *sdb = scsi_in(sc);
- mem_bidi_ptr = (void *)sdb->table.sgl;
+ mem_bidi_ptr = sdb->table.sgl;
sg_no_bidi = sdb->table.nents;
}
} else {
@@ -384,7 +384,7 @@ static int tcm_loop_device_reset(struct scsi_cmnd *sc)
/*
* Allocate the LUN_RESET TMR
*/
- se_cmd->se_tmr_req = core_tmr_alloc_req(se_cmd, (void *)tl_tmr,
+ se_cmd->se_tmr_req = core_tmr_alloc_req(se_cmd, tl_tmr,
TMR_LUN_RESET);
if (IS_ERR(se_cmd->se_tmr_req))
goto release;
@@ -904,7 +904,7 @@ static int tcm_loop_queue_status(struct se_cmd *se_cmd)
((se_cmd->se_cmd_flags & SCF_TRANSPORT_TASK_SENSE) ||
(se_cmd->se_cmd_flags & SCF_EMULATED_TASK_SENSE))) {
- memcpy((void *)sc->sense_buffer, (void *)se_cmd->sense_buffer,
+ memcpy(sc->sense_buffer, se_cmd->sense_buffer,
SCSI_SENSE_BUFFERSIZE);
sc->result = SAM_STAT_CHECK_CONDITION;
set_driver_byte(sc, DRIVER_SENSE);
@@ -1054,7 +1054,7 @@ static int tcm_loop_make_nexus(
* transport_register_session()
*/
__transport_register_session(se_tpg, tl_nexus->se_sess->se_node_acl,
- tl_nexus->se_sess, (void *)tl_nexus);
+ tl_nexus->se_sess, tl_nexus);
tl_tpg->tl_hba->tl_nexus = tl_nexus;
printk(KERN_INFO "TCM_Loop_ConfigFS: Established I_T Nexus to emulated"
" %s Initiator Port: %s\n", tcm_loop_dump_proto_id(tl_hba),
@@ -1242,7 +1242,7 @@ struct se_portal_group *tcm_loop_make_naa_tpg(
* Register the tl_tpg as a emulated SAS TCM Target Endpoint
*/
ret = core_tpg_register(&tcm_loop_fabric_configfs->tf_ops,
- wwn, &tl_tpg->tl_se_tpg, (void *)tl_tpg,
+ wwn, &tl_tpg->tl_se_tpg, tl_tpg,
TRANSPORT_TPG_TYPE_NORMAL);
if (ret < 0)
return ERR_PTR(-ENOMEM);
diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c
index 64418ef..ac7f765 100644
--- a/drivers/target/target_core_configfs.c
+++ b/drivers/target/target_core_configfs.c
@@ -2037,7 +2037,7 @@ static ssize_t target_core_dev_show(struct config_item *item,
if (!(tc_attr->show))
return -EINVAL;
- return tc_attr->show((void *)se_dev, page);
+ return tc_attr->show(se_dev, page);
}
static ssize_t target_core_dev_store(struct config_item *item,
@@ -2053,7 +2053,7 @@ static ssize_t target_core_dev_store(struct config_item *item,
if (!(tc_attr->store))
return -EINVAL;
- return tc_attr->store((void *)se_dev, page, count);
+ return tc_attr->store(se_dev, page, count);
}
static struct configfs_item_operations target_core_dev_item_ops = {
diff --git a/drivers/target/target_core_file.c b/drivers/target/target_core_file.c
index 0c44bc0..46825ce 100644
--- a/drivers/target/target_core_file.c
+++ b/drivers/target/target_core_file.c
@@ -223,7 +223,7 @@ static struct se_device *fd_create_virtdevice(
dev_limits.queue_depth = FD_DEVICE_QUEUE_DEPTH;
dev = transport_add_device_to_core_hba(hba, &fileio_template,
- se_dev, dev_flags, (void *)fd_dev,
+ se_dev, dev_flags, fd_dev,
&dev_limits, "FILEIO", FD_VERSION);
if (!(dev))
goto fail;
diff --git a/drivers/target/target_core_iblock.c b/drivers/target/target_core_iblock.c
index fb15987..b0bafe4 100644
--- a/drivers/target/target_core_iblock.c
+++ b/drivers/target/target_core_iblock.c
@@ -74,7 +74,7 @@ static int iblock_attach_hba(struct se_hba *hba, u32 host_id)
ib_host->iblock_host_id = host_id;
- hba->hba_ptr = (void *) ib_host;
+ hba->hba_ptr = ib_host;
printk(KERN_INFO "CORE_HBA[%d] - TCM iBlock HBA Driver %s on"
" Generic Target Core Stack %s\n", hba->hba_id,
@@ -172,7 +172,7 @@ static struct se_device *iblock_create_virtdevice(
ib_dev->ibd_bd = bd;
dev = transport_add_device_to_core_hba(hba,
- &iblock_template, se_dev, dev_flags, (void *)ib_dev,
+ &iblock_template, se_dev, dev_flags, ib_dev,
&dev_limits, "IBLOCK", IBLOCK_VERSION);
if (!(dev))
goto failed;
@@ -593,7 +593,7 @@ static struct bio *iblock_get_bio(
DEBUG_IBLOCK("Allocated bio: %p task_size: %u\n", bio, task->task_size);
bio->bi_bdev = ib_dev->ibd_bd;
- bio->bi_private = (void *) task;
+ bio->bi_private = task;
bio->bi_destructor = iblock_bio_destructor;
bio->bi_end_io = &iblock_bio_done;
bio->bi_sector = lba;
diff --git a/drivers/target/target_core_pscsi.c b/drivers/target/target_core_pscsi.c
index 44a79a5..823da67 100644
--- a/drivers/target/target_core_pscsi.c
+++ b/drivers/target/target_core_pscsi.c
@@ -72,7 +72,7 @@ static int pscsi_attach_hba(struct se_hba *hba, u32 host_id)
phv->phv_host_id = host_id;
phv->phv_mode = PHV_VIRUTAL_HOST_ID;
- hba->hba_ptr = (void *)phv;
+ hba->hba_ptr = phv;
printk(KERN_INFO "CORE_HBA[%d] - TCM SCSI HBA Driver %s on"
" Generic Target Core Stack %s\n", hba->hba_id,
@@ -355,7 +355,7 @@ static struct se_device *pscsi_add_device_to_list(
pdv->pdv_sd = sd;
dev = transport_add_device_to_core_hba(hba, &pscsi_template,
- se_dev, dev_flags, (void *)pdv,
+ se_dev, dev_flags, pdv,
&dev_limits, NULL, NULL);
if (!(dev)) {
pdv->pdv_sd = NULL;
@@ -394,7 +394,7 @@ static void *pscsi_allocate_virtdevice(struct se_hba *hba, const char *name)
pdv->pdv_se_hba = hba;
printk(KERN_INFO "PSCSI: Allocated pdv: %p for %s\n", pdv, name);
- return (void *)pdv;
+ return pdv;
}
/*
@@ -812,7 +812,7 @@ static inline void pscsi_blk_init_request(
* also set the end_io_data pointer.to struct se_task.
*/
req->end_io = pscsi_req_done;
- req->end_io_data = (void *)task;
+ req->end_io_data = task;
/*
* Load the referenced struct se_task's SCSI CDB into
* include/linux/blkdev.h:struct request->cmd
@@ -822,7 +822,7 @@ static inline void pscsi_blk_init_request(
/*
* Setup pointer for outgoing sense data.
*/
- req->sense = (void *)&pt->pscsi_sense[0];
+ req->sense = &pt->pscsi_sense[0];
req->sense_len = 0;
}
diff --git a/drivers/target/target_core_rd.c b/drivers/target/target_core_rd.c
index fbf06c3..63cd5a7 100644
--- a/drivers/target/target_core_rd.c
+++ b/drivers/target/target_core_rd.c
@@ -66,7 +66,7 @@ static int rd_attach_hba(struct se_hba *hba, u32 host_id)
rd_host->rd_host_id = host_id;
- hba->hba_ptr = (void *) rd_host;
+ hba->hba_ptr = rd_host;
printk(KERN_INFO "CORE_HBA[%d] - TCM Ramdisk HBA Driver %s on"
" Generic Target Core Stack %s\n", hba->hba_id,
@@ -271,7 +271,7 @@ static struct se_device *rd_create_virtdevice(
dev = transport_add_device_to_core_hba(hba,
(rd_dev->rd_direct) ? &rd_dr_template :
- &rd_mcp_template, se_dev, dev_flags, (void *)rd_dev,
+ &rd_mcp_template, se_dev, dev_flags, rd_dev,
&dev_limits, prod, rev);
if (!(dev))
goto fail;
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 6f5d4df..6198a7e 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -201,7 +201,7 @@ struct kmem_cache *t10_alua_tg_pt_gp_mem_cache;
typedef int (*map_func_t)(struct se_task *, u32);
static int transport_generic_write_pending(struct se_cmd *);
-static int transport_processing_thread(void *);
+static int transport_processing_thread(void *param);
static int __transport_execute_tasks(struct se_device *dev);
static void transport_complete_task_attr(struct se_cmd *cmd);
static void transport_direct_request_timeout(struct se_cmd *cmd);
@@ -1548,7 +1548,7 @@ struct se_device *transport_add_device_to_core_hba(
transport_init_queue_obj(&dev->dev_queue_obj);
dev->dev_flags = device_flags;
dev->dev_status |= TRANSPORT_DEVICE_DEACTIVATED;
- dev->dev_ptr = (void *) transport_dev;
+ dev->dev_ptr = transport_dev;
dev->se_hba = hba;
dev->se_sub_dev = se_dev;
dev->transport = transport;
@@ -2908,7 +2908,7 @@ static int transport_get_sense_data(struct se_cmd *cmd)
offset = cmd->se_tfo->set_fabric_sense_len(cmd,
TRANSPORT_SENSE_BUFFER);
- memcpy((void *)&buffer[offset], (void *)sense_buffer,
+ memcpy(&buffer[offset], sense_buffer,
TRANSPORT_SENSE_BUFFER);
cmd->scsi_status = task->task_scsi_status;
/* Automatically padded */
@@ -5379,7 +5379,7 @@ int transport_clear_lun_from_sessions(struct se_lun *lun)
{
struct task_struct *kt;
- kt = kthread_run(transport_clear_lun_thread, (void *)lun,
+ kt = kthread_run(transport_clear_lun_thread, lun,
"tcm_cl_%u", lun->unpacked_lun);
if (IS_ERR(kt)) {
printk(KERN_ERR "Unable to start clear_lun thread\n");
diff --git a/drivers/target/tcm_fc/tfc_conf.c b/drivers/target/tcm_fc/tfc_conf.c
index 8c5067c..58e4745 100644
--- a/drivers/target/tcm_fc/tfc_conf.c
+++ b/drivers/target/tcm_fc/tfc_conf.c
@@ -331,7 +331,7 @@ static struct se_portal_group *ft_add_tpg(
transport_init_queue_obj(&tpg->qobj);
ret = core_tpg_register(&ft_configfs->tf_ops, wwn, &tpg->se_tpg,
- (void *)tpg, TRANSPORT_TPG_TYPE_NORMAL);
+ tpg, TRANSPORT_TPG_TYPE_NORMAL);
if (ret < 0) {
kfree(tpg);
return NULL;
--
1.7.6
^ permalink raw reply related [flat|nested] 38+ messages in thread
* Re: [PATCH 005/103] loopback: Only build tcm_loop as module
2011-07-21 7:04 ` [PATCH 005/103] loopback: Only build tcm_loop as module Nicholas A. Bellinger
@ 2011-07-21 21:18 ` Roland Dreier
2011-07-22 2:54 ` Nicholas A. Bellinger
0 siblings, 1 reply; 38+ messages in thread
From: Roland Dreier @ 2011-07-21 21:18 UTC (permalink / raw)
To: Nicholas A. Bellinger; +Cc: target-devel, linux-scsi, Stefan Hajnoczi
Nick, please drop this patch. It doesn't make any sense -- after my fix that
you have later in your series, tcm_loop works perfectly well built-in (and in
fact is useful for quick debugging, along with "qemu -kernel").
- R.
On Thu, Jul 21, 2011 at 12:04 AM, Nicholas A. Bellinger
<nab@linux-iscsi.org> wrote:
> From: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
>
> The tcm_loop must be built as a module. It cannot be built in.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
> Signed-off-by: Nicholas A. Bellinger <nab@risingtidesystems.com>
> ---
> drivers/target/loopback/Kconfig | 3 ++-
> 1 files changed, 2 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/target/loopback/Kconfig b/drivers/target/loopback/Kconfig
> index 57dcbc2..1bd6a65 100644
> --- a/drivers/target/loopback/Kconfig
> +++ b/drivers/target/loopback/Kconfig
> @@ -1,7 +1,8 @@
> config LOOPBACK_TARGET
> tristate "TCM Virtual SAS target and Linux/SCSI LDD fabric loopback module"
> + depends on m
> help
> - Say Y here to enable the TCM Virtual SAS target and Linux/SCSI LLD
> + Say M here to enable the TCM Virtual SAS target and Linux/SCSI LLD
> fabric loopback module.
>
> config LOOPBACK_TARGET_CDB_DEBUG
> --
> 1.7.6
>
> --
> To unsubscribe from this list: send the line "unsubscribe target-devel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
--
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] 38+ messages in thread
* Re: [PATCH 005/103] loopback: Only build tcm_loop as module
2011-07-21 21:18 ` Roland Dreier
@ 2011-07-22 2:54 ` Nicholas A. Bellinger
0 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-22 2:54 UTC (permalink / raw)
To: Roland Dreier; +Cc: target-devel, linux-scsi, Stefan Hajnoczi
On Thu, 2011-07-21 at 14:18 -0700, Roland Dreier wrote:
> Nick, please drop this patch. It doesn't make any sense -- after my fix that
> you have later in your series, tcm_loop works perfectly well built-in (and in
> fact is useful for quick debugging, along with "qemu -kernel").
>
Thanks Roland, dropping this patch now.
--nab
> - R.
>
> On Thu, Jul 21, 2011 at 12:04 AM, Nicholas A. Bellinger
> <nab@linux-iscsi.org> wrote:
> > From: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
> >
> > The tcm_loop must be built as a module. It cannot be built in.
> >
> > Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
> > Signed-off-by: Nicholas A. Bellinger <nab@risingtidesystems.com>
> > ---
> > drivers/target/loopback/Kconfig | 3 ++-
> > 1 files changed, 2 insertions(+), 1 deletions(-)
> >
> > diff --git a/drivers/target/loopback/Kconfig b/drivers/target/loopback/Kconfig
> > index 57dcbc2..1bd6a65 100644
> > --- a/drivers/target/loopback/Kconfig
> > +++ b/drivers/target/loopback/Kconfig
> > @@ -1,7 +1,8 @@
> > config LOOPBACK_TARGET
> > tristate "TCM Virtual SAS target and Linux/SCSI LDD fabric loopback module"
> > + depends on m
> > help
> > - Say Y here to enable the TCM Virtual SAS target and Linux/SCSI LLD
> > + Say M here to enable the TCM Virtual SAS target and Linux/SCSI LLD
> > fabric loopback module.
> >
> > config LOOPBACK_TARGET_CDB_DEBUG
> > --
> > 1.7.6
> >
> > --
> > To unsubscribe from this list: send the line "unsubscribe target-devel" 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] 38+ messages in thread
* Re: [PATCH 018/103] target: Remove struct se_global
2011-07-21 7:04 ` [PATCH 018/103] target: Remove struct se_global Nicholas A. Bellinger
@ 2011-07-23 8:38 ` Fubo Chen
2011-07-23 9:17 ` Nicholas A. Bellinger
2011-07-23 8:40 ` Fubo Chen
1 sibling, 1 reply; 38+ messages in thread
From: Fubo Chen @ 2011-07-23 8:38 UTC (permalink / raw)
To: linux-iscsi-target-dev
Cc: target-devel, linux-scsi, Andy Grover, Nicholas A. Bellinger
On Thu, Jul 21, 2011 at 9:04 AM, Nicholas A. Bellinger
<nab@linux-iscsi.org> wrote:
> diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c
> index 03cda0e..82c8e9a 100644
> --- a/drivers/target/target_core_configfs.c
> +++ b/drivers/target/target_core_configfs.c
> @@ -37,6 +37,7 @@
> #include <linux/parser.h>
> #include <linux/syscalls.h>
> #include <linux/configfs.h>
> +#include <linux/spinlock.h>
>
> #include <target/target_core_base.h>
> #include <target/target_core_device.h>
> @@ -52,6 +53,8 @@
> #include "target_core_rd.h"
> #include "target_core_stat.h"
>
> +extern struct t10_alua_lu_gp *default_lu_gp;
I see here an extern declaration in a .c file. Shouldn't instead that
be moved to a header file ?
Fubo.
--
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] 38+ messages in thread
* Re: [PATCH 018/103] target: Remove struct se_global
2011-07-21 7:04 ` [PATCH 018/103] target: Remove struct se_global Nicholas A. Bellinger
2011-07-23 8:38 ` Fubo Chen
@ 2011-07-23 8:40 ` Fubo Chen
1 sibling, 0 replies; 38+ messages in thread
From: Fubo Chen @ 2011-07-23 8:40 UTC (permalink / raw)
To: linux-iscsi-target-dev
Cc: target-devel, linux-scsi, Andy Grover, Nicholas A. Bellinger
On Thu, Jul 21, 2011 at 9:04 AM, Nicholas A. Bellinger
<nab@linux-iscsi.org> wrote:
> From: Andy Grover <agrover@redhat.com>
> diff --git a/drivers/target/target_core_tpg.c b/drivers/target/target_core_tpg.c
> index dc024c1..4e2bee9 100644
> --- a/drivers/target/target_core_tpg.c
> +++ b/drivers/target/target_core_tpg.c
> @@ -44,6 +44,12 @@
> #include <target/target_core_fabric_ops.h>
>
> #include "target_core_hba.h"
> +#include "target_core_stat.h"
> +
> +extern struct se_device *g_lun0_dev;
Same here: another extern declaraion in a C file.
Fubo.
--
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] 38+ messages in thread
* Re: [PATCH 018/103] target: Remove struct se_global
2011-07-23 8:38 ` Fubo Chen
@ 2011-07-23 9:17 ` Nicholas A. Bellinger
2011-07-23 13:36 ` Christoph Hellwig
0 siblings, 1 reply; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-23 9:17 UTC (permalink / raw)
To: Fubo Chen; +Cc: linux-iscsi-target-dev, target-devel, linux-scsi, Andy Grover
On Sat, 2011-07-23 at 10:38 +0200, Fubo Chen wrote:
> On Thu, Jul 21, 2011 at 9:04 AM, Nicholas A. Bellinger
> <nab@linux-iscsi.org> wrote:
> > diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c
> > index 03cda0e..82c8e9a 100644
> > --- a/drivers/target/target_core_configfs.c
> > +++ b/drivers/target/target_core_configfs.c
> > @@ -37,6 +37,7 @@
> > #include <linux/parser.h>
> > #include <linux/syscalls.h>
> > #include <linux/configfs.h>
> > +#include <linux/spinlock.h>
> >
> > #include <target/target_core_base.h>
> > #include <target/target_core_device.h>
> > @@ -52,6 +53,8 @@
> > #include "target_core_rd.h"
> > #include "target_core_stat.h"
> >
> > +extern struct t10_alua_lu_gp *default_lu_gp;
>
> I see here an extern declaration in a .c file. Shouldn't instead that
> be moved to a header file ?
>
I don't really have a strong preference here either way, and will defer
to Andy and Christoph for mainline style consideration on these two
particular bits..
Thanks for your review!
--nab
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 018/103] target: Remove struct se_global
2011-07-23 9:17 ` Nicholas A. Bellinger
@ 2011-07-23 13:36 ` Christoph Hellwig
2011-07-23 21:26 ` Nicholas A. Bellinger
0 siblings, 1 reply; 38+ messages in thread
From: Christoph Hellwig @ 2011-07-23 13:36 UTC (permalink / raw)
To: Nicholas A. Bellinger
Cc: Fubo Chen, linux-iscsi-target-dev, target-devel, linux-scsi, Andy Grover
On Sat, Jul 23, 2011 at 02:17:17AM -0700, Nicholas A. Bellinger wrote:
> > > #include "target_core_rd.h"
> > > #include "target_core_stat.h"
> > >
> > > +extern struct t10_alua_lu_gp *default_lu_gp;
> >
> > I see here an extern declaration in a .c file. Shouldn't instead that
> > be moved to a header file ?
> >
>
> I don't really have a strong preference here either way, and will defer
> to Andy and Christoph for mainline style consideration on these two
> particular bits..
externs should always be in headers. In fact sparse will complain about
having it in the .c file.
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 018/103] target: Remove struct se_global
2011-07-23 13:36 ` Christoph Hellwig
@ 2011-07-23 21:26 ` Nicholas A. Bellinger
0 siblings, 0 replies; 38+ messages in thread
From: Nicholas A. Bellinger @ 2011-07-23 21:26 UTC (permalink / raw)
To: Christoph Hellwig
Cc: Fubo Chen, linux-iscsi-target-dev, target-devel, linux-scsi, Andy Grover
On Sat, 2011-07-23 at 09:36 -0400, Christoph Hellwig wrote:
> On Sat, Jul 23, 2011 at 02:17:17AM -0700, Nicholas A. Bellinger wrote:
> > > > #include "target_core_rd.h"
> > > > #include "target_core_stat.h"
> > > >
> > > > +extern struct t10_alua_lu_gp *default_lu_gp;
> > >
> > > I see here an extern declaration in a .c file. Shouldn't instead that
> > > be moved to a header file ?
> > >
> >
> > I don't really have a strong preference here either way, and will defer
> > to Andy and Christoph for mainline style consideration on these two
> > particular bits..
>
> externs should always be in headers. In fact sparse will complain about
> having it in the .c file.
>
Ok, fixing this up now in lio-core-2.6.git and will get this included
into the initial merge.
^ permalink raw reply [flat|nested] 38+ messages in thread
end of thread, other threads:[~2011-07-23 21:35 UTC | newest]
Thread overview: 38+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2011-07-21 7:04 [PATCH 000/103] target: Updates for v3.1-rc1 Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 001/103] target: Check LUN numbers in transport_get_lun_for_[cmd,tmr] Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 002/103] target: Minor header comment fixes Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 003/103] target: use MAINTENANCE_IN and MAINTENANCE_OUT definitions in scsi.h Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 004/103] target: Remove unnecessary *cdb transport_get_lun_for_cmd parameter Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 005/103] loopback: Only build tcm_loop as module Nicholas A. Bellinger
2011-07-21 21:18 ` Roland Dreier
2011-07-22 2:54 ` Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 006/103] target: Remove unused su_group usage in fabric register/dergister Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 007/103] target: use errno values instead of returning -1 for everything Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 008/103] target: Rename transport_calc_sg_num to transport_init_task_sg Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 009/103] target: Fix leak in error path in transport_init_task_sg Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 010/103] target: Remove pscsi_get_sh() in _pscsi.c Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 011/103] target: Make two runtime checks into WARN_ONs Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 012/103] tcm_loop: off by one in tcm_loop_make_naa_tpg() Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 013/103] target: Remove hba queue depth and convert to spin_lock_irq usage Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 014/103] target: dev->dev_status_queue_obj is unused Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 015/103] target: Make struct se_queue_req.cmd type struct se_cmd * Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 016/103] target: Remove __transport_get_qr_from_queue() Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 017/103] target: Rename se_dev->g_se_dev_list to se_dev_node Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 018/103] target: Remove struct se_global Nicholas A. Bellinger
2011-07-23 8:38 ` Fubo Chen
2011-07-23 9:17 ` Nicholas A. Bellinger
2011-07-23 13:36 ` Christoph Hellwig
2011-07-23 21:26 ` Nicholas A. Bellinger
2011-07-23 8:40 ` Fubo Chen
2011-07-21 7:04 ` [PATCH 019/103] target: Simplify scsi mib index table code Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 020/103] target: Make dev_queue_obj a member of se_device instead of a pointer Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 021/103] target: remove extraneous returns at end of void functions Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 022/103] target: Ensure transport_dump_vpd_ident_type returns null-terminated str Nicholas A. Bellinger
2011-07-21 7:04 ` [PATCH 023/103] target: Function pointers don't need to use '&' to be assigned Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 024/103] target: Fix comment in __transport_execute_tasks() Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 025/103] target: Misc style cleanups Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 026/103] target: rename struct pr_reservation_template to pr_reservation Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 027/103] target: Remove #defines that just perform indirection Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 028/103] target: Inline transport_get_task_from_execute_queue() Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 029/103] target: Fix double test of inquiry_prod Nicholas A. Bellinger
2011-07-21 7:05 ` [PATCH 030/103] target: Remove unneeded casts to void* Nicholas A. Bellinger
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.