I hit a crash in qla2x00_abort_all_cmds() if the qla2xxx module is
unloaded right after it is loaded. I debugged this down to the abort
handling improperly treating a command of type SRB_ADISC_CMD as if it
had a bsg_job to complete when that command actually uses the iocb_cmd
part of the union. (I guess to hit this one has to unload the module
while the async FC initialization is still in progress)
It seems we should only look for a bsg_job if type is SRB_ELS_CMD_RPT,
SRB_ELS_CMD_HST or SRB_CT_CMD, so switch the test to make that explicit.
Signed-off-by: Roland Dreier <roland@purestorage.com>
Acked-by: Chad Dupuis <chad.dupuis@qlogic.com>
Cc: stable@kernel.org
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* git://bedivere.hansenpartnership.com/git/scsi-rc-fixes-2.6:
[SCSI] 3w-9xxx: fix iommu_iova leak
[SCSI] cxgb3i: convert cdev->l2opt to use rcu to prevent NULL dereference
[SCSI] scsi: qla4xxx needs libiscsi.o
[SCSI] libsas: fix failure to revalidate domain for anything but the first expander child.
[SCSI] aacraid: reset should disable MSI interrupt
There are numerous broken references to Documentation files (in other
Documentation files, in comments, etc.). These broken references are
caused by typo's in the references, and by renames or removals of the
Documentation files. Some broken references are simply odd.
Fix these broken references, sometimes by dropping the irrelevant text
they were part of.
Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Following reports on the list, it looks like the 3e-9xxx driver will leak dma
mappings every time we get a transient queueing error back from the card.
This is because it maps the sg list in the routine that sends the command, but
doesn't unmap again in the transient failure path (even though the command is
sent back to the block layer). Fix by unmapping before returning the status.
Reported-by: Chris Boot <bootc@bootc.net>
Tested-by: Chris Boot <bootc@bootc.net>
Acked-by: Adam Radford <aradford@gmail.com>
Cc: stable@kernel.org
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
This oops was reported recently:
d:mon> e
cpu 0xd: Vector: 300 (Data Access) at [c0000000fd4c7120]
pc: d00000000076f194: .t3_l2t_get+0x44/0x524 [cxgb3]
lr: d000000000b02108: .init_act_open+0x150/0x3d4 [cxgb3i]
sp: c0000000fd4c73a0
msr: 8000000000009032
dar: 0
dsisr: 40000000
current = 0xc0000000fd640d40
paca = 0xc00000000054ff80
pid = 5085, comm = iscsid
d:mon> t
[c0000000fd4c7450] d000000000b02108 .init_act_open+0x150/0x3d4 [cxgb3i]
[c0000000fd4c7500] d000000000e45378 .cxgbi_ep_connect+0x784/0x8e8 [libcxgbi]
[c0000000fd4c7650] d000000000db33f0 .iscsi_if_rx+0x71c/0xb18
[scsi_transport_iscsi2]
[c0000000fd4c7740] c000000000370c9c .netlink_data_ready+0x40/0xa4
[c0000000fd4c77c0] c00000000036f010 .netlink_sendskb+0x4c/0x9c
[c0000000fd4c7850] c000000000370c18 .netlink_sendmsg+0x358/0x39c
[c0000000fd4c7950] c00000000033be24 .sock_sendmsg+0x114/0x1b8
[c0000000fd4c7b50] c00000000033d208 .sys_sendmsg+0x218/0x2ac
[c0000000fd4c7d70] c00000000033f55c .sys_socketcall+0x228/0x27c
[c0000000fd4c7e30] c0000000000086a4 syscall_exit+0x0/0x40
--- Exception: c01 (System Call) at 00000080da560cfc
The root cause was an EEH error, which sent us down the offload_close path in
the cxgb3 driver, which in turn sets cdev->l2opt to NULL, without regard for
upper layer driver (like the cxgbi drivers) which might have execution contexts
in the middle of its use. The result is the oops above, when t3_l2t_get attempts
to dereference L2DATA(cdev)->nentries in arp_hash right after the EEH error handler sets it to NULL.
The fix is to prevent the setting of the NULL pointer until after there are no
further users of it. The t3cdev->l2opt pointer is now converted to be an rcu
pointer and the L2DATA macro is now called under the protection of the
rcu_read_lock(). When the EEH error path:
t3_adapter_error->offload_close->cxgb3_offload_deactivate
Is exectured, setting of that l2opt pointer to NULL, is now gated on an rcu
quiescence point, preventing, allowing L2DATA callers to safely check for a NULL
pointer without concern that the underlying data will be freeded before the
pointer is dereferenced.
This has been tested by the reporter and shown to fix the reproted oops
[nhorman: fix up unitinialised variable reported by Dan Carpenter]
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Karen Xie <kxie@chelsio.com>
Cc: stable@kernel.org
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
sector_t can be different types, so cast it to its largest possible
type.
drivers/scsi/qla2xxx/qla_isr.c:1509:5: warning: format '%lx' expects type 'long unsigned int', but argument 5 has type 'sector_t'
Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
SCSI_ISCI needs to select SCSI_SAS_HOST_SMP to ensure that all
needed symbols are available to it.
Fixes this build error:
ERROR: "try_test_sas_gpio_gp_bit" [drivers/scsi/isci/isci.ko] undefined!
Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When the ipr driver decides to dump the adapter, it changes the
sdt_state to GET_DUMP, then prepares the adapter so that the dump
can be read. However, if the ipr worker thread wakes up for some
reason before the driver has put the adapter in a state where it
can succesfully dump the adapter, the driver will start dumping
the adapter too early, which can potentially trigger a BUG check
in the pci config blocking API. Fix this by adding a new
sdt_state to differentiate between the ipr driver wanting to dump
the adapter in the near future and wanting to dump the adapter now.
Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Gracefully handle bnx2fc_map_sg failure, so that queuecommand returns host busy
and SCSI-ml can retry the IO.
Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
The OSD protocol calls for all kind of security levels that use
CRYPTO_HMAC and SH1, but the current code only supports NO_SEC,
which does not use any of these.
Remove a wrong FIXME that calls for them. Thanks Maxin for
reporting on this.
Reported-by: "Maxin B. John" <maxin.john@gmail.com>
Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
Sizeof a pointer-typed expression returns the size of the pointer, not that
of the pointed data.
The semantic patch that fixes this problem is as follows:
(http://coccinelle.lip6.fr/)
// <smpl>
@@
expression *e;
type T;
identifier f;
@@
f(...,(T)e,...,
-sizeof(e)
+sizeof(*e)
,...)
// </smpl>
Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Mutual exclusion is redundant here because all the paths in the call graph
leading to esp_driver_ops.send_dma_cmd() happen under spin_lock_irqsave/
spin_lock_irqrestore. Remove it.
Tested on a Mac Quadra 660av and a Mac LC 630.
Signed-off-by: Finn Thain <fthain@telegraphics.com.au>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Support added for controllers capable of multi reply queues.
The following are the modifications to the driver to support NUMA.
1) Create the new structure adapter_reply_queue to contain the reply queue
info for every msix vector. This object will contain a
reply_post_host_index, reply_post_free for each instance, msix_index, among
other parameters. We will track all the reply queues on a link list called
ioc->reply_queue_list. Each reply queue is aligned with each IRQ, and is
passed to the interrupt via the bus_id parameter.
(2) The driver will figure out the msix_vector_count from the PCIe MSIX
capabilities register instead of the IOC Facts->MaxMSIxVectors. This is
because the firmware is not filling in this field until the driver has
already registered MSIX support.
(3) If the ioc_facts reports that the controller is MSIX compatible in the
capabilities, then the driver will request for multiple irqs. This count
is calculated based on the minimum between the online cpus available and
the ioc->msix_vector_count. This count is reported to firmware in the
ioc_init request.
(4) New routines were added _base_free_irq and _base_request_irq, so
registering and freeing msix vectors were done thru simple function API.
(5) The new routine _base_assign_reply_queues was added to align the msix
indexes across cpus. This will initialize the array called
ioc->cpu_msix_table. This array is looked up on every MPI request so the
MSIxIndex is set appropriately.
(6) A new shost sysfs attribute was added to report the reply_queue_count.
(7) User needs to set the affinity cpu mask, so the interrupts occur on the
same cpu that sent the original request.
Signed-off-by: Nagalakshmi Nandigama <nagalakshmi.nandigama@lsi.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
This patch adds couple more Vendor/Product IDs for RDAC.. There are no
functional changes.
Signed-off-by: Babu Moger <babu.moger@netapp.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Basic support to initialize the gpio unit, accept an incomming
SAS_GPIO_REG_TX_GP bitstream, and translate it to the ODx.n fields in
the hardware registers. If register indexes outside the supported range
are specified in the SMP frame we simply accept the write and return how
many registers (SFF-8485) were written (libsas reports this as residue
in the request).
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Add SFF-8485 v0.7 / SAS-1 smp-write-gpio register support to libsas.
Defer SAS-2 support unless/until it defines an sgpio interface.
Minimum implementation needed to get the lights blinking.
try_test_sas_gpio_gp_bit() provides a common method to parse the
incoming write data (raw bitstream), and the to_sas_gpio_gp_bit() helper
routine can be used as a basis for the set/clear operations for the
'read' implementation. Host implementations parse as many bits
(ODx.[012]) as are locally supported and report the number of registers
successfully written. If the submitted data overruns the internal
number of registers available report the write as a success with the
number of bytes remaining reported in ->resid_len.
Example (assuming an active backplane) set the "identify" pattern for
the first 21 devices:
smp_write_gpio --count=2 --data=92,49,24,92,24,92,49,24 -t 4 --index=1 /dev/bsg/sas_hostX
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
In an enclosure model where there are chaining expanders to a large body
of storage, it was discovered that libsas, responding to a broadcast
event change, would only revalidate the domain of first child expander
in the list.
The issue is that the pointer value to the discovered source device was
used to break out of the loop, rather than the content of the pointer.
This still remains non-compliant as the revalidate domain code is
supposed to loop through all child expanders, and not stop at the first
one it finds that reports a change count. However, the design of this
routine does not allow multiple device discoveries and that would be a
more complicated set of patches reserved for another day. We are fixing
the glaring bug rather than refactoring the code.
Signed-off-by: Mark Salyzyn <msalyzyn@us.xyratex.com>
Cc: stable@kernel.org
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
qla4xxx now uses iscsi_boot_sysfs to export the targets used
for boot to sysfs. It needs to select that config option
to make sure that module is also built.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Firmware asserts when the same CQE is armed twice. This scenario happens during
RSCN stress tests as driver incorrects arms the CQ after the session is
offloaded.
Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
It is not required to hold rtnl_lock and bnx2fc_dev_lock when calling
bnx2fc_if_destroy, as the locking is only required to serialize creation and
deletion of fcoe instances. More importantly, this unnecessary locking causes
deadlock as bnx2fc_if_destroy calls fc_remove_host holding rtnl_lock.
Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
When bnx2fc receives an UNREGISTER event on a vlan interface it calls
destroy on all interfaces that matches the physical interface. Add
vlan_id check to destroy only the vlan interface that generated the
event.
Signed-off-by: Nithin Nayak Sujir <nsujir@broadcom.com>
Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
ABTS was not issued for timed out REC, as REC completion handler exits out if
the IO completed. Check for timed out REC and issue ABTS before proceeding with
further processing in REC completion handler. Also, initialize rec_retry and
srr_retry before starting the IO.
Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Link up event is generated to the driver even before vlan discovery has
started. Because of this driver can send discovery solicitation on a stale
vlan. Call fcoe_ctlr_link_up() only when the driver is in enabled state, which
implies the vlan discovery is complete before sending solicitation.
Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
If the max receive frame size is changed during link down, the driver uses the
same value after linkup unless it is reset to default.
Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
This patch does several things:
- introduces __ethtool_get_settings which is called from ethtool code and
from drivers as well. Put ASSERT_RTNL there.
- dev_ethtool_get_settings() is replaced by __ethtool_get_settings()
- changes calling in drivers so rtnl locking is respected. In
iboe_get_rate was previously ->get_settings() called unlocked. This
fixes it. Also prb_calc_retire_blk_tmo() in af_packet.c had the same
problem. Also fixed by calling __dev_get_by_index() instead of
dev_get_by_index() and holding rtnl_lock for both calls.
- introduces rtnl_lock in bnx2fc_vport_create() and fcoe_vport_create()
so bnx2fc_if_create() and fcoe_if_create() are called locked as they
are from other places.
- use __ethtool_get_settings() in bonding code
Signed-off-by: Jiri Pirko <jpirko@redhat.com>
v2->v3:
-removed dev_ethtool_get_settings()
-added ASSERT_RTNL into __ethtool_get_settings()
-prb_calc_retire_blk_tmo - use __dev_get_by_index() and lock
around it and __ethtool_get_settings() call
v1->v2:
add missing export_symbol
Reviewed-by: Ben Hutchings <bhutchings@solarflare.com> [except FCoE bits]
Acked-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
It was pointed out by 'make versioncheck' that some includes of
linux/version.h are not needed in drivers/scsi/.
This patch removes them.
Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
This is a resend from the original, changing the title from PATCH to
RFC(since this is a review for commit, and I should have put that the first go around).
and also removing some of the commit's with ia64 and bash since it is significant.
let me know if I might have missed anything etc..
Signed-off-by: Justin P. Mattock <justinmattock@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Since printk_ratelimit() shouldn't be used anymore (see comment in
include/linux/printk.h), replace it with printk_ratelimited.
Signed-off-by: Christian Dietrich <christian.dietrich@informatik.uni-erlangen.de>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
* git://bedivere.hansenpartnership.com/git/scsi-rc-fixes-2.6: (25 commits)
[SCSI] bnx2i: Fixed the endian on TTT for NOP out transmission
[SCSI] libfc: fix referencing to fc_fcp_pkt from the frame pointer via fr_fsp()
[SCSI] libfc: block SCSI eh thread for blocked rports
[SCSI] libfc: fix fc_eh_host_reset
[SCSI] fcoe: Fix deadlock between fip's recv_work and rtnl
[SCSI] qla2xxx: Update version number to 8.03.07.07-k.
[SCSI] qla2xxx: Set the task attributes after memsetting fcp cmnd.
[SCSI] qla2xxx: Correct inadvertent loop state transitions during port-update handling.
[SCSI] qla2xxx: Save and restore irq in the response queue interrupt handler.
[SCSI] qla2xxx: Double check for command completion if abort mailbox command fails.
[SCSI] qla2xxx: Acquire hardware lock while manipulating dsd list.
[SCSI] qla2xxx: Fix qla24xx revision check while enabling interrupts.
[SCSI] qla2xxx: T10 DIF - Fix incorrect error reporting.
[SCSI] qla2xxx: T10 DIF - Handle uninitalized sectors.
[SCSI] hpsa: fix physical device lun and target numbering problem
[SCSI] hpsa: fix problem that OBDR devices are not detected
[SCSI] isci: add version number
[SCSI] isci: fix event-get pointer increment
[SCSI] isci: dynamic interrupt coalescing
[SCSI] isci: Leave requests alone if already terminating.
...
When CONFIG_NET is disabled, SCSI_QLA_ISCSI selects SCSI_ISCSI_ATTRS,
which uses network interfaces, so the build fails with multiple errors:
warning: (ISCSI_TCP && SCSI_CXGB3_ISCSI && SCSI_CXGB4_ISCSI && SCSI_QLA_ISCSI && INFINIBAND_ISER) selects SCSI_ISCSI_ATTRS which has unmet direct dependencies (SCSI && NET)
ERROR: "skb_trim" [drivers/scsi/scsi_transport_iscsi.ko] undefined!
ERROR: "netlink_kernel_create" [drivers/scsi/scsi_transport_iscsi.ko] undefined!
ERROR: "netlink_kernel_release" [drivers/scsi/scsi_transport_iscsi.ko] undefined!
...
so make SCSI_QLA_ISCSI also depend on NET to prevent the build errors.
Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Cc: Ravi Anand <ravi.anand@qlogic.com>
Cc: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Cc: iscsi-driver@qlogic.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The alua device handler starts the first retry after 10 seconds,
and increases it times 10 for each round.
This leads to an unnecessary delay. This patch modifies it to
start after one second, and increase by a factor of two.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
For Target Portal Group IDs occupying the full 2 bytes in the
RTPG response, the following group_id check in the alua_rtpg
routine always fails in scsi_dh_alua.c:
if (h->group_id == (ucp[2] << 8) + ucp[3]) {
This causes the ALUA handler to wrongly identify the AAS of
a specified device as well as incorrectly interpreting the
supported AAS of the target as seen by the following entries
in the /var/log/messages:
"alua: port group 3ea state A supports tousna"
"alua: port group 3e9 state A supports tousna"
This is because 'ucp' is wrongly declared in alua_rtpg as
a character pointer instead of an unsigned character pointer.
Signed-off-by: Martin George <marting@netapp.com>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Avoid attaching a hardware handler to a device which is
already scheduled for deletion.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
When activating a patch we should always update the TPGS state
as it might have changed in between.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Some device handler types are not tied to the vendor/model
but rather to a specific capability. Eg ALUA is supported
if the 'TPGS' setting in the standard inquiry is set.
This patch implements a 'match' callback for device handler
which supersedes the original vendor/model lookup and
implements the callback for the ALUA handler.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Fixup some kernel-doc comments to reference to the
correct function name.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Instead of issuing a standard inquiry from within the
alua device handler we can evaluate the TPGS setting from
the existing inquiry data of the sdev and save us the I/O.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
->queuecommand must return either 0, or one of the SCSI_MLQUEUE_* return
values. Non-transient errors are indicated by setting cmd->result before
calling ->scsi_done and returning 0. Fix libsas to adhere to this calling
convention. Note that the DID_ERROR for returns from the low-level driver
might not be correct for all cases, but it's the best we can do with
the current layering in libsas. I also suspect that the pre-existing
handling of -SAS_QUEUE_FULL should really be SCSI_MLQUEUE_HOST_BUSY, but
I'll leave that for a separate change.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
