pandora-kernel.git
11 years ago[SCSI] lpfc 8.3.34: Fix parameter field in CQE to mask for LOCAL_REJECT status
James Smart [Tue, 14 Aug 2012 18:25:43 +0000 (14:25 -0400)]
[SCSI] lpfc 8.3.34: Fix parameter field in CQE to mask for LOCAL_REJECT status

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.34: Fix number of IO channels to match CPUs
James Smart [Tue, 14 Aug 2012 18:25:36 +0000 (14:25 -0400)]
[SCSI] lpfc 8.3.34: Fix number of IO channels to match CPUs

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.34: Add SLI-4 V1 Capacity and Resource Descriptor support
James Smart [Tue, 14 Aug 2012 18:25:29 +0000 (14:25 -0400)]
[SCSI] lpfc 8.3.34: Add SLI-4 V1 Capacity and Resource Descriptor support

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.34: Add LOGO support after ABTS compliance
James Smart [Tue, 14 Aug 2012 18:25:21 +0000 (14:25 -0400)]
[SCSI] lpfc 8.3.34: Add LOGO support after ABTS compliance

Make compliant with FC specs by sending LOGO after ABTS timeouts

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] scsi_lib: Set the device state from transport-offline to running
Vikas Chaudhary [Thu, 9 Aug 2012 08:51:30 +0000 (04:51 -0400)]
[SCSI] scsi_lib: Set the device state from transport-offline to running

FC and iSCSI class set SCSI devices to transport-offline state after
fast_io_fail/replacement_timeout has fired, but after relogin, function
scsi_internal_device_unblock() is not setting scsi device state to running.
Due to this the devices even after being relogged in remain offline.

Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Reviewed-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] qla4xxx: Update driver version to 5.02.00-k19
Vikas Chaudhary [Tue, 7 Aug 2012 11:57:17 +0000 (07:57 -0400)]
[SCSI] qla4xxx: Update driver version to 5.02.00-k19

Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] qla4xxx: Properly handle SCSI underrun while processing status IOCBs.
Lalit Chandivade [Tue, 7 Aug 2012 11:57:16 +0000 (07:57 -0400)]
[SCSI] qla4xxx: Properly handle SCSI underrun while processing status IOCBs.

The current code would incorrectly return a DID_OK for a
CHECK CONDITION with Recovered error sense key causing incorrect
completion of a command when there is a dropped frame.

Signed-off-by: Lalit Chandivade <lalit.chandivade@qlogic.com>
Signed-off-by: Tej Parkash <tej.parkash@qlogic.com>
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Reviewed-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] qla4xxx: Fix multiple conn login event issue during session recovery.
Manish Rangankar [Tue, 7 Aug 2012 11:57:15 +0000 (07:57 -0400)]
[SCSI] qla4xxx: Fix multiple conn login event issue during session recovery.

During iscsid session recovery driver sends multiple ISCSI_CONN_STATE_LOGGED_IN
event from qla4xxx_conn_start() and qla4xxx_ddb_change(), which causes iscsid
to crash.

Signed-off-by: Manish Rangankar <manish.rangankar@qlogic.com>
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Reviewed-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] qla4xxx: Fix gcc warning for x86 system
Vikas Chaudhary [Tue, 7 Aug 2012 11:57:14 +0000 (07:57 -0400)]
[SCSI] qla4xxx: Fix gcc warning for x86 system

Fix warning:-
drivers/scsi/qla4xxx/ql4_nx.c:1867:2: warning: format ‘%lx’ expects argument of type ‘long unsigned int’, but argument 5 has type ‘uint32_t’ [-Wformat]

Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Reviewed-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] qla4xxx: Fix memory corruption issue in qla4xxx_ep_connect.
Manish Rangankar [Tue, 7 Aug 2012 11:57:13 +0000 (07:57 -0400)]
[SCSI] qla4xxx: Fix memory corruption issue in qla4xxx_ep_connect.

In qla4xxx_ep_connect(), qla_ep->dst_addr and dst_addr are type
struct sockaddr. We are copying sizeof(struct sockaddr_in6) bytes
from dst_addr to qla_ep->dst_addr which is 12 bytes larger. This
will cause memory corruption. So we change qla_ep->dst_addr to
struct sockaddr_storage which is of 128 byte, large enough to
hold sizeof(struct sockaddr_in6).

Signed-off-by: Manish Rangankar <manish.rangankar@qlogic.com>
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Reviewed-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] mvsas: Fix oops when ata commond timeout.
Jianpeng Ma [Sat, 4 Aug 2012 02:34:14 +0000 (10:34 +0800)]
[SCSI] mvsas: Fix oops when ata commond timeout.

Kernel message follows:

[  511.712011] sd 11:0:0:0: [sdf] command ffff8800a4e81400 timed out
[  511.712022] sas: Enter sas_scsi_recover_host busy: 1 failed: 1
[  511.712024] sas: trying to find task 0xffff8800a4d24c80
[  511.712026] sas: sas_scsi_find_task: aborting task 0xffff8800a4d24c80
[  511.712029] drivers/scsi/mvsas/mv_sas.c 1631:mvs_abort_task()
mvi=ffff8800b5300000 task=ffff8800a4d24c80 slot=ffff8800b5325038
slot_idx=x0
[  511.712035] BUG: unable to handle kernel NULL pointer dereference at
0000000000000058
[  511.712040] IP: [<ffffffff815f8c0c>] _raw_spin_lock_irqsave+0xc/0x30
[  511.712047] PGD 0
[  511.712049] Oops: 0002 [#1] SMP
[  511.712052] Modules linked in: mvsas libsas scsi_transport_sas
raid456 async_pq async_xor xor async_memcpy async_raid6_recov raid6_pq
async_tx [last unloaded: mvsas]
[  511.712062] CPU 3
[  511.712066] Pid: 7322, comm: scsi_eh_11 Not tainted 3.5.0+ #106 To Be
Filled By O.E.M. To Be Filled By O.E.M./To be filled by O.E.M.
[  511.712068] RIP: 0010:[<ffffffff815f8c0c>]  [<ffffffff815f8c0c>]
_raw_spin_lock_irqsave+0xc/0x30
[  511.712073] RSP: 0018:ffff880098d3bcb0  EFLAGS: 00010086
[  511.712074] RAX: 0000000000000286 RBX: 0000000000000058 RCX:
00000000000000c3
[  511.712076] RDX: 0000000000000100 RSI: 0000000000000046 RDI:
0000000000000058
[  511.712078] RBP: ffff880098d3bcb0 R08: 000000000000000a R09:
0000000000000000
[  511.712080] R10: 00000000000004e8 R11: 00000000000004e7 R12:
ffff8800a4d24c80
[  511.712082] R13: 0000000000000050 R14: ffff8800b5325038 R15:
ffff8800a4eafe00
[  511.712084] FS:  0000000000000000(0000) GS:ffff8800bdb80000(0000)
knlGS:0000000000000000
[  511.712086] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[  511.712088] CR2: 0000000000000058 CR3: 00000000a4ce6000 CR4:
00000000000407e0
[  511.712090] DR0: 0000000000000000 DR1: 0000000000000000 DR2:
0000000000000000
[  511.712091] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7:
0000000000000400
[  511.712093] Process scsi_eh_11 (pid: 7322, threadinfo
ffff880098d3a000, task ffff8800a61dde40)
[  511.712095] Stack:
[  511.712096]  ffff880098d3bce0 ffffffff81060683 ffff880000000000
0000000000000000
[  511.712099]  ffff8800a4d24c80 ffff8800b5300000 ffff880098d3bcf0
ffffffffa0076a88
[  511.712102]  ffff880098d3bd50 ffffffffa0079bb5 ffff880000000000
ffff880000000018
[  511.712106] Call Trace:
[  511.712110]  [<ffffffff81060683>] complete+0x23/0x60
[  511.712115]  [<ffffffffa0076a88>] mvs_tmf_timedout+0x18/0x20 [mvsas]
[  511.712119]  [<ffffffffa0079bb5>] mvs_slot_complete+0x765/0x7d0
[mvsas]
[  511.712125]  [<ffffffffa005a17d>] sas_scsi_recover_host+0x55d/0xdb0
[libsas]
[  511.712128]  [<ffffffff8106d600>] ? idle_balance+0xe0/0x130
[  511.712133]  [<ffffffff813b150c>] scsi_error_handler+0xcc/0x470
[  511.712136]  [<ffffffff815f7ad0>] ? __schedule+0x370/0x730
[  511.712139]  [<ffffffff8105f728>] ? __wake_up_common+0x58/0x90
[  511.712142]  [<ffffffff813b1440>] ? scsi_eh_get_sense+0x110/0x110
[  511.712146]  [<ffffffff810571be>] kthread+0x8e/0xa0
[  511.712150]  [<ffffffff816015f4>] kernel_thread_helper+0x4/0x10
[  511.712153]  [<ffffffff81057130>] ? flush_kthread_work+0x120/0x120
[  511.712156]  [<ffffffff816015f0>] ? gs_change+0xb/0xb
[  511.712157] Code: 8a 00 01 00 00 89 d0 f0 66 0f b1 0f 66 39 d0 0f 94
c0 0f b6 c0 5d c3 0f 1f 84 00 00 00 00 00 55 48 89 e5 9c 58 fa ba 00 01
00 00 <f0> 66 0f c1 17 0f b6 ce 38 d1 74 11 0f 1f 84 00 00 00 00 00 f3
[  511.712191] RIP  [<ffffffff815f8c0c>] _raw_spin_lock_irqsave+0xc/0x30
[  511.712194]  RSP <ffff880098d3bcb0>
[  511.712196] CR2: 0000000000000058
[  511.712198] ---[ end trace a781c7b1e65db92c ]---

Signed-off-by: Jianpeng Ma <majianpeng@gmail.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Update lpfc version for 8.3.33 driver release
James Smart [Fri, 3 Aug 2012 16:37:17 +0000 (12:37 -0400)]
[SCSI] lpfc 8.3.33: Update lpfc version for 8.3.33 driver release

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Fixed incomplete list of SLI4 commands with extended timeout...
James Smart [Fri, 3 Aug 2012 16:37:08 +0000 (12:37 -0400)]
[SCSI] lpfc 8.3.33: Fixed incomplete list of SLI4 commands with extended timeout value

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Fixed debugfs queInfo to include queue stats
James Smart [Fri, 3 Aug 2012 16:37:00 +0000 (12:37 -0400)]
[SCSI] lpfc 8.3.33: Fixed debugfs queInfo to include queue stats

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Add lpfc_fcp_look_ahead module parameter
James Smart [Fri, 3 Aug 2012 16:36:52 +0000 (12:36 -0400)]
[SCSI] lpfc 8.3.33: Add lpfc_fcp_look_ahead module parameter

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Make I/O to hw queue distribution algorithm a module parameter
James Smart [Fri, 3 Aug 2012 16:36:42 +0000 (12:36 -0400)]
[SCSI] lpfc 8.3.33: Make I/O to hw queue distribution algorithm a module parameter

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Change Naming convention for SLI4 Interrupt vector
James Smart [Fri, 3 Aug 2012 16:36:33 +0000 (12:36 -0400)]
[SCSI] lpfc 8.3.33: Change Naming convention for SLI4 Interrupt vector

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Allow per-hba interrupt rate tuning
James Smart [Fri, 3 Aug 2012 16:36:24 +0000 (12:36 -0400)]
[SCSI] lpfc 8.3.33: Allow per-hba interrupt rate tuning

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Tie parallel I/O queues into separate MSIX vectors
James Smart [Fri, 3 Aug 2012 16:36:13 +0000 (12:36 -0400)]
[SCSI] lpfc 8.3.33: Tie parallel I/O queues into separate MSIX vectors

Add fcp_io_channel module attribute to control amount of parallel I/O queues

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Add Interrupts per second stats via debugfs
James Smart [Fri, 3 Aug 2012 16:36:03 +0000 (12:36 -0400)]
[SCSI] lpfc 8.3.33: Add Interrupts per second stats via debugfs

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Parallelize SLI-4 Q distribution
James Smart [Fri, 3 Aug 2012 16:35:54 +0000 (12:35 -0400)]
[SCSI] lpfc 8.3.33: Parallelize SLI-4 Q distribution

Commonize SLI-3/4 Ring/Queue framework, to keep SLI-3 compatibility
Parallelize SLI-4 Q distribution - to use multiple posting/completion queues

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Misc changes to optimize critical path
James Smart [Fri, 3 Aug 2012 16:35:44 +0000 (12:35 -0400)]
[SCSI] lpfc 8.3.33: Misc changes to optimize critical path

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Formally separate lpfc_sli_ring SLI-3 and SLI-4 variantions
James Smart [Fri, 3 Aug 2012 16:35:34 +0000 (12:35 -0400)]
[SCSI] lpfc 8.3.33: Formally separate lpfc_sli_ring SLI-3 and SLI-4 variantions

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Convert to no SCSI host lock in queuecommand
James Smart [Fri, 3 Aug 2012 16:35:24 +0000 (12:35 -0400)]
[SCSI] lpfc 8.3.33: Convert to no SCSI host lock in queuecommand

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Add debugfs interface to display SLI queue information
James Smart [Fri, 3 Aug 2012 16:35:13 +0000 (12:35 -0400)]
[SCSI] lpfc 8.3.33: Add debugfs interface to display SLI queue information

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Fix error when remote port switches address
James Smart [Fri, 3 Aug 2012 16:35:03 +0000 (12:35 -0400)]
[SCSI] lpfc 8.3.33: Fix error when remote port switches address

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Fix scsi_eh escalation panic by checking the proper return status
James Smart [Fri, 3 Aug 2012 16:34:54 +0000 (12:34 -0400)]
[SCSI] lpfc 8.3.33: Fix scsi_eh escalation panic by checking the proper return status

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Check data transfer amount on write commands
James Smart [Fri, 3 Aug 2012 16:34:44 +0000 (12:34 -0400)]
[SCSI] lpfc 8.3.33: Check data transfer amount on write commands

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Fix error of not reiniting mbx cmd before reissue
James Smart [Fri, 3 Aug 2012 16:34:29 +0000 (12:34 -0400)]
[SCSI] lpfc 8.3.33: Fix error of not reiniting mbx cmd before reissue

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Fix bug with rrq_pool not being destroyed during driver removal.
James Smart [Fri, 3 Aug 2012 16:34:15 +0000 (12:34 -0400)]
[SCSI] lpfc 8.3.33: Fix bug with rrq_pool not being destroyed during driver removal.

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: Enable attachment to OCe14000 adapters
James Smart [Fri, 3 Aug 2012 16:42:51 +0000 (12:42 -0400)]
[SCSI] lpfc 8.3.33: Enable attachment to OCe14000 adapters

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] lpfc 8.3.33: When doing loopback testing, set the diag valid bit
James Smart [Fri, 3 Aug 2012 16:32:52 +0000 (12:32 -0400)]
[SCSI] lpfc 8.3.33: When doing loopback testing, set the diag valid bit

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] ibmvscsi: Add maintainer for IBM virtual SCSI/FC drivers
Robert Jennings [Tue, 31 Jul 2012 17:34:36 +0000 (12:34 -0500)]
[SCSI] ibmvscsi: Add maintainer for IBM virtual SCSI/FC drivers

Add a MAINTAINERS entry for the IBM Power Virtual SCSI and FC device
drivers.

Signed-off-by: Robert Jennings <rcj@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] ipr: remove an unneeded check
Dan Carpenter [Mon, 30 Jul 2012 08:18:22 +0000 (11:18 +0300)]
[SCSI] ipr: remove an unneeded check

"rc" is always zero here, so there is no need to check.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] ipr: missing unlock before a return
Dan Carpenter [Mon, 30 Jul 2012 08:16:35 +0000 (11:16 +0300)]
[SCSI] ipr: missing unlock before a return

We recently changed the locking in this function, but this return was
missed.  It needs an unlock and the IRQs need to be restored.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] add DID_TARGET_FAILURE and DID_NEXUS_FAILURE to hostbyte_table in constants.c
Moger, Babu [Fri, 27 Jul 2012 17:28:40 +0000 (17:28 +0000)]
[SCSI] add DID_TARGET_FAILURE and DID_NEXUS_FAILURE to hostbyte_table in constants.c

Without this patch, scsi_show_result prints hostbyte as invalid for statuses
that are not defined in hostbyte_table (when scsi logging is enabled).

Signed-off-by: Babu Moger <babu.moger@netapp.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] hpsa: use ioremap_nocache instead of ioremap
Stephen M. Cameron [Thu, 26 Jul 2012 16:34:23 +0000 (11:34 -0500)]
[SCSI] hpsa: use ioremap_nocache instead of ioremap

I think ioremap() ends up being equivalent to ioremap_nocache
by default, but we should signal our intent that these mappings
should be non-cacheable.

Signed-off-by: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] hpsa: fix incorrect abort diagnostic message
Stephen M. Cameron [Thu, 26 Jul 2012 16:34:17 +0000 (11:34 -0500)]
[SCSI] hpsa: fix incorrect abort diagnostic message

In the abort handler, when asked to abort a command which
is not known to the driver, SUCCESS is returned, but the
diagnostic message incorrectly indicates the abort failed.

Signed-off-by: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] hpsa: Use LUN reset instead of target reset
Stephen M. Cameron [Thu, 26 Jul 2012 16:34:10 +0000 (11:34 -0500)]
[SCSI] hpsa: Use LUN reset instead of target reset

It turns out Smart Array logical drives do not support target
reset and when the target reset fails, the logical drive will
be taken off line.  Symptoms look like this:

hpsa 0000:03:00.0: Abort request on C1:B0:T0:L0
hpsa 0000:03:00.0: resetting device 1:0:0:0
hpsa 0000:03:00.0: cp ffff880037c56000 is reported invalid (probably means target device no longer present)
hpsa 0000:03:00.0: resetting device failed.
sd 1:0:0:0: Device offlined - not ready after error recovery
sd 1:0:0:0: rejecting I/O to offline device
EXT3-fs error (device sdb1): read_block_bitmap:

LUN reset is supported though, and is what we should be using.
Target reset is also disruptive in shared SAS situations,
for example, an external MSA1210m which does support target
reset attached to Smart Arrays in multiple hosts -- a target
reset from one host is disruptive to other hosts as all LUNs
on the target will be reset and will abort all outstanding i/os
back to all the attached hosts.  So we should use LUN reset,
not target reset.

Tested this with Smart Array logical drives and with tape drives.
Not sure how this bug survived since 2009, except it must be very
rare for a Smart Array to require more than 30s to complete a request.

Signed-off-by: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Cc: stable@vger.kernel.org
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] megaraid_sas: Version and Changelog update
adam radford [Wed, 18 Jul 2012 01:21:01 +0000 (18:21 -0700)]
[SCSI] megaraid_sas: Version and Changelog update

Signed-off-by: Adam Radford <aradford@gmail.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] megaraid_sas: Add resetwaittime module parameter
adam radford [Wed, 18 Jul 2012 01:20:24 +0000 (18:20 -0700)]
[SCSI] megaraid_sas: Add resetwaittime module parameter

This allows a user to adjust the wait time in seconds after I/O timeout before
resetting the adapter.

Signed-off-by: Adam Radford <aradford@gmail.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] megaraid_sas: Add throttlequeuedepth module parameter
adam radford [Wed, 18 Jul 2012 01:20:03 +0000 (18:20 -0700)]
[SCSI] megaraid_sas: Add throttlequeuedepth module parameter

This allows a user to adjust the queue depth of the adapter when throttled due
to I/O timeout.

Signed-off-by: Adam Radford <aradford@gmail.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] ipr: Driver version 2.5.4
Brian King [Tue, 17 Jul 2012 13:15:33 +0000 (08:15 -0500)]
[SCSI] ipr: Driver version 2.5.4

Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] ipr: Reduce interrupt lock time
Brian King [Tue, 17 Jul 2012 13:14:40 +0000 (08:14 -0500)]
[SCSI] ipr: Reduce interrupt lock time

Reduce the amount of time the host lock is held in the interrupt handler
for improved performance.

[jejb: fix up checkpatch noise]
Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] ipr: Reduce queuecommand lock time
Brian King [Tue, 17 Jul 2012 13:13:52 +0000 (08:13 -0500)]
[SCSI] ipr: Reduce queuecommand lock time

Reduce the amount of time the host lock is held in queuecommand
for improved performance.

[jejb: fix up checkpatch noise]
Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] mpt2sas: Bump driver vesion to 14.100.00.00
sreekanth.reddy@lsi.com [Tue, 17 Jul 2012 10:26:15 +0000 (15:56 +0530)]
[SCSI] mpt2sas: Bump driver vesion to 14.100.00.00

Signed-off-by: Sreekanth Reddy <sreekanth.reddy@lsi.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] mpt2sas: Fix for With post diag reset same set of device gets added, removed...
sreekanth.reddy@lsi.com [Tue, 17 Jul 2012 10:26:13 +0000 (15:56 +0530)]
[SCSI] mpt2sas: Fix for With post diag reset same set of device gets added, removed and then again gets added with new target ids

When device discovery is disabled during driver load time using module
parameter "disable_discovery=1" and when diag reset is issued then from logs,
it is observed that the devices get added, removed and then added with new
target ids.

So, in order to limit this turn-off the code which is deleting and devices
across host reset when the disable_discovery module parameter is turned on.

Signed-off-by: Sreekanth Reddy <sreekanth.reddy@lsi.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] mpt2sas: Fix for staged device discovery functionality of driver not working
sreekanth.reddy@lsi.com [Tue, 17 Jul 2012 10:26:12 +0000 (15:56 +0530)]
[SCSI] mpt2sas: Fix for staged device discovery functionality of driver not working

This patch provides a command line option to disable "Port enable" during
the driver load.

The objective of this command line option is to load the driver and do
all the necessary initialization excluding port enable(i.e. delay
device discovery)

Signed-off-by: Sreekanth Reddy <sreekanth.reddy@lsi.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] mpt2sas : MPI 2.0 Rev V(2.0.14) specification
sreekanth.reddy@lsi.com [Tue, 17 Jul 2012 10:26:10 +0000 (15:56 +0530)]
[SCSI] mpt2sas : MPI 2.0 Rev V(2.0.14) specification

Changeset in MPI 2.0 Rev V(2.0.14) specification

        1) Bumped MPI2_HEADER_VERSION_UNIT.
        2) Added a product specific range to event values.
        3) Added clarification to Direct-Attached SAS PHY Power condition.
        4) Updated timing requirements for performing Hard Reset.

Signed-off-by: Sreekanth Reddy <sreekanth.reddy@lsi.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] mpt2sas: Fix for max_sectors warning message is stating the incorrect range
sreekanth.reddy@lsi.com [Tue, 17 Jul 2012 10:26:09 +0000 (15:56 +0530)]
[SCSI] mpt2sas: Fix for max_sectors warning message is stating the incorrect range

When specifying the command line option "max_sectors" less than 64, then
warning message should provide correct upper boundary value 32767 instead of
8192.

Signed-off-by: Sreekanth Reddy <sreekanth.reddy@lsi.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] mpt2sas: Provide sysfs attribute to report Backup Rail Monitor Status
sreekanth.reddy@lsi.com [Tue, 17 Jul 2012 10:26:07 +0000 (15:56 +0530)]
[SCSI] mpt2sas: Provide sysfs attribute to report Backup Rail Monitor Status

A new sysfs shost attribute called "BMR_status" is implemented to
report Backup Rail Monitor status.

This attribute is located in the path
        /sys/class/scsi_host/host#/BMR_status

when reading this adapter attribute, then driver will output the state
of GPIO[24]. It returns "0" if BMR is healthy and it returns "1" for failure.

if it returns an empty string then it means that there was an error while
obtaining the BMR status. Then check dmesg for what error has occured.

Signed-off-by: Sreekanth Reddy <sreekanth.reddy@lsi.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] mpt2sas: To include more Intel Branding
sreekanth.reddy@lsi.com [Tue, 17 Jul 2012 10:26:05 +0000 (15:56 +0530)]
[SCSI] mpt2sas: To include more Intel Branding

Updating the customer branding string for "SSD 910 Series" controller

Signed-off-by: Sreekanth Reddy <sreekanth.reddy@lsi.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] mpt2sas: 2012 source code copyright
sreekanth.reddy@lsi.com [Tue, 17 Jul 2012 10:26:03 +0000 (15:56 +0530)]
[SCSI] mpt2sas: 2012 source code copyright

The Copyright String in all the drivers sources were changed to 2012

Signed-off-by: Sreekanth Reddy <sreekanth.reddy@lsi.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] libsas, ipr: cleanup ata_host flags initialization via ata_host_init
Dan Williams [Tue, 10 Jul 2012 04:06:08 +0000 (21:06 -0700)]
[SCSI] libsas, ipr: cleanup ata_host flags initialization via ata_host_init

libsas and ipr pass flags to ata_host_init that are meant for the port.

ata_host flags:
ATA_HOST_SIMPLEX = (1 << 0), /* Host is simplex, one DMA channel per host only */
ATA_HOST_STARTED = (1 << 1), /* Host started */
ATA_HOST_PARALLEL_SCAN = (1 << 2), /* Ports on this host can be scanned in parallel */
ATA_HOST_IGNORE_ATA = (1 << 3), /* Ignore ATA devices on this host. */

flags passed by libsas:
ATA_FLAG_SATA = (1 << 1),
ATA_FLAG_PIO_DMA = (1 << 7), /* PIO cmds via DMA */
ATA_FLAG_NCQ = (1 << 10), /* host supports NCQ */

The only one that aliases is ATA_HOST_STARTED which is a 'don't care' in
the libsas and ipr cases since ata_hosts from these sources are not
registered with libata.

Reported-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Acked-by: Brian King <brking@us.ibm.com>
Acked-by: Jeff Garzik <jgarzik@redhat.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] isci: implement suspend/resume support
Artur Wojcik [Fri, 22 Jun 2012 06:41:56 +0000 (23:41 -0700)]
[SCSI] isci: implement suspend/resume support

Provide a "simple-dev-pm-ops" implementation that shuts down the domain
and the device on suspend, and resumes the device and the domain on
resume.  All of the mechanics of restoring domain connectivity are
handled by libsas once isci has notified libsas that all links should be
back up.  libsas is in charge of handling links that did not resume, or
resumed out of order.

Signed-off-by: Artur Wojcik <artur.wojcik@intel.com>
Signed-off-by: Jacek Danecki <jacek.danecki@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] libsas: suspend / resume support
Dan Williams [Fri, 22 Jun 2012 06:41:51 +0000 (23:41 -0700)]
[SCSI] libsas: suspend / resume support

libsas power management routines to suspend and recover the sas domain
based on a model where the lldd is allowed and expected to be
"forgetful".

sas_suspend_ha - disable event processing allowing the lldd to take down
                 links without concern for causing hotplug events.
                 Regardless of whether the lldd actually posts link down
                 messages libsas notifies the lldd that all
                 domain_devices are gone.

sas_prep_resume_ha - on the way back up before the lldd starts link
                     training clean out any spurious events that were
                     generated on the way down, and re-enable event
                     processing

sas_resume_ha - after the lldd has started and decided that all phys
have posted link-up events this routine is called to let
libsas start it's own timeout of any phys that did not
resume.  After the timeout an lldd can cancel the
                phy teardown by posting a link-up event.

Storage for ex_change_count (u16) and phy_change_count (u8) are changed
to int so they can be set to -1 to indicate 'invalidated'.

Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Reviewed-by: Jacek Danecki <jacek.danecki@intel.com>
Tested-by: Maciej Patelczyk <maciej.patelczyk@intel.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] libata: export ata_port suspend/resume infrastructure for sas
Dan Williams [Fri, 22 Jun 2012 06:41:46 +0000 (23:41 -0700)]
[SCSI] libata: export ata_port suspend/resume infrastructure for sas

Reuse ata_port_{suspend|resume}_common for sas.  This path is chosen
over adding coordination between ata-tranport and sas-transport because
libsas wants to revalidate the domain at resume-time at the host level.
It can not validate links have resumed properly until libata has had a
chance to perform its revalidation, and any sane placing of an ata_port
in the sas-transport model would delay it's resumption until after the
host.

Export the common portion of port suspend/resume (bypass pm_runtime),
and allow sas to perform these operations asynchronously (similar to the
libsas async-ata probe implmentation).  Async operation is determined by
having an external, rather than stack based, location for storing the
result of the operation.

Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Reviewed-by: Jacek Danecki <jacek.danecki@intel.com>
Acked-by: Jeff Garzik <jgarzik@redhat.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years ago[SCSI] libata: reset once
Dan Williams [Fri, 22 Jun 2012 06:41:41 +0000 (23:41 -0700)]
[SCSI] libata: reset once

Hotplug testing with libsas currently encounters a 55 second wait for
link recovery to give up.  In the case where the user trusts the
response time of their devices permit the recovery attempts to be
limited to one.

Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Acked-by: Jeff Garzik <jgarzik@redhat.com>
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
11 years agoLinux 3.6-rc2 v3.6-rc2
Linus Torvalds [Thu, 16 Aug 2012 21:51:24 +0000 (14:51 -0700)]
Linux 3.6-rc2

11 years agoautofs4 - fix get_next_positive_subdir()
Ian Kent [Mon, 6 Aug 2012 01:37:47 +0000 (09:37 +0800)]
autofs4 - fix get_next_positive_subdir()

Following a report of a crash during an automount expire I found that
the locking in fs/autofs4/expire.c:get_next_positive_subdir() was wrong.
Not only is the locking wrong but the function is more complex than it
needs to be.

The function is meant to calculate (and dget) the next entry in the list
of directories contained in the root of an autofs mount point (an autofs
indirect mount to be precise). The main problem was that the d_lock of
the owner of the list was not being taken when walking the list, which
lead to list corruption under load. The only other lock that needs to
be taken is against the next dentry candidate so it can be checked for
usability.

Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
11 years agoMerge tag 'vfio-for-v3.6-rc1' of git://github.com/awilliam/linux-vfio
Linus Torvalds [Thu, 16 Aug 2012 18:47:42 +0000 (11:47 -0700)]
Merge tag 'vfio-for-v3.6-rc1' of git://github.com/awilliam/linux-vfio

Pull VFIO fix from Alex Williamson:
 "Just a trivial patch to include vfio.h in the installed headers so we
  can complete userspace integration into QEMU."

* tag 'vfio-for-v3.6-rc1' of git://github.com/awilliam/linux-vfio:
  vfio: Include vfio.h in installed headers

11 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi...
Linus Torvalds [Thu, 16 Aug 2012 18:46:31 +0000 (11:46 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/mszeredi/fuse

Pull fuse updates from Miklos Szeredi.

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse:
  fuse: verify all ioctl retry iov elements
  fuse: add missing INIT flag descriptions
  fuse: add missing INIT flags
  fuse: update attributes on aio_read
  fuse: invalidate inode mapping if mtime changes
  fuse: add FUSE_AUTO_INVAL_DATA init flag

11 years agoMerge tag 'stable/for-linus-3.6-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 16 Aug 2012 18:31:59 +0000 (11:31 -0700)]
Merge tag 'stable/for-linus-3.6-rc1-tag' of git://git./linux/kernel/git/konrad/xen

Pull Xen fix from Konrad Rzeszutek Wilk:
 "Way back in v3.5 we added a mechanism to populate back pages that were
  released (they overlapped with MMIO regions), but neglected to reserve
  the proper amount of virtual space for extend_brk to work properly.

  Coincidentally some other commit aligned the _brk space to larger area
  so I didn't trigger this until it was run on a machine with more than
  2GB of MMIO space."

 * On machines with large MMIO/PCI E820 spaces we fail to boot b/c
   we failed to pre-allocate large enough virtual space for extend_brk.

* tag 'stable/for-linus-3.6-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
  xen/p2m: Reserve 8MB of _brk space for P2M leafs when populating back.

11 years agoMerge tag 'sh-for-linus' of git://github.com/pmundt/linux-sh
Linus Torvalds [Thu, 16 Aug 2012 18:31:29 +0000 (11:31 -0700)]
Merge tag 'sh-for-linus' of git://github.com/pmundt/linux-sh

Pull SuperH fixes from Paul Mundt.

* tag 'sh-for-linus' of git://github.com/pmundt/linux-sh:
  sh: intc: Handle domain association for sparseirq pre-allocated vectors.
  sh: sh7269: Fix LCD pinmux
  sh: dma: fix request_irq usage

11 years agoMAINTAINERS: update address for Dan Williams
Dan Williams [Thu, 16 Aug 2012 02:20:02 +0000 (19:20 -0700)]
MAINTAINERS: update address for Dan Williams

Moved to djbw@fb.com

Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Vinod Koul <vinod.koul@linux.intel.com>
Signed-off-by: Dan Williams <djbw@fb.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
11 years agoscripts/decodecode: Fixup trapping instruction marker
Borislav Petkov [Wed, 15 Aug 2012 11:00:51 +0000 (13:00 +0200)]
scripts/decodecode: Fixup trapping instruction marker

When dumping "Code: " sections from an oops, the trapping instruction
%rip points to can be a string copy

  2b:*  f3 a5                   rep movsl %ds:(%rsi),%es:(%rdi)

and the line contain a bunch of ":".  Current "cut" selects only the and
the second field output looks funnily overlaid this:

  2b:*  f3 a5                   rep movsl %ds     <-- trapping instruction:(%rsi),%es:(%rdi

Fix this by selecting the remaining fields too.

Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: linux-kbuild@vger.kernel.org
Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
11 years agoMerge branch 'fixes' of git://git.infradead.org/users/vkoul/slave-dma
Linus Torvalds [Thu, 16 Aug 2012 18:13:16 +0000 (11:13 -0700)]
Merge branch 'fixes' of git://git.infradead.org/users/vkoul/slave-dma

Pull two slave-dmaengine fixes from Vinod Koul:
 "One fixes the correct use of clock API in imx driver and the other
  enables clock for tegra driver, which is used for other tegra driver
  conversion to dmanegine in -next."

* 'fixes' of git://git.infradead.org/users/vkoul/slave-dma:
  dma: tegra: enable/disable dma clock
  dma: imx-dma: Fix kernel crash due to missing clock conversion

11 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Thu, 16 Aug 2012 18:08:32 +0000 (11:08 -0700)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

Pull more drm fixes from Dave Airlie:
 "Just some intel and nouveau ones this time, intel has more edp panel
  fixes for macbooks and nouveau has a suspend/resume regression fix in
  there."

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  drm/i915: Apply post-sync write for pipe control invalidates
  drm/i915: reorder edp disabling to fix ivb MacBook Air
  drm/nv86/fifo: suspend fix
  drm/nouveau: disable copy engine on NVAF
  nouveau: fixup scanout enable in nvc0_pm
  drm/nouveau/aux: mask off higher bits of auxch index in i2c table entry
  drm/nvd0/disp: mask off high 16 bit of negative cursor x-coordinate
  drm/i915: ensure i2c adapter is all set before adding it
  drm/i915: ignore eDP bpc settings from vbt
  drm/i915: Fix blank panel at reopening lid
  drm/nve0/fifo: add support for the flip completion swmthd

11 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc
Linus Torvalds [Thu, 16 Aug 2012 00:07:01 +0000 (17:07 -0700)]
Merge git://git./linux/kernel/git/davem/sparc

Pull two sparc fixes from David S. Miller.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:
  sparc64: Be less verbose during vmemmap population.
  sparc64: do not clobber personality flags in sys_sparc64_personality()

11 years agoMerge branch 'drm-nouveau-fixes' of git://git.freedesktop.org/git/nouveau/linux-2...
Dave Airlie [Wed, 15 Aug 2012 10:31:22 +0000 (20:31 +1000)]
Merge branch 'drm-nouveau-fixes' of git://git.freedesktop.org/git/nouveau/linux-2.6 into drm-fixes

* 'drm-nouveau-fixes' of git://git.freedesktop.org/git/nouveau/linux-2.6:
  drm/nv86/fifo: suspend fix
  drm/nouveau: disable copy engine on NVAF
  nouveau: fixup scanout enable in nvc0_pm
  drm/nouveau/aux: mask off higher bits of auxch index in i2c table entry
  drm/nvd0/disp: mask off high 16 bit of negative cursor x-coordinate
  drm/nve0/fifo: add support for the flip completion swmthd

11 years agoMerge branch 'drm-intel-fixes' of git://people.freedesktop.org/~danvet/drm-intel...
Dave Airlie [Wed, 15 Aug 2012 10:27:51 +0000 (20:27 +1000)]
Merge branch 'drm-intel-fixes' of git://people.freedesktop.org/~danvet/drm-intel into drm-fixes

Daniel Vetter writes:

"A few important fixers:
- fix various lvds backlight issues, regressed in 3.6 (Takashi Iwai)
- make the retina mbp work (ignore bogus edp bpc value in vbt)
- fix a gmbus regression introduced in (iirc) 3.4 (Jani Nikula)
- fix an edp panel power sequence regression, fixes the new macbook air
- apply the tlb invalidate w/a

Otherwise we still have another gmbus regression (patches are awaiting
tested-bys) and there's something odd going with some rare systems not
entering rc6 often enough (and hence blowing through too much power).  It
seems to be a timing-related issue and can be mitigated by frobbing the
magic tuning parameters. We're still working on that one. Also, we still
have some fallout from the hw context support, but you can only hit that
with mesa master."

* 'drm-intel-fixes' of git://people.freedesktop.org/~danvet/drm-intel:
  drm/i915: Apply post-sync write for pipe control invalidates
  drm/i915: reorder edp disabling to fix ivb MacBook Air
  drm/i915: ensure i2c adapter is all set before adding it
  drm/i915: ignore eDP bpc settings from vbt
  drm/i915: Fix blank panel at reopening lid

11 years agosparc64: Be less verbose during vmemmap population.
David S. Miller [Wed, 15 Aug 2012 07:37:29 +0000 (00:37 -0700)]
sparc64: Be less verbose during vmemmap population.

On a 2-node machine with 256GB of ram we get 512 lines of
console output, which is just too much.

This mimicks Yinghai Lu's x86 commit c2b91e2eec9678dbda274e906cc32ea8f711da3b
(x86_64/mm: check and print vmemmap allocation continuous) except that
we aren't ever going to get contiguous block pointers in between calls
so just print when the virtual address or node changes.

This decreases the output by an order of 16.

Also demote this to KERN_DEBUG.

Signed-off-by: David S. Miller <davem@davemloft.net>
11 years agodrm/i915: Apply post-sync write for pipe control invalidates
Chris Wilson [Fri, 10 Aug 2012 09:18:10 +0000 (10:18 +0100)]
drm/i915: Apply post-sync write for pipe control invalidates

When invalidating the TLBs it is documentated as requiring a post-sync
write. Failure to do so seems to result in a GPU hang.

Exposure to this hang on IVB seems to be a result of removing the extra
stalls required for SNB pipecontrol workarounds:

commit 6c6cf5aa9c583478b19e23149feaa92d01fb8c2d
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Fri Jul 20 18:02:28 2012 +0100

    drm/i915: Only apply the SNB pipe control w/a to gen6

Note: Manually switch the pipe_control cmd to 4 dwords to avoid a
(silent) functional conflict with -next. This way will get a loud (but
conflict with next (since the scratch_addr has been deleted there).

Reported-and-tested-by: yex.tian@intel.com
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=53322
Acked-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: added note about merge conflict with -next.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
11 years agodrm/i915: reorder edp disabling to fix ivb MacBook Air
Daniel Vetter [Sun, 12 Aug 2012 20:17:14 +0000 (22:17 +0200)]
drm/i915: reorder edp disabling to fix ivb MacBook Air

eDP is tons of fun. It turns out that at least the new MacBook Air 5,1
model absolutely doesn't like the new force vdd dance we've introduced
in

commit 6cb49835da0426f69a2931bc2a0a8156344b0e41
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Sun May 20 17:14:50 2012 +0200

    drm/i915: enable vdd when switching off the eDP panel

But that patch also tried to fix some neat edp sequence issue with the
force_vdd timings. Closer inspection reveals that we've raised
force_vdd only to do the aux channel communication dp_sink_dpms. If we
move the edp_panel_off below that, we don't need any force_vdd for the
disable sequence, which makes the Air happy.

Unfortunately the reporter of the original bug that the above commit
fixed is travelling, so we can't test whether this regresses things.
But my theory is that since we don't check for any power-off ->
force_vdd-on delays in edp_panel_vdd_on, this was the actual
root-cause of this failure. With that force_vdd dance completely
eliminated, I'm hopeful the original bug stays fixed, too.

For reference the old bug, which hopefully doesn't get broken by this:

https://bugzilla.kernel.org/show_bug.cgi?id=43163

In any case, regression fixers win over plain bugfixes, so this needs
to go in asap.

v2: The crucial pieces seems to be to clear the force_vdd flag
uncoditionally, too, in edp_panel_off. Looks like this is left behind
by the firmware somehow.

v3: The Apple firmware seems to switch off the panel on it's own, hence
we still need to keep force_vdd on, but properly clear it when switching
the panel off.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=45671
Tested-by: Roberto Romer <sildurin@gmail.com>
Tested-by: Daniel Wagner <wagi@monom.org>
Tested-by: Keith Packard <keithp@keithp.com>
Cc: stable@vger.kernel.org
Cc: Keith Packard <keithp@keithp.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
11 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux
Linus Torvalds [Tue, 14 Aug 2012 04:58:59 +0000 (07:58 +0300)]
Merge branch 'for-linus' of git://git./linux/kernel/git/s390/linux

Pull s390 patches from Martin Schwidefsky:
 "Included are bug fixes and a patch to enable system call filtering
  with BPF."

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
  s390/compat: fix mmap compat system calls
  s390/compat: fix compat wrappers for process_vm system calls
  s390: do not clobber personality flags in sys_32_personality()
  s390/seccomp: add support for system call filtering using BPF
  s390/sclp_sdias: Add missing break and "fall through"
  s390/mm: remove MAX_PHYSADDR_BITS define

11 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Tue, 14 Aug 2012 04:52:41 +0000 (07:52 +0300)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

Pull drm fixes from Dave Airlie:
 "Radeon and intel fixes mostly, one fix to the mgag200 driver to not
  hang on certain server variants."

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux: (32 commits)
  drm/radeon: fix typo in function header comment
  drm/radeon/kms: implement timestamp userspace query (v2)
  drm/radeon/kms: add MSAA texture support for r600-evergreen
  drm/radeon/kms: reorder code in r600_check_texture_resource
  drm/radeon: fence virtual address and free it once idle v4
  drm/radeon: fix some missing parens in asic macros
  drm/radeon: add some new SI pci ids
  drm/radeon: fix ordering in pll picking on dce4+
  drm/radeon: do not reenable crtc after moving vram start address
  drm/radeon: fix bank tiling parameters on cayman
  drm/radeon: fix bank tiling parameters on evergreen
  drm/radeon: fix bank tiling parameters on SI
  drm/radeon: properly handle crtc powergating
  drm/radeon: properly handle SS overrides on TN (v2)
  drm/radeon/dce4+: set a more reasonable cursor watermark
  drm/radeon: fix handling for ddc type 5 on combios
  drm/mgag200: fix G200ER pll picking algorithm
  drm/edid: Fix potential memory leak in edid_load()
  drm/udl: Use ERR_CAST inlined function instead of ERR_PTR(PTR_ERR(.. [1]
  drm/radeon/kms: allow "invalid" DB formats as a means to disable DB
  ...

11 years agoGPIO: gpio-pxa: fix building without CONFIG_OF
Arnd Bergmann [Mon, 13 Aug 2012 14:36:10 +0000 (14:36 +0000)]
GPIO: gpio-pxa: fix building without CONFIG_OF

Commit 7212157267 ("GPIO: gpio-pxa: fix devicetree functions") added an
"xlate" function pointer to the irq_domain_ops, but this function is nor
declared or defined anywhere when CONFIG_OF is disabled, causing the
build error:

  drivers/gpio/gpio-pxa.c:532:11: error: 'irq_domain_xlate_twocell' undeclared here (not in a function)

Extending the DT-only code section to cover the irq_domain_ops and the
pxa_gpio_dt_ids solves this problem and makes it clearer which code is
actually used without DT.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
11 years agodrm/nv86/fifo: suspend fix
Maxim Levitsky [Mon, 13 Aug 2012 23:22:07 +0000 (02:22 +0300)]
drm/nv86/fifo: suspend fix

This fix is a backport from the reworked nouveau driver.  It masks off the
engines we're not expecting to use before attempting a channel kickoff.

Signed-off-by: Maxim Levitsky <maximlevitsky@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
11 years agodrm/nouveau: disable copy engine on NVAF
Henrik Rydberg [Sat, 4 Aug 2012 06:00:45 +0000 (08:00 +0200)]
drm/nouveau: disable copy engine on NVAF

The copy engine exhibits random memory corruption in at least one case, the
GeForce 320M (nv50, 0xaf) in the MacBookAir3,1.

This patch omits creating the engine for the specific chipset, falling back
to M2MF, which kills the symptoms.

Signed-off-by: Henrik Rydberg <rydberg@euromail.se>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
11 years agonouveau: fixup scanout enable in nvc0_pm
Maarten Lankhorst [Mon, 30 Jul 2012 10:03:30 +0000 (12:03 +0200)]
nouveau: fixup scanout enable in nvc0_pm

Fixes screen being black after changing performance level.

Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
Cc: stable@vger.kernel.org [3.5+]
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
11 years agodrm/nouveau/aux: mask off higher bits of auxch index in i2c table entry
Ben Skeggs [Fri, 27 Jul 2012 04:18:03 +0000 (14:18 +1000)]
drm/nouveau/aux: mask off higher bits of auxch index in i2c table entry

At least partially fixes DP output detection on W530.  Not sure if more
issues remain, or if my adaptor is just behaving weirdly (it does that
sometimes).

In any case, this patch is necessary.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
11 years agodrm/nvd0/disp: mask off high 16 bit of negative cursor x-coordinate
Christoph Bumiller [Thu, 26 Jul 2012 18:53:19 +0000 (20:53 +0200)]
drm/nvd0/disp: mask off high 16 bit of negative cursor x-coordinate

Signed-off-by: Christoph Bumiller <e0425955@student.tuwien.ac.at>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Cc: stable@vger.kernel.org
11 years agoMerge branch 'drm-fixes-3.6' of git://people.freedesktop.org/~agd5f/linux into drm...
Dave Airlie [Mon, 13 Aug 2012 23:25:01 +0000 (09:25 +1000)]
Merge branch 'drm-fixes-3.6' of git://people.freedesktop.org/~agd5f/linux into drm-fixes

Alex Deucher writes:
This is the current set of radeon fixes for 3.6.  Nothing too major.

Highlights:
- various display fixes
- some SI fixes
- new SI pci ids
- major VM fix
- CS checker support for MSAA

I've tested on a number of cards across generations and noticed no problems.

* 'drm-fixes-3.6' of git://people.freedesktop.org/~agd5f/linux:
  drm/radeon: fix typo in function header comment
  drm/radeon/kms: implement timestamp userspace query (v2)
  drm/radeon/kms: add MSAA texture support for r600-evergreen
  drm/radeon/kms: reorder code in r600_check_texture_resource
  drm/radeon: fence virtual address and free it once idle v4
  drm/radeon: fix some missing parens in asic macros
  drm/radeon: add some new SI pci ids
  drm/radeon: fix ordering in pll picking on dce4+
  drm/radeon: do not reenable crtc after moving vram start address
  drm/radeon: fix bank tiling parameters on cayman
  drm/radeon: fix bank tiling parameters on evergreen
  drm/radeon: fix bank tiling parameters on SI
  drm/radeon: properly handle crtc powergating
  drm/radeon: properly handle SS overrides on TN (v2)
  drm/radeon/dce4+: set a more reasonable cursor watermark
  drm/radeon: fix handling for ddc type 5 on combios

11 years agodrm/i915: ensure i2c adapter is all set before adding it
Jani Nikula [Mon, 13 Aug 2012 14:33:02 +0000 (17:33 +0300)]
drm/i915: ensure i2c adapter is all set before adding it

i2c_add_adapter() may do i2c transfers on the bus to detect supported
devices. Therefore the adapter needs to be all set before adding it. This
was not the case for the bit-banging fallback, resulting in an oops if the
device detection GMBUS transfers timed out. Fix the issue by calling
i2c_add_adapter() only after intel_gpio_setup().

LKML-Reference: <5021F00B.7000503@ionic.de>
Tested-by: Mihai Moldovan <ionic@ionic.de>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
11 years agodrm/radeon: fix typo in function header comment
Dmitrii Cherkasov [Mon, 13 Aug 2012 14:53:29 +0000 (10:53 -0400)]
drm/radeon: fix typo in function header comment

Signed-off-by: Dmitrii Cherkasov <DCherkasov@luxsoft.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
11 years agodrm/radeon/kms: implement timestamp userspace query (v2)
Marek Olšák [Thu, 9 Aug 2012 14:34:17 +0000 (16:34 +0200)]
drm/radeon/kms: implement timestamp userspace query (v2)

Returns a snapshot of the GPU clock counter.  Needed
for certain OpenGL extensions.

v2: agd5f
- address Jerome's comments
- add function documentation

Signed-off-by: Marek Olšák <maraeo@gmail.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
11 years agodrm/radeon/kms: add MSAA texture support for r600-evergreen
Marek Olšák [Thu, 9 Aug 2012 14:34:16 +0000 (16:34 +0200)]
drm/radeon/kms: add MSAA texture support for r600-evergreen

Most of the checking seems to be in place already. As you can see,
log2(number of samples) resides in LAST_LEVEL.

This is required for MSAA support (namely for depth-stencil resolve and
blitting between MSAA resources).

Signed-off-by: Marek Olšák <maraeo@gmail.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
11 years agodrm/radeon/kms: reorder code in r600_check_texture_resource
Marek Olšák [Thu, 9 Aug 2012 14:34:15 +0000 (16:34 +0200)]
drm/radeon/kms: reorder code in r600_check_texture_resource

Signed-off-by: Marek Olšák <maraeo@gmail.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
11 years agodrm/radeon: fence virtual address and free it once idle v4
Jerome Glisse [Mon, 6 Aug 2012 16:32:21 +0000 (12:32 -0400)]
drm/radeon: fence virtual address and free it once idle v4

Virtual address need to be fenced to know when we can safely remove it.
This patch also properly clear the pagetable. Previously it was
serouisly broken.

Kernel 3.5/3.4 need a similar patch but adapted for difference in mutex locking.

v2: For to update pagetable when unbinding bo (don't bailout if
    bo_va->valid is true).
v3: Add kernel 3.5/3.4 comment.
v4: Fix compilation warnings.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
11 years agodrm/radeon: fix some missing parens in asic macros
Alex Deucher [Fri, 3 Aug 2012 15:50:54 +0000 (11:50 -0400)]
drm/radeon: fix some missing parens in asic macros

Better safe than sorry.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
11 years agodrm/radeon: add some new SI pci ids
Alex Deucher [Mon, 6 Aug 2012 14:03:59 +0000 (10:03 -0400)]
drm/radeon: add some new SI pci ids

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
11 years agodrm/radeon: fix ordering in pll picking on dce4+
Alex Deucher [Mon, 6 Aug 2012 21:06:03 +0000 (17:06 -0400)]
drm/radeon: fix ordering in pll picking on dce4+

No functional change, but re-order the cases so they
evaluate properly due to the way the DCE macros work.

Noticed by kallisti5 on IRC.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
11 years agodrm/radeon: do not reenable crtc after moving vram start address
Jerome Glisse [Fri, 27 Jul 2012 20:32:24 +0000 (16:32 -0400)]
drm/radeon: do not reenable crtc after moving vram start address

It seems we can not update the crtc scanout address. After disabling
crtc, update to base address do not take effect after crtc being
reenable leading to at least frame being scanout from the old crtc
base address. Disabling crtc display request lead to same behavior.

So after changing the vram address if we don't keep crtc disabled
we will have the GPU trying to read some random system memory address
with some iommu this will broke the crtc engine and will lead to
broken display and iommu error message.

So to avoid this, disable crtc. For flicker less boot we will need
to avoid moving the vram start address.

This patch should also fix :

https://bugs.freedesktop.org/show_bug.cgi?id=42373

Cc: <stable@vger.kernel.org>
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
11 years agodrm/radeon: fix bank tiling parameters on cayman
Alex Deucher [Tue, 31 Jul 2012 15:05:11 +0000 (11:05 -0400)]
drm/radeon: fix bank tiling parameters on cayman

Handle the 16 bank case.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
11 years agodrm/radeon: fix bank tiling parameters on evergreen
Alex Deucher [Tue, 31 Jul 2012 15:01:10 +0000 (11:01 -0400)]
drm/radeon: fix bank tiling parameters on evergreen

Handle the 16 bank case.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
11 years agodrm/radeon: fix bank tiling parameters on SI
Christian König [Tue, 31 Jul 2012 11:48:51 +0000 (13:48 +0200)]
drm/radeon: fix bank tiling parameters on SI

The sixteen bank case wasn't handled here, leading to GPU
crashes because of userspace miscalculation.

Signed-off-by: Christian König <deathsimple@vodafone.de>
Cc: stable@vger.kernel.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
11 years agodrm/radeon: properly handle crtc powergating
Alex Deucher [Thu, 26 Jul 2012 17:38:52 +0000 (13:38 -0400)]
drm/radeon: properly handle crtc powergating

Need to make sure the crtc is gated on before modesetting.
Explicitly gate the crtc on in prepare() and set a flag
so that the dpms functions don't gate it off during
mode set.

Noticed by sylware on IRC.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
11 years agodrm/radeon: properly handle SS overrides on TN (v2)
Alex Deucher [Wed, 25 Jul 2012 16:32:59 +0000 (12:32 -0400)]
drm/radeon: properly handle SS overrides on TN (v2)

The IntegratedSystemInfo table changed versions
on TN.  Update the SS override lookup to handle it.

v2: fix copy-paste typo.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
11 years agodrm/radeon/dce4+: set a more reasonable cursor watermark
Alex Deucher [Tue, 10 Jul 2012 19:20:24 +0000 (15:20 -0400)]
drm/radeon/dce4+: set a more reasonable cursor watermark

Set a more reasonable default cursor watermark. The
recommended default value is 4.  This should reduce
urgency requests to the MC form the display hw.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
11 years agodrm/radeon: fix handling for ddc type 5 on combios
Alex Deucher [Fri, 20 Jul 2012 00:11:44 +0000 (20:11 -0400)]
drm/radeon: fix handling for ddc type 5 on combios

When ddc type is 5, need to look up the i2c channel
in the i2c table.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>