Multipath is best at handling transport errors. If it gets a device
error then there is not much the multipath layer can do. It will just
access the same device but from a different path.
This patch breaks up failfast into device, transport and driver errors.
The multipath layers (md and dm mutlipath) only ask the lower levels to
fast fail transport errors. The user of failfast, read ahead, will ask
to fast fail on all errors.
Note that blk_noretry_request will return true if any failfast bit
is set. This allows drivers that do not support the multipath failfast
bits to continue to fail on any failfast error like before. Drivers
like scsi that are able to fail fast specific errors can check
for the specific fail fast type. In the next patch I will convert
scsi.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Cc: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
EMC Symmetrix Subsystem Control I/O through CKD dasd requires a
specific parameter list sent to the array via a Perform Subsystem
Function CCW. The Symmetrix response is retrieved from the array
via a Read Subsystem Data CCW.
Signed-off-by: Nigel Hislop <hislop_nigel@emc.com>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The xpram driver uses a single block device queue for all of its
devices so far. With recent kernels removing xpram module fails to
clean up all sysfs files. The next time the xpram module is loaded
you'll get warnings:
WARNING: at fs/sysfs/dir.c:463 sysfs_add_one+0x5e/0x64()
sysfs: duplicate filename '35:0' can not be created
Modules linked in: xpram(+) [last unloaded: xpram]
Followed by the usual WARN_ON output, followed by an error message
from kobject_add_internal, followed by a badness in genhd. Allocating
a block queue per device fixes this.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
In the unsolicited interupt handler fake IRBs from CIO have to be
ignored because there is nothing to do.
The function dump_sense should not be called if there is no sense
data available.
Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The DCSS block device driver is modified to add >2G DCSSs support and
allow a DCSS block device to map to a set of contiguous DCSSs. The
extmem code is also modified to use new Diagnose x'64' subcodes for
>2G DCSSs.
Signed-off-by: Hongjie Yang <hongjie@us.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Convert most s390 users setting bus_id to dev_set_name().
css and ccw busses are deferred since they need some special
treatment.
Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
* Implement disk_devt() and part_devt() and use them to directly
access devt instead of computing it from ->major and ->first_minor.
Note that all references to ->major and ->first_minor outside of
block layer is used to determine devt of the disk (the part0) and as
->major and ->first_minor will continue to represent devt for the
disk, converting these users aren't strictly necessary. However,
convert them for consistency.
* Implement disk_max_parts() to avoid directly deferencing
genhd->minors.
* Update bdget_disk() such that it doesn't assume consecutive minor
space.
* Move devt computation from register_disk() to add_disk() and make it
the only one (all other usages use the initially determined value).
These changes clean up the code and will help disk->part dereference
fix and extended block device numbers.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Fix this compile bug:
CC drivers/s390/block/dcssblk.o
drivers/s390/block/dcssblk.c: In function 'dcssblk_add_store':
drivers/s390/block/dcssblk.c:387: error: implicit declaration of function 'dcssblk_get_segment_by_name'
drivers/s390/block/dcssblk.c:389: error: label 'release_gd' used but not defined
make[1]: *** [drivers/s390/block/dcssblk.o] Error 1
make: *** [drivers/s390/block/] Error 2
Cc: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Concurrently adding the same segment may lead to duplicate device_register()
calls, which will trigger an error in kobject code:
"... don't try to register things with the same name in the same directory".
This patch adds a check to avoid this.
Signed-off-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Calling a ccw driver's notify function without the ccw device lock
held opens up a race window between discovery and handling of a change
in the device operational state. As a result, the device driver may
encounter unexpected device malfunction, leading to out-of-retry
situations or similar.
Remove race by extending the ccw device lock from state change
discovery to the calling of the notify function.
Signed-off-by: Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The Perform Subsystem Function/Prepare for Read Subsystem Data
command requires 12 bytes of parameter data, but the respective data
structure dasd_psf_prssd_data has a length of 16 bytes.
Current storage servers ignore the obsolete bytes, but older models
fail to execute the command and report an incorrect length error.
This causes the device initilization for these devices to fail.
To fix this problem we need to correct the dasd_psf_prssd_data
structure and shorten it to the correct length.
Reported-by: Ivan Warren <ivan@vmfacility.fr>
Reviewed-by: Ivan Warren <ivan@vmfacility.fr>
Tested-by: Ivan Warren <ivan@vmfacility.fr>
CC: stable <stable@kernel.org>
Signed-off-by: Stefan Weinhuber <wein@de.ibm.com>
In case of error, functions dasd_kmalloc_request and idal_buffer_alloc
return an ERR pointer, but never return the NULL pointer. So after a
call to one of these functions, a NULL test should be replaced by an
IS_ERR test.
A simplified version of the semantic patch that makes this change is
as follows:
(http://www.emn.fr/x-info/coccinelle/)
// <smpl>
@correct_null_test@
expression x,E;
statement S1, S2;
@@
x =
(
dasd_kmalloc_request(...)
|
idal_buffer_alloc(...)
)
<... when != x = E
if (
(
- x@p2 != NULL
+ ! IS_ERR ( x )
|
- x@p2 == NULL
+ IS_ERR( x )
)
)
S1
else S2
...>
? x = E;
// </smpl>
Signed-off-by: Julien Brunel <brunel@diku.dk>
Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
When z/VM provides two virtual devices (minidisks) that reside on the
same real device, both will receive the configuration data from the
real device and thus get the same uid. To fix this problem, z/VM
provides an additional configuration data record that allows to
distinguish between minidisks.
z/VM APAR VM64273 needs be installed so this fix has an effect.
Signed-off-by: Stefan Weinhuber <wein@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
return value -ENOTSUPP is not valid in userspace context, use
-EOPNOTSUPP instead
Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Add missing schedule_bh and check that there is 32 bit sense data.
Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Extend the scsw data structure to the format required by fcx. Also
provide helper functions for easier access to fields which are present
in both the traditional as well as the modified format.
Signed-off-by: Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Use a generic wait_queue to prevent the wait_queue in dasd_sleep_on_
functions from being referenced by callback_data while it does not
exist any more.
Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
When the dasd_int_handler is called with an error code instead of
an irb, the associated request should be restarted. This handling
was missing from the -ETIMEDOUT case. In fact it should be done in
any case.
Signed-off-by: Stefan Weinhuber <wein@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
We should use 'const char *' in the busid functions since the
strings are not modified anyway.
Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Use proc_create() to make sure that ->proc_fops be setup before gluing PDE to
main tree.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Remove proc_root export. Creation and removal works well if parent PDE is
supplied as NULL -- it worked always that way.
So, one useless export removed and consistency added, some drivers created
PDEs with &proc_root as parent but removed them as NULL and so on.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alter the block device ->direct_access() API to work with the new
get_xip_mem() API (that requires both kaddr and pfn are returned).
Some architectures will not do the right thing in their virt_to_page() for use
by XIP (to translate from the kernel virtual address returned by
direct_access(), to a user mappable pfn in XIP's page fault handler.
However, we can't switch it to just return the pfn and not the kaddr, because
we have no good way to get a kva from a pfn, and XIP requires the kva for its
read(2) and write(2) handlers. So we have to return both.
Signed-off-by: Jared Hulbert <jaredeh@gmail.com>
Signed-off-by: Nick Piggin <npiggin@suse.de>
Cc: Carsten Otte <cotte@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: linux-mm@kvack.org
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Move the function that prints the segment warning messages found in the
monreader driver and the dcssblk driver to the extmem base code.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Most noteable part of this commit is the new local header file entry.h
which contains all the function declarations of functions that get only
called from asm code or are arch internal. That way we can avoid extern
declarations in C files.
This is more or less the same that was done for sparc64.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
__FUNCTION__ is gcc-specific, use __func__
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Now the system reports system information messages (SIM) to the user.
The System Reference Code (SRC) which is reported to the user gives
the abbility to lookup the reason of the SIM online in the
documentation of the storage server.
Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
allocating dasd_fba_private without GFP_DMA results in IO error
during read device characteristics of a FBA disk
Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
I compiled the kernel without deadline, and the dasd code exits the old
scheduler (CFQ), fails to load the new one (deadline), and then things just
hang - with one of these (sorry about the weird chars - I copy & pasted it
from a 3270 console):
dasd(eckd): 0.0.0151: 3390/0A(CU:3990/01) Cyl:3338 Head:15 Sec:224
------------ cut here ------------
Badness at kernel/mutex.c:134
Modules linked in: dasd_eckd_mod dasd_mod
CPU: 0 Not tainted 2.6.25-rc3 #9
Process exe (pid: 538, task: 000000000d172000, ksp: 000000000d21ef88)
Krnl PSW : 0404000180000000 000000000022fb5c (mutex_lock_nested+0x2a4/0x2cc)
R:0 T:1 IO:0 EX:0 Key:0 M:1 W:0 P:0 AS:0 CC:0 PM:0 EA:3
Krnl GPRS: 0000000000024218 000000000076fc78 0000000000000000 000000000000000f
000000000022f92e 0000000000449898 000000000f921c00 000003e000162590
00000000001539c4 000000000d172000 070000007fffffff 000000000d21f400
000000000f8f2560 00000000002413f8 000000000022fb44 000000000d21f400
Krnl Code: 000000000022fb50: bf2f1000 icm %r2,15,0(%r1)
000000000022fb54: a774fef6 brc 7,22f940
000000000022fb58: a7f40001 brc 15,22fb5a
>000000000022fb5c: a7f4fef2 brc 15,22f940
000000000022fb60: c0e5fffa112a brasl %r14,171db4
000000000022fb66: 1222 ltr %r2,%r2
000000000022fb68: a784fedb brc 8,22f91e
000000000022fb6c: c010002a0086 larl %r1,76fc78
Call Trace:
(<000000000022f92e> mutex_lock_nested+0x76/0x2cc)
<00000000001539c4> elevator_exit+0x38/0x80
<0000000000156ffe> blk_cleanup_queue+0x62/0x7c
<000003e0001d5414> dasd_change_state+0xe0/0x8ec
<000003e0001d5cae> dasd_set_target_state+0x8e/0x9c
<000003e0001d5f74> dasd_generic_set_online+0x160/0x284
<000003e00011e83a> dasd_eckd_set_online+0x2e/0x40
<0000000000199bf4> ccw_device_set_online+0x170/0x2c0
<0000000000199d9e> online_store_recog_and_online+0x5a/0x14c
<000000000019a08a> online_store+0xbe/0x2ec
<000000000018456c> dev_attr_store+0x38/0x58
<000000000010efbc> sysfs_write_file+0x130/0x190
<00000000000af582> vfs_write+0xb2/0x160
<00000000000afc7c> sys_write+0x54/0x9c
<0000000000025e16> sys32_write+0x2e/0x50
<0000000000024218> sysc_noemu+0x10/0x16
<0000000077e82bd2> 0x77e82bd2
Set elevator pointer to NULL in order to avoid double elevator_exit
calls when elevator_init call for deadline iosched fails.
Also make sure the dasd device driver depends on IOSCHED_DEADLINE so
the default IO scheduler of the dasd driver is present.
Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Using the /proc/dasd/devices interface leaves the reference counter
of alias devices in an inconsistent state. A process that tries to set
such a device offline afterwards will hang.
The dasd_devices_show function returns immediately for alias devices
and this code path was missing a dasd_put_device call.
Signed-off-by: Stefan Weinhuber <wein@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
When a request fails that was started on an alias device then the
first recovery step is to retry it on the base device. If the
recovery request fails again with the same symptoms, the next step
should not be a simple retry, but should be a proper recovery based
on sense data, etc. To do so, the dasd recovery functions need to
recognize the alias recovery step in the erp chain by comparing
the start devices.
Signed-off-by: Stefan Weinhuber <wein@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
After setting the status of the cqr and releasing the lock for the
block cqr queue, we call the cqr callback function, which will usually
just trigger the dasd_block_tasklet. But when the tasklet is already
running the cqr might be processed before we invoke the callback
function. In rare cases the callback pointer may already be invalid
by the time we want to call it, which will result in a panic.
Solution: Call the callback function first and then release the lock.
Signed-off-by: Stefan Weinhuber <wein@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
In case a dcss segment cannot be loaded blk_cleanup_queue
will be called before blk_queue_make_request, leaving the
struct work unplug_work of the request queue uninitialized
before it is used.
That leads also to the lockdep message below.
To avoid that call blk_queue_make_request right after the
request_queue has been allocated.
This makes sure that the struct work is always initialized
before it is used.
INFO: trying to register non-static key.
the code is fine but needs lockdep annotation.
turning off the locking correctness validator.
CPU: 2 Not tainted 2.6.24 #6
Process swapper (pid: 1, task: 000000000f854038, ksp: 000000000f85f980)
040000000f85f860 000000000f85f880 0000000000000002 0000000000000000
000000000f85f920 000000000f85f898 000000000f85f898 000000000001622e
0000000000000000 000000000f85f980 0000000000000000 0000000000000000
000000000f85f880 000000000000000c 000000000f85f880 000000000f85f8f0
0000000000342908 000000000001622e 000000000f85f880 000000000f85f8d0
Call Trace:
([<000000000001619e>] show_trace+0xda/0x104)
[<0000000000016288>] show_stack+0xc0/0xf8
[<00000000000163d0>] dump_stack+0xb0/0xc0
[<000000000006e4ea>] __lock_acquire+0x47e/0x1160
[<000000000006f27c>] lock_acquire+0xb0/0xd8
[<000000000005a522>] __cancel_work_timer+0x9e/0x240
[<000000000005a72e>] cancel_work_sync+0x2a/0x3c
[<0000000000165c46>] kblockd_flush_work+0x26/0x34
[<0000000000169034>] blk_sync_queue+0x38/0x48
[<0000000000169080>] blk_release_queue+0x3c/0xa8
[<000000000017bce8>] kobject_cleanup+0x58/0xac
[<000000000017bd66>] kobject_release+0x2a/0x38
[<000000000017d28e>] kref_put+0x6e/0x94
[<000000000017bc80>] kobject_put+0x38/0x48
[<00000000001653be>] blk_put_queue+0x2a/0x38
[<0000000000168fee>] blk_cleanup_queue+0x82/0x90
[<0000000000213e7e>] dcssblk_add_store+0x34e/0x700
[<00000000005243b8>] dcssblk_init+0x1a0/0x308
[<000000000050a3c2>] kernel_init+0x1b2/0x3a4
[<000000000001ac82>] kernel_thread_starter+0x6/0xc
[<000000000001ac7c>] kernel_thread_starter+0x0/0xc
INFO: lockdep is turned off.
Cc: Gerald Schaefer <geraldsc@de.ibm.com>
Cc: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
When an alias device is set offline while it is in use this may
result in a panic in the cleanup part of the dasd_block_tasklet.
The problem here is that there may exist some ccw requests that were
originally created for the alias device and transferred to the base
device when the alias was set offline. When these request are
cleaned up later, the discipline pointer in the alias device may not
be valid anymore. To fix this use the base device discipline to find
the cleanup function.
Signed-off-by: Stefan Weinhuber <wein@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Adding interface control check (ifcc) handling in error recovery.
First retry up to 255 times and if all retries fail try an alternate
path if possible.
Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
This patch converts s390 to use blk_end_request interfaces.
Related 'uptodate' arguments are converted to 'error'.
As a result, the interfaces of internal functions below are changed:
o dasd_end_request
o tapeblock_end_request
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: linux390@de.ibm.com
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Add time to the 'expires' value to avoid a loop caused by the cqr
termination function
Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Parallel access volumes (PAV) is a storage server feature, that allows
to start multiple channel programs on the same DASD in parallel. It
defines alias devices which can be used as alternative paths to the
same disk. With the old base PAV support we only needed rudimentary
functionality in the DASD device driver. As the mapping between base
and alias devices was static, we just had to export an identifier
(uid) and could leave the combining of devices to external layers
like a device mapper multipath.
Now hyper PAV removes the requirement to dedicate alias devices to
specific base devices. Instead each alias devices can be combined with
multiple base device on a per request basis. This requires full
support by the DASD device driver as now each channel program itself
has to identify the target base device.
The changes to the dasd device driver and the ECKD discipline are:
- Separate subchannel device representation (dasd_device) from block
device representation (dasd_block). Only base devices are block
devices.
- Gather information about base and alias devices and possible
combinations.
- For each request decide which dasd_device should be used (base or
alias) and build specific channel program.
- Support summary unit checks, which allow the storage server to
upgrade / downgrade between base and hyper PAV at runtime (support
is mandatory).
Signed-off-by: Stefan Weinhuber <wein@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
single list_head variable initialized with LIST_HEAD_INIT could almost
always can be replaced with LIST_HEAD declaration, this shrinks the code
and looks better.
Signed-off-by: Denis Cheng <crquan@gmail.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Using the return value of ccw_device_set_online as return value for
dasd_generic_probe() causes the DASD to fail setting online
Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>