Vasu Dev [Wed, 29 Jul 2009 00:33:37 +0000 (17:33 -0700)]
[SCSI] libfc: fix a circular locking warning during sending RRQ
Currently the fc_exch_rrq is called with fc_exch's ex_lock held.
The fc_exch_rrq allocates new exch and that requires taking
ex_lock again after EM lock. This locking order causes warning,
see more details on this warning at :-
This patch fixes this by dropping the ex_lock before calling
fc_exch_rrq().
The fc_exch_rrq needs to grab ex_lock lock again to schedule
RRQ retry and in the meanwhile fc_exch_reset could occur before
ex_lock is grabbed inside fc_exch_rrq. So to handle this case,
this patch adds additional check to detect fc_exch_reset after
ex_lock acquired and in case the fc_exch_reset occurred then
abandons the RRQ retry and releases the exch.
Signed-off-by: Vasu Dev <vasu.dev@intel.com> Signed-off-by: Robert Love <robert.w.love@intel.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
[SCSI] qla4xxx: Remove hiwat code so scsi eh does not get escalated when we can make progress
Removed unnecessary hiwat code to free up the number available IOCBs.
Eliminates unnecessary eh_ escalations due to inability to obtain IOCB
pkt for marker.
v2.
- Remove define not used anymore and fix req_q_coun accounting.
Signed-off-by: Karen Higgins <karen.higgins@qlogic.com>
[michaelc: ported patch from qlogic.com driver to upstream] Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
[SCSI] qla4xxx: Fix srb lookup in qla4xxx_eh_device_reset
eh_device_reset may be called from scsi error handler or sg_reset, etc.
When called from sg_reset, there will not be an associated srb. The
driver should lookup the corresponding device handle given information
from the supplied cmd structure and should not assume that there exists
an srb.
Signed-off-by: Karen Higgins <karen.higgins@qlogic.com> Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Mike Christie [Wed, 15 Jul 2009 20:02:59 +0000 (15:02 -0500)]
[SCSI] qla4xxx: add timeout handler
Recently dm-multipath began calling blk_abort_queue. This causes all the
commands/request running on the path to have the timeout function called.
If a path does go down, and the LLD returns DID_*, dm-multpiath will eventually
get this error and begin to call the cmd timeout handler. qla4xxx currently
does not set a timed out handler and so the default one could return
BLK_EH_NOT_HANDLED and end up firing the scsi eh and stopping IO to all
paths on the host when only one path is affected.
For software and offload iscsi we have a timed out handler already.
This patch adds a driver specific one to qla4xxx because there
are some ddb->state and session->state and command completion races
that are better handled in the LLD.
This also handles the problem where if the session is down,
we do not need the scsi eh to run until the transport code
has tried to reconnect us.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
[SCSI] qla4xxx: Correct Extended Sense Data Errors
Fixed sense data errors occurring above the first 32 bytes,
as required by some third party applications. Sense data
in the first 32 bytes has always been correct.
Patch updated to use srb data variables instead of scsi command
scratchpad data area, as scratchpad area is already used.
Also, corrected debug print alignment bug in dump_buffer routine.
Changed KERN_DEBUG to KERN_INFO in printk statements in this routine.
Changed version number to 5.01.00-k9
Signed-off-by: Karen Higgins <karen.higgins@qlogic.com>
[michaelc: fixed checkpath.pl errors] Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Mike Christie [Wed, 15 Jul 2009 20:02:57 +0000 (15:02 -0500)]
[SCSI] libiscsi: disable bh in and abort handler.
The session lock can be held in the scsi eh thread or the completion
paths run from the net softirq. This disables bhs in iscsi_eh_abort when
taking the session lock.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Under certain conditions it is possible that a WKA port ist not opened
within the expected timeframe of half a second. In this situation
the WKA port remains in the state OPENING preventing any succeding
request to open the port. This led to unrecoverable remote ports.
Fixing this by always setting an appropriate WKA port status before
leaving the function and removing the timeout value here since it's
not needed here because the general timeout processing would deal
with it if required.
[SCSI] zfcp: Add port only once to FC transport class
When calling fc_remote_port_add make sure to not call it again before
fc_remote_port_delete has been called. In other words, ensure to
create a new fc_rport, then delete it, then create a new one again.
Depending on interruptions on some storage systems, the complete
channel can stall which looks like an outbound queue stall to Linux.
When trying to acquire a free SBAL for a non-SCSI command, zfcp waits
for 5 seconds for a free slot to appear. This is the right place to
detect a queue stall: If the wait times out, we assume a stalled queue
and try to recover this.
The overall strategy should be to trigger the erp from specific
events, and not try an overall escalation from one failed port to a
full-blown queue recovery. If we manage to send a command, the status
codes for this command or a timeout will trigger the right follow-on
actions.
If an action fails, retry it until the erp count exceeds the
threshold. If there is something fundamentally wrong, the FSF layer
will trigger a more appropriate action depending on the FSF status
codes.
The followup for successful actions is a different followup than
retrying failed actions, so split the code two functions to make this
clear.
[SCSI] zfcp: Use unchained mode for small ct and els requests
The ELS ADISC and the GID_PN requests sent from zfcp fit into
unchained FSF requests. Change the FSF allocation logic to use
unchained requests whenever possible where everything fits in one
SBAL. This avoids acquiring more SBALs than necessary, especially
during zfcp recovery when things might be stalled.
[SCSI] zfcp: Use correct flags for zfcp_erp_notify
zfcp_erp_notify uses the ZFCP_ERP_STATUS_* flags, so it is
ZFCP_STATUS_ERP_LOWMEM instead of ZFCP_ERP_NOMEM. Signalling
ZFCP_ERP_FAILED is not necessary, the missing d_id will show that the
nameserver did not return the d_id.
We should not modify the port status after triggering an ERP action
for the port. It is not guaranteed which status is finally active
when the ERP action is performed. This can lead to situations which
are unwanted and hard to debug in case of a failure.
Tom Peng [Wed, 1 Jul 2009 12:37:26 +0000 (20:37 +0800)]
[SCSI] libsas: reuse the original port when hotplugging phys in wide ports
There's a hotplug problem in the way libsas allocates ports: it loops over the
available ports first trying to add to an existing for a wide port and
otherwise allocating the next free port. This scheme only works if the port
array is packed from zero, which fails if a port gets hot unplugged and the
array becomes sparse. In that case, a new port is formed even if there's a
wide port it should be part of. Fix this by creating two loops over all the
ports: the first to see if the phy should be part of a wide port and the
second to form a new port in an empty port slot.
Signed-off-by: Tom Peng <tom_peng@usish.com> Signed-off-by: Jack Wang <jack_wang@usish.com> Signed-off-by: Lindar Liu <lindar_liu@usish.com> Cc: Stable Tree <stable@kernel.org> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Merge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev
* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
ahci: add device ID for 82801JI sata controller
drivers/ata: Move a dereference below a NULL test
libata: implement and use HORKAGE_NOSETXFER, take#2
libata: fix follow-up SRST failure path
libata: implement and use HORKAGE_NOSETXFER, take#2
PIONEER DVD-RW DVRTD08 times out SETXFER if no media is present. The
device is SATA and simply skipping SETXFER works around the problem.
Implement ATA_HORKAGE_NOSETXFER and apply it to the device.
Reported by Moritz Rigler in the following thread.
http://thread.gmane.org/gmane.linux.ide/36790
and by Lars in bko#9540.
Updated to whine and ignore NOSETXFER if PATA component is detected as
suggested by Alan Cox.
Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Moritz Rigler <linux-ide@momail.e4ward.com> Reported-by: Lars <lars21ce@gmx.de> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
ata_eh_reset() was missing error return handling after follow-up SRST
allowing EH to continue the normal probing path after reset failure.
This was discovered while testing new WD 2TB drives which take longer
than 10 secs to spin up and cause the first follow-up SRST to time
out.
Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Merge branch 'i2c-for-2631-rc3' of git://aeryn.fluff.org.uk/bjdooks/linux
* 'i2c-for-2631-rc3' of git://aeryn.fluff.org.uk/bjdooks/linux:
i2c: Use resource_size
i2c-davinci: behave with i2cdetect
i2c-davinci: convert clock usage after clkdev conversion
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
mlx4_core: Add new ConnectX EN PCI ID 0x6764
mlx4_core: Handle multi-physical function devices
Merge branch 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
hrtimer: Fix migration expiry check
hrtimer: migration: do not check expiry time on current CPU
Merge branch 'core-fixes-for-linus-2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'core-fixes-for-linus-2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
futexes: Fix infinite loop in get_futex_key() on huge page
Merge branch 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
tracing/function-profiler: do not free per cpu variable stat
tracing/events: Move TRACE_SYSTEM outside of include guard
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6:
Revert "NET: Fix locking issues in PPP, 6pack, mkiss and strip line disciplines."
skbuff.h: Fix comment for NET_IP_ALIGN
drivers/net: using spin_lock_irqsave() in net_send_packet()
NET: phy_device, fix lock imbalance
gre: fix ToS/DiffServ inherit bug
igb: gcc-3.4.6 fix
atlx: duplicate testing of MCAST flag
NET: Fix locking issues in PPP, 6pack, mkiss and strip line disciplines.
netdev: restore MTU change operation
netdev: restore MAC address set and validate operations
sit: fix regression: do not release skb->dst before xmit
net: ip_push_pending_frames() fix
net: sk_prot_alloc() should not blindly overwrite memory
David Brownell [Mon, 6 Jul 2009 22:48:36 +0000 (15:48 -0700)]
i2c-davinci: behave with i2cdetect
Make i2c-davinci cope properly with "i2cdetect": don't spew
syslog spam on perfectly normal behaviors, or respond to any
address other than the one reserved for the SMBus host.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Ben Dooks <ben-linux@fluff.org>
As Alan Cox explained, the TTY layer changes that went recently
to get rid of the tty->low_latency stuff fixes this already,
and even for -stable it's the ->low_latency changes that should
go in to fix this, rather than this patch.
Signed-off-by: David S. Miller <davem@davemloft.net>
Dongdong Deng [Sun, 12 Jul 2009 20:27:06 +0000 (20:27 +0000)]
drivers/net: using spin_lock_irqsave() in net_send_packet()
spin_unlock_irq() will enable interrupt in net_send_packet(),
this patch changes it to spin_lock_irqsave/spin_lock_irqrestore,
so that it doesn't enable interrupts when already disabled,
and netconsole would work properly over cs89x0/isa-skeleton.
Andreas Jaggi [Tue, 14 Jul 2009 16:35:59 +0000 (09:35 -0700)]
gre: fix ToS/DiffServ inherit bug
Fixes two bugs:
- ToS/DiffServ inheritance was unintentionally activated when using impair fixed ToS values
- ECN bit was lost during ToS/DiffServ inheritance
Signed-off-by: Andreas Jaggi <aj@open.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Merge branch 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4
* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
jbd2: fix race between write_metadata_buffer and get_write_access
ext4: Fix ext4_mb_initialize_context() to initialize all fields
ext4: fix null handler of ioctls in no journal mode
ext4: Fix buffer head reference leak in no-journal mode
ext4: Move __ext4_journalled_writepage() to avoid forward declaration
ext4: Fix mmap/truncate race when blocksize < pagesize && !nodellaoc
ext4: Fix mmap/truncate race when blocksize < pagesize && delayed allocation
ext4: Don't look at buffer_heads outside i_size.
ext4: Fix goal inum check in the inode allocator
ext4: fix no journal corruption with locale-gen
ext4: Calculate required journal credits for inserting an extent properly
ext4: Fix truncation of symlinks after failed write
jbd2: Fix a race between checkpointing code and journal_get_write_access()
ext4: Use rcu_barrier() on module unload.
ext4: naturally align struct ext4_allocation_request
ext4: mark several more functions in mballoc.c as noinline
ext4: Fix potential reclaim deadlock when truncating partial block
jbd2: Remove GFP_ATOMIC kmalloc from inside spinlock critical region
ext4: Fix type warning on 64-bit platforms in tracing events header
jbd2: fix race between write_metadata_buffer and get_write_access
The function jbd2_journal_write_metadata_buffer() calls
jbd_unlock_bh_state(bh_in) too early; this could potentially allow
another thread to call get_write_access on the buffer head, modify the
data, and dirty it, and allowing the wrong data to be written into the
journal. Fortunately, if we lose this race, the only time this will
actually cause filesystem corruption is if there is a system crash or
other unclean shutdown of the system before the next commit can take
place.
Eric Dumazet [Mon, 13 Jul 2009 18:11:41 +0000 (11:11 -0700)]
igb: gcc-3.4.6 fix
forward declaration of inline function should be avoided, or
old gcc cannot compile.
Reported-by: Teck Choon Giam <giamteckchoon@gmail.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging-2.6:
Staging: stlc45xx: convert config_interface to bss_info_changed, fixing a build error
Staging: comedi: s626: use subvendor:subdevice ids for SAA7146 board
Staging: prevent rtl8192su from crashing dev_ioctl in SIOCGIWNAME
Staging: prevent rtl8187se from crashing dev_ioctl() in SIOCGIWNAME
Staging: rtl8192su: convert to net_device_ops
Staging: serqt_usb2: declare qt_open static in serqt_usb2
Staging: serqt_usb2: fix qt_close parameters in serqt_usb2
Staging: comedi: jr3_pci.c: add required includes
Staging: meilhaus: add email address to TODO
Staging: rspiusb: use NULL virtual address instead of a bogus one
Staging: vt6655: compile fix
Staging: rt2870: Add USB ID for Sitecom WL-608
Dave Kleikamp [Mon, 13 Jul 2009 16:02:24 +0000 (11:02 -0500)]
update JFS entry in MAINTAINERS
JFS hasn't really been supported for a while. It's still maintained,
but saying it's supported is a stretch. Updating my preferred email
address as well.
Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
ext4: Fix ext4_mb_initialize_context() to initialize all fields
Pavel Roskin pointed out that kmemcheck indicated that
ext4_mb_store_history() was accessing uninitialized values of
ac->ac_tail and ac->ac_buddy leading to garbage in the mballoc
history. Fix this by initializing the entire structure to all zeros
first.
Also, two fields were getting doubly initialized by the caller of
ext4_mb_initialize_context, so remove them for efficiency's sake.
Peng Tao [Mon, 13 Jul 2009 13:30:17 +0000 (09:30 -0400)]
ext4: fix null handler of ioctls in no journal mode
The EXT4_IOC_GROUP_ADD and EXT4_IOC_GROUP_EXTEND ioctls should not
flush the journal in no_journal mode. Otherwise, running resize2fs on
a mounted no_journal partition triggers the following error messages:
BUG: unable to handle kernel NULL pointer dereference at 00000014
IP: [<c039d282>] _spin_lock+0x8/0x19
*pde = 00000000
Oops: 0002 [#1] SMP
Signed-off-by: Peng Tao <bergwolf@gmail.com> Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
ext4: Fix buffer head reference leak in no-journal mode
We found a problem with buffer head reference leaks when using an ext4
partition without a journal. In particular, calls to ext4_forget() would
not to a brelse() on the input buffer head, which will cause pages they
belong to to not be reclaimable.
Further investigation showed that all places where ext4_journal_forget() and
ext4_journal_revoke() are called are subject to the same problem. The patch
below changes __ext4_journal_forget/__ext4_journal_revoke to do an explicit
release of the buffer head when the journal handle isn't valid.
x86, apic: Fix false positive section mismatch in numaq_32.c
The variable apic_numaq placed in noninit section references the
function wakeup_secondary_cpu_via_nmi(), which is in __cpuinit
section. Thus causes a section mismatch warning. To avoid such
mismatch we mark apic_numaq as __refdata.
We were warned by the following warning:
WARNING: arch/x86/kernel/built-in.o(.data+0x932c): Section mismatch in
reference from the variable apic_numaq to the function
.cpuinit.text:wakeup_secondary_cpu_via_nmi()
x86: Fix false positive section mismatch in es7000_32.c
The variable apic_es7000_cluster references the function __cpuinit
wakeup_secondary_cpu_via_mip() from a noninit section. So we've been
warned by the following warning. To avoid possible collision between
init/noninit, its best to mark the variable as __refdata.
We were warned by the following warning:
LD arch/x86/kernel/apic/built-in.o
WARNING: arch/x86/kernel/apic/built-in.o(.data+0x198c): Section
mismatch in reference from the variable apic_es7000_cluster to the
function .cpuinit.text:wakeup_secondary_cpu_via_mip()
Steven Rostedt [Thu, 4 Jun 2009 04:55:45 +0000 (00:55 -0400)]
tracing/function-profiler: do not free per cpu variable stat
The per cpu variable stat is freeded if we fail to allocate a name
on start up. This was due to stat at first being allocated in the
initial design. But since then, it has become a static per cpu variable
but the free on error was not removed.
Also added __init annotation to the function that this is in.
[ Impact: prevent possible memory corruption on low mem at boot up ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Li Zefan [Mon, 13 Jul 2009 02:33:21 +0000 (10:33 +0800)]
tracing/events: Move TRACE_SYSTEM outside of include guard
If TRACE_INCLDUE_FILE is defined, <trace/events/TRACE_INCLUDE_FILE.h>
will be included and compiled, otherwise it will be
<trace/events/TRACE_SYSTEM.h>
So TRACE_SYSTEM should be defined outside of #if proctection,
just like TRACE_INCLUDE_FILE.
The issue was, that the locking code in mkiss was assuming it was only
ever being called in process or bh context. Fixed by converting the
involved locking code to use irq-safe locks.
Review of other networking line disciplines shows that 6pack, both sync
and async PPP and STRIP have similar issues. The ppp_async one is the
most interesting one as it sorts out half of the issue as far back as
2004 in commit http://git.kernel.org/?p=linux/kernel/git/tglx/history.git;a=commitdiff;h=2996d8deaeddd01820691a872550dc0cfba0c37d
Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Reported-by: Guido Trentalancia <guido@trentalancia.com> Signed-off-by: David S. Miller <davem@davemloft.net>
This patch (as1262) fixes a bug in usbfs: It refuses to accept
zero-length transfers, and it insists that the buffer pointer be valid
even if there is no data being transferred.
The patch also consolidates a bunch of repetitive access_ok() checks
into a single check, which incidentally fixes the lack of such a check
for Isochronous URBs.
Alan Stern [Mon, 29 Jun 2009 18:34:59 +0000 (14:34 -0400)]
USB: EHCI: report actual_length for iso transfers
This patch (as1259b) makes ehci-hcd return the total number of bytes
transferred in urb->actual_length for Isochronous transfers.
Until now, the actual_length value was unaccountably left at 0.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Acked-by: David Brownell <dbrownell@users.sourceforge.net> Cc: stable <stable@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Thu, 9 Jul 2009 16:58:43 +0000 (12:58 -0400)]
USB: option: remove unnecessary and erroneous code
This patch (as1264) removes a bunch of unnecessary and erroneous stuff
from the option USB-serial driver. Clearly there's no need to verify
that the device pointer stored in the URBs is right or to store the
same pointer over again. After all, the pointer can't change once it
has been set up.
There's also no need to call usb_clear_halt for the IN endpoint
multiple times -- in fact, doing so is an error since every time after
the first there will be active URBs queued for that endpoint. Since
the Clear-Halts don't appear to be needed at all, the patch simply
removes them.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Thu, 9 Jul 2009 16:59:57 +0000 (12:59 -0400)]
USB: cypress_m8: remove invalid Clear-Halt
This patch (as1265) removes an erroneous call to usb_clear_halt from
the cypress_m8 driver. The call isn't valid because it is made from
interrupt context whereas usb_clear_halt is a blocking routine.
Presumably the code has never been executed; if it did it would cause
an oops. So instead treat -EPIPE like any other sort of unexplained
error.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Sergei Shtylyov [Fri, 10 Jul 2009 17:02:44 +0000 (20:02 +0300)]
USB: musb_host: undo incorrect change in musb_advance_schedule()
Commit c9cd06b3d6ea825c62e277def929cc4315802b48 (musb_host: refactor
URB giveback) included due to my overlook the change incorrect in the
context of the current kernel -- undo it.
Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com> Acked-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Daniel Mack [Fri, 10 Jul 2009 09:04:58 +0000 (11:04 +0200)]
USB: fix LANGID=0 regression
commit b7af0bb ("USB: allow malformed LANGID descriptors") broke support
for devices without string descriptor support.
Reporting string descriptors is optional to USB devices, and a device
lets us know it can't deal with strings by responding to the LANGID
request with a STALL token.
The kernel handled that correctly before b7af0bb came in, but failed
hard if the LANGID was reported but broken. More than that, if a device
was not able to provide string descriptors, the LANGID was retrieved
over and over again at each string read request.
This patch changes the behaviour so that
a) the LANGID is only queried once
b) devices which can't handle string requests are not asked again
c) devices with malformed LANGID values have a sane fallback to 0x0409
Signed-off-by: Daniel Mack <daniel@caiaq.de> Acked-by: Alan Stern <stern@rowland.harvard.edu> Cc: stable <stable@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Tue, 7 Jul 2009 13:50:14 +0000 (09:50 -0400)]
USB: Sierra: fix oops upon device close
This patch (as1263) fixes a mixup that occurred when conflicting
patches for the sierra driver were merged incorrectly. The former
sierra_shutdown routine should have been become sierra_release, not
sierra_disconnect.
The symptom this fixes is an oops when the device file is closed after
a Sierra device has been unplugged (Bugzilla #13675).
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Tested-by: Peter Naulls <peter@mushroomnetworks.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Commit 32ebbe7b6ad44ae9c276419710b56de6ba705303 which filters the
SCSI REZERO command in option_ms based on a SCSI INQUIRY with a vendor
of Option breaks my Option Icon 225 (0af0:6971). This device returns a
vendor of ZCOPTION for the ZeroCD device. The following trivial patch
fixes things for me.
Signed-Off-By: Jonathan McDowell <noodles@earth.li> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Randy Dunlap [Fri, 3 Jul 2009 20:26:57 +0000 (13:26 -0700)]
USB: gadget audio: select SND_PCM
Fix USB gadget audio: select SND_PCM, like many other sound
drivers do, to fix build errors:
drivers/built-in.o: In function `f_audio_playback_work':
audio.c:(.text+0x15a3e7): undefined reference to `snd_pcm_kernel_ioctl'
audio.c:(.text+0x15a471): undefined reference to `snd_pcm_lib_write'
drivers/built-in.o: In function `_snd_pcm_hw_param_set':
audio.c:(.text+0x15aca7): undefined reference to `snd_interval_refine'
drivers/built-in.o: In function `gaudio_setup':
(.init.text+0x12adf): undefined reference to `_snd_pcm_hw_params_any'
drivers/built-in.o: In function `gaudio_setup':
(.init.text+0x12b43): undefined reference to `snd_pcm_kernel_ioctl'
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Acked-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Martin Geleynse [Thu, 2 Jul 2009 17:10:35 +0000 (13:10 -0400)]
USB: ftdi: support NDI devices
It enhances the driver for FTDI-based USB serial adapters to recognize and
support Northern Digital Inc (NDI) measurement equipment. NDI has been
providing this patch for various kernel flavors for several years and we would
like to see these changes built in to the driver so that our equipement works
without the need for customers to patch the kernel themselves.
The patch makes small modifications to 2 files: ./drivers/usb/serial/ftdi_sio.c
and ./drivers/usb/serial/ftdi_sio.h. It accomplishes 3 things:
1. Define the VID and PIDs to allow the driver to recognize the NDI devices.
2. Map the 19200 baud rate setting to our higher baud rate of 1.2Mb
We would have chosen to map 38400 to the higher rate, similar to what
several other vendors have done, but some of our legacy customers actually
use 38400, therefore we remap 19200 to the higher rate.
3. We set the default transmit latency in the FTDI chip to 1ms for our devices.
Our devices are typically polled at 60Hz and the default ftdi latency
seriously affects turn-around time and results in missed data frames. We
have created a modprobe option that allows this setting to be increased.
This has proven necessary particularly in some virtualized environments.
Signed-off-by: Martin P. Geleynse <mgeleyns@ndigital.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Revert USB: usbfs: deprecate and hide option for !embedded
This reverts commit cc71329b3b89b4a5be849b617f2c4f151f0b9213, so that
Red Hat machines can boot properly. It seems that the Red Hat initrd
code tries to watch the /proc/bus/usb/devices file to monitor usb
devices showing up. While this task is prone to lots of races and does
not show the true state of the system, they seem to like it.
So for now, don't move this option under the EMBEDDED config option.
Cc: Scott James Remnant <scott@canonical.com> Cc: Kay Sievers <kay.sievers@vrfy.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Peter Jones <pjones@redhat.com> Cc: Jeff Chua <jeff.chua.linux@gmail.com> Cc: Dave Jones <davej@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Randy Dunlap [Thu, 18 Jun 2009 17:39:11 +0000 (10:39 -0700)]
USB: usb.h: fix kernel-doc notation
Fix usb.h kernel-doc warnings:
Warning(include/linux/usb.h:918): Excess struct/union/enum/typedef member 'nodename' description in 'usb_device_driver'
Warning(include/linux/usb.h:939): No description found for parameter 'nodename'
Warning(include/linux/usb.h:1219): No description found for parameter 'sg'
Warning(include/linux/usb.h:1219): No description found for parameter 'num_sgs'
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
David Brownell [Fri, 19 Jun 2009 10:09:04 +0000 (03:09 -0700)]
USB: RNDIS gadget, fix issues talking from PXA
The reworked Ethernet gadget has an RNDIS interop problem when used
with the CDC subset driver ... e.g. on PXA 2xx and 3xx hardware,
which currently has a hard time talking to MS-Windows hosts.
The issue is that Microsoft requires USB_CLASS_COMM. Fix by tweaking
the CDC subset driver to not switch to USB_CLASS_VENDOR_SPEC if RNDIS
is used in some other device configuration.
[ UPDATED: some "statements" were comma-terminated; fix that. ]
USB: serial: FTDI with product code FB80 and vendor id 0403
It seems an USB device with vendor id 0403 and product code FB80 has an
FTDI serial io chip as well: http://ftdichip.com/Drivers/D2XX.htm
This device in fact is a true random generantor by comsci:
http://comscire.com/Products/R2000KU/
So the following patch should add support for this device if I am
correct. Not tested as I do not own this device (I would like support in
the kernel so that my entropybroker application (which distributes
entrop data (random values) between servers and clients)).
From: Folkert van Heusden <folkert@vanheusden.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Jason Wessel [Mon, 22 Jun 2009 16:32:20 +0000 (11:32 -0500)]
USB: console: Fix regression in usb console on kernel boot
The commit 335f8514f200e63d689113d29cb7253a5c282967 introduced a
regression which stopped usb consoles from working correctly as a
kernel boot console as well as interactive login device.
The addition of the serial_close() which in turn calls
tty_port_close_start() will change the reference count of port.count
and warn about it. The usb console code had previously incremented
the port.count to indicate it was making use of the device as a
console and the forced change causes a double open on the usb device
which leads to a non obvious kernel oops later on when the tty is
freed.
To fix the problem instead make use of port->console to track if the
port is in fact an active console port to avoid double initialization
of the usb serial device. The port.count is incremented and
decremented only with in the scope of usb_console_setup() for the
purpose of the low level driver initialization.
Signed-off-by: Jason Wessel <jason.wessel@windriver.com> Acked-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Mon, 29 Jun 2009 21:36:14 +0000 (17:36 -0400)]
USB: EHCI: check for STALL before other errors
This patch (as1257) revises the way ehci-hcd detects STALLs. The
logic is a little peculiar because there's no hardware status bit
specifically meant to indicate a STALL. You just have to guess that a
STALL was received if the BABBLE bit (which is fatal) isn't set and
the transfer stopped before all its retries were used up.
The existing code doesn't do this properly, because it tests for MMF
(Missed MicroFrame) and DBE (Data Buffer Error) before testing the
retry counter. Thus, if a transaction gets either MMF or DBE the
corresponding flag is set and the transaction is retried. If the
second attempt receives a STALL then -EPIPE is the correct return
value. But the existing code would see the MMF or DBE flag instead
and return -EPROTO, -ENOSR, or -ECOMM.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Acked-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Mon, 29 Jun 2009 14:47:30 +0000 (10:47 -0400)]
USB: EHCI: use the new clear_tt_buffer interface
This patch (as1256) changes ehci-hcd and all the other drivers in the
EHCI family to make use of the new clear_tt_buffer callbacks. When a
Clear-TT-Buffer request is in progress for a QH, the QH is not allowed
to be linked into the async schedule until the request is finished.
At that time, if there are any URBs queued for the QH, it is linked
into the async schedule.
Alan Stern [Mon, 29 Jun 2009 14:43:32 +0000 (10:43 -0400)]
USB: fix the clear_tt_buffer interface
This patch (as1255) updates the interface for calling
usb_hub_clear_tt_buffer(). Even the name of the function is changed!
When an async URB (i.e., Control or Bulk) going through a high-speed
hub to a non-high-speed device is cancelled or fails, the hub's
Transaction Translator buffer may be left busy still trying to
complete the transaction. The buffer has to be cleared; that's what
usb_hub_clear_tt_buffer() does.
It isn't safe to send any more URBs to the same endpoint until the TT
buffer is fully clear. Therefore the HCD needs to be told when the
Clear-TT-Buffer request has finished. This patch adds a callback
method to struct hc_driver for that purpose, and makes the hub driver
invoke the callback at the proper time.
The patch also changes a couple of names; "hub_tt_kevent" and
"tt.kevent" now look rather antiquated.
This patch introduces a work around for cdc-acm devices which are
low speed contrary to the specification, which requires bulk endpoints
which are banned in low speed and converted by usbcore to virtual
interrupt endpoints if they are used nevertheless.
David Brownell [Wed, 1 Jul 2009 10:36:16 +0000 (03:36 -0700)]
USB: musb: silence "suspend as a_wait_vrise is_active" msgs
Get rid of some obnoxious and inappropriate messaging, mostly on
DaVinci, when usbcore tries to autosuspend a root hub if just a
mini/micro-A connector is connected. Symptom: endless stream of
messages reading like:
musb_bus_suspend 2221: trying to suspend as a_wait_vrise is_active=1
Improve that musb bus suspend primitive a bit. Take advantage of
this call to update the OTG state machine if appropriate, moving
the device out of the A_WAIT_VRISE state. There's basically no
timer for that state transition just now, except with tusb6010;
that can make trouble.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
David Brownell [Wed, 1 Jul 2009 10:43:58 +0000 (03:43 -0700)]
USB: gadget: pxa25x compiler warning fix
Fix config-dependent compiler warning:
CC drivers/usb/gadget/pxa25x_udc.o
drivers/usb/gadget/pxa25x_udc.c: In function 'pxa25x_udc_irq':
drivers/usb/gadget/pxa25x_udc.c:1806: warning: array subscript is above array bounds
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Kevin Hilman [Wed, 1 Jul 2009 10:33:46 +0000 (03:33 -0700)]
USB: musb: davinci dm6446evm GPIO renumbering
Numbering for GPIOs on the pcf857x chips on the dm644x EVM board
changed when DaVinci chips with more GPIOs were supported. Update
the GPIO number used for nVBUS_DRV.
Longer term, we need a better abstraction of board-specific setup in
this code so we're not hard-coding board specific GPIOs into the
driver, but for now this at least gets it back to working with
mainline davinci core code.
Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
[ dbrownell@users.sourceforge.net: remove OID_802_3_MAXIMUM_LIST_SIZE
setting ... it was bogus, this code only handles one entry, not 32.
And we don't know what would break if we lied about that... ]
Signed-off-by: Helen Chen <helen.chen@intel.com> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Arnd Bergmann [Fri, 19 Jun 2009 12:35:34 +0000 (05:35 -0700)]
USB: buildfix ppc randconfig
We could just make the USB_OHCI_HCD_PPC_OF option implicit
and selected only if at least one of USB_OHCI_HCD_PPC_OF_BE
and USB_OHCI_HCD_PPC_OF_LE are set.
[ dbrownell@users.sourceforge.net: fix patch manglation and dependencies ]