| Age | Commit message (Collapse) | Author |
|---|
|
This is a minor change.
Up until kernel 2.6.32, getsockopt(fd, SOL_PACKET, PACKET_STATISTICS,
...) would return total and dropped packets since its last invocation. The
introduction of socket queue overflow reporting [1] changed drop
rate calculation in the normal packet socket path, but not when using a
packet ring. As a result, the getsockopt now returns different statistics
depending on the reception method used. With a ring, it still returns the
count since the last call, as counts are incremented in tpacket_rcv and
reset in getsockopt. Without a ring, it returns 0 if no drops occurred
since the last getsockopt and the total drops over the lifespan of
the socket otherwise. The culprit is this line in packet_rcv, executed
on a drop:
drop_n_acct:
po->stats.tp_drops = atomic_inc_return(&sk->sk_drops);
As it shows, the new drop number it taken from the socket drop counter,
which is not reset at getsockopt. I put together a small example
that demonstrates the issue [2]. It runs for 10 seconds and overflows
the queue/ring on every odd second. The reported drop rates are:
ring: 16, 0, 16, 0, 16, ...
non-ring: 0, 15, 0, 30, 0, 46, 0, 60, 0 , 74.
Note how the even ring counts monotonically increase. Because the
getsockopt adds tp_drops to tp_packets, total counts are similarly
reported cumulatively. Long story short, reinstating the original code, as
the below patch does, fixes the issue at the cost of additional per-packet
cycles. Another solution that does not introduce per-packet overhead
is be to keep the current data path, record the value of sk_drops at
getsockopt() at call N in a new field in struct packetsock and subtract
that when reporting at call N+1. I'll be happy to code that, instead,
it's just more messy.
[1] http://patchwork.ozlabs.org/patch/35665/
[2] http://kernel.googlecode.com/files/test-packetsock-getstatistics.c
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
If a VM is saved and restored (or migrated) the netback driver will no
longer process any Tx packets from the frontend. xenvif_up() does not
schedule the processing of any pending Tx requests from the front end
because the carrier is off. Without this initial kick the frontend
just adds Tx requests to the ring without raising an event (until the
ring is full).
This was caused by 47103041e91794acdbc6165da0ae288d844c820b (net:
xen-netback: convert to hw_features) which reordered the calls to
xenvif_up() and netif_carrier_on() in xenvif_connect().
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
During a test where a pair of bonding interfaces using ARP monitoring
were both brought up and torn down (with an rmmod) repeatedly, a panic
in the timer code was noticed. I tracked this down and determined that
any of the bonding functions that ran as workqueue handlers and requeued
more work might not properly exit when the module was removed.
There was a flag protected by the bond lock called kill_timers that is
set when the interface goes down or the module is removed, but many of
the functions that monitor link status now unlock the bond lock to take
rtnl first. There is a chance that another CPU running the rmmod could
get the lock and set kill_timers after the first check has passed.
This patch does not allow any function to queue work that will make
itself run unless kill_timers is not set. I also noticed while doing
this work that bond_resend_igmp_join_requests did not have a check for
kill_timers, so I added the needed call there as well.
Signed-off-by: Andy Gospodarek <andy@greyhouse.net>
Reported-by: Liang Zheng <lzheng@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The commit aabdcb0b553b9c9547b1a506b34d55a764745870 ("can bcm: fix tx_setup
off-by-one errors") fixed only a part of the original problem reported by
Andre Naujoks. It turned out that the original code needed to be re-ordered
to reduce complexity and to finally fix the reported frame counting issues.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
In the rds_iw_mr_pool struct the free_pinned field keeps track of
memory pinned by free MRs. While this field is incremented properly
upon allocation, it is never decremented upon unmapping. This would
cause the rds_rdma module to crash the kernel upon unloading, by
triggering the BUG_ON in the rds_iw_destroy_mr_pool function.
This change keeps track of the MRs that become unpinned, so that
free_pinned can be decremented appropriately.
Signed-off-by: Jonathan Lallinger <jonathan@ogc.us>
Signed-off-by: Steve Wise <swise@ogc.us>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Signed-off-by: Roy.Li <rongqing.li@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
If request_irq fails, the ibmveth driver will overwrite
the rc and end up returning a successful rc on its open
function, resulting in an oops later when a packet gets
sent and buffers are not allocated due to the failed open.
Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
ipv6_ac_list and ipv6_fl_list from listening socket are inadvertently
shared with new socket created for connection.
Signed-off-by: Zheng Yan <zheng.z.yan@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Fix EEH recovery on new P Series platform by
requesting fundamental reset.
Signed-off-by: Divy Le Ray <divy@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This patch fixes two off-by-one errors that canceled each other out.
Checking for the same condition two times in bcm_tx_timeout_tsklet() reduced
the count of frames to be sent by one. This did not show up the first time
tx_setup is invoked as an additional frame is sent due to TX_ANNONCE.
Invoking a second tx_setup on the same item led to a reduced (by 1) number of
sent frames.
Reported-by: Andre Naujoks <nautsch@gmail.com>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
I got:
Generating server: Tehuti.onmicrosoft.com
baum@tehutinetworks.net
#< #5.1.1 smtp;550 5.1.1 RESOLVER.ADR.RecipNotFound; not found> #SMTP#
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Cc: Alexander Indenbaum <baum@tehutinetworks.net>
Cc: Andy Gospodarek <andy@greyhouse.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: netdev@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The driver has two warning messages that might be triggered
by normal use cases. When they appear, the messages give the
impression of a never ending series of errors.
This commit changes them to debug messages instead.
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The IEEE 1588 standard defines two kinds of messages, event and general
messages. Event messages require time stamping, and general do not. When
using UDP transport, two separate ports are used for the two message
types.
The BPF designed to recognize event messages incorrectly classifies L2
general messages as event messages. This commit fixes the issue by
extending the filter to check the message type field for L2 PTP packets.
Event messages are be distinguished from general messages by testing
the "general" bit.
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Cc: <stable@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Doing it just before starting to call into cpu_idle() made a sick kind
of sense only because the original bug we fixed (see commit
288d5abec831: "Boot up with usermodehelper disabled") was about problems
with some scheduler data structures not being initialized, and they had
better be initialized at that point.
But it really didn't make any other conceptual sense, and doing it after
the initial "schedule()" call for the idle thread actually opened up a
race: what if the main initialization thread did everything without
needing to sleep, and got all the way into user land too? Without
actually having scheduled back to the idle thread?
Now, in normal circumstances that doesn't ever happen, but it looks like
Richard Cochran triggered exactly that on his ARM IXP4xx machines:
"I have some ARM IXP4xx based machines that use the two on chip MAC
ports (aka NPEs). The NPE needs a firmware in order to function.
Ever since the following commit [that 288d5abec831 one], it is no
longer possible to bring up the interfaces during the init scripts."
with a call trace showing an ioctl coming from user space. Richard says:
"The init is busybox, and the startup script does mount, syslogd, and
then ifup, so that all can go by quickly."
The fix is to move the usermodehelper_enable() into the main 'init'
thread, and just put it after we've done all our initcalls. By then,
everything really should be up, but we've obviously not actually started
the user-mode portion of init yet.
Reported-and-tested-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
* git://github.com/davem330/net:
ipv6-multicast: Fix memory leak in IPv6 multicast.
ipv6: check return value for dst_alloc
net: check return value for dst_alloc
ipv6-multicast: Fix memory leak in input path.
bnx2x: add missing break in bnx2x_dcbnl_get_cap
bnx2x: fix WOL by enablement PME in config space
bnx2x: fix hw attention handling
net: fix a typo in Documentation/networking/scaling.txt
ath9k: Fix a dma warning/memory leak
rtlwifi: rtl8192cu: Fix unitialized struct
iwlagn: fix dangling scan request
batman-adv: do_bcast has to be true for broadcast packets only
cfg80211: Fix validation of AKM suites
iwlegacy: do not use interruptible waits
iwlegacy: fix command queue timeout
ath9k_hw: Fix Rx DMA stuck for AR9003 chips
|
|
* git://bedivere.hansenpartnership.com/git/scsi-rc-fixes-2.6:
[SCSI] 3w-9xxx: fix iommu_iova leak
[SCSI] cxgb3i: convert cdev->l2opt to use rcu to prevent NULL dereference
[SCSI] scsi: qla4xxx needs libiscsi.o
[SCSI] libsas: fix failure to revalidate domain for anything but the first expander child.
[SCSI] aacraid: reset should disable MSI interrupt
|
|
* 'for-linus' of git://git.kernel.dk/linux-block:
block: Free queue resources at blk_release_queue()
|
|
* 'writeback-for-linus' of git://github.com/fengguang/linux:
writeback: show raw dirtied_when in trace writeback_single_inode
|
|
A kernel crash is observed when a mounted ext3/ext4 filesystem is
physically removed. The problem is that blk_cleanup_queue() frees up
some resources eg by calling elevator_exit(), which are not checked for
in normal operation. So we should rather move these calls to the
destructor function blk_release_queue() as at that point all remaining
references are gone. However, in doing so we have to ensure that any
externally supplied queue_lock is disconnected as the driver might free
up the lock after the call of blk_cleanup_queue(),
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
|
|
|
|
* 'for-linus' of git://github.com/tiwai/sound:
ASoC: ssm2602: Re-enable oscillator after suspend
ALSA: usb-audio: Check for possible chip NULL pointer before clearing probing flag
ALSA: hda/realtek - Don't detect LO jack when identical with HP
ALSA: hda/realtek - Avoid bogus HP-pin assignment
ALSA: HDA: No power nids on 92HD93
ASoC: omap-mcbsp: Do not attempt to change DAI sysclk if stream is active
|
|
* 'pm-fixes' of git://github.com/rjwysocki/linux-pm:
PM / Clocks: Do not acquire a mutex under a spinlock
|
|
into for-davem
|
|
If reg_vif_xmit cannot find a routing entry, be sure to
free the skb before returning the error.
Signed-off-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
return value of dst_alloc must be checked before use
Signed-off-by: Madalin Bucur <madalin.bucur@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
return value of dst_alloc must be checked before use
Signed-off-by: Madalin Bucur <madalin.bucur@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Have to free the skb before returning if we fail
the fib lookup.
Signed-off-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use register name to initialize attention mask
Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
That flag no longer makes sense, since we don't look up automount points
as eagerly any more. Additionally, it turns out that the NO_AUTOMOUNT
handling was buggy to begin with: it would avoid automounting even for
cases where we really *needed* to do the automount handling, and could
return ENOENT for autofs entries that hadn't been instantiated yet.
With our new non-eager automount semantics, one discussion has been
about adding a AT_AUTOMOUNT flag to vfs_fstatat (and thus the
newfstatat() and fstatat64() system calls), but it's probably not worth
it: you can always force at least directory automounting by simply
adding the final '/' to the filename, which works for *all* of the stat
family system calls, old and new.
So AT_NO_AUTOMOUNT (and thus LOOKUP_NO_AUTOMOUNT) really were just a
result of our bad default behavior.
Acked-by: Ian Kent <raven@themaw.net>
Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
Currently the the internal oscillator is powered down when entering BIAS_OFF
state, but not re-enabled when going back to BIAS_STANDBY. As a result the
CODEC will stop working after suspend if the internal oscillator is used to
generate the sysclock signal. This patch fixes it by clearing the appropriate
bit in the power down register when the CODEC is re-enabled.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Cc: stable@kernel.org
|
|
The concensus seems to be that system calls such as stat() etc should
not trigger an automount. Neither should the l* versions.
This patch therefore adds a LOOKUP_AUTOMOUNT flag to tag those lookups
that _should_ trigger an automount on the last path element.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
[ Edited to leave out the cases that are already covered by LOOKUP_OPEN,
LOOKUP_DIRECTORY and LOOKUP_CREATE - all of which also fundamentally
force automounting for their own reasons - Linus ]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
Since we've now turned around and made LOOKUP_FOLLOW *not* force an
automount, we want to add the ability to force an automount event on
lookup even if we don't happen to have one of the other flags that force
it implicitly (LOOKUP_OPEN, LOOKUP_DIRECTORY, LOOKUP_PARENT..)
Most cases will never want to use this, since you'd normally want to
delay automounting as long as possible, which usually implies
LOOKUP_OPEN (when we open a file or directory, we really cannot avoid
the automount any more).
But Trond argued sufficiently forcefully that at a minimum bind mounting
a file and quotactl will want to force the automount lookup. Some other
cases (like nfs_follow_remote_path()) could use it too, although
LOOKUP_DIRECTORY would work there as well.
This commit just adds the flag and logic, no users yet, though. It also
doesn't actually touch the LOOKUP_NO_AUTOMOUNT flag that is related, and
was made irrelevant by the same change that made us not follow on
LOOKUP_FOLLOW.
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Ian Kent <raven@themaw.net>
Cc: Jeff Layton <jlayton@redhat.com>
Cc: Miklos Szeredi <miklos@szeredi.hu>
Cc: David Howells <dhowells@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Greg KH <gregkh@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
* 'samsung-fixes-3' of git://github.com/kgene/linux-samsung:
ARM: EXYNOS4: Rename sclk_cam clocks for FIMC driver
ARM: S5PV210: Rename sclk_cam clocks for FIMC media driver
ARM: S5P: fix incorrect loop iterator usage on gpio-interrupt
ARM: S3C2443: Fix bit-reset in setrate of clk_armdiv
|
|
The sclk_cam clocks are now controlled by the top level FIMC media
device driver bound to "s5p-fimc-md" platform device.
Rename sclk_cam clocks so they accessible by the corresponding
driver.
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
The sclk_cam clocks are now controlled by the top level FIMC media
device driver bound to "s5p-fimc-md" platform device.
Rename sclk_cam clocks so they accessible by the corresponding
driver.
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
* 'hwmon-for-linus' of git://github.com/groeck/linux:
hwmon: (coretemp) remove struct platform_data * parameter from create_core_data()
hwmon: (coretemp) constify static data
hwmon: (coretemp) don't use kernel assigned CPU number as platform device ID
hwmon: (ds620) Fix handling of negative temperatures
hwmon: (w83791d) rename prototype parameter from 'register' to 'reg'
hwmon: (coretemp) Don't use threshold registers for tempX_max
hwmon: (coretemp) Let the user force TjMax
hwmon: (coretemp) Drop duplicate function get_pkg_tjmax
|
|
* 'kvm-updates/3.1' of git://github.com/avikivity/kvm:
KVM: x86 emulator: fix Src2CL decode
KVM: MMU: fix incorrect return of spte
|
|
http://ftp.arm.linux.org.uk/pub/linux/arm/kernel/git-cur/linux-2.6-arm
* 'fixes' of http://ftp.arm.linux.org.uk/pub/linux/arm/kernel/git-cur/linux-2.6-arm:
ARM: 7099/1: futex: preserve oldval in SMP __futex_atomic_op
ARM: dma-mapping: free allocated page if unable to map
ARM: fix vmlinux.lds.S discarding sections
ARM: nommu: fix warning with checksyscalls.sh
ARM: 7091/1: errata: D-cache line maintenance operation by MVA may not succeed
|
|
proper dma_unmapping and freeing of skb's has to be done in the rx
cleanup for EDMA chipsets when the device is unloaded and this also
seems to address the following warning which shows up occasionally when
the device is unloaded
Call Trace:
[<c0148cd2>] warn_slowpath_common+0x72/0xa0
[<c03b669c>] ? dma_debug_device_change+0x19c/0x200
[<c03b669c>] ? dma_debug_device_change+0x19c/0x200
[<c0148da3>] warn_slowpath_fmt+0x33/0x40
[<c03b669c>] dma_debug_device_change+0x19c/0x200
[<c0657f12>] notifier_call_chain+0x82/0xb0
[<c0171370>] __blocking_notifier_call_chain+0x60/0x90
[<c01713bf>] blocking_notifier_call_chain+0x1f/0x30
[<c044f594>] __device_release_driver+0xa4/0xc0
[<c044f647>] driver_detach+0x97/0xa0
[<c044e65c>] bus_remove_driver+0x6c/0xe0
[<c029af0b>] ? sysfs_addrm_finish+0x4b/0x60
[<c0450109>] driver_unregister+0x49/0x80
[<c0299f54>] ? sysfs_remove_file+0x14/0x20
[<c03c3ab2>] pci_unregister_driver+0x32/0x80
[<f92c2162>] ath_pci_exit+0x12/0x20 [ath9k]
[<f92c8467>] ath9k_exit+0x17/0x36 [ath9k]
[<c06523cd>] ? mutex_unlock+0xd/0x10
[<c018e27f>] sys_delete_module+0x13f/0x200
[<c02139bb>] ? sys_munmap+0x4b/0x60
[<c06547c5>] ? restore_all+0xf/0xf
[<c0657a20>] ? spurious_fault+0xe0/0xe0
[<c01832f4>] ? trace_hardirqs_on_caller+0xf4/0x180
[<c065b863>] sysenter_do_call+0x12/0x38
---[ end trace 16e1c1521c06bcf9 ]---
Mapped at:
[<c03b7938>] debug_dma_map_page+0x48/0x120
[<f92ba3e8>] ath_rx_init+0x3f8/0x4b0 [ath9k]
[<f92b5ae4>] ath9k_init_device+0x4c4/0x7b0 [ath9k]
[<f92c2813>] ath_pci_probe+0x263/0x330 [ath9k]
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
|
|
Driver rtl8192cu assigns a new struct rtl_tcb_desc object, but fails to
clear it.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Stable <stable@kernel.org> [2.6.39+]
Signed-off-by: John W. Linville <linville@tuxdriver.com>
|
|
If iwl_scan_initiate() fails for any reason,
priv->scan_request and priv->scan_vif are left
dangling. This can lead to a crash later when
iwl_bg_scan_completed() tries to run a pending
scan request.
In practice, this seems to be very rare due to
the STATUS_SCANNING check earlier. That check,
however, is wrong -- it should allow a scan to
be queued when a reset/roc scan is going on.
When a normal scan is already going on, a new
one can't be issued by mac80211, so that code
can be removed completely. I introduced this
bug when adding off-channel support in commit
266af4c745952e9bebf687dd68af58df553cb59d.
Cc: stable@kernel.org [3.0]
Reported-by: Peng Yan <peng.yan@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
|
|
Commit b7ab83e (PM: Use spinlock instead of mutex in clock
management functions) introduced a regression causing clocks_mutex
to be acquired under a spinlock. This happens because
pm_clk_suspend() and pm_clk_resume() call pm_clk_acquire() under
pcd->lock, but pm_clk_acquire() executes clk_get() which causes
clocks_mutex to be acquired. Similarly, __pm_clk_remove(),
executed under pcd->lock, calls clk_put(), which also causes
clocks_mutex to be acquired.
To fix those problems make pm_clk_add() call pm_clk_acquire(), so
that pm_clk_suspend() and pm_clk_resume() don't have to do that.
Change pm_clk_remove() and pm_clk_destroy() to separate
modifications of the pcd->clock_list list from the actual removal of
PM clock entry objects done by __pm_clk_remove().
Reported-and-tested-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Russell King <rmk+kernel@arm.linux.org.uk>
|
|
Following reports on the list, it looks like the 3e-9xxx driver will leak dma
mappings every time we get a transient queueing error back from the card.
This is because it maps the sg list in the routine that sends the command, but
doesn't unmap again in the transient failure path (even though the command is
sent back to the block layer). Fix by unmapping before returning the status.
Reported-by: Chris Boot <bootc@bootc.net>
Tested-by: Chris Boot <bootc@bootc.net>
Acked-by: Adam Radford <aradford@gmail.com>
Cc: stable@kernel.org
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
|
|
This oops was reported recently:
d:mon> e
cpu 0xd: Vector: 300 (Data Access) at [c0000000fd4c7120]
pc: d00000000076f194: .t3_l2t_get+0x44/0x524 [cxgb3]
lr: d000000000b02108: .init_act_open+0x150/0x3d4 [cxgb3i]
sp: c0000000fd4c73a0
msr: 8000000000009032
dar: 0
dsisr: 40000000
current = 0xc0000000fd640d40
paca = 0xc00000000054ff80
pid = 5085, comm = iscsid
d:mon> t
[c0000000fd4c7450] d000000000b02108 .init_act_open+0x150/0x3d4 [cxgb3i]
[c0000000fd4c7500] d000000000e45378 .cxgbi_ep_connect+0x784/0x8e8 [libcxgbi]
[c0000000fd4c7650] d000000000db33f0 .iscsi_if_rx+0x71c/0xb18
[scsi_transport_iscsi2]
[c0000000fd4c7740] c000000000370c9c .netlink_data_ready+0x40/0xa4
[c0000000fd4c77c0] c00000000036f010 .netlink_sendskb+0x4c/0x9c
[c0000000fd4c7850] c000000000370c18 .netlink_sendmsg+0x358/0x39c
[c0000000fd4c7950] c00000000033be24 .sock_sendmsg+0x114/0x1b8
[c0000000fd4c7b50] c00000000033d208 .sys_sendmsg+0x218/0x2ac
[c0000000fd4c7d70] c00000000033f55c .sys_socketcall+0x228/0x27c
[c0000000fd4c7e30] c0000000000086a4 syscall_exit+0x0/0x40
--- Exception: c01 (System Call) at 00000080da560cfc
The root cause was an EEH error, which sent us down the offload_close path in
the cxgb3 driver, which in turn sets cdev->l2opt to NULL, without regard for
upper layer driver (like the cxgbi drivers) which might have execution contexts
in the middle of its use. The result is the oops above, when t3_l2t_get attempts
to dereference L2DATA(cdev)->nentries in arp_hash right after the EEH error handler sets it to NULL.
The fix is to prevent the setting of the NULL pointer until after there are no
further users of it. The t3cdev->l2opt pointer is now converted to be an rcu
pointer and the L2DATA macro is now called under the protection of the
rcu_read_lock(). When the EEH error path:
t3_adapter_error->offload_close->cxgb3_offload_deactivate
Is exectured, setting of that l2opt pointer to NULL, is now gated on an rcu
quiescence point, preventing, allowing L2DATA callers to safely check for a NULL
pointer without concern that the underlying data will be freeded before the
pointer is dereferenced.
This has been tested by the reporter and shown to fix the reproted oops
[nhorman: fix up unitinialised variable reported by Dan Carpenter]
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Karen Xie <kxie@chelsio.com>
Cc: stable@kernel.org
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
|
