The ext4_move_extents() functions checks with BUG_ON() whether the
exchanged blocks count accords with request blocks count. But, if the
target range (orig_start + len) includes sparse block(s), 'moved_len'
(exchanged blocks count) does not agree with 'len' (request blocks
count), since sparse block is not counted in 'moved_len'. This causes
us to hit the BUG_ON(), even though the function succeeded.
The mext_check_arguments() function in move_extents.c has wrong
comparisons. orig_start which is passed from user-space is block
unit, but i_size of inode is byte unit, therefore the checks do not
work fine. This mis-check leads to the overflow of 'len' and then
hits BUG_ON() in ext4_move_extents(). The patch fixes this issue.
We need to flush the write cache unconditionally in ->fsync, otherwise
writes into already allocated blocks can get lost. Writes into fully
allocated files are very common when using disk images for
virtualization, and without this fix can easily lose data after
an fdatasync, which is the typical implementation for a cache flush on
the virtual drive.
Signed-off-by: Christoph Hellwig <hch@lst.de> Acked-by: Eric Sandeen <sandeen@redhat.com> Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
To solve a lock inversion problem, we implement part of the
range_cyclic algorithm in ext4_da_writepages(). (See commit 2acf2c26
for more details.)
As part of that change wbc->range_start was modified by ext4's
writepages function, which causes its callers to get confused since
they aren't expecting the filesystem to modify it. The simplest fix
is to save and restore wbc->range_start in ext4_da_writepages.
In ext4_link we need to check using EXT4_LINK_MAX, and not
EXT4_DIR_LINK_MAX(), since ext4_link() is creating hard links of
regular files, and not directories.
We need to unlock the new inode before iput. This patch fixes the
following warning when calling chattr +e to migrate a file to use
extents. It also fixes problems in when e4defrag attempts to
defragment an inode.
A user reported that although his root ext4 filesystem was mounting
fine, other filesystems would not mount, with the:
"Filesystem with huge files cannot be mounted RDWR without CONFIG_LBDAF"
error on his 32-bit box built without CONFIG_LBDAF. This is because
the test at mount time for this situation was not being re-checked
on remount, and the normal boot process makes an ro->rw transition,
so this was being missed.
Refactor to make a common helper function to test the filesystem
features against the type of mount request (RO vs. RW) so that we
stay consistent.
Addresses Red-Hat-Bugzilla: #517650
Signed-off-by: Eric Sandeen <sandeen@redhat.com> Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
ext4 will happily mount a > 16T filesystem on a 32-bit box, but
this is not safe; writes to the block device will wrap past 16T
and the page cache can't index past 16T (232 index * 4k pages).
Adding another test to the existing "too many sectors" test
should do the trick.
Add a comment, a relevant return value, and fix the reference
to the CONFIG_LBD(AF) option as well.
Jan Kara [Tue, 18 Aug 2009 02:17:20 +0000 (22:17 -0400)]
ext4: Fix possible deadlock between ext4_truncate() and ext4_get_blocks()
During truncate we are sometimes forced to start a new transaction as
the amount of blocks to be journaled is both quite large and hard to
predict. So far we restarted a transaction while holding i_data_sem
and that violates lock ordering because i_data_sem ranks below a
transaction start (and it can lead to a real deadlock with
ext4_get_blocks() mapping blocks in some page while having a
transaction open).
We fix the problem by dropping the i_data_sem before restarting the
transaction and acquire it afterwards. It's slightly subtle that this
works:
1) By the time ext4_truncate() is called, all the page cache for the
truncated part of the file is dropped so get_block() should not be
called on it (we only have to invalidate extent cache after we
reacquire i_data_sem because some extent from not-truncated part could
extend also into the part we are going to truncate).
2) Writes, migrate or defrag hold i_mutex so they are stopped for all
the time of the truncate.
This bug has been found and analyzed by Theodore Tso <tytso@mit.edu>.
Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: "Theodore Ts'o" <tytso@mit.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
lockdep annotation for a transaction start has been at the end of
jbd2_journal_start(). But a transaction is also started from
jbd2_journal_restart(). Move the lockdep annotation to start_this_handle()
which covers both cases.
Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: "Theodore Ts'o" <tytso@mit.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Currently the group preallocation code tries to find a large (512)
free block from which to do per-cpu group allocation for small files.
The problem with this scheme is that it leaves the filesystem horribly
fragmented. In the worst case, if the filesystem is unmounted and
remounted (after a system shutdown, for example) we forget the fact
that wee were using a particular (now-partially filled) 512 block
extent. So the next time we try to allocate space for a small file,
we will find *another* completely free 512 block chunk to allocate
small files. Given that there are 32,768 blocks in a block group,
after 64 iterations of "mount, write one 4k file in a directory,
unmount", the block group will have 64 files, each separated by 511
blocks, and the block group will no longer have any free 512
completely free chunks of blocks for group preallocation space.
So if we try to allocate blocks for a file that has been closed, such
that we know the final size of the file, and the filesystem is not
busy, avoid using group preallocation.
The logic around sbi->s_mb_last_group and sbi->s_mb_last_start was all
screwed up. These fields were getting unconditionally all the time,
set even when stream allocation had not taken place, and if they were
being used when the file was smaller than s_mb_stream_request, which
is when the allocation should _not_ be doing stream allocation.
Fix this by determining whether or not we stream allocation should
take place once, in ext4_mb_group_or_file(), and setting a flag which
gets used in ext4_mb_regular_allocator() and ext4_mb_use_best_found().
This simplifies the code and assures that we are consistently using
(or not using) the stream allocation logic.
move_extent_par_page calls a_ops->write_begin() to increase journal
handler's reference count. However, if either mext_replace_branches()
or ext4_get_block fails, the increased reference count isn't
decreased. This will cause a later attempt to umount of the fs to hang
forever. The patch addresses the issue by calling ext4_journal_stop()
if page is not NULL (which means a_ops->write_end() isn't invoked).
fix jiffie rounding in jbd commit timer setup code. Rounding down
could cause the timer to be fired before the corresponding transaction
has expired. That transaction can stay not committed forever if no
new transaction is created or expicit sync/umount happens.
Signed-off-by: Alex Zhuravlev (Tomas) <alex.zhuravlev@sun.com> Signed-off-by: Andreas Dilger <adilger@sun.com> Signed-off-by: "Theodore Ts'o" <tytso@mit.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The allocation of the ext4_group_info array was moved to a new
function ext4_mb_add_group_info() in commit 5f21b0e6 so that online
resize would use a common (and correct) codepath. Unfortunately, the
call to the new ext4_mb_add_group_info() function was added without
removing the code which originally allocated the array. This caused a
memory leak each time an ext4 filesystem was mounted.
The fix is simple; remove the code that did the original allocation,
since it is no longer needed.
Keysyms stored in key_map[] are not simply K() values, but U(K()) values,
as can be seen in the KDSKBENT ioctl handler. The kernel-generated
braille keysyms thus need a U() call too.
Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Dmitry Torokhov <dtor@mail.ru> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Return temperature in milidegree instead of degree, as sysfs-api requires
the temperature in milidegree.
Signed-off-by: Peter Feuerer <peter@piie.net> Tested-by: Borislav Petkov <petkovbb@gmail.com> Cc: Andreas Mohr <andi@lisas.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Peter Feuerer <peter@piie.net> Cc: Andreas Mohr <andi@lisas.de> Cc: Borislav Petkov <petkovbb@googlemail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
A bug was seen on boards using a PLX 8518 switch device which advertises
AER on each of it's transparent bridges. The AER driver was loaded for
each bridge and this driver tried to access the AER source ID register
whenever an interrupt occured on the shared PCI INTX lines. The source
ID register does not exist on non root port PCIE device's which
advertise AER and trying to access this register causes a unsupported
request error on the bridge. Thus, when the next interrupt occurs,
another error is found and the non existent source ID register is
accessed again, and so it goes on.
The result is a spammed dmesg with unsupported request PCI express
errors on the bridge device that the AER driver is loaded against.
Many boards have a bug-free ns16550 compatible serial port, which we should
register as PORT_16550A. This introduces a new value "ns16550a" for the
compatible property of of_serial to let a firmware choose that model instead
of using the crippled PORT_16550 mode.
When the framebuffer driver does not publish detailed timing information
for the current video mode, the correct value for the pixclock field is
zero, not -1.
Since pixclock is actually unsigned, the value -1 would be interpreted
as 4294967295 picoseconds (i.e., about 4 milliseconds) by
register_framebuffer() and userspace programs.
This patch allows X.org's fbdev driver to work.
Signed-off-by: Clemens Ladisch <clemens@ladisch.de> Tested-by: Paulius Zaleckas <paulius.zaleckas@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
- Apply Borislav Petkov's patch (convert the fancmd[] array to a real
struct thus disambiguating command handling and making code more
readable.)
- Add BIOS product to BIOS table as AOA110 and AOA150 have different
register values
- Add force_product parameter to allow forcing different product
- fix linker warning caused by "acerhdf_drv" not being named
"acerhdf_driver"
Signed-off-by: Peter Feuerer <peter@piie.net> Cc: Andreas Mohr <andi@lisas.de> Acked-by: Borislav Petkov <petkovbb@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com> Cc: Adrian von Bidder <avbidder@fortytwo.ch> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Fix userspace_device list corruption. The corruption was caused by
clients not being removed when adapters with such clients were
themselves removed. Something like the following would trigger it
(assuming i2c-stub gets adapter number 3):
* Set correct xpd curve indices for high/low gain curves during
rfbuffer setup on RF5112B with both calibration curves available.
* Don't return zero min power when we have the same pcdac value
twice because it breaks interpolation. Instead return the right
x barrier as we do when we have equal power levels for 2 different
pcdac values.
Signed-off-by: Nick Kossifidis <mickflemm@gmail.com> Acked-by: Bob Copeland <me@bobcopeland.com> Signed-off-by: John W. Linville <linville@tuxdriver.com> Cc: Dan Williams <dcbw@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Before bringing up a sky2 interface up ethtool reports
"Link detected: yes". Do as ixgbe does and netif_carrier_off() on
probe().
Signed-off-by: Brandon Philips <bphilips@suse.de> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
b43 allocates a bouncebuffer, if the supplied TX skb is in an invalid
memory range for DMA.
However, this is broken in that it fails to copy over some metadata to the
new skb.
This patch fixes three problems:
* Failure to adjust the ieee80211_tx_info pointer to the new buffer.
This results in a kmemcheck warning.
* Failure to copy the skb cb, which contains ieee80211_tx_info, to the new skb.
This results in breakage of various TX-status postprocessing (Rate control).
* Failure to transfer the queue mapping.
This results in the wrong queue being stopped on saturation and can result in queue overflow.
Signed-off-by: Michael Buesch <mb@bu3sch.de> Tested-by: Christian Casteyde <casteyde.christian@free.fr> Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Commit v2.6.28-rc1~717^2~109^2~2 was slightly incomplete; not all
instances of par->match->family were changed to par->family.
References: http://bugzilla.netfilter.org/show_bug.cgi?id=610 Signed-off-by: Jan Engelhardt <jengelh@medozas.de> Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Vitezslav Samel discovered that since 2.6.30.4+ active FTP can not work
over NAT. The "cause" of the problem was a fix of unacknowledged data
detection with NAT (commit a3a9f79e361e864f0e9d75ebe2a0cb43d17c4272).
However, actually, that fix uncovered a long standing bug in TCP conntrack:
when NAT was enabled, we simply updated the max of the right edge of
the segments we have seen (td_end), by the offset NAT produced with
changing IP/port in the data. However, we did not update the other parameter
(td_maxend) which is affected by the NAT offset. Thus that could drift
away from the correct value and thus resulted breaking active FTP.
The patch below fixes the issue by *not* updating the conntrack parameters
from NAT, but instead taking into account the NAT offsets in conntrack in a
consistent way. (Updating from NAT would be more harder and expensive because
it'd need to re-calculate parameters we already calculated in conntrack.)
Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu> Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The comment says that limits are cached for 60 seconds but the code
actually caches them for only 2 seconds. Align the code on the
comment, as 60 seconds makes more sense.
Signed-off-by: Jean Delvare <khali@linux-fr.org> Acked-by: Hans de Goede <hdegoede@redhat.com> Cc: Jordan Crouse <jordan@cosmicpenguin.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The logic of temperature fault flags is wrong, it shows faults when
there are none and vice versa. Fix it.
I can't believe this has been broken since the driver was added, 8
months ago, basically breaking temp1 and temp3, and nobody ever
complained.
Signed-off-by: Jean Delvare <khali@linux-fr.org> Acked-by: Hans de Goede <hdegoede@redhat.com> Cc: Jordan Crouse <jordan@cosmicpenguin.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
commit 0762b8bde9729f10f8e6249809660ff2ec3ad735
(from 14 months ago) introduced a use-after-free bug which has just
recently started manifesting in my md testing.
I tried git bisect to find out what caused the bug to start
manifesting, and it could have been the recent change to
blk_unregister_queue (48c0d4d4c04) but the results were inconclusive.
This patch certainly fixes my symptoms and looks correct as the two
calls are now in the same order as elsewhere in that function.
Fix soft-lockup in hso.c which is triggered on SMP machine when
modem is removed while file descriptor(s) under /dev are still open:
old version called kref_put() too early which resulted in destroying
hso_serial and hso_device objects which were still used later on.
Signed-off-by: Antti Kaijanmäki <antti.kaijanmaki@nomovok.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: David S. Miller <davem@davemloft.net>
The loop in perf_ctx_adjust_freq checks the frequency of sampling
event counters, and adjusts the event interval and unthrottles the
event if required, and resets the interrupt count for the event.
However, at present it only looks at group leaders.
This means that a sampling event that is not a group leader will
eventually get throttled, once its interrupt count reaches
sysctl_perf_event_sample_rate/HZ --- and that is guaranteed to
happen, if the event is active for long enough, since the interrupt
count never gets reset. Once it is throttled it never gets
unthrottled, so it basically just stops working at that point.
This fixes it by making perf_ctx_adjust_freq use ctx->event_list
rather than ctx->group_list. The existing spin_lock/spin_unlock
around the loop makes it unnecessary to put rcu_read_lock/
rcu_read_unlock around the list_for_each_entry_rcu().
Reported-by: Mark W. Krentel <krentel@cs.rice.edu> Signed-off-by: Paul Mackerras <paulus@samba.org> Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <19157.26731.855609.165622@cargo.ozlabs.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
commit 4706b349f was a forward port of a fix that was needed
for SLES10. But in fact it is not needed in mainline because
the earlier commit dd00a99e7a fixes the same problem in a
better way.
Further, this commit introduces a bug in the way it interacts with
the automatic read-error-correction. If, after a read error is
successfully corrected, the same disk is chosen to re-read - the
re-read won't be attempted but an error will be returned instead.
After reverting that commit, there is the possibility that a
read error on a read-only array (where read errors cannot
be corrected as that requires a write) will repeatedly read the same
device and continue to get an error.
So in the "Array is readonly" case, fail the drive immediately on
a read error.
On the parisc architecture we face for each and every loaded kernel module
this kernel "badness warning":
sysfs: cannot create duplicate filename '/module/ac97_bus/sections/.text'
Badness at fs/sysfs/dir.c:487
Reason for that is, that on parisc all kernel modules do have multiple
.text sections due to the usage of the -ffunction-sections compiler flag
which is needed to reach all jump targets on this platform.
An objdump on such a kernel module gives:
Sections:
Idx Name Size VMA LMA File off Algn
0 .note.gnu.build-id 00000024000000000000000000000034 2**2
CONTENTS, ALLOC, LOAD, READONLY, DATA
1 .text 00000000000000000000000000000058 2**0
CONTENTS, ALLOC, LOAD, READONLY, CODE
2 .text.ac97_bus_match 0000001c000000000000000000000058 2**2
CONTENTS, ALLOC, LOAD, READONLY, CODE
3 .text 000000000000000000000000000000d4 2**0
CONTENTS, ALLOC, LOAD, READONLY, CODE
...
Since the .text sections are empty (size of 0 bytes) and won't be
loaded by the kernel module loader anyway, I don't see a reason
why such sections need to be listed under
/sys/module/<module_name>/sections/<section_name> either.
The attached patch does solve this issue by not exporting section
names which are empty.
This fixes bugzilla http://bugzilla.kernel.org/show_bug.cgi?id=14703
mmc_remove_host() will cause the mmc core to switch off the bus power by
eventually calling pxamci_set_ios(). This function uses the regulator or
the GPIO which have been freed already.
a quirky chipset needs periodic schedules to run for a minimum
time before they can be disabled again. This enforces the requirement
with a time stamp and a calculated delay
Signed-off-by: Oliver Neukum <oliver@neukum.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The use of urb->actual_length to update tx_outstanding_bytes
implicitly assumes that the number of bytes actually written is the
same as the number of bytes we tried to write. On error that
assumption is violated so just use transfer_buffer_length the number
of bytes we intended to write to the device.
If an error occurs we need to fall through and call
usb_serial_port_softint to wake up processes waiting in
tty_wait_until_sent.
Signed-off-by: Eric W. Biederman <ebiederm@aristanetworks.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
- fixed shared interrupt bug reported by Vadim Lobanov
- fixed possible warning oops on driver unload when connected
- prevent interrupt flood in PIO mode ("modprobe amd5536udc use_dma=0")
when using gadget ether
Signed-off-by: Thomas Dahlmann <dahlmann.thomas@arcor.de> Cc: Robert Richter <robert.richter@amd.com> Cc: David Brownell <david-b@pacbell.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The driver incorrectly cancels the mass-storage device CSW request
(which leads to device reset) due to giving back URB at the head of
endpoint's queue after sending each STALL handshake; stop doing that
and start checking for the queue being non-empty before stalling an
endpoint and disallowing stall in such case in musb_gadget_set_halt()
like the other gadget drivers do.
Moreover, the driver starts Rx request despite of the endpoint being
halted -- fix this by moving the SendStall bit check from musb_g_rx()
to rxstate(). And we also sometimes get into rxstate() with DMA still
active after clearing an endpoint's halt (not clear why), so bail out
in this case, similarly to what txstate() does...
While at it, also do the following changes :
- in musb_gadget_set_halt(), remove pointless Tx FIFO flushing (the
driver does not allow stalling with non-empty Tx FIFO anyway);
- in rxstate(), stop pointlessly zeroing the 'csr' variable;
- in musb_gadget_set_halt(), move the 'done' label to a more proper
place;
- in musb_g_rx(), eliminate the 'done' label completely...
Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This patch (as1304) fixes a regression in ehci-hcd. Evidently some
hubs don't handle Clear-TT-Buffer requests correctly, so we should
avoid sending them when they don't appear to be absolutely necessary.
The reported symptom is that output on a downstream audio device cuts
out because the hub stops relaying isochronous packets.
The patch prevents Clear-TT-Buffer requests from being sent following
a STALL handshake. In theory a STALL indicates either that the
downstream device sent a STALL or that no matching TT buffer could be
found. In either case, the transfer is completed and the TT buffer
does not remain busy, so it doesn't need to be cleared.
Also, the patch fixes a minor flaw in the code that actually sends the
Clear-TT-Buffer requests. Although the pipe direction isn't really
used for control transfers, it should be a Send rather than a Receive.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Reported-by: Javier Kohen <jkohen@users.sourceforge.net> CC: David Brownell <david-b@pacbell.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
"[CPUFREQ] cpumask: avoid playing with cpus_allowed in speedstep-ich.c"
changed the code to mistakenly pass the current cpu as the "processor"
argument of speedstep_get_frequency(), whereas it should be the type of
the processor.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Acked-by: Dominik Brodowski <linux@brodo.de> Reported-by: Dave Mueller <dave.mueller@gmx.ch> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Dave Jones <davej@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
ipv4 ip_frag_reasm(), fully replace 'dev_net(dev)' with 'net', defined
previously patched into 2.6.29.
Between 2.6.28.10 and 2.6.29, net/ipv4/ip_fragment.c was patched,
changing from dev_net(dev) to container_of(...). Unfortunately the goto
section (out_fail) on oversized packets inside ip_frag_reasm() didn't
get touched up as well. Oversized IP packets cause a NULL pointer
dereference and immediate hang.
I discovered this running openvasd and my previous email on this is
titled: NULL pointer dereference at 2.6.32-rc8:net/ipv4/ip_fragment.c:566
Signed-off-by: David Ford <david@blue-labs.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The Hauppauge WinTV HVR-1150 retail boards require the FORCE_TS_VALID bit
to be set in order to function properly. This change will work on the early
revisions on the board as well, but the final revision will not function
without this change.
When FORCE_TS_VALID mode is enabled, the saa713x will accept MPEG TS input
without requiring TS_VALID set high. This is required for some new boards
to function properly, due to the hardware design implementation.
The configuration is toggled within the board setup configuration. Boards
that do not have this bit set will function as before with no change.
> We allocate and zero cpu_isolated_map after the isolcpus
> __setup option has run. This means cpu_isolated_map always
> ends up empty and if CPUMASK_OFFSTACK is enabled we write to a
> cpumask that hasn't been allocated.
I introduced this regression in 49557e620339cb13 (sched: Fix
boot crash by zalloc()ing most of the cpu masks).
Use the bootmem allocator if they set isolcpus=, otherwise
allocate and zero like normal.
Reported-by: Anton Blanchard <anton@samba.org> Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Cc: peterz@infradead.org Cc: Linus Torvalds <torvalds@linux-foundation.org>
LKML-Reference: <200912021409.17013.rusty@rustcorp.com.au> Signed-off-by: Ingo Molnar <mingo@elte.hu> Tested-by: Anton Blanchard <anton@samba.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
`>>' has a higher precedence than `?' so src2 evaluated to
either 16 or 0 dependent on the bits set in rs2.
Signed-off-by: Roel Kluin <roel.kluin@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Page buffers containing packets with an incorrect checksum or using a
protocol not handled by hardware checksum offload were previously not
passed to LRO. The conversion to GRO changed this, but did not set
the ip_summed value accordingly.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Modify bonding hash transmit policies to use the psource MAC address of
the packet instead of the MAC address configured for the bonding device.
The old sitation conflicts with the documentation.
Signed-off-by: Jasper Spaans <spaans@fox-it.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: Jay Vosburgh <fubar@us.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
On UDP sockets, we must call skb_free_datagram() with socket locked,
or risk sk_forward_alloc corruption. This requirement is not respected
in SUNRPC.
Add a convenient helper, skb_free_datagram_locked() and use it in SUNRPC
Reported-by: Francis Moreau <francis.moro@gmail.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Pedit will reject a policy that is large because it
uses the wrong structure in the policy validation.
This fixes it.
Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
If ace_load_firmware() fails, ace_init() cleans up but still returns
0, leading to an oops as seen in <http://bugs.debian.org/521383>.
It should pass the error code up.
Signed-off-by: Ben Hutchings <ben@decadent.org.uk> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
udp_poll() can in some circumstances drop frames with incorrect checksums.
Problem is we now have to lock the socket while dropping frames, or risk
sk_forward corruption.
This bug is present since commit 95766fff6b9a78d1
([UDP]: Add memory accounting.)
While we are at it, we can correct ioctl(SIOCINQ) to also drop bad frames.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The VIDIOC_G_STD ioctl may not be present in the case of radio receivers.
In that case G_STD will return an error. The v4l1-compat layer should not
attempt to propagate that error to the caller, instead it should be
ignored.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com> Signed-off-by: Michael Krufky <mkrufky@linuxtv.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
In commit 0de51088e6a82bc8413d3ca9e28bbca2788b5b53, we introduced the
use of acpi-cpufreq on VIA/Centaur CPU's by removing a vendor check for
VENDOR_INTEL. However, as it turns out, at least the Nano CPU's also
need the PDC (processor driver capabilities) handshake in order to
activate the methods required for acpi-cpufreq.
Since arch_acpi_processor_init_pdc() contains another vendor check for
Intel, the PDC is not initialized on VIA CPU's. The resulting behavior
of a current mainline kernel on such systems is: acpi-cpufreq
loads and it indicates CPU frequency changes. However, the CPU stays at
a single frequency
This trivial patch ensures that init_intel_pdc() is called on Intel and
VIA/Centaur CPU's alike.
Signed-off-by: Harald Welte <HaraldWelte@viatech.com> Signed-off-by: Dave Jones <davej@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
As shown in http://thread.gmane.org/gmane.linux.kernel.wireless.general/36497,
mac80211 has a bug that allows a call to the TX routine after the queues have
been stopped. This situation will only occur under extreme stress. Although
b43 does not crash when this condition occurs, it does generate a WARN_ON and
also logs a queue overrun message. This patch recognizes b43 is not at fault
and logs a message only when the most verbose debugging mode is enabled. In
the unlikely event that the queue is not stopped when the DMA queue becomes
full, then a warning is issued.
During testing of this patch with one output stream running repeated tcpperf
writes and a second running a flood ping, this routine was entered with
the DMA ring stopped about once per hour. The condition where the DMA queue is
full but the ring has not been stopped has never been seen by me.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net> Signed-off-by: John W. Linville <linville@tuxdriver.com> Cc: Michael Buesch <mb@bu3sch.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Lennert Buytenhek noticed that delBA handling in mac80211
was broken and has remotely triggerable problems, some of
which are due to some code shuffling I did that ended up
changing the order in which things were done -- this was
mac80211: A-MPDU Tx add delBA from recipient support
The first problem is that I moved a BUG_ON before various
checks -- thereby making it possible to hit. As the comment
indicates, the BUG_ON can be removed since the ampdu_action
callback must already exist when the state is != IDLE.
The second problem isn't easily exploitable but there's a
race condition due to unconditionally setting the state to
OPERATIONAL when a delBA frame is received, even when no
aggregation session was ever initiated. All the drivers
accept stopping the session even then, but that opens a
race window where crashes could happen before the driver
accepts it. Right now, a WARN_ON may happen with non-HT
drivers, while the race opens only for HT drivers.
For this case, there are two things necessary to fix it:
1) don't process spurious delBA frames, and be more careful
about the session state; don't drop the lock
2) HT drivers need to be prepared to handle a session stop
even before the session was really started -- this is
true for all drivers (that support aggregation) but
iwlwifi which can be fixed easily. The other HT drivers
(ath9k and ar9170) are behaving properly already.
Reported-by: Lennert Buytenhek <buytenh@marvell.com> Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Lennert Buytenhek noticed a remotely triggerable problem
in mac80211, which is due to some code shuffling I did
that ended up changing the order in which things were
done -- this was in
When we get a stream suspend event force the power down since otherwise
the stream would remain marked as active. In future we'll probably want
to make this stream-specific and add an interface to make the power down
of other widgets optional in order to support leaving bypass paths
active while suspending the processor.
Reported-by: Joonyoung Shim <jy0922.shim@samsung.com> Tested-by: Joonyoung Shim <jy0922.shim@samsung.com> Acked-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
also holds for fuse_create, however, the same kind of check was missing there.
As an impact of this bug, open(newfile, O_RDWR|O_CREAT|O_DIRECT) fails, but a
stub newfile will remain if the fuse server handled the implied FUSE_CREATE
request appropriately.
Other impact: in the above situation ima_file_free() will complain to open/free
imbalance if CONFIG_IMA is set.
Changeset a65318bf3afc93ce49227e849d213799b072c5fd (NFSv4: Simplify some
cache consistency post-op GETATTRs) incorrectly changed the getattr
bitmap for readdir().
This causes the readdir() function to fail to return a
fileid/inode number, which again exposed a bug in the NFS readdir code that
causes spurious ENOENT errors to appear in applications (see
http://bugzilla.kernel.org/show_bug.cgi?id=14541).
The immediate band aid is to revert the incorrect bitmap change, but more
long term, we should change the NFS readdir code to cope with the
fact that NFSv4 servers are not required to support fileids/inode numbers.
While running fsstress tests on the NFSv4 mounted ext3 and ext4
filesystem, the following call trace was generated on the nfs
server machine.
Replace GFP_KERNEL with GFP_NOFS in ima_iint_insert() to avoid a
potential deadlock.
=================================
[ INFO: inconsistent lock state ]
2.6.31-31.el6.x86_64 #1
---------------------------------
inconsistent {RECLAIM_FS-ON-W} -> {IN-RECLAIM_FS-W} usage.
kswapd2/75 [HC0[0]:SC0[0]:HE1:SE1] takes:
(jbd2_handle){+.+.?.}, at: [<ffffffff811edd5e>] jbd2_journal_start+0xfe/0x13f
{RECLAIM_FS-ON-W} state was registered at:
[<ffffffff81091e40>] mark_held_locks+0x65/0x99
[<ffffffff81091f31>] lockdep_trace_alloc+0xbd/0xf5
[<ffffffff81126fdd>] kmem_cache_alloc+0x40/0x185
[<ffffffff812344d7>] ima_iint_insert+0x3d/0xf1
[<ffffffff812345b0>] ima_inode_alloc+0x25/0x44
[<ffffffff811484ac>] inode_init_always+0xec/0x271
[<ffffffff81148682>] alloc_inode+0x51/0xa1
[<ffffffff81148700>] new_inode+0x2e/0x94
[<ffffffff811b2f08>] ext4_new_inode+0xb8/0xdc9
[<ffffffff811be611>] ext4_create+0xcf/0x175
[<ffffffff8113e2cd>] vfs_create+0x82/0xb8
[<ffffffff8113f337>] do_filp_open+0x32c/0x9ee
[<ffffffff811309b9>] do_sys_open+0x6c/0x12c
[<ffffffff81130adc>] sys_open+0x2e/0x44
[<ffffffff81011e42>] system_call_fastpath+0x16/0x1b
[<ffffffffffffffff>] 0xffffffffffffffff
irq event stamp: 90371
hardirqs last enabled at (90371): [<ffffffff8112708d>]
kmem_cache_alloc+0xf0/0x185
hardirqs last disabled at (90370): [<ffffffff81127026>]
kmem_cache_alloc+0x89/0x185
softirqs last enabled at (89492): [<ffffffff81068ecf>]
__do_softirq+0x1bf/0x1eb
softirqs last disabled at (89477): [<ffffffff8101312c>] call_softirq+0x1c/0x30
other info that might help us debug this:
2 locks held by kswapd2/75:
#0: (shrinker_rwsem){++++..}, at: [<ffffffff810f98ba>] shrink_slab+0x44/0x177
#1: (&type->s_umount_key#25){++++..}, at: [<ffffffff811450ba>]
Reported-by: Muni P. Beerakam <mbeeraka@in.ibm.com> Reported-by: Amit K. Arora <amitarora@in.ibm.com> Signed-off-by: Mimi Zohar <zohar@us.ibm.com> Signed-off-by: James Morris <jmorris@namei.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
When 802.11g was introduced, we had RTS/CTS and CTS-to-Self protection
mechanisms. In an HT Beacon, HT stations use the "Operating Mode" field
in the HT Information Element to determine whether or not to use
protection.
The Operating Mode field has 4 possible settings: 0-3:
Mode 0: If all stations in the BSS are 20/40 MHz HT capable, or if the
BSS is 20/40 MHz capable, or if all stations in the BSS are 20 MHz HT
stations in a 20 MHz BSS
Mode 1: used if there are non-HT stations or APs using the primary or
secondary channels
Mode 2: if only HT stations are associated in the BSS and at least one
20 MHz HT station is associated.
Mode 3: used if one or more non-HT stations are associated in the BSS.
When in operating modes 1 or 3, and the Use_Protection field is 1 in the
Beacon's ERP IE, all HT transmissions must be protected using RTS/CTS or
CTS-to-Self.
By default, CTS-to-self is the preferred protection mechanism for less
overhead and higher throughput; but using the full RTS/CTS will better
protect the inner exchange from interference, especially in
highly-congested environment.
For 6000 series WIFI NIC, RTS/CTS protection mechanism is the
recommended choice for HT traffic based on the HW design.
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The scratchpad_free() function uses xhci->page_size to free some memory
with pci_free_consistent(). However, the page_size is set to zero before
the call, causing kernel oopses on driver unload. Call scratchpad_free()
before setting xhci->page_size to zero.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com> Acked-by: John Youn <John.Youn@synopsys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The trb_in_td() function in the xHCI driver is supposed to translate a
physical transfer buffer request (TRB) into a virtual pointer to the ring
segment that TRB is in.
Unfortunately, a mistake in this function may cause endless loops as the
driver searches through the linked list of ring segments over and over
again. Fix a couple bugs that may lead to loops or bad output:
1. Bail out if we get a NULL pointer when translating the segment's
private structure and the starting DMA address of the segment chunk. If
this happens, we've been handed a starting TRB pointer from a different
ring.
2. Make sure the function works when there's multiple segments in the
ring. In the while loop to search through the ring segments, use the
current segment variable (cur_seg), rather than the starting segment
variable (start_seg) that is passed in.
3. Stop searching the ring if we've run through all the segments in the
ring.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
If the xHCI driver fails during the memory initialization, xhci->ir_set
may not be a valid pointer. Check that it points to valid DMA'able memory
before writing to that address during the memory freeing process.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Add D-Link DWM-162-U5 device id 1e0e:ce16 into option driver. The device
has 4 interfaces, of which 1 is handled by storage and the other 3 by
option driver.
The device appears first as CD-only 05c6:2100 device and must be
switched to 1e0e:ce16 mode either by using "eject CD" or usb_modeswitch.
This patch (as1299b) fixes a bug in an error-handling path of usbmon's
binary interface. The storage area for URB data is divided into
fixed-size blocks. If an URB's data can't be copied, the area
reserved for it should be decreased to the size of the truncated
information (rounded up to a block boundary). Rounding up the amount
to be removed and subtracting it from the reserved size is definitely
the wrong thing to do.
Also, when the data for an isochronous URB can't be copied, we can
still copy the isoc packet descriptors. In fact the current code does
copy the descriptors, but then sets the capture length to 0 so they
remain inaccessible. The capture length should be reduced to the
length of the descriptors, not set to 0.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Acked-by: Pete Zaitcev <zaitcev@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
projects / karo-tx-linux.git / log