Linus Torvalds [Sat, 15 Jun 2013 05:25:04 +0000 (19:25 -1000)]
Merge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
Pull powerpc fixes from Benjamin Herrenschmidt:
"So here are 3 fixes still for 3.10. Fixes are simple, bugs are nasty
(though not recent regressions, nasty enough) and all targeted at
stable"
* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:
powerpc: Fix missing/delayed calls to irq_work
powerpc: Fix emulation of illegal instructions on PowerNV platform
powerpc: Fix stack overflow crash in resume_kernel when ftracing
David Daney [Fri, 14 Jun 2013 18:13:59 +0000 (11:13 -0700)]
smp.h: Use local_irq_{save,restore}() in !SMP version of on_each_cpu().
Thanks to commit f91eb62f71b3 ("init: scream bloody murder if interrupts
are enabled too early"), "bloody murder" is now being screamed.
With a MIPS OCTEON config, we use on_each_cpu() in our
irq_chip.irq_bus_sync_unlock() function. This gets called in early as a
result of the time_init() call. Because the !SMP version of
on_each_cpu() unconditionally enables irqs, we get:
WARNING: at init/main.c:560 start_kernel+0x250/0x410()
Interrupts were enabled early
CPU: 0 PID: 0 Comm: swapper Not tainted 3.10.0-rc5-Cavium-Octeon+ #801
Call Trace:
show_stack+0x68/0x80
warn_slowpath_common+0x78/0xb0
warn_slowpath_fmt+0x38/0x48
start_kernel+0x250/0x410
Suggested fix: Do what we already do in the SMP version of
on_each_cpu(), and use local_irq_save/local_irq_restore. Because we
need a flags variable, make it a static inline to avoid name space
issues.
[ Change from v1: Convert on_each_cpu to a static inline function, add
#include <linux/irqflags.h> to avoid build breakage on some files.
on_each_cpu_mask() and on_each_cpu_cond() suffer the same problem as
on_each_cpu(), but they are not causing !SMP bugs for me, so I will
defer changing them to a less urgent patch. ]
Linus Torvalds [Sat, 15 Jun 2013 05:18:56 +0000 (19:18 -1000)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull VFS fixes from Al Viro:
"Several fixes + obvious cleanup (you've missed a couple of open-coded
can_lookup() back then)"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
snd_pcm_link(): fix a leak...
use can_lookup() instead of direct checks of ->i_op->lookup
move exit_task_namespaces() outside of exit_notify()
fput: task_work_add() can fail if the caller has passed exit_task_work()
ncpfs: fix rmdir returns Device or resource busy
Linus Torvalds [Sat, 15 Jun 2013 05:16:31 +0000 (19:16 -1000)]
Merge tag 'for-linus-v3.10-rc6' of git://oss.sgi.com/xfs/xfs
Pull xfs fixes from Ben Myers:
- Remove noisy warnings about experimental support which spams the logs
- Add padding to align directory and attr structures correctly
- Set block number on child buffer on a root btree split
- Disable verifiers during log recovery for non-CRC filesystems
* tag 'for-linus-v3.10-rc6' of git://oss.sgi.com/xfs/xfs:
xfs: don't shutdown log recovery on validation errors
xfs: ensure btree root split sets blkno correctly
xfs: fix implicit padding in directory and attr CRC formats
xfs: don't emit v5 superblock warnings on write
Linus Torvalds [Sat, 15 Jun 2013 05:15:36 +0000 (19:15 -1000)]
Merge tag 'char-misc-3.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc
Pull char / misc fixes from Greg Kroah-Hartman:
"Here are some small mei driver fixes for 3.10-rc6 that fix some
reported problems"
* tag 'char-misc-3.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
mei: me: clear interrupts on the resume path
mei: nfc: fix nfc device freeing
mei: init: Flush scheduled work before resetting the device
Linus Torvalds [Sat, 15 Jun 2013 05:14:39 +0000 (19:14 -1000)]
Merge tag 'usb-3.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Pull USB fixes from Greg Kroah-Hartman:
"Here are some small USB driver fixes that resolve some reported
problems for 3.10-rc6
Nothing major, just 3 USB serial driver fixes, and two chipidea fixes"
* tag 'usb-3.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
usb: chipidea: fix id change handling
usb: chipidea: fix no transceiver case
USB: pl2303: fix device initialisation at open
USB: spcp8x5: fix device initialisation at open
USB: f81232: fix device initialisation at open
When replaying interrupts (as a result of the interrupt occurring
while soft-disabled), in the case of the decrementer, we are exclusively
testing for a pending timer target. However we also use decrementer
interrupts to trigger the new "irq_work", which in this case would
be missed.
This change the logic to force a replay in both cases of a timer
boundary reached and a decrementer interrupt having actually occurred
while disabled. The former test is still useful to catch cases where
a CPU having been hard-disabled for a long time completely misses the
interrupt due to a decrementer rollover.
Paul Mackerras [Fri, 14 Jun 2013 10:07:41 +0000 (20:07 +1000)]
powerpc: Fix emulation of illegal instructions on PowerNV platform
Normally, the kernel emulates a few instructions that are unimplemented
on some processors (e.g. the old dcba instruction), or privileged (e.g.
mfpvr). The emulation of unimplemented instructions is currently not
working on the PowerNV platform. The reason is that on these machines,
unimplemented and illegal instructions cause a hypervisor emulation
assist interrupt, rather than a program interrupt as on older CPUs.
Our vector for the emulation assist interrupt just calls
program_check_exception() directly, without setting the bit in SRR1
that indicates an illegal instruction interrupt. This fixes it by
making the emulation assist interrupt set that bit before calling
program_check_interrupt(). With this, old programs that use no-longer
implemented instructions such as dcba now work again.
If we look at current's stack (paca->__current->stack) we see it is
equal to c0000002ecab0000. Our stack is 16K, and comparing to
paca->kstack (c0000002ecab3e30) we can see that we have overflowed our
kernel stack. This leads to us writing over our struct thread_info, and
in this case we have corrupted thread_info->flags and set
_TIF_EMULATE_STACK_STORE.
__ppc64_runlatch_on() is called from RUNLATCH_ON in the exception entry
path. At that point the irq state is not consistent, ie. interrupts are
hard disabled (by the exception entry), but the paca soft-enabled flag
may be out of sync.
This leads to the local_irq_restore() in trace_graph_entry() actually
enabling interrupts, which we do not want. Because we have not yet
reprogrammed the decrementer we immediately take another decrementer
exception, and recurse.
The fix is twofold. Firstly make sure we call DISABLE_INTS before
calling RUNLATCH_ON. The badly named DISABLE_INTS actually reconciles
the irq state in the paca with the hardware, making it safe again to
call local_irq_save/restore().
Although that should be sufficient to fix the bug, we also mark the
runlatch routines as notrace. They are called very early in the
exception entry and we are asking for trouble tracing them. They are
also fairly uninteresting and tracing them just adds unnecessary
overhead.
Oleg Nesterov [Fri, 14 Jun 2013 19:09:49 +0000 (21:09 +0200)]
move exit_task_namespaces() outside of exit_notify()
exit_notify() does exit_task_namespaces() after
forget_original_parent(). This was needed to ensure that ->nsproxy
can't be cleared prematurely, an exiting child we are going to
reparent can do do_notify_parent() and use the parent's (ours) pid_ns.
However, after 32084504 "pidns: use task_active_pid_ns in
do_notify_parent" ->nsproxy != NULL is no longer needed, we rely
on task_active_pid_ns().
Move exit_task_namespaces() from exit_notify() to do_exit(), after
exit_fs() and before exit_task_work().
This solves the problem reported by Andrey, free_ipc_ns()->shm_destroy()
does fput() which needs task_work_add().
Note: this particular problem can be fixed if we change fput(), and
that change makes sense anyway. But there is another reason to move
the callsite. The original reason for exit_task_namespaces() from
the middle of exit_notify() was subtle and it has already gone away,
now this looks confusing. And this allows us do simplify exit_notify(),
we can avoid unlock/lock(tasklist) and we can use ->exit_state instead
of PF_EXITING in forget_original_parent().
Oleg Nesterov [Fri, 14 Jun 2013 19:09:47 +0000 (21:09 +0200)]
fput: task_work_add() can fail if the caller has passed exit_task_work()
fput() assumes that it can't be called after exit_task_work() but
this is not true, for example free_ipc_ns()->shm_destroy() can do
this. In this case fput() silently leaks the file.
Change it to fallback to delayed_fput_work if task_work_add() fails.
The patch looks complicated but it is not, it changes the code from
if (PF_KTHREAD) {
schedule_work(...);
return;
}
task_work_add(...)
to
if (!PF_KTHREAD) {
if (!task_work_add(...))
return;
/* fallback */
}
schedule_work(...);
As for shm_destroy() in particular, we could make another fix but I
think this change makes sense anyway. There could be another similar
user, it is not safe to assume that task_work_add() can't fail.
Daniel Borkmann [Fri, 14 Jun 2013 16:24:07 +0000 (18:24 +0200)]
net: sctp: sctp_association_init: put refs in reverse order
In case we need to bail out for whatever reason during assoc
init, we call sctp_endpoint_put() and then sock_put(), however,
we've hold both refs in reverse, non-symmetric order, so first
sctp_endpoint_hold() and then sock_hold().
Reverse this, so that in an error case we have sock_put() and then
sctp_endpoint_put(). Actually shouldn't matter too much, since both
cleanup paths do the right thing, but that way, it is more consistent
with the rest of the code.
Daniel Borkmann [Fri, 14 Jun 2013 16:24:06 +0000 (18:24 +0200)]
net: sctp: minor: remove variable in sctp_init_sock
It's only used at this one time, so we could remove it as well.
This is valid and also makes it more explicit/obvious that in case
of error the sp->ep is NULL here, i.e. for the sctp_destroy_sock()
check that was recently added.
Daniel Borkmann [Fri, 14 Jun 2013 16:24:05 +0000 (18:24 +0200)]
net: sctp: sctp_sf_do_prm_asoc: do SCTP_CMD_INIT_CHOOSE_TRANSPORT first
While this currently cannot trigger any NULL pointer dereference in
sctp_seq_dump_local_addrs(), better change the order of commands to
prevent a future bug to happen. Although we first add SCTP_CMD_NEW_ASOC
and then set the SCTP_CMD_INIT_CHOOSE_TRANSPORT, it is okay for now,
since this primitive is only called by sctp_connect() or sctp_sendmsg()
with sctp_assoc_add_peer() set first. However, lets do this precaution
and first set the transport and then add it to the association hashlist
to prevent in future something to possibly triggering this.
Daniel Borkmann [Fri, 14 Jun 2013 16:24:04 +0000 (18:24 +0200)]
net: sctp: sideeffect: throw BUG if primary_path is NULL
This clearly states a BUG somewhere in the SCTP code as e.g. fixed once
in f28156335 ("sctp: Use correct sideffect command in duplicate cookie
handling"). If this ever happens, throw a trace in the sideeffect engine
where assocs clearly must have a primary_path assigned.
When in sctp_seq_dump_local_addrs() also throw a WARN and bail out since
we do not need to panic for printing this one asterisk. Also, it will
avoid the not so obvious case when primary != NULL test passes and at a
later point in time triggering a NULL ptr dereference caused by primary.
While at it, also fix up the white space.
David S. Miller [Fri, 14 Jun 2013 22:31:22 +0000 (15:31 -0700)]
Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/jesse/openvswitch
Jesse Gross says:
====================
A few miscellaneous improvements and cleanups before the GRE tunnel
integration series. Intended for net-next/3.11.
====================
Jesse Gross [Mon, 13 May 2013 15:15:26 +0000 (08:15 -0700)]
openvswitch: Immediately exit on error in ovs_vport_cmd_set().
It is an error to try to change the type of a vport using the set
command. However, while we check that this is an error, we still
proceed to allocate memory which then gets freed immediately.
This stops processing after noticing the error, which does not
actually fix a bug but is more correct.
Dave Chinner [Wed, 12 Jun 2013 02:19:06 +0000 (12:19 +1000)]
xfs: don't shutdown log recovery on validation errors
Unfortunately, we cannot guarantee that items logged multiple times
and replayed by log recovery do not take objects back in time. When
they are taken back in time, the go into an intermediate state which
is corrupt, and hence verification that occurs on this intermediate
state causes log recovery to abort with a corruption shutdown.
Instead of causing a shutdown and unmountable filesystem, don't
verify post-recovery items before they are written to disk. This is
less than optimal, but there is no way to detect this issue for
non-CRC filesystems If log recovery successfully completes, this
will be undone and the object will be consistent by subsequent
transactions that are replayed, so in most cases we don't need to
take drastic action.
For CRC enabled filesystems, leave the verifiers in place - we need
to call them to recalculate the CRCs on the objects anyway. This
recovery problem can be solved for such filesystems - we have a LSN
stamped in all metadata at writeback time that we can to determine
whether the item should be replayed or not. This is a separate piece
of work, so is not addressed by this patch.
Dave Chinner [Wed, 12 Jun 2013 02:19:08 +0000 (12:19 +1000)]
xfs: ensure btree root split sets blkno correctly
For CRC enabled filesystems, the BMBT is rooted in an inode, so it
passes through a different code path on root splits than the
freespace and inode btrees. This is much less traversed by xfstests
than the other trees. When testing on a 1k block size filesystem,
I've been seeing ASSERT failures in generic/234 like:
which are generally preceded by a lblock check failure. I noticed
this in the bmbt stats:
$ pminfo -f xfs.btree.block_map
xfs.btree.block_map.lookup
value 39135
xfs.btree.block_map.compare
value 268432
xfs.btree.block_map.insrec
value 15786
xfs.btree.block_map.delrec
value 13884
xfs.btree.block_map.newroot
value 2
xfs.btree.block_map.killroot
value 0
.....
Very little coverage of root splits and merges. Indeed, on a 4k
filesystem, block_map.newroot and block_map.killroot are both zero.
i.e. the code is not exercised at all, and it's the only generic
btree infrastructure operation that is not exercised by a default run
of xfstests.
Turns out that on a 1k filesystem, generic/234 accounts for one of
those two root splits, and that is somewhat of a smoking gun. In
fact, it's the same problem we saw in the directory/attr code where
headers are memcpy()d from one block to another without updating the
self describing metadata.
Simple fix - when copying the header out of the root block, make
sure the block number is updated correctly.
Dave Chinner [Wed, 12 Jun 2013 02:19:07 +0000 (12:19 +1000)]
xfs: fix implicit padding in directory and attr CRC formats
Michael L. Semon has been testing CRC patches on a 32 bit system and
been seeing assert failures in the directory code from xfs/080.
Thanks to Michael's heroic efforts with printk debugging, we found
that the problem was that the last free space being left in the
directory structure was too small to fit a unused tag structure and
it was being corrupted and attempting to log a region out of bounds.
Hence the assert failure looked something like:
.....
#5 calling xfs_dir2_data_log_unused() 36 32
#1 4092 4095 4096
#2 8182 8183 4096
XFS: Assertion failed: first <= last && last < BBTOB(bp->b_length), file: fs/xfs/xfs_trans_buf.c, line: 568
Where #1 showed the first region of the dup being logged (i.e. the
last 4 bytes of a directory buffer) and #2 shows the corrupt values
being calculated from the length of the dup entry which overflowed
the size of the buffer.
It turns out that the problem was not in the logging code, nor in
the freespace handling code. It is an initial condition bug that
only shows up on 32 bit systems. When a new buffer is initialised,
where's the freespace that is set up:
Note the offset of the first region being logged? It's 60 bytes into
the buffer. Once I saw that, I pretty much knew that the bug was
going to be caused by this.
Essentially, all direct entries are rounded to 8 bytes in length,
and all entries start with an 8 byte alignment. This means that we
can decode inplace as variables are naturally aligned. With the
directory data supposedly starting on a 8 byte boundary, and all
entries padded to 8 bytes, the minimum freespace in a directory
block is supposed to be 8 bytes, which is large enough to fit a
unused data entry structure (6 bytes in size). The fact we only have
4 bytes of free space indicates a directory data block alignment
problem.
And what do you know - there's an implicit hole in the directory
data block header for the CRC format, which means the header is 60
byte on 32 bit intel systems and 64 bytes on 64 bit systems. Needs
padding. And while looking at the structures, I found the same
problem in the attr leaf header. Fix them both.
Note that this only affects 32 bit systems with CRCs enabled.
Everything else is just fine. Note that CRC enabled filesystems created
before this fix on such systems will not be readable with this fix
applied.
Dave Chinner [Mon, 27 May 2013 06:38:19 +0000 (16:38 +1000)]
xfs: don't emit v5 superblock warnings on write
We write the superblock every 30s or so which results in the
verifier being called. Right now that results in this output
every 30s:
XFS (vda): Version 5 superblock detected. This kernel has EXPERIMENTAL support enabled!
Use of these features in this kernel is at your own risk!
And spamming the logs.
We don't need to check for whether we support v5 superblocks or
whether there are feature bits we don't support set as these are
only relevant when we first mount the filesytem. i.e. on superblock
read. Hence for the write verification we can just skip all the
checks (and hence verbose output) altogether.
Linus Torvalds [Fri, 14 Jun 2013 05:34:14 +0000 (22:34 -0700)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs
Pull btrfs fixes from Chris Mason:
"This is an assortment of crash fixes"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
Btrfs: stop all workers before cleaning up roots
Btrfs: fix use-after-free bug during umount
Btrfs: init relocate extent_io_tree with a mapping
btrfs: Drop inode if inode root is NULL
Btrfs: don't delete fs_roots until after we cleanup the transaction
Samuel Ortiz [Mon, 10 Jun 2013 07:10:25 +0000 (10:10 +0300)]
mei: init: Flush scheduled work before resetting the device
Flushing pending work items before resetting the device makes more
sense than doing so afterwards. Some of them, like e.g. the NFC
initialization one, find themselves with client IDs changed after
the reset, eventually leading to trigger a client.c:mei_me_cl_by_id()
warning after a few modprobe/rmmod cycles.
Neil Horman [Wed, 12 Jun 2013 18:26:44 +0000 (14:26 -0400)]
sctp: fully initialize sctp_outq in sctp_outq_init
In commit 2f94aabd9f6c925d77aecb3ff020f1cc12ed8f86
(refactor sctp_outq_teardown to insure proper re-initalization)
we modified sctp_outq_teardown to use sctp_outq_init to fully re-initalize the
outq structure. Steve West recently asked me why I removed the q->error = 0
initalization from sctp_outq_teardown. I did so because I was operating under
the impression that sctp_outq_init would properly initalize that value for us,
but it doesn't. sctp_outq_init operates under the assumption that the outq
struct is all 0's (as it is when called from sctp_association_init), but using
it in __sctp_outq_teardown violates that assumption. We should do a memset in
sctp_outq_init to ensure that the entire structure is in a known state there
instead.
Florian Fainelli [Wed, 12 Jun 2013 19:53:05 +0000 (20:53 +0100)]
bcm63xx_enet: add support Broadcom BCM6345 Ethernet
This patch adds support for the Broadcom BCM6345 SoC Ethernet. BCM6345
has a slightly different and older DMA engine which requires the
following modifications:
- the width of the DMA channels on BCM6345 is 64 bytes vs 16 bytes,
which means that the helpers enet_dma{c,s} need to account for this
channel width and we can no longer use macros
- BCM6345 DMA engine does not have any internal SRAM for transfering
buffers
- BCM6345 buffer allocation and flow control is not per-channel but
global (done in RSET_ENETDMA)
- the DMA engine bits are right-shifted by 3 compared to other DMA
generations
- the DMA enable/interrupt masks are a little different (we need to
enabled more bits for 6345)
- some register have the same meaning but are offsetted in the ENET_DMAC
space so a lookup table is required to return the proper offset
The MAC itself is identical and requires no modifications to work.
Willem de Bruijn [Thu, 13 Jun 2013 19:29:38 +0000 (15:29 -0400)]
net-rps: fixes for rps flow limit
Caught by sparse:
- __rcu: missing annotation to sd->flow_limit
- __user: direct access in cpumask_scnprintf
Also
- add endline character when printing bitmap if room in buffer
- avoid bucket overflow by reducing FLOW_LIMIT_HISTORY
The last item warrants some explanation. The hashtable buckets are
subject to overflow if FLOW_LIMIT_HISTORY is larger than or equal
to bucket size, since all packets may end up in a single bucket. The
current (rather arbitrary) history value of 256 happens to match the
buffer size (u8).
As a result, with a single flow, the first 128 packets are accepted
(correct), the second 128 packets dropped (correct) and then the
history[] array has filled, so that each subsequent new packet
causes an increment in the bucket for new_flow plus a decrement
for old_flow: a steady state.
This is fine if packets are dropped, as the steady state goes away
as soon as a mix of traffic reappears. But, because the 256th packet
overflowed the bucket to 0: no packets are dropped.
Instead of explicitly adding an overflow check, this patch changes
FLOW_LIMIT_HISTORY to never be able to overflow a single bucket.
Neil Horman [Thu, 13 Jun 2013 19:31:28 +0000 (15:31 -0400)]
tulip: Properly check dma mapping result
Tulip throws an error when dma debugging is enabled, as it doesn't properly
check dma mapping results with dma_mapping_error() durring tx ring refills.
Easy fix, just add it in, and drop the frame if the mapping is bad
Linus Torvalds [Thu, 13 Jun 2013 22:32:17 +0000 (15:32 -0700)]
Merge tag 'devicetree-for-linus' of git://git.secretlab.ca/git/linux
Pull device tree bug fixes from Grant Likely:
"This branch contains the following bug fixes:
- Fix locking vs. interrupts. Bug caught by lockdep checks
- Fix parsing of cpp #line directive output by dtc
- Fix 'make clean' for dtc temporary files.
There is also a commit that regenerates the dtc lexer and parser files
with Bison 2.5. The only purpose of this commit is to separate the
functional change in the dtc bug fix from the code generation change
caused by a different Bison version"
* tag 'devicetree-for-linus' of git://git.secretlab.ca/git/linux:
dtc: ensure #line directives don't consume data from the next line
dtc: Update generated files to output from Bison 2.5
of: Fix locking vs. interrupts
kbuild: make sure we clean up DTB temporary files
Grant Likely [Thu, 13 Jun 2013 11:57:44 +0000 (12:57 +0100)]
dtc: ensure #line directives don't consume data from the next line
Previously, the #line parsing regex ended with ({WS}+[0-9]+)?. The {WS}
could match line-break characters. If the #line directive did not contain
the optional flags field at the end, this could cause any integer data on
the next line to be consumed as part of the #line directive parsing. This
could cause syntax errors (i.e. #line parsing consuming the leading 0
from a hex literal 0x1234, leaving x1234 to be parsed as cell data,
which is a syntax error), or invalid compilation results (i.e. simply
consuming literal 1234 as part of the #line processing, thus removing it
from the cell data).
Fix this by replacing {WS} with [ \t] so that it can't match line-breaks.
Convert all instances of {WS}, even though the other instances should be
irrelevant for any well-formed #line directive. This is done for
consistency and ultimate safety.
Grant Likely [Thu, 13 Jun 2013 12:00:43 +0000 (13:00 +0100)]
dtc: Update generated files to output from Bison 2.5
This patch merely updates the generated dtc parser and lexer files to
the output generated by Bison 2.5. The previous versions were generated
from version 2.4.1. The only reason for this commit is to minimize the
diff on the next commit which fixes a bug in the DTC #line directive
parsing. Otherwise the Bison changes would be intermingled with the
functional changes.
The OF code uses irqsafe locks everywhere except in a handful of functions
for no obvious reasons. Since the conversion from the old rwlocks, this
now triggers lockdep warnings when used at interrupt time. At least one
driver (ibmvscsi) seems to be doing that from softirq context.
This converts the few non-irqsafe locks into irqsafe ones, making them
consistent with the rest of the code.
Linus Torvalds [Thu, 13 Jun 2013 20:09:50 +0000 (13:09 -0700)]
Merge tag 'acpi-3.10-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull ACPI fix from Rafael Wysocki:
"This is an alternative fix for the regression introduced in 3.9 whose
previous fix had to be reverted right before 3.10-rc5, because it
broke one of the Tony's machines.
In this one the check is confined to the ACPI video driver (which is
the only one causing the problem to happen in the first place) and the
Tony's box shouldn't even notice it.
- ACPI fix for an issue causing ACPI video driver to attempt to bind
to devices it shouldn't touch from Rafael J Wysocki."
* tag 'acpi-3.10-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
ACPI / video: Do not bind to device objects with a scan handler
Linus Torvalds [Thu, 13 Jun 2013 20:08:51 +0000 (13:08 -0700)]
Merge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Peter Anvin:
"Another set of fixes, the biggest bit of this is yet another tweak to
the UEFI anti-bricking code; apparently we finally got some feedback
from Samsung as to what makes at least their systems fail. This set
should actually fix the boot regressions that some other systems (e.g.
SGI) have exhibited.
Other than that, there is a patch to avoid a panic with particularly
unhappy memory layouts and two minor protocol fixes which may or may
not be manifest bugs"
* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86: Fix typo in kexec register clearing
x86, relocs: Move __vvar_page from S_ABS to S_REL
Modify UEFI anti-bricking code
x86: Fix adjust_range_size_mask calling position
Linus Torvalds [Thu, 13 Jun 2013 19:36:42 +0000 (12:36 -0700)]
Merge branch 'rcu/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu
Pull RCU fixes from Paul McKenney:
"I must confess that this past merge window was not RCU's best showing.
This series contains three more fixes for RCU regressions:
1. A fix to __DECLARE_TRACE_RCU() that causes it to act as an
interrupt from idle rather than as a task switch from idle.
This change is needed due to the recent use of _rcuidle()
tracepoints that can be invoked from interrupt handlers as well
as from idle. Without this fix, invoking _rcuidle() tracepoints
from interrupt handlers results in splats and (more seriously)
confusion on RCU's part as to whether a given CPU is idle or not.
This confusion can in turn result in too-short grace periods and
therefore random memory corruption.
2. A fix to a subtle deadlock that could result due to RCU doing
a wakeup while holding one of its rcu_node structure's locks.
Although the probability of occurrence is low, it really
does happen. The fix, courtesy of Steven Rostedt, uses
irq_work_queue() to avoid the deadlock.
3. A fix to a silent deadlock (invisible to lockdep) due to the
interaction of timeouts posted by RCU debug code enabled by
CONFIG_PROVE_RCU_DELAY=y, grace-period initialization, and CPU
hotplug operations. This will not occur in production kernels,
but really does occur in randconfig testing. Diagnosis courtesy
of Steven Rostedt"
* 'rcu/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu:
rcu: Fix deadlock with CPU hotplug, RCU GP init, and timer migration
rcu: Don't call wakeup() with rcu_node structure ->lock held
trace: Allow idle-safe tracepoints to be called from irq
Linus Torvalds [Thu, 13 Jun 2013 18:02:31 +0000 (11:02 -0700)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux
Pull s390 fixes from Martin Schwidefsky:
"Three kvm related memory management fixes, a fix for show_trace, a fix
for early console output and a patch from Ben to help prevent compile
errors in regard to irq functions (or our lack thereof)"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
s390/pci: Implement IRQ functions if !PCI
s390/sclp: fix new line detection
s390/pgtable: make pgste lock an explicit barrier
s390/pgtable: Save pgste during modify_prot_start/commit
s390/dumpstack: fix address ranges for asynchronous and panic stack
s390/pgtable: Fix guest overindication for change bit
Linus Torvalds [Thu, 13 Jun 2013 17:18:33 +0000 (10:18 -0700)]
Merge tag 'asoc-v3.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound
Pull ASoC sound updates from Mark Brown:
"Takashi is travelling at the minute and it'd be good to get the
MAINTAINERS update in here merged so sending directly.
As well as the usual driver specifics we've got a couple of core fixes
here, one fixing capabilities for unidirectional streams and the other
fixing suspend while audio streams are active.
The suspend fix is a little involved but mostly as a result of
removing some special casing that was doing the wrong thing."
* tag 'asoc-v3.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound:
ASoC: tlv320aic3x: Remove deadlock from snd_soc_dapm_put_volsw_aic3x()
ASoC: dapm: Treat DAI widgets like AIF widgets for power
ASoC: arizona: Correct AEC loopback enable
ASoC: pcm: Require both CODEC and CPU support when declaring stream caps
MAINTAINERS: Remove myself from Wolfson maintainers
ASoC: wm8994: Ensure microphone detection state is reset on removal
ASoC: wm8994: Avoid leaking pm_runtime reference on removed jack race
ASoC: cs42l52: fix hp_gain_enum shift value.
ASoC: cs42l52: use correct PCM mixer TLV dB scale to match datasheet.
Linus Torvalds [Thu, 13 Jun 2013 17:13:29 +0000 (10:13 -0700)]
Merge tag 'md-3.10-fixes' of git://neil.brown.name/md
Pull md bugfixes from Neil Brown:
"A few bugfixes for md
Some tagged for -stable"
* tag 'md-3.10-fixes' of git://neil.brown.name/md:
md/raid1,5,10: Disable WRITE SAME until a recovery strategy is in place
md/raid1,raid10: use freeze_array in place of raise_barrier in various places.
md/raid1: consider WRITE as successful only if at least one non-Faulty and non-rebuilding drive completed it.
md: md_stop_writes() should always freeze recovery.
Daniel Drake [Tue, 11 Jun 2013 19:40:20 +0000 (15:40 -0400)]
Bluetooth: btmrvl: fix thread stopping race
There is currently a race condition in the btmrvl_remove_card() which
is causing hangs on suspend for OLPC. When the race occurs,
kthread_stop() never returns.
The problem is that btmrvl_service_main_thread() calls kthread_should_stop()
and then does a fair number of things before restarting the loop and
sleeping.
If the thread gets stopped after kthread_should_stop() is checked, but
before the sleep happens, the thread will go to sleep and won't necessarily
be woken up.
Move the kthread_should_stop() check into a race-free place.
Johan Hedberg [Thu, 13 Jun 2013 08:01:13 +0000 (11:01 +0300)]
Bluetooth: Fix conditions for HCI_Delete_Stored_Link_Key
Even though the HCI_Delete_Stored_Link_Key command is mandatory for 1.1
and later controllers some controllers do not seem to support it
properly as was witnessed by one Broadcom based controller:
< HCI Command: Delete Stored Link Key (0x03|0x0012) plen 7
bdaddr 00:00:00:00:00:00 all 1
> HCI Event: Command Complete (0x0e) plen 4
Delete Stored Link Key (0x03|0x0012) ncmd 1
status 0x11 deleted 0
Error: Unsupported Feature or Parameter Value
Luckily this same controller also doesn't list the command in its
supported commands bit mask (counting from 0 bit 7 of octet 6):
< HCI Command: Read Local Supported Commands (0x04|0x0002) plen 0
> HCI Event: Command Complete (0x0e) plen 68
Read Local Supported Commands (0x04|0x0002) ncmd 1
status 0x00
Commands: ffffffffffff1ffffffffffff30fffff3f
Therefore, it makes sense to move sending of HCI_Delete_Stored_Link_Key
to after receiving the supported commands response and to only send it
if its respective bit in the mask is set. The downside of this is that
we no longer send the HCI_Delete_Stored_Link_Key command for Bluetooth
1.1 controllers since HCI_Read_Local_Supported_Command was introduced in
version 1.2, but this is an acceptable penalty as the command in
question shouldn't affect critical behavior.
Bluetooth: Fix crash in l2cap_build_cmd() with small MTU
If a too small MTU value is set with ioctl(HCISETACLMTU) or by a bogus
controller, memory corruption happens due to a memcpy() call with
negative length.
Fix this crash on either incoming or outgoing connections with a MTU
smaller than L2CAP_HDR_SIZE + L2CAP_CMD_HDR_SIZE:
Josh Triplett [Thu, 13 Jun 2013 00:26:37 +0000 (17:26 -0700)]
turbostat: Increase output buffer size to accommodate C8-C10
On platforms with C8-C10 support, the additional C-states cause
turbostat to overrun its output buffer of 128 bytes per CPU. Increase
this to 256 bytes per CPU.
[ As a bugfix, this should go into 3.10; however, since the C8-C10
support didn't go in until after 3.9, this need not go into any stable
kernel. ]
H. Peter Anvin [Thu, 13 Jun 2013 15:59:23 +0000 (08:59 -0700)]
Merge tag 'efi-urgent' into x86/urgent
* More tweaking to the EFI variable anti-bricking algorithm. Quite a
few users were reporting boot regressions in v3.9. This has now been
fixed with a more accurate "minimum storage requirement to avoid
bricking" value from Samsung (5K instead of 50%) and code to trigger
garbage collection when we near our limit - Matthew Garrett.
net: sh_eth: fix incorrect RX length error if R8A7740
This patch fixes an issue that the driver increments the "RX length error"
on every buffer in sh_eth_rx() if the R8A7740.
This patch also adds a description about the Receive Frame Status bits.
Eric Dumazet [Fri, 7 Jun 2013 20:26:05 +0000 (13:26 -0700)]
ip_tunnel: remove __net_init/exit from exported functions
If CONFIG_NET_NS is not set then __net_init is the same as __init and
__net_exit is the same as __exit. These functions will be removed from
memory after the module loads or is removed. Functions that are exported
for use by other functions should never be labeled for removal.
Bug introduced by commit c54419321455631079c
("GRE: Refactor GRE tunneling code.")
MDIO driver should resume before CPSW ethernet driver so that CPSW connect
to the phy and start tx/rx ethernet packets, changing the suspend/resume
apis with suspend_late/resume_early.
Saurabh Mohan [Tue, 11 Jun 2013 00:45:10 +0000 (17:45 -0700)]
net/ipv4: ip_vti clear skb cb before tunneling.
If users apply shaper to vti tunnel then it will cause a kernel crash. The
problem seems to be due to the vti_tunnel_xmit function not clearing
skb->opt field before passing the packet to xfrm tunneling code.
Yuchung Cheng [Tue, 11 Jun 2013 22:35:32 +0000 (15:35 -0700)]
tcp: properly send new data in fast recovery in first RTT
Linux sends new unset data during disorder and recovery state if all
(suspected) lost packets have been retransmitted ( RFC5681, section
3.2 step 1 & 2, RFC3517 section 4, NexSeg() Rule 2). One requirement
is to keep the receive window about twice the estimated sender's
congestion window (tcp_rcv_space_adjust()), assuming the fast
retransmits repair the losses in the next round trip.
But currently it's not the case on the first round trip in either
normal or Fast Open connection, beucase the initial receive window
is identical to (expected) sender's initial congestion window. The
fix is to double it.
Sergei Shtylyov [Tue, 11 Jun 2013 23:07:29 +0000 (03:07 +0400)]
sh_eth: remove '__maybe_unused' annotations
Now that the SoC specific support is no longer done with help of #ifdef'fery,
we no longer need '__maybe_unused' annotations to sh_eth_select_mii() and
sh_eth_set_duplex()...
Nithin Sujir [Wed, 12 Jun 2013 18:08:59 +0000 (11:08 -0700)]
tg3: Wait for boot code to finish after power on
Some systems that don't need wake-on-lan may choose to power down the
chip on system standby. Upon resume, the power on causes the boot code
to startup and initialize the hardware. On one new platform, this is
causing the device to go into a bad state due to a race between the
driver and boot code, once every several hundred resumes. The same race
exists on open since we come up from a power on.
This patch adds a wait for boot code signature at the beginning of
tg3_init_hw() which is common to both cases. If there has not been a
power-off or the boot code has already completed, the signature will be
present and poll_fw() returns immediately. Also return immediately if
the device does not have firmware.
Guillaume Nault [Wed, 12 Jun 2013 14:07:23 +0000 (16:07 +0200)]
l2tp: Fix PPP header erasure and memory leak
Copy user data after PPP framing header. This prevents erasure of the
added PPP header and avoids leaking two bytes of uninitialised memory
at the end of skb's data buffer.
bonding: fix igmp_retrans type and two related races
First the type of igmp_retrans (which is the actual counter of
igmp_resend parameter) is changed to u8 to be able to store values up
to 255 (as per documentation). There are two races that were hidden
there and which are easy to trigger after the previous fix, the first is
between bond_resend_igmp_join_requests and bond_change_active_slave
where igmp_retrans is set and can be altered by the periodic. The second
race condition is between multiple running instances of the periodic
(upon execution it can be scheduled again for immediate execution which
can cause the counter to go < 0 which in the unsigned case leads to
unnecessary igmp retransmissions).
Since in bond_change_active_slave bond->lock is held for reading and
curr_slave_lock for writing, we use curr_slave_lock for mutual
exclusion. We can't drop them as there're cases where RTNL is not held
when bond_change_active_slave is called. RCU is unlocked in
bond_resend_igmp_join_requests before getting curr_slave_lock since we
don't need it there and it's pointless to delay.
The decrement is moved inside the "if" block because if we decrement
unconditionally there's still a possibility for a race condition although
it is much more difficult to hit (many changes have to happen in
a very short period in order to trigger) which in the case of 3 parallel
running instances of this function and igmp_retrans == 1
(with check bond->igmp_retrans-- > 1) is:
f1 passes, doesn't re-schedule, but decrements - igmp_retrans = 0
f2 then passes, doesn't re-schedule, but decrements - igmp_retrans = 255
f3 does the unnecessary retransmissions.
bonding: reset master mac on first enslave failure
If the bond device is supposed to get the first slave's MAC address and
the first enslavement fails then we need to reset the master's MAC
otherwise it will stay the same as the failed slave device. We do it
after err_undo_flags since that is the first place where the MAC can be
changed and we check if it should've been the first slave and if the
bond's MAC was set to it because that err place is used by multiple
locations prior to changing the master's MAC address.
Flavio Leitner [Tue, 11 Jun 2013 21:09:29 +0000 (23:09 +0200)]
net: make all team port device link events urgent
Since team functionality relies heavily on userspace daemon, we need to
deliver event to userspace via Netlink as quick as possible. So make all
team port device link events urgent.
Daniel Borkmann [Wed, 12 Jun 2013 14:02:27 +0000 (16:02 +0200)]
packet: packet_getname_spkt: make sure string is always 0-terminated
uaddr->sa_data is exactly of size 14, which is hard-coded here and
passed as a size argument to strncpy(). A device name can be of size
IFNAMSIZ (== 16), meaning we might leave the destination string
unterminated. Thus, use strlcpy() and also sizeof() while we're
at it. We need to memset the data area beforehand, since strlcpy
does not padd the remaining buffer with zeroes for user space, so
that we do not possibly leak anything.
Wu Fengguang [Wed, 12 Jun 2013 13:04:16 +0000 (21:04 +0800)]
net: ping_check_bind_addr() etc. can be static
net/ipv4/ping.c:286:5: sparse: symbol 'ping_check_bind_addr' was not declared. Should it be static?
net/ipv4/ping.c:355:6: sparse: symbol 'ping_set_saddr' was not declared. Should it be static?
net/ipv4/ping.c:370:6: sparse: symbol 'ping_clear_saddr' was not declared. Should it be static?
net/ipv6/ping.c:60:5: sparse: symbol 'dummy_ipv6_recv_error' was not declared. Should it be static?
net/ipv6/ping.c:64:5: sparse: symbol 'dummy_ip6_datagram_recv_ctl' was not declared. Should it be static?
net/ipv6/ping.c:69:5: sparse: symbol 'dummy_icmpv6_err_convert' was not declared. Should it be static?
net/ipv6/ping.c:73:6: sparse: symbol 'dummy_ipv6_icmp_error' was not declared. Should it be static?
net/ipv6/ping.c:75:5: sparse: symbol 'dummy_ipv6_chk_addr' was not declared. Should it be static?
net/ipv6/ping.c:201:5: sparse: symbol 'ping_v6_seq_show' was not declared. Should it be static?
Dinh Nguyen [Wed, 12 Jun 2013 16:05:03 +0000 (11:05 -0500)]
net: ethernet: stmicro: stmmac: Fix compile error when STMMAC_XMIT_DEBUG used
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c: In function:
stmmac_xmit drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:1902:74:
error: expected ) before __func__
David S. Miller [Thu, 13 Jun 2013 08:26:54 +0000 (01:26 -0700)]
Merge tag 'batman-adv-fix-for-davem' of git://git.open-mesh.org/linux-merge
Included change:
- fix "rtnl locked" concurrent executions by using rtnl_lock instead of
rtnl_trylock. This fix enables batman-adv initialisation to do not fail just
because somewhere else in the system another code path is holding the rtnl
lock. It is easy to see the problem when batman-adv is trying to start
together with other networking components.
- fix the routing protocol forwarding policy by enhancing the duplicate control
packet detection. When the right circumstances trigger the issue, some nodes in
the network become totally unreachable, so breaking the mesh connectivity.
- fix the Bridge Loop Avoidance component by not running the originator address
change handling routine when the component is disabled. The routine was
generating useless packets that were sent over the network.
Jan Beulich [Tue, 11 Jun 2013 10:00:34 +0000 (11:00 +0100)]
xen-netback: don't de-reference vif pointer after having called xenvif_put()
When putting vif-s on the rx notify list, calling xenvif_put() must be
deferred until after the removal from the list and the issuing of the
notification, as both operations dereference the pointer.
Changing this got me to notice that the "irq" variable was effectively
unused (and was of too narrow type anyway).
Ben Hutchings [Mon, 10 Jun 2013 16:34:13 +0000 (17:34 +0100)]
cxgb4: Do not set net_device::dev_id to VI index
net_device::dev_id should not be used merely to indicate a VI index,
as it affects the way the local part of IPv6 addresses is normally
generated.
This field was intended for use where multiple devices may share a
single assigned MAC address and need to have different IPv6 addresses.
T4 VIs each have their own MAC address.
Simon Horman [Thu, 13 Jun 2013 07:04:33 +0000 (16:04 +0900)]
sctp: Correct byte order of access to skb->{network, transport}_header
Corrects an byte order conflict introduced by 158874cac61245b84e939c92c53db7000122b7b0
("sctp: Correct access to skb->{network, transport}_header").
The values in question are host byte order.
However, with a non passthrough device we never set promisc on open,
even if NOPROMISC is off. As a result:
If userspace clears NOPROMISC on open, then does not clear it on a
netlink command, promisc counter is not decremented on stop and there
will be no way to clear it once macvlan is detached.
If userspace does not clear NOPROMISC on open, then sets NOPROMISC on a
netlink command, promisc counter will be decremented from 0 and overflow
to fffffffff with no way to clear promisc.
To fix, simply ignore NOPROMISC flag in a netlink command for
non-passthrough devices, same as we do at open/close.
Since we touch this code anyway - check dev_set_promiscuity return code
and pass it to users (though an error here is unlikely).
Gao feng [Thu, 13 Jun 2013 02:05:38 +0000 (10:05 +0800)]
netlink: make compare exist all the time
Commit da12c90e099789a63073fc82a19542ce54d4efb9
"netlink: Add compare function for netlink_table"
only set compare at the time we create kernel netlink,
and reset compare to NULL at the time we finially
release netlink socket, but netlink_lookup wants
the compare exist always.
So we should set compare after we allocate nl_table,
and never reset it. make comapre exist all the time.
H. Peter Anvin [Wed, 12 Jun 2013 14:37:43 +0000 (07:37 -0700)]
md/raid1,5,10: Disable WRITE SAME until a recovery strategy is in place
There are cases where the kernel will believe that the WRITE SAME
command is supported by a block device which does not, in fact,
support WRITE SAME. This currently happens for SATA drivers behind a
SAS controller, but there are probably a hundred other ways that can
happen, including drive firmware bugs.
After receiving an error for WRITE SAME the block layer will retry the
request as a plain write of zeroes, but mdraid will consider the
failure as fatal and consider the drive failed. This has the effect
that all the mirrors containing a specific set of data are each
offlined in very rapid succession resulting in data loss.
However, just bouncing the request back up to the block layer isn't
ideal either, because the whole initial request-retry sequence should
be inside the write bitmap fence, which probably means that md needs
to do its own conversion of WRITE SAME to write zero.
Until the failure scenario has been sorted out, disable WRITE SAME for
raid1, raid5, and raid10.
[neilb: added raid5]
This patch is appropriate for any -stable since 3.7 when write_same
support was added.
NeilBrown [Wed, 12 Jun 2013 01:01:22 +0000 (11:01 +1000)]
md/raid1,raid10: use freeze_array in place of raise_barrier in various places.
Various places in raid1 and raid10 are calling raise_barrier when they
really should call freeze_array.
The former is only intended to be called from "make_request".
The later has extra checks for 'nr_queued' and makes a call to
flush_pending_writes(), so it is safe to call it from within the
management thread.
Using raise_barrier will sometimes deadlock. Using freeze_array
should not.
As 'freeze_array' currently expects one request to be pending (in
handle_read_error - the only previous caller), we need to pass
it the number of pending requests (extra) to ignore.
The deadlock was made particularly noticeable by commits 050b66152f87c7 (raid10) and 6b740b8d79252f13 (raid1) which
appeared in 3.4, so the fix is appropriate for any -stable
kernel since then.
This patch probably won't apply directly to some early kernels and
will need to be applied by hand.
Alex Lyakas [Tue, 4 Jun 2013 17:42:21 +0000 (20:42 +0300)]
md/raid1: consider WRITE as successful only if at least one non-Faulty and non-rebuilding drive completed it.
Without that fix, the following scenario could happen:
- RAID1 with drives A and B; drive B was freshly-added and is rebuilding
- Drive A fails
- WRITE request arrives to the array. It is failed by drive A, so
r1_bio is marked as R1BIO_WriteError, but the rebuilding drive B
succeeds in writing it, so the same r1_bio is marked as
R1BIO_Uptodate.
- r1_bio arrives to handle_write_finished, badblocks are disabled,
md_error()->error() does nothing because we don't fail the last drive
of raid1
- raid_end_bio_io() calls call_bio_endio()
- As a result, in call_bio_endio():
if (!test_bit(R1BIO_Uptodate, &r1_bio->state))
clear_bit(BIO_UPTODATE, &bio->bi_flags);
this code doesn't clear the BIO_UPTODATE flag, and the whole master
WRITE succeeds, back to the upper layer.
So we returned success to the upper layer, even though we had written
the data onto the rebuilding drive only. But when we want to read the
data back, we would not read from the rebuilding drive, so this data
is lost.
[neilb - applied identical change to raid10 as well]
This bug can result in lost data, so it is suitable for any
-stable kernel.
NeilBrown [Wed, 8 May 2013 23:48:30 +0000 (09:48 +1000)]
md: md_stop_writes() should always freeze recovery.
__md_stop_writes() will currently sometimes freeze recovery.
So any caller must be ready for that to happen, and indeed they are.
However if __md_stop_writes() doesn't freeze_recovery, then
a recovery could start before mddev_suspend() is called, which
could be awkward. This can particularly cause problems or dm-raid.
So change __md_stop_writes() to always freeze recovery. This is safe
and more predicatable.
1) Fix dump iterator in nfnl_acct_dump() and ctnl_timeout_dump() to
dump all objects properly, from Pablo Neira Ayuso.
2) xt_TCPMSS must use the default MSS of 536 when no MSS TCP option is
present. Fix from Phil Oester.
3) qdisc_get_rtab() looks for an existing matching rate table and uses
that instead of creating a new one. However, it's key matching is
incomplete, it fails to check to make sure the ->data[] array is
identical too. Fix from Eric Dumazet.
4) ip_vs_dest_entry isn't fully initialized before copying back to
userspace, fix from Dan Carpenter.
5) Fix ubuf reference counting regression in vhost_net, from Jason
Wang.
6) When sock_diag dumps a socket filter back to userspace, we have to
translate it out of the kernel's internal representation first.
From Nicolas Dichtel.
7) davinci_mdio holds a spinlock while calling pm_runtime, which
sleeps. Fix from Sebastian Siewior.
8) Timeout check in sh_eth_check_reset is off by one, from Sergei
Shtylyov.
9) If sctp socket init fails, we can NULL deref during cleanup. Fix
from Daniel Borkmann.
10) netlink_mmap() does not propagate errors properly, from Patrick
McHardy.
11) Disable powersave and use minstrel by default in ath9k. From Sujith
Manoharan.
12) Fix a regression in that SOCK_ZEROCOPY is not set on tuntap sockets
which prevents vhost from being able to use zerocopy. From Jason
Wang.
13) Fix race between port lookup and TX path in team driver, from Jiri
Pirko.
14) Missing length checks in bluetooth L2CAP packet parsing, from Johan
Hedberg.
15) rtlwifi fails to connect to networking using any encryption method
other than WPA2. Fix from Larry Finger.
16) Fix iwlegacy build due to incorrect CONFIG_* ifdeffing for power
management stuff. From Yijing Wang.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (35 commits)
b43: stop format string leaking into error msgs
ath9k: Use minstrel rate control by default
Revert "ath9k_hw: Update rx gain initval to improve rx sensitivity"
ath9k: Disable PowerSave by default
net: wireless: iwlegacy: fix build error for il_pm_ops
rtlwifi: Fix a false leak indication for PCI devices
wl12xx/wl18xx: scan all 5ghz channels
wl12xx: increase minimum singlerole firmware version required
wl12xx: fix minimum required firmware version for wl127x multirole
rtlwifi: rtl8192cu: Fix problem in connecting to WEP or WPA(1) networks
mwifiex: debugfs: Fix out of bounds array access
Bluetooth: Fix mgmt handling of power on failures
Bluetooth: Fix missing length checks for L2CAP signalling PDUs
Bluetooth: btmrvl: support Marvell Bluetooth device SD8897
Bluetooth: Fix checks for LE support on LE-only controllers
team: fix checks in team_get_first_port_txable_rcu()
team: move add to port list before port enablement
team: check return value of team_get_port_by_index_rcu() for NULL
tuntap: set SOCK_ZEROCOPY flag during open
netlink: fix error propagation in netlink_mmap()
...
Linus Torvalds [Wed, 12 Jun 2013 23:42:39 +0000 (16:42 -0700)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-block
Pull block layer fixes from Jens Axboe:
"Outside of bcache (which really isn't super big), these are all
few-liners. There are a few important fixes in here:
- Fix blk pm sleeping when holding the queue lock
- A small collection of bcache fixes that have been done and tested
since bcache was included in this merge window.
- A fix for a raid5 regression introduced with the bio changes.
- Two important fixes for mtip32xx, fixing an oops and potential data
corruption (or hang) due to wrong bio iteration on stacked devices."
* 'for-linus' of git://git.kernel.dk/linux-block:
scatterlist: sg_set_buf() argument must be in linear mapping
raid5: Initialize bi_vcnt
pktcdvd: silence static checker warning
block: remove refs to XD disks from documentation
blkpm: avoid sleep when holding queue lock
mtip32xx: Correctly handle bio->bi_idx != 0 conditions
mtip32xx: Fix NULL pointer dereference during module unload
bcache: Fix error handling in init code
bcache: clarify free/available/unused space
bcache: drop "select CLOSURES"
bcache: Fix incompatible pointer type warning
The lockless reclaim hierarchy iterator currently has a misplaced
barrier that can lead to use-after-free crashes.
The reclaim hierarchy iterator consist of a sequence count and a
position pointer that are read and written locklessly, with memory
barriers enforcing ordering.
The write side sets the position pointer first, then updates the
sequence count to "publish" the new position. Likewise, the read side
must read the sequence count first, then the position. If the sequence
count is up to date, it's guaranteed that the position is up to date as
well:
writer: reader:
iter->position = position if iter->sequence == expected:
smp_wmb() smp_rmb()
iter->sequence = sequence position = iter->position
However, the read side barrier is currently misplaced, which can lead to
dereferencing stale position pointers that no longer point to valid
memory. Fix this.
Akinobu Mita [Wed, 12 Jun 2013 21:05:08 +0000 (14:05 -0700)]
frontswap: fix incorrect zeroing and allocation size for frontswap_map
The bitmap accessed by bitops must have enough size to hold the required
numbers of bits rounded up to a multiple of BITS_PER_LONG. And the
bitmap must not be zeroed by memset() if the number of bits cleared is
not a multiple of BITS_PER_LONG.
This fixes incorrect zeroing and allocation size for frontswap_map. The
incorrect zeroing part doesn't cause any problem because frontswap_map
is freed just after zeroing. But the wrongly calculated allocation size
may cause the problem.
For 32bit systems, the allocation size of frontswap_map is about twice
as large as required size. For 64bit systems, the allocation size is
smaller than requeired if the number of bits is not a multiple of
BITS_PER_LONG.
Chen Gang [Wed, 12 Jun 2013 21:05:07 +0000 (14:05 -0700)]
kernel/audit_tree.c:audit_add_tree_rule(): protect `rule' from kill_rules()
audit_add_tree_rule() must set 'rule->tree = NULL;' firstly, to protect
the rule itself freed in kill_rules().
The reason is when it is killed, the 'rule' itself may have already
released, we should not access it. one example: we add a rule to an
inode, just at the same time the other task is deleting this inode.
The work flow for adding a rule:
audit_receive() -> (need audit_cmd_mutex lock)
audit_receive_skb() ->
audit_receive_msg() ->
audit_receive_filter() ->
audit_add_rule() ->
audit_add_tree_rule() -> (need audit_filter_mutex lock)
...
unlock audit_filter_mutex
get_tree()
...
iterate_mounts() -> (iterate all related inodes)
tag_mount() ->
tag_trunk() ->
create_trunk() -> (assume it is 1st rule)
fsnotify_add_mark() ->
fsnotify_add_inode_mark() -> (add mark to inode->i_fsnotify_marks)
...
get_tree(); (each inode will get one)
...
lock audit_filter_mutex
The work flow for deleting an inode:
__destroy_inode() ->
fsnotify_inode_delete() ->
__fsnotify_inode_delete() ->
fsnotify_clear_marks_by_inode() -> (get mark from inode->i_fsnotify_marks)
fsnotify_destroy_mark() ->
fsnotify_destroy_mark_locked() ->
audit_tree_freeing_mark() ->
evict_chunk() ->
...
tree->goner = 1
...
kill_rules() -> (assume current->audit_context == NULL)
call_rcu() -> (rule->tree != NULL)
audit_free_rule_rcu() ->
audit_free_rule()
...
audit_schedule_prune() -> (assume current->audit_context == NULL)
kthread_run() -> (need audit_cmd_mutex and audit_filter_mutex lock)
prune_one() -> (delete it from prue_list)
put_tree(); (match the original get_tree above)
Naoya Horiguchi [Wed, 12 Jun 2013 21:05:04 +0000 (14:05 -0700)]
mm: migration: add migrate_entry_wait_huge()
When we have a page fault for the address which is backed by a hugepage
under migration, the kernel can't wait correctly and do busy looping on
hugepage fault until the migration finishes. As a result, users who try
to kick hugepage migration (via soft offlining, for example) occasionally
experience long delay or soft lockup.
This is because pte_offset_map_lock() can't get a correct migration entry
or a correct page table lock for hugepage. This patch introduces
migration_entry_wait_huge() to solve this.
mm/page_alloc.c: fix watermark check in __zone_watermark_ok()
The watermark check consists of two sub-checks. The first one is:
if (free_pages <= min + lowmem_reserve)
return false;
The check assures that there is minimal amount of RAM in the zone. If
CMA is used then the free_pages is reduced by the number of free pages
in CMA prior to the over-mentioned check.
if (!(alloc_flags & ALLOC_CMA))
free_pages -= zone_page_state(z, NR_FREE_CMA_PAGES);
This prevents the zone from being drained from pages available for
non-movable allocations.
The second check prevents the zone from getting too fragmented.
for (o = 0; o < order; o++) {
free_pages -= z->free_area[o].nr_free << o;
min >>= 1;
if (free_pages <= min)
return false;
}
The field z->free_area[o].nr_free is equal to the number of free pages
including free CMA pages. Therefore the CMA pages are subtracted twice.
This may cause a false positive fail of __zone_watermark_ok() if the CMA
area gets strongly fragmented. In such a case there are many 0-order
free pages located in CMA. Those pages are subtracted twice therefore
they will quickly drain free_pages during the check against
fragmentation. The test fails even though there are many free non-cma
pages in the zone.
This patch fixes this issue by subtracting CMA pages only for a purpose of
(free_pages <= min + lowmem_reserve) check.
Laura said:
We were observing allocation failures of higher order pages (order 5 =
128K typically) under tight memory conditions resulting in driver
failure. The output from the page allocation failure showed plenty of
free pages of the appropriate order/type/zone and mostly CMA pages in
the lower orders.
For full disclosure, we still observed some page allocation failures
even after applying the patch but the number was drastically reduced and
those failures were attributed to fragmentation/other system issues.
projects / linux.git / log