Ramiro Voicu hits the BUG_ON(!pte_none(*pte)) in zeromap_pte_range: kernel
bugzilla 7645.  Right: read_zero_pagealigned uses down_read of mmap_sem,
but another thread's racing read of /dev/zero, or a normal fault, can
easily set that pte again, in between zap_page_range and zeromap_page_range
getting there.  It's been wrong ever since 2.4.3.

The simple fix is to use down_write instead, but that would serialize reads
of /dev/zero more than at present: perhaps some app would be badly
affected.  So instead let zeromap_page_range return the error instead of
BUG_ON, and read_zero_pagealigned break to the slower clear_user loop in
that case - there's no need to optimize for it.

Use -EEXIST for when a pte is found: BUG_ON in mmap_zero (the other user of
zeromap_page_range), though it really isn't interesting there.  And since
mmap_zero wants -EAGAIN for out-of-memory, the zeromaps better return that
than -ENOMEM.
Signed-off-by: Hugh Dickins <hugh@veritas.com>
Cc: Ramiro Voicu: <Ramiro.Voicu@cern.ch>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

The SHA384 block size should be 128 bytes, not 96 bytes.  This was
spotted by Andrew Donofrio.

This breaks HMAC which uses the block size during setup and the final
calculation.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

As CBC is the default chaining method for cryptoloop, we should select
it from cryptoloop to ease the transition.  Spotted by Rene Herman.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

GLIBC uses them etc.

They are guarded by ifndef __KERNEL__ so nobody will start
accidently using them in the kernel again, it's just for
userspace.
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

At least some nforce cards continue to send legacy interrupts when MSI
is enabled, and these interrupts are treated as unhandled by the
kernel. This patch disables legacy interrupts explicitly when enabling
MSI mode.

The correct fix is to change the MSI infrastructure to disable legacy
interrupts when enabling MSI, but this is potentially risky if the
device isn't PCI-2.3 or is quirky, so the correct fix is going into
mainline, while patches like this one go into -stable.

Legend has it that it is most correct to disable legacy interrupts
before enabling MSI, but the mainline patch does it in the other
order, and this patch is "obviously" the same as mainline.
Signed-off-by: Daniel Barkalow <barkalow@iabervon.org>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: Greg KH <gregkh@suse.de>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

2.6.19  stopped booting (or booted based on build/config) on our x86_64
systems due to a bug introduced in 2.6.19.  check_nmi_watchdog schedules an
IPI on all cpus to  busy wait on a flag, but fails to set the busywait
flag if NMI functionality is disabled.  This causes the secondary cpus
to spin in an endless loop, causing the kernel bootup to hang.
Depending upon the build, the  busywait flag got overwritten (stack variable)
and caused  the kernel to bootup on certain builds.  Following patch fixes
the bug by setting the busywait flag before returning from check_nmi_watchdog.
I guess using a stack variable is not good here as the calling function could
potentially return while the busy wait loop is still spinning on the flag.

AK: I redid the patch significantly to be cleaner
Signed-off-by: Ravikiran Thirumalai <kiran@scalex86.org>
Signed-off-by: Shai Fultheim <shai@scalex86.org>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

The m32r kernel 2.6.18-rc1 or after cause build errors of "unknown isa
configuration" for userspace application programs, such as glibc, gdb, etc.

This is because the recent kernel do not include linux/config.h not to expose
kernel headers for userspace.

To fix the above compile errors, this patch fixes two headers ptrace.h and
sigcontext.h for m32r and makes them platform-independent.
Signed-off-by: Hirokazu Takata <takata@linux-m32r.org>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

It is possible to have tasklets get scheduled before softirqd has had a chance
to spawn on all CPUs.  This is totally harmless; after success during action
CPU_UP_PREPARE, action CPU_ONLINE will be called, which immediately wakes
softirqd on the appropriate CPU to process the already pending tasklets.  So
there is no danger of having a missed wakeup for any tasklets that were
already pending.

In particular, i386 is affected by this during startup, and is visible when
using a very large initrd; during the time it takes for the initrd to be
decompressed, a timer IRQ can come in and schedule RCU callbacks.  It is also
possible that resending of a hardware IRQ via a softirq triggers the same bug.

Because of different timing conditions, this shows up in all emulators and
virtual machines tested, including Xen, VMware, Virtual PC, and Qemu.  It is
also possible to trigger on native hardware with a large enough initrd,
although I don't have a reliable case demonstrating that.
Signed-off-by: Zachary Amsden <zach@vmware.com>
Cc: <caglar@pardus.org.tr>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

When kernel is compiled with old version of autofs (CONFIG_AUTOFS_FS), and
new (observed at least with 5.x.x) automount deamon is started, kernel
correctly reports incompatible version of kernel and userland daemon, but
then screws things up instead of correct handling of the error:

 autofs: kernel does not match daemon version
 =====================================
 [ BUG: bad unlock balance detected! ]
 -------------------------------------
 automount/4199 is trying to release lock (&type->s_umount_key) at:
 [<c0163b9e>] get_sb_nodev+0x76/0xa4
 but there are no more locks to release!

 other info that might help us debug this:
 no locks held by automount/4199.

 stack backtrace:
  [<c0103b15>] dump_trace+0x68/0x1b2
  [<c0103c77>] show_trace_log_lvl+0x18/0x2c
  [<c01041db>] show_trace+0xf/0x11
  [<c010424d>] dump_stack+0x12/0x14
  [<c012e02c>] print_unlock_inbalance_bug+0xe7/0xf3
  [<c012fd4f>] lock_release+0x8d/0x164
  [<c012b452>] up_write+0x14/0x27
  [<c0163b9e>] get_sb_nodev+0x76/0xa4
  [<c0163689>] vfs_kern_mount+0x83/0xf6
  [<c016373e>] do_kern_mount+0x2d/0x3e
  [<c017513f>] do_mount+0x607/0x67a
  [<c0175224>] sys_mount+0x72/0xa4
  [<c0102b96>] sysenter_past_esp+0x5f/0x99
 DWARF2 unwinder stuck at sysenter_past_esp+0x5f/0x99
 Leftover inexact backtrace:
  =======================

and then deadlock comes.

The problem: autofs_fill_super() returns EINVAL to get_sb_nodev(), but
before that, it calls kill_anon_super() to destroy the superblock which
won't be needed.  This is however way too soon to call kill_anon_super(),
because get_sb_nodev() has to perform its own cleanup of the superblock
first (deactivate_super(), etc.).  The correct time to call
kill_anon_super() is in the autofs_kill_sb() callback, which is called by
deactivate_super() at proper time, when the superblock is ready to be
killed.

I can see the same faulty codepath also in autofs4.  This patch solves
issues in both filesystems in a same way - it postpones the
kill_anon_super() until the proper time is signalized by deactivate_super()
calling the kill_sb() callback.

[raven@themaw.net: update comment]
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Acked-by: Ian Kent <raven@themaw.net>
Cc: <stable@kernel.org>
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

The 'testproc' swsusp debug mode thaws tasks twice in a row, which is _very_
confusing.  Fix that.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Pavel Machek <pavel@ucw.cz>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

OpenVZ Linux kernel team has found a problem with mounting in compat mode.

Simple command "mount -t smbfs ..." on Fedora Core 5 distro in 32-bit mode
leads to oops:

Unable to handle kernel NULL pointer dereference at 0000000000000000 RIP:
[<ffffffff802bc7c6>] compat_sys_mount+0xd6/0x290
PGD 34d48067 PUD 34d03067 PMD 0
Oops: 0000 [1] SMP
CPU: 0
Modules linked in: iptable_nat simfs smbfs ip_nat ip_conntrack vzdquota
parport_pc lp parport 8021q bridge llc vznetdev vzmon nfs lockd sunrpc vzdev
iptable_filter af_packet xt_length ipt_ttl xt_tcpmss ipt_TCPMSS
iptable_mangle xt_limit ipt_tos ipt_REJECT ip_tables x_tables thermal
processor fan button battery asus_acpi ac uhci_hcd ehci_hcd usbcore i2c_i801
i2c_core e100 mii floppy ide_cd cdrom
Pid: 14656, comm: mount
RIP: 0060:[<ffffffff802bc7c6>]  [<ffffffff802bc7c6>]
compat_sys_mount+0xd6/0x290
RSP: 0000:ffff810034d31f38  EFLAGS: 00010292
RAX: 000000000000002c RBX: 0000000000000000 RCX: 0000000000000000
RDX: ffff810034c86bc0 RSI: 0000000000000096 RDI: ffffffff8061fc90
RBP: ffff810034d31f78 R08: 0000000000000000 R09: 000000000000000d
R10: ffff810034d31e58 R11: 0000000000000001 R12: ffff810039dc3000
R13: 000000000805ea48 R14: 0000000000000000 R15: 00000000c0ed0000
FS:  0000000000000000(0000) GS:ffffffff80749000(0033) knlGS:00000000b7d556b0
CS:  0060 DS: 007b ES: 007b CR0: 000000008005003b
CR2: 0000000000000000 CR3: 0000000034d43000 CR4: 00000000000006e0
Process mount (pid: 14656, veid=300, threadinfo ffff810034d30000, task
ffff810034c86bc0)
Stack:  0000000000000000 ffff810034dd0000 ffff810034e4a000 000000000805ea48
 0000000000000000 0000000000000000 0000000000000000 0000000000000000
 000000000805ea48 ffffffff8021e64e 0000000000000000 0000000000000000
Call Trace:
 [<ffffffff8021e64e>] ia32_sysret+0x0/0xa

Code: 83 3b 06 0f 85 41 01 00 00 0f b7 43 0c 89 43 14 0f b7 43 0a
RIP  [<ffffffff802bc7c6>] compat_sys_mount+0xd6/0x290
 RSP <ffff810034d31f38>
CR2: 0000000000000000

The problem is that data_page pointer can be NULL, so we should skip data
conversion in this case.
Signed-off-by: Andrey Mirkin <amirkin@openvz.org>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

Fix http://bugzilla.kernel.org/show_bug.cgi?id=7606

WARNING: "drm_sman_set_manager" [drivers/char/drm/sis.ko] undefined!

Cc: <daniel-silveira@gee.inatel.br>
Cc: Dave Airlie <airlied@linux.ie>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

With CONFIG_SMP=n:

drivers/input/ff-memless.c:384: warning: implicit declaration of function 'local_bh_disable'
drivers/input/ff-memless.c:393: warning: implicit declaration of function 'local_bh_enable'

Really linux/spinlock.h should include linux/interrupt.h.  But interrupt.h
includes sched.h which will need spinlock.h.

So the patch breaks the _bh declarations out into a separate header and
includes it in bothj interrupt.h and spinlock.h.

Cc: "Randy.Dunlap" <rdunlap@xenotime.net>
Cc: Andi Kleen <ak@suse.de>
Cc: <stable@kernel.org>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

Restore API compatibility due to bits moved from rtnetlink.h to
separate headers.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

We must reserve SAR + MAX_HEADER bytes for IrLMP to fit in.
This fixes an oops reported (and fixed) by Jeet Chaudhuri, when max_sdu_size
is greater than 0.
Signed-off-by: Samuel Ortiz <samuel@sortiz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

We grab a reference to the route's inetpeer entry but
forget to release it in xfrm4_dst_destroy().

Bug discovered by Kazunori MIYAZAWA <kazunori@miyazawa.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

The PhidgetServo causes an Oops when any of its sysfs attributes are read
or written too, making the driver useless.
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

Currently the behaviour of disable_xfrm is inconsistent between
locally generated and forwarded packets. For locally generated
packets disable_xfrm disables the policy lookup if it is set on
the output device, for forwarded traffic however it looks at the
input device. This makes it impossible to disable xfrm on all
devices but a dummy device and use normal routing to direct
traffic to that device.

Always use the output device when checking disable_xfrm.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

ip_summed changes last summer had missed that one.  As the result,
we have ip_summed interpreted as CHECKSUM_PARTIAL now.  IOW,
->csum is interpreted as offset of checksum in the packet.  net/core/*
will both read and modify the value as that offset, with obvious
reasons.  At the very least it's a remote memory corruptor.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

On Sat, Dec 02, 2006 at 11:47:44PM +0300, Alexey Dobriyan wrote:
> David Binderman compiled 2.6.19 with icc and grepped for "was set but never
> used". Many warnings are on
> 	http://coderock.org/kj/unused-2.6.19-fs

Heh, the very first line:
fs/exec.c(1465): remark #593: variable "flag" was set but never used

fs/exec.c:
  1477		/*
  1478		 *	We cannot trust fsuid as being the "true" uid of the
  1479		 *	process nor do we know its entire history. We only know it
  1480		 *	was tainted so we dump it as root in mode 2.
  1481		 */
  1482		if (mm->dumpable == 2) {	/* Setuid core dump mode */
  1483			flag = O_EXCL;		/* Stop rewrite attacks */
  1484			current->fsuid = 0;	/* Dump root private */
  1485		}

And then filp_open follows with "flag" totally ignored.
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

ib_ucm_cleanup_events() holds file_mutex while calling ib_destroy_cm_id().
This can deadlock since ib_destroy_cm_id() flushes event handlers, and
ib_ucm_event_handler() needs file_mutex, too.  Therefore, drop the
file_mutex during the call to ib_destroy_cm_id().
Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Acked-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

Routine ieee80211softmac_wx_set_mlme has one return that fails
to release a mutex acquired at entry.
Signed-off-by: Maxime Austruy <maxime@tralhalla.org>
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

The attached patch resolves an issue where a IP DNATed packet with a
martian source is forwarded while it's better to drop it. It also
resolves messages complaining about ip forwarding being disabled while
it's actually enabled. Thanks to lepton <ytht.net@gmail.com> for
reporting this problem.

This is probably a candidate for the -stable release.
Signed-off-by: Bart De Schuymer <bdschuym@pandora.be>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

In compat mode, matches and targets valid hooks checks always successful due
to not initialized e->comefrom field yet. This patch separates this checks from
translation code and moves them after mark_source_chains() call, where these
marks are initialized.
Signed-off-by: Dmitry Mishin <dim@openvz.org>
Signed-off-by; Patrick McHardy <kaber@trash.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

Commit 590bdf7f introduced a regression
in match/target hook validation. mark_source_chains builds a bitmask
for each rule representing the hooks it can be reached from, which is
then used by the matches and targets to make sure they are only called
from valid hooks. The patch moved the match/target specific validation
before the mark_source_chains call, at which point the mask is always zero.

This patch returns back to the old order and moves the standard checks
to mark_source_chains. This allows to get rid of a special case for
standard targets as a nice side-effect.
Signed-off-by: Dmitry Mishin <dim@openvz.org>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

The following patch fixes the failure of sunhme drivers on x86 hosts
due to missing pci_enable_device() and pci_set_master() calls, lost
during code refactoring. It has been filed as bugzilla bug #7502 [0]
and Debian bug #397460 [1].

[0] http://bugzilla.kernel.org/show_bug.cgi?id=7502
[1] http://bugs.debian.org/397460Signed-off-by: Jurij Smakov <jurij@wooyd.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

Not returning -EINVAL, because someone might want to use the value
zero in some future gact_prob algorithm?
Signed-off-by: Kim Nordlund <kim.nordlund@nokia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

This reverts commit 281ea49b,
which broke ACPI Interrupt source overrides that move
the SCI from one IRQ in PIC mode to another in IOAPIC mode.

If the SCI shared an interrupt line with another device,
this would result in a "irq 18: nobody cared" type failure.

http://bugzilla.kernel.org/show_bug.cgi?id=7601Signed-off-by: Len Brown <len.brown@intel.com>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

As CBC is the default chaining method for cryptoloop, we should select
it from cryptoloop to ease the transition.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

The tc actions increased the size of struct tc_police, which broke
compatibility with old iproute binaries since both the act_police
and the old NET_CLS_POLICE code check for an exact size match.

Since the new members are not even used, the simple fix is to also
accept the size of the old structure. Dumping is not affected since
old userspace will receive a bigger structure, which is handled fine.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Acked-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

No need to revisit a chain we'd already finished with during
the check for current hook.  It's either instant loop (which
we'd just detected) or a duplicate work.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

We need that for iterator to work; existing check had been too weak.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

We need to verify that
	a) we are not too close to the end of buffer to dereference
	b) next entry we'll be checking won't be _before_ our

While we are at it, don't subtract unrelated pointers...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

In the scan section of ieee80211softmac, network transmits are disabled.
When SoftMAC re-enables transmits, it may override the wishes of a driver
that may have very good reasons for disabling transmits. At least one failure
in bcm43xx can be traced to this problem. In addition, several unexplained
problems may arise from the unexpected enabling of transmits.
Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

MAX_HEADER does not include the ipv6 header length in it,
so we need to add it in explicitly.

With help from YOSHIFUJI Hideaki.
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

It's all good.

This changes the type of variable "i" in rtl8169_init_one()
from "unsigned int" to "int". "i" is checked for < 0 later,
which can never happen for "unsigned". This results in broken
error handling.
Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: Francois Romieu <romieu@fr.zoreil.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

* 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm:
  [ARM] 3943/1: share declaration of struct pxa2xx_udc_mach_info between multiple platforms
  [ARM] Export smp_call_function()
  [ARM] Add PM_LEGACY defaults