Make sure that a bio doesn't contain NULL pages in the front of its vec,
if a device bounces a bio that doesn't start from 0.

Problem noted by Mark Haverkamp.
Signed-off-by: Jens Axboe <axboe@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

The new pipe buffer merging doesn't check for the "no readers" case, so
with small writes that coalesce, we may not get a timely SIGPIPE/EPIPE
notification to the writer.
Signed-off-by: Andrea Arcangeli <andrea@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

The pipe code has traditionally not cared about which end-point of the
pipe you are polling, meaning that if you poll the write-only end of a
pipe, it will still set the "this pipe is readable" bits if there is
data to be read on the read side.

That makes no sense, and together with the new bigger buffers breaks
python-twisted.

Based on debugging/patch by Andrea Arcangeli and testcase from
Thomas Crhak

srat's node_to_pxm() references pxm2node[] after init. so pxm2node[]
should not be __initdata.

Error: ./arch/x86_64/mm/srat.o .text refers to 0000000000000008 R_X86_64_32S      .init.data
Error: ./arch/x86_64/mm/srat.o .text refers to 0000000000000015 R_X86_64_32S      .init.data
Signed-off-by: Randy Dunlap <rddunlap@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

into ppc970.osdl.org:/home/torvalds/v2.6/linux

We should report the total bytes in the whole receive
queue, not just the first packet, in these cases.

Reported by Uwe Bonnes.
Signed-off-by: David S. Miller <davem@davemloft.net>

Spotted by itkes@fat.imed.msu.ru, the fn_hash_dump_bucket() main
loop does not increment 'i' properly, and thus routes will not
be listed, when the test 'i < s_i' passes.

The bug was added when the code was converted over to
hlist_for_each_entry() by your's truly.
Signed-off-by: David S. Miller <davem@davemloft.net>

into ppc970.osdl.org:/home/torvalds/v2.6/linux

into ppc970.osdl.org:/home/torvalds/v2.6/linux

It can only be set by root, but let's not tempt people to do
things that can't work.

setup_APIC_timer is only called in __init context and uses __initdata

wait_timer_tick refers to the __init functions wait_8254_wraparound
or wait_hpet_tick, hence must be __initdata.

scsi_dev_flags is referred to in
module_param_string(dev_flags, scsi_dev_flags, sizeof(scsi_dev_flags), 0);

cfq_init() calls __init cfq_slab_setup and hence must be __init itself

also made it static

parport_init_mode is referred to in int __devinit sio_via_probe().

I have just tested kernel version 2.6.11-rc5 and noticed it is not
possible to do an USB transfer by submitting an URB to an output
endpoint.  This breaks newest versions of libusb and thus SANE, gphoto2,
and a lot of software.

The bug was introduced in version 2.6.11-rc1 and is due to a wrong
comparison. 
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

This ain't pretty. Real fix under discussion.

modedb can not be __init because fb_find_mode() (which is not __init)
may get db == NULL.  fb_find_mode() is called from modules. 
Signed-off-by: Olaf Hering <olh@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

into nuts.davemloft.net:/disk1/BK/net-2.6

into ppc970.osdl.org:/home/torvalds/v2.6/linux

We were passing the size of the region in pages, where as
remap_pfn_range expected the size in bytes.
Signed-off-by: Russell King <rmk@arm.linux.org.uk>

The cast to void still triggers the gcc "warning: ignoring return value"
warning, which is dumb of it.  So ignore the return value differently.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Here's a patch for 2.6.11-rc5 that a lot of cdc-acm driver users are
clamoring for.

There's a bug introduced in a cleanup which will lead to a race making
reopenings fail.  This fix is by Alexander Lykanov.
Signed-off-by: Oliver Neukum <oliver@neukum.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

The attached patch makes keyctl() use the correct argument when invoking
the KEYCTL_JOIN_SESSION_KEYRING function.

I'm not sure how this evaded testing before, but I suspect the compiler was
kind and made both argument registers hold the same value.

Thanks to Kevin Coffman <kwc@citi.umich.edu> for spotting this.
Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

This fixes a potentially bad (although very rarely triggered) bug in the
ppc64 hugepage code.

hpte_update() did not correctly calculate the address for hugepages, so
pte_clear() (which we use for hugepage ptes as well as normal ones)
would not correctly flush the hash page table entry.  Under the right
circumstances this could potentially lead to duplicate hash entries,
which is very bad.

davem's upcoming patch to pass the virtual address directly to set_pte()
and its ilk will obsolete this, but this is bad enough it should
probably be fixed in the meantime.
Signed-off-by: David Gibson <dwg@au1.ibm.com>
Acked-by: William Irwin <wli@holomorphy.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Revert to previous way of handling masked_irq argument.
Reported to fix problems with shared PCI IRQs.
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

* always call __ide_do_rw_disk() in ide_do_rw_disk()
* modify ide_hwif_t->rw_disk hook accordingly
* update and cleanup hpt372n_rw_disk()
  (the only user of ide_hwif_t->rw_disk hook)
* make __ide_do_rw_disk() static + fix comment
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

It's a bit difficult to have a value and a sign bit in a
1-bit field.

Fix (90) boolean/bitfield sparse warnings:
include/linux/ide.h:937:18: warning: dubious one-bit signed bitfield
include/linux/ide.h:939:17: warning: dubious one-bit signed bitfield
Signed-off-by: Randy Dunlap <rddunlap@osdl.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

into nuts.davemloft.net:/disk1/BK/net-2.6

From Randy Dunlap
Signed-off-by: David S. Miller <davem@davemloft.net>

This is the best we've got: We cannot release and re-grab lock,
since checkentry() is called before ip_tables.c grabs ipt_mutex.  
We also cannot grab the hashtable spinlock, since htable_create will 
call vmalloc, and that can sleep.  And we cannot just re-search
the list of htable's in htable_create(), since then we would
create duplicate proc files.
Signed-off-by: Harald Welte <laforge@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

From Nishanth Aravamudan <nacc@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Do it in addrconf_ifdown.  This fixes OOPSes on shutdown
with 2.6.10
Signed-off-by: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

The path for loopback is:
LOCAL_OUT: conntrack defrags
POST_ROUTING: conntrack refrags
PRE_ROUTING: skip conntrack defrag because skb->nfct != NULL
PRE_ROUTING: NAT gets hit by fragments

Always defrag on loopback if NAT is compiled in.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Acked-by: Rusty Russel <rusty@rustcorp.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

into pobox.com:/garz/repo/net-drivers-2.6

Factor out some code
Signed-off-by: Francois Romieu <romieu@fr.zoreil.com>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>

IRQ races during change of mtu
- NAPI poll must be enabled prior to IRQ activation or the IRQ handler
  will not know what to do with an incoming packet;
- rtl8169_down() needs to try twice to sync with the IRQ handler when
  it is not issued under !netif_running() protection.

Both changes make it safe to request a change of mtu on a live device.
Signed-off-by: Francois Romieu <romieu@fr.zoreil.com>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>

Some futex functions do get_user calls while holding mmap_sem for
reading.  If get_user() faults, and another thread happens to be in mmap
(or somewhere else holding waiting on down_write for the same
semaphore), then do_page_fault will deadlock.  Most architectures seem
to be exposed to this.

To avoid it, make sure the page is available.  If not, release the
semaphore, fault it in and retry.

I also found another exposure by inspection, moving some of the code
around avoids the possible deadlock there.
Signed-off-by: Olof Johansson <olof@austin.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

into ppc970.osdl.org:/home/torvalds/v2.6/linux