This patch adds skb_recycle_check(), which can be used by a network
driver after transmitting an skb to check whether this skb can be
recycled as a receive buffer.
skb_recycle_check() checks that the skb is not shared or cloned, and
that it is linear and its head portion large enough (as determined by
the driver) to be recycled as a receive buffer. If these conditions
are met, it does any necessary reference count dropping and cleans
up the skbuff as if it just came from __alloc_skb().
Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The following actions are possible:
tcp_v6_rcv
skb->dev = NULL;
tcp_v6_do_rcv
tcp_v6_hnd_req
tcp_check_req
req->rsk_ops->send_ack == tcp_v6_send_ack
So, skb->dev can be NULL in tcp_v6_send_ack. We must obtain namespace
from dst entry.
Thanks to Vitaliy Gusev <vgusev@openvz.org> for initial problem finding
in IPv4 code.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
* Add support for AR2417 (include pci id) since my previous patch doesn't sit on top of base.c/ath5k.h anymore.
* Update module version to 0.6.0
Changes-Licensed-under: ISC
Signed-Off-by: Nick Kossifidis <mickflemm@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
* Fix srev reporting during attach
Changes-Licensed-under: ISC
Signed-Off-by: Nick Kossifidis <mickflemm@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
* Use QUIET mechanism to drain tx buffer on PCU for newer chips
* Make sure that INTPEND is really 1 and not 0xffffffff while checking for pending interrupts
Changes-Licensed-under: ISC
Signed-Off-by: Nick Kossifidis <mickflemm@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
* Use new SREV values and PHY srevs to identify radio type durring attach
Changes-Licensed-under: ISC
Signed-Off-by: Nick Kossifidis <mickflemm@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
One of the spin-on-condition loops in routine do_dummy_tx always exits before
the condition is satisfied. The hardware might be left in an inconsistent
state that might be the cause of the PHY transmission errors seen by some
users.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
* Update registers
* Update SREV values and add some PHY srevs
* Prepare ath5k.h for newer radios etc
Thanks to Atheros 's HAL source we now know for sure how many parts we have
and what their SREV values are. We also have some updates on registers. Prepare
ath5k for some major updates ;-)
My previous mail had 2 more patches following (git log misusage), sorry for double
posting ;-(
Changes-Licensed-under: ISC
Signed-Off-by: Nick Kossifidis <mickflemm@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
According to the newly-released Atheros HAL code, asserting the
TSF reset bit will toggle a hardware internal state, resulting in a
spurious reset on the next chip reset. Whenever we force a TSF bit,
write the bit twice to clear the internal signal.
Signed-off-by: Bob Copeland <me@bobcopeland.com>
Acked-by: Nick Kossifidis <mickflemm@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
When CONFIG_MAC80211_MESH=n and CONFIG_MAC80211_NOINLINE=y,
gcc doesn't optimize out a call to ieee80211_rx_h_mesh_fwding,
even if the previous comparison is always false in this case.
This leads to the following errors during modpost:
ERROR: "mpp_path_lookup" [net/mac80211/mac80211.ko] undefined!
ERROR: "mpp_path_add" [net/mac80211/mac80211.ko] undefined!
Fix by removing the possibility of uninlining
ieee80211_rx_h_mesh_fwding rx handler.
Signed-off-by: Davide Pesavento <davidepesa@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Fix by disabling rt2x00 rfkill support when rt2x00 is built-in and rfkill has been modularized, and
a similar scheme for the relationship between leds_class and rt2x00..
Also, give a warning to the end-user when rfkill-/leds-support is disabled this way, so that the
end-user has at least some clues on what is going on.
Proper fixing required some general updates of the Kconfig-structure for the rt2x00 driver, whereby
internal configuration symbols had to be moved to after the user-visible configuration symbols.
Signed-off-by: Gertjan van Wingerde <gwingerde@kpnplanet.nl>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This patch addresses comments from Dan Williams about the patch
committed as "libertas: Improvements on automatic tx power control via
SIOCSIWTXPOW."
Signed-off-by: Anna Neal <anna@cozybit.com>
Signed-off-by: Javier Cardona <javier@cozybit.com>
Signed-off-by: Dan Williams <dcbw@redhat.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This patch removes wme_tx_queue and wme_rx_queue from struct sta_info
and from the debugfs sub-structure of struct sta_info
in net/mac80211/sta_info.h, as they are useless and not used.
Signed-off-by: Rami Rosen <ramirose@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Yet another BCM4306 card with the Bluetooth Coexistence SPROM programming
error has been found.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
In "mac80211: make master iface not wireless" I accidentally
forgot to include these changes ... leading to the expected
BUG_ON errors.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Use correct DMA_MASK: 4964 and 5000 support 36 bit addresses for
pci express memory access.
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This patch enables power save setting from config (iwconfig power)
The sysfs power_level interface is still preserved as it has
mac80211 power implementation is not yet rich enough.
Signed-off-by: Ester Kummer <ester.kummer@intel.com>
Reviewed-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The patch adds HW bug W/A FH_RCSR_CHNL0_RX_IGNORE_RXF_EMPTY so that we
can enable again interrupt coalescing. It also uses named constants for
open code.
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The command
make C=2 CF="-D__CHECK_ENDIAN__" drivers/net/wireless/p54/
generates the following warnings:
.../p54common.c:152:38: warning: incorrect type in argument 1 (different base types)
.../p54common.c:152:38: expected restricted __be32 const [usertype] *p
.../p54common.c:152:38: got unsigned int *<noident>
.../p54common.c:184:15: warning: restricted __le32 degrades to integer
.../p54common.c:185:29: warning: cast to restricted __le16
.../p54common.c:309:11: warning: symbol 'p54_rf_chips' was not declared.
Should it be static?
.../p54common.c:313:5: warning: symbol 'p54_parse_eeprom' was not declared.
Should it be static?
.../p54common.c:620:43: warning: incorrect type in argument 3 (different base types)
.../p54common.c:620:43: expected unsigned long [unsigned] [usertype] len
.../p54common.c:620:43: got restricted __le16 [usertype] len
.../p54common.c:780:41: warning: restricted __le16 degrades to integer
.../p54common.c:781:32: warning: restricted __le16 degrades to integer
.../p54common.c:1250:28: warning: incorrect type in argument 2 (different base types)
.../p54common.c:1250:28: expected unsigned short [unsigned] [usertype] filter_type
.../p54common.c:1250:28: got restricted __le16 [usertype] filter_type
.../p54common.c:1252:28: warning: incorrect type in argument 2 (different base types)
.../p54common.c:1252:28: expected unsigned short [unsigned] [usertype] filter_type
.../p54common.c:1252:28: got restricted __le16 [usertype] filter_type
.../p54common.c:1257:42: warning: incorrect type in argument 2 (different base types)
.../p54common.c:1257:42: expected unsigned short [unsigned] [usertype] filter_type
.../p54common.c:1257:42: got restricted __le16
.../p54common.c:1260:42: warning: incorrect type in argument 2 (different base types)
.../p54common.c:1260:42: expected unsigned short [unsigned] [usertype] filter_type
.../p54common.c:1260:42: got restricted __le16
.../p54usb.c:228:10: warning: restricted __le32 degrades to integer
.../p54usb.c:228:23: warning: restricted __le32 degrades to integer
.../p54usb.c:228:7: warning: incorrect type in assignment (different base types)
.../p54usb.c:228:7: expected restricted __le32 [assigned] [usertype] chk
.../p54usb.c:228:7: got unsigned int
.../p54usb.c:221:8: warning: symbol 'p54u_lm87_chksum' was not declared.
Should it be static?
All of the above have been fixed. One question, however, remains: In struct
bootrec, the array "data" is treated in many places as native CPU order, but
it may be little-endian everywhere. As far as I can tell, this driver has only
been used with little-endian hardware.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
An additional BCM4306 has been found with the Bluetooth coexistence
SPROM coding error.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Since call to function sctp_sf_abort_violation() need paramter 'arg' with
'struct sctp_chunk' type, it will read the chunk type and chunk length from
the chunk_hdr member of chunk. But call to sctp_sf_violation_paramlen()
always with 'struct sctp_paramhdr' type's parameter, it will be passed to
sctp_sf_abort_violation(). This may cause kernel panic.
sctp_sf_violation_paramlen()
|-- sctp_sf_abort_violation()
|-- sctp_make_abort_violation()
This patch fixed this problem. This patch also fix two place which called
sctp_sf_violation_paramlen() with wrong paramter type.
Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix grammar errors spotted by Randy Dunlap,
and adds some more details.
Signed-off-by: Remi Denis-Courmont <remi.denis-courmont@nokia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add more docbook comments to network device functions and cleanup
the comments.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
dev_change_name and netdev_drivername should use const char on
parameters that are read-only input values. The strcpy to newname is
not needed since newname is not used later in function.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We're never supposed to shrink the headroom or tailroom. In fact,
shrinking the headroom is a fatal action.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
There's a race between mm->owner assignment and swapoff, more easily
seen when task slab poisoning is turned on. The condition occurs when
try_to_unuse() runs in parallel with an exiting task. A similar race
can occur with callers of get_task_mm(), such as /proc/<pid>/<mmstats>
or ptrace or page migration.
CPU0 CPU1
try_to_unuse
looks at mm = task0->mm
increments mm->mm_users
task 0 exits
mm->owner needs to be updated, but no
new owner is found (mm_users > 1, but
no other task has task->mm = task0->mm)
mm_update_next_owner() leaves
mmput(mm) decrements mm->mm_users
task0 freed
dereferencing mm->owner fails
The fix is to notify the subsystem via mm_owner_changed callback(),
if no new owner is found, by specifying the new task as NULL.
Jiri Slaby:
mm->owner was set to NULL prior to calling cgroup_mm_owner_callbacks(), but
must be set after that, so as not to pass NULL as old owner causing oops.
Daisuke Nishimura:
mm_update_next_owner() may set mm->owner to NULL, but mem_cgroup_from_task()
and its callers need to take account of this situation to avoid oops.
Hugh Dickins:
Lockdep warning and hang below exec_mmap() when testing these patches.
exit_mm() up_reads mmap_sem before calling mm_update_next_owner(),
so exec_mmap() now needs to do the same. And with that repositioning,
there's now no point in mm_need_new_owner() allowing for NULL mm.
Reported-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Signed-off-by: Hugh Dickins <hugh@veritas.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Paul Menage <menage@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* git://git.kernel.org/pub/scm/linux/kernel/git/bart/ide-2.6:
cdrom: update ioctl documentation
ide: note that IDE generic may prevent other drivers from attaching
ide-tape: fix vendor strings
Swarm: Fix crash due to missing initialization
* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
[SSB] Initialise dma_mask for SSB_BUSTYPE_SSB devices
[MIPS] BCM47xx: Fix build error due to missing PCI functions
[MIPS] IP27: Switch to dynamic interrupt routing avoding panic on error.
[MIPS] au1000: Make sure GPIO value is zero or one
* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jwessel/linux-2.6-kgdb:
kgdboc,tty: Fix tty polling search to use name correctly
kgdb, x86_64: fix PS CS SS registers in gdb serial
kgdb, x86_64: gdb serial has BX and DX reversed
kgdb, x86, arm, mips, powerpc: ignore user space single stepping
kgdb: could not write to the last of valid memory with kgdb
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
ALSA: ASoC: Fix another cs4270 error path
ALSA: make the CS4270 driver a new-style I2C driver
* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-rc-fixes-2.6:
[SCSI] qlogicpti: fix sg list traversal error in continuation entries
[SCSI] Fix hang with split requests
[SCSI] qla2xxx: Defer enablement of RISC interrupts until ISP initialization completes.
Commit f072181e64 ("kconfig: drop the
""trying to assign nonexistent symbol" warning") simply dropped the
warnings, but it does a little more than that, it also marks the current
.config as needed saving, so add this back.
Signed-off-by: Roman Zippel <zippel@linux-m68k.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Recent changes to oldconfig have mixed up the silentoldconfig handling,
so this fixes that by clearly separating that special mode, e.g.
KCONFIG_NOSILENTUPDATE is only relevant here, the .config is written as
needed.
This will also properly close Bug 11230.
Signed-off-by: Roman Zippel <zippel@linux-m68k.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The VFS interface for the 'd_compare()' is a bit special (read: 'odd'),
because it really just essentially replaces a memcmp(). The filesystem
is supposed to just compare the two names with whatever case-independent
or other function.
And when I say 'is supposed to', I obviously mean that 'procfs does odd
things, and actually looks at the dentry that we don't even pass down,
rather than just the name'. Which results in problems, because we
actually call d_compare before we have even verified that the dentry is
still hashed at all.
And that causes a problm since the inode that procfs looks at may have
been free'd and the d_inode pointer is NULL. procfs just assumes that
all dentries are positive, since procfs itself never generates a
negative one. But memory pressure will still result in the dentry
getting torn down, and as it is removed by RCU, it still remains visible
on some lists - and to d_compare.
If the filesystem just did a name comparison, we wouldn't care. And we
could just fix procfs to know about negative dentries too. But rather
than have the low-level filesystems know about internal VFS details,
just move the check for a unhashed dentry up a bit, so that we will only
call d_compare on dentries that are still active.
The actual oops this caused didn't look like a NULL pointer dereference
because procfs did a 'container_of(inode, struct proc_inode, vfs_inode)'
to get at its internal proc_inode information from the inode pointer,
and accessed a field below the inode. So the oops would look something
like
BUG: unable to handle kernel paging request at fffffffffffffff0
IP: [<ffffffff802bc6c6>] proc_sys_compare+0x36/0x50
and was seen on both x86-64 (Alexey Dobriyan and Hugh Dickins) and
ppc64 (Hugh Dickins).
Reported-by: Alexey Dobriyan <adobriyan@gmail.com>
Acked-by: Hugh Dickins <hugh@veritas.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Reviewed-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-of-by: Linus Torvalds <torvalds@linux-foundation.org>
Conversion to new-style i2c driver missed the error path of the
probe function. Fix it.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: Timur Tabi <timur@freescale.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Update the CS4270 ALSA device driver to use the new-style I2C interface.
Starting with the 2.6.27 PowerPC kernel, I2C devices that have entries in the
device trees can no longer be probed by old-style I2C drivers. The device
tree for Freescale MPC8610 HPCD has included an entry for the CS4270 since
2.6.25, but that entry was previously ignored by the PowerPC I2C subsystem.
Since that's no longer the case, the best solution is to update the CS4270
driver to a new-style interface, rather than try to revert the behavior of
new PowerPC I2C subsystem.
Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Boaz writes:
"I've reviewed all patches since Matthew's, and I find one small
problem.
In the load_cmd() there is a compound loop where the first 4 sg's are
set then the rest are set into a memory structure in group of 7 sg's.
Well the second 7-group and on is a bug because sg pointer does not advance.
This is a fall out from Jens's patch."
The reporter, Meelis Roos <mroos@ut.ee>, verified that this patch
does indeed fix his problem with qlogicpti.
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Commit 2fd673ecf0 which tried to remove
hardreset for generic accidentally removed it for all flavors as all
others were inheriting from nv_generic_ops. This patch reinstates
nv_hardreset() and puts it into nv_common_ops which all flavors
inherit from. nv_generic_ops now inherits from nv_common_ops and
overrides .hardreset to ATA_OP_NULL.
While at it, explain why nv_hardreset and ATA_OP_NULL override are
necessary.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Lennert Buytenhek