11 years agobonding: fix enslavement error unwinds
Jay Vosburgh [Sat, 3 May 2008 01:06:02 +0000]
bonding: fix enslavement error unwinds

As part of:

commit c2edacf80e155ef54ae4774379d461b60896bc2e
Author: Jay Vosburgh <fubar@us.ibm.com>
Date:   Mon Jul 9 10:42:47 2007 -0700

    bonding / ipv6: no addrconf for slaves separately from master

two steps were rearranged in the enslavement process: netdev_set_master
is now before the call to dev_open to open the slave.

This patch updates the error cases and unwind process at the
end of bond_enslave to match the new order.  Without this patch, it is
possible for the enslavement to fail, but leave the slave with IFF_SLAVE
set in its flags.

Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

11 years agobonding: Deadlock between bonding_store_bonds and bond_destroy_sysfs.
Pavel Emelyanov [Sat, 3 May 2008 00:49:39 +0000]
bonding: Deadlock between bonding_store_bonds and bond_destroy_sysfs.

The sysfs layer has an internal protection, that ensures, that
all the process sitting inside ->sore/->show callback exits
before the appropriate entry is unregistered (the calltraces
are rather big, but I can provide them if required).

On the other hand, bonding takes rtnl_lock in
a) the bonding_store_bonds, i.e. in ->store callback,
b) module exit before calling the sysfs unregister routines.

Thus, the classical AB-BA deadlock may occur. To reproduce run
# while :; do modprobe bonding; rmmod bonding; done
and
# while :; do echo '+bond%d' > /sys/class/net/bonding_masters ; done
in parallel.

The fix is to move the bond_destroy_sysfs out of the rtnl_lock,
but _before_ bond_free_all to make sure no bonding devices exist
after module unload.

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Acked-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

11 years agobonding: fix error unwind in bonding_store_bonds
Jay Vosburgh [Sat, 3 May 2008 00:49:38 +0000]
bonding: fix error unwind in bonding_store_bonds

Fixed an error unwind in bonding_store_bonds that didn't release
the locks it held, and consolidated unwinds into a common block at the
end of the function.  Bug reported by Pavel Emelyanov <xemul@openvz.org>,
who provided a different fix.

Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

11 years agobonding: Do not call free_netdev for already registered device.
Pavel Emelyanov [Sat, 3 May 2008 00:49:37 +0000]
bonding: Do not call free_netdev for already registered device.

If the call to bond_create_sysfs_entry in bond_create fails, the
proper rollback is to call unregister_netdevice, not free_netdev.
Otherwise - kernel BUG at net/core/dev.c:4057!

Checked with artificial failures injected into bond_create_sysfs_entry.

Pavel's original patch modified by Jay Vosburgh to move code around
for clarity (remove goto-hopping within the unwind block).

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

11 years agonet_cls_act: act_simple dont ignore realloc code
Jamal Hadi Salim [Tue, 6 May 2008 07:10:24 +0000]
net_cls_act: act_simple dont ignore realloc code

reallocation of the policy data was being ignored. It could fail.
Simplify so that there is no need for reallocating.

Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoiwlwifi: make IWLWIFI a tristate
Adrian Bunk [Tue, 6 May 2008 07:04:47 +0000]
iwlwifi: make IWLWIFI a tristate

IWLWIFI should be a tristate so that if IWLCORE and/or IWL3945 are m
and none of them is y kbuild doesn't create an empty
drivers/net/wireless/built-in.o

This patch also removes the pointless "default n".

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoRevert "atm: Do not free already unregistered net device."
David S. Miller [Tue, 6 May 2008 07:00:16 +0000]
Revert "atm: Do not free already unregistered net device."

This reverts commit 65e4113684e50cee75357ce10dc9026b0929e4e9.

Unlike the other cases Pavel fixed, this case did not
setup a netdev->destructor of free_netdev, therefore this
change was not correct.

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

11 years agodccp: return -EINVAL on invalid feature length
Chris Wright [Mon, 5 May 2008 20:50:24 +0000]
dccp: return -EINVAL on invalid feature length

dccp_feat_change() validates length and on error is returning 1.
This happens to work since call chain is checking for 0 == success,
but this is returned to userspace, so make it a real error value.

Signed-off-by: Chris Wright <chrisw@sous-sol.org>
Acked-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoirda: fix !PNP support for drivers/net/irda/smsc-ircc2.c
Ingo Molnar [Mon, 5 May 2008 08:06:54 +0000]
irda: fix !PNP support for drivers/net/irda/smsc-ircc2.c

x86.git testing found this build bug on v2.6.26-rc1:

  ERROR: "pnp_get_resource" [drivers/net/irda/smsc-ircc2.ko] undefined!
  make[1]: *** [__modpost] Error 1
  make: *** [modules] Error 2

the driver did not anticipate the case of !CONFIG_PNP which is rare but
still possible. Instead of restricting the driver to PNP-only in the
Kconfig space, add the (trivial) dummy struct pnp_driver - this is that
other drivers use in the !PNP case too.

The driver itself can in theory be initialized on !PNP too in certain
cases, via smsc_ircc_legacy_probe().

Patch only minimally build tested, i dont have this hardware.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoirda: fix !PNP support in drivers/net/irda/nsc-ircc.c
Ingo Molnar [Mon, 5 May 2008 08:04:06 +0000]
irda: fix !PNP support in drivers/net/irda/nsc-ircc.c

x86.git testing found the following build failure in latest -git:

 drivers/built-in.o: In function `nsc_ircc_pnp_probe':
 nsc-ircc.c:(.text+0xdf1b6): undefined reference to `pnp_get_resource'
 nsc-ircc.c:(.text+0xdf1d4): undefined reference to `pnp_get_resource'
 nsc-ircc.c:(.text+0xdf1ee): undefined reference to `pnp_get_resource'
 nsc-ircc.c:(.text+0xdf237): undefined reference to `pnp_get_resource'
 nsc-ircc.c:(.text+0xdf24c): undefined reference to `pnp_get_resource'
 drivers/built-in.o:nsc-ircc.c:(.text+0xdf266): more undefined references to `pnp_get_resource' follow
 make: *** [.tmp_vmlinux1] Error 1

triggered via this config:

  http://redhat.com/~mingo/misc/config-Sat_May__3_20_53_13_CEST_2008.bad

while generally most users will have PNP enabled, drivers can support
non-PNP build mode too - and most drivers implement it. That is typically
done by providing a dummy pnp_driver structure that will not probe anything.

The fallback routines in the driver will handle this dumber mode of
operation too.

This patch implements that. I have not tested whether this actually
works on real hardware so take care. It does resolve the build bug.

[ Another solution that is used by a few drivers is to exclude the driver
  in the Kconfig if PNP is disabled, via "depends on PNP", but this would
  limit the availability of the driver needlessly. ]

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonet_cls_act: Make act_simple use of netlink policy.
Jamal Hadi Salim [Mon, 5 May 2008 07:22:35 +0000]
net_cls_act: Make act_simple use of netlink policy.

Convert to netlink helpers by using netlink policy validation.
As a side effect fixes a leak.

Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoip: Use inline function dst_metric() instead of direct access to dst->metric[]
Satoru SATOH [Mon, 5 May 2008 05:14:42 +0000]
ip: Use inline function dst_metric() instead of direct access to dst->metric[]

There are functions to refer to the value of dst->metric[THE_METRIC-1]
directly without use of a inline function "dst_metric" defined in
net/dst.h.

The following patch changes them to use the inline function
consistently.

Signed-off-by: Satoru SATOH <satoru.satoh@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoip: Make use of the inline function dst_metric_locked()
Satoru SATOH [Mon, 5 May 2008 05:12:43 +0000]
ip: Make use of the inline function dst_metric_locked()

Signed-off-by: Satoru SATOH <satoru.satoh@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoatm: Bad locking on br2684_devs modifications.
Pavel Emelyanov [Mon, 5 May 2008 01:00:36 +0000]
atm: Bad locking on br2684_devs modifications.

The list_del happens under read-locked devs_lock.

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoatm: Do not free already unregistered net device.
Pavel Emelyanov [Mon, 5 May 2008 01:00:05 +0000]
atm: Do not free already unregistered net device.

Both br2684_push and br2684_exit do so, but unregister_netdev()
releases the device itself.

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agomac80211: Do not free net device after it is unregistered.
Pavel Emelyanov [Mon, 5 May 2008 00:59:30 +0000]
mac80211: Do not free net device after it is unregistered.

The error path in ieee80211_register_hw() may call the unregister_netdev()
and right after it - the free_netdev(), which is wrong, since the
unregister releases the device itself.

So the proposed fix is to NULL the local->mdev after unregister is done
and check this before calling free_netdev().

I checked - no code uses the local->mdev after unregister in this error
path (but even if some did this would be a BUG).

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agobridge: Consolidate error paths in br_add_bridge().
Pavel Emelyanov [Mon, 5 May 2008 00:58:07 +0000]
bridge: Consolidate error paths in br_add_bridge().

This actually had to be merged with the patch #1, but I decided not to
mix two changes in one patch.

There are already two calls to free_netdev() in there, so merge them
into one.

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agobridge: Net device leak in br_add_bridge().
Pavel Emelyanov [Mon, 5 May 2008 00:57:29 +0000]
bridge: Net device leak in br_add_bridge().

In case the register_netdevice() call fails the device is leaked,
since the out: label is just rtnl_unlock()+return.

Free the device.

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoniu: Fix probing regression for maramba on-board chips.
David S. Miller [Sun, 4 May 2008 08:34:31 +0000]
niu: Fix probing regression for maramba on-board chips.

Changeset 7f7c4072ea552f97a0898331322f71986a97299c ("niu: Determine
the # of ports from the card's VPD data") caused maramba on-board
NIU ports to stop probing properly.

The old code had a fallback that would use a num_ports value of
4 if all the probing methods failed, but that was removed.

This restores the fallback of 4 ports, to get things working
again.

Bump driver version and release date.

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

11 years agolapbeth: Release ->ethdev when unregistering device.
David S. Miller [Sun, 4 May 2008 04:10:58 +0000]
lapbeth: Release ->ethdev when unregistering device.

Otherwise it leaks forever.

Based upon a report by Roland <devzero@web.de>

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

11 years agoxfrm: convert empty xfrm_audit_* macros to functions
Marcin Slusarz [Sun, 4 May 2008 04:03:01 +0000]
xfrm: convert empty xfrm_audit_* macros to functions

it removes these warnings when CONFIG_AUDITSYSCALL is unset:

net/xfrm/xfrm_user.c: In function 'xfrm_add_sa':
net/xfrm/xfrm_user.c:412: warning: unused variable 'sid'
net/xfrm/xfrm_user.c:411: warning: unused variable 'sessionid'
net/xfrm/xfrm_user.c:410: warning: unused variable 'loginuid'
net/xfrm/xfrm_user.c: In function 'xfrm_del_sa':
net/xfrm/xfrm_user.c:485: warning: unused variable 'sid'
net/xfrm/xfrm_user.c:484: warning: unused variable 'sessionid'
net/xfrm/xfrm_user.c:483: warning: unused variable 'loginuid'
net/xfrm/xfrm_user.c: In function 'xfrm_add_policy':
net/xfrm/xfrm_user.c:1132: warning: unused variable 'sid'
net/xfrm/xfrm_user.c:1131: warning: unused variable 'sessionid'
net/xfrm/xfrm_user.c:1130: warning: unused variable 'loginuid'
net/xfrm/xfrm_user.c: In function 'xfrm_get_policy':
net/xfrm/xfrm_user.c:1382: warning: unused variable 'sid'
net/xfrm/xfrm_user.c:1381: warning: unused variable 'sessionid'
net/xfrm/xfrm_user.c:1380: warning: unused variable 'loginuid'
net/xfrm/xfrm_user.c: In function 'xfrm_add_pol_expire':
net/xfrm/xfrm_user.c:1620: warning: unused variable 'sid'
net/xfrm/xfrm_user.c:1619: warning: unused variable 'sessionid'
net/xfrm/xfrm_user.c:1618: warning: unused variable 'loginuid'
net/xfrm/xfrm_user.c: In function 'xfrm_add_sa_expire':
net/xfrm/xfrm_user.c:1658: warning: unused variable 'sid'
net/xfrm/xfrm_user.c:1657: warning: unused variable 'sessionid'
net/xfrm/xfrm_user.c:1656: warning: unused variable 'loginuid'

Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonet: Fix useless comment reference loop.
Johannes Berg [Sun, 4 May 2008 03:56:42 +0000]
net: Fix useless comment reference loop.

include/linux/skbuff.h says:
        /* These elements must be at the end, see alloc_skb() for details.  */

net/core/skbuff.c says:
* See comment in sk_buff definition, just before the 'tail' member

This patch contains my guess as to the actual reason rather than a
dead comment reference loop.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agosch_htb: remove from event queue in htb_parent_to_leaf()
Jarek Poplawski [Sun, 4 May 2008 03:46:29 +0000]
sch_htb: remove from event queue in htb_parent_to_leaf()

There is lack of removing a class from the event queue while changing
from parent to leaf which can cause corruption of this rb tree. This
patch fixes a bug introduced by my patch: "sch_htb: turn intermediate
classes into leaves" commit: 160d5e10f87b1dc88fd9b84b31b1718e0fd76398.

Many thanks to Jan 'yanek' Bortl for finding a way to reproduce this
rare bug and narrowing the test case, which made possible proper
diagnosing.

This patch is recommended for all kernels starting from 2.6.20.

Reported-and-tested-by: Jan 'yanek' Bortl <yanek@ya.bofh.cz>
Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agorose: Wrong list_lock argument in rose_node seqops
Bernard Pidoux [Sat, 3 May 2008 00:03:22 +0000]
rose: Wrong list_lock argument in rose_node seqops

In rose_node_start() as well as in rose_node_stop() __acquires() and
spin_lock_bh() were wrongly passing rose_neigh_list_lock instead of
rose_node_list_lock arguments.

Signed-off-by: Bernard Pidoux <f6bvp@amsat.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonetns: Fix reassembly timer to use the right namespace
Daniel Lezcano [Sat, 3 May 2008 00:02:03 +0000]
netns: Fix reassembly timer to use the right namespace

This trivial fix retrieves the network namespace from frag queue
and use it to get the network device in the right namespace.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonetns: Fix device renaming for sysfs
Daniel Lezcano [Sat, 3 May 2008 00:00:58 +0000]
netns: Fix device renaming for sysfs

When a netdev is moved across namespaces with the
'dev_change_net_namespace' function, the 'device_rename' function is
used to fixup kobject and refresh the sysfs tree. The device_rename
function will call kobject_rename and this one will check if there is
an object with the same name and this is the case because we are
renaming the object with the same name.

The use of 'device_rename' seems for me wrong because we usually don't
rename it but just move it across namespaces. As we just want to do a
mini "netdev_[un]register", IMO the functions
'netdev_[un]register_kobject' should be used instead, like an usual
network device [un]registering.

This patch replace device_rename by netdev_unregister_kobject,
followed by netdev_register_kobject.

The netdev_register_kobject will call device_initialize and will raise
a warning indicating the device was already initialized. In order to
fix that, I split the device initialization into a separate function
and use it together with 'netdev_register_kobject' into
register_netdevice. So we can safely call 'netdev_register_kobject' in
'dev_change_net_namespace'.

This fix will allow to properly use the sysfs per namespace which is
coming from -mm tree.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Acked-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agobnx2: Update version to 1.7.5.
Michael Chan [Fri, 2 May 2008 23:58:18 +0000]
bnx2: Update version to 1.7.5.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agobnx2: Update RV2P firmware for 5709.
Michael Chan [Fri, 2 May 2008 23:57:59 +0000]
bnx2: Update RV2P firmware for 5709.

The new RV2P firmware fixes 2 issues:

1. The jumbo rx buffer page size is now configurable and set to the
   proper PAGE_SIZE.  Before, it was assumed to be always 4K.

2. Driver sometimes would crash when receiving jumbo packets mixed
   with firmware management packets.  This was caused by the old
   firmware DMA'ing to the wrong address.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agobnx2: Zero out context memory for 5709.
Michael Chan [Fri, 2 May 2008 23:57:26 +0000]
bnx2: Zero out context memory for 5709.

We should zero out the context memory for 5709 before each reset.  When
we resume after suspend for example, the memory may not be zero and the
chip may not function correctly.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agobnx2: Fix register test on 5709.
Michael Chan [Fri, 2 May 2008 23:57:08 +0000]
bnx2: Fix register test on 5709.

The register BNX2_CTX_STATUS (0x1004) should be skipped on 5709 as it
contains reserved bits.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agobnx2: Fix remote PHY initial link state.
Michael Chan [Fri, 2 May 2008 23:56:44 +0000]
bnx2: Fix remote PHY initial link state.

On some remote PHY blade systems, the driver receives no initial link
interrupt.  As a result, the GMII/MII MAC mode does not get setup properly.
To fix this problem, we add an initial poll of the link state after chip
reset.

With this change, the setting of the initial carrier state in the init
code can be eliminated.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agobnx2: Refine remote PHY locking.
Michael Chan [Fri, 2 May 2008 23:56:16 +0000]
bnx2: Refine remote PHY locking.

bnx2_set_remote_link() should be called under bp->phy_lock to protect
against concurrent polling and interrupt calls.  This change is needed
by the next patch which will add one initial poll of the remote PHY
link status.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agobridge: forwarding table information for >256 devices
Stephen Hemminger [Fri, 2 May 2008 23:53:33 +0000]
bridge: forwarding table information for >256 devices

The forwarding table binary interface (my bad choice), only exposes
the port number of the first 8 bits. The bridge code was limited to
256 ports at the time, but now the kernel supports up 1024 ports, so
the upper bits are lost when doing:

   brctl showmacs

The fix is to squeeze the extra bits into small hole left in data
structure, to maintain binary compatiablity.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agotg3: Update version to 3.92
Matt Carlson [Fri, 2 May 2008 23:49:50 +0000]
tg3: Update version to 3.92

This patch updates the version number to 3.92.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agotg3: Add link state reporting to UMP firmware
Matt Carlson [Fri, 2 May 2008 23:49:29 +0000]
tg3: Add link state reporting to UMP firmware

All variants of the 5714, 5715, and 5780 offer a feature called the
"Universal Management Port".  This feature is implemented in firmware
and is largely transparent to the driver, except...

It turns out that the UMP firmware needs to know the current status
of the link.  Because the firmware cannot touch the PHY registers while
the driver is in control of the device, it needs the driver to report
link status changes through an additional handshaking mechanism.
Without this handshake, it has been observed in the field that the UMP
firmware will not operate correctly.

This patch implements the new handshake with the UMP firmware.  Since
the handshake uses the same mechanism ASF heartbeats use, code was
added to detect and wait for completion of a pending previous event.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agotg3: Fix ethtool loopback test for 5761 BX devices
Matt Carlson [Fri, 2 May 2008 23:48:59 +0000]
tg3: Fix ethtool loopback test for 5761 BX devices

A CPMU related loopback test bug existed for AX revisions of the 5761.
While that errata has been fixed, the CPMU still slows down the core
clock too far to run the loopback test successfully.  This patch
disables the CPMU LINK_SPEED mode just like we do with the AX
revisions of the 5761 and all revisions of the 5784.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agotg3: Fix 5761 NVRAM sizes
Matt Carlson [Fri, 2 May 2008 23:48:36 +0000]
tg3: Fix 5761 NVRAM sizes

The 5761 NVRAM sizes assigned to the nvram_size member are half as big
as they should be.  This patch corrects the NVRAM sizes and replaces
the hardcoded constants with preprocessor constants for readability.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agotg3: Use constant 500KHz MI clock on adapters with a CPMU
Matt Carlson [Fri, 2 May 2008 23:47:53 +0000]
tg3: Use constant 500KHz MI clock on adapters with a CPMU

The MI clock is not configured correctly on adapters with the CPMU
present.  The tg3 driver has code which statically sets the MI clock to
be a fraction of the speed at which the core clock is running.
However, the CPMU can change the adapter's core clock frequency based
on operating conditions.  Consequently, the MI will run slow when the
core's clock has been slowed down.

There is a new 500KHz constant frequency clock available on adapters
with a CPMU.  This patch removes the static core clock scaling and
configures the MI clock to use this new 500KHz clock instead.

Running the MI clock at slower speeds will not directly result in data
corruption, but it does challenge the PHY read and write routine timeouts.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agohci_usb.h: fix hard-to-trigger race
Pavel Machek [Fri, 2 May 2008 23:45:10 +0000]
hci_usb.h: fix hard-to-trigger race

If someone tries to _urb_unlink while _urb_queue_head is running, he'll see
_urb->queue == NULL and fail to do any locking.  Prevent that from happening
by strategically placed barriers.

Signed-off-by: Pavel Machek <pavel@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agodccp: ccid2.c, ccid3.c use clamp(), clamp_t()
Harvey Harrison [Fri, 2 May 2008 23:44:07 +0000]
dccp: ccid2.c, ccid3.c use clamp(), clamp_t()

Makes the intention of the nested min/max clear.

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonet: remove NR_CPUS arrays in net/core/dev.c
Mike Travis [Fri, 2 May 2008 23:43:08 +0000]
net: remove NR_CPUS arrays in net/core/dev.c

Remove the fixed size channels[NR_CPUS] array in net/core/dev.c and
dynamically allocate array based on nr_cpu_ids.

Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonet: use get/put_unaligned_* helpers
Harvey Harrison [Fri, 2 May 2008 23:26:16 +0000]
net: use get/put_unaligned_* helpers

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agobluetooth: use get/put_unaligned_* helpers
Harvey Harrison [Fri, 2 May 2008 23:25:46 +0000]
bluetooth: use get/put_unaligned_* helpers

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoirda: use get_unaligned_* helpers
Harvey Harrison [Fri, 2 May 2008 23:21:52 +0000]
irda: use get_unaligned_* helpers

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonet: Add a WARN_ON_ONCE() to the transmit timeout function
Arjan van de Ven [Fri, 2 May 2008 23:21:07 +0000]
net: Add a WARN_ON_ONCE() to the transmit timeout function

WARN_ON_ONCE() gives a stack trace including the full module list.
Having this in the kernel dump for the timeout case in the
generic netdev watchdog will help us see quicker which driver
is involved. It also allows us to collect statistics
and patterns in terms of which drivers have this event occuring.

Suggested by Andrew Morton

Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonet: Add missing braces to multi-statement if()s
Ilpo Järvinen [Fri, 2 May 2008 23:20:10 +0000]
net: Add missing braces to multi-statement if()s

One finds all kinds of crazy things with some shell pipelining.

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonetns: assign PDE->data before gluing entry into /proc tree
Denis V. Lunev [Fri, 2 May 2008 11:12:41 +0000]
netns: assign PDE->data before gluing entry into /proc tree

In this unfortunate case, proc_mkdir_mode wrapper can't be used anymore and
this is no way to reuse proc_create_data due to nlinks assignment. So,
copy the code from proc_mkdir and assign PDE->data at the appropriate
moment.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonetfilter: assign PDE->data before gluing PDE into /proc tree
Denis V. Lunev [Fri, 2 May 2008 11:11:52 +0000]
netfilter: assign PDE->data before gluing PDE into /proc tree

Replace proc_net_fops_create with proc_create_data.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Acked-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonetfilter: assign PDE->fops before gluing PDE into /proc tree
Denis V. Lunev [Fri, 2 May 2008 11:10:57 +0000]
netfilter: assign PDE->fops before gluing PDE into /proc tree

Replace create_proc_entry with specially created for this purpose proc_create.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Acked-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoipv4: assign PDE->data before gluing PDE into /proc tree
Denis V. Lunev [Fri, 2 May 2008 11:10:08 +0000]
ipv4: assign PDE->data before gluing PDE into /proc tree

The check for PDE->data != NULL becomes useless after the replacement
of proc_net_fops_create with proc_create_data.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agovlan: assign PDE->data before gluing PDE into /proc tree
Denis V. Lunev [Fri, 2 May 2008 11:09:11 +0000]
vlan: assign PDE->data before gluing PDE into /proc tree

Simply replace proc_create and further data assigned with proc_create_data.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Acked-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoatm: assign PDE->data before gluing PDE into /proc tree
Denis V. Lunev [Fri, 2 May 2008 11:08:30 +0000]
atm: assign PDE->data before gluing PDE into /proc tree

Simply replace proc_create and further data assigned with proc_create_data.
proc_atm_dev_ops holds proper referrence.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoipv6: assign PDE->data before gluing PDE into /proc tree
Denis V. Lunev [Fri, 2 May 2008 09:46:55 +0000]
ipv6: assign PDE->data before gluing PDE into /proc tree

Simply replace proc_create and further data assigned with proc_create_data.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonet: assign PDE->data before gluing PDE into /proc tree
Denis V. Lunev [Fri, 2 May 2008 09:46:22 +0000]
net: assign PDE->data before gluing PDE into /proc tree

Simply replace proc_create and further data assigned with proc_create_data.
Additionally, there is no need to assign NULL to PDE->data after creation,
/proc generic has already done this for us.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agonetfilter: assign PDE->data before gluing PDE into /proc tree
Denis V. Lunev [Fri, 2 May 2008 09:45:42 +0000]
netfilter: assign PDE->data before gluing PDE into /proc tree

Simply replace proc_create and further data assigned with proc_create_data.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Acked-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agosunrpc: assign PDE->data before gluing PDE into /proc tree
Denis V. Lunev [Fri, 2 May 2008 09:44:36 +0000]
sunrpc: assign PDE->data before gluing PDE into /proc tree

Simply replace proc_create and further data assigned with proc_create_data.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoMerge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/linville/wireles...
David S. Miller [Thu, 1 May 2008 23:04:59 +0000]
Merge branch 'master' of /linux/kernel/git/linville/wireless-2.6

11 years agort2x00: Fix quality/activity led handling
Ivo van Doorn [Mon, 21 Apr 2008 17:01:09 +0000]
rt2x00: Fix quality/activity led handling

There was an obvious typo in LED structure
initialization which caused the radio and quality/activity
leds to be incorrectly initialized which resulted in
the leds not being enabled.

Additionally add the rt2x00led_led_activity() handler
that will enable TX/RX activity leds when the radio
is being enabled.

Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agort2x00: Don't enable short preamble for 1MBs
Ivo van Doorn [Mon, 21 Apr 2008 17:00:17 +0000]
rt2x00: Don't enable short preamble for 1MBs

The timing settings for 1MBs should exclude
the short preamble bit since that only applies
to 2MBs, 5.5MBs and 11MBs.

Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agoMake linux/wireless.h be able to compile
Kirill A. Shutemov [Tue, 22 Apr 2008 13:38:55 +0000]
Make linux/wireless.h be able to compile

Signed-off-by: Kirill A. Shutemov <kirill@shutemov.name>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agoiwlwifi: fix debug messages during scanning
Bill Moss [Thu, 17 Apr 2008 23:03:40 +0000]
iwlwifi: fix debug messages during scanning

direct_mask will be set when we are not associated and requesting a
direct scan. The second debug print will be confusing as priv->essid
is not set at that time and it will thus print "<hidden>" while it is
known to which AP a direct scan is requested - as previous debug message
also indicates.

Now make all debugging consistent.

Signed-off-by: Bill Moss <bmoss@clemson.edu>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agoiwlwifi: fix current channel is not scanned
Guy Cohen [Mon, 21 Apr 2008 22:41:57 +0000]
iwlwifi: fix current channel is not scanned

All channels should be scanned, including the current channel
when the client is associated.

Removed also unused flag to scan only active channels.

Signed-off-by: Guy Cohen <guy.cohen@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agob43: Fix some TX/RX locking issues
Michael Buesch [Fri, 25 Apr 2008 17:29:08 +0000]
b43: Fix some TX/RX locking issues

This fixes some TX/RX related locking issues.
With this patch applied, some of the PHY transmission errors are fixed.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agonet: fix returning void-valued expression warnings
Harvey Harrison [Thu, 1 May 2008 09:47:38 +0000]
net: fix returning void-valued expression warnings

drivers/net/8390.c:37:2: warning: returning void-valued expression
drivers/net/bnx2.c:1635:3: warning: returning void-valued expression
drivers/net/xen-netfront.c:1806:2: warning: returning void-valued expression
net/ipv4/tcp_hybla.c:105:3: warning: returning void-valued expression
net/ipv4/tcp_vegas.c:171:3: warning: returning void-valued expression
net/ipv4/tcp_veno.c:123:3: warning: returning void-valued expression
net/sysctl_net.c:85:2: warning: returning void-valued expression

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Acked-by: Alan Cox <alan@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agoMerge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/linville/wireles...
David S. Miller [Thu, 1 May 2008 09:06:32 +0000]
Merge branch 'master' of /linux/kernel/git/linville/wireless-2.6

11 years agoUpdate .gitignore to include include/linux/bounds.h
Theodore Ts'o [Thu, 1 May 2008 01:55:48 +0000]
Update .gitignore to include include/linux/bounds.h

(which is autogenerated by kbuild)

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

11 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
Linus Torvalds [Thu, 1 May 2008 03:13:22 +0000]
Merge git://git./linux/kernel/git/davem/net-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6:
  ipv6: Compilation fix for compat MCAST_MSFILTER sockopts.

11 years agoFix dnotify/close race
Al Viro [Thu, 1 May 2008 02:52:22 +0000]
Fix dnotify/close race

We have a race between fcntl() and close() that can lead to
dnotify_struct inserted into inode's list *after* the last descriptor
had been gone from current->files.

Since that's the only point where dnotify_struct gets evicted, we are
screwed - it will stick around indefinitely.  Even after struct file in
question is gone and freed.  Worse, we can trigger send_sigio() on it at
any later point, which allows to send an arbitrary signal to arbitrary
process if we manage to apply enough memory pressure to get the page
that used to host that struct file and fill it with the right pattern...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

11 years agox86: Mark OPTIMIZE_INLINING broken
Linus Torvalds [Thu, 1 May 2008 02:50:03 +0000]
x86: Mark OPTIMIZE_INLINING broken

So Ingo finally did figure out why UML broke with this option: UML
passes gcc the -fno-unit-at-a-time flag, and apparently that wreaks
havoc with gcc's inlining.

We could turn off -fno-unit-at-a-time for UML for gcc4+ (which is what
x86 does), but there's bad blood about this whole option, and it does
show that the thing is just fragile as heck.

So let tempers cool, and disable the thing, and we can revisit the
decision later.

Cc: Adrian Bunk <bunk@kernel.org>
Cc: David Miller <davem@davemloft.net>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

11 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/x86/linux...
Linus Torvalds [Thu, 1 May 2008 02:31:52 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/x86/linux-2.6-x86-fixes3

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/x86/linux-2.6-x86-fixes3: (21 commits)
  x86: numaq fix
  x86: 8K stacks by default
  x86: ioremap ram check fix
  x86: fix HT cpu booting on 32-bit
  x86: optimize inlining off
  x86: CONFIG_X86_ELAN fix
  x86: Kconfig fix
  x86 PAT: fix performance drop for glx, use UC minus for ioremap(), ioremap_nocache() and pci_mmap_page_range()
  x86: use defconfigs from x86/configs/*
  toshiba: use ioremap_cached
  revert: "x86: ioremap(), extend check to all RAM pages"
  x86: don't bother printing compat vdso address
  fix: x86: support for new UV apic
  x86: fix early-BUG message
  x86: iommu_sac_force can become static
  x86: add proper header for reboot_force
  x86 VISWS: build fix
  x86, voyager: fix ioremap_nocache()
  hpet: fix
  x86: unexport kmap_atomic_to_page
  ...

11 years agolibertas: fix use-before-check violation
Holger Schurig [Tue, 29 Apr 2008 08:07:56 +0000]
libertas: fix use-before-check violation

According to Coverity (kudo's to Adrian Bunk), we had one use-before-check
bug in libe libertas driver. This patch fixes this issue.

Signed-off-by: Holger Schurig <hs4233@mail.mn-solutions.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agob43: Fix dual-PHY devices
Michael Buesch [Sun, 27 Apr 2008 17:06:18 +0000]
b43: Fix dual-PHY devices

This fixes operation of dual-PHY (A/B/G) devices.
Do not anounce the A-PHY to mac80211, as that's not supported, yet.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agomac80211: incorrect shift direction
Harvey Harrison [Sun, 27 Apr 2008 10:48:40 +0000]
mac80211: incorrect shift direction

Looks like  5d2cdcd4e85c5187db30a6b29f79fbbe59f39f78 ("mac80211: get a
TKIP phase key from skb") got the shifts wrong.

Noticed by sparse:
net/mac80211/tkip.c:234:25: warning: right shift by bigger than source value
net/mac80211/tkip.c:235:25: warning: right shift by bigger than source value
net/mac80211/tkip.c:236:25: warning: right shift by bigger than source value

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agomac80211: insert WDS peer after adding interface
Johannes Berg [Thu, 24 Apr 2008 12:18:37 +0000]
mac80211: insert WDS peer after adding interface

This reorders the open code so that WDS peer STA info entries
are added after the corresponding interface is added to the
driver so that driver callbacks aren't invoked out of order.
Also make any master device startup fatal.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agomac80211: don't allow invalid WDS peer addresses
Johannes Berg [Thu, 24 Apr 2008 12:16:36 +0000]
mac80211: don't allow invalid WDS peer addresses

Rather than just disallowing the zero address, disallow all
invalid ones.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agomac80211: assign conf.beacon_control for mesh
Johannes Berg [Wed, 23 Apr 2008 21:35:09 +0000]
mac80211: assign conf.beacon_control for mesh

Drivers can rightfully assume that they get a beacon_control
if the beacon is set.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agomac80211: use 4-byte mesh sequence number
Luis Carlos Cobo [Wed, 23 Apr 2008 19:15:29 +0000]
mac80211: use 4-byte mesh sequence number

This follows the new 802.11s/D2.0 draft.

Signed-off-by: Luis Carlos Cobo <luisca@cozybit.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

11 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-2.6
Linus Torvalds [Thu, 1 May 2008 00:05:21 +0000]
Merge git://git./linux/kernel/git/gregkh/driver-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-2.6:
  klist: fix coding style errors in klist.h and klist.c
  driver core: remove no longer used "struct class_device"
  pcmcia: remove pccard_sysfs_interface warnings
  devres: support addresses greater than an unsigned long via dev_ioremap
  kobject: do not copy vargs, just pass them around
  sysfs: sysfs_update_group stub for CONFIG_SYSFS=n
  DEBUGFS: Correct location of debugfs API documentation.
  driver core: warn about duplicate driver names on the same bus
  klist: implement klist_add_{after|before}()
  klist: implement KLIST_INIT() and DEFINE_KLIST()
  sysfs: Disallow truncation of files in sysfs

11 years agoklist: fix coding style errors in klist.h and klist.c
Greg Kroah-Hartman [Wed, 30 Apr 2008 23:43:45 +0000]
klist: fix coding style errors in klist.h and klist.c

Finally clean up the odd spacing in these files.

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

11 years agodriver core: remove no longer used "struct class_device"
Kay Sievers [Wed, 12 Mar 2008 19:47:35 +0000]
driver core: remove no longer used "struct class_device"

Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

11 years agopcmcia: remove pccard_sysfs_interface warnings
David Brownell [Mon, 28 Apr 2008 08:03:20 +0000]
pcmcia: remove pccard_sysfs_interface warnings

Make the PCMCIA core stop using class_interface to hide socket attribute
registration.  This removes the associated section mismatch warnings, and
helps get to the point where that mechanism can finally be removed.

Simplify that attribute registration by using an attribute_group.
This is a net shrink in object size.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

11 years agodevres: support addresses greater than an unsigned long via dev_ioremap
Kumar Gala [Tue, 29 Apr 2008 15:25:48 +0000]
devres: support addresses greater than an unsigned long via dev_ioremap

Use a resource_size_t instead of unsigned long since some arch's are
capable of having ioremap deal with addresses greater than the size of a
unsigned long.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Cc: Tejun Heo <htejun@gmail.com>
Cc: Jeff Garzik <jgarzik@pobox.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

11 years agokobject: do not copy vargs, just pass them around
Kay Sievers [Wed, 30 Apr 2008 00:06:29 +0000]
kobject: do not copy vargs, just pass them around

This prevents a few unneeded copies.

Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

11 years agosysfs: sysfs_update_group stub for CONFIG_SYSFS=n
Randy Dunlap [Wed, 30 Apr 2008 16:01:17 +0000]
sysfs: sysfs_update_group stub for CONFIG_SYSFS=n

scsi_transport_spi uses sysfs_update_group() when CONFIG_SYSFS=n,
so provide a stub for it.

next-20080423/drivers/scsi/scsi_transport_spi.c:1467: error: implicit declaration of function 'sysfs_update_group'
make[3]: *** [drivers/scsi/scsi_transport_spi.o] Error 1

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

11 years agoDEBUGFS: Correct location of debugfs API documentation.
Robert P. J. Day [Fri, 25 Apr 2008 12:52:51 +0000]
DEBUGFS: Correct location of debugfs API documentation.

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

11 years agodriver core: warn about duplicate driver names on the same bus
Stas Sergeev [Sat, 26 Apr 2008 15:52:35 +0000]
driver core: warn about duplicate driver names on the same bus

Currently an attempt to register multiple
drivers with the same name causes the
stack trace with some cryptic error message.
The attached patch adds the necessary check
and the clear error message.

Signed-off-by: Stas Sergeev <stsp@aknet.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

11 years agoklist: implement klist_add_{after|before}()
Tejun Heo [Tue, 22 Apr 2008 09:58:46 +0000]
klist: implement klist_add_{after|before}()

Add klist_add_after() and klist_add_before() which puts a new node
after and before an existing node, respectively.  This is useful for
callers which need to keep klist ordered.  Note that synchronizing
between simultaneous additions for ordering is the caller's
responsibility.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

11 years agoklist: implement KLIST_INIT() and DEFINE_KLIST()
Tejun Heo [Fri, 25 Apr 2008 18:16:04 +0000]
klist: implement KLIST_INIT() and DEFINE_KLIST()

klist is missing static initializers and definition helper.  Add them.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

11 years agosysfs: Disallow truncation of files in sysfs
Ben Hutchings [Mon, 28 Apr 2008 14:59:58 +0000]
sysfs: Disallow truncation of files in sysfs

sysfs allows attribute files to be truncated, e.g. using ftruncate(), with the
expected effect on their inode.   For most attributes, this doesn't change the
"real" size of the file i.e. how much can be read from it.  However, the
parameter validation for reading and writing binary attribute files is based
on the inode size and not the size specified in the file's bin_attribute, so it
can be broken by this. For example, if we try using dd to write to such a file:

# pwd
/sys/bus/pci/devices/0000:08:00.0
# ls -l config
-rw-r--r--  1 root root 4096 Feb  1 17:35 config
# dd if=/dev/zero of=config bs=4 count=1
1+0 records in
1+0 records out
# ls -l config
-rw-r--r--  1 root root 0 Feb  1 17:50 config
# dd if=/dev/zero of=config bs=4 count=1 seek=128
dd: writing `config': No space left on device
1+0 records in
0+0 records out

Also, after truncation to 0, parameter validation for read and write is
disabled.  Most bin_attribute read and write methods also validate the size and
offset, but for some this will allow out-of-range access.  This may be a
security issue, though access to such files is often limited to root.  In any
case, the validation should remain for safety's sake!)

This was previously reported in Bugzilla as bug 9867.

sysfs should ignore size changes or else refuse them (by returning -EINVAL).
This patch makes it ignore them.

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

11 years agoFix ACPI vs proc_create_data() mismerge
Alexey Dobriyan [Thu, 1 May 2008 00:10:02 +0000]
Fix ACPI vs proc_create_data() mismerge

acpi_device_dir() is NULL until all files are createst, so everyting is
created in straight in /proc/ and creation code warns.

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

11 years agoipv6: Compilation fix for compat MCAST_MSFILTER sockopts.
Pavel Emelyanov [Wed, 30 Apr 2008 21:49:54 +0000]
ipv6: Compilation fix for compat MCAST_MSFILTER sockopts.

The last hunk from the commit dae50295 (ipv4/ipv6 compat: Fix SSM
applications on 64bit kernels.) escaped from the compat_ipv6_setsockopt
to the ipv6_getsockopt (I guess due to patch smartness wrt searching
for context) thus breaking 32-bit and 64-bit-without-compat compilation.

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Acked-by: David L Stevens <dlstevens@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 years agox86: numaq fix
Ingo Molnar [Wed, 30 Apr 2008 21:05:52 +0000]
x86: numaq fix

do not override the existing pci-y rule when adding visws or
numaq rules.

Signed-off-by: Ingo Molnar <mingo@elte.hu>

11 years agox86: 8K stacks by default
Ingo Molnar [Wed, 30 Apr 2008 18:45:40 +0000]
x86: 8K stacks by default

Switch back to 8K stacks as the safer default. Out-of-memory
situations are less problematic than silent and hard to debug
stack corruption.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

11 years agox86: ioremap ram check fix
Andres Salomon [Wed, 30 Apr 2008 15:30:24 +0000]
x86: ioremap ram check fix

bdd3cee2e4b7279457139058615ced6c2b41e7de (x86: ioremap(), extend check
to all RAM pages) breaks OLPC's ioremap call.  The ioremap that OLPC uses is:

        romsig = ioremap(0xffffffc0, 16);

The commit that breaks it is basically:

-       for (pfn = phys_addr >> PAGE_SHIFT; pfn < max_pfn_mapped &&
-            (pfn << PAGE_SHIFT) < last_addr; pfn++) {
+       for (pfn = phys_addr >> PAGE_SHIFT;
+                               (pfn << PAGE_SHIFT) < last_addr; pfn++) {
+

Previously, the 'pfn < max_pfn_mapped' check would've caused us to not
enter the loop.  Removing that check means we loop infinitely.  The
reason for that is because pfn is 0xfffff, and last_addr is 0xffffffcf.
The remaining check that is used to exit the loop is not sufficient;
when pfn<<PAGE_SHIFT is 0xfffff000, that is less than 0xffffffcf; when
we increment pfn and it overflows (pfn == 0x100000), pfn<<PAGE_SHIFT
ends up being 0.  That, of course, is less than last_addr.  In effect,
pfn<<PAGE_SHIFT is never lower than last_addr.

The simple fix for this is to limit the last_addr check to the PAGE_MASK;
a patch is below.

Signed-off-by: Andres Salomon <dilinger@debian.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

11 years agox86: fix HT cpu booting on 32-bit
Hugh Dickins [Wed, 30 Apr 2008 15:17:46 +0000]
x86: fix HT cpu booting on 32-bit

Since recent smpboot 32/64-bit merge, my dual Xeon with HT has been
booting only 2 of its 4 cpus (when running an i386 kernel; but x86_64
is okay).  J.A. Magallón reports the same.

 native_cpu_up: bad cpu 2
 native_cpu_up: bad cpu 3

The mach-default cpu_present_to_apicid() was just returning cpu number
(2, 3) instead of apicid (6, 7): looks like we now need the x86_64 code
even for the i386 case.

Comparing with other versions of cpu_present_to_apicid(), it seems a
good idea to include an NR_CPUS test too, since cpu_present() doesn't
include that; but that wasn't a problem here, and may no problem at all.

Prior to that smpboot merge, my Xeon booted the two HT siblings on one
physical first, then the two siblings on the other physical after - when
i386, but alternated them when x86_64.  Since the merge, the x86_64
sequence is unchanged, but the i386 sequence is now like x86_64.

I prefer this consistency, and I prefer the new sequence: booting with
maxcpus=2 then uses the independent physicals without HT sharing.

Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

11 years agox86: optimize inlining off
Ingo Molnar [Wed, 30 Apr 2008 08:29:13 +0000]
x86: optimize inlining off

default to inline optimizing off.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

11 years agox86: CONFIG_X86_ELAN fix
Ingo Molnar [Wed, 30 Apr 2008 06:58:27 +0000]
x86: CONFIG_X86_ELAN fix

move the X86_CPU section out of the !X86_ELAN branch.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

11 years agox86: Kconfig fix
Ingo Molnar [Wed, 30 Apr 2008 06:48:45 +0000]
x86: Kconfig fix

Andrew noticed that OPTIMIZE_INLINING appeared in the toplevel
menu - fix it.

Reported-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

11 years agox86 PAT: fix performance drop for glx, use UC minus for ioremap(), ioremap_nocache...
Suresh Siddha [Sat, 26 Apr 2008 00:07:22 +0000]
x86 PAT: fix performance drop for glx, use UC minus for ioremap(), ioremap_nocache() and pci_mmap_page_range()

Use UC_MINUS for ioremap(), ioremap_nocache() instead of strong UC.
Once all the X drivers move to ioremap_wc(), we can go back to strong
UC semantics for ioremap() and ioremap_nocache().

To avoid attribute aliasing issues, pci_mmap_page_range() will also
use UC_MINUS for default non write-combining mapping request.

Next steps:
a) change all the video drivers using ioremap() or ioremap_nocache()
   and adding WC MTTR using mttr_add() to ioremap_wc()

b) for strict usage, we can go back to strong uc semantics
   for ioremap() and ioremap_nocache() after some grace period for
   completing step-a.

c) user level X server needs to use the appropriate method for setting
   up WC mapping (like using resourceX_wc sysfs file instead of
   adding MTRR for WC and using /dev/mem or resourceX under /sys)

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

11 years agox86: use defconfigs from x86/configs/*
Sam Ravnborg [Tue, 29 Apr 2008 10:48:15 +0000]
x86: use defconfigs from x86/configs/*

Daniel Drake <dsd@gentoo.org> reported:

In 2.6.23, if you unpacked a kernel source tarball and then
ran "make menuconfig" you'd be presented with this message:
    # using defaults found in arch/i386/defconfig

and the default options would be set.

The same thing in 2.6.24 does not give you any "using defaults" message, and
the default config options within menuconfig are rather blank (e.g. no PCI
support). You can work around this by explicitly running "make defconfig"
before menuconfig, but it would be nice to have the behaviour the way it was
for 2.6.23 (and the way it still is for other archs).

Fixed by adding a x86 specific defconfig list to Kconfig.

Fixes: http://bugzilla.kernel.org/show_bug.cgi?id=10470
Tested-by: dsd@gentoo.org
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>