Pick an ioctl number for EXT4_IOC_MIGRATE that won't conflict with
other ext4 ioctl's. Since there haven't been any major userspace
users of this ioctl, we can afford to change this now, to avoid
potential problems later.
Also, reorder the ioctl numbers in ext4.h to avoid this sort of
mistake in the future.
ext4: hook the ext3 migration interface to the EXT4_IOC_SETFLAGS ioctl
This patch hooks the ext3 to ext4 migrate interface to
EXT4_IOC_SETFLAGS ioctl. The userspace interface is via chattr +e. We
only allow setting extent flags. Clearing extent flag (migrating from
ext4 to ext3) is not supported.
Signed-off-by: Dmitry Baryshkov <dbaryshkov@gmail.com> Acked-by: Eric Miao <eric.miao@marvell.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
[ARM] 5246/1: tosa: add proper clock alias for tc6393xb clock
Add clock alias for clock that is used by tc6393xb device on tosa.
As that chip plays pretty major part in tosa life and is currently
disabled, this is 2.4.27 material.
Signed-off-by: Dmitry Baryshkov <dbaryshkov@gmail.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Andrew Vasquez [Fri, 12 Sep 2008 05:19:45 +0000 (22:19 -0700)]
[SCSI] qla2xxx: Defer enablement of RISC interrupts until ISP initialization completes.
Josip Rodin noted
(http://article.gmane.org/gmane.linux.ports.sparc/10152) the
driver oopsing during registration of an rport to the
FC-transport layer with a backtrace indicating a dereferencing of
an shost->shost_data equal to NULL. David Miller identified a
small window in driver logic where this could happen:
> Look at how the driver registers the IRQ handler before the host has
> been registered with the SCSI layer.
>
> That leads to a window of time where the shost hasn't been setup
> fully, yet ISRs can come in and trigger DPC thread events, such as
> loop resyncs, which expect the transport area to be setup.
>
> But it won't be setup, because scsi_add_host() hasn't finished yet.
>
> Note that in Josip's crash log, we don't even see the
>
> qla_printk(KERN_INFO, ha, "\n"
> " QLogic Fibre Channel HBA Driver: %s\n"
> " QLogic %s - %s\n"
> " ISP%04X: %s @ %s hdma%c, host#=%ld, fw=%s\n",
> ...
>
> message yet.
>
> Which means that the crash occurs between qla2x00_request_irqs()
> and printing that message.
Close this window by enabling RISC interrupts after the host has
been registered with the SCSI midlayer.
Reported-by: Josip Rodin <joy@entuzijast.net> Cc: Stable Tree <stable@kernel.org> Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Alexander Duyck [Sat, 13 Sep 2008 00:57:23 +0000 (17:57 -0700)]
multiq: Further multiqueue cleanup
This patch resolves a few issues found with multiq including wording
suggestions and a problem seen in the allocation of queues.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Sat, 13 Sep 2008 00:56:50 +0000 (17:56 -0700)]
skbedit: Fix a typo in the documentation
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Tony Lindgren [Fri, 12 Sep 2008 23:32:15 +0000 (16:32 -0700)]
ARM: OMAP: Fix cpu detection
At some point while adding omap2 support, IS_OMAP_CLASS broke as pointed out
by Russell King. Fix this to work as it used to work, and modify omap24xx and
omap34xx detection to work accordingly.
Also rename OMAP3430_REV_ES2_2 to OMAP3430_REV_ES3_0.
Alexander Duyck [Fri, 12 Sep 2008 23:30:20 +0000 (16:30 -0700)]
pkt_action: add new action skbedit
This new action will have the ability to change the priority and/or
queue_mapping fields on an sk_buff.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Fri, 12 Sep 2008 23:29:34 +0000 (16:29 -0700)]
pkt_sched: Add multiqueue scheduler support
This patch is intended to add a qdisc to support the new tx multiqueue
architecture by providing a band for each hardware queue. By doing
this it is possible to support a different qdisc per physical hardware
queue.
This qdisc uses the skb->queue_mapping to select which band to place
the traffic onto. It then uses a round robin w/ a check to see if the
subqueue is stopped to determine which band to dequeue the packet from.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
inet6_rsk() is called on a struct request_sock * before we
have checked whether the socket is an ipv6 socket or a ipv6-
mapped ipv4 socket. The access that triggers this is the
inet_rsk(rsk)->inet6_rsk_offset dereference in inet6_rsk().
This is arguably not a critical error as the inet6_rsk_offset
is only used to compute a pointer which is never really used
(in the code path in question) anyway. But it might be a
latent error, so let's fix it.
Spotted by kmemcheck.
Signed-off-by: Vegard Nossum <vegard.nossum@gmail.com> Acked-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Benjamin Thery [Fri, 12 Sep 2008 23:16:37 +0000 (16:16 -0700)]
net: fix scheduling of dst_gc_task by __dst_free
The dst garbage collector dst_gc_task() may not be scheduled as we
expect it to be in __dst_free().
Indeed, when the dst_gc_timer was replaced by the delayed_work
dst_gc_work, the mod_timer() call used to schedule the garbage
collector at an earlier date was replaced by a schedule_delayed_work()
(see commit 86bba269d08f0c545ae76c90b56727f65d62d57f).
But, the behaviour of mod_timer() and schedule_delayed_work() is
different in the way they handle the delay.
mod_timer() stops the timer and re-arm it with the new given delay,
whereas schedule_delayed_work() only check if the work is already
queued in the workqueue (and queue it (with delay) if it is not)
BUT it does NOT take into account the new delay (even if the new delay
is earlier in time).
schedule_delayed_work() returns 0 if it didn't queue the work,
but we don't check the return code in __dst_free().
If I understand the code in __dst_free() correctly, we want dst_gc_task
to be queued after DST_GC_INC jiffies if we pass the test (and not in
some undetermined time in the future), so I think we should add a call
to cancel_delayed_work() before schedule_delayed_work(). Patch below.
Or we should at least test the return code of schedule_delayed_work(),
and reset the values of dst_garbage.timer_inc and dst_garbage.timer_expires
back to their former values if schedule_delayed_work() failed.
Otherwise the subsequent calls to __dst_free will test the wrong values
and assume wrong thing about when the garbage collector is supposed to
be scheduled.
dst_gc_task() also calls schedule_delayed_work() without checking
its return code (or calling cancel_scheduled_work() first), but it
should fine there: dst_gc_task is the routine of the delayed_work, so
no dst_gc_work should be pending in the queue when it's running.
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net> Acked-by: Eric Dumazet <dada1@cosmosbay.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Santwona Behera [Fri, 12 Sep 2008 23:04:26 +0000 (16:04 -0700)]
niu: panic on reset
The reset_task function in the niu driver does not reset the tx and rx
buffers properly. This leads to panic on reset. This patch is a
modified implementation of the previously posted fix.
Signed-off-by: Santwona Behera <santwona.behera@sun.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Paul Walmsley [Wed, 10 Sep 2008 16:48:04 +0000 (10:48 -0600)]
OMAP2 clockdomain: add virt_opp_clkdm
Every OMAP2/3 clock will need to be associated with a clockdomain.
However, the "virt_prcm_set" clock has no clockdomain, since it is a
virtual clock without any hardware referents. So, create a new
clockdomain, "virt_clkdm", for this clock. This clockdomain should be
reusable for OMAP3 virtual clock nodes.
Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
Paul Walmsley [Wed, 10 Sep 2008 16:48:01 +0000 (10:48 -0600)]
OMAP2/3 clock: note clockdomains for remaining clocks
Fill in clockdomains for all remaining clocks. On OMAP2, these mostly
involve IVA and DSP clocks; on OMAP3, these mostly involve some core_l4_clkdm
devices and the secure peripherals.
Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
Each DPLL exists in its own powerdomain (cf 34xx TRM figure 4-18) and
clockdomain; so, create powerdomain and clockdomain structures for them.
These are used in a following patch for DPLL-related clocks.
Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
Paul Walmsley [Wed, 10 Sep 2008 16:47:46 +0000 (10:47 -0600)]
OMAP2/3 clock: convert wkup_clkdm CM clocks to cm_clkdm
Convert existing wkup_clkdm clocks that should be in the CM clockdomain
to cm_clkdm. (A later patch will add CM clockdomain associations for
unassociated clocks.)
Paul Walmsley [Wed, 10 Sep 2008 16:47:39 +0000 (10:47 -0600)]
OMAP2/3 clock: convert wkup_clkdm PRM clocks to prm_clkdm
Convert existing wkup_clkdm clocks that should be in the PRM clockdomain
to prm_clkdm. (A later patch will add PRM clockdomain associations for
unassociated clocks.)
Viktor Rosendahl [Thu, 11 Sep 2008 16:20:42 +0000 (19:20 +0300)]
lowmem: Remove the possibility to stack the module
The mod_reg_security() function that is used to stack modules has been removed:
"LSMs wishing to stack capability as a secondary module should do so
explicitly". Thus, let's remove the possibility to stack in the init function.
The functionality could probably be reintroduced later if somebody has time to
implement it.
Signed-off-by: Viktor Rosendahl <viktor.rosendahl@nokia.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
Viktor Rosendahl [Thu, 11 Sep 2008 16:20:40 +0000 (19:20 +0300)]
lowmem: add compatibility code
This is a patch that will reintroduce the old sysctl lowmem variables as read
only values in /proc/sys/vm. Some parts of the userland (at least libosso) are
reading these and it would probably greatly ease the transition to the new
scheme with memory limits in pages. We can revert this ugly patch when userland
has moved to the new scheme.
Signed-off-by: Viktor Rosendahl <viktor.rosendahl@nokia.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
Viktor Rosendahl [Thu, 11 Sep 2008 16:20:39 +0000 (19:20 +0300)]
lowmem: Security modules can no longer be removed.
The unregister_security() and mod_unreg_security() functions were removed
because the upstream developers believe that it should not be possible to remove
security modules, "It is used only by out-of-tree modules, which are often
binary-only, illegal, abusive of the API and dangerous, e.g. silently
re-vectoring SELinux."
Signed-off-by: Viktor Rosendahl <viktor.rosendahl@nokia.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
Viktor Rosendahl [Thu, 11 Sep 2008 16:20:36 +0000 (19:20 +0300)]
lowmem: remove an unecessary local variable
This "changed" variable is kind of deprecated. We used to have a spinlock
[which was later deemed unecessary] around the first if clause in the
*_watermark_state() functions; the variable was needed in order to move the
operation in the second if clause out of the lock section.
Signed-off-by: Viktor Rosendahl <viktor.rosendahl@nokia.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
Viktor Rosendahl [Thu, 11 Sep 2008 16:20:35 +0000 (19:20 +0300)]
lowmem: get rid of the percentages
The purpose of this patch is to get rid of the lame percentage calculations in
lowmem.c. Currently, the limits (measured in pages) are calculated from
percentages every time somebody from userspace requests some memory. With this
patch, the limits are in pages and furthermore, they are specified as minimum
amount of "free pages", instead of as maximum amount of "used pages". "Free"
means free pages or such pages that can easily be freed by the VM system.
This patch changes the names of the lowmem sysctl limits in /proc/sys/vm:
lowmem_deny_watermark => lowmem_deny_watermark_pages
lowmem_notify_low => lowmem_notify_low_pages
lowmem_notify_high => lowmem_notify_high_pages
The following read only value disappears:
lowmem_used_pages
The following read only value is introduced:
lowmem_free_pages
The old value can be calculated from userspace with the following pseudocode:
Note that the value of lowmem_free_pages is only accurate when the system is
rather low on memory (otherwise it's a bit too pessimistic), this was true also
for the lowmem_used_pages metric.
Signed-off-by: Viktor Rosendahl <viktor.rosendahl@nokia.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
Viktor Rosendahl [Thu, 11 Sep 2008 16:20:34 +0000 (19:20 +0300)]
lowmem: remove unecessary usage of proc_dointvec_minmax()
As far as I can understand, using proc_dointvec_minmax() instead of
proc_dointvec only makes sense if you supply the min and/or max values in the
extra1 and extra2 fields of the struct.
Signed-off-by: Viktor Rosendahl <viktor.rosendahl@nokia.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
Henrik Saari [Thu, 11 Sep 2008 16:20:33 +0000 (19:20 +0300)]
Security: Kernel OOM-killer and allocation denial use different rules
Security: Kernel OOM-killer and allocation denial use different rules
Original patch created by Leonid Moiseichuk <leonid.moiseichuk@nokia.com>
Fixes the issue with the oom killer and lowmem module being inconsistent;
processes that are important enough to be protected from the oom killer should
not be denied memory either.
Signed-off-by: Henrik Saari <henrik.saari@nokia.com> Signed-off-by: Viktor Rosendahl <viktor.rosendahl@nokia.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
David Brownell [Thu, 11 Sep 2008 09:48:39 +0000 (02:48 -0700)]
twl4030: improve rtc device setup
Make the twl4030 RTC initialization follow the driver model better.
The platform device is created (if needed) as part of twl4030 setup
instead of as a board-specific thing, and fits properly into the
driver model tree.
The only minor glitch here is on boards using platform_data to pass
MSECURE initialization hooks to the RTC driver. The right solution
probably just removes that platform_data, and might even make Linux
follow the principle of "least privilege" (at the hardware level!)
instead of always holding this signal high.
Pending better handling of MSECURE, this patch just does what the
RTC init does, but earlier: MSECURE is always high, so the RTC time
and calendar registers can be updated (and presumably a bunch of
non-RTC privileged operations will be allowed too).
Yet to be done: set up the IRQ resource; make the rtc driver use
that IRQ resource; and properly issue wakeup alarms.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Tony Lindgren <tony@atomide.com>
David Brownell [Thu, 11 Sep 2008 06:01:24 +0000 (23:01 -0700)]
mach-omap2: fix more arch_initcall() breakage
Remove more bogus arch_initcall() logic in mach-omap2/board-xyx.c files.
They broke a multi-OMAP build I did, at *RUN TIME* not build time, since
it tried to do the i2c init for every board linked in the kernel.
Remember, init_machine() entries run at arch_initcall() time; that's
where any board-specific init logic should normally go. Any initcalls
in the mach-*/*c files should normally be guarded by tests to make sure
they only run on the relevant hardware (board, cpu). Better yet, get
rid of the initcalls; init_machine() can *explicitly* call the right
version of that code, and pass in board-specific config data; and there
are hooks that can handle cpu-specific stuff too.
A quick glance suggests most of the remaining initcall logic in the
mach-omap2 directory is similarly broken... this patch gets rid of
one frequently-cloned idiom, it should help.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Tony Lindgren <tony@atomide.com>
Paul Mundt [Fri, 12 Sep 2008 14:27:46 +0000 (23:27 +0900)]
sh: Tidy up ELF core dumps.
These have been using overrides for ELF_CORE_COPY_TASK_REGS and
ELF_CORE_COPY_FPREGS while the generic versions can be used instead.
Presently the pt_regs are also duplicated across elf_core_copy_regs()
and elf_core_copy_task_regs(), this switches to simply copying out
through elf_core_copy_regs() instead.
Such a crash is possible if the CPU in this system is a 64-bit
processor which doesn't support NX (ie, old Intel P4 -based64-bit
processors).
Certainly, if we support such processors, then we should start with
_PAGE_NX initially clear in __supported_pte_flags, and then set it once
we've established that the processor does indeed support NX. That will
prevent early_ioremap - or anything else - from trying to set it.
The simple fix is to simply call check_efer() earlier.
Reported-by: Russ Anderson <rja@sgi.com> Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
The panic code called the kexec code which called mutex_trylock() which
called spin_lock_mutex() which then stupidly went and blurted a load of
debug stuff because of in_interrupt().
Keep the lock debug code from escallating an already crappy situation.
David S. Miller [Fri, 12 Sep 2008 06:57:40 +0000 (23:57 -0700)]
sparc64: Fix sparse warnings in pci.c
1) Declare pci_poke_* in pci_impl.h
2) of_create_pci_dev() should be static
3) ->setup_msi_irq() wants an unsigned int pointer not a plain
int one
4) void value expression return in arch_teardown_msi_irq()
Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Fri, 12 Sep 2008 03:17:05 +0000 (20:17 -0700)]
vlan: vlan device not reading gso max size of parent.
The vlan devices are not reading the gso max size of the parent device. As
a result devices that do not support 64K max gso size are currently
failing.
This issue is seen on 2.6.26 kernels as well and the same patch should be
able to be applied without any issues.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
/**
* nla_ok - check if the netlink attribute fits into the remaining bytes
* @nla: netlink attribute
* @remaining: number of bytes remaining in attribute stream
*/
static inline int nla_ok(const struct nlattr *nla, int remaining)
{
return remaining >= sizeof(*nla) &&
nla->nla_len >= sizeof(*nla) &&
nla->nla_len <= remaining;
}
It turns out that remaining can become negative due to alignment in
nla_next(). But GCC promotes "remaining" to unsigned in the test
against sizeof(*nla) above. Therefore the test succeeds, and the
nla_for_each_attr() may access memory outside the received buffer.
A short example illustrating this point is here:
#include <stdio.h>
main(void)
{
printf("%d\n", -1 >= sizeof(int));
}
...which prints "1".
This patch adds a cast in front of the sizeof so that GCC will make
a signed comparison and fix the illegal memory dereference. With the
patch applied, there is no kmemcheck report.
Signed-off-by: Vegard Nossum <vegard.nossum@gmail.com> Acked-by: Thomas Graf <tgraf@suug.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
[Bluetooth] Fix regression from using default link policy
To speed up the Simple Pairing connection setup, the support for the
default link policy has been enabled. This is in contrast to settings
the link policy on every connection setup. Using the default link policy
is the preferred way since there is no need to dynamically change it for
every connection.
For backward compatibility reason and to support old userspace the
HCISETLINKPOL ioctl has been switched over to using hci_request() to
issue the HCI command for setting the default link policy instead of
just storing it in the HCI device structure.
However the hci_request() can only be issued when the device is
brought up. If used on a device that is registered, but still down
it will timeout and fail. This is problematic since the command is
put on the TX queue and the Bluetooth core tries to submit it to
hardware that is not ready yet. The timeout for these requests is
10 seconds and this causes a significant regression when setting up
a new device.
The userspace can perfectly handle a failure of the HCISETLINKPOL
ioctl and will re-submit it later, but the 10 seconds delay causes
a problem. So in case hci_request() is called on a device that is
still down, just fail it with ENETDOWN to indicate what happens.
Tony Lindgren [Fri, 12 Sep 2008 00:23:52 +0000 (17:23 -0700)]
ARM: OMAP: Clean-up MMC device init
Clean-up MMC device init:
- Initialize devices in mach-omap1/devices.c and mach-omap2/devices.c
instead of plat-omap/devices.c
- Remove old struct omap_mmc_config, use struct omap_mmc_platform_data instead
Note that this removes the second omap-hsmmc controller init for overo.
This needs to be fixed properly by changing hsmmc_init() to accept
board specific struct omap_mmc_platform_data.
Anna Neal [Thu, 11 Sep 2008 18:17:25 +0000 (11:17 -0700)]
libertas: Improvements on automatic tx power control via SIOCSIWTXPOW.
iwconfig txpower can now be used to set tx power to fixed or auto. If set to
auto the default firmware settings are used.
The command CMD_802_11_PA_CFG is only sent to older firmware, as Dan Williams
noted the command was no longer supported in firmware V9+.
Signed-off-by: Anna Neal <anna@cozybit.com> Signed-off-by: Javier Cardona <javier@cozybit.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
mac80211: Reorder debugfs calls during netdev deinit
ieee80211_free_keys() must be called before
ieee80211_debugfs_remove_netdev() in order to make sure that the
possible default_key symlink is removed before attempting to
remove the netdev debugfs directory.
Signed-off-by: Jouni Malinen <jouni.malinen@atheros.com> Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Johannes Berg [Tue, 9 Sep 2008 12:55:09 +0000 (14:55 +0200)]
mac80211: make ieee80211_rx_h_mgmt more readable
That function isn't exactly easy to read especially since it
does something in an if branch that continues after the if
because the else returns. Express it in a more readable way.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Johannes Berg [Tue, 9 Sep 2008 12:49:03 +0000 (14:49 +0200)]
mac80211: move spectrum management code out
Like the HT code, this doesn't depend on the STA-mode implementation
and can be handled entirely independently. There's only stub code
for now, but when it gets filled having it in its own file will be
beneficial.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>