Since commit c7f404b ('vfs: new superblock methods to override
/proc/*/mount{s,info}'), nfs_path() is used to generate the mounted
device name reported back to userland.
nfs_path() always generates a trailing slash when the given dentry is
the root of an NFS mount, but userland may expect the original device
name to be returned verbatim (as it used to be). Make this
canonicalisation optional and change the callers accordingly.
[[email protected]: use flag instead of bool argument]
Reported-and-tested-by: Chris Hiestand <[email protected]>
Reference: http://bugs.debian.org/669314
Signed-off-by: Ben Hutchings <[email protected]>
Cc: <[email protected]> # v2.6.39+
Signed-off-by: Jonathan Nieder <[email protected]>
Signed-off-by: Trond Myklebust <[email protected]>
These arrays are accessed by iteration in
llc_exec_station_trans_actions(). There must not be any zero-filled
gaps in them, so the explicit indices are pointless.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
We only ever put one skb on the send queue, and then immediately
send it. Remove the queue and call dev_queue_xmit() directly.
This leaves struct llc_station empty, so remove that as well.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
We only ever put one skb on the event queue, and then immediately
process it. Remove the queue and fold together the related functions,
removing several blatantly false comments.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
The initial state is UP and there is no way to enter the other states
as the required event type is never generated. Delete all states,
event types, and other dead code. The only thing left is handling
of the XID and TEST commands.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
softsynth_read() reads a character at a time from the init string;
when it finds the null terminator it sets the initialized flag but
then repeats the last character.
Additionally, if the read() buffer is not big enough for the init
string, the next read() will start reading from the beginning again.
So the caller may never progress to reading anything else.
Replace the simple initialized flag with the current position in
the init string, carried over between calls. Switch to reading
real data once this reaches the null terminator.
(This assumes that the length of the init string can't change, which
seems to be the case. Really, the string and position belong together
in a per-file private struct.)
Tested-by: Samuel Thibault <[email protected]>
Signed-off-by: Ben Hutchings <[email protected]>
Cc: stable <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
ite_dev::rdev is currently initialised in ite_probe() after
rc_register_device() returns. If a newly registered device is opened
quickly enough, we may enable interrupts and try to use ite_dev::rdev
before it has been initialised. Move it up to the earliest point we
can, right after calling rc_allocate_device().
Reported-and-tested-by: YunQiang Su <[email protected]>
Signed-off-by: Ben Hutchings <[email protected]>
Cc: [email protected]
Signed-off-by: Mauro Carvalho Chehab <[email protected]>
The JMicron JMB362 controller supports AHCI only, but some revisions
use the IDE class code. These need to be matched by device ID.
These additions have apparently been included by QNAP in their NAS
devices using these controllers.
References: http://bugs.debian.org/634180
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
There is a new convention, used by systemd and supported by most
distributions, to put basic OS release information in /etc/os-release.
Added some additional error checking on strdup()
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: K. Y. Srinivasan <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
hv_kvp_daemon currently does not check whether fread() or fwrite()
succeed. Add the necessary checks. Also, remove the incorrect use of
feof() before fread().
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: K. Y. Srinivasan <[email protected]>
Cc: [email protected]
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Linux native exit codes are 8-bit unsigned values. exit(-1) results
in an exit code of 255, which is usually reserved for shells reporting
'command not found'. Use the portable value EXIT_FAILURE. (Not that
this matters much for a daemon.)
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: K. Y. Srinivasan <[email protected]>
Cc: [email protected]
Signed-off-by: Greg Kroah-Hartman <[email protected]>
When we use a separate output directory, we add util/ to the include
path for the generated C files. However, this is currently added to the
end of the path, behind /usr/include/slang and /usr/include/gtk-2.0 if
use of the respective libraries is enabled. Thus the '#include
"../perf.h"' in util/parse-events.l can actually include
/usr/include/perf.h if it exists.
Move '-Iutil/' ahead of all the other preprocessor options.
Reported-by: Sedat Dilek <[email protected]>
Signed-off-by: Ben Hutchings <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Paul Mackerras <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: Sedat Dilek <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
When registering the handlers, any state they rely on must be
completely initialised first. When unregistering, we must wait until
they are definitely no longer running. llc_rcv() must also avoid
reading the handler pointers again after checking for NULL.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Otherwise the station packet handler will remain registered even though
the module is unloaded.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
llc_station_init() creates and processes an event skb with no effect
other than to change the state from DOWN to UP. Allocation failure is
reported, but then ignored by its caller, llc2_init(). Remove this
possibility by simply initialising the state as UP.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
ZSMALLOC is tristate, but the code has no MODULE_LICENSE and since it
depends on GPL-only symbols it cannot be loaded as a module. This in
turn breaks zram which now depends on it. I assume it's meant to be
Dual BSD/GPL like the other z-stuff.
There is also no module_exit, which will make it impossible to unload.
Add the appropriate module_init and module_exit declarations suggested
by comments.
Reported-by: Christian Ohm <[email protected]>
References: http://bugs.debian.org/677273
Cc: [email protected] # v3.4
Signed-off-by: Ben Hutchings <[email protected]>
Reviewed-by: Jonathan Nieder <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
fintek-cir, ite-cir and nuvoton-cir may try to free an I/O region
and/or IRQ handler that was never allocated after a failure in their
respective probe functions. Add and use separate labels on the
failure path so they will do the right cleanup after each possible
point of failure.
Compile-tested only.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Mauro Carvalho Chehab <[email protected]>
Currently cfg80211 fails to create a "phy80211" symlink in sysfs from
the net device to the wiphy device. The latter needs to be registered
first.
Compile-tested only.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Currently cfg80211 fails to create a "phy80211" symlink in sysfs from
the net device to the wiphy device. The latter needs to be registered
first.
Compile-tested only.
Reported-by: Cesare Leonardi <[email protected]>
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Commit 554cdaefd1 ('ARM: orion5x: Refactor
mpp code to use common orion platform mpp.') seems to have accidentally
inverted the GPIO valid bits for MPP9 (only). For the mv2120 platform
which uses MPP9 as a GPIO LED device, this results in the error:
[ 12.711476] leds-gpio: probe of leds-gpio failed with error -22
Reported-by: Henry von Tresckow <[email protected]>
References: http://bugs.debian.org/667446
Signed-off-by: Ben Hutchings <[email protected]>
Cc: [email protected] [v3.0+]
Tested-by: Hans Henry von Tresckow <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
This function is defined for use in exec, not in modules.
No other architecture exports its implementation.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
We must not allow the input buffer length to change while we're
shuffling the buffer contents. We also mustn't clear the WDM_READ
flag after more data might have arrived. Therefore move both of these
into the spinlocked region at the bottom of wdm_read().
When reading desc->length without holding the iuspin lock, use
ACCESS_ONCE() to ensure the compiler doesn't re-read it with
inconsistent results.
Signed-off-by: Ben Hutchings <[email protected]>
Tested-by: Bjørn Mork <[email protected]>
Cc: Oliver Neukum <[email protected]>
Cc: stable <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Commit b66b8b9a4a ('intel-idle: convert
to x86_cpu_id auto probing') added a distinction between Nehalem and
Westemere processors and changed auto_demotion_disable_flags for the
former to 0. This was not explained in the commit message, so change
it back.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Thomas Renninger <[email protected]>
Acked-by: H. Peter Anvin <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Commit fa8031aefe ('cpufreq: Add support
for x86 cpuinfo auto loading v4') added a device ID table to this
driver, but didn't declare it as the module device ID table.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Thomas Renninger <[email protected]>
Acked-by: H. Peter Anvin <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Commit fa8031aefe ('cpufreq: Add support
for x86 cpuinfo auto loading v4') seems to have inadvertently changed
the matched CPU family number from 6 to 7. Change it back.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Thomas Renninger <[email protected]>
Acked-by: H. Peter Anvin <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Commit b66b8b9a4a ('intel-idle: convert
to x86_cpu_id auto probing') put two entries for model 0x2f
(Westmere-EX Xeon) in the device ID table and left out model 0x2e
(Nehalem-EX Xeon).
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Thomas Renninger <[email protected]>
Acked-by: H. Peter Anvin <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
We currently include commas on both sides of the feature ID in a
modalias, but this prevents the lowest numbered feature of a CPU from
being matched. Since all feature IDs have the same length, we do not
need to worry about substring matches, so omit commas from the
modalias entirely.
Avoid generating multiple adjacent wildcards when there is no
feature ID to match.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Thomas Renninger <[email protected]>
Acked-by: H. Peter Anvin <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
This parameter blob and generator program have been moved to the
linux-firmware.git repository.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Dan Williams <[email protected]>
Signed-off-by: James Bottomley <[email protected]>
frv, h8300, m68k, microblaze, openrisc, score, um and xtensa currently
do not register a CPU device. Add the config option GENERIC_CPU_DEVICES
which causes a generic CPU device to be registered for each present CPU,
and make all these architectures select it.
Richard Weinberger <[email protected]> covered UML and suggested using
per_cpu.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
cpu_dev_init() is only called from driver_init(), which does not check
its return value. Therefore make cpu_dev_init() return void.
We must register the CPU subsystem, so panic if this fails.
If sched_create_sysfs_power_savings_entries() fails, the damage is
contained, so ignore this (as before).
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Commit 5b7c840667 ('ipv4: correct IGMP
behavior on v3 query during v2-compatibility mode') added yet another
case for query parsing, which can result in max_delay = 0. Substitute
a value of 1, as in the usual v3 case.
Reported-by: Simon McVittie <[email protected]>
References: http://bugs.debian.org/654876
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Currently the value reported for max_batch_time is really the
value of min_batch_time.
Reported-by: Russell Coker <[email protected]>
Signed-off-by: Ben Hutchings <[email protected]>
This allows uswsusp built for i386 to run on an x86_64 kernel (tested
with Debian package version 1.0+20110509-2).
References: http://bugs.debian.org/502816
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Rafael J. Wysocki <[email protected]>
David and I are sharing maintenance of this repository. Patches
should be sent to both of us.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Commit ded19addf9 ('pasemic_mac*: Move
the PA Semi driver') inadvertently split pasemi_mac into two separate
modules with unresolved symbols. Change it back into a single module.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Support for specific hardware belongs under drivers/net/ not net/.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Lennert Buytenhek <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Any headers included by drivers should be under include/, and
any definitions they use are not really private to the core as
the name "dsa_priv.h" suggests.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Lennert Buytenhek <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
I mistakenly exported functions from slave.c that are only called from
dsa.c, part of the same module.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Lennert Buytenhek <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
We should provide topology information to userland even if it's not
very interesting. The current code appears to work properly for !SMP
(tested on i386).
Reference: http://bugs.debian.org/649216
Reported-by: Marcus Osdoba <[email protected]>
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Change the kconfig types to tristate and adjust the condition for
declaring net_device::dsa_ptr to allow for this.
Adjust the makefile so that if NET_DSA_MV88E6123_61_65=y and
NET_DSA_MV88E6131=m or vice versa then both drivers are built-in. We
could leave these options as bool and make NET_DSA_MV88E6XXX a
user-selected option, but that would break existing configurations.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
These drivers share a lot of code, so if we make them modular they
should be built into the same module. Therefore, link them together
and merge their respective module init and exit functions.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
These files have circular dependencies, so if we make DSA modular then
they must be built into the same module. Therefore, link them
together and merge their respective module init and exit functions.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
eth_type_trans() will use these functions if DSA is enabled, which
blocks building DSA as a module.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
commit 88491d8103 ("drivers/net: Kconfig
& Makefile cleanup") changed the type of these options to bool, but
they select code that could (and still can) be built as modules.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
lirc_serial_probe() must fail if request_irq() returns an error, even if
it isn't EBUSY or EINVAL,
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Mauro Carvalho Chehab <[email protected]>
Device not found? ENODEV, not EINVAL.
Write to read-only device? EPERM, not EBADF.
Invalid argument? EINVAL, not ENOSYS.
Unsupported ioctl? ENOIOCTLCMD, not ENOSYS.
Another function returned an error code? Use that, don't replace it.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Mauro Carvalho Chehab <[email protected]>
A resume function cannot remove the device it is resuming!
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Mauro Carvalho Chehab <[email protected]>
Failure to allocate the I/O region leaves the IRQ allocated.
A later failure leaves them both allocated.
Reported-by: Torsten Crass <[email protected]>
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Mauro Carvalho Chehab <[email protected]>
Currently the module init function registers a platform_device and
only then allocates its IRQ and I/O region. This allows allocation to
race with the device's suspend() function. Instead, allocate
resources in the platform driver's probe() function and free them in
the remove() function.
The module exit function removes the platform device before the
character device that provides access to it. Change it to reverse the
order of initialisation.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Mauro Carvalho Chehab <[email protected]>
This file introduced in 2.6.32.47 currently fails to compile:
arch/sparc/kernel/sigutil_64.c: In function 'save_fpu_state':
arch/sparc/kernel/sigutil_64.c:25: error: 'EFAULT' undeclared (first use in this function)
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Even after commit 5478755616
("block: check for proper length of iov entries earlier ...")
we still won't check for zero-length entries after an unaligned
entry. Remove the break-statement, so all entries are checked.
Signed-off-by: Ben Hutchings <[email protected]>
Cc: [email protected]
Signed-off-by: Jens Axboe <[email protected]>
Use of the GPL or a compatible licence doesn't necessarily make the code
any good. We already consider staging modules to be suspect, and this
should also be true for out-of-tree modules which may receive very
little review.
Signed-off-by: Ben Hutchings <[email protected]>
Reviewed-by: Dave Jones <[email protected]>
Acked-by: Greg Kroah-Hartman <[email protected]>
Signed-off-by: Rusty Russell <[email protected]> (patched oops-tracing.txt)
Dynamic debugging is currently disabled for tainted modules, except
for TAINT_CRAP. This prevents use of dynamic debugging for
out-of-tree modules once the next patch is applied.
This condition was apparently intended to avoid a crash if a force-
loaded module has an incompatible definition of dynamic debug
structures. However, a administrator that forces us to load a module
is claiming that it *is* compatible even though it fails our version
checks. If they are mistaken, there are any number of ways the module
could crash the system.
As a side-effect, proprietary and other tainted modules can now use
dynamic_debug.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Mathieu Desnoyers <[email protected]>
Signed-off-by: Rusty Russell <[email protected]>
Commit 9676001559
("ALSA: fm801: add error handling if auto-detect fails") seems to
break systems that were previously working without a tuner.
As a bonus, this should fix init and cleanup for the case where the
tuner is explicitly disabled.
Reported-and-tested-by: Hor Jiun Shyong <[email protected]>
References: http://bugs.debian.org/641946
Signed-off-by: Ben Hutchings <[email protected]>
Cc: [email protected] [v3.0+]
Signed-off-by: Takashi Iwai <[email protected]>
Commit 9676001559
("ALSA: fm801: add error handling if auto-detect fails") added
incorrect error handling.
Once we have successfully called snd_device_new(), the cleanup
function fm801_free() will automatically be called by snd_card_free()
and we must *not* also call fm801_free() directly.
Reported-by: Hor Jiun Shyong <[email protected]>
References: http://bugs.debian.org/641946
Signed-off-by: Ben Hutchings <[email protected]>
Cc: [email protected] [v3.0+]
Signed-off-by: Takashi Iwai <[email protected]>
Since GPIOLIB is optional on ia64, GENERIC_GPIO must not be selected by
default. If GPIOLIB is enabled, it will select GENERIC_GPIO.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Tony Luck <[email protected]>
<linux/can/bcm.h> uses type canid_t, defined in <linux/can.h>.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Oliver Hartkopp <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Various headers use union nf_inet_addr, defined in <linux/netfilter.h>.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Patrick McHardy <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Various headers use INT_MIN and INT_MAX, which are defined for
userland in <limits.h>.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Patrick McHardy <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
<linux/tipc_config.h> defines inline functions using ntohs() etc.
For userland these are defined in <arpa/inet.h>.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
These types are guaranteed to be defined by <linux/types.h> for
both userland and kernel, unlike u_intN_t.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Patrick McHardy <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Complete the work started with commit
6602a4baf4 ('net: Make userland include
of netlink.h more sane').
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
<linux/types.h> defines __kernel_pid_t for userland; pid_t is
defined elsewhere (and potentially differently).
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
These headers use the ax25_address type defined in <linux/ax25.h>.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Ralf Baechle <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
<linux/if_ppox.h> uses ETH_ALEN, defined in <linux/if_ether.h>.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
gas used to accept (and ignore?) .size directives which referred to
undefined symbols, as these do. In binutils 2.21 these are treated
as fatal errors.
The issue in proc-arm7tdmi.S was also fixed independently by Peter
Chubb.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Uwe Kleine-König <[email protected]>
Signed-off-by: Russell King <[email protected]>
The driver can be built as either a PCI or platform (OF) driver. It
might make sense to built it as both (though that is not currently
possible) but it certainly doesn't make sense to build it as neither!
Add dependencies and add a choice group to ensure that exactly one of
FB_MB862XX_PCI_GDC and FB_MB862XX_LIME is selected.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Anatolij Gustschin <[email protected]>
Signed-off-by: Paul Mundt <[email protected]>
Commit b987812b3f left
crash_kexec_wait_realmode() undefined for UP.
Commit 7c7a81b53e defined it for UP but
left it undefined for 32-bit SMP.
Seems like people are getting confused by nested #ifdef's, so move the
definitions of crash_kexec_wait_realmode() after the #ifdef CONFIG_SMP
section.
Compile-tested with 32-bit UP, 32-bit SMP and 64-bit SMP configurations.
Signed-off-by: Ben Hutchings <[email protected]>
Tested-by: Paul Gortmaker <[email protected]>
Signed-off-by: Benjamin Herrenschmidt <[email protected]>
Commit b987812b3f left
crash_kexec_wait_realmode() undefined for UP.
Commit 7c7a81b53e defined it for UP but
left it undefined for 32-bit SMP.
Seems like people are getting confused by nested #ifdef's, so move the
definitions of crash_kexec_wait_realmode() after the #ifdef CONFIG_SMP
section.
Compile-tested with 32-bit UP, 32-bit SMP and 64-bit SMP configurations.
Signed-off-by: Ben Hutchings <[email protected]>
Tested-by: Paul Gortmaker <[email protected]>
Signed-off-by: Benjamin Herrenschmidt <[email protected]>
Commit b02d0ed677 ('xhci: Change
hcd_priv into a pointer') added calls to kzalloc() and kfree() in
xhci-pci.c. On most architectures <linux/slab.h> is indirectly
included, but on some it is not.
Signed-off-by: Ben Hutchings <[email protected]>
Cc: Sarah Sharp <[email protected]>,
Signed-off-by: Greg Kroah-Hartman <[email protected]>
gas used to accept (and ignore?) .size directives which referred to
undefined symbols, as this does. In binutils 2.21 these are treated
as fatal errors.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Eric Miao <[email protected]>
Commit 40aee729b3 ('kconfig: fix default value for choice input')
fixed some cases where kconfig would select the wrong option from a
choice with a single valid option and thus enter an infinite loop.
However, this broke the test for user input of the form 'N?', because
when kconfig selects the single valid option the input is zero-length
and the test will read the byte before the input buffer. If this
happens to contain '?' (as it will in a mips build on Debian unstable
today) then kconfig again enters an infinite loop.
Signed-off-by: Ben Hutchings <[email protected]>
Cc: [email protected] [2.6.17+]
Signed-off-by: Linus Torvalds <[email protected]>
My changes in commit 4d42d417be were
written some time before the introduction of FLAG_POINTTOPOINT, so
didn't include that flag in the new driver_info. Change the new
driver_info to be consistent.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Some RNDIS devices don't respond on the control channel until polled
on the status channel. In particular, this was reported to be the
case for the 2Wire HomePortal 1000SW and for some Windows Mobile
devices.
This is roughly based on a patch by John Carr <[email protected]>
which is currently applied by Mandriva.
Reported-by: Mark Glassberg <[email protected]>
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
gas used to accept (and ignore?) .size directives which referred to
undefined symbols, as this does. In binutils 2.21 these are treated
as errors.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Hans-Christian Egtvedt <[email protected]>
commit 40aee729b3 ('kconfig: fix default
value for choice input') fixed some cases where kconfig would select
the wrong option from a choice with a single valid option and thus
enter an infinite loop.
However, this broke the test for user input of the form 'N?', because
when kconfig selects the single valid option the input is zero-length
and the test will read the byte before the input buffer. If this
happens to contain '?' (as it will in a mips build on Debian unstable
today) then kconfig again enters an infinite loop.
Signed-off-by: Ben Hutchings <[email protected]>
Cc: [email protected] [2.6.17+]
Signed-off-by: Michal Marek <[email protected]>
via-ircc has been passing a NULL pointer to DMA allocation functions,
which is completely invalid and results in a BUG on PowerPC. Now
that we always have the device pointer available, pass it in.
Reference: http://bugs.debian.org/619450
Reported-by: Andrew Buckeridge <[email protected]>
Signed-off-by: Ben Hutchings <[email protected]>
Tested-by: Andrew Buckeridge <[email protected]> [against 2.6.32]
Signed-off-by: David S. Miller <[email protected]>
via-ircc still maintains its own array of device pointers in Linux 2.4
style. Worse, it always uses index 0, so it will crash if there are
multiple suitable devices in the system.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Define some constant offsets for CALL_REQUEST based on the description
at <http://www.techfest.com/networking/wan/x25plp.htm> and the
definition of ROSE as using 10-digit (5-byte) addresses. Use them
consistently. Validate all implicit and explicit facilities lengths.
Validate the address length byte rather than either trusting or
assuming its value.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
gas used to accept (and ignore?) .size directives which referred to
undefined symbols, as this does. In binutils 2.21 these are treated
as errors.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Commit e2cda32264 ("thp: add pmd mangling generic functions") replaced
some macros in <asm-generic/pgtable.h> with inline functions.
If the functions are to be defined (not all architectures need them)
then struct vm_area_struct must be defined first. So include
<linux/mm_types.h>.
Fixes a build failure seen in Debian:
CC [M] drivers/media/dvb/mantis/mantis_pci.o
In file included from arch/arm/include/asm/pgtable.h:460,
from drivers/media/dvb/mantis/mantis_pci.c:25:
include/asm-generic/pgtable.h: In function 'ptep_test_and_clear_young':
include/asm-generic/pgtable.h:29: error: dereferencing pointer to incomplete type
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
FREQ is a ridiculously short name for a platform-specific macro in a
generic header, and it now conflicts with an enumeration in the
gspca/ov519 driver.
Also delete conditional reference to ixp4xx_get_board_tick_rate()
which is not defined anywhere.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: Krzysztof Hałasa <[email protected]>
Ben Hutchings