We have two different transfer methods on i.MX: FIQ and DMA. Since
the merge of the ASoC multicomponent support the DMA device is lost.
Add it again. Also, imx_ssi_dai_probe has to be called for !AC97
aswell.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Acked-by: Liam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
In order to not touch the driver file for different xtal usage,
push the clkin value to board file and calculate the register
value instead of hardcoding it.
Signed-off-by: Bob Liu <lliubbo@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Felipe Balbi <balbi@ti.com>
We need to restart the timer in order to recognize USB devices in
host-only mode.
Signed-off-by: Bob Liu <lliubbo@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Split the USB MMR init steps out into a helper func that both the platform
init and the resume code may call.
Then while suspending, the gpio_vrsel will change from high to low which
will generate a wakeup event and resume the system immediately, so we need
to manually drive it low before we sleep.
Signed-off-by: Bob Liu <lliubbo@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Since clock support is optional across processors, don't make the whole
musb pm paths depend upon it. Just conditionalize the clock accesses.
Signed-off-by: Bob Liu <lliubbo@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Felipe Balbi <balbi@ti.com>
The new MUSB power code needs musb_read_txhubport() to
return a value (so stub it as 0 like the other Blackfin
hub funcs).
Signed-off-by: Ian Jeffray <ian@jeffray.co.uk>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Commit 9f445cb29918dc488b7a9a92ef018599cce33df7[USB: musb: disable
double buffering for older RTL versions] tries to disable double
buffer mode by writing endpoint hw max packet size to TXMAP/RXMAP.
First the approach can break full speed and cause overflow problems.
We should always set those registers with the actual max packet size
from endpoint descriptor.
Second, the problem describe by commit 9f445cb299
was caused by musb gadget driver; nothing to do with RTL revision as
originaly suspected.
The real fix to the problem is to always use actual max packet
size from endpoint descriptor to config TXMAP/RXMAP registers.
Cc: Cliff Cai <cliff.cai@analog.com>
Cc: David Brownell <dbrownell@users.sourceforge.net>
Cc: Anand Gadiyar <gadiyar@ti.com>
Cc: Mike Frysinger <vapier@gentoo.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: stable@kernel.org
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Some actions like musb_platform_exit are only performed on module
removal and not on shutdown, which results in PHY being left enabled
on reboot at least. This is sometimes causing strange failures after
reboot (observed on OMAP3 pandora board), when DEVCTL does not report
VBUS state correctly due to unknown reasons (possibly because of
communication issues between musb IP and PHY). Running
musb_platform_exit before reset seems to resolve that issue.
Move some exit code from musb_remove() to musb_shutdown() so that it
is performed on both module removal and shutdown/reset. Also convert
the host check so that it doesn't need #ifdef.
Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
In function musb_gadget_setup() call put_device()
when device_register() fails.
Signed-off-by: Rahul Ruikar <rahul.ruikar@gmail.com>
Acked-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
1, In Rx double buffer case, FIFO may have two packets, so
rxstate should be called to unload fifo if RXPKTRDY is set
even the current request has not been completed.
2, Commit 633ba7876b96ec339ef685357e2f7c60b5a8ce85
introduces autoclear to support double buffer in dma mode 0,
so remove clearing RXPKTRDY manually for dma mode 0.
3, Commit c7af6b29ffeffbeb28caf39e5b2ce29b11807c7d may break
dma mode 1 for non-doublebuffer endpoint, fix it.
With this patch, either usbtest #5 or g_file_storage(writing
file to device in usb host) or g_ether have been tested OK in
double buffer case(using fifo mode 3). Also, this patch has
been verified that single buffer case can't be broken.
Cc: David Brownell <dbrownell@users.sourceforge.net>
Cc: Anand Gadiyar <gadiyar@ti.com>
Cc: Mike Frysinger <vapier@gentoo.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Full rewrite of the P2M DMA Engine to support scatter gather and take
advantage of some of the features of the hardware. This includes using
repeat DMA operations and line-mode transfers (for copying OSG and
video display buffers).
What isn't working: For some reason, descriptor mode is not working. I've
implemented a psuedo version (still has one-interrupt per DMA operation),
but we would get huge improvements if we could hand off a ring of
descriptors to the P2M and get back one interrupt when it was done with
all of them.
Documentation is very vague on this, and even the ODM example code
half attempts to get it right, but comments it out of the driver
because it just doesn't work *sigh*
Converts all v4l2 to dma-sg. So long slow dma-contiguous, but hello
more interrupts :(
Signed-off-by: Ben Collins <bcollins@bluecherry.net>
This patch is based on Dan's original patch. His original description is
below:
Smatch complained about a couple checking for NULL after dereferencing
bugs. I'm not super familiar with the code so I did the conservative
thing and move the dereferences after the checks.
The dereferences in cifs_lock() and cifs_fsync() were added in
ba00ba64cf "cifs: make various routines use the cifsFileInfo->tcon
pointer". The dereference in find_writable_file() was added in
6508d904e6 "cifs: have find_readable/writable_file filter by fsuid".
The comments there say it's possible to trigger the NULL dereference
under stress.
Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
Noticed while reviewing (late) the rbtree conversion patchset (which has been merged
already).
Cc: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Suresh Jayaraman <sjayaraman@suse.de>
Signed-off-by: Steve French <sfrench@us.ibm.com>
We were using nlmsg_find_attr() to look up the bytecode by attribute when
auditing, but then just using the first attribute when actually running
bytecode. So, if we received a message with two attribute elements, where only
the second had type INET_DIAG_REQ_BYTECODE, we would validate and run different
bytecode strings.
Fix this by consistently using nlmsg_find_attr everywhere.
Signed-off-by: Nelson Elhage <nelhage@ksplice.com>
Signed-off-by: Thomas Graf <tgraf@infradead.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This will let us use it on a nlmsghdr stored inside a netlink_callback.
Signed-off-by: Nelson Elhage <nelhage@ksplice.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
After commit ebc0ffae5 (RCU conversion of fib_lookup()),
fib_result_assign() should not change fib refcounts anymore.
Thanks to Michael who did the bisection and bug report.
Reported-by: Michael Ellerman <michael@ellerman.id.au>
Tested-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
a00b10c360 "Only enforce fence limits inside the GTT" also
added a fenceable/mappable disdinction when binding/pinning buffers.
This only complicates the code with no pratical gain:
- In execbuffer this matters on for g33/pineview, as this is the only
chip that needs fences and has an unmappable gtt area. But fences
are only possible in the mappable part of the gtt, so need_fence
implies need_mappable. And need_mappable is only set independantly
with relocations which implies (for sane userspace) that the buffer
is untiled.
- The overlay code is only really used on i8xx, which doesn't have
unmappable gtt. And it doesn't support tiled buffers, currently.
- For all other buffers it's a bug to pass in a tiled bo.
In short, this disdinction doesn't have any practical gain.
I've also reverted mapping the overlay and context pages as possibly
unmappable. It's not worth being overtly clever here, all the big
gains from unmappable are for execbuf bos.
Also add a comment for a clever optimization that confused me
while reading the original patch by Chris Wilson.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
In a00b10c360 "Only enforce fence limits inside the GTT"
Chris Wilson implemented an optimization to only pin framebuffers
as mappable for crtc_set_base (but not for pageflips). This breaks
the abi, eg: A double buffering mesa client might leave the last
framebuffer in unmappable space on close. A subsequent glReadPix
by a frontbuffer rendering client then goes boom. My pretty anal
mappable/unmappable consistency checking detected this, see
https://bugs.freedesktop.org/show_bug.cgi?id=31286
Chris Wilson tried to fix this in 085ce26437 by pinning
tiled framebuffers into mappable space. This
a) renders the original optimization of not forcing framebuffers
for pageflipping clients into mappable pointless because all our
scanout buffers are tiled by default.
b) doesn't solve the problem for untiled framebuffers.
So kill this. Emperically it's no gain anyway because framebuffers are
being reused by the ddx and hence there's no chance for them to get
constanly bounced between mappable and unmappable.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
The function shash_async_import did not initialise the descriptor
correctly prior to calling the underlying shash import function.
This patch adds the required initialisation.
Reported-by: Miloslav Trmac <mitr@redhat.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
On certain VIA chipsets AES-CBC requires the input/output to be
a multiple of 64 bytes. We had a workaround for this but it was
buggy as it sent the whole input for processing when it is meant
to only send the initial number of blocks which makes the rest
a multiple of 64 bytes.
As expected this causes memory corruption whenever the workaround
kicks in.
Reported-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Many of the IBM Terminal keyboards from the 1980s and early 1990s communicate
using a protocol similar, but not identical to the AT keyboard protocol.
(Models known to be like this include 6110344, 6110668, 1390876, 1386887, and
possibly others.)
When the connector is rewired or adapter to an AT-DIN or PS/2 connector, they
can be connected to a standard PC, with three caveats:
a) They can only use scancode set 3; requests to use anything else are
quietly ignored.
b) The AT Command to request Make, Break and Repeat codes is not properly
interpreted.
c) The top function keys on a 122 key keyboard, and the arrow/edit keys in
the middle of the board send non-standard scancodes.
C) is easily taken care of in userspace, by use of setkeycodes
B) can be taken care of by a userspace hack (that makes the kernel complain
in dmesg)
A) is fixable in theory, but on the keyboard i tested on (6110668), it seems
to be detected unoverridably as Set 2, causing userspace oddities that make
it harder to fix C).
Enclosed is a small patch to the kernel that fixes A) and B) in the kernel,
making it much easier to fix C) in userspace. It adds a single kernel
command line parameter that overrides the detection that sets these boards
as set 2, and instead of sending the Make-break-repeat command to the
keyboard, it sends the make-break command, which is properly recognized by
these keyboards. Software level key repeating seems to make up for the lack
of hardware repeat codes perfectly.
Without manually setting the command line parameter (tentatively named
atkbd.terminal), this code has no effect, and the driver works exactly as
before.
See also:
http://www.seasip.info/VintagePC/ibm_1390876.htmlhttp://www.seasip.info/VintagePC/ibm_6110344.htmlhttp://geekhack.org/showwiki.php?title=Island:7306
Signed-off-by: Erika Quinn <erikas.aubade@gmail.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
From: Rob Herring <rob.herring@smooth-stone.com>
The timer-sp h/w used on versatile platforms can also be used for other
platforms, so move it to a common location.
Signed-off-by: Rob Herring <rob.herring@smooth-stone.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
The SWP instruction was deprecated in the ARMv6 architecture,
superseded by the LDREX/STREX family of instructions for
load-linked/store-conditional operations. The ARMv7 multiprocessing
extensions mandate that SWP/SWPB instructions are treated as undefined
from reset, with the ability to enable them through the System Control
Register SW bit.
This patch adds the alternative solution to emulate the SWP and SWPB
instructions using LDREX/STREX sequences, and log statistics to
/proc/cpu/swp_emulation. To correctly deal with copy-on-write, it also
modifies cpu_v7_set_pte_ext to change the mappings to priviliged RO when
user RO.
Signed-off-by: Leif Lindholm <leif.lindholm@arm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Kirill A. Shutemov <kirill@shutemov.name>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
This patch removes the domain switching functionality via the set_fs and
__switch_to functions on cores that have a TLS register.
Currently, the ioremap and vmalloc areas share the same level 1 page
tables and therefore have the same domain (DOMAIN_KERNEL). When the
kernel domain is modified from Client to Manager (via the __set_fs or in
the __switch_to function), the XN (eXecute Never) bit is overridden and
newer CPUs can speculatively prefetch the ioremap'ed memory.
Linux performs the kernel domain switching to allow user-specific
functions (copy_to/from_user, get/put_user etc.) to access kernel
memory. In order for these functions to work with the kernel domain set
to Client, the patch modifies the LDRT/STRT and related instructions to
the LDR/STR ones.
The user pages access rights are also modified for kernel read-only
access rather than read/write so that the copy-on-write mechanism still
works. CPU_USE_DOMAINS gets disabled only if the hardware has a TLS register
(CPU_32v6K is defined) since writing the TLS value to the high vectors page
isn't possible.
The user addresses passed to the kernel are checked by the access_ok()
function so that they do not point to the kernel space.
Tested-by: Anton Vorontsov <cbouatmailru@gmail.com>
Cc: Tony Lindgren <tony@atomide.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
The Sony VPCZ1 doesn't support active multiplexing and trying to enable
it causes keyboard to stop working. Since most (all?) VAIOs do not have
external PS/2 ports nor they implement active multiplexing properly, and
trying to enable MUX usually messes up keyboard/touchpad, let's simply
disable MUX probing based on board name (VAIO).
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
We do access_ok checks on all ring values on an ioctl,
so we don't need to redo them on each access.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This is broken from 97ef1bdd0b.
Let's set the correct bit for LLC+MLC and LLC only.
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Cc: stable@kernel.org
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
This restores cache behavior for default AGP_USER_MEMORY as
uncached, and leave default AGP_USER_CACHED_MEMORY as LLC only.
I've seen different cache behavior on one sandybridge desktop CPU vs.
another mobile CPU. Until we figure out how to detect the real cache
config, restore back to the original behavior now.
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Cc: stable@kernel.org
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
We should enable FDI normal training on Sandybridge/CPT system
as well.
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
[ickle: removed unrelated chunks]
Cc: stable@kernel.org
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Given that __in_29bit_mode() is a constant for the non-PMB case, we can
simply use the PMB-facing version of phys_addr_mask() and drop the other
variants.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Now that nommu selects 32BIT we run in to the situation where SH-2A
supports an uncached identity mapping by way of the BSC, while the SH-2
does not. This provides stubs for the PC manglers and tidies up some of
the system*.h mess in the process.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
The nommu code has regressed somewhat in that 29BIT gets set for the
SH-2/2A configs regardless of the fact that they are really 32BIT sans
MMU or PMB. This does a bit of tidying to get nommu properly selecting
32BIT as it was before.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
There was a leftover inw() used here that really just wants to be a
__raw_readw() instead. Convert it over.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Presently the extern inline case results in a compiler warning on ARM due
to the memory barrier definition used in the I/O routines. These
ultimately all want to be static inline anyways, so just convert them all
in place.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
As non-PFC chips are added that may support IRQs, pass through to the
generic helper. This follows the the SH change.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
