Cause is a misplaced bracket.
The code
strlen(buf+1)
will be two bytes less than
strlen(buf)+1
The +1 is in this code to reserve space for an additional space character.
[ralf@linux-mips.org: Thomas' original patch fixed the issue only for
Yosemite but the same bug exists also in Emma.]
Signed-off-by: Thomas Jarosch <thomas.jarosch@intra2net.com>
Patchwork: https://patchwork.linux-mips.org/patch/2861/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Otherwise the fh changing the master control won't get the inactive state
change event for the slave controls.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Setting sev->fh to NULL causes problems for the del op added in the next
patch of this series, since this op needs a way to get to its own data
structures, and typically this will be done by using container_of on an
embedded v4l2_fh struct.
The reason the original code is setting sev->fh to NULL is to signal
to users of the event framework that the unsubscription has happened,
but since their is no shared lock between the event framework and users
of it, this is inherently racy, and it also turns out to be unnecessary
as long as both the event framework and the user of the framework do their
own locking properly and the user guarantees that it holds no references
to the subcribed_event structure after its del operation has been called.
This is best explained by looking at the only code currently checking for
sev->fh being set to NULL on unsubscribe, which is the v4l2-ctrls.c send_event
function. Here is the relevant code from v4l2-ctrls: send_event():
if (sev->fh && (sev->fh != fh ||
(sev->flags & V4L2_EVENT_SUB_FL_ALLOW_FEEDBACK)))
v4l2_event_queue_fh(sev->fh, &ev);
Now lets say that v4l2_event_unsubscribe and v4l2-ctrls: send_event() race
on the same sev, then the following could happens:
1) send_event checks sev->fh, finds it is not NULL
<thread switch>
2) v4l2_event_unsubscribe sets sev->fh NULL
3) v4l2_event_unsubscribe calls v4l2_ctrls del_event function, this blocks
as the thread calling send_event holds the ctrl_lock
<thread switch>
4) send_event calls v4l2_event_queue_fh(sev->fh, &ev) which not is equivalent
to calling: v4l2_event_queue_fh(NULL, &ev)
5) oops, NULL pointer deref.
Now again without setting sev->fh to NULL in v4l2_event_unsubscribe and
without the (now senseless since always true) sev->fh != NULL check in
1) send_event is about to call v4l2_event_queue_fh(sev->fh, &ev)
<thread switch>
2) v4l2_event_unsubscribe removes sev->list from the fh->subscribed list
<thread switch>
3) send_event calls v4l2_event_queue_fh(sev->fh, &ev)
4) v4l2_event_queue_fh blocks on the fh_lock spinlock
<thread switch>
5) v4l2_event_unsubscribe unlocks the fh_lock spinlock
6) v4l2_event_unsubscribe calls v4l2_ctrls del_event function, this blocks
as the thread calling send_event holds the ctrl_lock
<thread switch>
8) v4l2_event_queue_fh takes the fh_lock
7) v4l2_event_queue_fh calls v4l2_event_subscribed, does not find it since
sev->list has been removed from fh->subscribed already -> does nothing
9) v4l2_event_queue_fh releases the fh_lock
10) the caller of send_event releases the ctrl lock (mutex)
<thread switch>
11) v4l2_ctrls del_event takes the ctrl lock
12) v4l2_ctrls del_event removes sev->node from the ev_subs list
13) v4l2_ctrls del_event releases the ctrl lock
14) v4l2_event_unsubscribe frees the sev, to which no references are being
held anymore
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The kev pointers inside the pending events queue (the available queue) of the
fh point to data inside the sev, unsubscribing frees the sev, thus making these
pointers point to freed memory!
This patch fixes these dangling pointers in the available queue by removing
all matching pending events on unsubscription.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
64933337e3
[media] tvp5150: Add video format registers configuration values
Added constants for each video standard supported by TVP5150, so this patch
get rid of the magic numbers.
Signed-off-by: Javier Martinez Canillas <martinez.javier@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
G2D is a 2D graphics accelerator engine present in the s5p family
of Samsung SoCs. It is capable of bitblt and raster operations on
images having dimensions of up to 8000x8000.
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Allow drivers to optimise out the register cache sync if they didn't need
to do one. If the hardware is desynced from the register cache (by power
loss for example) then the driver should call regcache_mark_dirty() to
let the core know about this.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Give regcache_lzo_block_count() a copy of the map so that when we decide
we want to make the LZO cache more controllable we can more easily do so.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
There seem to be lots of regmap-using devices with very similar interrupt
controllers with a small bank of interrupt registers and mask registers
with an interrupt per bit. This won't cover everything but it's a good
start.
Each chip supplies a base for the status registers, a base for the mask
registers, an optional base for writing acknowledgements (which may be the
same as the status registers) and an array of bits within each of these
register banks which indicate the interrupt.
There is an assumption that the bit for each interrupt will be the same
in each of the register bank.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
MFC uses the multi-plane API, but it reported single-plane
when querying capabilities.
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Queued list was not reset on REQBUFS(0) call. This caused to enqueue a
freed buffer to the driver.
Reported-by: Angela Wan <angela.j.wan@gmail.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
v4l2_planes[plane].length field was not initialized for userptr buffers.
This patch fixes this issue.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
CC: Pawel Osciak <pawel@osciak.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
snd_soc_update_bits already does read-modify-write,
no need to read the register before calling snd_soc_update_bits.
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
FSI2 can control valid data position, like
package in front/back or stream mode (16bit x 2).
But current fsi driver is assuming it was in-back.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
__buffer_in_use() might be called for empty/uninitialized buffer in the
following scenario: REQBUF(n, USER_PTR), QUERYBUF(). This patch fixes
kernel ops in such case.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
CC: Pawel Osciak <pawel@osciak.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The gpio_base for the PMIC on the CPU module was being incorrectly set to
be the same as that for the CODEC causing the two GPIO drivers to collide.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
The conversion to use SAMSUNG_GPIO_EXTRA rather than S3C24XX_GPIO_EXTRA
broke a number of S3C64xx boards which had been using the symbols provided
to reserve a range of GPIOs for board specific use. Fix this by adding
equivalent symbols for the new shared driver and updating the relevant
boards to use the new symbols.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
The conversion of the S3C64XX GPIOs to the generic Samsung GPIOs mean that
rather than using the previous S3C24XX_GPIO_EXTRA configuration the driver
now uses SAMSUNG_GPIO_EXTRA.
Since SAMSUNG_GPIO_EXTRA requires the arch to include the extra space in
ARCH_NR_GPIOs add it on to BOARD_NR_GPIOs (in case boards are relying on
the existing reservation).
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
ak4642 was tested by ms7724se board
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
This patch fixup below warning
${linux}/sound/soc/sh/fsi.c:442:3:\
warning: passing argument 1 of '__fsi_reg_read' makes pointer\
from integer without a cast
${linux}/sound/soc/sh/fsi.c:517:3: \
warning: passing argument 1 of '__fsi_reg_write' makes pointer\
from integer without a cast
${linux}/sound/soc/sh/fsi.c:663:3: \
warning: passing argument 1 of '__fsi_reg_mask_set' makes pointer\
from integer without a cast
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
I've been seeing memory leaks on my system in the form of large
(300-400MB) GEM objects created by now-dead processes laying around
clogging up memory. I usually notice when it gets to about 1.2GB of
them. Hopefully this clears up the issue, but I just found this bug
by inspection.
Signed-off-by: Eric Anholt <eric@anholt.net>
Cc: stable@kernel.org
Signed-off-by: Keith Packard <keithp@keithp.com>
Kernels with no iommu support cannot ever need the Ironlake
work-around, so never enable it in that case.
Might be better to completely remove the work-around from the kernel
in this case?
Signed-off-by: Keith Packard <keithp@keithp.com>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
The semi-colon is a typo here and it makes the if statement
unconditional.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Resetting DISPC when a DISPC output is enabled causes the DSS to go into an
inconsistent state. Thus if the bootloader has enabled a display, the hwmod code
cannot reset the DISPC module just like that, but the outputs need to be
disabled first.
Add function dispc_disable_outputs() which disables all active overlay manager
and ensure all frame transfers are completed.
Modify omap_dss_reset() to call this function and clear DSS_CONTROL,
DSS_SDI_CONTROL and DSS_PLL_CONTROL so that DSS is in a clean state when the
DSS2 driver starts.
This resolves the hang issue(caused by a L3 error during boot) seen on the
beagle board C3, which has a factory bootloader that enables display. The issue
is resolved with this patch.
Thanks to Tomi and Sricharan for some additional testing.
Acked-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Tested-by: R, Sricharan <r.sricharan@ti.com>
Signed-off-by: Archit Taneja <archit@ti.com>
[paul@pwsan.com: restructured code, removed omap_{read,write}l(), removed
cpu_is_omap*() calls and converted to dev_attr]
Signed-off-by: Paul Walmsley <paul@pwsan.com>
This patch adds a custom DSS reset function used on OMAPs from OMAP2
forward.
The function doesn't actually do a reset, it only waits for the reset to
complete. The reason for this is that on OMAP4 there is no possibility
to do a SW reset, and on OMAP2/3 doing a SW reset for dss_core resets
all the other DSS modules also, thus breaking the HWMOD model where
every DSS module is handled independently.
This fixes the problem with DSS reset on OMAP4, caused by the fact that
because there's no SW reset for dss_core on OMAP4, the HWMOD framework
doesn't try to reset dss_core and thus the DSS clocks were never enabled
at the same time. This causes causes the HWMOD reset to fail for
dss_dispc and dss_rfbi.
The common reset function will also allow us to fix another problem in
the future: before doing a reset we need to disable DSS outputs, which
are in some cases enabled by the bootloader, as otherwise DSS HW seems
to get more or less stuck, requiring a power reset to recover.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
[paul@pwsan.com: modified to build arch/arm/mach-omap2/display.o
unconditionally to avoid an error when !CONFIG_OMAP2_DSS]
Signed-off-by: Paul Walmsley <paul@pwsan.com>
OMAP2/3 dss_core has a reset status flag in sysstatus register. Add
SYSS_HAS_RESET_STATUS flag to HWMOD data so it can be used.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
The OMAP4 HWMOD data currently contains errors with DSS clocks:
dss_hdmi and dss_venc have their main_clks wrong. The clocks should be
dss_48mhz_clk and dss_tv_clk, respectively.
These problems were temporarily fixed with the DSS patches
9ede365aa6 ("HACK: OMAP: DSS2: clk hack
for OMAP2/3"), and df5d3ed23c ("OMAP:
DSS2: HDMI: fix hdmi clock name"), which can be reverted after this
patch.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
DSS needs all DSS clocks to be enabled to be able to finish reset
properly. Before v3.1-rc1 the omapdss driver was managing clocks and
resets correctly. However, when omapdss started using runtime PM at
v3.1-rc1, the responsibility for the reset moved to HWMOD framework.
HWMOD framework does not currently enable all the DSS clocks when
resetting the DSS hardware. This causes the HWMOD frameworks boot-time
reset to fail, possibly leaving the DSS hardware in undefined state.
This patch sets HWMOD_CONTROL_OPT_CLKS_IN_RESET for dss_core. The flag
is actually not used on OMAP4, because dss_core hardware does not have
soft-reset functionality and thus the HWMOD framework never resets nor
waits for the reset to finish.
However, while the flag is not strictly needed currently, I think it
represents the HW correctly: all the DSS clocks should be enabled after
power-on to allow DSS hardware to finish its reset.
A custom reset function will be added in the following patches which
manages this correctly for OMAP4.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Remove the dss_dss_clk from dss_core's opt-clocks. dss_dss_clk already
defined as the dss main_clk, and thus is not needed as an opt-clock.
Remove opt-clocks for dss_dispc, as dispc only uses the main_clk.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
The OMAP3 HWMOD data currently contains these errors with DSS clocks:
- dss_rfbi is missing ick opt-clock, which is needed for RFBI to
calculate timings
- dss_dsi is missing ick and sys_clk
- dss_venc is missing dss_96m_fck opt-clock, which is required on
OMAP3430
- dss_venc's interface and main clocks are wrong, causing VENC to fail
to start
These problems were temporarily fixed with a DSS patch
9ede365aa6 ("HACK: OMAP: DSS2: clk hack
for OMAP2/3"), which can be reverted after this patch (and the similar
patches for other OMAPs).
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
DSS needs all DSS clocks to be enabled to be able to finish reset
properly. Before v3.1-rc1 the omapdss driver was managing clocks and
resets correctly. However, when omapdss started using runtime PM at
v3.1-rc1, the responsibility for the reset moved to HWMOD framework.
HWMOD framework does not currently enable all the DSS clocks when
resetting the DSS hardware. This hasn't caused any problems so far, but
we may just have been lucky.
dss_core's opt-clocks is also missing dss_96m_fck, which is a DSS clock
present only on OMAP3430, and thus required on OMAP3430 to finish the
reset.
This patch sets HWMOD_CONTROL_OPT_CLKS_IN_RESET and adds the dss_96m_fck
opt-clock for dss_core in OMAP3 HWMOD data, fixing the issue.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
[paul@pwsan.com: merged duplicate .flags fields]
Signed-off-by: Paul Walmsley <paul@pwsan.com>
The OMAP2xxx HWMOD data currently contains two errors with DSS clocks:
- dss_rfbi is missing ick opt-clock, which is needed for RFBI to
calculate timings
- dss_venc's interface and main clocks are wrong, causing VENC to fail
to start
These problems were temporarily fixed with a DSS patch
9ede365aa6 ("HACK: OMAP: DSS2: clk hack
for OMAP2/3"), which can be reverted after this patch (and the similar
patches for other OMAPs).
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
DSS needs all DSS clocks to be enabled to be able to finish reset
properly. Before v3.1-rc1 the omapdss driver was managing clocks and
resets correctly. However, when omapdss started using runtime PM at
v3.1-rc1, the responsibility for the reset moved to HWMOD framework.
HWMOD framework does not currently enable all the DSS clocks when
resetting the DSS hardware. This hasn't caused any problems so far, but
we may just have been lucky.
This patch sets HWMOD_CONTROL_OPT_CLKS_IN_RESET for dss_core in OMAP2xxx
HWMOD data, fixing the issue.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
[paul@pwsan.com: merged duplicate .flags fields]
Signed-off-by: Paul Walmsley <paul@pwsan.com>
This patch adds read-ahead capability to GFS2's
directory hash table management. It greatly improves
performance for some directory operations. For example:
In one of my file systems that has 1000 directories, each
of which has 1000 files, time to execute a recursive
ls (time ls -fR /mnt/gfs2 > /dev/null) was reduced
from 2m2.814s on a stock kernel to 0m45.938s.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Christoph has split up REQ_PRIO from REQ_META. That means that
we can drop REQ_PRIO from places where is it not needed. I'm
not at all sure that the combination WRITE_FLUSH_FUA | REQ_PRIO
makes any kind of sense, anyway.
In addition, I've added REQ_META to one place in the code where
it was missing. REQ_PRIO has been left for read/writes triggered
by glock acquisition and writeback only. We can adjust it again
if required, but these are the most important points from a
performance perspective.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
'sampling_rate_max' was removed with commit ef598549 ("[...] Remove
deprecated sysfs file sampling_rate_max"), so its line can be dropped
from governors.txt. And 'show_sampling_rate_min' is a typo: the sysfs
file is called 'sampling_rate_min'.
Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Providing a pincfg fix for VAIO-TT with ALC889 codec to work with the
auto-parser, and drop the static configuration.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Now that model=ultra is supported well by the auto-parser, we can get rid
of the whole alc262_quirks.c and its related codes.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
A machine like Q1-ultra which has only a single HP but no mic-jack, we
can re-task the headhpone as an external mic jack. This was done formerly
in ALC262 model=ultra quirk, and now the auto-parser supports this mode.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
