Allocating new block group is easy when the disk has plenty of space.
But things get difficult as the disk fills up, especially if
the FS has been run through btrfs-vol -b. The balance operation
is likely to make the total bytes available on the device greater
than the largest extent we'll actually be able to allocate.
But the device extent allocation code incorrectly assumes that a device
with 5G free will be able to allocate a 5G extent. It isn't normally a
problem because device extents don't get freed unless btrfs-vol -b
is run.
This fixes the device extent allocator to remember the largest free
extent it can find, and then uses that value as a fallback.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Btrfs allocates individual extents from block groups, and each
block group has a specific type. It may hold metadata, data
mirrored or striped etc.
When we balance space (btrfs-vol -b) or remove a drive (btrfs-vol -r)
we free block groups. Once a block group is freed, the space it was
using on the device may be available for use by new block groups.
btrfs_remove_block_group was clearing the flag that said
'our devices are full, don't even try to allocate new block groups',
but it was only clearing that flag for a specific type of block group.
This commit clears the full flag for all of the types of block groups,
making it much more likely that we'll be able to balance space when
the drive is close to full.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The commit_transaction call to wait_ordered_extents when snap_pending
passes nocow_only=1 to process only NOCOW or PREALLOC extents. This isn't
correct for the 'flushoncommit' mode, as it skips extents we just started
IO on in start_delalloc_inodes.
So, in the flushoncommit case, wait on all ordered extents. Otherwise,
only pass the nocow_only flag to wait_ordered_extents if snap_pending.
Signed-off-by: Sage Weil <sage@newdream.net>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
When enabling debug with v4l_debug set to 2 or greater, the driver
OOPS's on startup. Checks dev pointer before dereferencing, in
order to prevent this OOPS.
Signed-off-by: Michael Krufky <mkrufky@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
IOCTLS will never get handled if we dont connect
video_ioctl2 to mpeg_fops.ioctl
Signed-off-by: Michael Krufky <mkrufky@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Both tvaudio and the tuner share i2c address 0x42. The tvaudio module can
check whether it really is a tda9840, but the tuner can't. So the tvaudio
module must be loaded before the tuner module. This was also the case for
2.6.29, but the order was swapped in 2.6.30.
Thanks to Krzysztof Grygiencz for reporting and testing this.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
A recent patch didn't take into account that frontends can be compiled into
the kernel. Or that frontends compiled as modules can't be used by the
b2c2 driver if it is not a module itself.
Some frontends require multiple drivers, e.g. a demod driver and a tuner
driver. The code for the frontend support was getting added if the demod
driver was available. Change this to also require any needed tuner or SEC
drivers as well.
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
This is not just pure cosmetic, since the order affects the out-of-tree
module build at the -hg development tree.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Request demodulator auto-detect transmission parameters in case of
garbage parameters provided by application for compatibility.
That's needed at least for MPlayer compatibility currently.
Thanks to Jelle de Jong for reporting issue and providing SSH access to
Devin for debugging.
Thanks to Devin Heitmueller for hard debug work he did to find that bug.
Cc: Devin Heitmueller <dheitmueller@kernellabs.com>
Cc: Jelle de Jong <jelledejong@powercraft.nl>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The bug was introduced when adding the light frequency control
Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
When the change was introduced in the zl10353 for the i2c gate behavior, this
broke the HVR-900 which was not behind a gate. Use a version of the zl10353
config profile that indicates the tuner is not behind such a gate.
Without this patch the first tune succeeds, but subsequent tuning attempts
will fail.
The change also renames the terratec zl10353 profile I wrote to be more
generic, since it is shared by the non-terratec device.
Thanks to Michael Krufky for providing a HVR-900 and DVB-T environment to test
with.
Cc: Michael Krufky <mkrufky@kernellabs.com>
Signed-off-by: Devin Heitmueller <dheitmueller@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The Pinnacle Hybrid Pro 320e was missing a demod config for the xc3028, which
is required for digital tuning to work properly. Add the missing profile.
Thanks to Andreas Lunderhage for testing patches and providing a remote debug
environment.
Cc: Andreas Lunderhage <lunderhage@home.se>
Signed-off-by: Devin Heitmueller <dheitmueller@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
In cases where the device has a generic Empia USB ID, the call in the
precard setup phase did not set the tuner GPIO. As a result, the tuner may
not be taken out of reset before attempting initialization in the analog
driver.
This problem was not seen before with the EVGA inDtube, since that particular
board has the analog GPIO setup to include taking the tuner out of reset.
Thanks to Andreas Lunderhage for testing patches and providing a remote debug
environment for the Pinnacle 320e.
Cc: Andreas Lunderhage <lunderhage@home.se>
Signed-off-by: Devin Heitmueller <dheitmueller@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Set the GPIO properly for the analog side of the Pinnacle Hybrid Pro, or else
the emp202 doesn't get detected properly.
Signed-off-by: Devin Heitmueller <dheitmueller@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Setup the GPIOs properly and enable support for the DVB side of the Pinnacle
Hybrid Pro USB stick.
Thanks to Andreas Lunderhage for testing patches and providing a remote debug
environment.
Cc: Andreas Lunderhage <lunderhage@home.se>
Signed-off-by: Devin Heitmueller <dheitmueller@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Andy walls pointed out that we were passing 0x5d to the TUNER_GO register,
instead of 0x01. Set the register properly (note the code did still work with
the incorrect value, so this does not address a regression).
Thanks to Andy Walls for noticing the issue.
Cc: Andy Walls <awalls@radix.net>
Signed-off-by: Devin Heitmueller <dheitmueller@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The Terratec Cinergy T XS USB can have either a zl10353 or an mt352. Add
support for the MT352 variant.
Thanks to Jelle de Jong for providing a unit to test/debug with.
Cc: Jelle de Jong <jelledejong@powercraft.nl>
Signed-off-by: Devin Heitmueller <dheitmueller@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Depending on the video input format, vinmode/vinctl needs adjustments.
For TV, this is not relevant, since the supported decoders output data
at the same format. However, webcam sensors may have different formats,
so, this needs to be adjusted based on the device.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
In order to properly estimate fps, mt9v011 sensor driver needs to know
what is the used frequency on the sensor cristal. Adds the proper fields
and initialization code for specifying the cristal frequency.
Also, based on experimentation, it was noticed that the Silvercrest is
outputing data at 7 fps. This means that it should be using a 6.3 MHz
cristal. This information needs to be double checked later, by opening
the device. Anyway, by using this value for xtal, at least now we have
the correct fps report.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Since frames per second is a function of cristal frequency, and this is
device-specific, add a function that allows adjusting it, via
subdev->core->s_config callback.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
It is possible to adjust the fps rate by changing some register values.
This is function of the connected Xtal at the camera sensor, being a 27
MHz cristal needed, in order to support 640x480 at 30 fps.
For now, it will only calculate the values for fps. Later patches may
introduce V4L2 ioctls, to allow frequency rate adjustments.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
While trying to fix an mt9v001 webcam, I noticed that HSCALE/VSCALE do
work with em28xx + webcam. The issue is that the scaling setup depends
on the number of visible rows/cols of the input image.
With mt9v011 (Silvercrest), the resolution is 640x480. So, the scaling
is different from a normal TV image (720x480 on NTSC). This were causing
a wrong scaling and a previous patch disabled scaling.
As each sensor have their different resolution setting, the xres/yres
should be adjusted accordingly with the input sensor.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
With the previous approach, autodetection were working only for the two
generic entries (em275x and em2820 unknown ones). So, if someone would
try to force probing an specific device, the code would not properly run
the autodetection code.
With the new approach, the sensor autodetection will be run not only for
the two generic entries, but also do webcam specific ones.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Instead of using em28xx board decoder field for storing sensor information,
let's use instead a separate field for it.
Also, as sensors are currently autodetected, there's no need of having
it at the boards description. So, move it to the main em28xx struct.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Webcams in general don't have eeprom. So, the sensor hint code should be
called to properly detect what sensor is inside.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
By having the webcam devices marked as such, it will help the em28xx
driver to do the right thing on those devices.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Just renames the flag, to use a clearer name. Later patches will use
this flag to properly set some drivers behaviors for webcams.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
A couple of erroneous register writes snuck in that made the image go haywire. Remove these.
Many thanks to Grégory Lardière for finding this out
Signed-off-by: Erik Andrén <erik.andren@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
All hdcs registers use bit 0 as a read/write flag and needs to be shifted one bit to the left. This wasn't accounted for when doing a sequence of writes.
Signed-off-by: Erik Andrén <erik.andren@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
btrfs_split_leaf and btrfs_del_items can end up in a loop
where one is constantly spliting a given leaf and the other
is constantly merging it back with the adjacent nodes.
There is a better fix for this, but in the interest of something
small, this patch just changes btrfs_del_items back to balancing less
often.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Check objectid of item before checking the item type, otherwise we may return
zero for a key that is actually too low.
Signed-off-by: Yan Zheng <zheng.yan@oracle.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
find_free_dev_extent does not properly handle the case where
the device is not complete free, and there is a free extent
at the beginning of the device.
Signed-off-by: Yan Zheng <zheng.yan@oracle.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
comp_keys is duplicating what is done in btrfs_comp_cpu_keys, so just
call it.
Signed-off-by: Diego Calleja <diegocg@gmail.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The master mute switch is wrongly implemented as checking the pointer
instead of its value, thus it can be never muted. This patch fixes
the issue.
Reference: Novell bnc#404873
https://bugzilla.novell.com/show_bug.cgi?id=404873
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Cc: <stable@kernel.org>
This patch moves the caching of the block group off to a kthread in order to
allow people to allocate sooner. Instead of blocking up behind the caching
mutex, we instead kick of the caching kthread, and then attempt to make an
allocation. If we cannot, we wait on the block groups caching waitqueue, which
the caching kthread will wake the waiting threads up everytime it finds 2 meg
worth of space, and then again when its finished caching. This is how I tested
the speedup from this
mkfs the disk
mount the disk
fill the disk up with fs_mark
unmount the disk
mount the disk
time touch /mnt/foo
Without my changes this took 11 seconds on my box, with these changes it now
takes 1 second.
Another change thats been put in place is we lock the super mirror's in the
pinned extent map in order to keep us from adding that stuff as free space when
caching the block group. This doesn't really change anything else as far as the
pinned extent map is concerned, since for actual pinned extents we use
EXTENT_DIRTY, but it does mean that when we unmount we have to go in and unlock
those extents to keep from leaking memory.
I've also added a check where when we are reading block groups from disk, if the
amount of space used == the size of the block group, we go ahead and mark the
block group as cached. This drastically reduces the amount of time it takes to
cache the block groups. Using the same test as above, except doing a dd to a
file and then unmounting, it used to take 33 seconds to umount, now it takes 3
seconds.
This version uses the commit_root in the caching kthread, and then keeps track
of how many async caching threads are running at any given time so if one of the
async threads is still running as we cross transactions we can wait until its
finished before handling the pinned extents. Thank you,
Signed-off-by: Josef Bacik <jbacik@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Currently btrfs has a problem where it can use a ridiculous amount of RAM simply
tracking free space. As free space gets fragmented, we end up with thousands of
entries on an rb-tree per block group, which usually spans 1 gig of area. Since
we currently don't ever flush free space cache back to disk this gets to be a
bit unweildly on large fs's with lots of fragmentation.
This patch solves this problem by using PAGE_SIZE bitmaps for parts of the free
space cache. Initially we calculate a threshold of extent entries we can
handle, which is however many extent entries we can cram into 16k of ram. The
maximum amount of RAM that should ever be used to track 1 gigabyte of diskspace
will be 32k of RAM, which scales much better than we did before.
Once we pass the extent threshold, we start adding bitmaps and using those
instead for tracking the free space. This patch also makes it so that any free
space thats less than 4 * sectorsize we go ahead and put into a bitmap. This is
nice since we try and allocate out of the front of a block group, so if the
front of a block group is heavily fragmented and then has a huge chunk of free
space at the end, we go ahead and add the fragmented areas to bitmaps and use a
normal extent entry to track the big chunk at the back of the block group.
I've also taken the opportunity to revamp how we search for free space.
Previously we indexed free space via an offset indexed rb tree and a bytes
indexed rb tree. I've dropped the bytes indexed rb tree and use only the offset
indexed rb tree. This cuts the number of tree operations we were doing
previously down by half, and gives us a little bit of a better allocation
pattern since we will always start from a specific offset and search forward
from there, instead of searching for the size we need and try and get it as
close as possible to the offset we want.
I've given this a healthy amount of testing pre-new format stuff, as well as
post-new format stuff. I've booted up my fedora box which is installed on btrfs
with this patch and ran with it for a few days without issues. I've not seen
any performance regressions in any of my tests.
Since the last patch Yan Zheng fixed a problem where we could have overlapping
entries, so updating their offset inline would cause problems. Thanks,
Signed-off-by: Josef Bacik <jbacik@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Every time a request is enqueued or there is some work outstanding
from the ap_tasklet, the ap_poll_timer is scheduled again.
Unfortunately it was permanently called. It looked as if it was
started in the past and thus imediately expired.
This has been changed. First it is checked if the hrtimer is already
expired. Then the expiring time is forwarded and the timer restarted.
Signed-off-by: Felix Beck <felix.beck@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The combination of noexec=on and a clock_gettime call with clock id
CLOCK_THREAD_CPUTIME_ID is broken. The vdso code switches to the
access register mode to get access to the per-cpu data structure to
execute the magic ectg instruction. After the ectg instruction the
code always switches back to the primary mode but for noexec=on the
correct mode is the secondary mode. The effect of the bug is that the
user space program looses the access to all mappings without PROT_EXEC,
e.g. the stack. The problem is fixed by restoring the mode that has
been active before the switch to the access register mode.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
vdso per cpu area allocation in smp_prepare_cpus() happens with GFP_KERNEL
but irqs disabled. Triggers this one:
Badness at kernel/lockdep.c:2280
Modules linked in:
CPU: 0 Not tainted 2.6.30 #2
Process swapper (pid: 1, task: 000000003fe88000, ksp: 000000003fe87eb8)
Krnl PSW : 0400c00180000000 0000000000083360 (lockdep_trace_alloc+0xec/0xf8)
[...]
Call Trace:
([<00000000000832b6>] lockdep_trace_alloc+0x42/0xf8)
[<00000000000b1880>] __alloc_pages_internal+0x3e8/0x5c4
[<00000000000b1b4a>] __get_free_pages+0x3a/0xb0
[<0000000000026546>] vdso_alloc_per_cpu+0x6a/0x18c
[<00000000005eff82>] smp_prepare_cpus+0x322/0x594
[<00000000005e8232>] kernel_init+0x76/0x398
[<000000000001bb1e>] kernel_thread_starter+0x6/0xc
[<000000000001bb18>] kernel_thread_starter+0x0/0xc
Fix this by moving the allocation out of the irqs disabled section.
Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
swsusp_arch_suspend() actually saves all cpu register contents on
hibernation.
Machine checks must be disabled since swsusp_arch_suspend() stores
register contents to their lowcore save areas. That's the same
place where register contents on machine checks would be saved.
To avoid register corruption disable machine checks.
We must also disable machine checks in the new psw mask for
program checks, since swsusp_arch_suspend() may generate program
checks.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Our swsusp_arch_suspend() backend implementation disables prefixing
by setting the contents of the prefix register to 0.
However afterwards common code functions are called which might
access percpu data structures.
Since the lowcore contains e.g. the percpu base pointer this isn't
a good idea. So fix this by copying the hibernating cpu's lowcore to
absolute address zero.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Timer interrupts are excluded from being disabled during suspend. The
clock events code manages the disabling of clock events on its own
because the timer interrupt needs to be functional before the resume
code reenables the device interrupts.
The mfgpt timer request its interrupt without setting the IRQF_TIMER
flag so suspend_device_irqs() disables it as well which results in a
fatal resume failure.
Adding IRQF_TIMER to the interupt flags when requesting the mrgpt
timer interrupt solves the problem.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
LKML-Reference: <new-submission>
Cc: Andres Salomon <dilinger@debian.org>
Cc: stable@kernel.org
