tracing/function-return-tracer: call prepare_ftrace_return by registers
Impact: Optimize a bit the function return tracer
This patch changes the calling convention of prepare_ftrace_return to
pass its arguments by register. This will optimize it a bit and
prepare it to support dynamic tracing.
tracing/function-return-tracer: make the function return tracer lockless
Impact: remove spinlocks and irq disabling in function return tracer.
I've tried to figure out all of the race condition that could happen
when the tracer pushes or pops a return address trace to/from the
current thread_info.
Theory:
_ One thread can only execute on one cpu at a time. So this code
doesn't need to be SMP-safe. Just drop the spinlock.
_ The only race could happen between the current thread and an
interrupt. If an interrupt is raised, it will increase the index of
the return stack storage and then execute until the end of the
tracing to finally free the index it used. We don't need to disable
irqs.
This is theorical. In practice, I've tested it with a two-core SMP and
had no problem at all. Perhaps -tip testing could confirm it.
Steven Rostedt [Wed, 12 Nov 2008 20:24:24 +0000 (15:24 -0500)]
ftrace: add tracer called branch
Impact: added new branch tracer
Currently the tracing of branch profiling (unlikelys and likelys hit)
is only activated by the iter_ctrl. This patch adds a tracer called
"branch" that will just trace the branch profiling. The advantage
of adding this tracer is that it can be added to the ftrace selftests
on startup.
Signed-off-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Steven Rostedt [Wed, 12 Nov 2008 20:24:24 +0000 (15:24 -0500)]
trace: rename unlikely profiler to branch profiler
Impact: name change of unlikely tracer and profiler
Ingo Molnar suggested changing the config from UNLIKELY_PROFILE
to BRANCH_PROFILING. I never did like the "unlikely" name so I
went one step farther, and renamed all the unlikely configurations
to a "BRANCH" variant.
Signed-off-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
also clean up the ugly ifdefs in arch/x86/kernel/vsyscall_64.c by
creating DISABLE_UNLIKELY_PROFILE facility for code to turn off
instrumentation on a per file basis.
Steven Rostedt [Wed, 12 Nov 2008 05:14:40 +0000 (00:14 -0500)]
tracing: likely/unlikely branch annotation tracer
Impact: new likely/unlikely branch tracer
This patch adds a way to record the instances of the likely() and unlikely()
branch condition annotations.
When "unlikely" is set in /debugfs/tracing/iter_ctrl the unlikely conditions
will be added to any of the ftrace tracers. The change takes effect when
a new tracer is passed into the current_tracer file.
Which shows the normal tracer heading, as well as whether the condition was
correct "[correct]" or was mistaken "[INCORRECT]", followed by the function,
file name and line number.
Signed-off-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Steven Rostedt [Wed, 12 Nov 2008 05:14:39 +0000 (00:14 -0500)]
tracing: profile likely and unlikely annotations
Impact: new unlikely/likely profiler
Andrew Morton recently suggested having an in-kernel way to profile
likely and unlikely macros. This patch achieves that goal.
When configured, every(*) likely and unlikely macro gets a counter attached
to it. When the condition is hit, the hit and misses of that condition
are recorded. These numbers can later be retrieved by:
/debugfs/tracing/profile_likely - All likely markers
/debugfs/tracing/profile_unlikely - All unlikely markers.
As you can see by the above, there's a bit of work to do in rethinking
the use of some unlikelys and likelys. Note: the unlikely case had 71 hits
that were more than 25%.
Note: After submitting my first version of this patch, Andrew Morton
showed me a version written by Daniel Walker, where I picked up
the following ideas from:
1) Using __builtin_constant_p to avoid profiling fixed values.
2) Using __FILE__ instead of instruction pointers.
3) Using the preprocessor to stop all profiling of likely
annotations from vsyscall_64.c.
Thanks to Andrew Morton, Arjan van de Ven, Theodore Tso and Ingo Molnar
for their feed back on this patch.
(*) Not ever unlikely is recorded, those that are used by vsyscalls
(a few of them) had to have profiling disabled.
Signed-off-by: Steven Rostedt <srostedt@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Theodore Tso <tytso@mit.edu> Cc: Arjan van de Ven <arjan@infradead.org> Cc: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Steven Rostedt [Wed, 12 Nov 2008 05:01:27 +0000 (00:01 -0500)]
ring-buffer: no preempt for sched_clock()
Impact: disable preemption when calling sched_clock()
The ring_buffer_time_stamp still uses sched_clock as its counter.
But it is a bug to call it with preemption enabled. This requirement
should not be pushed to the ring_buffer_time_stamp callers, so
the ring_buffer_time_stamp needs to disable preemption when calling
sched_clock.
Signed-off-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
tracing/fastboot: Use the ring-buffer timestamp for initcall entries
Impact: Split the boot tracer entries in two parts: call and return
Now that we are using the sched tracer from the boot tracer, we want
to use the same timestamp than the ring-buffer to have consistent time
captures between sched events and initcall events.
So we get rid of the old time capture by the boot tracer and split the
initcall events in two parts: call and return. This way we have the
ring buffer timestamp of both.
tracing/fastboot: move boot tracer structs and funcs into their own header.
Impact: Cleanups on the boot tracer and ftrace
This patch bring some cleanups about the boot tracer headers. The
functions and structures of this tracer have nothing related to ftrace
and should have so their own header file.
Steven Rostedt [Tue, 11 Nov 2008 20:28:41 +0000 (15:28 -0500)]
ring-buffer: clean up warn ons
Impact: Restructure WARN_ONs in ring_buffer.c
The current WARN_ON macros in ring_buffer.c are quite ugly.
This patch cleans them up and uses a single RB_WARN_ON that returns
the value of the condition. This allows the caller to abort the
function if the condition is true.
Signed-off-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Steven Rostedt [Tue, 11 Nov 2008 20:01:42 +0000 (15:01 -0500)]
ring-buffer: buffer record on/off switch
Impact: enable/disable ring buffer recording API added
Several kernel developers have requested that there be a way to stop
recording into the ring buffers with a simple switch that can also
be enabled from userspace. This patch addes a new kernel API to the
ring buffers called:
tracing_on()
tracing_off()
When tracing_off() is called, all ring buffers will not be able to record
into their buffers.
tracing_on() will enable the ring buffers again.
These two act like an on/off switch. That is, there is no counting of the
number of times tracing_off or tracing_on has been called.
A new file is added to the debugfs/tracing directory called
tracing_on
This allows for userspace applications to also flip the switch.
echo 0 > debugfs/tracing/tracing_on
disables the tracing.
echo 1 > /debugfs/tracing/tracing_on
enables it.
Note, this does not disable or enable any tracers. It only sets or clears
a flag that needs to be set in order for the ring buffers to write to
their buffers. It is a global flag, and affects all ring buffers.
The buffers start out with tracing_on enabled.
There are now three flags that control recording into the buffers:
tracing_on: which affects all ring buffer tracers.
buffer->record_disabled: which affects an allocated buffer, which may be set
if an anomaly is detected, and tracing is disabled.
cpu_buffer->record_disabled: which is set by tracing_stop() or if an
anomaly is detected. tracing_start can not reenable this if
an anomaly occurred.
The userspace debugfs/tracing/tracing_enabled is implemented with
tracing_stop() but the user space code can not enable it if the kernel
called tracing_stop().
Userspace can enable the tracing_on even if the kernel disabled it.
It is just a switch used to stop tracing if a condition was hit.
tracing_on is not for protecting critical areas in the kernel nor is
it for stopping tracing if an anomaly occurred. This is because userspace
can reenable it at any time.
Side effect: With this patch, I discovered a dead variable in ftrace.c
called tracing_on. This patch removes it.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Linus Torvalds [Tue, 11 Nov 2008 18:53:50 +0000 (10:53 -0800)]
Merge branch 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
timers: handle HRTIMER_CB_IRQSAFE_UNLOCKED correctly from softirq context
nohz: disable tick_nohz_kick_tick() for now
irq: call __irq_enter() before calling the tick_idle_check
x86: HPET: enter hpet_interrupt_handler with interrupts disabled
x86: HPET: read from HPET_Tn_CMP() not HPET_T0_CMP
x86: HPET: convert WARN_ON to WARN_ON_ONCE
Linus Torvalds [Tue, 11 Nov 2008 18:52:25 +0000 (10:52 -0800)]
Merge branch 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
sched: release buddies on yield
fix for account_group_exec_runtime(), make sure ->signal can't be freed under rq->lock
sched: clean up debug info
Linus Torvalds [Tue, 11 Nov 2008 18:51:50 +0000 (10:51 -0800)]
Merge branch 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
ring-buffer: prevent infinite looping on time stamping
ftrace: disable tracing on resize
ftrace: fix breakage in bin_fmt results
ftrace: ftrace.txt version update
ftrace: update txt document
Steven Rostedt [Tue, 11 Nov 2008 17:47:44 +0000 (18:47 +0100)]
ring-buffer: add reader lock
Impact: serialize reader accesses to individual CPU ring buffers
The code in the ring buffer expects only one reader at a time, but currently
it puts that requirement on the caller. This is not strong enough, and this
patch adds a "reader_lock" that serializes the access to the reader API
of the ring buffer.
Signed-off-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Linus Torvalds [Tue, 11 Nov 2008 17:32:58 +0000 (09:32 -0800)]
Merge branch 'for-linus' of git://oss.sgi.com/xfs/xfs
* 'for-linus' of git://oss.sgi.com/xfs/xfs:
[XFS] XFS: Check for valid transaction headers in recovery
[XFS] handle memory allocation failures during log initialisation
[XFS] Account for allocated blocks when expanding directories
[XFS] Wait for all I/O on truncate to zero file size
[XFS] Fix use-after-free with log and quotas
Linus Torvalds [Tue, 11 Nov 2008 17:31:32 +0000 (09:31 -0800)]
Merge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mfasheh/ocfs2
* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mfasheh/ocfs2: (21 commits)
ocfs2: Check search result in ocfs2_xattr_block_get()
ocfs2: fix printk related build warnings in xattr.c
ocfs2: truncate outstanding block after direct io failure
ocfs2/xattr: Proper hash collision handle in bucket division
ocfs2: return 0 in page_mkwrite to let VFS retry.
ocfs2: Set journal descriptor to NULL after journal shutdown
ocfs2: Fix check of return value of ocfs2_start_trans() in xattr.c.
ocfs2: Let inode be really deleted when ocfs2_mknod_locked() fails
ocfs2: Fix checking of return value of new_inode()
ocfs2: Fix check of return value of ocfs2_start_trans()
ocfs2: Fix some typos in xattr annotations.
ocfs2: Remove unused ocfs2_restore_xattr_block().
ocfs2: Don't repeat ocfs2_xattr_block_find()
ocfs2: Specify appropriate journal access for new xattr buckets.
ocfs2: Check errors from ocfs2_xattr_update_xattr_search()
ocfs2: Don't return -EFAULT from a corrupt xattr entry.
ocfs2: Check xattr block signatures properly.
ocfs2: add handler_map array bounds checking
ocfs2: remove duplicate definition in xattr
ocfs2: fix function declaration and definition in xattr
...
Linus Torvalds [Tue, 11 Nov 2008 17:25:21 +0000 (09:25 -0800)]
Merge branch 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-2.6
* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-2.6: (35 commits)
V4L/DVB (9516): cx18: Move DVB buffer transfer handling from irq handler to work_queue
V4L/DVB (9557): gspca: Small changes for the sensor HV7131B in zc3xx.
V4L/DVB (9556): gspca: Bad init sequence for sensor HV7131B in zc3xx.
V4L/DVB (9549): gspca: Fix a typo in one of gspca chips name.
V4L/DVB (9515): cx18: Use correct Mailbox IRQ Ack values and misc IRQ handling cleanup
V4L/DVB (9493): kconfig patch
V4L/DVB (9527): af9015: fix compile warnings
V4L/DVB (9524): af9013: fix bug in status reading
V4L/DVB (9511): cx18: Mark CX18_CPU_DE_RELEASE_MDL as a slow API call
V4L/DVB (9510): cx18: Fix write retries for registers that always change - part 2.
V4L/DVB (9506): ivtv/cx18: fix test whether modules should be loaded or not.
V4L/DVB (9499): cx88-mpeg: final fix for analogue only compilation + de-alloc fix
V4L/DVB (9496): cx88-blackbird: bugfix: cx88-blackbird-mpeg-users
V4L/DVB (9495): cx88-blackbird: bugfix: cx88-blackbird-poll-fix
V4L/DVB (9494): anysee: initialize anysee_usb_mutex statically
V4L/DVB (9492): unplug oops from dvb_frontend_init...
V4L/DVB (9486): ivtv/ivtvfb: no longer experimental
V4L/DVB (9485): ivtv: remove incorrect V4L1 & tvaudio dependency
V4L/DVB (9482): Documentation, especially regarding audio and informational links
V4L/DVB (9475): cx18: Disable write retries for registers that always change - part 1.
...
Linus Torvalds [Tue, 11 Nov 2008 17:22:24 +0000 (09:22 -0800)]
Merge branch 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6
* 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6:
drm/i915: Move legacy breadcrumb out of the reserved status page area
drm/i915: Filter pci devices based on PCI_CLASS_DISPLAY_VGA
drm/radeon: map registers at load time
drm: Remove infrastructure for supporting i915's vblank swapping.
i915: Remove racy delayed vblank swap ioctl.
i915: Don't whine when pci_enable_msi() fails.
i915: Don't attempt to short-circuit object_wait_rendering by checking domains.
i915: Clean up sarea pointers on leavevt
i915: Save/restore MCHBAR_RENDER_STANDBY on GM965/GM45
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6:
dsa: fix master interface allmulti/promisc handling
dsa: fix skb->pkt_type when mac address of slave interface differs
net: fix setting of skb->tail in skb_recycle_check()
net: fix /proc/net/snmp as memory corruptor
mac80211: fix a buffer overrun in station debug code
netfilter: payload_len is be16, add size of struct rather than size of pointer
ipv6: fix ip6_mr_init error path
[4/4] dca: fixup initialization dependency
[3/4] I/OAT: fix async_tx.callback checking
[2/4] I/OAT: fix dma_pin_iovec_pages() error handling
[1/4] I/OAT: fix channel resources free for not allocated channels
ssb: Fix DMA-API compilation for non-PCI systems
SSB: hide empty sub menu
vlan: Fix typos in proc output string
[netdrvr] usb/hso: Cleanup rfkill error handling
sfc: Correct address of gPXE boot configuration in EEPROM
el3_common_init() should be __devinit, not __init
hso: rfkill type should be WWAN
mlx4_en: Start port error flow bug fix
af_key: mark policy as dead before destroying
Ingo Molnar [Tue, 11 Nov 2008 10:57:02 +0000 (11:57 +0100)]
tracing: function return tracer, build fix
fix:
arch/x86/kernel/ftrace.c: In function 'ftrace_return_to_handler':
arch/x86/kernel/ftrace.c:112: error: implicit declaration of function 'cpu_clock'
cpu_clock() is implicitly included via a number of ways, but its real
location is sched.h. (Build failure is triggerable if enough other
kernel components are turned off.)
the [parent_replaced] is used in an =rm fashion, so that constraint
is correct in isolation - but [parent_old] aliases register %0 and uses
it in an addressing mode that is only valid with registers - so change
the constraint from =rm to =r.
Andy Walls [Wed, 5 Nov 2008 03:49:14 +0000 (00:49 -0300)]
V4L/DVB (9516): cx18: Move DVB buffer transfer handling from irq handler to work_queue
cx18: Move DVB buffer transfer handling from irq handler to work_queue thread.
In order to properly lock the epu2cpu mailbox for driver to CX23418 commands,
the DVB/TS buffer handling needs to be moved from the IRQ handler and IRQ
context to a work queue. This work_queue implmentation is strikingly similar
to the ivtv implementation - for better or worse.
Signed-off-by: Andy Walls <awalls@radix.net> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Andy Walls [Wed, 5 Nov 2008 01:02:23 +0000 (22:02 -0300)]
V4L/DVB (9515): cx18: Use correct Mailbox IRQ Ack values and misc IRQ handling cleanup
cx18: Use correct Mailbox IRQ Ack values and misc IRQ handling cleanup.
The SCB field definitions for Ack IRQ's for mailboxes were inconsistent with
the bitmasks being loaded into those SCB fields and the SW2 Ack IRQ handling
logic. Renamed fields in SCB to make things consistent and did misc IRQ
handling cleanups: removing legacy ivtv dma_reg_lock, HPU IRQ flags, etc.
Signed-off-by: Andy Walls <awalls@radix.net> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Frederic CAND [Wed, 29 Oct 2008 17:37:49 +0000 (14:37 -0300)]
V4L/DVB (9493): kconfig patch
Ok I made a patch that converts gspca kconfig file to a more standard=
one, with tabs + 2 white spaces, so that if a warning is added it still
compiles
please find it attached
Andy Walls [Sat, 1 Nov 2008 04:07:36 +0000 (01:07 -0300)]
V4L/DVB (9511): cx18: Mark CX18_CPU_DE_RELEASE_MDL as a slow API call
cx18: Mark CX18_CPU_DE_RELEASE_MDL as a slow API call.
Give the encoder time to complete the MDL release before destroying the
encoder internal task. This avoids an encoder lockup on the next digital
capture and error messages about buffers being returned for an inactive
encoder task handle.
Signed-off-by: Andy Walls <awalls@radix.net> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Andy Walls [Fri, 31 Oct 2008 23:49:12 +0000 (20:49 -0300)]
V4L/DVB (9510): cx18: Fix write retries for registers that always change - part 2.
cx18: Fix write retries for registers that always change - part 2.
Some registers, especially interrupt related ones, will never read
back the value just written. Modified interrupt register readback
checks to make sure the intended effect was achieved.
Signed-off-by: Andy Walls <awalls@radix.net> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
anysee_usb_mutex is initialized at every time the anysee device is probed.
If the second anysee device is probed while anysee_usb_mutex is locked by
the first anysee device, the mutex is broken.
This patch fixes by initialize anysee_usb_mutex statically rather
than initialize at probe time.
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
V4L/DVB (9492): unplug oops from dvb_frontend_init...
When inadvertently hot-unplugging a WT-220U USB DVB-T receiver with
2.6.24, I was met with an oops [1]. The problem is relevant to
2.6.25/26-rc also.
dvb_frontend_init() was called either from re-creation of the kdvb-fe0
thread - seems unlikely, or someone called
dvb_frontend_reinitialise(), causing this path in the thread - really
unlikely, as I can't find any call-site for it.
Either way, quite a number of drivers call dvb_usb_generic_rw() [2]
without checking the validity of the relevant member in the
dvb_usb_device struct - which had changed. Having dvb_usb_generic_rw()
sanity-check and fail (rather than loading from 0x120) seems
reasonable defensive programming [3], in light of it being called in
this way.
The problem with this, is that drivers don't check the return code of
the init call [4]. Does it make sense to cook a patch which allows the
failure to be propagated back up, or am I missing something else?
ivtv used tvaudio in the past and at the time tvaudio required V4L1.
Since tvaudio is no longer dependent on V4L1 and since ivtv actually
no longer uses tvaudio at all, this is no removed from Kconfig.
Without this patch ivtv won't be build if V4L1 is disabled.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Andy Walls [Sun, 26 Oct 2008 02:27:06 +0000 (23:27 -0300)]
V4L/DVB (9475): cx18: Disable write retries for registers that always change - part 1.
cx18: Disable write retries for registers that always change - part 1.
Interrupt related registers will likely not read back the value we just wrote.
Disable retries for these registers for now to avoid accidently discarding
interrupts. More intelligent read back verification criteria are needed for
these and other registers (e.g. GPIO line registers), which will be addressed in
subsequent changes.
Signed-off-by: Andy Walls <awalls@radix.net> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Jean Delvare [Fri, 24 Oct 2008 18:08:28 +0000 (15:08 -0300)]
V4L/DVB (9372): Minor fixes to the saa7110 driver
* Apparently the author of the saa7110 driver was confused by the
number of outputs returned by DECODER_GET_CAPABILITIES. Of course a
decoder chip has no analog ouputs, but it must have at least one
digital output.
* Fix an off-by-one error when checking the input value of
DECODER_SET_INPUT.
Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Rafael Diniz [Sat, 25 Oct 2008 02:07:57 +0000 (23:07 -0300)]
V4L/DVB (9368): VBI fix for cx88 cards
The attached patch fix VBI support cx88 card.
I'm running a capture for hours, getting the closed caption from it[1], and
it's working perfect - the output is the same of a bttv card.
Please apply this patch as soon as possible.
[1] - using zvbi-ntsc-cc of zvbi project.
Signed-off-by: Rafael Diniz <diniz@wimobilis.com.br> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
V4L/DVB (9357): cx88-dvb: Fix Oops in case i2c bus failed to register
There already is an report at kernel bugzilla about this issue:
http://bugzilla.kernel.org/show_bug.cgi?id=9455
When enabling extra checks for the i2c-bus of cx88 based cards by
loading i2c_algo_bit with bit_test=1 this may trigger an oops
when loading cx88_dvb.
This is caused by the extra check code that detects that the
sda-line is stuck high and thus does not register the i2c-bus.
cx88-dvb however does not check if the i2c-bus is valid and just
uses core->i2c_adap to attach dvb frontend modules.
This leads to an oops at the first call to i2c_transfer:
Impact: driver could possibly stomp on resources outside of its scope
{mchehab@redhat.com: I got two versions of the same patch (identical,
except for whitespacing). One authored by Andy Burns and another
authored by Suresh Siddha. Due to that, I'm applying the one that has
less CodingStyle errors. I'm also adding both comments and the SOB's for
both patches, since they are both interesting}
BAR base is located in the middle of the 4K page and the hardcoded
size argument makes the request span two pages causing the conflict.
Fix the hard coded size argument in ioremap().
Andy Burns commented:
I have already sent this patch on the linux-dvb list, but it didn't get
much attention, so re-sending direct, I hope you all don't mind.
While attempting to run mythtv in a xen domU, I encountered problems
loading the driver for my saa7134 card, with an error from ioremap().
This error was due to the driver allocating an incorrectly sized mmio
area, which was trapped by xen's permission checks, but this would go
un-noticed on a kernel without xen.
My card has a 1K sized mmio area, I've had information that other cards
have 2K areas, perhaps others have different sizes, yet the driver
always attempts to map 4K. I realise that the granularity of mapping is
the page size, which typically would be 4K, but unless the card's base
address happens to fall on a 4K boundary (mine does not) then the
base+4K will end up spanning two pages, and this is when the error
occurs under xen.
My patch uses the pci_resource_len macro to determine the size required
for the user's particular card, instead of the hardcoded 4K value. I've
tested with a couple of printk() inside ioremap() that the start address
and size do get rounded to the closest page boundary.
With this patch I am able to successfully load the saa7134 driver and
run mythtv under xen with my card, subject to correct pollirq settings
in case of shared IRQ, I am still seeing occasional DMA panics, which I
think are related to swiotlb handling by dom0/domU, usually the panic
occurs when changing mux, once tuned to a mux, 12 hour continuous
recordings are possible without errors.
Jonathan Corbet [Fri, 17 Oct 2008 23:19:29 +0000 (20:19 -0300)]
V4L/DVB (9355): de-BKL cafe_ccic.c
Remove lock_kernel() call from cafe_ccic.c
Commit d56dc61265d2527a63ab5b0f03199a43cd89ca36 added lock_kernel()
calls to cafe_ccic.c. But that driver was written with proper locking
and does not need the BKL, so take it back out.
Signed-off-by: Jonathan Corbet <corbet@lwn.net> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
As reported by David Ellingsworth:
> I'm not sure if it matters or not, but the ibmcam driver in the
> Mauro's linux-2.6 git tree in the for_linus branch is currently
> broken.
uvd is equal to NULL during most of ibmcam_probe. Due to that, an OOPS is
generated at dev_info. This patch replaces uvd->dev->dev to dev->dev
inside this routine.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com> Reviewed-by: David Ellingsworth <david@identd.dyndns.org>
Manu Abraham [Mon, 20 Oct 2008 21:14:14 +0000 (18:14 -0300)]
V4L/DVB (9346): Optimization: Enable gate in a symmetric/disciplined way,
rather than implementing different ways leading to confusion.
This allows multiple gate_enable/disable's in the tuner_read/write
functions, thereby lesser number of I/O operations throughout,
eventually leading to better results. As a side effect demods that
detect the STOP bit for auto closing of the gate can be avoided, thereby
a very minimal gain in disabling the auto detect feature as well.
Improves readability on the device control.
Signed-off-by: Manu Abraham <manu@linuxtv.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Jean Delvare [Sat, 18 Oct 2008 14:05:58 +0000 (11:05 -0300)]
V4L/DVB (9337b): remove tuner-3036 and dpc7146 drivers from feature-removal-schedule.txt
The tuner-3036 and dpc7146 drivers have been deleted now so we can
remove the corresponding entries from feature-removal-schedule.txt.
(Thanks for doing this, BTW.)
Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Gautham R Shenoy [Sat, 25 Oct 2008 04:52:38 +0000 (10:22 +0530)]
timers: handle HRTIMER_CB_IRQSAFE_UNLOCKED correctly from softirq context
Impact: fix incorrect locking triggered during hotplug-intense stress-tests
While migrating the the CB_IRQSAFE_UNLOCKED timers during a cpu-offline,
we queue them on the cb_pending list, so that they won't go
stale.
Thus, when the callbacks of the timers run from the softirq context,
they could run into potential deadlocks, since these callbacks
assume that they're running with irq's disabled, thereby annoying
lockdep!
Fix this by emulating hardirq context while running these callbacks from
the hrtimer softirq.
=================================
[ INFO: inconsistent lock state ]
2.6.27 #2
--------------------------------
inconsistent {in-hardirq-W} -> {hardirq-on-W} usage.
ksoftirqd/0/4 [HC0[0]:SC1[1]:HE1:SE0] takes:
(&rq->lock){++..}, at: [<c011db84>] sched_rt_period_timer+0x9e/0x1fc
{in-hardirq-W} state was registered at:
[<c014103c>] __lock_acquire+0x549/0x121e
[<c0107890>] native_sched_clock+0x88/0x99
[<c013aa12>] clocksource_get_next+0x39/0x3f
[<c0139abc>] update_wall_time+0x616/0x7df
[<c0141d6b>] lock_acquire+0x5a/0x74
[<c0121724>] scheduler_tick+0x3a/0x18d
[<c047ed45>] _spin_lock+0x1c/0x45
[<c0121724>] scheduler_tick+0x3a/0x18d
[<c0121724>] scheduler_tick+0x3a/0x18d
[<c012c436>] update_process_times+0x3a/0x44
[<c013c044>] tick_periodic+0x63/0x6d
[<c013c062>] tick_handle_periodic+0x14/0x5e
[<c010568c>] timer_interrupt+0x44/0x4a
[<c0150c9f>] handle_IRQ_event+0x13/0x3d
[<c0151c14>] handle_level_irq+0x79/0xbd
[<c0105634>] do_IRQ+0x69/0x7d
[<c01041e4>] common_interrupt+0x28/0x30
[<c047007b>] aac_probe_one+0x1a3/0x3f3
[<c047ec2d>] _spin_unlock_irqrestore+0x36/0x39
[<c01512b4>] setup_irq+0x1be/0x1f9
[<c065d70b>] start_kernel+0x259/0x2c5
[<ffffffff>] 0xffffffff
irq event stamp: 50102
hardirqs last enabled at (50102): [<c047ebf4>] _spin_unlock_irq+0x20/0x23
hardirqs last disabled at (50101): [<c047edc2>] _spin_lock_irq+0xa/0x4b
softirqs last enabled at (50088): [<c0128ba6>] do_softirq+0x37/0x4d
softirqs last disabled at (50099): [<c0128ba6>] do_softirq+0x37/0x4d
other info that might help us debug this:
no locks held by ksoftirqd/0/4.
tracing: add a tracer to catch execution time of kernel functions
Impact: add new tracing plugin which can trace full (entry+exit) function calls
This tracer uses the low level function return ftrace plugin to
measure the execution time of the kernel functions.
The first field is the caller of the function, the second is the
measured function, and the last one is the execution time in
nanoseconds.
- v3:
- HAVE_FUNCTION_RET_TRACER have been added. Each arch that support ftrace return
should enable it.
- ftrace_return_stub becomes ftrace_stub.
- CONFIG_FUNCTION_RET_TRACER depends now on CONFIG_FUNCTION_TRACER
- Return traces printing can be used for other tracers on trace.c
- Adapt to the new tracing API (no more ctrl_update callback)
- Correct the check of "disabled" during insertion.
- Minor changes...
tracing, x86: add low level support for ftrace return tracing
Impact: add infrastructure for function-return tracing
Add low level support for ftrace return tracing.
This plug-in stores return addresses on the thread_info structure of
the current task.
The index of the current return address is initialized when the task
is the first one (init) and when a process forks (the child). It is
not needed when a task does a sys_execve because after this syscall,
it still needs to return on the kernel functions it called.
Note that the code of return_to_handler has been suggested by Steven
Rostedt as almost all of the ideas of improvements in this V3.
For purpose of security, arch/x86/kernel/process_32.c is not traced
because __switch_to() changes the current task during its execution.
That could cause inconsistency in the stored return address of this
function even if I didn't have any crash after testing with tracing on
this function enabled.
Paul Mackerras [Tue, 11 Nov 2008 08:39:26 +0000 (19:39 +1100)]
powerpc: Update desktop/server defconfigs
Turned off CONFIG_PCI_LEGACY and turned on EXT4, and otherwise mostly
took the defaults. This also updates ppc6xx_defconfig, which covers
the 6xx/7xx/7xxx-based embedded boards.
Andreas Schwab [Thu, 6 Nov 2008 00:49:00 +0000 (00:49 +0000)]
powerpc: Fix msr check in compat_sys_swapcontext
The new context may not be 16-byte aligned, so the real address of the
mcontext structure should be read from the uc_regs pointer instead of
directly using the (unaligned) uc_mcontext field.
Signed-off-by: Andreas Schwab <schwab@suse.de> Signed-off-by: Paul Mackerras <paulus@samba.org>
Steven Rostedt [Tue, 11 Nov 2008 04:07:30 +0000 (23:07 -0500)]
ring-buffer: replace most bug ons with warn on and disable buffer
This patch replaces most of the BUG_ONs in the ring_buffer code with
RB_WARN_ON variants. It adds some more variants as needed for the
replacement. This lets the buffer die nicely and still warn the user.
One BUG_ON remains in the code, and that is because it detects a
bad pointer passed in by the calling function, and not a bug by
the ring buffer code itself.
Signed-off-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Steven Rostedt [Tue, 11 Nov 2008 04:07:30 +0000 (23:07 -0500)]
ftrace: prevent ftrace_special from recursion
Impact: stop ftrace_special from recursion
The ftrace_special is used to help debug areas of the kernel.
Because of this, if it is put in certain locations, the fact that
it allows recursion can become a problem if the kernel developer
using does not realize that.
This patch changes ftrace_special to not allow recursion into itself
to make it more robust.
It also changes from preempt disable interrupts disable to prevent
any loss of trace entries.
Signed-off-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Keith Packard [Sat, 8 Nov 2008 01:44:14 +0000 (11:44 +1000)]
drm/i915: Move legacy breadcrumb out of the reserved status page area
Addresses in the hardware status page below index 0x20 are reserved for use
by the hardware. The legacy breadcrumb was sitting at index 5. Move it to
index 0x21, and make sure everyone uses the defined value instead of
hard-coded constants.
Signed-off-by: Keith Packard <keithp@keithp.com> Signed-off-by: Dave Airlie <airlied@linux.ie>
Dave Airlie [Tue, 11 Nov 2008 08:02:12 +0000 (18:02 +1000)]
drm/i915: Filter pci devices based on PCI_CLASS_DISPLAY_VGA
This fixes hangs on 855-class hardware by avoiding double attachment of the
driver due to the stub second head device having the same pci id as the real
device.
Other DRM drivers probably want this treatment as well, but I'm applying it
just to this one for safety. But we should clean up the drm_pciids.h mess
now so that each driver has its own pci id list header in its own directory.
Lets do that in the next release.
Signed-off-by: Eric Anholt <eric@anholt.net> Signed-off-by: Dave Airlie <airlied@redhat.com>
Tejun Heo [Tue, 4 Nov 2008 08:08:40 +0000 (17:08 +0900)]
libata: fix last_reset timestamp handling
ehc->last_reset is used to ensure that resets are not issued too
close to each other. It's initialized to jiffies minus one minute
on EH entry. However, when new links are initialized after PMP is
probed, new links have zero for this timestamp resulting in long wait
depending on the current jiffies.
This patch makes last_set considered iff ATA_EHI_DID_RESET is set, in
which case last_reset is always initialized. As an added precaution,
WARN_ON() is added so that warning is printed if last_reset is
in future.
This problem is spotted and debugged by Shane Huang.
Roland Dreier [Tue, 4 Nov 2008 18:34:48 +0000 (10:34 -0800)]
libata: Avoid overflow in ata_tf_read_block() when tf->hba_lbal > 127
Phillip O'Donnell <phillip.odonnell@gmail.com> pointed out that the same
sign extension bug that was fixed in commit ba14a9c2 ("libata: Avoid
overflow in ata_tf_to_lba48() when tf->hba_lbal > 127") also appears to
exist in ata_tf_read_block(). Fix this by adding a cast to u64.
Signed-off-by: Roland Dreier <rolandd@cisco.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Dave Airlie [Tue, 11 Nov 2008 07:56:16 +0000 (17:56 +1000)]
drm/radeon: map registers at load time
Now that the radeon driver has suspend/resume functions, it needs to map its
registers at load time or it will likely crash if a suspend operation occurs
before the driver has been initialized.
This patch moves the register mapping code from firstopen to load and makes
the mapping into a _DRM_DRIVER one so that the core won't remove it at
lastclose time.
Fixes (at least partially) kernel bz #11891.
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> Signed-off-by: Dave Airlie <airlied@linux.ie>
Eric Anholt [Tue, 4 Nov 2008 20:01:24 +0000 (12:01 -0800)]
i915: Remove racy delayed vblank swap ioctl.
When userland detected that this ioctl was supported (by version number check),
it used it in a racy way -- dispatch delayed swap, wait for vblank, continue
rendering. As there was no mechanism for it to wait for the swap to finish,
sometimes it would render before the swap and garbage would be displayed on
the screen.
By removing the ioctl and returning -EINVAL, userland returns to its previous,
correct rendering path of waiting for a vblank then dispatching a swap. The
only path that could have used this ioctl correctly was page flipping, which
relied on only one client running and emitting wait-for-vblank-before-rendering
in the command stream. That path also falls back correctly, at the performance
cost of not being able to queue up rendering before the flip occurs.
Signed-off-by: Eric Anholt <eric@anholt.net> Signed-off-by: Dave Airlie <airlied@linux.ie>
Owen Taylor [Mon, 3 Nov 2008 22:38:17 +0000 (14:38 -0800)]
i915: Don't attempt to short-circuit object_wait_rendering by checking domains.
This could return early when reading after writing a buffer, if somebody
had already put it on the flushing list (write domains are 0, but still
active), leading to glReadPixels failure.
Signed-off-by: Eric Anholt <eric@anholt.net> Signed-off-by: Dave Airlie <airlied@linux.ie>
Keith Packard [Mon, 3 Nov 2008 07:38:20 +0000 (23:38 -0800)]
i915: Clean up sarea pointers on leavevt
This corresponds to the setup of the sarea pointers in DMA initialization,
though neither is exactly the point at which the sarea is set up or torn down.
Signed-off-by: Keith Packard <keithp@keithp.com> Signed-off-by: Eric Anholt <eric@anholt.net> Signed-off-by: Dave Airlie <airlied@linux.ie>