pandora-kernel.git
14 years agotracing/trace_stack: Cleanup for trace_lookup_stack()
Xiao Guangrong [Thu, 16 Jul 2009 06:17:11 +0000 (14:17 +0800)]
tracing/trace_stack: Cleanup for trace_lookup_stack()

We can directly use %pF input format instead of sprint_symbol()
and %s input format.

Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Reviewed-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
14 years agotracing/function: Simplify __ftrace_replace_code()
Xiao Guangrong [Wed, 15 Jul 2009 04:32:15 +0000 (12:32 +0800)]
tracing/function: Simplify __ftrace_replace_code()

Rewrite the __ftrace_replace_code() function, simplify it, but don't
change the code's logic.

First, we get the state we want to set, if the record has the same
state, then do nothing, otherwise enable/disable it.

Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Reviewed-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
14 years agoring_buffer: Fix warning while ignoring cmpxchg return value
Lai Jiangshan [Wed, 15 Jul 2009 08:27:30 +0000 (16:27 +0800)]
ring_buffer: Fix warning while ignoring cmpxchg return value

kernel/trace/ring_buffer.c: In function 'rb_tail_page_update':
kernel/trace/ring_buffer.c:849: warning: value computed is not used
kernel/trace/ring_buffer.c:850: warning: value computed is not used

Add "(void)"s to fix this warning, because we don't need here to handle
the fail case of cmpxchg, it's fine if an interrupt already did the
job.

Changed from V1:
  Add a comment(which is written by Steven) for it.

Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
14 years agoMerge branch 'tip/tracing/ring-buffer-3' of git://git.kernel.org/pub/scm/linux/kernel...
Ingo Molnar [Fri, 10 Jul 2009 11:30:06 +0000 (13:30 +0200)]
Merge branch 'tip/tracing/ring-buffer-3' of git://git./linux/kernel/git/rostedt/linux-2.6-trace into tracing/core

14 years agotracing/workqueues: Add refcnt to struct cpu_workqueue_stats
Lai Jiangshan [Mon, 6 Jul 2009 08:10:23 +0000 (16:10 +0800)]
tracing/workqueues: Add refcnt to struct cpu_workqueue_stats

The stat entries can be freed when the stat file is being read.
The worse is, the ptr can be freed immediately after it's returned
from workqueue_stat_start/next().

Add a refcnt to struct cpu_workqueue_stats to avoid use-after-free.

Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Acked-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A51B16F.6010608@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotracing/stat: Add stat_release() callback
Lai Jiangshan [Mon, 6 Jul 2009 08:10:18 +0000 (16:10 +0800)]
tracing/stat: Add stat_release() callback

Add stat_release() callback to struct tracer_stat, so a stat tracer
can release it's entries after the stat file has been read out.

Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Lai Jiangshan <laijs@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A51B16A.6020708@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agokmemtrace: Rename some functions
Li Zefan [Mon, 6 Jul 2009 08:15:04 +0000 (16:15 +0800)]
kmemtrace: Rename some functions

So we have:

 - kmemtrace_print_alloc/free() for kmemtrace default output

 - kmemtrace_print_alloc/free_user() for binary output used
   by kmemtrace-user.

Suggested-by: Eduard - Gabriel Munteanu <eduard.munteanu@linux360.ro>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Acked-by: Pekka Enberg <penberg@cs.helsinki.fi>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A51B288.70505@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotracing/kmemtrace: Use the %pf format
Frederic Weisbecker [Thu, 9 Jul 2009 02:46:30 +0000 (04:46 +0200)]
tracing/kmemtrace: Use the %pf format

Remove the obsolete seq_print_ip_sym() usage and replace it
by the %pf format in order to print function symbols.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Reviewed-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Lai Jiangshan <laijs@cn.fujitsu.com>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
Cc: Eduard - Gabriel Munteanu <eduard.munteanu@linux360.ro>
LKML-Reference: <1247107590-6428-3-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotracing/function-graph-tracer: Use the %pf format
Frederic Weisbecker [Thu, 9 Jul 2009 02:46:29 +0000 (04:46 +0200)]
tracing/function-graph-tracer: Use the %pf format

Remove the obsolete seq_print_ip_sym() usage and replace it
by the %pf format in order to print function symbols.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Reviewed-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Lai Jiangshan <laijs@cn.fujitsu.com>
LKML-Reference: <1247107590-6428-2-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotracing/filter: Remove empty subsystem and its directory
Xiao Guangrong [Thu, 9 Jul 2009 08:22:22 +0000 (16:22 +0800)]
tracing/filter: Remove empty subsystem and its directory

Remove empty subsystem and its directory when module unload.

Before patch:
 # rmmod trace-events-sample.ko
 # ls sample
 enable  filter

After patch:
 # rmmod trace-events-sample.ko
 # ls sample
 ls: cannot access sample: No such file or directory

Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Acked-by: Tom Zanussi <tzanussi@gmail.com>
Reviewed-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A55A8BE.9010707@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotracing/filter: Remove preds from struct event_subsystem
Xiao Guangrong [Thu, 9 Jul 2009 08:20:12 +0000 (16:20 +0800)]
tracing/filter: Remove preds from struct event_subsystem

No need to save preds to event_subsystem, because it's not used.

Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Acked-by: Tom Zanussi <tzanussi@gmail.com>
Reviewed-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A55A83C.1030005@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoring-buffer: add design document
Steven Rostedt [Wed, 10 Jun 2009 19:45:41 +0000 (15:45 -0400)]
ring-buffer: add design document

This adds the design document for the ring buffer and also
explains how it is designed to have lockless writes.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
14 years agoring-buffer: make lockless
Steven Rostedt [Fri, 27 Mar 2009 15:00:29 +0000 (11:00 -0400)]
ring-buffer: make lockless

This patch converts the ring buffers into a completely lockless
buffer recording system. The read side still takes locks since
we still serialize readers. But the writers are the ones that
must be lockless (those can happen in NMIs).

The main change is to the "head_page" pointer. We write to the
tail, and read from the head. The "head_page" pointer in the cpu
buffer is now just a reference to where to look. The real head
page is now kept in the head_page->list->prev->next pointer.
That is, in the list head of the previous page we set flags.

The list pages are allocated to be aligned such that the lowest
significant bits are always zero pointing to the list. This gives
us play to put in flags to their pointers.

bit 0: set when the page is a head page
bit 1: set when the writer is moving the page (for overwrite mode)

cmpxchg is used to update the pointer.

When the writer wraps the buffer and the tail meets the head,
in overwrite mode, the writer must move the head page forward.
It first uses cmpxchg to change the pointer flag from 1 to 2.
Once this is done, the reader on another CPU will not take the
page from the buffer.

The writers need to protect against interrupts (we don't bother with
disabling interrupts because NMIs are allowed to write too).

After the writer sets the pointer flag to 2, it takes care to
manage interrupts coming in. This is discribed in detail within the
comments of the code.

 Changes in version 2:
  - Let reader reset entries value of header page.
  - Fix tail page passing commit page on reader page test.
  - Always increment entries and write counter in rb_tail_page_update
  - Add safety check in rb_set_commit_to_write to break out of infinite loop
  - add mask in rb_is_reader_page

[ Impact: lock free writing to the ring buffer ]

Signed-off-by: Steven Rostedt <srostedt@redhat.com>
14 years agoring-buffer: make the buffer a true circular link list
Steven Rostedt [Mon, 30 Mar 2009 19:32:01 +0000 (15:32 -0400)]
ring-buffer: make the buffer a true circular link list

This patch changes the ring buffer data pages from using a link list
head pointer, to making each buffer page point to another buffer page
and never back to a "head".

This makes the handling of the ring buffer less complex, since the
traversing of the ring buffer pages no longer needs to account for the
head pointer.

This change also is needed to make the ring buffer lockless.

[
  Changes in version 2:

  - Added change that Lai Jiangshan mentioned.

  From: Lai Jiangshan <laijs@cn.fujitsu.com>
  Date: Thu, 11 Jun 2009 11:25:48 +0800
  LKML-Reference: <4A30793C.6090208@cn.fujitsu.com>

  I'm not sure whether these 4 lines:
bpage = list_entry(pages.next, struct buffer_page, list);
list_del_init(&bpage->list);
cpu_buffer->pages = &bpage->list;

list_splice(&pages, cpu_buffer->pages);
  equal to these 2 lines:
  cpu_buffer->pages = pages.next;
  list_del(&pages);

  If there are equivalent, I think the second one
  are simpler. It may be not a really necessarily cleanup.

  What I asked is: if there are equivalent, could you use these two line:
  cpu_buffer->pages = pages.next;
list_del(&pages);
]

[ Impact: simplify the ring buffer to help make it lockless ]

Signed-off-by: Steven Rostedt <srostedt@redhat.com>
14 years agokmemtrace: Print binary output only if 'bin' option is set
Li Zefan [Fri, 3 Jul 2009 09:34:24 +0000 (17:34 +0800)]
kmemtrace: Print binary output only if 'bin' option is set

Currently by default the output of kmemtrace is binary format instead
of human-readable output.

This patch makes the following changes:

  - We'll see human-readable output by default
  - We'll see binary output if 'bin' option is set

Note: you may probably need to explicitly disable context-info binary
      output:

# echo 0 > options/context-info
# echo 1 > options/bin
# cat trace_pipe

v2:
- use %pF to print call_site

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Acked-by: Pekka Enberg <penberg@cs.helsinki.fi>
Acked-by: Eduard - Gabriel Munteanu <eduard.munteanu@linux360.ro>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A4DD0A0.5060500@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotracing/events: Add trace_event boot option
Li Zefan [Wed, 1 Jul 2009 02:47:05 +0000 (10:47 +0800)]
tracing/events: Add trace_event boot option

We already have ftrace= boot option, and this adds a similar
boot option for trace events, so allow trace events to be
enabled at boot, for boot debugging purpose.

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A4ACE29.3010407@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agokprobes: Clean up insn_pages by using list instead of hlist
Masami Hiramatsu [Tue, 30 Jun 2009 21:08:14 +0000 (17:08 -0400)]
kprobes: Clean up insn_pages by using list instead of hlist

Use struct list instead of struct hlist for managing
insn_pages, because insn_pages doesn't use hash table.

Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com>
Acked-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: Jim Keniston <jkenisto@us.ibm.com>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
LKML-Reference: <20090630210814.17851.64651.stgit@localhost.localdomain>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agokprobes: No need to unlock kprobe_insn_mutex
Masami Hiramatsu [Tue, 30 Jun 2009 21:08:09 +0000 (17:08 -0400)]
kprobes: No need to unlock kprobe_insn_mutex

Remove needless kprobe_insn_mutex unlocking during safety check
in garbage collection, because if someone releases a dirty slot
during safety check (which ensures other cpus doesn't execute
all dirty slots), the safety check must be fail. So, we need to
hold the mutex while checking safety.

Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: Jim Keniston <jkenisto@us.ibm.com>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
LKML-Reference: <20090630210809.17851.28781.stgit@localhost.localdomain>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotracing/fastboot: Document the need of initcall_debug
Li Zefan [Mon, 29 Jun 2009 07:55:10 +0000 (15:55 +0800)]
tracing/fastboot: Document the need of initcall_debug

To use boot tracer, one should pass initcall_debug as well as
ftrace=initcall to the command line.

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A48735E.9050002@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotrace_export: Repair missed fields
Lai Jiangshan [Fri, 26 Jun 2009 09:07:02 +0000 (17:07 +0800)]
trace_export: Repair missed fields

Some fields for struct ftrace_graph_ret are missed
when they are exported to user.

Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A448FB6.5000302@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotracing: Fix stack tracer sysctl handling
Li Zefan [Fri, 26 Jun 2009 08:55:51 +0000 (16:55 +0800)]
tracing: Fix stack tracer sysctl handling

This made my machine completely frozen:

  # echo 1 > /proc/sys/kernel/stack_tracer_enabled
  # echo 2 > /proc/sys/kernel/stack_tracer_enabled

The cause is register_ftrace_function() was called twice.

Also fix ftrace_enabled sysctl, though seems nothing bad happened
as I tested it.

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A448D17.9010305@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoftrace: Fix the output of profile
Li Zefan [Fri, 26 Jun 2009 03:15:37 +0000 (11:15 +0800)]
ftrace: Fix the output of profile

The first entry of the ftrace profile was always skipped when
reading trace_stat/functionX.

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A443D59.4080307@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoring-buffer: Make it generally available
Paul Mundt [Thu, 25 Jun 2009 05:30:12 +0000 (14:30 +0900)]
ring-buffer: Make it generally available

In hunting down the cause for the hwlat_detector ring buffer spew in
my failed -next builds it became obvious that folks are now treating
ring_buffer as something that is generic independent of tracing and thus,
suitable for public driver consumption.

Given that there are only a few minor areas in ring_buffer that have any
reliance on CONFIG_TRACING or CONFIG_FUNCTION_TRACER, provide stubs for
those and make it generally available.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Cc: Jon Masters <jcm@jonmasters.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <20090625053012.GB19944@linux-sh.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoftrace: Remove duplicate newline
Li Zefan [Thu, 25 Jun 2009 06:05:27 +0000 (14:05 +0800)]
ftrace: Remove duplicate newline

Before:
  # echo 'sys_open:traceon:' > set_ftrace_filter
  # echo 'sys_close:traceoff:5' > set_ftrace_filter
  # cat set_ftrace_filter
  #### all functions enabled ####
  sys_open:traceon:unlimited

  sys_close:traceoff:count=0

After:
  # cat set_ftrace_filter
  #### all functions enabled ####
  sys_open:traceon:unlimited
  sys_close:traceoff:count=0

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A4313A7.7030105@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotracing: Fix trace_buf_size boot option
Li Zefan [Wed, 24 Jun 2009 09:33:15 +0000 (17:33 +0800)]
tracing: Fix trace_buf_size boot option

We should be able to specify [KMG] when setting trace_buf_size
boot option, as documented in kernel-parameters.txt

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A41F2DB.4020102@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoftrace: Fix t_hash_start()
Li Zefan [Wed, 24 Jun 2009 01:54:54 +0000 (09:54 +0800)]
ftrace: Fix t_hash_start()

When the output of set_ftrace_filter is larger than PAGE_SIZE,
t_hash_start() will be called the 2nd time, and then we start
from the head of a hlist, which is wrong and causes some entries
to be outputed twice.

The worse is, if the hlist is large enough, reading set_ftrace_filter
won't stop but in a dead loop.

Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A41876E.2060407@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoftrace: Don't manipulate @pos in t_start()
Li Zefan [Wed, 24 Jun 2009 01:54:19 +0000 (09:54 +0800)]
ftrace: Don't manipulate @pos in t_start()

It's rather confusing that in t_start(), in some cases @pos is
incremented, and in some cases it's decremented and then incremented.

This patch rewrites t_start() in a much more general way.

Thus we fix a bug that if ftrace_filtered == 1, functions have tracer
hooks won't be printed, because the branch is always unreachable:

static void *t_start(...)
{
...
if (!p)
return t_hash_start(m, pos);
return p;
}

Before:
  # echo 'sys_open' > /mnt/tracing/set_ftrace_filter
  # echo 'sys_write:traceon:4' >> /mnt/tracing/set_ftrace_filter
  sys_open

After:
  # echo 'sys_open' > /mnt/tracing/set_ftrace_filter
  # echo 'sys_write:traceon:4' >> /mnt/tracing/set_ftrace_filter
  sys_open
  sys_write:traceon:count=4

Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A41874B.4090507@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoftrace: Don't increment @pos in g_start()
Li Zefan [Wed, 24 Jun 2009 01:54:00 +0000 (09:54 +0800)]
ftrace: Don't increment @pos in g_start()

It's wrong to increment @pos in g_start(). It causes some entries
lost when reading set_graph_function, if the output of the file
is larger than PAGE_SIZE.

Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A418738.7090401@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotracing: Reset iterator in t_start()
Li Zefan [Wed, 24 Jun 2009 01:53:44 +0000 (09:53 +0800)]
tracing: Reset iterator in t_start()

The iterator is m->private, but it's not reset to trace_types in
t_start(). If the output is larger than PAGE_SIZE and t_start()
is called the 2nd time, things will go wrong.

Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A418728.5020506@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotrace_stat: Don't increment @pos in seq start()
Li Zefan [Wed, 24 Jun 2009 01:53:26 +0000 (09:53 +0800)]
trace_stat: Don't increment @pos in seq start()

It's wrong to increment @pos in stat_seq_start(). It causes some
stat entries lost when reading stat file, if the output of the file
is larger than PAGE_SIZE.

Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A418716.90209@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotracing_bprintk: Don't increment @pos in t_start()
Li Zefan [Wed, 24 Jun 2009 01:52:58 +0000 (09:52 +0800)]
tracing_bprintk: Don't increment @pos in t_start()

It's wrong to increment @pos in t_start(), otherwise we'll lose
some entries when reading printk_formats, if the output is larger
than PAGE_SIZE.

Reported-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A4186FA.1020106@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agotracing/events: Don't increment @pos in s_start()
Li Zefan [Wed, 24 Jun 2009 01:52:29 +0000 (09:52 +0800)]
tracing/events: Don't increment @pos in s_start()

While testing syscall tracepoints posted by Jason, I found 3 entries
were missing when reading available_events. The output size of
available_events is < 4 pages, which means we lost 1 entry per page.

The cause is, it's wrong to increment @pos in s_start().

Actually there's another bug here -- reading avaiable_events/set_events
can race with module unload:

  # cat available_events               |
      s_start()                        |
      s_stop()                         |
                                       | # rmmod foo.ko
      s_start()                        |
        call = list_entry(m->private)  |

@call might be freed and accessing it will lead to crash.

Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A4186DD.6090405@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agonvidiafb: fix boot-time printk string
Mikael Pettersson [Tue, 23 Jun 2009 19:37:07 +0000 (12:37 -0700)]
nvidiafb: fix boot-time printk string

On bootup nvidiafb prints the following on my Apple G5:

nvidiafb: CRTC 1appears to have a CRT attached

There should be a space between the '1' and the 'appears'.  Add it.

Signed-off-by: Mikael Pettersson <mikpe@it.uu.se>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoh8/300: fix incorrect "select" directives in arch/h8300/Kconfig.cpu.
Robert P. J. Day [Tue, 23 Jun 2009 19:37:06 +0000 (12:37 -0700)]
h8/300: fix incorrect "select" directives in arch/h8300/Kconfig.cpu.

Fix the incorrect "select" directives by dropping the "CONFIG_"
prefixes, and correcting the typo "H8S2768" to "H8S2678".

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Acked-by: Yoshinori Sato <ysato@users.sourceforge.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agomm: fix handling of pagesets for downed cpus
Dimitri Sivanich [Tue, 23 Jun 2009 19:37:04 +0000 (12:37 -0700)]
mm: fix handling of pagesets for downed cpus

After downing/upping a cpu, an attempt to set
/proc/sys/vm/percpu_pagelist_fraction results in an oops in
percpu_pagelist_fraction_sysctl_handler().

If a processor is downed then we need to set the pageset pointer back to
the boot pageset.

Updates of the high water marks should not access pagesets of unpopulated
zones (those pointer go to the boot pagesets which would be no longer
functional if their size would be increased beyond zero).

Signed-off-by: Dimitri Sivanich <sivanich@sgi.com>
Signed-off-by: Christoph Lameter <cl@linux-foundation.org>
Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agormap: fixup page_referenced() for nommu systems
Mike Frysinger [Tue, 23 Jun 2009 19:37:01 +0000 (12:37 -0700)]
rmap: fixup page_referenced() for nommu systems

After the recent changes that went into mm/vmscan.c to overhaul stuff, we
ended up with these warnings on no-mmu systems:

  mm/vmscan.c: In function `shrink_page_list':
  mm/vmscan.c:580: warning: unused variable `vm_flags'
  mm/vmscan.c: In function `shrink_active_list':
  mm/vmscan.c:1294: warning: `vm_flags' may be used uninitialized in this function
  mm/vmscan.c:1242: note: `vm_flags' was declared here

This is because the no-mmu function defines page_referenced() to work on
the first argument only (the page).  It does not clear the vm_flags given
to it because for no-mmu systems, they never actually get utilized.  Since
that is no longer strictly true, we need to set vm_flags to 0 like
everyone else so gcc can do proper dead code elimination without annoying
us with unused warnings.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Cc: David Howells <dhowells@redhat.com>
Acked-by: David McCullough <davidm@snapgear.com>
Cc: Greg Ungerer <gerg@uclinux.org>
Cc: Paul Mundt <lethal@linux-sh.org>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agomm: pass mm to grab_swap_token
Hugh Dickins [Tue, 23 Jun 2009 19:36:58 +0000 (12:36 -0700)]
mm: pass mm to grab_swap_token

If a kthread happens to use get_user_pages() on an mm (as KSM does),
there's a chance that it will end up trying to read in a swap page, then
oops in grab_swap_token() because the kthread has no mm: GUP passes down
the right mm, so grab_swap_token() ought to be using it.

We have not identified a stronger case than KSM's daemon (not yet in
mainline), but the issue must have come up before, since RHEL has included
a fix for this for years (though a different fix, they just back out of
grab_swap_token if current->mm is unset: which is what we first proposed,
but using the right mm here seems more correct).

Reported-by: Izik Eidus <ieidus@redhat.com>
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Acked-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6
Linus Torvalds [Tue, 23 Jun 2009 18:52:16 +0000 (11:52 -0700)]
Merge git://git./linux/kernel/git/lethal/sh-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6:
  sh: Fix up HAVE_PERF_COUNTERS typo.
  sh: Fix up more dma-mapping fallout.
  sh: SH7786 SMP support.
  sh: Wire up the uncached fixmap on sh64 as well.
  sh: Use local TLB flush in set_pte_phys().
  sh: Provide cpu_idle_wait() to fix up cpuidle/SMP build.

14 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland...
Linus Torvalds [Tue, 23 Jun 2009 18:36:01 +0000 (11:36 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/roland/infiniband

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
  RDMA: Add __init/__exit macros to addr.c and cma.c
  IB/ehca: Bump version number
  mlx4_core: Fix dma_sync_single_for_cpu() with matching for_device() calls
  IB/mthca: Replace dma_sync_single() use with proper functions
  RDMA/nes: Fix FIN state handling under error conditions
  RDMA/nes: Fix max_qp_init_rd_atom returned from query device
  IB/ehca: Ensure that guid_entry index is not negative
  IB/ehca: Tolerate dynamic memory operations before driver load

14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6
Linus Torvalds [Tue, 23 Jun 2009 18:35:37 +0000 (11:35 -0700)]
Merge git://git./linux/kernel/git/jejb/scsi-misc-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6: (38 commits)
  fusion: mptsas, fix lock imbalance
  [SCSI] scsi_transport_fc: replace BUS_ID_SIZE by fixed count
  sd, sr: fix Driver 'sd' needs updating message
  scsi_transport_iscsi: return -EOVERFLOW for Too many iscsi targets
  fc_transport: Selective return value from BSG timeout function
  fc_transport: The softirq_done function registration for BSG request
  sym53c8xx: ratelimit parity errors
  explain the hidden scsi_wait_scan Kconfig variable
  ibmvfc: Fix endless PRLI loop in discovery
  ibmvfc: Process async events before command responses
  libfc: Add runtime debugging with debug_logging module parameter
  libfcoe: Add runtime debugging with module param debug_logging
  fcoe: Add runtime debug logging with module parameter debug_logging
  scsi_debug: Add support for physical block exponent and alignment
  cnic: add NETDEV_1000 and NETDEVICES to Kconfig select
  cnic: Fix __symbol_get() build error.
  Revert "[SCSI] cnic: fix error: implicit declaration of function ‘__symbol_get’"
  ipr: differentiate pci-x and pci-e based adapters
  ipr: add test for MSI interrupt support
  scsi_transport_spi: Blacklist Ultrium-3 tape for IU transfers
  ...

14 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arnd/asm...
Linus Torvalds [Tue, 23 Jun 2009 18:34:24 +0000 (11:34 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/arnd/asm-generic

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arnd/asm-generic:
  asm-generic: add dummy pgprot_noncached()
  lib/checksum.c: fix endianess bug
  asm-generic: hook up new system calls
  asm-generic: list Arnd as asm-generic maintainer
  asm-generic: drop HARDIRQ_BITS definition from hardirq.h
  asm-generic: uaccess: fix up local access_ok() usage
  asm-generic: uaccess: add missing access_ok() check to strnlen_user()

14 years agoMerge branch 'kmemleak' of git://linux-arm.org/linux-2.6
Linus Torvalds [Tue, 23 Jun 2009 18:25:04 +0000 (11:25 -0700)]
Merge branch 'kmemleak' of git://linux-arm.org/linux-2.6

* 'kmemleak' of git://linux-arm.org/linux-2.6:
  kmemleak: Do not force the slab debugging Kconfig options
  kmemleak: use pr_fmt

14 years agomm: don't rely on flags coincidence
Hugh Dickins [Tue, 23 Jun 2009 12:52:49 +0000 (13:52 +0100)]
mm: don't rely on flags coincidence

Indeed FOLL_WRITE matches FAULT_FLAG_WRITE, matches GUP_FLAGS_WRITE,
and it's tempting to devise a set of Grand Unified Paging flags;
but not today.  So until then, let's rely upon the compiler to spot
the coincidence, "rather than have that subtle dependency and a
comment for it" - as you remarked in another context yesterday.

Signed-off-by: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Acked-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agohugetlb: fault flags instead of write_access
Hugh Dickins [Tue, 23 Jun 2009 12:49:05 +0000 (13:49 +0100)]
hugetlb: fault flags instead of write_access

handle_mm_fault() is now passing fault flags rather than write_access
down to hugetlb_fault(), so better recognize that in hugetlb_fault(),
and in hugetlb_no_page().

Signed-off-by: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Acked-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoMerge branch 'for-linus' of git://www.jni.nu/cris
Linus Torvalds [Tue, 23 Jun 2009 17:47:01 +0000 (10:47 -0700)]
Merge branch 'for-linus' of git://www.jni.nu/cris

* 'for-linus' of git://www.jni.nu/cris:
  CRISv10: remove redundant tests on unsigned
  CRISv32: irq.c - Move end brace outside #endif
  CRISv32: Fix potential null reference in cryptocop driver.
  CRISv32: Add arch optimized strcmp.
  CRIS: assignment/is equal confusion

14 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
Linus Torvalds [Tue, 23 Jun 2009 17:45:39 +0000 (10:45 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/tiwai/sound-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
  ALSA: hda - Make jack-plug notification selectable
  ALSA: ctxfi - Add PM support
  sound: seq_midi_event: fix decoding of (N)RPN events
  ALSA: hda - Add digital-mic support to ALC262 auto model
  ALSA: hda - Fix check of input source type for realtek codecs
  ALSA: hda - Add quirk for Sony VAIO Z21MN
  ALSA: hda - Get back Input Source for ALC262 toshiba-s06 model
  ALSA: hda - Fix unsigned comparison in patch_sigmatel.c
  ALSA: via82xx: add option to disable 500ms delay in snd_via82xx_codec_wait
  sound: fix check for return value in snd_pcm_hw_refine
  ALSA: ctxfi - Allow unknown PCI SSIDs
  ASoC: Blackfin: update the bf5xx_i2s_resume parameters
  ASoC: Blackfin: keep better track of SPORT configuration state

14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/wim/linux-2.6-watchdog
Linus Torvalds [Tue, 23 Jun 2009 17:41:17 +0000 (10:41 -0700)]
Merge git://git./linux/kernel/git/wim/linux-2.6-watchdog

* git://git.kernel.org/pub/scm/linux/kernel/git/wim/linux-2.6-watchdog:
  [WATCHDOG] wdt_pci.c: remove #ifdef CONFIG_WDT_501_PCI
  [WATCHDOG] hpwdt: Add NMI priority option
  [WATCHDOG] OMAP fixes: enable clock in probe, trigger timer reload
  [WATCHDOG] add bcm47xx watchdog driver
  [WATCHDOG] Freescale STMP: watchdog driver
  [WATCHDOG] twl4030 watchdog driver
  [WATCHDOG] U300 COH 901 327 watchdog driver
  [WATCHDOG] Add pnx833x_wdt

14 years agoMerge branch 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab...
Linus Torvalds [Tue, 23 Jun 2009 17:40:19 +0000 (10:40 -0700)]
Merge branch 'for_linus' of git://git./linux/kernel/git/mchehab/linux-2.6

* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-2.6: (62 commits)
  V4L/DVB (12131): BUGFIX: An incorrect Carrier Recovery Loop optimization table was being
  V4L/DVB (12130): Fix a redundant compiler warning
  V4L/DVB (12003): v4l2: Move bounding code outside I2C ifdef block
  V4L/DVB (11913): cx231xx: TRY_FMT should not actually set anything
  V4L/DVB (11912): em28xx: Use v4l bounding/alignment function
  V4L/DVB (11911): cx231xx: Use v4l bounding/alignment function
  V4L/DVB (11910): mt9: Use v4l bounding/alignment function
  V4L/DVB (11909): cx23885: Use v4l bounding/alignment function
  V4L/DVB (11908): w8968cf: Use v4l bounding/alignment function
  V4L/DVB (11907): cx88: Use v4l bounding/alignment function
  V4L/DVB (11906): saa7134: Use v4l bounding/alignment function
  V4L/DVB (11905): vivi: Use v4l bounding/alignment function
  V4L/DVB (11904): zoran: Use v4l bounding/alignment functiob
  V4L/DVB (11903): sh_mobile_ceu_camera: Use v4l bounding/alignment function
  V4L/DVB (11902): pxa-camera: Use v4l bounding/alignment function
  V4L/DVB (11901): v4l2: Create helper function for bounding and aligning images
  V4L/DVB (12128): v4l2: update framework documentation.
  V4L/DVB (12125): v4l2: add new s_config subdev ops and v4l2_i2c_new_subdev_cfg/board calls
  V4L/DVB (12122): pvrusb2: De-obfuscate code which handles routing schemes
  V4L/DVB (12121): pvrusb2: Improve handling of routing schemes
  ...

14 years agoMerge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzi...
Linus Torvalds [Tue, 23 Jun 2009 17:39:36 +0000 (10:39 -0700)]
Merge branch 'upstream-linus' of git://git./linux/kernel/git/jgarzik/libata-dev

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  libata: don't set IORDY for reset
  sata_fsl: Add power mgmt support
  [libata] PATA driver for CF interface on AT91SAM9260 SoC
  [libata] beautify module parameters

14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/ide-2.6
Linus Torvalds [Tue, 23 Jun 2009 17:39:00 +0000 (10:39 -0700)]
Merge git://git./linux/kernel/git/davem/ide-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/ide-2.6:
  ide cmd64x: Remove serialize setting.
  ide: Take over as maintainer.

14 years agoMerge branches 'ehca', 'misc', 'mlx4', 'mthca' and 'nes' into for-linus
Roland Dreier [Tue, 23 Jun 2009 17:38:47 +0000 (10:38 -0700)]
Merge branches 'ehca', 'misc', 'mlx4', 'mthca' and 'nes' into for-linus

14 years agoRDMA: Add __init/__exit macros to addr.c and cma.c
Peter Huewe [Tue, 23 Jun 2009 17:38:42 +0000 (10:38 -0700)]
RDMA: Add __init/__exit macros to addr.c and cma.c

Add __init and __exit annotations to the module_init/module_exit
functions from drivers/infiniband/core/addr.c and cma.c.

Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Acked-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
14 years agoIB/ehca: Bump version number
Alexander Schmidt [Tue, 23 Jun 2009 17:30:04 +0000 (10:30 -0700)]
IB/ehca: Bump version number

Increment version number for DMEM toleration.

Signed-off-by: Alexander Schmidt <alexs@linux.vnet.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
14 years agomm: fix incorrect page removal from LRU
KAMEZAWA Hiroyuki [Mon, 22 Jun 2009 23:57:55 +0000 (08:57 +0900)]
mm: fix incorrect page removal from LRU

The isolated page is "cursor_page" not "page".

This could cause LRU list corruption under memory pressure, caught by
CONFIG_DEBUG_LIST.

Reported-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Reviewed-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Tested-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agosh: Fix up HAVE_PERF_COUNTERS typo.
Paul Mundt [Tue, 23 Jun 2009 16:41:05 +0000 (01:41 +0900)]
sh: Fix up HAVE_PERF_COUNTERS typo.

That's HAVE_PERF_COUNTERS not HAVE_PERF_COUNTER. This was right
initially but I seem to have screwed it up while re-typing it out
by hand on another machine when I checked it in. Hmph.

Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
14 years agokmemleak: Do not force the slab debugging Kconfig options
Catalin Marinas [Tue, 23 Jun 2009 13:40:27 +0000 (14:40 +0100)]
kmemleak: Do not force the slab debugging Kconfig options

Selecting DEBUG_SLAB or SLUB_DEBUG by the KMEMLEAK menu entry may cause
issues with other dependencies (KMEMCHECK). These configuration options
aren't strictly needed by kmemleak but they may increase the chances of
finding leaks. This patch also updates the KMEMLEAK config entry help
text.

Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Pekka Enberg <penberg@cs.helsinki.fi>
14 years agokmemleak: use pr_fmt
Joe Perches [Tue, 23 Jun 2009 13:40:26 +0000 (14:40 +0100)]
kmemleak: use pr_fmt

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
14 years agoasm-generic: add dummy pgprot_noncached()
Paul Mundt [Tue, 23 Jun 2009 11:51:19 +0000 (13:51 +0200)]
asm-generic: add dummy pgprot_noncached()

Most architectures now provide a pgprot_noncached(), the
remaining ones can simply use an dummy default implementation,
except for cris and xtensa, which should override the
default appropriately.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Jesper Nilsson <jesper.nilsson@axis.com>
Cc: Chris Zankel <chris@zankel.net>
Cc: Magnus Damm <magnus.damm@gmail.com>
14 years agosh: Fix up more dma-mapping fallout.
Paul Mundt [Tue, 23 Jun 2009 08:36:23 +0000 (17:36 +0900)]
sh: Fix up more dma-mapping fallout.

commit dbe6f1869188b6e04e38aa861dd198befb08bcd7
("dma-mapping: mark dma_sync_single and dma_sync_sg as deprecated"
conveniently broke every single SH build.

In the future it would be great if people could at least bother
figuring out how to use grep.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
14 years agosh: SH7786 SMP support.
Paul Mundt [Fri, 19 Jun 2009 07:00:00 +0000 (16:00 +0900)]
sh: SH7786 SMP support.

SH7786 is roughly identical to SH-X3 proto SMP, though there are only 2
CPUs. This just wraps in to the existing SH-X3 SMP code with some minor
changes for SH7786, including wiring up the IPIs properly, enabling
IRQ_PER_CPU, and so forth.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
14 years agosh: Wire up the uncached fixmap on sh64 as well.
Paul Mundt [Fri, 19 Jun 2009 06:39:47 +0000 (15:39 +0900)]
sh: Wire up the uncached fixmap on sh64 as well.

Now that sh64 also can use the uncached section, wire up the fixmap for
it as well.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
14 years agosh: Use local TLB flush in set_pte_phys().
Paul Mundt [Fri, 19 Jun 2009 06:37:11 +0000 (15:37 +0900)]
sh: Use local TLB flush in set_pte_phys().

set_pte_phys() presently uses the global flush_tlb_one(), which locks on
SMP trying to do the IPI. As we have not even initialized the other CPUs
at this point, switch to the local_ variant so the flush happens on the
boot CPU.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
14 years agosh: Provide cpu_idle_wait() to fix up cpuidle/SMP build.
Paul Mundt [Fri, 19 Jun 2009 05:40:51 +0000 (14:40 +0900)]
sh: Provide cpu_idle_wait() to fix up cpuidle/SMP build.

Crib the x86 cpu_idle_wait() implementation and shove it in with the
idle code, subsequently enabling ARCH_HAS_CPU_IDLE_WAIT.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
14 years agoCRISv10: remove redundant tests on unsigned
Roel Kluin [Tue, 23 Jun 2009 08:04:14 +0000 (10:04 +0200)]
CRISv10: remove redundant tests on unsigned

Since dmanr is unsigned, negatives are wrapped and caught by the other test.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
Acked-by: Jesper Nilsson <jesper.nilsson@axis.com>
14 years ago[WATCHDOG] wdt_pci.c: remove #ifdef CONFIG_WDT_501_PCI
Wim Van Sebroeck [Fri, 19 Jun 2009 09:32:57 +0000 (09:32 +0000)]
[WATCHDOG] wdt_pci.c: remove #ifdef CONFIG_WDT_501_PCI

Change the wdt_pci.c watchdog driver so that the code is the same for
both the PCI-WDT500 as the PCI-WDT501 card. The selection of the card
is now being done via the module parameter: 'type' instead of the
config option CONFIG_WDT_501_PCI.

Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
14 years ago[WATCHDOG] hpwdt: Add NMI priority option
Tom Mingarelli [Thu, 18 Jun 2009 23:28:57 +0000 (23:28 +0000)]
[WATCHDOG] hpwdt: Add NMI priority option

Add a priority option so that the user can choose if we do the NMI
first or last.

Signed-off-by: Thomas Mingarelli <thomas.mingarelli@hp.com>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
14 years ago[WATCHDOG] OMAP fixes: enable clock in probe, trigger timer reload
Ulrik Bech Hald [Fri, 12 Jun 2009 21:18:32 +0000 (16:18 -0500)]
[WATCHDOG] OMAP fixes: enable clock in probe, trigger timer reload

This patch contains two fixes:

1)In omap_wdt_probe() the watchdog is reset and disabled. This
requires register access and the clks needs to be enabled temporarily

2)In omap_wdt_open() the timer register needs to be reloaded
to trigger a new timer value (the default of 60s)

Tested on OMAP34xx platform (Zoom1)

Reviewed-by: Kevin Hilman <khilman@ti.deeprootsystems.com>
Signed-off-by: Ulrik Bech Hald <ubh@ti.com>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
14 years agoV4L/DVB (12131): BUGFIX: An incorrect Carrier Recovery Loop optimization table was...
Manu Abraham [Thu, 18 Jun 2009 07:50:53 +0000 (04:50 -0300)]
V4L/DVB (12131): BUGFIX: An incorrect Carrier Recovery Loop optimization table was being

loaded for a given chip version. This would cause the optimization in
tuning not to be applied and thus a failed expectation, in tuning speed
increment. The patch swaps the tables in use. It also fixes a possible
one in a million condition where state->dev_ver implies an older Cut
(Cut < 2.0, eventhough the driver doesn't attach to any Cut older than
2.0) or even negative (due to a bad I2C bus master driver) for the card
combination.

Thanks to Mauro Carvalho Chehab <mchehab@infradead.org> for pointing
out the issue at large.

Signed-off-by: Manu Abraham <manu@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12130): Fix a redundant compiler warning
Manu Abraham [Wed, 17 Jun 2009 19:48:17 +0000 (16:48 -0300)]
V4L/DVB (12130): Fix a redundant compiler warning

drivers/media/dvb/frontends/stv090x.c: In function ‘stv090x_optimize_carloop_short’:
drivers/media/dvb/frontends/stv090x.c:2677: warning: ‘short_crl’ may be used uninitialized in this function

Signed-off-by: Manu Abraham <manu@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12003): v4l2: Move bounding code outside I2C ifdef block
Trent Piepho [Fri, 12 Jun 2009 19:31:29 +0000 (16:31 -0300)]
V4L/DVB (12003): v4l2: Move bounding code outside I2C ifdef block

On Fri, 12 Jun 2009, Randy Dunlap wrote:
> From: Randy Dunlap <randy.dunlap@oracle.com>
>
> Move v4l_bound_align_image() outside of an #ifdef CONFIG_I2C block
> so that it is always built.  Fixes a build error:

clamp_align() should be moved as well, since it's only used by
v4l_bound_align_image().  I'm attaching an alternate version that fixes
this.  Labeled the endif too.

Reported-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11913): cx231xx: TRY_FMT should not actually set anything
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11913): cx231xx: TRY_FMT should not actually set anything

In the TRY_FMT handler the function get_scale() is called to find what the
scaler hardware will produce for a requested size.

The problem is that get_scale(struct cx231xx *dev, ..., unsigned int *vscale,
unsigned int *hscale) saves the calculated scale values into both the
pointer arguments and into dev's hscale and vscale fields.  TRY_FMT shouldn't
actually change anything in the device state.

The code to in get_scale() that writes to dev->[hv]scale can just be
deleted.  In all cases when dev's fields should be modified, get_scale()
was called with get_scale(dev, ..., &dev->hscale, &dev->vscale), so dev was
getting updated anyway.

This didn't actually cause a problem because nothing ever actually made use
of the hscale and vscale fields.  I changed cx231xx_resolution_set() to use
those fields rather than re-calculate them with a call to get_scale().

Updating [hv]scale in cx231xx_resolution_set() isn't necessary because
every call of cx231xx_resolution_set() was already preceded by a call to
get_scale() or setting the [hv]scale fields, so they will be always be
up-to-date w.r.t. width and height.

Removing the call to get_scale() from cx231xx_resolution_set() allowed
making get_scale() a static function, which is a good thing for something
with such a short name.  There is already another function with the same
name in the em28xx driver, but that one is static.

Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11912): em28xx: Use v4l bounding/alignment function
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11912): em28xx: Use v4l bounding/alignment function

The v4l function has a better algorithm for aligning image size.

It appears that the em2800 can only scale by 50% or 100%, i.e. the only
heights supported might be 240 and 480.  In that case the old code would
set any height other than 240 to 480.  Request 240 get 240, but request 239
and then you get 480.  Change it to round to the nearest supported value.

Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11911): cx231xx: Use v4l bounding/alignment function
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11911): cx231xx: Use v4l bounding/alignment function

The v4l function has a better algorithm for aligning image size.

Cc: Srinivasa Deevi <srinivasa.deevi@conexant.com>
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11910): mt9: Use v4l bounding/alignment function
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11910): mt9: Use v4l bounding/alignment function

The v4l function has a better algorithm for aligning image size.

Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11909): cx23885: Use v4l bounding/alignment function
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11909): cx23885: Use v4l bounding/alignment function

The v4l function has a better algorithm for aligning image size.

Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11908): w8968cf: Use v4l bounding/alignment function
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11908): w8968cf: Use v4l bounding/alignment function

The v4l function has a better algorithm for aligning image size.

The existing code was casting pointers to u32 and to unsigned int into
pointers to u16.  This could mess up if someone passed in an image size
greater than 65,535 and on big-endian platforms it won't work at all.

The existing bounding code would shrink an image if it was too big, but
returned ERANGE if it was too small.  The code will not shrink or expand as
necessary.

Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11907): cx88: Use v4l bounding/alignment function
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11907): cx88: Use v4l bounding/alignment function

The v4l function has a better algorithm for aligning image size.

Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11906): saa7134: Use v4l bounding/alignment function
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11906): saa7134: Use v4l bounding/alignment function

The v4l function has a better algorithm for aligning image size.

Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11905): vivi: Use v4l bounding/alignment function
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11905): vivi: Use v4l bounding/alignment function

The v4l function has a better algorithm for aligning image size.

Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11904): zoran: Use v4l bounding/alignment functiob
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11904): zoran: Use v4l bounding/alignment functiob

The v4l function has a better algorithm for aligning image size.

Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11903): sh_mobile_ceu_camera: Use v4l bounding/alignment function
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11903): sh_mobile_ceu_camera: Use v4l bounding/alignment function

The v4l function has a better algorithm for aligning image size.

Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11902): pxa-camera: Use v4l bounding/alignment function
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11902): pxa-camera: Use v4l bounding/alignment function

The v4l function has a better algorithm for aligning image size.

For instance the old code would change 159x243 into 156x240 to meet the
alignment requirements.  The new function will use 160x243, which is a lot
closer to what was asked for originally.

Cc: Robert Jarzmik <robert.jarzmik@free.fr>
Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (11901): v4l2: Create helper function for bounding and aligning images
Trent Piepho [Sun, 31 May 2009 00:45:46 +0000 (21:45 -0300)]
V4L/DVB (11901): v4l2: Create helper function for bounding and aligning images

Most hardware has limits on minimum and maximum image dimensions and also
requirements about alignment.  For example, image width must be even or a
multiple of four.  Some hardware has requirements that the total image size
(width * height) be a multiple of some power of two.

v4l_bound_align_image() will enforce min and max width and height, power of
two alignment on width and height, and power of two alignment on total
image size.

It uses an efficient algorithm that will try to find the "closest" image
size that meets the requirements.

Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12128): v4l2: update framework documentation.
Hans Verkuil [Tue, 9 Jun 2009 20:29:29 +0000 (17:29 -0300)]
V4L/DVB (12128): v4l2: update framework documentation.

Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12125): v4l2: add new s_config subdev ops and v4l2_i2c_new_subdev_cfg/board...
Hans Verkuil [Tue, 9 Jun 2009 20:12:33 +0000 (17:12 -0300)]
V4L/DVB (12125): v4l2: add new s_config subdev ops and v4l2_i2c_new_subdev_cfg/board calls

Add a new s_config core ops call: this is called with the irq and platform
data to be used to initialize the subdev.

Added new v4l2_i2c_new_subdev_cfg and v4l2_i2c_new_subdev_board calls
that allows you to pass these new arguments.

The existing v4l2_i2c_new_subdev functions were modified to also call
s_config.

In the future the existing v4l2_i2c_new_subdev functions will be replaced
by a single v4l2_i2c_new_subdev function similar to v4l2_i2c_new_subdev_cfg
but without the irq and platform_data arguments.

Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12122): pvrusb2: De-obfuscate code which handles routing schemes
Mike Isely [Sat, 20 Jun 2009 17:57:24 +0000 (14:57 -0300)]
V4L/DVB (12122): pvrusb2: De-obfuscate code which handles routing schemes

This change does not change any outward behavior; it merely chops down
some large if-conditions with embedded assignments into something a
little more maintainable for others (I of course never had a problem
with this...).

Signed-off-by: Mike Isely <isely@pobox.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12121): pvrusb2: Improve handling of routing schemes
Mike Isely [Sat, 20 Jun 2009 17:55:31 +0000 (14:55 -0300)]
V4L/DVB (12121): pvrusb2: Improve handling of routing schemes

The pvrusb2 driver has a concept of "routing scheme" which defines
which physical inputs should be connected based on application's
choice of logical input.  The correct "routing scheme" depends on the
specific device since different devices might wire up their muxes

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12120): pvrusb2: Change initial default frequency setting
Mike Isely [Sat, 20 Jun 2009 17:51:29 +0000 (14:51 -0300)]
V4L/DVB (12120): pvrusb2: Change initial default frequency setting

Change default frequency to be US Broadcast channel 3 - with the
transition to d igital the previous value has now become useless.
This change is PURELY to help with my testing (I need to set some kind
of default so it might as well be some thing usable).

Signed-off-by: Mike Isely <isely@pobox.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12119): pvrusb2: Re-fix hardware scaling on video standard change
Mike Isely [Sat, 20 Jun 2009 17:50:14 +0000 (14:50 -0300)]
V4L/DVB (12119): pvrusb2: Re-fix hardware scaling on video standard change

The cx25840 module's VBI initialization logic uses the current video
standard as part of its internal algorithm.  This therefore means that
we probably need to make sure that the correct video standard has been
set before initializing VBI.  (Normally we would not care about VBI,
but as described in an earlier changeset, VBI must be initialized
correctly on the cx25840 in order for the chip's hardware scaler to
operate correctly.)

It's kind of messy to force the video standard to be set before
initializing VBI (mainly because we can't know what the app really
wants that early in the initialization process).  So this patch does
the next best thing: VBI is re-initialized after any point where the
video standard has been set.

Signed-off-by: Mike Isely <isely@pobox.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12118): pvrusb2: Fix hardware scaling when used with cx25840
Mike Isely [Sat, 20 Jun 2009 17:45:52 +0000 (14:45 -0300)]
V4L/DVB (12118): pvrusb2: Fix hardware scaling when used with cx25840

The cx25840 module requires that its VBI initialization entry point be
called in order for hardware-scaled video capture to work properly -
even if we don't care about VBI.  Making this behavior even more
subtle is that if the capture resolution is set to 720x480 - which is
the default that the pvrusb2 driver sets up - then the cx25840
bypasses the hardware scaler.  Therefore this problem does not
manifest itself until some other resolution, e.g. 640x480, is tried.
MythTV typically defaults to 640x480 or 480x480, which means that
things break whenever the driver is used with MythTV.

This all has been known for a while (since at least Nov 2006), but
recent changes in the pvrusb2 driver (specifically in regards to
sub-device support) caused this to break again.  VBI initialization
must happen *after* the chip's firmware is loaded, not before.  With
this fix, 24xxx devices work correctly again.

A related fix that is part of this changeset is that now we
re-initialize VBI any time after we issue a reset to the cx25840
driver.  Issuing a chip reset erases the state that the VBI setup
previously did.  Until the HVR-1950 came along this subtlety went
unnoticed, because the pvrusb2 driver previously never issued such a
reset.  But with the HVR-1950 we have to do that reset in order to
correctly transition from digital back to analog mode - and since the
HVR-1950 always starts in digital mode (required for the DVB side to
initialize correctly) then this device has never had a chance to work
correctly in analog mode!  Analog capture on the HVR-1950 has been
broken this *ENTIRE* time.  I had missed it until now because I've
usually been testing at the default 720x480 resolution which does not
require scaling...  What fun.  By re-initializing VBI after a cx25840
chip reset, correct behavior is restored.

Signed-off-by: Mike Isely <isely@pobox.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12116): cx23885: ensure correct IF freq is used on HVR1200 & HVR1700
Michael Krufky [Thu, 21 May 2009 15:49:28 +0000 (12:49 -0300)]
V4L/DVB (12116): cx23885: ensure correct IF freq is used on HVR1200 & HVR1700

Ensure that we're programming the tda18271 tuner with the correct
IF frequencies to match the programming of the TDA10048 DVB-T demod
for the HVR1200 and HVR1700 products.

Signed-off-by: Michael Krufky <mkrufky@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12115): tda10048: add missing entry to pll_tab for 3.8 MHz IF
Michael Krufky [Fri, 19 Jun 2009 19:21:37 +0000 (16:21 -0300)]
V4L/DVB (12115): tda10048: add missing entry to pll_tab for 3.8 MHz IF

Thanks for Terry Wu for pointing out the missing entry.

Cc: Terry Wu <terrywu2009@gmail.com>
Signed-off-by: Michael Krufky <mkrufky@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12112): cx231xx: fix uninitialized variable.
Hans Verkuil [Sat, 20 Jun 2009 12:21:37 +0000 (09:21 -0300)]
V4L/DVB (12112): cx231xx: fix uninitialized variable.

The variable 'rc' could be used uninitialized in the cx231xx_capture_start
function. Sri informed me that it should be initialized to -1.

Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12111): tcm825x: remove incorrect __exit_p wrapper
Hans Verkuil [Sat, 20 Jun 2009 12:18:34 +0000 (09:18 -0300)]
V4L/DVB (12111): tcm825x: remove incorrect __exit_p wrapper

tcm825x_remove is not necessarily called on module exit, it can also be
called when the i2c_adapter is removed. While the i2c adapter might never
be removed on an embedded system, in practice this sensor driver can also
be used in e.g. a USB webcam where this is a perfectly acceptable thing
to do.

Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12109): radio-tea5764: fix incorrect rxsubchans value
Hans Verkuil [Sat, 20 Jun 2009 09:29:12 +0000 (06:29 -0300)]
V4L/DVB (12109): radio-tea5764: fix incorrect rxsubchans value

rxsubchans was only set when stereo was detected, otherwise it was
left to 0 instead of setting it to mono.

Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12108): v4l2-i2c-drv.h: add comment describing when not to use this header.
Hans Verkuil [Sun, 14 Jun 2009 10:12:11 +0000 (07:12 -0300)]
V4L/DVB (12108): v4l2-i2c-drv.h: add comment describing when not to use this header.

Make it very clear that this header should not be used for i2c drivers that
do not need to be compiled for pre-2.6.26 kernels.

As soon as the minimum supported kernel in the v4l-dvb repository becomes
2.6.26 or up, then this header should be removed entirely.

Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12107): smscoreapi: fix compile warning
Hans Verkuil [Sun, 14 Jun 2009 09:47:35 +0000 (06:47 -0300)]
V4L/DVB (12107): smscoreapi: fix compile warning

gcc 4.3.1 generates this warning:

v4l/smscoreapi.c: In function 'smscore_gpio_configure':
v4l/smscoreapi.c:1481: warning: 'GroupNum' may be used uninitialized in this function
v4l/smscoreapi.c:1480: warning: 'TranslatedPinNum' may be used uninitialized in this function

While in practice this will not happen, it is something that the compiler
can't determine. Initializing these two local variables to 0 suppresses
this warning.

Cc: Udi Atar <udi.linuxtv@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12104): ivtv/cx18: fix regression: class controls are no longer seen
Hans Verkuil [Sun, 14 Jun 2009 08:20:21 +0000 (05:20 -0300)]
V4L/DVB (12104): ivtv/cx18: fix regression: class controls are no longer seen

A previous change (v4l2-common: remove v4l2_ctrl_query_fill_std) broke
the handling of class controls in VIDIOC_QUERYCTRL. The MPEG class control
was broken for all drivers that use the cx2341x module and the USER class
control was broken for ivtv and cx18.

This change adds back proper class control support.

Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12102): em28xx: add Remote control support for EVGA inDtube
Devin Heitmueller [Sun, 21 Jun 2009 00:34:42 +0000 (21:34 -0300)]
V4L/DVB (12102): em28xx: add Remote control support for EVGA inDtube

Add an IR profile for the EVGA inDtube remote control (which is an NEC type
remote)

Signed-off-by: Devin Heitmueller <dheitmueller@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
14 years agoV4L/DVB (12101): em28xx: add support for EVGA inDtube
Devin Heitmueller [Fri, 19 Jun 2009 03:33:54 +0000 (00:33 -0300)]
V4L/DVB (12101): em28xx: add support for EVGA inDtube

Add support for the EVGA inDtube.  Both ATSC and analog side validated as
fully functional.

Thanks to Jake Crimmins from EVGA for providing the correct GPIO info.
Thanks to Alan Hagge for doing all the device testing.
Thanks to Greg Williamson for providing hardware for testing.

Cc: Jake Crimmins <jcrimmins@evga.com>
Cc: Alan Hagge <ahagge@gmail.com>
Cc: Greg Williamson <cheeseboy16@gmail.com>
Signed-off-by: Devin Heitmueller <dheitmueller@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>