This patch enables readers to optimistically spin when the
rspin_threshold is non-zero. That threshold value should only
be set when the lock owners of the rwsem are unlikely to go to
sleep. Otherwise enabling reader spinning may make the performance
worse in some cases.
On a 4-socket Haswell mach
The default reader spining threshold is current set to 4096. However,
the right reader spinning threshold may vary from one system to
another and among the different architectures. This patch adds a new
kernel boot parameter to modify the threshold value. This enables
better tailoring to the needs
On Thu, Aug 18, 2016 at 7:26 PM, Minchan Kim wrote:
> Hi Michal,
>
> On Thu, Aug 18, 2016 at 08:01:04PM +0200, Michal Hocko wrote:
>> On Thu 18-08-16 10:47:57, Sonny Rao wrote:
>> > On Thu, Aug 18, 2016 at 12:44 AM, Michal Hocko wrote:
>> > > On Wed 17-08-16 11:57:56, Sonny Rao wrote:
>> [...]
>>
On Thu, Aug 18, 2016 at 11:01 AM, Michal Hocko wrote:
> On Thu 18-08-16 10:47:57, Sonny Rao wrote:
>> On Thu, Aug 18, 2016 at 12:44 AM, Michal Hocko wrote:
>> > On Wed 17-08-16 11:57:56, Sonny Rao wrote:
> [...]
>> >> 2) User space OOM handling -- we'd rather do a more graceful shutdown
>> >> tha
On Thu, Aug 18, 2016 at 2:05 PM, Robert Foss wrote:
>
>
> On 2016-08-18 02:01 PM, Michal Hocko wrote:
>>
>> On Thu 18-08-16 10:47:57, Sonny Rao wrote:
>>>
>>> On Thu, Aug 18, 2016 at 12:44 AM, Michal Hocko wrote:
On Wed 17-08-16 11:57:56, Sonny Rao wrote:
>>
>> [...]
>
> 2) User
2016-08-19 5:11 GMT+08:00 Waiman Long :
> When the count value is in between 0 and RWSEM_WAITING_BIAS, there
> are 2 possibilities.
> Either a writer is present and there is no waiter
count = 0x0001
>or there are waiters and readers. There is no easy way to
count = 0x000X
However, RWSEM
The table there is too big and doesn't have format hints for
LaTeX output.
Fix it.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/dev-subdev.rst | 9 +
1 file changed, 9 insertions(+)
diff --git a/Documentation/media/uapi/v4l/dev-subdev.rst
b/Documentation/media
This table has several troubles:
- a duplicated "struct" on its name;
- a reference to a V4L version 1 struct that will never
point to something (as we got rid of V4L1 API a long
time ago);
- misses hints for LaTeX output (column size and longtable
Fix mosto fo the tables there in order to make them fit at the
page size.
There are, however, two exceptions: RGB and YUV big tables,
where adding the raw latex adjustbox caused the tables to not
be properly formatted. I suspect that the problem is because
those are long tables, but not really sur
One of the tables has wrong column dimentions. Fix it.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/dev-rds.rst | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/Documentation/media/uapi/v4l/dev-rds.rst
b/Documentation/media/uapi/v4l/dev-rds.rst
index
Manually adjust the table width for it to look nice on PDF output.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/control.rst | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/Documentation/media/uapi/v4l/control.rst
b/Documentation/media/uapi/v4l/con
Adjust the table to fit at the LaTeX and PDF outputs, just like
what was done with pixfmt-packed-rgb.rst.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/pixfmt-packed-yuv.rst | 26 ++
1 file changed, 17 insertions(+), 9 deletions(-)
diff --git a/Docume
LaTeX doesn't handle too well auto-width on tables, and ReST
markup requires an special tag to give it the needed hints.
As we're using A4 paper, we have 17cm of useful spaces. As
most media tables have widths, let's use it to generate the
needed via the following perl script:
my ($line_size, $ta
There's one note there that it is indented for no good reason.
Fix it.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/dev-overlay.rst | 10 +-
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/Documentation/media/uapi/v4l/dev-overlay.rst
b/Documentatio
Adjust the tables to fit at the LaTeX and PDF outputs.
Previously, we were displaying the long table in landscape,
but it makes harder to read on displays.
This time, let's use the adjustbox to shrink the size of those
long tables, as the table size can still be visible on screen,
and it is a way
Lots of tables at extended-controls.rst need explicit hints for
LaTeX to adjust their widths. Provide that.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/extended-controls.rst | 63 --
1 file changed, 36 insertions(+), 27 deletions(-)
diff --git a/Doc
This one fix most of the issues with LaTeX output for the media book at
the V4L2 book, before "Function Reference". So, up to page 181 of a 969
pages document.
There are still some things that I was not able to fix on this set:
- the ".. notes::" tag inside a table column are getting the wrong si
The table columns are not properly displayed. Also, some
tables are too big to fit into just one page. So, fix them,
in order to better display the tables.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/buffer.rst | 20 ++--
1 file changed, 14 insertions(+)
Just like on dvb-raw-vbi.rst, the LaTeX output doesn't work
well with cell spans. Also, this is actually a note, so, move
it to a footnote.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/dev-sliced-vbi.rst | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff -
There is an extra column just before eack pack of bits, to
improve table reading, but the header file didn't take this
into account.
Fix it.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/pixfmt-packed-rgb.rst | 78 --
1 file changed, 42 insertions(+),
Those hints are wrong, and doesn't really improve the look
of those tables. So, keep them only when they're useful.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/pixfmt-grey.rst | 7 ---
Documentation/media/uapi/v4l/pixfmt-m420.rst | 6 --
Docume
Add the needed tags to fix LaTeX output of the tables there.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/dev-raw-vbi.rst | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/Documentation/media/uapi/v4l/dev-raw-vbi.rst
b/Documentation/media/uapi/v4
There's a bug with LaTeX output on flat-tables with Sphinx 1.4.5
that prevents references at a cell span to be broken. As the
text is indeed too long, it makes sense to place the reference
to the pictures showing the VBI limits as a footnote.
That makes the text easier to read and also solves the
Better format the tables in a way that will fit inside the
page.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/dev-sliced-vbi.rst | 34 +
1 file changed, 24 insertions(+), 10 deletions(-)
diff --git a/Documentation/media/uapi/v4l/dev-sliced-vbi.rs
Originally, each image were one page big, causing them to be
displayed on separate pages at the PDF output. Re-generate
them from the gif files.
Signed-off-by: Mauro Carvalho Chehab
---
.../media/uapi/v4l/dev-raw-vbi_files/vbi_525.pdf | Bin 3395 -> 3706 bytes
.../media/uapi/v4l/dev-raw-vbi
On Thu, Aug 18, 2016 at 05:46:30PM -0600, Jonathan Corbet wrote:
> So would the old hats be happier with a patch that looks like this? The
> quality of the formatted output suffers slightly, but it's not a big
> deal...
I think this a major improvement. The only thing that still strikes me
as ra
On Thu, Aug 18, 2016 at 12:44 AM, Michal Hocko wrote:
> On Wed 17-08-16 11:57:56, Sonny Rao wrote:
>> On Wed, Aug 17, 2016 at 6:03 AM, Michal Hocko wrote:
>> > On Wed 17-08-16 11:31:25, Jann Horn wrote:
> [...]
>> >> That's at least 30.43% + 9.12% + 7.66% = 47.21% of the task's kernel
>> >> time
Add column hints for LaTeX to format columns on the tables inside
pixfmt-002.rst and pixfmt-006.rst.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/media/uapi/v4l/pixfmt-002.rst | 4 +++-
Documentation/media/uapi/v4l/pixfmt-006.rst | 5 +
2 files changed, 8 insertions(+), 1 deletion(
When the count value is in between 0 and RWSEM_WAITING_BIAS, there
are 2 possibilities. Either a writer is present and there is no
waiter or there are waiters and readers. There is no easy way to
know which is true unless the wait_lock is taken.
This patch changes the RWSEM_WAITING_BIAS from 0xfff
Reader optimistic spinning will be disabled once the rspin_enabled
count reaches 0. After that, it cannot be re-enabled. This may cause
an eligible rwsem locked out of reader spinning because of a series
of unfortunate events.
This patch looks at the regular writer-on-writer spinning history. If
t
v3->v4:
- Rebased to the latest tip tree due to changes to rwsem-xadd.c.
- Update the OSQ patch to fix race condition.
v2->v3:
- Used smp_acquire__after_ctrl_dep() to provide acquire barrier.
- Added the following new patches:
1) make rwsem_spin_on_owner() return a tristate value.
2) rea
Currently, when down_read() fails, the active read locking isn't undone
until the rwsem_down_read_failed() function grabs the wait_lock. If the
wait_lock is contended, it may takes a while to get the lock. During
that period, writer lock stealing will be disabled because of the
active read lock.
T
Almost all the macro definitions in the various architecture specific
rwsem.h header files are essentially the same. This patch moves all
of them into a common header asm-generic/rwsem_types.h to eliminate
the duplication.
Suggested-by: Peter Zijlstra (Intel)
Signed-off-by: Waiman Long
---
arch
The osq_lock() and osq_unlock() function may not provide the necessary
acquire and release barrier in some cases. This patch makes sure
that the proper barriers are provided when osq_lock() is successful
or when osq_unlock() is called.
Suggested-by: Peter Zijlstra (Intel)
Signed-off-by: Waiman Lo
Hi Michal,
On Thu, Aug 18, 2016 at 08:01:04PM +0200, Michal Hocko wrote:
> On Thu 18-08-16 10:47:57, Sonny Rao wrote:
> > On Thu, Aug 18, 2016 at 12:44 AM, Michal Hocko wrote:
> > > On Wed 17-08-16 11:57:56, Sonny Rao wrote:
> [...]
> > >> 2) User space OOM handling -- we'd rather do a more grace
On Tue, 9 Aug 2016 01:31:06 -0700
Christoph Hellwig wrote:
> On Tue, Aug 09, 2016 at 10:28:38AM +0200, Daniel Vetter wrote:
> > The point is to make the docs more discoverable by being able to
> > cross-link them. Old hats like us don't need that, but it definitely
> > has value in bringing new f
On Thu, 18 Aug 2016 11:11:46 +0800
Zhou Wenjian wrote:
Thank you for working to improve the documentation!
> * We generally don' have to bring up a SMP kernel just to capture the
>dump. Hence generally it is useful either to build a UP dump-capture
>kernel or specify maxcpus=1 option wh
On Sat, 13 Aug 2016 18:01:13 +0800
Icenowy Zheng wrote:
> Now, the A83T and A64 SoC user manuals are available.
> Update the documentation to add the links.
>
> An updated version of A83T datasheet is also included now.
Applied to the docs tree, thanks.
jon
--
To unsubscribe from this list: se
Em Thu, 18 Aug 2016 13:15:29 -0300
Mauro Carvalho Chehab escreveu:
> This one fix most of the issues with LaTeX output for the media book at
> the V4L2 book, before "Function Reference". So, up to page 181 of a 969
> pages document.
>
> There are still some things that I was not able to fix on t
On Wed, 3 Aug 2016 00:40:34 +0200
Eugeniu Rosca wrote:
> Improper menuconfig usage leads to empty menu entries.
> zconfdump() is able to reveal some real-life examples:
> - menuconfig VFIO_NOIOMMU
> - menuconfig RESET_CONTROLLER
> - menuconfig SND_ARM
>
> To avoid future occurrences of those, im
Sphinx supports LaTeX output. Sometimes, it is interesting to
call it directly, instead of also generating a PDF. As it comes
for free, add a target for it.
Signed-off-by: Mauro Carvalho Chehab
---
Documentation/DocBook/Makefile | 1 +
Documentation/Makefile.sphinx | 7 ++-
Makefile
+++ Petr Mladek [18/08/16 11:57 +0200]:
On Wed 2016-08-17 20:58:30, Jessica Yu wrote:
Document usage of arch-specific elf sections in livepatch as well
as implementation of arch-specific code.
Signed-off-by: Jessica Yu
---
Documentation/livepatch/module-elf-format.txt | 20 ---
On 2016-08-18 02:01 PM, Michal Hocko wrote:
On Thu 18-08-16 10:47:57, Sonny Rao wrote:
On Thu, Aug 18, 2016 at 12:44 AM, Michal Hocko wrote:
On Wed 17-08-16 11:57:56, Sonny Rao wrote:
[...]
2) User space OOM handling -- we'd rather do a more graceful shutdown
than let the kernel's OOM kill
On Thu, 18 Aug 2016 10:54:19 +0200
Ricardo Ribalda Delgado wrote:
> It is not a very good practice to define the IOCTL manually instead of
> using the header file.
>
> Fix it on the documentation example.
Makes sense to me; applied to the docs tree, thanks.
jon
--
To unsubscribe from this list
This patch modifies rwsem_spin_on_owner() to return a tri-state value
to better reflect the state of lock holder which enables us to make a
better decision of what to do next.
Signed-off-by: Waiman Long
---
kernel/locking/rwsem-xadd.c | 14 +-
1 files changed, 9 insertions(+), 5 de
On Mon, 15 Aug 2016 16:08:23 +0200
Markus Heiser wrote:
> this is my approach to eliminate some distortions we have with the c/cpp
> Sphinx
> domains. The C domain is simple: it assumes that all functions, enums, etc
> are global, e. g. there should be just one function called "ioctl", or "open"
On Sat, 13 Aug 2016 16:12:42 +0200
Markus Heiser wrote:
> Add a generic way to build only a reST sub-folder with or
> without a individual *build-theme*.
>
> * control *sub-folders* by environment SPHINXDIRS
> * control *build-theme* by environment SPHINX_CONF
>
> Folders with a conf.py file, m
When the rwsem is owned by reader, writers stop optimistic spinning
simply because there is no easy way to figure out if all the readers
are actively running or not. However, there are scenarios where
the readers are unlikely to sleep and optimistic spinning can help
performance.
This patch provid
I've applied the whole series (with the small Documentation tweak
suggested by Petr) to livepatching.git#for-4.9/klp-paravirt-alternatives
Thanks,
--
Jiri Kosina
SUSE Labs
--
To unsubscribe from this list: send the line "unsubscribe linux-doc" in
the body of a message to majord...@vger.kernel.
+++ Petr Mladek [18/08/16 11:51 +0200]:
On Wed 2016-08-17 20:58:29, Jessica Yu wrote:
Implement arch_klp_init_object_loaded() for x86, which applies
alternatives/paravirt patches. This fixes the order in which relocations
and alternatives/paravirt patches are applied.
Previously, if a patch mod
On Thu 18-08-16 10:47:57, Sonny Rao wrote:
> On Thu, Aug 18, 2016 at 12:44 AM, Michal Hocko wrote:
> > On Wed 17-08-16 11:57:56, Sonny Rao wrote:
[...]
> >> 2) User space OOM handling -- we'd rather do a more graceful shutdown
> >> than let the kernel's OOM killer activate and need to gather this
Hi Jonathan,
Thanks for your reply.
On 08/19/2016 01:18 AM, Jonathan Corbet wrote:
On Thu, 18 Aug 2016 11:11:46 +0800
Zhou Wenjian wrote:
Thank you for working to improve the documentation!
* We generally don' have to bring up a SMP kernel just to capture the
dump. Hence generally it
Move the rwsem_down_read_failed() function down to below the
optimistic spinning section before enabling optimistic spinning for
the readers. It is because the rwsem_down_read_failed() function will
call rwsem_optimistic_spin() in later patch.
There is no change in code.
Signed-off-by: Waiman Lon
On Tue, 16 Aug 2016 09:30:31 -0700
Laura Abbott wrote:
> KASAN has been supported on arm64 since 39d114ddc682 ("arm64: add KASAN
> support"). Update the docs to indicate this.
This creates a conflict with my patch sphinxifying this file, of course,
but I'll cope. Applied to the docs tree, thank
The following changes since commit
29b4817d4018df78086157ea3a55c1d9424a7cfc:
Linux 4.8-rc1 (2016-08-07 18:18:00 -0700)
are available in the git repository at:
git://git.lwn.net/linux.git tags/4.8-doc-fixes
for you to fetch changes up to d1669c8288a2c86daa6fd59c7fb938c08589d053:
doc-rst:
From: "winton.liu" <18502523...@163.com>
reservation-object-reserve-shared and other API already defined.
delete duplication definitions in fence, it's already defined in reservation.
Signed-off-by: winton.liu <18502523...@163.com>
---
Documentation/DocBook/device-drivers.tmpl |2 --
1 file
On Wed, 17 Aug 2016, Jessica Yu wrote:
> Hi,
>
> A few months ago, Chris Arges reported a bug involving alternatives/paravirt
> patching that was discussed here [1] and here [2]. To briefly summarize the
> bug, patch modules that contained .altinstructions or .parainstructions
> sections would br
On Wed 2016-08-17 20:58:30, Jessica Yu wrote:
> Document usage of arch-specific elf sections in livepatch as well
> as implementation of arch-specific code.
>
> Signed-off-by: Jessica Yu
> ---
> Documentation/livepatch/module-elf-format.txt | 20
> 1 file changed, 16 inserti
On Wed 2016-08-17 20:58:28, Jessica Yu wrote:
> Introduce arch_klp_init_object_loaded() to complete any additional
> arch-specific tasks during patching. Architecture code may override this
> function.
>
> Signed-off-by: Jessica Yu
Reviewed-by: Petr Mladek
Best Regards,
Petr
--
To unsubscribe
On Wed 2016-08-17 20:58:29, Jessica Yu wrote:
> Implement arch_klp_init_object_loaded() for x86, which applies
> alternatives/paravirt patches. This fixes the order in which relocations
> and alternatives/paravirt patches are applied.
>
> Previously, if a patch module had alternatives or paravirt
On Wed, Aug 17, 2016 at 04:26:42PM +0100, Catalin Marinas wrote:
> On Wed, Aug 17, 2016 at 04:32:23PM +0200, Dr. Philipp Tomsich wrote:
> > On 17 Aug 2016, at 16:29, Catalin Marinas wrote:
> > > On Wed, Aug 17, 2016 at 02:54:59PM +0200, Dr. Philipp Tomsich wrote:
> > >> On 17 Aug 2016, at 14:48, Y
It is not a very good practice to define the IOCTL manually instead of
using the header file.
Fix it on the documentation example.
Signed-off-by: Ricardo Ribalda Delgado
---
Documentation/serial/serial-rs485.txt | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/Documentat
On Wed 17-08-16 11:57:56, Sonny Rao wrote:
> On Wed, Aug 17, 2016 at 6:03 AM, Michal Hocko wrote:
> > On Wed 17-08-16 11:31:25, Jann Horn wrote:
[...]
> >> That's at least 30.43% + 9.12% + 7.66% = 47.21% of the task's kernel
> >> time spent on evaluating format strings. The new interface
> >> woul
On 08/18/2016 02:47 PM, Dave Young wrote:
On 08/17/16 at 07:36pm, Joe Perches wrote:
On Thu, 2016-08-18 at 10:31 +0800, Zhou Wenjian wrote:
nr_cpus can help to save memory. So we should remind user of it.
trivia:
diff --git a/Documentation/kdump/kdump.txt b/Documentation/kdump/kdump.txt
[]
64 matches
Mail list logo
