This is where the dynamically allocated task_autonuma structure is
being handled.
This is the structure holding the per-thread NUMA statistics generated
by the NUMA hinting page faults. This per-thread NUMA statistical
information is needed by sched_autonuma_balance to make optimal NUMA
balancing
Until THP native migration is implemented it's safer to boost
khugepaged scanning rate because all memory migration are splitting
the hugepages. So the regular rate of scanning becomes too low when
lots of memory is migrated.
Signed-off-by: Andrea Arcangeli
---
mm/huge_memory.c |6 ++
1
Add the config options to allow building the kernel with AutoNUMA.
If CONFIG_AUTONUMA_DEFAULT_ENABLED is "=y", then
/sys/kernel/mm/autonuma/enabled will be equal to 1, and AutoNUMA will
be enabled automatically at boot.
Signed-off-by: Andrea Arcangeli
---
arch/Kconfig |3 +++
arch/x86/K
When pages are collapsed try to keep the last_nid information from one
of the original pages.
Signed-off-by: Andrea Arcangeli
---
mm/huge_memory.c | 14 ++
1 files changed, 14 insertions(+), 0 deletions(-)
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index a65590f..0d2a12f 100
Move the AutoNUMA per page information from the "struct page" to a
separate page_autonuma data structure allocated in the memsection
(with sparsemem) or in the pgdat (with flatmem).
This is done to avoid growing the size of "struct page". The
page_autonuma data is only allocated if the kernel is b
This algorithm takes as input the statistical information filled by the
knuma_scand (mm->mm_autonuma) and by the NUMA hinting page faults
(p->task_autonuma), evaluates it for the current scheduled task, and
compares it against every other running process to see if it should
move the current task to
On Wed, 22 Aug 2012, Konrad Rzeszutek Wilk wrote:
> On Wed, Aug 22, 2012 at 12:20:09PM +0100, Stefano Stabellini wrote:
> > Konrad,
> > I cannot see this patch anywhere in your trees. Did I miss it?
> > Or maybe it just fell through the cracks?
>
> Fell through the cracks.. Was there a new version
Hello everyone,
Before the Kernel Summit, I think it's good idea to post a new
AutoNUMA24 and to go through a new review cycle. The last review cycle
has been fundamental in improving the patchset. Thanks!
The objective of AutoNUMA is to be able to perform as close as
possible to (and sometime fa
In the special "pmd" mode of knuma_scand
(/sys/kernel/mm/autonuma/knuma_scand/pmd == 1), the pmd may be of numa
type (_PAGE_PRESENT not set), however the pte might be
present. Therefore, gup_pmd_range() must return 0 in this case to
avoid losing a NUMA hinting page fault during gup_fast.
Note: gup
>From 32 to 12 bytes, so the AutoNUMA memory footprint is reduced to
0.29% of RAM.
This however will fail to migrate pages above a 16 Terabyte offset
from the start of each node (migration failure isn't fatal, simply
those pages will not follow the CPU, a warning will be printed in the
log just on
This is where the numa hinting page faults are detected and are passed
over to the AutoNUMA core logic.
Signed-off-by: Andrea Arcangeli
---
include/linux/huge_mm.h |2 ++
mm/huge_memory.c| 18 ++
mm/memory.c | 31 +++
3 file
This defines the knuma_migrated queues. There is one knuma_migrated
per NUMA node with active CPUs. Pages are added to these queues
through the NUMA hinting page fault (memory follow CPU algorithm with
false sharing evaluation). The daemons are then woken up with a
certain hysteresis to migrate the
Link the AutoNUMA core and scheduler object files in the kernel if
CONFIG_AUTONUMA=y.
Signed-off-by: Andrea Arcangeli
---
kernel/sched/Makefile |1 +
mm/Makefile |1 +
2 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/kernel/sched/Makefile b/kernel/sched/Makefile
i
is_vma_temporary_stack() is needed by mm/autonuma.c too, and without
this the build breaks with CONFIG_TRANSPARENT_HUGEPAGE=n.
Reported-by: Petr Holasek
Acked-by: Rik van Riel
Signed-off-by: Andrea Arcangeli
---
include/linux/huge_mm.h |4 ++--
1 files changed, 2 insertions(+), 2 deletions
We will set these bitflags only when the pmd or pte is not present.
They work like PROTNONE but they identify a request for the numa
hinting page fault to trigger.
Because we want to be able to set these bitflags in any established
pte or pmd (while clearing the present bit at the same time) with
On 22/08/12 15:19, Thomas Gleixner wrote:
On Wed, 22 Aug 2012, Konrad Rzeszutek Wilk wrote:
On Tue, Aug 21, 2012 at 11:22:03PM +0200, Thomas Gleixner wrote:
On Tue, 21 Aug 2012, Attilio Rao wrote:
Differences with v1:
- The patch serie is re-arranged in a way that it helps re
These flags are the ones tweaked through sysfs, they control the
behavior of autonuma, from enabling disabling it, to selecting various
runtime options.
Signed-off-by: Andrea Arcangeli
---
include/linux/autonuma_flags.h | 129
1 files changed, 129 insert
This implements the following parts of autonuma:
o knuma_scand: daemon for setting pte_numa and pmd_numa while
gathering NUMA mm stats
o NUMA hinting page fault handler: queues pages for migration and
gathers NUMA task stats
o knuma_migrated: kernel threads that migrate memory from remote no
From: Vaidyanathan Srinivasan
* PMD flaging is not required in powerpc since large pages
are tracked in ptes.
* Yet to be tested with large pages
* This is an initial patch that partially works
* knuma_scand and numa hinting page faults works
* Page migration is yet to b
Remove the sysfs entry /sys/kernel/mm/autonuma/knuma_scand/pmd and
force the knuma_scand pmd mode off if
CONFIG_HAVE_ARCH_AUTONUMA_SCAN_PMD is not set by the architecture.
Enable AutoNUMA for PPC64.
Signed-off-by: Andrea Arcangeli
---
arch/Kconfig |3 +++
arch/powerpc/Kconfig |6
This function makes it easy to bind the per-node knuma_migrated
threads to their respective NUMA nodes. Those threads take memory from
the other nodes (in round robin with a incoming queue for each remote
node) and they move that memory to their local node.
Signed-off-by: Andrea Arcangeli
---
in
Header that defines the generic AutoNUMA specific functions.
All functions are defined unconditionally, but are only linked into
the kernel if CONFIG_AUTONUMA=y. When CONFIG_AUTONUMA=n, their call
sites are optimized away at build time (or the kernel wouldn't link).
Signed-off-by: Andrea Arcangel
On Wed, Aug 22, 2012 at 5:43 AM, Tiejun Chen wrote:
> +int kgdb_skipexception(int exception, struct pt_regs *regs)
> +{
> + if (kgdb_isremovedbreak(regs->nip))
> + return 1;
> +
> + return 0;
> +}
int kgdb_skipexception(int exception, struct pt_regs *regs)
{
ret
On Wed, Aug 22, 2012 at 03:12:46PM +0100, Jan Beulich wrote:
> >>> On 21.08.12 at 21:03, Konrad Rzeszutek Wilk
> >>> wrote:
> > On Tue, Aug 21, 2012 at 01:27:32PM -0400, Konrad Rzeszutek Wilk wrote:
> >> Jan, I thought something odd. Part of this code replaces this:
> >>
> >>memblock_reserve
This is where the mm_autonuma structure is being handled.
mm_autonuma holds the link for knuma_scand's list of mm structures to
scan and a pointer to the associated mm structure for knuma_scand's
convenience.
It also contains the per-mm NUMA statistics collected by knuma_scand
daemon. The per-mm
Working set estimation will only record memory that was recently used
and in turn will be eligible for automatic migration. It will ignore
memory that is never accessed by the process and that in turn will
never attempted to be migrated. This can increase NUMA convergence if
large areas of memory a
If an task_selected_nid has already been selected for the task, try to
allocate memory from it even if it's temporarily not the local
node. Chances are it's where most of its memory is already located and
where it will run in the future.
Acked-by: Rik van Riel
Signed-off-by: Andrea Arcangeli
---
Reduce the autonuma_migrate_head array entries from MAX_NUMNODES to
num_possible_nodes() or zero if autonuma is not possible.
Signed-off-by: Andrea Arcangeli
---
arch/x86/mm/numa.c |6 --
arch/x86/mm/numa_32.c |3 ++-
include/linux/memory_hotplug.h |3 ++-
in
When pages are freed abort any pending migration. If knuma_migrated
arrives first it will notice because get_page_unless_zero would fail.
You can safely ignore the #ifdef because a later patch (page_autonuma)
clears it.
Signed-off-by: Andrea Arcangeli
---
mm/page_alloc.c |4
1 files ch
Debug tweak.
Signed-off-by: Andrea Arcangeli
---
include/linux/autonuma.h | 19 +++
mm/page_alloc.c |3 ++-
2 files changed, 21 insertions(+), 1 deletions(-)
diff --git a/include/linux/autonuma.h b/include/linux/autonuma.h
index 1d87ecc..8a779e0 100644
--- a/inclu
Implement pte_numa and pmd_numa.
We must atomically set the numa bit and clear the present bit to
define a pte_numa or pmd_numa.
Once a pte or pmd has been set as pte_numa or pmd_numa, the next time
a thread touches a virtual address in the corresponding virtual range,
a NUMA hinting page fault w
On Wed, Aug 22, 2012 at 02:52:10PM +0200, Thomas Hellstrom wrote:
> Hi, Maarten,
> please see some comments inline.
>
> On 08/22/2012 01:50 PM, Maarten Lankhorst wrote:
> >Hey Dan,
> >
> >Op 16-08-12 01:12, Daniel Vetter schreef:
> >>Hi Maarten,
> >>
> >>Ok, here comes the promised review (finally
The i.MX cpufreq implementation uses the CPU_FREQ_TABLE helpers,
so it needs to select that code to be built. This problem has
apparently existed since the i.MX cpufreq code was first merged
in v3.6.37.
Building IMX without CPU_FREQ_TABLE results in:
arch/arm/plat-mxc/built-in.o: In function `mxc
The gpio leds driver can only be built if the generic LED code is
also enabled. Snowball selects LEDS_GPIO for an unknown reason
but it does not select LEDS_CLASS and also does not really require
the LED support at all, so we can just drop this dependency.
Without this patch, building without LEDS
Hi everyone,
Here are some more patches for bugs I found using
'make randconfig' in v3.6-rc+multiplatform. I can
to merge them through the arm-soc tree, but please
have a look and provide Acks.
Arnd
---
Cc: Artem Bityutskiy
Cc: Eric Miao
Cc: Kevin Hilman
Cc: Lee Jones
Cc: Linus Walle
The ksz9021rn_phy_fixup and mx6q_sabrelite functions try to
set up an ethernet phy if they can. They do check whether
phylib is enabled, but unfortunately the functions can only
be called from platform code if phylib is builtin, not
if it is a module
Without this patch, building with a modular phy
On i.MX6, we select ARM_CPU_SUSPEND when building with power management
support, but for some reason this was omitted on i.MX5. Normally we
build kernels for both together so the error only showed up in
randconfig tests.
Without this patch, building imx5 standalone results in:
arch/arm/mach-imx/b
This moves the imx5 pm code out of the list of unconditionally
compiled files for imx5, mirroring what we already do for imx6
and how it was done before the code was move from mach-mx5 to
mach-imx in v3.3.
Without this patch, building with CONFIG_PM disabled results in:
arch/arm/mach-imx/pm-imx5.
The new omap4 cpuidle implementation currently requires
ARCH_NEEDS_CPU_IDLE_COUPLED, which only works on SMP.
This patch makes it possible to build a non-SMP kernel
for that platform. This is not normally desired for
end-users but can be useful for testing.
Without this patch, building rand-0y2jS
Fix build error:
ERROR: "led_classdev_register" [drivers/media/radio/shark2.ko] undefined!
ERROR: "led_classdev_unregister" [drivers/media/radio/shark2.ko] undefined!
which is seen if RADIO_SHARK2 is enabled, but LEDS_CLASS is not.
Since RADIO_SHARK2 depends on NEW_LEDS and LEDS_CLASS, select bo
Hi,
Based on the previous discussion, in this version I propose only error
reporting fix ("overwrite recovery" is sparated out from this series.)
I think Fengguang's patch (patch 2 in this series) has a corner case
about inode cache drop, so I added patch 3 for it.
Shortlog and diffstat
Naoya
From: Wu Fengguang
This makes the EIO reports on write(), fsync(), or the NFS close()
sticky enough. The only way to get rid of it may be
echo 3 > /proc/sys/vm/drop_caches
Note that the impacted process will only be killed if it mapped the page.
XXX
via read()/write()/fsync() instead of
From: Jiang Liu
Commit 0d52f54e2ef64c189dedc332e680b2eb4a34590a (PCI / ACPI: Make acpiphp
ignore root bridges using PCIe native hotplug) added code that made the
acpiphp driver completely ignore PCIe root complexes for which the kernel
had been granted control of the native PCIe hotplug feature b
action_result() fails to print out "dirty" even if an error occurred on a
dirty pagecache, because when we check PageDirty in action_result() it was
cleared after page isolation even if it's dirty before error handling. This
can break some applications that monitor this message, so should be fixed.
"HWPOISON: report sticky EIO for poisoned file" still has a corner case
where we have possibilities of data lost. This is because in this fix
AS_HWPOISON is cleared when the inode cache is dropped.
For example, consider an application in which a process periodically
(every 10 minutes) writes some
For the sysctl work we should I think do something like this. At that point the
sysctl helpers can get/set values via the existing driver interfaces
From: Alan Cox
We want them split so that we can call them from setserial functionality
where we copy to/from user space and do the locking, but al
On Wednesday 22 August 2012, Nicolas Pitre wrote:
> On Wed, 22 Aug 2012, Arnd Bergmann wrote:
>
> > The mls instruction is not available in ARMv6K or below, so we
> > should make the test conditional on at least ARMv7. ldrex/strex
> > are available in ARMv6K or ARMv7, which we can test by checking
On 08/22/2012 04:57 PM, Hin-Tak Leung wrote:
Antti Palosaari wrote:
Hello Hiroshi,
On 08/21/2012 10:02 AM, Hiroshi Doyu wrote:
Antti Palosaari wrote @ Mon, 20 Aug 2012 23:29:34 +0200:
On 08/20/2012 02:14 PM, Hiroshi Doyu wrote:
Hi Antti,
Antti Palosaari wrote @ Sat, 18 Aug 2012 02:11:56
Hi,
I've a better fix for this here:
http://git.linuxtv.org/hgoede/gspca.git/shortlog/refs/heads/media-for_v3.6
I already send a pull-req for this to Mauro a while ago, Mauro?
Regards,
Hans
On 08/22/2012 05:16 PM, Guenter Roeck wrote:
Fix build error:
ERROR: "led_classdev_register" [drivers
- Allow xen_mapping_pagetable_reserve() to handle a start different from
pgt_buf_start, but still bigger than it.
- Add checks to xen_mapping_pagetable_reserve() and native_pagetable_reserve()
for verifying start and end are contained in the range
[pgt_buf_start, pgt_buf_top].
- In xen_mappin
When looking for documenting x86_init.mapping.pagetable_reserve, I realized
that it assumes start == pgt_buf_start. I think this is not semantically right
(even if with the current code this should not be a problem in practice) and
what we really want is to extend the logic in order to do the RO ->
The informations added on the hook are:
- Native behaviour
- Xen specific behaviour
- Logic behind the Xen specific behaviour
- setup function semantic
Even if in general it would be a good idea to separate implementation
from semantic, in this case it is beneficial to give a little bit of context
On Wed, 22 Aug 2012, Arnd Bergmann wrote:
> I've created this series some time ago, and updated it now to
> v3.6-rc1. The idea is to get us a big step closer to the
> single zImage kernel across multiple ARM platforms by
> untangling the duplicate header file names.
Thanks for picking up this tas
On Wed, 22 Aug 2012, Arnd Bergmann wrote:
> This is an attempt to simplify the duplicate header problem for
> multiplatform kernels, in multiple steps:
>
> 1. (this patch)
>for each arch/arm/mach-${MACHINE}/include/mach-${MACHINE}/*.h file,
>generate a arch/arm/include/generated/mach/*.h
On Wed, 22 Aug 2012, Arnd Bergmann wrote:
> This moves all headers to , and all
> headers to , as another step closer
> to building kernels for multiple platforms combined.
>
> A script to generate headers with the traditional names including
> the headers from the new place is already there, s
On Wed, Aug 22, 2012 at 05:22:26PM +0200, Hans de Goede wrote:
> Hi,
>
> I've a better fix for this here:
> http://git.linuxtv.org/hgoede/gspca.git/shortlog/refs/heads/media-for_v3.6
>
> I already send a pull-req for this to Mauro a while ago, Mauro?
>
Looks like it found its way into mainline i
On Wednesday 22 August 2012, Nicolas Pitre wrote:
> On Wed, 22 Aug 2012, Arnd Bergmann wrote:
> > There are two branches available in the arm-soc tree:
> >
> > 1. This series,
> >
> > http://git.kernel.org/?p=linux/kernel/git/arm/arm-soc.git;a=shortlog;h=refs/heads/testing/mach-headers
> >
On Wed, 22 Aug 2012, Arnd Bergmann wrote:
> These are lots of device drivers that include machine
> specific header files from ARM platforms and that are
> not easily scriptable. The changes have been found through
> manual inspection and should cause no visible changes
> because of the build scri
On Wed, Aug 22, 2012 at 8:43 PM, Arnd Bergmann wrote:
> The new omap4 cpuidle implementation currently requires
> ARCH_NEEDS_CPU_IDLE_COUPLED, which only works on SMP.
>
> This patch makes it possible to build a non-SMP kernel
> for that platform. This is not normally desired for
> end-users but c
On Wednesday 22 August 2012, Nicolas Pitre wrote:
> Ideally, files such as ...
>
> > rename from arch/arm/mach-at91/include/mach/at91_pio.h
> > rename to arch/arm/mach-at91/include/mach-at91/at91_pio.h
>
> should move next to the A91 GPIO driver file. but that require manual
> operations which
On Wed, 22 Aug 2012, Arnd Bergmann wrote:
> On Wednesday 22 August 2012, Nicolas Pitre wrote:
> > On Wed, 22 Aug 2012, Arnd Bergmann wrote:
> >
> > > The mls instruction is not available in ARMv6K or below, so we
> > > should make the test conditional on at least ARMv7. ldrex/strex
> > > are avai
On Wed, Aug 22, 2012 at 16:09:22, Chinmay V S wrote:
> Hmmm. Interesting. As i understand LIS331DLH provides 16bit data
> irrespective of the full-scale/sensitivity configuration. Hence we
> could effectively map +/-2G to +/-32768(signed 16bit 2's complement).
> According to the current-patch right
Hi Linus,
The following changes since commit 2d534926205db9ffce4bbbde67cb9b2cee4b835c:
Merge tag 'irqdomain-for-linus' of git://git.secretlab.ca/git/linux-2.6
(2012-07-31 20:44:03 -0700)
are available in the git repository at:
git://git.linux-nfs.org/projects/trondmy/linux-nfs.git tags/nf
Jaehoon,
On Tue, Aug 21, 2012 at 9:35 PM, Jaehoon Chung wrote:
>
> Hi Doug,
>
> I didn't know what purpose is.
> Why need to add the MMC_CAP2_NO_MMC?
> If card is SD or SDIO, mmc_attach_mmc(host) should not be entered.
> Could you explain to me in more detail?
Thanks for your feedback. In this
> -Original Message-
> From: Greg KH [mailto:gre...@linuxfoundation.org]
> Sent: Friday, August 17, 2012 12:51 PM
> To: KY Srinivasan
> Cc: linux-kernel@vger.kernel.org; de...@linuxdriverproject.org;
> virtualizat...@lists.osdl.org; o...@aepfle.de; a...@canonical.com;
> b...@decadent.org.
On Wed, Aug 22, 2012 at 12:55:25PM +0100, Lee Jones wrote:
> > I *have* asked you to communicate more clearly about what you're doing
> > but that doesn't mean to stop sending code, it means to have clearer
> > words around what you're sending.
> That's not how I interpreted your words:
> "What
>>> On 21.08.12 at 21:03, Konrad Rzeszutek Wilk wrote:
> On Tue, Aug 21, 2012 at 01:27:32PM -0400, Konrad Rzeszutek Wilk wrote:
>> On Mon, Aug 20, 2012 at 10:13:05AM -0400, Konrad Rzeszutek Wilk wrote:
>> > On Fri, Aug 17, 2012 at 06:35:12PM +0100, Stefano Stabellini wrote:
>> > > On Thu, 16 Aug 2
On 08/22, Ananth N Mavinakayanahalli wrote:
>
> +int arch_uprobe_analyze_insn(struct arch_uprobe *auprobe, struct mm_struct
> *mm, unsigned long addr)
> +{
> + unsigned int insn;
> +
> + if (addr & 0x03)
> + return -EINVAL;
> +
> + memcpy(&insn, auprobe->insn, MAX_UINSN_BYT
On Wed, Aug 22, 2012 at 03:18:45PM +0800, Dong Aisheng wrote:
> From: Dong Aisheng
>
> Using standard imx syscon API to access anatop register.
Acked-by: Mark Brown
With the conversion to regmap it'd also be good to convert the driver to
use the regmap helper functions for enable and voltage o
On 08/22/2012 12:15 AM, Eric Dumazet wrote:
This particular commit is the start of a patches batch that ended in the
generic TCP coalescing mechanism.
It is known to have problem on drivers doing skb_clone() in their rx
path.
Current kernels should be ok, because coalescing doesnt happen if th
On Aug 22, 2012, at 8:44 AM, Doug Anderson wrote:
> Jaehoon,
>
> On Tue, Aug 21, 2012 at 9:35 PM, Jaehoon Chung wrote:
>>
>> Hi Doug,
>>
>> I didn't know what purpose is.
>> Why need to add the MMC_CAP2_NO_MMC?
>> If card is SD or SDIO, mmc_attach_mmc(host) should not be entered.
>> Could yo
On Wed, Aug 22, 2012 at 03:18:42PM +0800, Dong Aisheng wrote:
> From: Dong Aisheng
> Add regmap based imx syscon driver.
Nice to see more regmap-mmio usage!
Reviwed-by: Mark Brown
from a regmap point of view.
> +int imx_syscon_write(struct device_node *np, u32 reg, u32 val)
> +{
> + str
On 08/22, Sebastian Andrzej Siewior wrote:
>
> On 08/22/2012 04:03 PM, Oleg Nesterov wrote:
>>
>> Sebastian, we have other uprobes patches in flight, I'll returns to
>> this after we push them.
>>
>> As I said, personally I mostly agree with this change... but may be
>> I'll try to convince you to
On Wed, Aug 22, 2012 at 09:15:22AM +0530, Vinod Koul wrote:
> On Tue, 2012-08-21 at 14:43 -0400, Matt Porter wrote:
> > Removes use of the DaVinci EDMA private DMA API and replaces
> > it with use of the DMA engine API.
> >
> > Signed-off-by: Matt Porter
> > ---
>
> > + struct dma_slav
On Wed, Aug 22, 2012 at 12:28 AM, Ingo Molnar wrote:
>
> * Luigi Semenzato wrote:
>
>> This fixes a long-standing bug caused by the lack of separate
>> COMM and EXEC record types, which makes "perf report" lose
>> track of symbols when a process renames itself.
>>
>> With this fix (suggested by S
If this is a must-to-do thing for Intel Panther Point platform, then
we need to make sure it's called on power up and resume. Yes, I think
moving the code below hc_init label should work and I think it's a
better solution than your original patch.
Yes it effects a lot of machines (Lenovo), I w
On Sunday, August 19, 2012 1:44 AM, Julia Lawall wrote:
> From: Julia Lawall
>
> Initialize return variable before exiting on an error path.
> Signed-off-by: Julia Lawall
>
> ---
> drivers/spi/spi-ep93xx.c |1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/drivers/spi/spi-ep93xx.c b/
On Aug 22, 2012, at 1:24 AM,
wrote:
> From: Xuelin Shi
>
> The RaidEngine is a new FSL hardware that used as hardware acceration
> for RAID5/6.
>
> This patch enables the RaidEngine functionality and provides hardware
> offloading capability for memcpy, xor and raid6 pq computation. It work
Hi Konrad,
the followings are the patches that I am proposing for Linux 3.7.
I am leaving out the bulk of the ARM patches for the moment.
Stefano Stabellini (6):
xen/events: fix unmask_evtchn for PV on HVM guests
xen: missing includes
xen: update xen_add_to_physmap interface
When unmask_evtchn is called, if we already have an event pending, we
just set evtchn_pending_sel waiting for local_irq_enable to be called.
That is because PV guests set the irq_enable pvops to
xen_irq_enable_direct in xen_setup_vcpu_info_placement:
xen_irq_enable_direct is implemented in assembly
Changes in v3:
- add missing pvclock-abi.h include to ia64 header files.
Changes in v2:
- remove pvclock hack;
- remove include linux/types.h from xen/interface/xen.h.
Signed-off-by: Stefano Stabellini
---
arch/ia64/include/asm/xen/interface.h |2 ++
arch/x86/include/asm/xen/interface.
This patch removes the "return -ENOSYS" for auto_translated_physmap
guests from privcmd_mmap, thus it allows ARM guests to issue privcmd
mmap calls. However privcmd mmap calls are still going to fail for HVM
and hybrid guests on x86 because the xen_remap_domain_mfn_range
implementation is currently
All the original Xen headers have xen_pfn_t as mfn and pfn type, however
when they have been imported in Linux, xen_pfn_t has been replaced with
unsigned long. That might work for x86 and ia64 but it does not for arm.
Bring back xen_pfn_t and let each architecture define xen_pfn_t as they
see fit.
On Wed, Aug 22, 2012 at 09:09:26AM +0530, Vinod Koul wrote:
> On Tue, 2012-08-21 at 14:43 -0400, Matt Porter wrote:
> > Add a DMA engine driver for the TI EDMA controller. This driver
> > is implemented as a wrapper around the existing DaVinci private
> > DMA implementation. This approach allows fo
Update struct xen_add_to_physmap to be in sync with Xen's version of the
structure.
The size field was introduced by:
changeset: 24164:707d27fe03e7
user:Jean Guyader
date:Fri Nov 18 13:42:08 2011 +
summary: mm: New XENMEM space, XENMAPSPACE_gmfn_range
According to the c
Reset the IRQ_NOAUTOEN and IRQ_NOREQUEST flags that are enabled by
default on ARM. If IRQ_NOAUTOEN is set, __setup_irq doesn't call
irq_startup, that is responsible for calling irq_unmask at startup time.
As a result event channels remain masked.
Signed-off-by: Stefano Stabellini
Acked-by: Konrad
In the old times, the whole idle task was considered
as an RCU quiescent state. But as RCU became more and
more successful overtime, some RCU read side critical
section have been added even in the code of some
architectures idle tasks, for tracing for example.
So nowadays, rcu_idle_enter() and rcu
In the old times, the whole idle task was considered
as an RCU quiescent state. But as RCU became more and
more successful overtime, some RCU read side critical
section have been added even in the code of some
architectures idle tasks, for tracing for example.
So nowadays, rcu_idle_enter() and rcu
In the old times, the whole idle task was considered
as an RCU quiescent state. But as RCU became more and
more successful overtime, some RCU read side critical
section have been added even in the code of some
architectures idle tasks, for tracing for example.
So nowadays, rcu_idle_enter() and rcu
In the old times, the whole idle task was considered
as an RCU quiescent state. But as RCU became more and
more successful overtime, some RCU read side critical
section have been added even in the code of some
architectures idle tasks, for tracing for example.
So nowadays, rcu_idle_enter() and rcu
In the old times, the whole idle task was considered
as an RCU quiescent state. But as RCU became more and
more successful overtime, some RCU read side critical
section have been added even in the code of some
architectures idle tasks, for tracing for example.
So nowadays, rcu_idle_enter() and rcu
Hi David,
Yes, I will be also attending LSS and would like to visit KS discussion as well.
Hope Rusty will send a schedule soon...
Regards,
Dmitry
On Wed, Aug 22, 2012 at 1:50 PM, David Howells wrote:
> Rusty Russell wrote:
>
>> > I've posted new versions of my module signing patches to my GI
In the old times, the whole idle task was considered
as an RCU quiescent state. But as RCU became more and
more successful overtime, some RCU read side critical
section have been added even in the code of some
architectures idle tasks, for tracing for example.
So nowadays, rcu_idle_enter() and rcu
So this fixes some potential RCU stalls in a bunch of architectures.
When rcu_idle_enter()/rcu_idle_exit() became a requirement, we forgot
to handle the architectures that don't support CONFIG_NO_HZ.
I guess the set should be dispatched into arch maintainer trees.
I'm sorry I haven't built tested
In the old times, the whole idle task was considered
as an RCU quiescent state. But as RCU became more and
more successful overtime, some RCU read side critical
section have been added even in the code of some
architectures idle tasks, for tracing for example.
So nowadays, rcu_idle_enter() and rcu
In the old times, the whole idle task was considered
as an RCU quiescent state. But as RCU became more and
more successful overtime, some RCU read side critical
section have been added even in the code of some
architectures idle tasks, for tracing for example.
So nowadays, rcu_idle_enter() and rcu
In the old times, the whole idle task was considered
as an RCU quiescent state. But as RCU became more and
more successful overtime, some RCU read side critical
section have been added even in the code of some
architectures idle tasks, for tracing for example.
So nowadays, rcu_idle_enter() and rcu
In the old times, the whole idle task was considered
as an RCU quiescent state. But as RCU became more and
more successful overtime, some RCU read side critical
section have been added even in the code of some
architectures idle tasks, for tracing for example.
So nowadays, rcu_idle_enter() and rcu
On Mon, Aug 20, 2012 at 9:40 PM, Cui, Dexuan wrote:
> Bjorn Helgaas wrote on 2012-08-21:
>> I am still concerned about reset_intel_82599_sfp_virtfn(). It looks
>> wrong and possibly unnecessary. It looks wrong because it sets
>> PCI_EXP_DEVCTL_BCR_FLR and blindly clears all other bits in
>> PCI
On Wed, Aug 22, 2012 at 02:03:41PM +0800, Xiao Guangrong wrote:
> On 08/21/2012 11:06 PM, Andrea Arcangeli wrote:
> > CPU0CPU1
> > oldpage[1] == 0 (both guest & host)
> > oldpage[0] = 1
> > trigger do_wp_page
>
> We always do ptep_clear_f
301 - 400 of 725 matches
Mail list logo
