Thank Thomas for your guidance
On Sun, Nov 27, 2022 at 1:05 AM Thomas Gleixner wrote:
>
> On Mon, Nov 21 2022 at 11:51, Zhouyi Zhou wrote:
> > During CPU-hotplug torture (CONFIG_NO_HZ_FULL=y), if we try to
> > offline tick_do_timer_cpu, the operation will fail because in
> > function tick_nohz_cp
On Sat, Nov 26, 2022 at 2:10 PM Thomas Weißschuh wrote:
>
> Commit 2df8220cc511 ("kbuild: build init/built-in.a just once") moved
> the usage of the define UTS_VERSION to the file version-timestamp.c.
With s/UTS_VERSION/UTS_RELEASE/,
Reviewed-by: Masahiro Yamada
>
> version-timestamp.c in tur
On Sat, Nov 26, 2022 at 2:10 PM Thomas Weißschuh wrote:
>
> Commit 7ad4bd887d27 ("powerpc/book3e: get rid of #include
> ")
> removed the usage of the define UTS_VERSION but forgot to drop the
> include.
>
> Fixes: 7ad4bd887d27 ("powerpc/book3e: get rid of #include
> ")
> Signed-off-by: Thomas We
On Sat, Nov 26, 2022 at 2:10 PM Thomas Weißschuh wrote:
>
> utsrelease.h is potentially generated on each build.
> By removing this unused include we can get rid of some spurious
> recompilations.
>
> Signed-off-by: Thomas Weißschuh
> ---
Reviewed-by: Masahiro Yamada
> drivers/base/firmware_
On Mon, Nov 21 2022 at 11:51, Zhouyi Zhou wrote:
> During CPU-hotplug torture (CONFIG_NO_HZ_FULL=y), if we try to
> offline tick_do_timer_cpu, the operation will fail because in
> function tick_nohz_cpu_down:
> ```
> if (tick_nohz_full_running && tick_do_timer_cpu == cpu)
> return -EBUSY;
> `
On Wednesday 02 November 2022 14:05:35 Christophe Leroy wrote:
> Le 01/11/2022 à 23:12, Pali Rohár a écrit :
> > On Sunday 09 October 2022 13:06:52 Pali Rohár wrote:
> >> On Monday 29 August 2022 10:54:51 Pali Rohár wrote:
> >>> On Sunday 28 August 2022 17:43:53 Christophe Leroy wrote:
> Le 28
On Wednesday 02 November 2022 00:25:03 Pali Rohár wrote:
> On Sunday 16 October 2022 16:59:53 Christophe Leroy wrote:
> > Hello,
> >
> > Le 16/10/2022 à 13:05, Pali Rohár a écrit :
> > > Hello Christophe! Do you have any other comments for this patch series?
> >
> > I'm AFK for two weeks, but as
PING?
On Tuesday 01 November 2022 23:43:48 Pali Rohár wrote:
> Michael, could you take this patch?
>
> On Sunday 09 October 2022 14:05:06 Pali Rohár wrote:
> > On Friday 30 September 2022 14:46:18 Pali Rohár wrote:
> > > + CC hwmon ML
> > >
> > > On Friday 30 September 2022 14:39:01 Pali Rohár w
PING?
On Tuesday 01 November 2022 23:26:03 Pali Rohár wrote:
> Hello! Gentle reminder...
>
> On Sunday 09 October 2022 13:25:55 Pali Rohár wrote:
> > Hello! Any comments on this? It would be nice to take these two patches
> > (or at least patch 2) to finally enable PPC_PCI_BUS_NUM_DOMAIN_DEPENDEN
PING?
On Tuesday 01 November 2022 23:19:09 Pali Rohár wrote:
> Hello! I would like to remind this patch.
>
> On Sunday 09 October 2022 13:08:08 Pali Rohár wrote:
> > Hello! Any opinion on this patch?
> >
> > On Saturday 20 August 2022 14:33:27 Pali Rohár wrote:
> > > If there's no PCI host bridg
On Tue, Nov 22, 2022 at 09:16:54PM +0100, Uwe Kleine-König wrote:
> On Tue, Nov 22, 2022 at 06:58:18PM +, Jonathan Cameron wrote:
> > Queued all of the below:
> > with one tweaked as per your suggestion and the highlighted one dropped on
> > basis
> > I was already carrying the equivalent - a
Le 26/11/2022 à 15:07, Thomas Weißschuh a écrit :
> On 2022-11-26 07:36+, Christophe Leroy wrote:
>> Le 26/11/2022 à 06:10, Thomas Weißschuh a écrit :
>>> Commit 7ad4bd887d27 ("powerpc/book3e: get rid of #include
>>> ")
>>> removed the usage of the define UTS_VERSION but forgot to drop the
>
On 2022-11-26 07:36+, Christophe Leroy wrote:
> Le 26/11/2022 à 06:10, Thomas Weißschuh a écrit :
>> Commit 7ad4bd887d27 ("powerpc/book3e: get rid of #include
>> ")
>> removed the usage of the define UTS_VERSION but forgot to drop the
>> include.
>
> What about:
> arch/powerpc/platforms/52xx/
Am 26.11.22 um 01:04 schrieb Randy Dunlap:
Make the offb (Open Firmware frame buffer) driver tristate,
i.e., so that it can be built as a loadable module.
However, it still depends on the setting of DRM_OFDRM
so that both of these drivers cannot be builtin at the same time
nor can one be built
This adds compile-time options that allow the EH lock hint bit to be
enabled or disabled, and adds some new options that may or may not
help matters. To help with experimentation and tuning.
Signed-off-by: Nicholas Piggin
---
arch/powerpc/include/asm/qspinlock.h | 61 ++--
Finding the owner or a queued waiter on a lock with a preempted vcpu
is indicative of an oversubscribed guest causing the lock to get into
trouble. Provide some options to detect this situation and have new
CPUs avoid queueing for a longer time (more steal iterations) to
minimise the problems cause
Provide an option that holds off queueing indefinitely while the lock
owner is preempted. This could reduce queueing latencies for very
overcommitted vcpu situations.
This is disabled by default.
Signed-off-by: Nicholas Piggin
---
arch/powerpc/lib/qspinlock.c | 77 +-
Allow for a reduction in the number of times a CPU from a different
node than the owner can attempt to steal the lock before queueing.
This could bias the transfer behaviour of the lock across the
machine and reduce NUMA crossings.
Signed-off-by: Nicholas Piggin
---
arch/powerpc/lib/qspinlock.c
Use the spin_begin/spin_cpu_relax/spin_end APIs in qspinlock, which helps
to prevent threads issuing a lot of expensive priority nops which may not
have much effect due to immediately executing low then medium priority.
Signed-off-by: Nicholas Piggin
---
arch/powerpc/lib/qspinlock.c | 39 +++
This change allows trylock to steal the lock. It also allows the initial
lock attempt to steal the lock rather than bailing out and going to the
slow path.
This gives trylock more strength: without this a continually-contended
lock will never permit a trylock to succeed. With this change, the
tryl
After the head of the queue acquires the lock, it releases the
next waiter in the queue to become the new head. Add an option
to prod the new head if its vCPU was preempted. This may only
have an effect if queue waiters are yielding.
Disable this option by default for now, i.e., no logical change.
Having all CPUs poll the lock word for the owner CPU that should be
yielded to defeats most of the purpose of using MCS queueing for
scalability. Yet it may be desirable for queued waiters to to yield
to a preempted owner.
With this change, queue waiters never sample the owner CPU directly
from th
If the head of queue is preventing stealing but it finds the owner vCPU
is preempted, it will yield its cycles to the owner which could cause it
to become preempted. Add an option to re-allow stealers before yielding,
and disallow them again after returning from the yield.
Disable this option by d
Queued waiters which are not at the head of the queue don't spin on
the lock word but their qnode lock word, waiting for the previous queued
CPU to release them. Add an option which allows these waiters to yield
to the previous CPU if its vCPU is preempted.
Signed-off-by: Nicholas Piggin
---
arc
Waiters spinning on the lock word should yield to the lock owner if the
vCPU is preempted. This improves performance when the hypervisor has
oversubscribed physical CPUs.
Signed-off-by: Nicholas Piggin
---
arch/powerpc/lib/qspinlock.c | 99 +++-
1 file changed, 87
Store the owner CPU number in the lock word so it may be yielded to,
as powerpc's paravirtualised simple spinlocks do.
Signed-off-by: Nicholas Piggin
---
arch/powerpc/include/asm/qspinlock.h | 12 ++--
arch/powerpc/include/asm/qspinlock_types.h | 12 +++-
arch/powerpc/lib/q
Give the queue head the ability to stop stealers. After a number of
spins without sucessfully acquiring the lock, the queue head sets
this, which halts stealing and will assure it is the next owner.
Signed-off-by: Nicholas Piggin
---
arch/powerpc/include/asm/qspinlock_types.h | 8 +++-
arch/pow
Allow new waiters to "steal" the lock before queueing. That is, to
acquire it while other CPUs have queued.
This particularly helps paravirt performance when physical CPUs are
oversubscribed, by keeping the lock from becoming a strict FIFO and
vCPU preemption causing queue train wrecks.
The new _
This uses more optimal ll/sc style access patterns (rather than
cmpxchg), and also sets the EH=1 lock hint on those operations
which acquire ownership of the lock.
Signed-off-by: Nicholas Piggin
---
arch/powerpc/include/asm/qspinlock.h | 24 +--
arch/powerpc/include/asm/qspinlock_types
The first 16 bits of the lock are only modified by the owner, and other
modifications always use atomic operations on the entire 32 bits, so
unlocks can use plain stores on the 16 bits. This is the same kind of
optimisation done by core qspinlock code.
Signed-off-by: Nicholas Piggin
---
arch/pow
This forms the basis of the qspinlock slow path.
Like generic qspinlocks and unlike the vanilla MCS algorithm, the lock
owner does not participate in the queue, only waiters. The first waiter
spins on the lock word, then when the lock is released it takes
ownership and unqueues the next waiter. Th
This replaces the generic queued spinlock code (like s390 does) with
our own implementation. There is an extra shim patch 1a to get the
series to apply.
Generic PV qspinlock code is causing latency / starvation regressions on
large systems that are resulting in hard lockups reported (mostly in
pat
32 matches
Mail list logo
