Since kvmppc_hv_find_lock_hpte() is called from both virtmode and
realmode, so it can trigger the deadlock.
Suppose the following scene:
Two physical cpuM, cpuN, two VM instances A, B, each VM has a group of vcpus.
If on cpuM, vcpu_A_1 holds bitlock X (HPTE_V_HVLOCK), then is switched out,
and o
v2->v3:
introduce kvmppc_hv_unlock_hpte() to pair with kvmppc_hv_find_lock_hpte()
and hide the preemption detail inside this pair from the callers
Liu Ping Fan (2):
powerpc: kvm: pair kvmppc_hv_find_lock_hpte with _unlock_hpte
powerpc: kvm: fix rare but potential deadlock scene
arch/powe
Highlight the lock pair for the reader. (and later it will the
place to hide the detail about preemption disable)
Signed-off-by: Liu Ping Fan
---
arch/powerpc/include/asm/kvm_book3s.h | 1 +
arch/powerpc/kvm/book3s_64_mmu_hv.c | 7 ++-
arch/powerpc/kvm/book3s_hv_rm_mmu.c | 13 +
On Fri, Nov 8, 2013 at 12:11 PM, Benjamin Herrenschmidt
wrote:
> On Fri, 2013-11-08 at 15:05 +1100, Benjamin Herrenschmidt wrote:
>> On Fri, 2013-11-08 at 04:10 +0100, Alexander Graf wrote:
>> > On 08.11.2013, at 03:44, Liu Ping Fan wrote:
>> >
>> > > syscall is a very common behavior inside gues
On Fri, Nov 8, 2013 at 11:10 AM, Alexander Graf wrote:
>
> On 08.11.2013, at 03:44, Liu Ping Fan wrote:
>
>> syscall is a very common behavior inside guest, and this patch
>> optimizes the path for the emulation of BOOK3S_INTERRUPT_SYSCALL,
>> so hypervisor can return to guest without heavy exit,
* Peter Zijlstra (pet...@infradead.org) wrote:
[...]
Hi Peter,
Looking at this simplified version of perf's ring buffer
synchronization, I get concerned about the following issue:
> /*
> * One important detail is that the kbuf part and the kbuf_writer() are
> * strictly per cpu and we can thu
On Fri, 2013-11-08 at 15:05 +1100, Benjamin Herrenschmidt wrote:
> On Fri, 2013-11-08 at 04:10 +0100, Alexander Graf wrote:
> > On 08.11.2013, at 03:44, Liu Ping Fan wrote:
> >
> > > syscall is a very common behavior inside guest, and this patch
> > > optimizes the path for the emulation of BOOK3
On Fri, 2013-11-08 at 04:10 +0100, Alexander Graf wrote:
> On 08.11.2013, at 03:44, Liu Ping Fan wrote:
>
> > syscall is a very common behavior inside guest, and this patch
> > optimizes the path for the emulation of BOOK3S_INTERRUPT_SYSCALL,
> > so hypervisor can return to guest without heavy ex
On 08.11.2013, at 03:44, Liu Ping Fan wrote:
> syscall is a very common behavior inside guest, and this patch
> optimizes the path for the emulation of BOOK3S_INTERRUPT_SYSCALL,
> so hypervisor can return to guest without heavy exit, i.e, no need
> to swap TLB, HTAB,.. etc
The syscall exit you
On Mon, Nov 4, 2013 at 6:31 PM, Hongbo Zhang wrote:
> Hi Vinod Koul and Dan Williams,
> Ping?
>
Not much to review from the dmaengine side, just one question below.
It would be helpful if you can send these to the new dmaengine
patchwork at dmaeng...@vger.kernel.org with the Acks you have already
syscall is a very common behavior inside guest, and this patch
optimizes the path for the emulation of BOOK3S_INTERRUPT_SYSCALL,
so hypervisor can return to guest without heavy exit, i.e, no need
to swap TLB, HTAB,.. etc
Signed-off-by: Liu Ping Fan
---
Compiled, but lack of bare metal, I have not
On Thu, Nov 7, 2013 at 6:06 PM, Alexander Graf wrote:
>
> On 07.11.2013, at 07:22, Liu Ping Fan wrote:
>
>> ret is assigned twice with the same value, so remove the later one.
>>
>> Signed-off-by: Liu Ping Fan
>> Acked-by: Paul Mackerras
>
> I suppose my last request for a patch description was
On Thu, Nov 7, 2013 at 5:54 PM, Alexander Graf wrote:
>
> On 07.11.2013, at 07:22, Liu Ping Fan wrote:
>
>> Since kvmppc_hv_find_lock_hpte() is called from both virtmode and
>> realmode, so it can trigger the deadlock.
>>
>> Suppose the following scene:
>>
>> Two physical cpuM, cpuN, two VM insta
On Thu, Nov 07, 2013 at 11:34:51AM -0600, Scott Wood wrote:
> On Thu, 2013-11-07 at 15:17 +0800, Kevin Hao wrote:
> > It makes no sense to initialize the mpic ipi for the SoC which has
> > doorbell support. So set the smp_85xx_ops.probe to NULL for this
> > case. Since the smp_85xx_ops.probe is als
On Thu, Nov 07, 2013 at 03:01:02AM +, Li Xiubo wrote:
> > > The SGTL5000 is based on regulators and when it is disabled, there
> > > will be an error returns directly while the SGTL5000 codec is probing.
> > What makes you say this?
> static int ldo_regulator_register(struct snd_soc_codec *
On Thu, 2013-11-07 at 15:17 +0800, Kevin Hao wrote:
> It makes no sense to initialize the mpic ipi for the SoC which has
> doorbell support. So set the smp_85xx_ops.probe to NULL for this
> case. Since the smp_85xx_ops.probe is also used in function
> smp_85xx_setup_cpu() to check if we need to inv
On Thu, 7 Nov 2013, Alistair Popple wrote:
> Thanks. Based on the discussion for the EHCI driver I would like to change
> the
> compatibility string to "usb-ochi" (instead of "ibm,akebono-ohci"). Are you
> still happy for me to add the Acked-by with the alternate compatibility (and
> of course
Hello Becky Bruce,
This is a semi-automatic email about new static checker warnings.
The patch 41151e77a4d9: "powerpc: Hugetlb for BookE" from Jun 28,
2011, leads to the following Smatch complaint:
arch/powerpc/mm/hugetlbpage-book3e.c:120 flush_hugetlb_page()
warn: variable dereference
On Thu, Nov 07, 2013 at 11:17:41AM +, Will Deacon wrote:
> Hi Peter,
>
> Couple of minor fixes on the arm64 side...
>
> On Wed, Nov 06, 2013 at 01:57:36PM +, Peter Zijlstra wrote:
> > --- a/arch/arm64/include/asm/barrier.h
> > +++ b/arch/arm64/include/asm/barrier.h
> > @@ -35,11 +35,59 @@
Unaligned stores take alignment exceptions on POWER7 running in little-endian.
This is a dumb little-endian base memcpy that prevents unaligned stores.
Once booted the feature fixup code switches over to the VMX copy loops
(which are already endian safe).
The question is what we do before that sw
2013/11/1 Arnd Bergmann :
> On Friday 01 November 2013, Alexander Popov wrote:
>> + * of_dma_xlate_by_chan_id - Translate dt property to DMA channel by
>> channel id
>> + * @dma_spec:pointer to DMA specifier as found in the device tree
>> + * @of_dma: pointer to DMA controller data
>
> Th
Hi Peter,
Couple of minor fixes on the arm64 side...
On Wed, Nov 06, 2013 at 01:57:36PM +, Peter Zijlstra wrote:
> --- a/arch/arm64/include/asm/barrier.h
> +++ b/arch/arm64/include/asm/barrier.h
> @@ -35,11 +35,59 @@
> #define smp_mb() barrier()
> #define smp_rmb() barrier()
> #
On 07.11.2013, at 07:22, Liu Ping Fan wrote:
> ret is assigned twice with the same value, so remove the later one.
>
> Signed-off-by: Liu Ping Fan
> Acked-by: Paul Mackerras
I suppose my last request for a patch description was slightly too abbreviated
:). Sorry about that.
Imagine you are
On 07.11.2013, at 07:22, Liu Ping Fan wrote:
> Since kvmppc_hv_find_lock_hpte() is called from both virtmode and
> realmode, so it can trigger the deadlock.
>
> Suppose the following scene:
>
> Two physical cpuM, cpuN, two VM instances A, B, each VM has a group of vcpus.
>
> If on cpuM, vcpu_
On Thu, 2013-11-07 at 09:14 +0100, Alexander Graf wrote:
> > And ? An explanation isn't going to be clearer than the code in that
> > case ...
>
> It's pretty non-obvious when you do a git show on that patch in 1 year
> from now, as the redundancy is out of scope of what the diff shows.
And ? How
Am 07.11.2013 um 08:55 schrieb Benjamin Herrenschmidt
:
> On Thu, 2013-11-07 at 08:52 +0100, Alexander Graf wrote:
>> Am 06.11.2013 um 20:58 schrieb Benjamin Herrenschmidt
>> :
>>
>>> On Wed, 2013-11-06 at 12:24 +0100, Alexander Graf wrote:
On 05.11.2013, at 08:42, Liu Ping Fan wrote:
>
26 matches
Mail list logo
