On Tue, 2 Jun 2015 16:42:31 +0100
Peter Maydell wrote:
> On 2 June 2015 at 16:32, Igor Mammedov wrote:
> > On Fri, 29 May 2015 16:41:01 +0300
> > Pavel Fedin wrote:
> >
> >> Hello!
> >>
> >> > Well KVM side should be fixed instead of driving us along wrong route.
> >>
> >> It can be fixed...
On Sun, 31 May 2015 11:03:53 +
Shlomo Pongratz wrote:
> Hi,
>
> See below.
>
> > -Original Message-
> > From: Pavel Fedin [mailto:p.fe...@samsung.com]
> > Sent: Friday, 29 May, 2015 9:45 AM
> > To: Shlomo Pongratz; qemu-devel@nongnu.org
> > Cc: peter.mayd...@linaro.org; 'Ashok Kumar
On 2 June 2015 at 16:32, Igor Mammedov wrote:
> On Fri, 29 May 2015 16:41:01 +0300
> Pavel Fedin wrote:
>
>> Hello!
>>
>> > Well KVM side should be fixed instead of driving us along wrong route.
>>
>> It can be fixed... Perhaps... If kernel developers acknowledge this is a
>> problem, which
>>
On Fri, 29 May 2015 16:41:01 +0300
Pavel Fedin wrote:
> Hello!
>
> > Well KVM side should be fixed instead of driving us along wrong route.
>
> It can be fixed... Perhaps... If kernel developers acknowledge this is a
> problem, which
> might not happen.
> But still we will have older kernel
Hi!
> You mean that I put your patch
https://lists.gnu.org/archive/html/qemu-devel/201505/msg04495.html
Yes. You can of course further modify it, like changing #define to a variable
which can
be set from outside.
> The first patch of my V3 patch series, with the sole exception of fixing the
Hi
> -Original Message-
> From: Pavel Fedin [mailto:p.fe...@samsung.com]
> Sent: Monday, 01 June, 2015 12:04 PM
> To: Shlomo Pongratz; qemu-devel@nongnu.org
> Cc: peter.mayd...@linaro.org; 'Ashok Kumar'
> Subject: RE: [Qemu-devel] [PATCH] Use Aff1 with mp
Hi again!
> I intend to release the V3 version during this week.
One more note. I remember that i suggested to merge ITS_CONTROL and
ITS_TRANSLATION into
one region in virt memory map. I was wrong, don't do this and leave them as
they are.
There is in-kernel ITS implementation suggested on AR
Hi!
> I think we should take Igor's comment into account. The CPUS_PER_CLUSTER
> should not be a
> constant, and maybe should be initialized in arm_cpu_initfn and
aarch64_{a53|a57}_initfn, as
> psci need to know it.
Yes, of course. Just replace a #define with something like:
int cpus_per_clu
Hi,
See below.
> -Original Message-
> From: Pavel Fedin [mailto:p.fe...@samsung.com]
> Sent: Friday, 29 May, 2015 9:45 AM
> To: Shlomo Pongratz; qemu-devel@nongnu.org
> Cc: peter.mayd...@linaro.org; 'Ashok Kumar'
> Subject: RE: [PATCH] Use Aff1 with mpidr
>
> Hi!
>
> > I see that you a
On 2015/5/30 1:37, Pavel Fedin wrote:
> Hi!
>
>> Well KVM side should be fixed instead of driving us along wrong route.
>
> I have studied the question a bit more, and i discovered that MPIDR access
> on ARM is not
> trapped by KVM. And guest would always get the same value as host would.
>
Hi!
> Well KVM side should be fixed instead of driving us along wrong route.
I have studied the question a bit more, and i discovered that MPIDR access on
ARM is not
trapped by KVM. And guest would always get the same value as host would.
Theoretically you
could modify the kernel so that PSCI
Hello!
> Well KVM side should be fixed instead of driving us along wrong route.
It can be fixed... Perhaps... If kernel developers acknowledge this is a
problem, which
might not happen.
But still we will have older kernels, and what? Don't you want 64-bit ARM KVM
work on
them?
Kind regards,
On Fri, 29 May 2015 15:26:57 +0300
Pavel Fedin wrote:
> Hello!
>
> > That's what I see as correct way to go
> > and along the way teach KVM to not derive mpidr encoding from vcpuid
> > and use QEMU provided mpidr value, that way mpidr would be consistent.
>
> Yes, and this would make modellin
Hello!
> That's what I see as correct way to go
> and along the way teach KVM to not derive mpidr encoding from vcpuid
> and use QEMU provided mpidr value, that way mpidr would be consistent.
Yes, and this would make modelling of real hardware working much better.
However,
unfortunately, i hav
On Fri, 29 May 2015 11:53:38 +0300
Pavel Fedin wrote:
> Hello!
>
> > > +/*
> > > + * When KVM is in use, psci is emulated in-kernel and not by qemu.
> > > + * In order for it to work correctly we should use correct MPIDR
> > > values,
> > > + * which appear to be inherited from
Hello!
> > +/*
> > + * When KVM is in use, psci is emulated in-kernel and not by qemu.
> > + * In order for it to work correctly we should use correct MPIDR
> > values,
> > + * which appear to be inherited from the host.
> why it must be inherited from host?
> Could you point out
On Fri, 22 May 2015 13:32:54 +0300
Pavel Fedin wrote:
> This is an improved and KVM-aware alternative to
> https://lists.gnu.org/archive/html/qemu-devel/2015-05/msg00942.html. Changes
> are:
> 1. MPIDR value (minus feature bits) is now directly stored in CPU instance.
> 2. Default assignment is
Hi!
> I see that you added mpidr to ARMCpu and initialized it in virt.c then you
> use it in
mpidr_read.
> Thus you must fix all other virtual machines in hw/arm not just virt.c as it
> is not
initialized for
> them.
The only change in virt.c is:
--- cut ---
diff --git a/hw/arm/virt.c b/hw/ar
Hi,
I see that you added mpidr to ARMCpu and initialized it in virt.c then you use
it in mpidr_read.
Thus you must fix all other virtual machines in hw/arm not just virt.c as it is
not initialized for them.
I have no other comments.
Best regards,
S.P.
> -Original Message-
> From: Pa
This is an improved and KVM-aware alternative to
https://lists.gnu.org/archive/html/qemu-devel/2015-05/msg00942.html. Changes
are:
1. MPIDR value (minus feature bits) is now directly stored in CPU instance.
2. Default assignment is based on original rule (8 CPUs per cluster).
3. If KVM is used, M
20 matches
Mail list logo
