linux-sh.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH 00/02] ARM: shmobile: r8a7790 8-core boot prototype
@ 2014-04-08 12:47 Magnus Damm
  2014-04-09  0:24 ` Simon Horman
                   ` (4 more replies)
  0 siblings, 5 replies; 6+ messages in thread
From: Magnus Damm @ 2014-04-08 12:47 UTC (permalink / raw)
  To: linux-sh

ARM: shmobile: r8a7790 8-core boot prototype

[PATCH 01/02] ARM: shmobile: Multicluster APMU prototype support
[PATCH 02/02] ARM: shmobile: r8a7790 CCI configuration

Prototype patches to allow 8-core boot on r8a7790. The first patch
unlocks multicluster support to allo SMP boot of all CPU cores, and
the second initializes the CCI-400 hardware.

Not for upstream merge.

Signed-off-by: Magnus Damm <damm@opensource.se>
---

 Applies on top of renesas-devel-v3.14-20140408

 arch/arm/mach-shmobile/platsmp-apmu.c |   16 ++++++++++++++--
 arch/arm/mach-shmobile/smp-r8a7790.c  |   15 +++++++++++++++
 2 files changed, 29 insertions(+), 2 deletions(-)

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH 00/02] ARM: shmobile: r8a7790 8-core boot prototype
  2014-04-08 12:47 [PATCH 00/02] ARM: shmobile: r8a7790 8-core boot prototype Magnus Damm
@ 2014-04-09  0:24 ` Simon Horman
  2014-04-09  4:11 ` Magnus Damm
                   ` (3 subsequent siblings)
  4 siblings, 0 replies; 6+ messages in thread
From: Simon Horman @ 2014-04-09  0:24 UTC (permalink / raw)
  To: linux-sh

On Tue, Apr 08, 2014 at 09:47:58PM +0900, Magnus Damm wrote:
> ARM: shmobile: r8a7790 8-core boot prototype
> 
> [PATCH 01/02] ARM: shmobile: Multicluster APMU prototype support
> [PATCH 02/02] ARM: shmobile: r8a7790 CCI configuration
> 
> Prototype patches to allow 8-core boot on r8a7790. The first patch
> unlocks multicluster support to allo SMP boot of all CPU cores, and
> the second initializes the CCI-400 hardware.

Hi Magnus,

I applied these patches on top of renesas-devel-v3.14-20140408,
compiled using the shmobile_defconfig, and booted a Lager board.
I was expecting to see 8 cores in /proc/cpuinfo but I only see 4.

Are my expectations wrong or am I missing something?

> Not for upstream merge.
> 
> Signed-off-by: Magnus Damm <damm@opensource.se>
> ---
> 
>  Applies on top of renesas-devel-v3.14-20140408
> 
>  arch/arm/mach-shmobile/platsmp-apmu.c |   16 ++++++++++++++--
>  arch/arm/mach-shmobile/smp-r8a7790.c  |   15 +++++++++++++++
>  2 files changed, 29 insertions(+), 2 deletions(-)
> 

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH 00/02] ARM: shmobile: r8a7790 8-core boot prototype
  2014-04-08 12:47 [PATCH 00/02] ARM: shmobile: r8a7790 8-core boot prototype Magnus Damm
  2014-04-09  0:24 ` Simon Horman
@ 2014-04-09  4:11 ` Magnus Damm
  2014-04-09  5:05 ` Simon Horman
                   ` (2 subsequent siblings)
  4 siblings, 0 replies; 6+ messages in thread
From: Magnus Damm @ 2014-04-09  4:11 UTC (permalink / raw)
  To: linux-sh

Hi Simon,

On Wed, Apr 9, 2014 at 9:24 AM, Simon Horman <horms@verge.net.au> wrote:
> On Tue, Apr 08, 2014 at 09:47:58PM +0900, Magnus Damm wrote:
>> ARM: shmobile: r8a7790 8-core boot prototype
>>
>> [PATCH 01/02] ARM: shmobile: Multicluster APMU prototype support
>> [PATCH 02/02] ARM: shmobile: r8a7790 CCI configuration
>>
>> Prototype patches to allow 8-core boot on r8a7790. The first patch
>> unlocks multicluster support to allo SMP boot of all CPU cores, and
>> the second initializes the CCI-400 hardware.
>
> Hi Magnus,
>
> I applied these patches on top of renesas-devel-v3.14-20140408,
> compiled using the shmobile_defconfig, and booted a Lager board.
> I was expecting to see 8 cores in /proc/cpuinfo but I only see 4.
>
> Are my expectations wrong or am I missing something?

The maximum number of CPUs in the kernel config may be limited to 4 by default?

Also, to use CA7 with upstream you need to disable the arch timer and
rely on CMT.

To clarify, I don't expect these to be merged upstream. Also, I don't
believe the standard kernel config needs to work out of the box.

Cheers,

/ magnus

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH 00/02] ARM: shmobile: r8a7790 8-core boot prototype
  2014-04-08 12:47 [PATCH 00/02] ARM: shmobile: r8a7790 8-core boot prototype Magnus Damm
  2014-04-09  0:24 ` Simon Horman
  2014-04-09  4:11 ` Magnus Damm
@ 2014-04-09  5:05 ` Simon Horman
  2014-04-09 23:57 ` Magnus Damm
  2014-04-10  5:48 ` Simon Horman
  4 siblings, 0 replies; 6+ messages in thread
From: Simon Horman @ 2014-04-09  5:05 UTC (permalink / raw)
  To: linux-sh

On Wed, Apr 09, 2014 at 01:11:43PM +0900, Magnus Damm wrote:
> Hi Simon,
> 
> On Wed, Apr 9, 2014 at 9:24 AM, Simon Horman <horms@verge.net.au> wrote:
> > On Tue, Apr 08, 2014 at 09:47:58PM +0900, Magnus Damm wrote:
> >> ARM: shmobile: r8a7790 8-core boot prototype
> >>
> >> [PATCH 01/02] ARM: shmobile: Multicluster APMU prototype support
> >> [PATCH 02/02] ARM: shmobile: r8a7790 CCI configuration
> >>
> >> Prototype patches to allow 8-core boot on r8a7790. The first patch
> >> unlocks multicluster support to allo SMP boot of all CPU cores, and
> >> the second initializes the CCI-400 hardware.
> >
> > Hi Magnus,
> >
> > I applied these patches on top of renesas-devel-v3.14-20140408,
> > compiled using the shmobile_defconfig, and booted a Lager board.
> > I was expecting to see 8 cores in /proc/cpuinfo but I only see 4.
> >
> > Are my expectations wrong or am I missing something?
> 
> The maximum number of CPUs in the kernel config may be limited to 4 by default?

The shmobile_defconfig yeilds CONFIG_NR_CPUS=8, which seems fine.

> Also, to use CA7 with upstream you need to disable the arch timer and
> rely on CMT.

Thanks I have disabled CONFIG_HAVE_ARM_ARCH_TIMER.

> To clarify, I don't expect these to be merged upstream. Also, I don't
> believe the standard kernel config needs to work out of the box.

That is fine. But even with the above tweak I still don't see 8 CPUs
so I do wonder what the trick to seeing them is.

...
CPU: Testing write buffer coherency: ok
CPU0: update cpu_power 1535
CPU0: thread -1, cpu 0, socket 0, mpidr 80000000
Setting up static identity map for 0x404141b8 - 0x40414210
CPU1: Booted secondary processor
CPU1: update cpu_power 1535
CPU1: thread -1, cpu 1, socket 0, mpidr 80000001
CPU2: Booted secondary processor
CPU2: update cpu_power 1535
CPU2: thread -1, cpu 2, socket 0, mpidr 80000002
CPU3: Booted secondary processor
CPU3: update cpu_power 1535
CPU3: thread -1, cpu 3, socket 0, mpidr 80000003
CPU4: failed to boot: -22
CPU5: failed to boot: -22
CPU6: failed to boot: -22
CPU7: failed to boot: -22
Brought up 4 CPUs
SMP: Total of 4 processors activated.
CPU: All CPU(s) started in SVC mode.
...

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH 00/02] ARM: shmobile: r8a7790 8-core boot prototype
  2014-04-08 12:47 [PATCH 00/02] ARM: shmobile: r8a7790 8-core boot prototype Magnus Damm
                   ` (2 preceding siblings ...)
  2014-04-09  5:05 ` Simon Horman
@ 2014-04-09 23:57 ` Magnus Damm
  2014-04-10  5:48 ` Simon Horman
  4 siblings, 0 replies; 6+ messages in thread
From: Magnus Damm @ 2014-04-09 23:57 UTC (permalink / raw)
  To: linux-sh

Hi Simon,

On Wed, Apr 9, 2014 at 2:05 PM, Simon Horman <horms@verge.net.au> wrote:
> On Wed, Apr 09, 2014 at 01:11:43PM +0900, Magnus Damm wrote:
>> Hi Simon,
>>
>> On Wed, Apr 9, 2014 at 9:24 AM, Simon Horman <horms@verge.net.au> wrote:
>> > On Tue, Apr 08, 2014 at 09:47:58PM +0900, Magnus Damm wrote:
>> >> ARM: shmobile: r8a7790 8-core boot prototype
>> >>
>> >> [PATCH 01/02] ARM: shmobile: Multicluster APMU prototype support
>> >> [PATCH 02/02] ARM: shmobile: r8a7790 CCI configuration
>> >>
>> >> Prototype patches to allow 8-core boot on r8a7790. The first patch
>> >> unlocks multicluster support to allo SMP boot of all CPU cores, and
>> >> the second initializes the CCI-400 hardware.
>> >
>> > Hi Magnus,
>> >
>> > I applied these patches on top of renesas-devel-v3.14-20140408,
>> > compiled using the shmobile_defconfig, and booted a Lager board.
>> > I was expecting to see 8 cores in /proc/cpuinfo but I only see 4.
>> >
>> > Are my expectations wrong or am I missing something?
>>
>> The maximum number of CPUs in the kernel config may be limited to 4 by default?
>
> The shmobile_defconfig yeilds CONFIG_NR_CPUS=8, which seems fine.
>
>> Also, to use CA7 with upstream you need to disable the arch timer and
>> rely on CMT.
>
> Thanks I have disabled CONFIG_HAVE_ARM_ARCH_TIMER.
>
>> To clarify, I don't expect these to be merged upstream. Also, I don't
>> believe the standard kernel config needs to work out of the box.
>
> That is fine. But even with the above tweak I still don't see 8 CPUs
> so I do wonder what the trick to seeing them is.
>
> ...
> CPU: Testing write buffer coherency: ok
> CPU0: update cpu_power 1535
> CPU0: thread -1, cpu 0, socket 0, mpidr 80000000
> Setting up static identity map for 0x404141b8 - 0x40414210
> CPU1: Booted secondary processor
> CPU1: update cpu_power 1535
> CPU1: thread -1, cpu 1, socket 0, mpidr 80000001
> CPU2: Booted secondary processor
> CPU2: update cpu_power 1535
> CPU2: thread -1, cpu 2, socket 0, mpidr 80000002
> CPU3: Booted secondary processor
> CPU3: update cpu_power 1535
> CPU3: thread -1, cpu 3, socket 0, mpidr 80000003
> CPU4: failed to boot: -22
> CPU5: failed to boot: -22
> CPU6: failed to boot: -22
> CPU7: failed to boot: -22
> Brought up 4 CPUs
> SMP: Total of 4 processors activated.
> CPU: All CPU(s) started in SVC mode.
> ...


Thanks for providing more information. I wonder, perhaps I forgot to
tell you to also pass "apmu=multicluster" on the kernel command line
to enable all cpus?

The unlocking of multiplatform support is implemented in:
[PATCH 01/02] ARM: shmobile: Multicluster APMU prototype support

Cheers,

/ magnus

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH 00/02] ARM: shmobile: r8a7790 8-core boot prototype
  2014-04-08 12:47 [PATCH 00/02] ARM: shmobile: r8a7790 8-core boot prototype Magnus Damm
                   ` (3 preceding siblings ...)
  2014-04-09 23:57 ` Magnus Damm
@ 2014-04-10  5:48 ` Simon Horman
  4 siblings, 0 replies; 6+ messages in thread
From: Simon Horman @ 2014-04-10  5:48 UTC (permalink / raw)
  To: linux-sh

On Thu, Apr 10, 2014 at 08:57:20AM +0900, Magnus Damm wrote:
> Hi Simon,
> 
> On Wed, Apr 9, 2014 at 2:05 PM, Simon Horman <horms@verge.net.au> wrote:
> > On Wed, Apr 09, 2014 at 01:11:43PM +0900, Magnus Damm wrote:
> >> Hi Simon,
> >>
> >> On Wed, Apr 9, 2014 at 9:24 AM, Simon Horman <horms@verge.net.au> wrote:
> >> > On Tue, Apr 08, 2014 at 09:47:58PM +0900, Magnus Damm wrote:
> >> >> ARM: shmobile: r8a7790 8-core boot prototype
> >> >>
> >> >> [PATCH 01/02] ARM: shmobile: Multicluster APMU prototype support
> >> >> [PATCH 02/02] ARM: shmobile: r8a7790 CCI configuration
> >> >>
> >> >> Prototype patches to allow 8-core boot on r8a7790. The first patch
> >> >> unlocks multicluster support to allo SMP boot of all CPU cores, and
> >> >> the second initializes the CCI-400 hardware.
> >> >
> >> > Hi Magnus,
> >> >
> >> > I applied these patches on top of renesas-devel-v3.14-20140408,
> >> > compiled using the shmobile_defconfig, and booted a Lager board.
> >> > I was expecting to see 8 cores in /proc/cpuinfo but I only see 4.
> >> >
> >> > Are my expectations wrong or am I missing something?
> >>
> >> The maximum number of CPUs in the kernel config may be limited to 4 by default?
> >
> > The shmobile_defconfig yeilds CONFIG_NR_CPUS=8, which seems fine.
> >
> >> Also, to use CA7 with upstream you need to disable the arch timer and
> >> rely on CMT.
> >
> > Thanks I have disabled CONFIG_HAVE_ARM_ARCH_TIMER.
> >
> >> To clarify, I don't expect these to be merged upstream. Also, I don't
> >> believe the standard kernel config needs to work out of the box.
> >
> > That is fine. But even with the above tweak I still don't see 8 CPUs
> > so I do wonder what the trick to seeing them is.
> >
> > ...
> > CPU: Testing write buffer coherency: ok
> > CPU0: update cpu_power 1535
> > CPU0: thread -1, cpu 0, socket 0, mpidr 80000000
> > Setting up static identity map for 0x404141b8 - 0x40414210
> > CPU1: Booted secondary processor
> > CPU1: update cpu_power 1535
> > CPU1: thread -1, cpu 1, socket 0, mpidr 80000001
> > CPU2: Booted secondary processor
> > CPU2: update cpu_power 1535
> > CPU2: thread -1, cpu 2, socket 0, mpidr 80000002
> > CPU3: Booted secondary processor
> > CPU3: update cpu_power 1535
> > CPU3: thread -1, cpu 3, socket 0, mpidr 80000003
> > CPU4: failed to boot: -22
> > CPU5: failed to boot: -22
> > CPU6: failed to boot: -22
> > CPU7: failed to boot: -22
> > Brought up 4 CPUs
> > SMP: Total of 4 processors activated.
> > CPU: All CPU(s) started in SVC mode.
> > ...
> 
> 
> Thanks for providing more information. I wonder, perhaps I forgot to
> tell you to also pass "apmu=multicluster" on the kernel command line
> to enable all cpus?

Thanks, that was the piece that I was missing.

> 
> The unlocking of multiplatform support is implemented in:
> [PATCH 01/02] ARM: shmobile: Multicluster APMU prototype support
> 
> Cheers,
> 
> / magnus
> --
> To unsubscribe from this list: send the line "unsubscribe linux-sh" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2014-04-10  5:48 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-04-08 12:47 [PATCH 00/02] ARM: shmobile: r8a7790 8-core boot prototype Magnus Damm
2014-04-09  0:24 ` Simon Horman
2014-04-09  4:11 ` Magnus Damm
2014-04-09  5:05 ` Simon Horman
2014-04-09 23:57 ` Magnus Damm
2014-04-10  5:48 ` Simon Horman

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).