Hi,

In this patch series we use winkle for offlined cores. I successfully
tested the working of this with subcore functionality.

Test scenario was as follows:
1. Set SMT mode to 1, Set subores-per-core to 1
2. Offline a core, in this case cpu 32 (sending it to winkle)
3. Set subcores-per-core to 4
4. Online the core
5. Start a guest (Topology 1 core 2 threads) on a subcore, in this case
on cpu 36

This works without any glitch.

Thanks,
Shreyas

On Monday 25 August 2014 11:31 PM, Shreyas B. Prabhu wrote:
> Fast sleep is an idle state, where the core and the L1 and L2
> caches are brought down to a threshold voltage. This also means that
> the communication between L2 and L3 caches have to be fenced. However
> the current P8 chips have a bug wherein this fencing between L2 and
> L3 caches get delayed by a cpu cycle. This can delay L3 response to
> the other cpus if they request for data during this time. Thus they
> would fetch the same data from the memory which could lead to data
> corruption if L3 cache is not flushed.
> Patch 4 adds support to work around this.
> 
> 'Deep Winkle' is a deeper idle state where core and private L2 are powered
> off. While it offers higher power savings, it is at the cost of losing
> hypervisor register state and higher latency.
> Patch 5-9 adds support for winkle and uses it for offline cpus.
> 
> Patch 1 - Moves parameters required discover idle states to a location 
> common to both cpuidle driver and powernv core code
> Patch 2 - Populates idle state details from device tree
> Patch 3 - Enables cpus to run guest after waking up from fastsleep/winkle
> 
> 
> Cc: Benjamin Herrenschmidt <b...@kernel.crashing.org>
> Cc: Paul Mackerras <pau...@samba.org>
> Cc: Michael Ellerman <m...@ellerman.id.au>
> Cc: Rafael J. Wysocki <r...@rjwysocki.net>
> Cc: Srivatsa S. Bhat <sriva...@mit.edu>
> Cc: Preeti U. Murthy <pre...@linux.vnet.ibm.com>
> Cc: Vaidyanathan Srinivasan <sva...@linux.vnet.ibm.com>
> Cc: Rob Herring <robh...@kernel.org>
> Cc: Grant Likely <grant.lik...@linaro.org>
> Cc: devicet...@vger.kernel.org
> Cc: linux...@vger.kernel.org
> Cc: linuxppc-...@lists.ozlabs.org
> 
> Preeti U Murthy (2):
>   cpuidle/powernv: Populate cpuidle state details by querying the
>     device-tree
>   powerpc/powernv/cpuidle: Add workaround to enable fastsleep
> 
> Shreyas B. Prabhu (6):
>   powerpc/kvm/book3s_hv: Enable CPUs to run guest after waking up from
>     fast-sleep
>   powerpc/powernv: Add OPAL call to save and restore
>   powerpc: Adding macro for accessing Thread Switch Control Register
>   powerpc/powernv: Add winkle infrastructure
>   powerpc/powernv: Discover and enable winkle
>   powerpc/powernv: Enter deepest supported idle state in offline
> 
> Srivatsa S. Bhat (1):
>   powerpc/powernv: Enable Offline CPUs to enter deep idle states
> 
>  arch/powerpc/include/asm/machdep.h             |   4 +
>  arch/powerpc/include/asm/opal.h                |  10 ++
>  arch/powerpc/include/asm/paca.h                |   3 +
>  arch/powerpc/include/asm/ppc-opcode.h          |   2 +
>  arch/powerpc/include/asm/processor.h           |   6 +-
>  arch/powerpc/include/asm/reg.h                 |   1 +
>  arch/powerpc/kernel/asm-offsets.c              |   1 +
>  arch/powerpc/kernel/exceptions-64s.S           |  37 ++---
>  arch/powerpc/kernel/idle.c                     |  30 ++++
>  arch/powerpc/kernel/idle_power7.S              |  83 +++++++++-
>  arch/powerpc/platforms/powernv/opal-wrappers.S |   2 +
>  arch/powerpc/platforms/powernv/powernv.h       |   8 +
>  arch/powerpc/platforms/powernv/setup.c         | 217 
> +++++++++++++++++++++++++
>  arch/powerpc/platforms/powernv/smp.c           |  13 +-
>  arch/powerpc/platforms/powernv/subcore.c       |  15 ++
>  drivers/cpuidle/cpuidle-powernv.c              |  40 ++++-
>  16 files changed, 439 insertions(+), 33 deletions(-)
> 

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Reply via email to