[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <bf5ab15a-1151-1163-e139-a3657614d0fd@arm.com>
Date: Wed, 14 Mar 2018 15:53:45 +0000
From: Marc Zyngier <marc.zyngier@....com>
To: James Hogan <jhogan@...nel.org>
Cc: Matt Redfearn <matt.redfearn@...s.com>,
Ralf Baechle <ralf@...ux-mips.org>,
Thomas Gleixner <tglx@...utronix.de>,
linux-mips@...ux-mips.org, Jason Cooper <jason@...edaemon.net>,
Dengcheng Zhu <dengcheng.zhu@...s.com>,
linux-kernel@...r.kernel.org,
Philippe Ombredanne <pombredanne@...b.com>,
Paul Burton <paul.burton@...s.com>,
Kate Stewart <kstewart@...uxfoundation.org>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: Re: [PATCH 0/6] irqchip/mips-gic: Enable & use VEIC mode if available
On 14/03/18 15:46, James Hogan wrote:
> On Wed, Mar 14, 2018 at 11:15:47AM +0000, Marc Zyngier wrote:
>> Hi Matt,
>>
>> On 05/01/18 10:31, Matt Redfearn wrote:
>>>
>>> This series enables the MIPS GIC driver to make use of the EIC
>>> mode supported in some MIPS cores. In this mode, the cores 6
>>> interrupt lines are switched to represent a vector number,
>>> 0..63. Currently all GIC interrupts are routed to a single CPU
>>> interrupt pin, but this is inefficient since we end up checking
>>> both local and shared interrupt flag registers for both local
>>> and shared interrupts. This introduces additional latency into
>>> the interrupt paths. With EIC mode this can be improved by
>>> using separate vectors for local and shared interrupts.
>>>
>>> This series is based on 4.15-rc6 and has been tested on Boston,
>>> Malta & SEAD3 MIPS platforms implementing a GIC with and
>>> without EIC mode supported in hardware.
>>
>> What the status of this series?
>
> FYI I've been meaning to test it with KVM, since host EIC I think
> will affect KVM & guest stuff.
OK. I'll park that until I hear from you guys.
Thanks,
M.
--
Jazz is not dead. It just smells funny...
Powered by blists - more mailing lists