[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <d5fbf81d-6823-4598-cd34-045a98b90770@ti.com>
Date: Wed, 7 Sep 2016 11:31:35 -0500
From: Suman Anna <s-anna@...com>
To: Santosh Shilimkar <santosh.shilimkar@...cle.com>,
Santosh Shilimkar <ssantosh@...nel.org>
CC: Russell King <linux@...linux.org.uk>,
Rob Herring <robh+dt@...nel.org>,
<linux-arm-kernel@...ts.infradead.org>,
<devicetree@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
Tero Kristo <t-kristo@...com>,
Murali Karicheri <m-karicheri2@...com>,
Vitaly Andrianov <vitalya@...com>
Subject: Re: [PATCH 0/5] Use mmio-sram driver for Keystone MSMC RAM
On 09/07/2016 11:25 AM, Santosh Shilimkar wrote:
> On 9/7/2016 9:22 AM, Suman Anna wrote:
>> Hi Santosh,
>>
>> On 09/07/2016 11:11 AM, Santosh Shilimkar wrote:
>>> Hi Suman,
>>>
>>> On 9/1/2016 3:58 PM, Suman Anna wrote:
>>>> Hi,
>>>>
>>>> The Keystone 2 family of SoCs have an on-chip RAM called the
>>>> Multicore Shared Memory (MSM) RAM. This RAM is accessible through
>>>> the Multicore Shared Memory Controller (MSMC). This series represents
>>>> these on-chip RAMs as sram nodes so that the memory allocations
>>>> can be managed by the in-kernel mmio-sram driver.
>>>>
>>>> The first 4 patches adds the basic SRAM nodes on each of the SoCs,
>>>> and the last patch enables the generic on-chip SRAM driver for
>>>> keystone defconfig.
>>>>
>>> The series looks good in general but I would like to understand
>>> the users of this memory in kernel. Is that going to be posted
>>> as a follow up patch ? Is the Power controller going to make
>>> use of this SRAM for PM code ?
>>
>> Yes, the users will eventually follow. Power Controller code is not
>> gonna be using this SRAM, it has its own RAM. This memory is gonna be
>> split between various functional features like IPC, OPTEE integration,
>> we already have the Boot Monitor code using this. We will have the
>> memory split by either having static child nodes or drivers requesting
>> the memory using gen_pool API.
>>
> OK. Its good to add the code at least with one active user of it.
> Since this has to anyway wait for another merge window, please post the
> users of it so that I can pull the combined patchset.
Yeah, will do. The first user will mostly be the corresponding child
nodes for the boot monitor.
regards
Suman
Powered by blists - more mailing lists