Re: [RESEND PATCH 2/2] arm: socfpga: Set share override bit of the l2 cache controller
From: Rob Herring
Date: Fri Feb 20 2015 - 08:55:17 EST
On Fri, Feb 20, 2015 at 1:15 AM, Dinh Nguyen <dinh.linux@xxxxxxxxx> wrote:
> Hi Rob,
> On 2/19/15 12:13 PM, Rob Herring wrote:
>> On Thu, Feb 19, 2015 at 11:06 AM, <dinguyen@xxxxxxxxxxxxxxxxxxxxx> wrote:
>>> From: Dinh Nguyen <dinguyen@xxxxxxxxxxxxxxxxxxxxx>
>>> By not having bit 22 set in the PL310 Auxiliary Control register (shared
>>> attribute override enable) has the side effect of transforming Normal
>>> Shared Non-cacheable reads into Cacheable no-allocate reads.
>>> Coherent DMA buffers in Linux always have a Cacheable alias via the
>>> kernel linear mapping and the processor can speculatively load cache
>>> lines into the PL310 controller. With bit 22 cleared, Non-cacheable
>>> reads would unexpectedly hit such cache lines leading to buffer
>> You really should be doing this in your bootloader.
> Can I ask what is your reasoning for doing this in the bootloader? It's
> seems like this is such a nice mechanism to do it here.
Primarily, this register is secure only and we try to avoid secure
mode setup in the kernel.
Russell also has had a patch to do this generically in his patch queue
forever. If we want this in the kernel, then we should apply that.
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/