Hi Mike,

On Thu, 18 Jun 2015 11:36:54 -0700
Michael Turquette [off-list ref] wrote:

Quoting Gregory CLEMENT (2015-06-17 04:57:08)

quoted

Hi Mike, Stephen,

On 16/06/2015 17:56, Boris Brezillon wrote:

quoted

The crypto SRAM, used by the armada 370 cpuidle code to workaround a bug
in the BootROM code, requires the crypto clk to be up and running.
Flag the crypto clk as IGNORE_UNUSED until we add the proper
infrastructure to define the crypto SRAM in the DT and reference the crypto
clk in this SRAM node.

Reported-by: Kevin Hilman <redacted>
Signed-off-by: Boris Brezillon <redacted>
Tested-by: Kevin Hilman <redacted>
Fixes: 29e623475c50267ee3aaa49ebb6cbe107bce8d2a ("clk: mvebu: add
missing CESA gate clk").
Cc: <redacted> #4.1

Acked-by: Gregory CLEMENT <redacted>

Would it possible to apply this fix on 4.1-rc8 ?

Hi Gregory,

I'm a bit confused by this. This patch fixes "clk: mvebu: add missing
CESA gate clk", right? That patch lives in clk-next. It was not pushed
to Linus' tree yet via a -fixes branch.

Thus I can just apply this patch to clk-next and the problem should be
resolved in clk-next, no?

Yes. I've tagged this patch as applicable on stable 4.1 in case it
didn't make it in the next -rc (which should be the last one before
4.1 is released).
I don't know if it's a good practice, but I've had several fixes which
missed the targeted release in the past, and tagging them as stable
helps in getting the patches applied afterwards.

Furthermore, I do not have 29e623475c50267ee3aaa49ebb6cbe107bce8d2a
anywhere in my tree. I'm not sure where this sha1 hash came from.

Oops, I mixed the commit hash in my local branch and the one in clk-next
(4d52b2acefdfceae0e47ed08324a96f511dc80b1).

Best Regards,

Boris

-- 
Boris Brezillon, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com