[PATCH v2] ARM: save/restore diagnostic register on Cortex-A9 suspend/resume

[Will Deacon]

On Tue, Jun 24, 2014 at 06:40:27PM +0100, Tomasz Figa wrote:
> Hi Will,

Hello,

> On 24.06.2014 18:33, Will Deacon wrote:
> > On Tue, Jun 24, 2014 at 05:28:54PM +0100, Tomasz Figa wrote:
> >> Hi Shawn,
> >>
> >> On 22.06.2014 12:15, Shawn Guo wrote:
> >>> The CP15 diagnostic register holds ARM errata bits on Cortex-A9, so it
> >>> needs to be saved/restored on suspend/resume.  Otherwise, the
> >>> effectiveness of errata workaround gets lost together with diagnostic
> >>> register bit across suspend/resume cycle.
> >>>
> >>> The patch adds a couple of Cortex-A9 specific suspend and resume
> >>> functions to handle the diagnostic register across suspend/resume cycle.
> >>
> >> [snip]
> >>
> >>> +ENTRY(cpu_ca9mp_do_resume)
> >>> +	ldmia	r0!, {r4}
> >>> +	mcr	p15, 0, r4, c15, c0, 1  @ Diagnostic register
> >>
> >> What about platforms running in non-secure mode in which the register is
> >> read-only?
> > 
> > On A9, it should be write-ignore. Are you seeing problems on a real SoC?
> 
> I'm observing a complete system hang on Exynos4412-based Trats2 board if
> I try to write this register in resume from system-wide sleep.

That's certainly unexpected; I just double-checked that non-secure accesses
are treated as read-only/write-ignore.

> Note that the board is running under secure firmware, but there is no
> support for suspend/resume of such boards in mainline yet, so I'm
> testing on a work in progress (but mostly finished) series that is yet
> to be sent.

Maybe it could be unrelated to this register. What happens if you change
the code to read/write a different (architected) cp15 register?

Will