Re: [PATCH v2] target/ppc: Generate storage interrupts for radix RC changes

["Nicholas Piggin" <npiggin@gmail.com>]

On Thu Jul 13, 2023 at 3:35 AM AEST, Shawn Anastasio wrote:
> On 7/12/23 11:56 AM, Cédric Le Goater wrote:
> > Hello Shawn,
> > 
> > On 7/12/23 18:13, Shawn Anastasio wrote:
> >> Change radix model to always generate a storage interrupt when the R/C
> >> bits are not set appropriately in a PTE instead of setting the bits
> >> itself.  According to the ISA both behaviors are valid, but in practice
> >> this change more closely matches behavior observed on the POWER9 CPU.
> > 
> > How did you spotted this dark corner case in emulation ? Do you have
> > MMU unit tests ?
>
> I'm currently porting Xen to Power and have been using QEMU's powernv
> model extensively for early bring up. I noticed the issue when my radix
> implementation worked in QEMU but failed on actual hardware since I
> didn't have a proper storage interrupt handler implemented.

Cool. This was on my todo list because we rely on it for nested HV
KVM too.

I actually didn't know about that odd effLIPD=0 exception, but it
looks right. How did you test that, by running with MSR[HV]=0 and
LPIDR=0 ?

For the patch,

Reviewed-by: Nicholas Piggin <npiggin@gmail.com>

Thanks,
Nick