2022-04-13 14:49:06

by Maciej W. Rozycki

[permalink] [raw]
Subject: Re: [PATCH v4 04/11] mips: use fallback for random_get_entropy() instead of zero

On Wed, 13 Apr 2022, Thomas Bogendoerfer wrote:

> > diff --git a/arch/mips/include/asm/timex.h b/arch/mips/include/asm/timex.h
> > index b05bb70a2e46..abc60a6395e3 100644
> > --- a/arch/mips/include/asm/timex.h
> > +++ b/arch/mips/include/asm/timex.h
> > @@ -94,7 +94,7 @@ static inline unsigned long random_get_entropy(void)
> > else if (likely(imp != PRID_IMP_R6000 && imp != PRID_IMP_R6000A))
> > return read_c0_random();
> > else
> > - return 0; /* no usable register */
> > + return random_get_entropy_fallback(); /* no usable register */
> > }
> > #define random_get_entropy random_get_entropy
> >
> > --
> > 2.35.1
>
> Acked-by: Thomas Bogendoerfer <[email protected]>

Or we could drop the PRID_IMP_R6000/A check and the final `else' clause
entirely, as we don't even pretend to support the R6k at all anymore, and
this is the final reference remaining. For one we no longer handle the
CPU in `cpu_probe_legacy' so any attempt to boot on such a CPU would
inevitably fail as no CPU options would be set (we probably should have a
`panic' or suchlike as the default case for the switch statement there).

Therefore I'm all for removing this piece instead, complementing commit
3b2db173f012 ("MIPS: Remove unused R6000 support"), where it should have
really happened.

Maciej