[PATCH 3/3] RISC-V: Add arch functions to support hibernation/suspend-to-disk

[JeeHeng Sia]

Hi,

In case if you are not aware, the latest series (v4) can be found at: https://patchwork.kernel.org/project/linux-riscv/list/?series=723586

Thanks
Regards
Jee Heng

> -----Original Message-----
> From: JeeHeng Sia
> Sent: Monday, 20 February, 2023 12:22 PM
> To: 'Pavel Machek' <pavel at ucw.cz>
> Cc: paul.walmsley at sifive.com; palmer at dabbelt.com; aou at eecs.berkeley.edu; linux-riscv at lists.infradead.org; linux-
> kernel at vger.kernel.org; Leyfoon Tan <leyfoon.tan at starfivetech.com>; Mason Huo <mason.huo at starfivetech.com>
> Subject: RE: [PATCH 3/3] RISC-V: Add arch functions to support hibernation/suspend-to-disk
> 
> Hi,
> 
> This is the old series, could you review the latest series at: https://patchwork.kernel.org/project/linux-riscv/list/?series=716143 ?
> 
> > -----Original Message-----
> > From: Pavel Machek <pavel at ucw.cz>
> > Sent: Saturday, 18 February, 2023 10:01 PM
> > To: JeeHeng Sia <jeeheng.sia at starfivetech.com>
> > Cc: paul.walmsley at sifive.com; palmer at dabbelt.com; aou at eecs.berkeley.edu; linux-riscv at lists.infradead.org; linux-
> > kernel at vger.kernel.org; Leyfoon Tan <leyfoon.tan at starfivetech.com>; Mason Huo <mason.huo at starfivetech.com>
> > Subject: Re: [PATCH 3/3] RISC-V: Add arch functions to support hibernation/suspend-to-disk
> >
> > Hi!
> >
> > > @@ -21,6 +21,11 @@ struct suspend_context {
> > >  #endif
> > >  };
> > >
> > > +/* This parameter will be assigned to 0 during resume and will be used by
> > > + * hibernation core for the subsequent resume sequence
> > > + */
> > > +extern int in_suspend;
> >
> > "/*" should be on separate line. (This occurs multiple times.)
> already resolved in v3 series.
> >
> > > +ENTRY(core_restore_code)
> > > +	/* switch to temp page table */
> > > +	csrw satp, s1
> > > +	sfence.vma
> > > +	beqz	s4, done
> > > +loop:
> > > +	/* The below code will restore the hibernated image. */
> > > +	ld	a1, HIBERN_PBE_ADDR(s4)
> > > +	ld	a0, HIBERN_PBE_ORIG(s4)
> > > +
> > > +	lui     a4, 0x1
> > > +	add     a4, a4, a0
> >
> > Do you have enough registers to move a4 = 0x1 assignment before the
> > loop:?
> I am sorry but I don't understand the question. Could you elaborate please?
> >
> > > +static int __init riscv_hibernate__init(void)
> > > +{
> > > +	hibernate_cpu_context = kcalloc(1, sizeof(struct suspend_context), GFP_KERNEL);
> > > +
> >
> > kzalloc?
> sure, will change it.
> >
> > > +	if (WARN_ON(!hibernate_cpu_context))
> > > +		return -ENOMEM;
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +early_initcall(riscv_hibernate__init);
> >
> > Does this really need to be done early?
> yes. To make sure it is initialized before next resume.
> >
> > Best regards,
> > 								Pavel
> > --
> > People of Russia, stop Putin before his war on Ukraine escalates.