RE: [PATCH 2/5] X86: Hyper-V: Enable IPI enlightenments

[KY Srinivasan]

> -----Original Message-----
> From: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
> Sent: Thursday, April 26, 2018 2:32 PM
> To: KY Srinivasan <kys@xxxxxxxxxxxxx>
> Cc: x86@xxxxxxxxxx; gregkh@xxxxxxxxxxxxxxxxxxx; linux-
> kernel@xxxxxxxxxxxxxxx; devel@xxxxxxxxxxxxxxxxxxxxxx; olaf@xxxxxxxxx;
> apw@xxxxxxxxxxxxx; jasowang@xxxxxxxxxx; tglx@xxxxxxxxxxxxx;
> hpa@xxxxxxxxx; Stephen Hemminger <sthemmin@xxxxxxxxxxxxx>; Michael
> Kelley (EOSG) <Michael.H.Kelley@xxxxxxxxxxxxx>; vkuznets@xxxxxxxxxx
> Subject: Re: [PATCH 2/5] X86: Hyper-V: Enable IPI enlightenments
> 
> On Wed, Apr 25, 2018 at 11:12:47AM -0700, kys@xxxxxxxxxxxxxxxxx wrote:
> > +/*
> > + * IPI implementation on Hyper-V.
> > + */
> > +
> > +static int __send_ipi_mask(const struct cpumask *mask, int vector)
> > +{
> > +	int cur_cpu, vcpu;
> > +	struct ipi_arg_non_ex **arg;
> > +	struct ipi_arg_non_ex *ipi_arg;
> > +	int ret = 1;
> 
> Not specifically related to this patch, but hv code sometimes returns 1
> on error or U64_MAX.  It's slightly magical.  Maybe
> HV_STATUS_INVALID_HYPERCALL_INPUT (3) would be more appropriate?
> Or we
> could make a new more generic error code:
> 
> #define HV_STATUS_INVALID        1

Good point. We will look at cleaning this up.

Regards,

K. Y