RE: [RFC PATCH v2 20/67] Hexagon instruction utility functions

From: Taylor Simpson <tsimpson@quicinc.com>
To: "bcain@codeaurora.org" <bcain@codeaurora.org>,
	"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>
Cc: "philmd@redhat.com" <philmd@redhat.com>,
	"riku.voipio@iki.fi" <riku.voipio@iki.fi>,
	"richard.henderson@linaro.org" <richard.henderson@linaro.org>,
	"laurent@vivier.eu" <laurent@vivier.eu>,
	"aleksandar.m.mail@gmail.com" <aleksandar.m.mail@gmail.com>
Subject: RE: [RFC PATCH v2 20/67] Hexagon instruction utility functions
Date: Thu, 9 Apr 2020 20:22:44 +0000	[thread overview]
Message-ID: <BYAPR02MB4886A123EB6D0DF1362F9003DEC10@BYAPR02MB4886.namprd02.prod.outlook.com> (raw)
In-Reply-To: <094101d60ea0$204c0ee0$60e42ca0$@codeaurora.org>

> -----Original Message-----
> From: bcain=codeaurora.org@mg.codeaurora.org
> <bcain=codeaurora.org@mg.codeaurora.org> On Behalf Of Brian Cain
> Sent: Thursday, April 9, 2020 1:53 PM
> To: Taylor Simpson <tsimpson@quicinc.com>; qemu-devel@nongnu.org
> Cc: riku.voipio@iki.fi; richard.henderson@linaro.org; laurent@vivier.eu;
> philmd@redhat.com; aleksandar.m.mail@gmail.com
> Subject: RE: [RFC PATCH v2 20/67] Hexagon instruction utility functions
>
> > -----Original Message-----
> > From: Qemu-devel <qemu-devel-
> > bounces+bcain=codeaurora.org@nongnu.org> On Behalf Of Taylor
> Simpson
> > Sent: Friday, February 28, 2020 10:43 AM
> > To: qemu-devel@nongnu.org
> > Cc: riku.voipio@iki.fi; richard.henderson@linaro.org; laurent@vivier.eu;
> > Taylor Simpson <tsimpson@quicinc.com>; philmd@redhat.com;
> > aleksandar.m.mail@gmail.com
> > Subject: [RFC PATCH v2 20/67] Hexagon instruction utility functions
> ...
> > +int arch_sf_invsqrt_common(size4s_t *Rs, size4s_t *Rd, int *adjust)
> > +{
> ...
> > +    } else if (r_class == FP_INFINITE) {
> > +        /* EJP: or put Inf in num fixup? */
> > +        RsV = fSFINFVAL(-1);
> > +        RdV = fSFINFVAL(-1);
> > +    } else if (r_class == FP_ZERO) {
> > +        /* EJP: or put zero in num fixup? */
> > +        RsV = RsV;
> > +        RdV = fSFONEVAL(0);
> ...
>
> This "RsV = RsV" looks like a logic error?  Presumably it's safe to remove --
> unless there's some other field that should get initialized here?  PeV maybe?

Probably a copy/paste error.  I will remove it.

Thanks,
Taylor