[llvm-commits] [PATCH] Incorrect VFP version supporting ARM Cortex-M4

[Jiangning Liu]

Jim,

ARM ARM says,

ARMv7-M FP extension is described as FPv4-SP. It defines a Floating Point
Unit (FPU) that provides single-precision floating-point instructions. It
supports:

* FP extension registers that software can view as either 32
single-precision or 16 doubleword registers
* single-precision floating-point arithmetic
* conversions between integer, single-precision floating-point, and
half-precision floating point formats
* data transfers of single-precision and doubleword registers.

FPv4-SP is a single-precision only variant of the VFPv4-D16 extension of the
ARMv7-A and ARMv7-R architecture profiles.

Hopefully this can answer your question.

Thanks,
-Jiangning

> -----Original Message-----
> From: Jim Grosbach [mailto:grosbach at apple.com]
> Sent: Friday, July 27, 2012 12:33 AM
> To: Jiangning Liu
> Cc: 'Renato Golin'; Jiangning Liu; llvm-commits at cs.uiuc.edu
> Subject: Re: [llvm-commits] [PATCH] Incorrect VFP version supporting ARM
> Cortex-M4
> 
> There are lots more instructions than just vfma/vfms that are enabled by
> VFP4. Are these all supported by Cortex-M4?
> 
> -Jim
> 
> On Jul 26, 2012, at 2:47 AM, Jiangning Liu <jiangning.liu at arm.com> wrote:
> 
> >
> >
> >> -----Original Message-----
> >> From: rengolin at gmail.com [mailto:rengolin at gmail.com] On Behalf Of
> >> Renato Golin
> >> Sent: Thursday, July 26, 2012 5:21 PM
> >> To: Jiangning Liu
> >> Cc: Jiangning Liu; Jim Grosbach; llvm-commits at cs.uiuc.edu for LLVM
> >> Subject: Re: [llvm-commits] [PATCH] Incorrect VFP version supporting
> >> ARM
> >> Cortex-M4
> >>
> >> On 26 July 2012 09:35, Jiangning Liu <Jiangning.Liu at arm.com> wrote:
> >>> PING... Who can help to review this piece of code change?
> >>
> >> Hi Jiangning,
> >>
> >> Cortex-A15 seems to have VFPv4-D32, while cortex-M4 has VFPv4-SP, how
> >> do you suppose to differentiate them?
> >>
> >
> > I expect FeatureVFPOnlySP is to support SP-D16, although the name
> > doesn't represent d16 yet. At present, Cortex-M4 should be the only
> > one using this definition.
> >
> > Thanks,
> > -Jiangning
> >
> >>
> >> --
> >> cheers,
> >> --renato
> >>
> >> http://systemcall.org/
> >
> >
> >
> 
>