Re: [PATCH 2/2] PR libgcc/59714 complex division is surprising on aarch64

2017-10-13 Thread Ramana Radhakrishnan
On Fri, Oct 13, 2017 at 10:25 PM, Richard Earnshaw wrote: > On 13/10/17 18:28, vladimir.mezent...@oracle.com wrote: >> On 10/12/2017 03:40 AM, Richard Earnshaw wrote: >>> On 12/10/17 06:21, vladimir.mezent...@oracle.com wrote: From: Vladimir Mezentsev FMA (floating-point multiply-a

Re: [PATCH 2/2] PR libgcc/59714 complex division is surprising on aarch64

2017-10-13 Thread Richard Earnshaw
On 13/10/17 18:28, vladimir.mezent...@oracle.com wrote: > On 10/12/2017 03:40 AM, Richard Earnshaw wrote: >> On 12/10/17 06:21, vladimir.mezent...@oracle.com wrote: >>> From: Vladimir Mezentsev >>> >>> FMA (floating-point multiply-add) instructions are supported on aarch64. >>> These instructions

Re: [PATCH 2/2] PR libgcc/59714 complex division is surprising on aarch64

2017-10-13 Thread vladimir . mezentsev
On 10/12/2017 03:40 AM, Richard Earnshaw wrote: > On 12/10/17 06:21, vladimir.mezent...@oracle.com wrote: >> From: Vladimir Mezentsev >> >> FMA (floating-point multiply-add) instructions are supported on aarch64. >> These instructions can produce different result if two operations executed >> sep

Re: [PATCH 2/2] PR libgcc/59714 complex division is surprising on aarch64

2017-10-12 Thread Richard Earnshaw
On 12/10/17 06:21, vladimir.mezent...@oracle.com wrote: > From: Vladimir Mezentsev > > FMA (floating-point multiply-add) instructions are supported on aarch64. > These instructions can produce different result if two operations executed > separately. > -ffp-contract=off doesn't allow the FMA ins

[PATCH 2/2] PR libgcc/59714 complex division is surprising on aarch64

2017-10-11 Thread vladimir . mezentsev
From: Vladimir Mezentsev FMA (floating-point multiply-add) instructions are supported on aarch64. These instructions can produce different result if two operations executed separately. -ffp-contract=off doesn't allow the FMA instructions. Tested on aarch64-linux-gnu. No regression. Two failed t