On Tue, Dec 25, 2018 at 12:15 PM Philippe Mathieu-Daudé wrote:
>
> Hi Emilio,
>
> On 12/25/18 8:03 AM, Emilio G. Cota wrote:
> > The added branch to the FMA ops is marked as unlikely and therefore
> > its impact on performance (measured with fp-bench) is within noise range
> > when measured on an
Hi Emilio,
On 12/25/18 8:03 AM, Emilio G. Cota wrote:
> The added branch to the FMA ops is marked as unlikely and therefore
> its impact on performance (measured with fp-bench) is within noise range
> when measured on an Intel(R) Xeon(R) Gold 6142 CPU @ 2.60GHz.
Did you forget to write here the r
The added branch to the FMA ops is marked as unlikely and therefore
its impact on performance (measured with fp-bench) is within noise range
when measured on an Intel(R) Xeon(R) Gold 6142 CPU @ 2.60GHz.
Reported-by: Laurent Desnogues
Signed-off-by: Emilio G. Cota
---
v2: drop the cpuid bits, sin
