Re: [go-nuts] Are float64 and float32 exact in Golang (like strictfp in Java)

2020-09-25 Thread Rob Pike
There is more than one Go compiler and the use of fused-multiply-add is not guaranteed. The Go spec permits FMA but does not require it. I suggest reading the "Floating-point operators" section of the specification for clarity here. -rob On Sat, Sep 26, 2020 at 9:23 AM Ian Lance Taylor wrote:

Re: [go-nuts] Are float64 and float32 exact in Golang (like strictfp in Java)

2020-09-25 Thread Ian Lance Taylor
On Fri, Sep 25, 2020 at 4:03 PM Ryan Keppel wrote: > > Does golang generate the SSE2 over the x87 floating point instructions on > x64? In Java terms, it's always using strictfp? You could check the assembly > I suppose. On amd64 the Go compiler uses SSE2. It does not use x87. However, the Go

[go-nuts] Are float64 and float32 exact in Golang (like strictfp in Java)

2020-09-25 Thread Ryan Keppel
Does golang generate the SSE2 over the x87 floating point instructions on x64? In Java terms, it's always using strictfp? You could check the assembly I suppose. -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and