Perfect! That feels much better. Thanks.
-Jordan
On Wednesday, February 28, 2018 at 4:38:57 PM UTC-7, Ian Lance Taylor wrote:
>
> On Wed, Feb 28, 2018 at 1:58 PM, <jorda...@gmail.com <javascript:>>
> wrote:
> >
> > What is the best way to get an unsafe.Pointer to a []byte's underlying
> > array?
>
> unsafe.Pointer(&b[0])
>
>
> > I have added this function to syscall/syscall_linux.go to support some
> of
> > our experiments. One of the new getsockopt calls returns a variable
> length
> > amount of data into the buffer, which has no real set size. As such,
> > userspace programs will have to retry with a larger buffer if the
> previous
> > try was not large enough.
> >
> > func GetsockoptString(fd, level, opt, max_size int) (string, error) {
> > var arr []byte = make([]byte, max_size, max_size)
> > vallen := _Socklen(max_size)
> >
> > // This is really gross and bad, but first entry in a slice is a pointer
> > into the array
> > t := (**byte)(unsafe.Pointer(&arr))
> >
> > err := getsockopt(fd, level, opt, unsafe.Pointer(*t), &vallen)
> >
> > return string(arr[:vallen]), err
> > }
> >
> >
> > This function works fine, but seems gross to me. Am I safe to always
> assume
> > that the first entry in a slice is always a pointer into the array for
> all
> > of Go1.x on Linux?
>
> No.
>
> Ian
>
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to golang-nuts+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
