Re: Make the vectoriser drop to strided accesses for stores with gaps

2019-01-08 Thread H.J. Lu
On Fri, Jul 20, 2018 at 3:57 AM Richard Sandiford wrote: > > We could vectorise: > > for (...) >{ > a[0] = ...; > a[1] = ...; > a[2] = ...; > a[3] = ...; > a += stride; >} > > (including the case when stride == 8) but not: > >

Re: Make the vectoriser drop to strided accesses for stores with gaps

2018-08-22 Thread Richard Sandiford
Richard Biener writes: > On Fri, Jul 20, 2018 at 12:57 PM Richard Sandiford > wrote: >> >> We could vectorise: >> >> for (...) >>{ >> a[0] = ...; >> a[1] = ...; >> a[2] = ...; >> a[3] = ...; >> a += stride; >>} >> >> (including the

Re: Make the vectoriser drop to strided accesses for stores with gaps

2018-07-20 Thread Richard Sandiford
Richard Biener writes: > On Fri, Jul 20, 2018 at 12:57 PM Richard Sandiford > wrote: >> >> We could vectorise: >> >> for (...) >>{ >> a[0] = ...; >> a[1] = ...; >> a[2] = ...; >> a[3] = ...; >> a += stride; >>} >> >> (including the

Re: Make the vectoriser drop to strided accesses for stores with gaps

2018-07-20 Thread Richard Biener
On Fri, Jul 20, 2018 at 12:57 PM Richard Sandiford wrote: > > We could vectorise: > > for (...) >{ > a[0] = ...; > a[1] = ...; > a[2] = ...; > a[3] = ...; > a += stride; >} > > (including the case when stride == 8) but not: > >