On Nov 12, 2010, at 6:39 AM, Joerg Sonnenberger wrote: > On Fri, Nov 12, 2010 at 03:25:04PM +0100, Johnny Billquist wrote: >> So what you are arguing is that MI needn't be so much MI anymore, >> and that supporting anything more than mainstream today is more to >> be considered a lucky accident than a desired goal? > > It is not about mainstream. Please tell me one architecture that has > been created in the last 10 years, supports at least 32bit address > space, virtual memory and doesn't support either CAS or LL/SC. I would > be very surprised if there is one. Like I said, there has been a lot of > research into scalable primitives and algorithms and processor design > has changed accordingly. The ras-like UP implementation of CAS is almost > free as long as your interrupt rate is much lower than the rate of CAS > operations. There are a lot of low hanging fruits for VAX like providing > proper assembler versions of the atomic add etc, which sound feasible > from the docs I found. There are other areas where VAX hasn't kept up > with the "mainstream" architectures. pmap_growkernel to name one with > practical implications.
atomic_add isn't possible since ADAWI is only 16 bits and then you have to worry about carry to the upper word. pmap_growkernel isn't really practical to the way the VAX MMU works since it needs contiguous physical pages for the system page table.
