Re:David Patterson Says It’s Time for New Computer Architect...
by solotronics on Monday September 17, 2018 @08:36PM
(#18011404)
probably a dumb question but if you exposed the low level memory manipulation wouldn't you just end up with something like assembly?
[ Parent
| on HN ]
Re:David Patterson Says It’s Time for New Computer Architect...
by Senderman on Monday September 17, 2018 @09:28PM
(#18011639)
I don't think it's a dumb question. And I think you're right: the closer you get to the way the CPU you're targeting likes its memory optimally managed, you're basically heading towards assembly.
I have notions that there can be something 'between' assembly and, for instance, C, but that's a daydream of mine without any details worked out yet.
Re:David Patterson Says It’s Time for New Computer Architect...
by setr on Monday September 17, 2018 @10:06PM
(#18011833)
I’m not sure you’re actually heading down the path of assembly in terms of the language itself, but you’re definitely moving towards the architectural lock-in of assembly.
But maybe thats fine if the language can express architectural-specific decisions (like memory layout) and abstract it away in a more sane fashion than ifdefsc and the rest of the codebase can maintain the pretense of portability.
Re:David Patterson Says It’s Time for New Computer Architect...
by klodolph on Monday September 17, 2018 @08:44PM
(#18011448)
Low-level memory work, if you care about things like data locations and exact layout of memory, is possible in practice with C. There are cases where you can run into aliasing problems, but controlling the actual memory layout is no problem at all. (There are some limitations.)