What's So Bad About Posix I/O?

Re:What's So Bad About Posix I/O? by aidenn0 on Tuesday September 12, 2017 @02:45AM (#15225284)

yes, the term "durable" was perhaps a poor choice of words, but the paragraph that followed made it clear that they were aware of the specific requirements (specifically mentioning making dirty caches available to all processes)

[ Parent | on HN ]

Re:What's So Bad About Posix I/O? by dom0 on Tuesday September 12, 2017 @05:05AM (#15225846)

Durable has a very specific meaning for I/O and is just not correct here.

POSIX does intentionally not specify any durability at all (e.g. a no-op fsync is explicitly permitted).

[ Parent | on HN ]

Re:What's So Bad About Posix I/O? by dullgiulio on Tuesday September 12, 2017 @05:32AM (#15225949)

And in practice it is close to that, due to cheating hardware with a cache.

[ Parent | on HN ]

Re:What's So Bad About Posix I/O? by AstralStorm on Tuesday September 12, 2017 @03:27AM (#15225445)

You can actually do that with mmap and if you want shm. (Mmap works on files well.) Of course then you have to deal with all the dirt yourself.

[ Parent | on HN ]