Re: [Qemu-devel] [RFC 1/2] pci-dma-api-v1

[Anthony Liguori <anthony@codemonkey.ws>]

Jamie Lokier wrote:
> Blue Swirl wrote:
>   
>>>  I wonder how can possibly aio_readv/writev be missing in posix aio?
>>>  Unbelievable. It'd be totally trivial to add those to glibc, much
>>>  easier infact than to pthread_create by hand, but how can we add a
>>>  dependency on a certain glibc version? Ironically it'll be more
>>>  user-friendly to add dependency on linux kernel-aio implementation
>>>  that is already available for ages and it's guaranteed to run faster
>>>  (or at least not slower).
>>>       
>> There's also lio_listio that provides for vectored AIO.
>>     
>
> I think lio_listio is the missing aio_readv/writev.
>
> It's more versatile, and that'll by why POSIX never bothered with
> aio_readv/writev.
>   

No.  lio_listio is a batch submission mechanism, not an operation.

Each request maintains the same behavior as if they were submitted 
independently.  This means if you're doing a write operation and you 
submit a vectored request via lio_listio, each portion of the vector has 
to complete before the other one.

But most importantly, pthreads doesn't use readv/writev to implement 
lio_listio as has been mentioned elsewhere.

Regards,

Anthony Liguori

> Doesn't explain why they didn't _start_ with aio_readv before
> inventing lio_listio, but there you go.  Unix history.
>
> -- Jamie
>
>
>