DiffArray Performance

Alastair Reid alastair at reid-consulting-uk.ltd.uk
Wed Nov 5 18:39:22 EST 2003


> The problem was whether DiffArrays should be thread-safe in the
> Concurrent Haskell sense, which means protecting access to the DiffArray
> with an MVar.

Can you give some estimate of the cost of using an MVar in this way?

Could some cheaper mechanism be provided for the special case of a mutex?
e.g., the moral equivalent of disabling interrupts might be cheaper (though 
I'd want to look at the diffarray code again before deciding if that style of 
solution were appropriate).

--
Alastair



More information about the Glasgow-haskell-users mailing list