Re: Rewriting the FIFO code

[Corinna Vinschen <corinna-cygwin at cygwin dot com>]

On Dec 26 14:00, Ken Brown wrote:
> On 12/26/2018 6:37 AM, Corinna Vinschen wrote:
> > On Dec 25 21:14, Ken Brown wrote:
> >> On 12/14/2018 8:43 AM, Ken Brown wrote:
> >>> I'll write
> >>> again when/if I've sorted it out.
> >>
> >> Hi Corinna,
> >>
> >> Here's a new start.  For now, at least, I'm only trying to accommodate one
> >> reader and several writers.  Maybe later I'll worry about more than one reader.
> >>
> >> The attached patch indicates the approach I have in mind.  There's much more to
> >> do, and I haven't thought about the duplex case yet.  And there isn't enough
> >> written for me to test it yet, except to make sure it compiles.  But I just want
> >> to see if you think the approach is reasonable before I continue.
> > 
> > Please go ahead.  I'm really excited that you're working on this.  FIFOs
> > are badly in need of a better approch anyway(*).  Turn the code upside
> > down, for all it's worth!
> 
> Thanks for the encouragement, and Happy Holidays to you too.  One question: In 
> the new AF_UNIX socket code you mostly used NT functions, but the existing FIFO 
> code uses Win32 functions.  Do you prefer NT functions for new code?

The NT functions have some advantages over the Win32 functions.
For instance, WaitNamedPipe is not interruptible, while
NtFsControlFile(FSCTL_PIPE_WAIT) can be called asynchronously
and then you can just wait for an event object via cygwait
(see fhandler_socket_unix::wait_pipe_thread).

So, in theory I'd prefer NT functions, but if you feel uncomfortable
and just want to implement away, feel free to go ahead with Win32
functions.


Corinna

-- 
Corinna Vinschen
Cygwin Maintainer

Attachment: signature.asc
Description: PGP signature