problem with readonly pinfo?

Matt Hargett matt@use.net
Thu Sep 18 18:09:00 GMT 2003


> On Wed, Sep 17, 2003 at 08:23:42PM +0200, Corinna Vinschen wrote:
> >On Wed, Sep 17, 2003 at 02:10:27PM -0400, Christopher Faylor wrote:
> >> On Wed, Sep 17, 2003 at 12:40:46PM -0400, Pierre A. Humblet wrote:
> >> >I was thinking about one event in catchem (sigproc) and the same kind
of
> >> >processing as what is done today. Don't know how compatible that is
with
> >> >your new implementation.
> >>
> >> That would be perfect for the old method.  I'm just blocking on a pipe
now.
> >> If MSFT actually implemented overlapped reads for pipes this wouldn't
be
> >> an issue.  I don't know why they limited things in this fashion.
> >
> >Weird enough, they implemented overlapped IO for named pipes.  Just for
> >anonymous pipes (and therefore not on 9x/Me) it's not implemented.
>
> Yeah.  Have you ever tested overlapped I/O on regular pipes on NT?  I
think
> I have but I don't remember the results.  I've suspected that anonymous
pipes
> are just named pipes under the hood.

You are correct, if my memory serves me correctly. They're both implemented
in NPFS.SYS (named pipe file system). Filemon from sysinternals.com can
monitor read/writes to named pipes, which is useful for debugging these
critters.



More information about the Cygwin-developers mailing list