Re: a couple proposed patches

[Christopher Faylor <cgf at redhat dot com>]

On Fri, May 09, 2003 at 12:50:48PM -0400, Joe Buehler wrote:
>Christopher Faylor wrote:
>
>>How could this possibly work?  You get a signal, keep reading, and
>>then another process which also has the pipe open wants to grab data from
>>the pipe.  But, it can't get it because the original process has sucked
>>it all in.  So you store the data in shared memory?  And have to worry
>>about who gets what data when?  Ugh.  That wouldn't work with native
>>windows apps anyway.
>
>That was one of the issues about the implementation.  Essentially the
>scenario involves two processes reading the same pipe, with one being
>interrupted on a read() so the other one can read().  Does anything do
>this?  And does it matter whether the first process actually read
>anything from the pipe?

It doesn't matter if anyone does this.  You're breaking a paradigm.  That's
unacceptable.

>Native apps are not involved in my patch.  All that is affected is Cygwin
>process reads.

Native apps are the most affected.  You could conceivably work around
the issue with cooperating cygwin processes.  Native, cygwin-unaware
processes wouldn't be able to deal with this behavior.

cgf