Mea Culpa (of sorts)
After backing up to older versions of CYGWIN stuff (and an older version of my working
code)
I discovered it was because I had added "better" error detection to the read in my
code.
Funny thing is under CYGWIN read returns -1 with errno = 11 (EAGAIN or EWOULDBLOCK),
but the same
code compiled an run under Gnu/Linux never does that.
Unfortunately I thought I had failed with my older and newest CYGWIN.
I WAS WRONG.
After adding "detecting" for this condition, it compiles and runs the same under
Latest CYGWIN and
Linux.
Sorry for your time and Thanks for your help.
PS: One annoying thing I've notice running CYGWIN "setup" is that is keeps
creating mounts for
/usr/lib and /usr/bin
I Don't want this as I have things in the real paths.
So every time I run a Setup I must umount /usr/lib /usr/bin
Is there a way to disable this "Feature" ?
-- joel
Earnie Boyd wrote:
>
> Joel wrote:
>
> > Hi Chris:
> >
> > I think I may have bumped into a new bug with the latest CYGWIN dll.
> > I have an application the does non canonical serial i/o that worked previously
>until today when I
> > updated the DLL.
> >
> > I am getting an (Resource temporarily unavailable) error back from the read on
>ttyS2.
> > (a side note: why isn't port numbering consistent with Gnu/Linux?)
> >
> > It may be related to the length of the write previous to the read, as I can read
>successfully when I
> > write only 19 bytes but fails when I write 75 bytes.
> >
> > I hope this makes sense, and Thanks for all your work on Cygwin
>
> This sounds familiar. You may want to check the archives. I was about a month or
>so ago. I don't
> remember the topic but I remember that Corinna answered it. Something to do with
>the speed with which
> reads and writes happen.
>
> Earnie.
>
> _________________________________________________________
> Do You Yahoo!?
> Get your free @yahoo.com address at http://mail.yahoo.com
--
Want to unsubscribe from this list?
Check out: http://cygwin.com/ml/#unsubscribe-simple