Hi,
Yes, by flush I mean a fflush. I've determined there are no memory problems
with my app, and that pTh raises the abort() signal only shortly after a
socket is closed (and probably reused)...hence the desire to try to flush
the socket.
For handling the socket, I'm using the typical
socket/connect/bind/poll/select mechanisms. Basically, I'd like to find a
way to ensure that the socket is closed and there is nothing left to read
or write, so the socket can be reused, without blocking the whole process.
Sincerely,
Brent Phillips
At 08:58 PM 7/24/2001 +0200, you wrote:
>On Tue, Jul 24, 2001, Brent Phillips wrote:
>
> > Is it possible to use the pTh event mechanism to prevent the whole process
> > from blocking on a flush? I'm primarily thinking of using it for sockets.
> > Would it matter if the socket is in blocking or non-blocking mode?
> > Ideally, I'd like to close the socket on a thread that blocks until the
> > flush is done, but I'm happy with any way that doesn't block the whole
> > process.
>
>What exactly do you mean by "flush"? A socket by default has no
>explicitly flushable buffer AFAIK. So I guess you are using stdio's
>fflush() on a fdopen'ed socket? What system calls are you using?
>
> Ralf S. Engelschall
> [EMAIL PROTECTED]
> www.engelschall.com
>______________________________________________________________________
>GNU Portable Threads (Pth) http://www.gnu.org/software/pth/
>User Support Mailing List [EMAIL PROTECTED]
>Automated List Manager (Majordomo) [EMAIL PROTECTED]
______________________________________________________________________
GNU Portable Threads (Pth) http://www.gnu.org/software/pth/
User Support Mailing List [EMAIL PROTECTED]
Automated List Manager (Majordomo) [EMAIL PROTECTED]
