Your patch has the wrong polarity (the code you are adding is marked with '-' rather 
than
'+'). Please post the corrected patch and I'll look at it.

Bill

----- Original Message -----
From: "Brad Nicholes" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>;
<[EMAIL PROTECTED]>
Sent: Wednesday, January 02, 2002 2:25 PM
Subject: [PATCH] Win32/NetWare sockets.c


> Any feedback would be appreciated.  If it looks OK then I will go ahead
> and check it in.
>
> Brad
>
> >>> "William A. Rowe, Jr." <[EMAIL PROTECTED]> Wednesday, January 02,
> 2002 12:12:59 PM >>>
> I'll look at this this afternoon ... but Mr's Stoddard and Trawick
> have
> a wee bit of insight about Win32 Sockets API [much more than myself]
> and
> might be interested as well.
>
> Bill
>
> ----- Original Message -----
> From: "Brad Nicholes" <[EMAIL PROTECTED]>
> To: <[EMAIL PROTECTED]>
> Sent: Wednesday, January 02, 2002 10:37 AM
> Subject: Changes to sockets.c
>
>
> > Bill,
> >     Would you mind taking a quick look at this before I check it in.
>
> > Basically I moved the call to accept() before allocating the
> > apr_socket_t structure so that I can do non-blocking accepts without
> > chewing up a huge chunk of memory on WSAEWOULDBLOCK's.  No need to
> > allocate memory if it would have blocked anyway or recieved some
> other
> > error.  FYI, there may be other places where the APR functions don't
> > accomodate a non-blocking scheme, but I haven't had the time to look
> > into it yet.
> >
> > thanks,
> > Brad
> >
>

Reply via email to