httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Bill Stoddard" <stodd...@raleigh.ibm.com>
Subject Re: ap_send_fb_length() oddness
Date Thu, 13 Apr 2000 21:40:04 GMT
BTW, this is preventing ap_finalize_request_protocol() from sending the final 0 byte
chunked encoding terminator in some cases.  I can't say that I understand chunked
encoding, but this line of code indicating the connection is aborted seems to be the
likely culprit.

Bill


> Here is a piece of code from ap_send_fb_length().  The section of code handles the
> blocking read on a CGI pipe.
>
> Should we be seting connection aborted (r->connection->aborted = 1) when the other
side
of
> the pipe is closed signaling EOF?  This is totally non-intuitive to me. Why mark the
> network connection as aborted just because the pipe closed down?
>
>
>
>             ap_bsetopt(fb, BO_TIMEOUT, &r->server->timeout);
>             rv = ap_bread(fb, buf, sizeof(buf), &n);
>             if (n == 0) {
>                 if (rv == APR_SUCCESS) {        /* eof */
>                     (void) ap_rflush(r);
>                 }
>                 r->connection->aborted = 1;  /* LOOK HERE... Should we be doing
this
here?
> */
>                 break;
>             }
>             ap_bsetopt(fb, BO_TIMEOUT, &zero_timeout);
>
>
>


Mime
View raw message