I think we are still trying to clean up too early
ssl_engine_ext.c:179
if (apr_table_get(r->connection->notes, "ssl") != NULL)
always fails in the logging phase. Since we have closed and set the "ssl" notes
back to NULL, apr_table_get doesn't distinguish between 'found, no value' and
'not found' :(
Bill
----- Original Message -----
From: "MATHIHALLI,MADHUSUDAN (HP-Cupertino,ex1)" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Monday, July 30, 2001 9:59 PM
Subject: RE: [patch] Trying to cleanup SSL filter logic
> I got into the same problem of keepalives.. it renegotiates everytime - it's
> pretty ugly :-(. I still need to work on it. I was thinking of eliminating
> the APR_BLOCK_READ in churn() before trying to resolve the keepalive
> problem.. I'd rather work on the keepalive now..
>
> -Madhu
>
>
> -----Original Message-----
> From: William A. Rowe, Jr. [mailto:[EMAIL PROTECTED]]
> Sent: Monday, July 30, 2001 7:39 PM
> To: [EMAIL PROTECTED]
> Subject: Re: [patch] Trying to cleanup SSL filter logic
>
>
> From: "MATHIHALLI,MADHUSUDAN (HP-Cupertino,ex1)"
> <[EMAIL PROTECTED]>
> Sent: Monday, July 30, 2001 9:27 PM
>
>
> > I've tried to cleanup some of the logic in the filter processing of
> > the SSL data. The changes include :
> > - eliminated the use of ssl_log - it used to cause seg faults during
> cleanup
> > since the conn_rec will no longer be valid.
> > - eliminated the "for (;;)" processing loop in ssl_io_filter_Output() -
> > we'll have to do that in churn_output() if required, so that any remaining
> > OpenSSL data (if available) is transferred before we call the
> > CloseConnection.
> > - Any remaining data in SSL should be cleaned up ideally in the
> > APR_BUCKET_IS_EOS() processing stage itself, as we close the SSL
> connection
> > here.
>
> Doug and I thank you for your hard work (we are debugging the same as your
> post
> came through :) I'll test in a minute.
>
> Question - what happens with keepalives. Don't we see an EOS on every
> single
> request, on the same open connection? Do we really want to close it up
> right here?
>
>
>
