cURL / Mailing Lists / curl-library / Single Mail

curl-library

Fwd: Re: test 91 fails - cause found?

From: Patrick Smith <patsmith_at_pobox.com>
Date: Wed, 04 Feb 2004 22:32:15 -0500

[More listserver problems? I sent the message below on Sunday, and
received a copy back. But it's not in the archives, so I'm wondering
whether it made it out to all the list members?]

-------- Original Message --------
Subject: Re: test 91 fails - cause found?
Date: Sun, 01 Feb 2004 23:37:36 -0500
From: Patrick Smith <patsmith_at_pobox.com>
To: curl-library_at_lists.sourceforge.net

Daniel Stenberg wrote:
> On Tue, 27 Jan 2004, Patrick Smith wrote:
>>By the way, even if the test normally succeeds on your system, I suspect you
>>can make it fail reliably by nice'ing the server.
>
> My take at repeating this problem failed, but it is getting late.

Oh well. Maybe if you increase the nice value?

> Here's my initial attempt at producing a fix for the test 91 failures. This
> makes libcurl retry an ECONNRESET reading if the connection is re-used and
> nothing was ever read from it...
>
> If you can repeat the test 91 failure, I'm interested in your comments on this
> patch.

The test still fails. :-(

In the patch, should sockerror be reset to 0 after a successful recv?

A quick look at the strace output shows pretty much the same sequence as
before - the recv fails with ECONNRESET and libcurl gives up,
complaining about an "empty reply". I don't have time right now to look
at things in more detail. Do you want the log and trace files?

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn
Received on 2004-02-05