apr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Branko Čibej <br...@xbc.nu>
Subject Re: CVS access weiredness
Date Fri, 24 Nov 2000 11:21:25 GMT

Brian Behlendorf wrote:

> On Thu, 23 Nov 2000, Branko [ISO-8859-2] Čibej wrote:
>> Brian Behlendorf wrote:
>>> On Thu, 23 Nov 2000, Branko [ISO-8859-2] Čibej wrote:
>>>> Is anybody else seeing the follwing with "cvs update" (over ssh):

>>> Should now be fixed.
>> Hmm. I'm still seeing it.
> Um, OK, so the problem report was:
>>     [lots of output snipped]
>>     cvs server: Updating apr/time/unix
>>     cvs server: Updating apr/time/win32
>>     cvs server: Updating apr/user
>>     cvs server: Updating apr/user/unix
>>     cvs [update aborted]: shutting down connection to apache.org: Invalid argument
> You didn't say if you were using the anoncvs pserver or over an SSH
> tunnel, 

Yes I did, and you even quoted it, see above. :-)

> So I don't have an answer for you, unless you can provide more
> debugging details.

At first I thought I would ignore it, as I saw the error in a cygwin 
session, but the same thing happens on Solaris 2.6. Except that on 
Solaris the error is EBADF.

> Also, apr/user/unix is the last directory in an update, so I'm wondering
> if it's finishing yet there's something about the way the connection ends
> that confuses your local cvs client?

Yep, that's what I was thinking, too. The fact APR builds on both boxes 
seems to corroborate this.

Now let me see ... Aha, :pserver: checkout works fine, so does :pserver: 
update ... oops, I got it. :ext: update from the top of the APR tree 
works fine, but reports that error when I update a Subversion tree, 
which links to APR. So this is obviously a CVS problem. Maybe it's 
confused by mixing :pserver: (svn) and :ext: (apr) in the same working 
copy tree.

I apologise for the spam. :-)

Brane Čibej
    home:   <brane@xbc.nu>             http://www.xbc.nu/brane/
    work:   <branko.cibej@hermes.si>   http://www.hermes-softlab.com/
     ACM:   <brane@acm.org>            http://www.acm.org/

View raw message