pcp
[Top] [All Lists]

Re: [pcp] another python qa failure

To: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Subject: Re: [pcp] another python qa failure
From: Nathan Scott <nathans@xxxxxxxxxx>
Date: Tue, 4 Aug 2015 19:21:02 -0400 (EDT)
Cc: PCP <pcp@xxxxxxxxxxx>
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <55C14011.6030500@xxxxxxxxxxxxxxxx>
References: <55C0687F.2090205@xxxxxxxxxxxxxxxx> <55C14011.6030500@xxxxxxxxxxxxxxxx>
Reply-to: Nathan Scott <nathans@xxxxxxxxxx>
Thread-index: ewGI8vwEMmhCi31UWtxi9FYG/1Y90g==
Thread-topic: another python qa failure
Hi Ken,

----- Original Message -----
> On 04/08/15 17:23, Ken McDonell wrote:
> > qa/739 this time on vm03 PCP 3.10.6 x86_64 Fedora 21
> 
> and the same test is failing on vm02 PCP 3.10.6 i686 openSUSE 13.2
> 
> 739 1s ... - output mismatch (see 739.out.bad)
> 14a15
> > ./739: line 32: 11035 Segmentation fault      $python
> > $here/src/test_pcp_options.py -a file param
> 
> so it looks like there is really something wrong here.

Sorry, ran outta time yesterday before replying.  This is something to
do with the python interpreter calling the pmOptions class destructor
unexpectedly (early?  before initialisation is complete, it seems).

Needs some more work in the python module code, I think - the destructor
is ensuring we don't leak memory, so does need to be called but only once
fully initialised.

cheers.

--
Nathan

<Prev in Thread] Current Thread [Next in Thread>