Bug in the Kerker mixing

[Matt W]

Hmm, OK, I'm running a snapshot development from a week or two ago

 CP2K| version string:                    CP2K version 2.7 (Development 
Version)
 CP2K| source code revision number:                                 
 git:8a8c4da

please check whether you can run the XAS regtest I tried above. If that 
works then please post a full working example that shows the problem.

Matt

On Friday, February 20, 2015 at 9:18:56 AM UTC, Ali Akbari wrote:
>
> Hi Matt, 
>
> Actually, I tested it with precompiled versions (the same input file). 
> Which version do you use?
>
> Best,
> Ali
>
>
>
>
>  
>
> On Friday, 20 February 2015 11:13:01 UTC+2, Matt W wrote:
>>
>> Hi,
>>
>> Kerker mixing works for me in the c8_broy_gapw_all_xashh.inp example in 
>> regtest-gapw-1 (I change the Broyden mixing to Kerker mixing in the input 
>> file)..
>>
>> I think the segfault is more likely indicating that your executable is 
>> not correctly compiled. Can you run the regtests?
>>
>> You could also download a precompiled executable from the website and 
>> test whether that worked for your input.
>>
>> Matt
>>
>> On Thursday, February 19, 2015 at 3:49:09 PM UTC, Ali Akbari wrote:
>>>
>>> Hi, 
>>>
>>> I have noticed that the Kereker mixing does not work any longer in the 
>>> XAS calculation and produces the following message.
>>>
>>> * Program received signal SIGSEGV: Segmentation fault - invalid memory 
>>> reference.*
>>>
>>> I have tested different versions of CP2K and this bug appears in 
>>> versions of 2.5 and 2.6.
>>>
>>> Best,
>>> Ali
>>>  
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20150220/749f0cf3/attachment.htm>