The ABINIT Users Mailing List ( CLOSED )

[Dan Ricinschi <ricinski@semi.ee.es.osaka-u.ac.jp>]

Dear abinit-ers,

I would appreciate your advice with the following problem. I 
attempted to restart/continue the structural optimization in the 
normal way, by setting irdwfk 1 and restartxf 1, when running abinip 
(v 5.2.4) with k-point parallelisation, but it appears to be 
impossible. I did some searches and this seems related to the 
following "known problem":

P416. (XG030522)
CRASH
Seen on Intel/PC with PGI compiler, in parallel
Test_v1#75,80  do not work
Problem to read XFhistory part of a wavefunction file.

Well, in my case Test_v1#75,80 do not crash, but the XFhistory is 
simply not read, and I get this message before the code starts the 
optimization from scratch:

 "gstate : reading   0 (x,f) history pairs from input wf file"

Is there any way that I could restart the optimization with abinip in 
a current version?

If not, would it be worth to try to naively sidestep this for 
immediate purposes e. g. by forcing nxfh=20 in gstate.F90 and/or 
outxfhist.F90 (and recompile in order to continue with Broyden step 
21), or full-fledged debugging is necessary for solving this?

I appologize if this has been discussed and I was not able to find 
it. Thank you in advance for your replies.

Dan Ricinschi