The ABINIT Users Mailing List ( CLOSED )

[Kunio Nishida <mm02191@murata.co.jp>]

Dear Abinit users,

Sometimes ago, we had asked NEC to do some benchmarks of
ABINIT(v3.4.4) on SX-6i.
In that process, they had discovered some dimension mismatch in
Src_9drive/invars2m.f:line211:

kpt_with_shift(:,:)=dtsets(idtset)%kpt(:,:)/dtsets(idtset)%kptnrm

should be

kpt_with_shift(:,:)=dtsets(idtset)%kpt(:,:nkpt)/dtsets(idtset)%kptnrm
                                          ~~~~
This mismatch seems not a problem on other platforms,
but causes unpredictable input-errors on SX possibly by overwriting
other arrays.

We hope this is an help.

regards,

------
| Kunio Nishida         | 2288 Ooshinohara, Yasu, Yasu, Shiga 520-2393 JAPAN 
| 
|                       | Yasu branch, Murata Mfg., Co. Ltd.                 |
| Functional Material   | phone:     +81-77-586-8363                         |
| Research Division     | facsimile: +81-77-587-1923                         |
| Murata Mfg., Co. Ltd. | email:     mm02191@murata.co.jp                    |