Displaying 4 results from an estimated 4 matches for "use_thread".
2020 May 26
2
Changing the BLAS from openblas on a F32 box
...anyone is interested, the issue Simon Wood reports with openblas
and *mgcv* is:
Issues:
** openblas 0.3.x x<7 is not thread safe if itself compiled for single thread
use and then called from multiple threads (unlike the reference BLAS, say).
0.2.20 appears to be OK. For 0.3.x x>6 make USE_THREAD=0 USE_LOCKING=1
to make openblas ensures thread safety (currently unclear if USE_LOCKING
will be default from 0.3.7).
and running `mgcv:::blas.thread.test()` (Simon's non-exported function
to test for problems) returns:
> mgcv:::blas.thread.test()
|...
2020 May 27
2
Changing the BLAS from openblas on a F32 box
...; > and *mgcv* is:
> >
> > Issues:
> >
> > ** openblas 0.3.x x<7 is not thread safe if itself compiled for single thread
> > use and then called from multiple threads (unlike the reference BLAS, say).
> > 0.2.20 appears to be OK. For 0.3.x x>6 make USE_THREAD=0 USE_LOCKING=1
> > to make openblas ensures thread safety (currently unclear if USE_LOCKING
> > will be default from 0.3.7).
>
> Does this mean that single-threaded openblas should be built with
> USE_LOCKING=1 by default? Is there any upstream recommendation about
>...
2020 May 27
0
Changing the BLAS from openblas on a F32 box
...imon Wood reports with openblas
> and *mgcv* is:
>
> Issues:
>
> ** openblas 0.3.x x<7 is not thread safe if itself compiled for single thread
> use and then called from multiple threads (unlike the reference BLAS, say).
> 0.2.20 appears to be OK. For 0.3.x x>6 make USE_THREAD=0 USE_LOCKING=1
> to make openblas ensures thread safety (currently unclear if USE_LOCKING
> will be default from 0.3.7).
Does this mean that single-threaded openblas should be built with
USE_LOCKING=1 by default? Is there any upstream recommendation about
this?
--
I?aki ?car
2020 May 27
0
Changing the BLAS from openblas on a F32 box
...; >
> > > Issues:
> > >
> > > ** openblas 0.3.x x<7 is not thread safe if itself compiled for single thread
> > > use and then called from multiple threads (unlike the reference BLAS, say).
> > > 0.2.20 appears to be OK. For 0.3.x x>6 make USE_THREAD=0 USE_LOCKING=1
> > > to make openblas ensures thread safety (currently unclear if USE_LOCKING
> > > will be default from 0.3.7).
> >
> > Does this mean that single-threaded openblas should be built with
> > USE_LOCKING=1 by default? Is there any upstream r...