problems with ACML and single precision

Satish Balay balay at mcs.anl.gov
Fri Jun 29 09:21:56 CDT 2007


On Fri, 29 Jun 2007, Roberto Gori wrote:

> i would use C.  I mean ... for the single precision i would like to
> use the acml library with the intel C compiler as in petsc 2.3.1.
> Is this impossibile for the current version or in the next future?

Not sure why it was disabled. You can reverse this change with the
attach patch, and see if it works for you.

cd petsc-2.3.3-p3/python/BuildSystem
patch -Np1 -R < ~/singleprecision.patch

Satish
-------------- next part --------------
tree 432027e85207
parent ea4189ae214c
author bsmith at anlext2wls147.wl.anl-external.org 1177707677 18000
committer bsmith at anlext2wls147.wl.anl-external.org 1177707677 18000
revision 1985

if configuring with --with-batch should not require locating mpiexec
diff --git a/config/packages/BlasLapack.py b/config/packages/BlasLapack.py
--- a/config/packages/BlasLapack.py
+++ b/config/packages/BlasLapack.py
@@ -472,8 +472,8 @@ class Configure(config.package.Package):
     self.executeTest(self.checkESSL)
     self.executeTest(self.checkPESSL)
     self.executeTest(self.checkMissing)
-    if (self.defaultPrecision == 'longdouble' or self.defaultPrecision == 'int') and not self.f2cpkg:
-      raise RuntimeError('Need to use --download-c-blas-lapack when using --with-precision=longdouble/int')
+    if (self.defaultPrecision == 'longdouble' or self.defaultPrecision == 'single') and not self.f2cpkg:
+      raise RuntimeError('Need to use --download-c-blas-lapack when using --with-precision=longdouble/single')
     return
 
 if __name__ == '__main__':


More information about the petsc-users mailing list