I've just about finished up the patch to allow us to configure 
subdomain_id_type with ./configure.  The current syntax I'm using is 
"--with-subdomain-id-bytes=<1|2|4>" defaulting to "2".

I'm running into an issue now with our regression test suite where I'm getting 
diffs on tests that use the mesh generation facilities when I reconfigure with 
a different size subdomain_id_types.  The issue is that in exodusII_io_helper 
we take any element in subdomain 0 and remap it to 
numeric_limits<subdomain_id_type>::max() because we can't write subdomain 0 to 
an exodusII file - *sigh*.

Possible solutions:
- Update MeshGeneration to initialize a subdomain to something other than zero 
(such as one or perhaps some other larger interger < 256)
- Do nothing (Have applications deal with it perhaps by remapping libmesh 
generated meshes to something they can manage)
Any thoughts/opinions?

Also, does anyone have a good way to assure that we get the right size type for 
*any* system?  Configure tests the sizes of the various types and records them 
but I couldn't come up with a clean way to get those sizes into the actual 
typedef without a lot of seemingly hackish CPreprocessor commands.  Right now 
I'm just making assumptions (see id_types.h)

Roy,
Finally this patch requires a new configure to be built - would you prefer a 
patch submission or is it OK to commit the tested patch with the new configure 
built?  You'll probably want to update your regression test suite to configure 
for all three sizes  I've attached the current patch minus configure.  It's 
formatted from git so apply it with "-p1"

Attachment: 0001-Configurable-subdomain_id_type.patch
Description: Binary data


Cody
------------------------------------------------------------------------------
EditLive Enterprise is the world's most technically advanced content
authoring tool. Experience the power of Track Changes, Inline Image
Editing and ensure content is compliant with Accessibility Checking.
http://p.sf.net/sfu/ephox-dev2dev
_______________________________________________
Libmesh-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/libmesh-devel

Reply via email to