New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
longlong-double-fix resurrection part 1 #52
Conversation
1 similar comment
Temporary close because Travis fouls up with race condition when both branch and PR. |
pdlctype => 'PDL_Anyval', | ||
realctype => 'double', | ||
ppforcetype => 'anyval', | ||
usenan => 0, # chm set this to 1, but then get compile error |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This comment needs to be more descriptive or at least the underlying issue should be looked at.
The changes in this look good to me. And it seems to increase the code coverage. |
… is not being used.
This first implementation has the actual ctype still being double. The idea is to replace all PDL_Double usages by PDL_Anyval to make sure nothing breaks. Then change the implementation of the PDL_Anyval C type. Ed J modified chm's code to set new type's "usenan" to 0, not 1
Since PDL_Anyval is currently the same as double, this should be a no-op...all tests still pass.
generic_qsort, generic_qsortvec, generic_qsort_ind, and generic_qsortvec_ind needed A added to the list of types to generate. For the underlying type-specific pdl_qsort* routines I used the corresponding pdl_qsort_*_D for now. This will need modification once the new PDL_Anyval type is introduced.
This should be consistent with the usage for the other PDL data types. What to do with the 8bytes and endian issues TBD.
WARNING: there are 2 different typemap files in Basic/Core which is confusing. It seems that they both need pretty much the same entries. The difference being that in the typemap.pdl file the usage for the PDL specific routines is given with class syntax (why?)
This is for setting any type of PDL value. Original code assumed (double) worked for everything. Now we need to handle (64bit ints) as well.
It looks good to me. Just waiting on Travis. |
No description provided.