Skip to content
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

Updating mpi-vector-mode branch with main #1595

Closed
wants to merge 63 commits into from
Closed

Updating mpi-vector-mode branch with main #1595

wants to merge 63 commits into from

Conversation

jlk9
Copy link
Collaborator

@jlk9 jlk9 commented Dec 20, 2023

wsmoses and others added 30 commits November 3, 2023 22:35
* Fix SCEV memory error

* Add smalltypeof to inactive
* [Blas] fix temporary allocation (and deallocation)

* fixups

* fix

* fix

* fix test
* Fix Blas diffuse on P^3 computation

* fixup
* Ring of springs integration test using modulo

* new bugs

* add phi fix

* Additional phi fix

---------

Co-authored-by: William S. Moses <gh@wsmoses.com>
- Clean up warnings, clang-format
- Clean up operation creation
- Remove unused code
- Use shadow type in LLVM AD impl
* Handle insert value of multi type

* Fix multi agg

* now with separated

* fix insertion index math

* Allow pointer in double addTo

* fix

* fix ins typetree

* fix erasure
tgymnich and others added 23 commits December 11, 2023 22:00
Co-authored-by: William Moses <gh@wsmoses.com>
* Use static clang plugin registration

* fix
* Zero sret copy

* fix undefv/null

* inactive mpfr_greater_p
* Nicer error for getindex

* Nice use in fwd handler
* [TypeAnalysis] improve memtransfer error handler

* also add better binop error
…xtenddatarestlengthone.cpp, ringspring3Drestlengthone.cpp
@jlk9 jlk9 requested a review from wsmoses December 20, 2023 00:06
@wsmoses
Copy link
Member

wsmoses commented Dec 20, 2023

@jlk9 I think you have this backwards, you need to merge your branch into main, not the other way around

wsmoses and others added 4 commits December 20, 2023 12:08
…1569)

* [mlir] more precise yet conservative aliasing analysis

- Update the points-to lattice to indicate that "other" pointers may
  have pointers to unknown location stored, but some specific pointers
  may have pointers to known locations.
- Add a trivial may/must write differentiation.
- The tests are *NOT* fully updated.

* [mlir] partially support unknown state in dense activity analysis

* some support is still missing in printers of activity analyses
* three tests are failing because they produce different results now
  (haven't investigated if the previous expectations were correct)

* [mlir] make tests pass

* Fix points-to-pointer analysis to return "other" value, which may be
  empty set (lattice bottom) rather than always returning "any"
  (lattice top).
* Don't get tricked by operator[] on maps default-initializing the value
  in `map[key] == deafult_value` comparison attempting to see if the map
  changed.
* Propagate changes in two places.
* Fix "other" in/out activity state in dense (pointer) activity
  analysis.

* Address review

* use i64 for data layout
@wsmoses wsmoses closed this Jan 4, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

10 participants