Commit Graph

6 Commits

Author SHA1 Message Date
River Riddle
ffde975e21 NFC: Move AffineOps dialect to the Dialect sub-directory.
PiperOrigin-RevId: 264482571
2019-08-20 15:36:39 -07:00
River Riddle
ba0fa92524 NFC: Move LLVMIR, SDBM, and StandardOps to the Dialect/ directory.
PiperOrigin-RevId: 264193915
2019-08-19 11:01:25 -07:00
Alex Zinenko
60965b4612 Move GPU dialect to {lib,include/mlir}/Dialect
Per tacit agreement, individual dialects should now live in lib/Dialect/Name
with headers in include/mlir/Dialect/Name and tests in test/Dialect/Name.

PiperOrigin-RevId: 259896851
2019-07-25 00:41:17 -07:00
Nicolas Vasilache
0002e2964d Move affine.for and affine.if to ODS
As the move to ODS is made, body and region names across affine and loop dialects are uniformized.

PiperOrigin-RevId: 258416590
2019-07-16 13:45:47 -07:00
Nicolas Vasilache
e78ea03b24 Replace linalg.for by loop.for
With the introduction of the Loop dialect, uses of the `linalg.for` operation can now be subsumed 1-to-1 by `loop.for`.
This CL performs the replacement and tests are updated accordingly.

PiperOrigin-RevId: 258322565
2019-07-16 13:44:57 -07:00
Alex Zinenko
80e2871087 Extend AffineToGPU to support Linalg loops
Extend the utility that converts affine loop nests to support other types of
loops by abstracting away common behavior through templates.  This also
slightly simplifies the existing Affine to GPU conversion by always passing in
the loop step as an additional kernel argument even though it is a known
constant.  If it is used, it will be propagated into the loop body by the
existing canonicalization pattern and can be further constant-folded, otherwise
it will be dropped by canonicalization.

This prepares for the common loop abstraction that will be used for converting
to GPU kernels, which is conceptually close to Linalg loops, while maintaining
the existing conversion operational.

PiperOrigin-RevId: 257172216
2019-07-09 05:26:50 -07:00