mlir-hlo/lib/Dialect/mhlo
Smit Hinsu bc7b6374c8 Fix handling of negative seeds in random number generator op kernels for XLA
Casting negative s32 number to u64 directly will have leading 1s in the representation which is not what we want to get a single u64 out of two s32 seeds. Fixed this by first getting unsigned number of the same bit-width.

PiperOrigin-RevId: 345902167
2020-12-05 18:55:41 -08:00
..
IR Fix handling of negative seeds in random number generator op kernels for XLA 2020-12-05 18:55:41 -08:00
transforms [kernel_gen] Lower max rank specialization from 6 to 5 2020-12-03 09:19:25 -08:00
CMakeLists.txt Add CMake files and lit configurations, enough for `ninja check-mlir-hlo` to pass on all the tests 2020-08-07 22:14:34 -07:00