3c8a4c6f47
Patch originally by Giorgis Georgakoudis (@ggeorgakoudis), typos and bugs introduced later by me. This patch allows us to remove redundant barriers if they are part of a "consecutive" pair of barriers in a basic block with no impacted memory effect (read or write) in-between them. Memory accesses to local (=thread private) or constant memory are allowed to appear. Technically we could also allow any other memory that is not used to share information between threads, e.g., the result of a malloc that is also not captured. However, it will be easier to do more reasoning once the code is put into an AA. That will also allow us to look through phis/selects reasonably. At that point we should also deal with calls, barriers in different blocks, and other complexities. Differential Revision: https://reviews.llvm.org/D118002
24 lines
830 B
ReStructuredText
24 lines
830 B
ReStructuredText
.. _omp190:
|
|
|
|
Redundant barrier eliminated. (device only)
|
|
====================================================================
|
|
|
|
This optimization remark indicates that analysis determined an aligned
|
|
barrier in the device code to be redundant. This can occur when state
|
|
updates that have been synchronized by the barrier were eliminated too.
|
|
See also "Co-Designing an OpenMP GPU Runtime and Optimizations for Near-Zero
|
|
Overhead Execution", IPDPS'22.
|
|
|
|
Example
|
|
-------
|
|
|
|
This optimization will trigger for most target regions if state initialization
|
|
was removed as a consequence of "state forwarding". This will trigger for
|
|
internal runtime functions so it requires enabling verbose remarks with
|
|
`-openmp-opt-verbose-remarks` (prefixed with `-mllvm` for use with clang).
|
|
|
|
Diagnostic Scope
|
|
----------------
|
|
|
|
OpenMP optimization remark.
|