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.

Sign up for GitHub

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

Jump to bottom

fix[next]: DaCe CFG analysis -> remove structurally inaccessible states #1598

Closed
wants to merge 1 commit into from
Closed

fix[next]: DaCe CFG analysis -> remove structurally inaccessible states #1598

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 10 additions & 0 deletions src/gt4py/next/program_processors/runners/dace_iterator/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,6 +24,8 @@
import numpy as np
from dace.sdfg import utils as sdutils
from dace.transformation.auto import auto_optimize as autoopt
from dace.transformation.pass_pipeline import Pipeline
from dace.transformation.passes import DeadStateElimination

import gt4py.next.iterator.ir as itir
from gt4py.next import common
Expand Down Expand Up @@ -298,6 +300,14 @@ def build_sdfg_from_itir(
# TODO(edopao): remove `inline_loop_blocks` when DaCe transformations support LoopRegion construct
sdutils.inline_loop_blocks(sdfg)

# TODO(kotsaloscv): remove the DeadStateElimination transformation once this issue is fixed: https://github.com/spcl/dace/issues/1625
# From DaCe v0.16 to v0.16.1, there were changes in the CFG analysis that appear to be incompatible
# with SDFGs that have structurally inaccessible states. The workaround involves calling DeadStateElimination first
# to remove those states so that they don't cause issues in subsequent passes.
dace_pipeline = Pipeline([DeadStateElimination()])
for sd in sdfg.all_sdfgs_recursive():
dace_pipeline.apply_pass(sd, {})

# run DaCe transformations to simplify the SDFG
sdfg.simplify()

Expand Down
Loading