luke
/
verilator
mirror of https://github.com/verilator/verilator.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
verilator/test_regress/t/t_opt_balance_cats.py

21 lines
716 B
Python
Raw Normal View History

Balance concatenations in DFG (#5598) The DFG peephole pass converts all associative trees into right leaning, which is good for simplifying pattern recognition, but can lead to an excessive amount of wide intermediate results being constructed for right leaning concatenations. Add a new pass to balance concatenation trees by trying to: - Create VL_EDATASIZE (32-bit) sub-terms, so words can then be packed easily afterwards - Try to ensure the operands of a concat are roughly the same width within a concatenation tree. This does not yield the shortest tree, but it ensures it has many sub-nodes that are small enough to fit into machine registers. This can eliminate a lot of wide intermediate results, which would need temporaries, and also increases ILP within sub-expressions (assuming the C compiler can't figure that out itself). This is over 2x run-time speedup on the high_perf configuration of VeeR EH2 (which you could arguably also get with -fno-dfg, but oh well).
2024-11-09 19:14:19 +01:00
#!/usr/bin/env python3
# DESCRIPTION: Verilator: Verilog Test driver/expect definition
#
# Copyright 2024 by Wilson Snyder. This program is free software; you
# can redistribute it and/or modify it under the terms of either the GNU
# Lesser General Public License Version 3 or the Perl Artistic License
# Version 2.0.
# SPDX-License-Identifier: LGPL-3.0-only OR Artistic-2.0
import vltest_bootstrap
test.scenarios('vlt')
Fix splitting of assignments to SC variables (#6329) (#6336) SC variables can only be assigned whole, so do not split up concatenation assignments in V3FuncOpt. Fixes #6329
2025-08-28 09:48:51 +02:00
test.compile(
verilator_flags2=["--stats", "--build", "--gate-stmts", "10000", "--expand-limit", "128"])
Balance concatenations in DFG (#5598) The DFG peephole pass converts all associative trees into right leaning, which is good for simplifying pattern recognition, but can lead to an excessive amount of wide intermediate results being constructed for right leaning concatenations. Add a new pass to balance concatenation trees by trying to: - Create VL_EDATASIZE (32-bit) sub-terms, so words can then be packed easily afterwards - Try to ensure the operands of a concat are roughly the same width within a concatenation tree. This does not yield the shortest tree, but it ensures it has many sub-nodes that are small enough to fit into machine registers. This can eliminate a lot of wide intermediate results, which would need temporaries, and also increases ILP within sub-expressions (assuming the C compiler can't figure that out itself). This is over 2x run-time speedup on the high_perf configuration of VeeR EH2 (which you could arguably also get with -fno-dfg, but oh well).
2024-11-09 19:14:19 +01:00
Move Concat balancing from DFG to FuncOpt (#5602) This means it applies more widely, e.g. inside sequential logic.
2024-11-10 18:23:11 +01:00
test.file_grep(test.stats, r'Optimizations, FuncOpt concat trees balanced\s+(\d+)', 1)
Split up assignments to wides with Concat on the RHS (#5599) Add a new pass to split up (recursively): foo = {l, r}; into the following, with the right indices, iff the concatenation straddles a wide word boundary. foo[_:_] = r; foo[_:_] = l; This eliminates more wide temporaries. Another 23% speedup on VeeR EH2 high_perf. Also brings the predicted stack size from 8M to 40k.
2024-11-10 16:51:59 +01:00
test.file_grep(test.stats, r'Optimizations, FuncOpt concat splits\s+(\d+)', 62)
Balance concatenations in DFG (#5598) The DFG peephole pass converts all associative trees into right leaning, which is good for simplifying pattern recognition, but can lead to an excessive amount of wide intermediate results being constructed for right leaning concatenations. Add a new pass to balance concatenation trees by trying to: - Create VL_EDATASIZE (32-bit) sub-terms, so words can then be packed easily afterwards - Try to ensure the operands of a concat are roughly the same width within a concatenation tree. This does not yield the shortest tree, but it ensures it has many sub-nodes that are small enough to fit into machine registers. This can eliminate a lot of wide intermediate results, which would need temporaries, and also increases ILP within sub-expressions (assuming the C compiler can't figure that out itself). This is over 2x run-time speedup on the high_perf configuration of VeeR EH2 (which you could arguably also get with -fno-dfg, but oh well).
2024-11-09 19:14:19 +01:00
test.passes()