Monomorphization: Add a flag to control the required improvement (#6837)

The argument is the minimum benefit we must see for us to decide to optimize, e.g.

--monomorphize --pass-arg=monomorphize-min-benefit@50

When the minimum benefit is 50% then if we reduce the cost by 50% through
monomorphization then we optimize there. 95% would only optimize when we
remove almost all the cost, etc.

In practice I see 95% will actually tend to reduce code size overall, as while we add
monomorphized versions of functions, we only do so when we remove a lot of
work and size, and after inlining we gain benefits. However, 50% or even lower can
lead to better benchmark results, in return for larger code size, just like with
inlining. To be careful, the default is set to 95%.

Previously we optimized whenever we saw any benefit at all, which is the same
as requiring a minimum benefit of 0%. Old tests have the flag applied in this PR
to set that value, so they do not change.

1 files changed, 4 insertions, 1 deletions

diff --git a/scripts/fuzz_opt.py b/scripts/fuzz_opt.py
index a4adb7b79..4087abed1 100755
--- a/scripts/fuzz_opt.py
+++ b/scripts/fuzz_opt.py
@@ -1566,7 +1566,10 @@ opt_choices = [
     ("--memory-packing",),
     ("--merge-blocks",),
     ('--merge-locals',),
-    ('--monomorphize',),
+    # test a few monomorphization levels, and also -always
+    ('--monomorphize', '--pass-arg=monomorphize-min-benefit@0'),
+    ('--monomorphize', '--pass-arg=monomorphize-min-benefit@50'),
+    ('--monomorphize', '--pass-arg=monomorphize-min-benefit@95'),
     ('--monomorphize-always',),
     ('--no-stack-ir',),
     ('--once-reduction',),