STR: run
wasm-opt a.wasm -all --roundtrip --print | md5sum
many times. Every 30 or so I see two different hashes.
Oddly this doesn't happen with --name-types done before the roundtrip, or with removing the roundtrip.
Two output wats are included in the zip. The diff is a minor ordering change:
11d10
< (type $ref|{mut:i32_mut:ref?|...0|_mut:i64_mut:i64_mut:i64_mut:ref?|{mut:i32}|}|_ref|{mut:i32}|_=>_none (func (param (ref ${mut:i32_mut:ref?|...0|_mut:i64_mut:i64_mut:i64_mut:ref?|{mut:i32}|}) (ref ${mut:i32}))))
12a12
> (type $ref|{mut:i32_mut:ref?|...0|_mut:i64_mut:i64_mut:i64_mut:ref?|{mut:i32}|}|_ref|{mut:i32}|_=>_none (func (param (ref ${mut:i32_mut:ref?|...0|_mut:i64_mut:i64_mut:i64_mut:ref?|{mut:i32}|}) (ref ${mut:i32}))))
cc @tlively
test.zip
STR: run
many times. Every 30 or so I see two different hashes.
Oddly this doesn't happen with
--name-typesdone before the roundtrip, or with removing the roundtrip.Two output wats are included in the zip. The diff is a minor ordering change:
cc @tlively
test.zip