Fuzzer: Distinguish traps from host limitations (#3801)

Host limitations are arbitrary and can be modified by optimizations, so
ignore them. For example, if the optimizer removes allocations then a
host limit on an allocation error may vanish. Or, an optimization that
removes recursion and replaces it with a loop may avoid a host limit
on call depth (that is not done currently, but might some day).

This removes a class of annoying false positives in the fuzzer.

2 files changed, 29 insertions, 1 deletions

diff --git a/test/passes/Oz_fuzz-exec_all-features.txt b/test/passes/Oz_fuzz-exec_all-features.txt
index 5d00a3db1..0da22edbd 100644
--- a/test/passes/Oz_fuzz-exec_all-features.txt
+++ b/test/passes/Oz_fuzz-exec_all-features.txt
@@ -33,7 +33,7 @@
 [LoggingExternalInterface logging 3]
 [trap cast error]
 [fuzz-exec] calling array-alloc-failure
-[trap allocation failure]
+[host limit allocation failure]
 (module
  (type $struct (struct (field (mut i32))))
  (type $void_func (func))
@@ -277,3 +277,16 @@
 [LoggingExternalInterface logging 3]
 [trap cast error]
 [fuzz-exec] calling array-alloc-failure
+ignoring comparison of ExecutionResults!
+[fuzz-exec] calling foo
+[host limit allocation failure]
+(module
+ (type $none_=>_i32 (func (result i32)))
+ (export "foo" (func $0))
+ (func $0 (; has Stack IR ;) (result i32)
+  (i32.const 0)
+ )
+)
+[fuzz-exec] calling foo
+[fuzz-exec] note result: foo => 0
+ignoring comparison of ExecutionResults!
diff --git a/test/passes/Oz_fuzz-exec_all-features.wast b/test/passes/Oz_fuzz-exec_all-features.wast
index 4092f6b34..c88133c3a 100644
--- a/test/passes/Oz_fuzz-exec_all-features.wast
+++ b/test/passes/Oz_fuzz-exec_all-features.wast
@@ -240,3 +240,18 @@
   )
  )
 )
+(module
+ (type $[mut:i8] (array (mut i8)))
+ (func "foo" (result i32)
+  ;; before opts this will trap on failing to allocate -1 >>> 0 bytes. after
+  ;; opts the unused value is removed so there is no trap, and a value is
+  ;; returned, which should not confuse the fuzzer.
+  (drop
+   (array.new_default_with_rtt $[mut:i8]
+    (i32.const -1)
+    (rtt.canon $[mut:i8])
+   )
+  )
+  (i32.const 0)
+ )
+)