1156532 - Crash [@ js::NativeObject::setDenseElementWithType] involving --unboxed-objects

Status: RESOLVED DUPLICATE of bug 1149498

(Core :: JavaScript Engine, defect)

Description

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

function a()
function() {};
function b()
function() {};
try {} catch (e) {}
try {} catch (e) {}
try {} catch (e) {}
try {} catch (e) {}
try {} catch (e) {}
try {} catch (e) {}
try {} catch (e) {}
try {} catch (e) {}
try {} catch (e) {}
try {} catch (e) {}
try {} catch (e) {}
try {} catch (e) {}
(function() {
    for (let d in []) {
        (function() {})
    }
}());
for (let w in [{}, {}, {
        x: 3
    }, {
        x: 3
    }, {
        x: 3
    }, {}, {
        x: 3
    }, {}, {}, {
        x: 3
    }, {
        x: 3
    }, {}, {
        x: 3
    }, {
        x: 3
    }, {}, {
        x: 3
    }, {}, {}, {
        x: 3
    }, {
        x: 3
    }, {
        x: 3
    }, {
        x: 3
    }, {
        x: 3
    }, {
        x: 3
    }, {
        x: 3
    }, {}, {
        x: 3
    }, {}, {
        x: 3
    }, {
        x: 3
    }, {}, {
        x: 3
    }, {}, {
        x: 3
    }]) {}

crashes js debug shell on m-c changeset 50b95032152c with --fuzzing-safe --no-threads --no-ion --unboxed-objects --gc-zeal=14 at js::NativeObject::setDenseElementWithType.

Configure options:

CC="clang -Qunused-arguments" CXX="clang++ -Qunused-arguments" AR=ar AUTOCONF=/usr/local/Cellar/autoconf213/2.13/bin/autoconf213 sh /Users/skywalker/trees/mozilla-central/js/src/configure --target=x86_64-apple-darwin12.5.0 --enable-debug --enable-nspr-build --enable-more-deterministic --with-ccache --enable-gczeal --enable-debug-symbols --disable-tests

python -u ~/fuzzing/js/compileShell.py -b "--enable-debug --enable-more-deterministic --enable-nspr-build" -r 50b95032152c

autoBisect is currently running.

Comment 1

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

Attachment: stack

(lldb) bt 5
* thread #1: tid = 0x166d88, 0x00000001002bb8ed js-dbg-64-dm-nsprBuild-darwin-50b95032152c`js::NativeObject::setDenseElementWithType(js::ExclusiveContext*, unsigned int, JS::Value const&) [inlined] JSObject::isSingleton(this=0x00000001039871c0) const at jsobj.h:154, queue = 'com.apple.main-thread', stop reason = EXC_BAD_ACCESS (code=2, address=0x1039871c0)
  * frame #0: 0x00000001002bb8ed js-dbg-64-dm-nsprBuild-darwin-50b95032152c`js::NativeObject::setDenseElementWithType(js::ExclusiveContext*, unsigned int, JS::Value const&) [inlined] JSObject::isSingleton(this=0x00000001039871c0) const at jsobj.h:154
    frame #1: 0x00000001002bb8ed js-dbg-64-dm-nsprBuild-darwin-50b95032152c`js::NativeObject::setDenseElementWithType(js::ExclusiveContext*, unsigned int, JS::Value const&) at TypeInference-inl.h:145
    frame #2: 0x00000001002bb8ed js-dbg-64-dm-nsprBuild-darwin-50b95032152c`js::NativeObject::setDenseElementWithType(js::ExclusiveContext*, unsigned int, JS::Value const&) + 40 at TypeInference-inl.h:170
    frame #3: 0x00000001002bb8c5 js-dbg-64-dm-nsprBuild-darwin-50b95032152c`js::NativeObject::setDenseElementWithType(this=0x0000000103959040, cx=0x0000000101fa5180, index=33, val=0x00007fff5fbfd248) + 53 at NativeObject-inl.h:72
    frame #4: 0x0000000100268a6b js-dbg-64-dm-nsprBuild-darwin-50b95032152c`js::NativeDefineProperty(js::ExclusiveContext*, JS::Handle<js::NativeObject*>, JS::Handle<jsid>, JS::Handle<JSPropertyDescriptor>, JS::ObjectOpResult&) + 116 at NativeObject.cpp:1214
(lldb)

Comment 2

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

autoBisect shows this is probably related to the following changeset:

The first bad revision is:
changeset:   https://hg.mozilla.org/mozilla-central/rev/76326ec30cd6
user:        Jeff Walden
date:        Wed Apr 15 10:55:39 2015 -0700
summary:     Bug 1154532 - Add ThrowRangeError and ThrowTypeError intrinsics to make self-hosted code's behavior clearer -- and also have each assert that error number and requested error type are consistent.  (It appears no self-hosted code throws SyntaxError, ReferenceError, or URIError yet, so no adding functions for those yet.)  r=till

Not sure if this is related to bug 1154532 or to --unboxed-objects, so keeping the needinfo? for Brian until it is clear that --unboxed-objects is not the issue.

Comment 3

[Brian Hackett [Laid off!]]

I can't reproduce this but it's probably a dupe of bug 1149498.

Comment 4

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

Can no longer reproduce using m-c rev 22a157f7feb7, assuming dupe to bug 1149498.