archives

« Bugzilla Issues Index

#27 — S15.1.1.1_A2_T1, S15.1.1.1_A2_T2, S15.1.1.2_A2_T1, S15.1.1.3_A2_T1, and S15.1.1.3_A2_T2 invalid (ES3 to ES5 change)

• bug_id: 27
• creation_ts: 2011-02-21 10:39:00 -0800
• short_desc: S15.1.1.1_A2_T1, S15.1.1.1_A2_T2, S15.1.1.2_A2_T1, S15.1.1.3_A2_T1, and S15.1.1.3_A2_T2 invalid (ES3 to ES5 change)
• delta_ts: 2011-09-24 16:43:50 -0700
• product: Test262
• component: ECMA-262 Tests
• version: unspecified
• rep_platform: All
• op_sys: All
• bug_status: RESOLVED
• resolution: FIXED
• priority: Normal
• bug_severity: normal
• everconfirmed: true
• reporter: Dave Fugate
• assigned_to: Mark S. Miller
• cc: ["andrebargull", "erights"]

---

• commentid: 46
• comment_count: 0
• who: Dave Fugate
• bug_when: 2011-02-21 10:39:59 -0800

These test cases assert that NaN/Infinity/etc are *not* readonly via assertions similar to:
NaN = true;
if (typeof(NaN) !== "boolean") {
$ERROR('#1: NaN = true; typeof(NaN) === "boolean". Actual: ' + (typeof(NaN)));
}

While this might have been valid in ES3, it's not in ES5. E.g., 15.1.1.1 states:
15.1.1.1 NaN
The value of NaN is NaN (see 8.5). This property has the attributes { [[Writable]]: false, [[Enumerable]]: false, [[Configurable]]: false }.
meaning that NaN can no longer be reassigned.

Disabling these tests.

---

• commentid: 450
• comment_count: 1
• who: André Bargull
• bug_when: 2011-09-11 08:48:58 -0700

S15.1.1.3_A2_T2 also calls "ERROR()" instead of "$ERROR()". Not really worth fixing this bug since the complete test is currently disabled, but I thought to mention it nonetheless...
---
undefined = true;
if (typeof(undefined) !== "boolean") {
ERROR('#1: undefined = true; typeof(undefined) === "boolean". Actual: ' + (typeof(undefined)));
}
---

---

• commentid: 464
• comment_count: 2
• who: Mark S. Miller
• bug_when: 2011-09-24 16:43:50 -0700

Fixed at http://hg.ecmascript.org/tests/test262/rev/7ee6fc6143d8