838148 - Add 63 missing jit-tests for various fixed JS bugs

Status: RESOLVED FIXED

(Core :: JavaScript Engine, defect)

Description

[Christian Holler (:decoder)]

Attachment: Patch

The attached patch adds 65 more jit-tests to the testsuite. These tests were automatically extracted from a list of fixed JS bugs that do have tests in the bugs but not in testsuite. The automation ensured that the bug reproduced on the original revision and also generated appropriate jit-test tags for all bugs. Only in a few cases, manual interaction was required to rewrite some tests.

This is part of a QA effort to bring down the number of fixed JS bugs that don't have tests checked in.

See also bug 831629.

Comment 1

[Terrence Cole [:terrence]]

Comment on attachment 710174 [details] [diff] [review]
Patch

Review of attachment 710174 [details] [diff] [review]:
-----------------------------------------------------------------

The ones I spot checked look good. As Luke would say: fire ze missiles!

::: js/src/jit-test/tests/auto-regress/bug487684.js
@@ +1,5 @@
> +// Binary: cache/js-dbg-32-86c8e18f20eb-linux
> +// Flags:
> +//
> +try {
> +  tracing(true);

|tracing()| isn't a shell function anymore, so I think the wrapping try is just catching the name error -- we have tons of other tests that check this case, so I think this test is probably not worth taking.

::: js/src/jit-test/tests/auto-regress/bug650618.js
@@ +2,5 @@
> +
> +// Binary: cache/js-dbg-32-242947d76f73-linux
> +// Flags:
> +//
> +print(evalcx("#1=@o"))

In case someone accidentally re-adds E4X, presumably.

Comment 2

[Jesse Ruderman]

>> +print(evalcx("#1=@o"))

> In case someone accidentally re-adds E4X, presumably.

And sharps.

Comment 3

[Christian Holler (:decoder)]

Attachment: Patch v2

I removed the two tests in question and marked in-testsuite-. Anyone stumbling over this: Feel free to remove any other unnecessary tests and mark the appropriate bugs.

I also marked one more test as being slow because win-debug try was orange due to this test timing out (win-debug always seems to be the slowest for some reason).

Keeping r+ from last patch.

Comment 4

[Christian Holler (:decoder)]

Landed:

https://hg.mozilla.org/integration/mozilla-inbound/rev/2e891e0db397

Comment 5

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/2e891e0db397