Bugfix in hydrogen array literal code generation. (issue 17334004)

1 view
Skip to first unread message

mvst...@chromium.org

unread,
Jun 19, 2013, 8:32:02 AM6/19/13
to yan...@chromium.org, v8-...@googlegroups.com
Reviewers: Yang,

Message:
Hi Yang, here is the change we discussed and a new test file. There is
actually
a test failing in mjsunit/array-literal-transitions.js that I need to solve,
though it's likely just adjusting expectations.

Description:
Bugfix in hydrogen array literal code generation.

If an array literal contains some non-constant elements, is of type SMI, and
then the boilerplate transitions to double or fast sometime after we've
crankshafted the code, then we could incorrectly store smis in double
arrays.

BUG=

Please review this at https://codereview.chromium.org/17334004/

SVN Base: https://v8.googlecode.com/svn/branches/bleeding_edge

Affected files:
M src/hydrogen.cc
A + test/mjsunit/array-literal-feedback.js


yan...@chromium.org

unread,
Jun 19, 2013, 8:35:05 AM6/19/13
to mvst...@chromium.org, v8-...@googlegroups.com
On 2013/06/19 12:32:02, mvstanton wrote:
> Hi Yang, here is the change we discussed and a new test file. There is
actually
> a test failing in mjsunit/array-literal-transitions.js that I need to
> solve,
> though it's likely just adjusting expectations.

LGTM once mjsunit/array-literal-transitions.js is fixed.

https://codereview.chromium.org/17334004/

yan...@chromium.org

unread,
Jun 19, 2013, 9:40:16 AM6/19/13
to mvst...@chromium.org, v8-...@googlegroups.com
On 2013/06/19 12:35:05, Yang wrote:
> On 2013/06/19 12:32:02, mvstanton wrote:
> > Hi Yang, here is the change we discussed and a new test file. There is
> actually
> > a test failing in mjsunit/array-literal-transitions.js that I need to
> solve,
> > though it's likely just adjusting expectations.

> LGTM once mjsunit/array-literal-transitions.js is fixed.

LGTM with a nit.

https://codereview.chromium.org/17334004/

yan...@chromium.org

unread,
Jun 19, 2013, 9:40:30 AM6/19/13
to mvst...@chromium.org, v8-...@googlegroups.com
... and here's the nit.


https://codereview.chromium.org/17334004/diff/9001/src/hydrogen.cc
File src/hydrogen.cc (right):

https://codereview.chromium.org/17334004/diff/9001/src/hydrogen.cc#newcode6068
src/hydrogen.cc:6068: // De-opt if elements kind changed from
boilerplate_elements_kind
add a period at the end of the comment.

https://codereview.chromium.org/17334004/

mvst...@chromium.org

unread,
Jun 19, 2013, 9:46:51 AM6/19/13
to yan...@chromium.org, v8-...@googlegroups.com
right on, thx, checking in..


https://codereview.chromium.org/17334004/diff/9001/src/hydrogen.cc
File src/hydrogen.cc (right):

https://codereview.chromium.org/17334004/diff/9001/src/hydrogen.cc#newcode6068
src/hydrogen.cc:6068: // De-opt if elements kind changed from
boilerplate_elements_kind
On 2013/06/19 13:40:30, Yang wrote:
> add a period at the end of the comment.

Done.

https://codereview.chromium.org/17334004/

mvst...@chromium.org

unread,
Jun 19, 2013, 9:48:57 AM6/19/13
to yan...@chromium.org, v8-...@googlegroups.com
Committed patchset #5 manually as r15207 (presubmit successful).

https://codereview.chromium.org/17334004/
Reply all
Reply to author
Forward
0 new messages