Bug 140804 - If we inline a nested function call, we should be able to do GCSE on closure variable accesses and object allocation sinking on the activation
Summary: If we inline a nested function call, we should be able to do GCSE on closure ...
Status: NEW
Alias: None
Product: WebKit
Classification: Unclassified
Component: JavaScriptCore (show other bugs)
Version: 528+ (Nightly build)
Hardware: All All
: P2 Normal
Assignee: Nobody
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2015-01-22 17:45 PST by Filip Pizlo
Modified: 2015-01-22 17:46 PST (History)
0 users

See Also:

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Filip Pizlo 2015-01-22 17:45:39 PST 
I'm pretty sure that this won't work right now because we don't reason about the activation that a NewFunction* operation references.  But we totally should do this.

This would be a poor man's way of doing single-and-loving-it analysis: http://www.cs.purdue.edu/homes/suresh/papers/popl98.ps.gz
Comment 1 Filip Pizlo 2015-01-22 17:46:04 PST 
One outcome of this is that it would allow us to get rid of more TDZ checks, since we could reason about the TDZ locally.