Bug 96567 - [V8] ScriptController::compileAndRunScript() can crash
Summary: [V8] ScriptController::compileAndRunScript() can crash
Status: RESOLVED WONTFIX
Alias: None
Product: WebKit
Classification: Unclassified
Component: WebCore JavaScript (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Kentaro Hara
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2012-09-12 15:12 PDT by Kentaro Hara
Modified: 2013-09-12 22:35 PDT (History)
4 users (show)

See Also:

Attachments
Patch (2.64 KB, patch)
2012-09-12 15:21 PDT, Kentaro Hara 		no flags Details | Formatted Diff | Diff
Patch (2.52 KB, patch)
2012-09-18 17:55 PDT, Kentaro Hara 		no flags Details | Formatted Diff | Diff
patch for landing (2.51 KB, patch)
2012-09-19 17:05 PDT, Kentaro Hara 		no flags Details | Formatted Diff | Diff
Show Obsolete (1) View All 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 Kentaro Hara 2012-09-12 15:12:42 PDT 
See chromium bug: http://code.google.com/p/chromium/issues/detail?id=146776

The root cause is the following code in ScriptSourceCode.h:

  CachedScript* cachedScript() const { return m_cachedScript.get(); }

Here m_cachedScript can be 0.
Comment 1 Kentaro Hara 2012-09-12 15:21:58 PDT 
Created attachment 163715 [details]
Patch
Comment 2 Adam Barth 2012-09-12 15:40:27 PDT 
Comment on attachment 163715 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=163715&action=review

> Source/WebCore/bindings/v8/ScriptSourceCode.h:67
> -    CachedScript* cachedScript() const { return m_cachedScript.get(); }
> +    CachedScript* cachedScript() const { return m_cachedScript ? m_cachedScript.get() : 0; }

What does it mean for m_cachedScript to be 0?  It is held as a member, not as a pointer.
Comment 3 Kentaro Hara 2012-09-12 15:50:49 PDT 
Comment on attachment 163715 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=163715&action=review

>> Source/WebCore/bindings/v8/ScriptSourceCode.h:67
>> +    CachedScript* cachedScript() const { return m_cachedScript ? m_cachedScript.get() : 0; }
> 
> What does it mean for m_cachedScript to be 0?  It is held as a member, not as a pointer.

Ah, good point... maybe I was confused. More investigation is needed.
Comment 4 Kentaro Hara 2012-09-18 17:55:16 PDT 
Created attachment 164640 [details]
Patch
Comment 5 Adam Barth 2012-09-19 11:24:50 PDT 
Comment on attachment 164640 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=164640&action=review

ok

> Source/WebCore/bindings/v8/ScriptSourceCode.cpp:53
> +    if (!scriptData.get())
> +        return nullptr;

There's no reason to call get() here.  OwnPtr has a bool operator.
Comment 6 Kentaro Hara 2012-09-19 17:05:33 PDT 
Created attachment 164805 [details]
patch for landing
Comment 7 WebKit Review Bot 2012-09-19 17:41:04 PDT 
Comment on attachment 164805 [details]
patch for landing

Clearing flags on attachment: 164805

Committed r129077: <http://trac.webkit.org/changeset/129077>
Comment 8 Kentaro Hara 2012-09-19 17:41:54 PDT 
Comment on attachment 164640 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=164640&action=review

>> Source/WebCore/bindings/v8/ScriptSourceCode.cpp:53
>> +        return nullptr;
> 
> There's no reason to call get() here.  OwnPtr has a bool operator.

Fixed. Thanks.
Comment 9 Anders Carlsson 2013-09-12 22:35:41 PDT 
V8 is gone.