Bug 106597 - We should be able to checkpoint and restore the HTMLTokenizer across threads
Summary: We should be able to checkpoint and restore the HTMLTokenizer across threads
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: New Bugs (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Adam Barth
URL:
Keywords:
Depends on:
Blocks: 106127
  Show dependency tree / graph
 
Reported: 2013-01-10 13:48 PST by Adam Barth
Modified: 2013-01-11 14:21 PST (History)
6 users (show)

See Also:

Attachments
Experiment---will need later (4.82 KB, patch)
2013-01-10 13:49 PST, Adam Barth 		no flags Details | Formatted Diff | Diff
Experiment---will need later (4.86 KB, patch)
2013-01-10 13:57 PST, Adam Barth 		no flags Details | Formatted Diff | Diff
Patch (8.99 KB, patch)
2013-01-10 17:41 PST, Tony Gentilcore 		no flags Details | Formatted Diff | Diff
Patch (6.92 KB, patch)
2013-01-11 10:31 PST, Tony Gentilcore 		no flags Details | Formatted Diff | Diff
Patch (6.93 KB, patch)
2013-01-11 10:44 PST, Tony Gentilcore 		no flags Details | Formatted Diff | Diff
Show Obsolete (4) 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 Adam Barth 2013-01-10 13:48:42 PST 
We should be able to checkpoint and restore the HTMLTokenizer across threads
Comment 1 Adam Barth 2013-01-10 13:49:36 PST 
Created attachment 182199 [details]
Experiment---will need later
Comment 2 Adam Barth 2013-01-10 13:57:01 PST 
Created attachment 182200 [details]
Experiment---will need later
Comment 3 Tony Gentilcore 2013-01-10 17:41:13 PST 
Created attachment 182231 [details]
Patch
Comment 4 Adam Barth 2013-01-10 17:48:56 PST 
Comment on attachment 182231 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=182231&action=review

> Source/WebCore/html/parser/HTMLTokenizer.cpp:143
> +    if (!m_appropriateEndTagName.isEmpty()) // Asserts because m_appropriateEndTagName is never cleared anywhere.

We probably just need to clear it.  We originally were very cheap about not clearing buffers in the tokenizer.
Comment 5 Tony Gentilcore 2013-01-11 10:31:49 PST 
Created attachment 182364 [details]
Patch
Comment 6 Build Bot 2013-01-11 10:38:05 PST 
Comment on attachment 182364 [details]
Patch

Attachment 182364 [details] did not pass mac-ews (mac):
Output: http://queues.webkit.org/results/15804436
Comment 7 Tony Gentilcore 2013-01-11 10:44:31 PST 
Created attachment 182369 [details]
Patch
Comment 8 Adam Barth 2013-01-11 10:54:23 PST 
Comment on attachment 182369 [details]
Patch

LGTM.  Perhaps we should have Eric review the patch since we collaborated on writing it?
Comment 9 Adam Barth 2013-01-11 14:16:33 PST 
Comment on attachment 182369 [details]
Patch

Eric seems to be absorbed in threading code, so I'll review this patch.
Comment 10 WebKit Review Bot 2013-01-11 14:21:37 PST 
Comment on attachment 182369 [details]
Patch

Clearing flags on attachment: 182369

Committed r139497: <http://trac.webkit.org/changeset/139497>
Comment 11 WebKit Review Bot 2013-01-11 14:21:41 PST 
All reviewed patches have been landed.  Closing bug.