Side navigation
#8337 closed bug (worksforme)
Opened February 21, 2011 12:15PM UTC
Closed March 30, 2011 06:08PM UTC
Last modified March 14, 2012 11:46AM UTC
Memory leak in jQuery 1.5 with IE8
| Reported by: | James Wiseman <jwiseman@scottishlife.co.uk> | Owned by: | James Wiseman <jwiseman@scottishlife.co.uk> |
|---|---|---|---|
| Priority: | undecided | Milestone: | 1.next |
| Component: | unfiled | Version: | 1.5 |
| Keywords: | needsreview | Cc: | |
| Blocked by: | Blocking: |
Description
Yes, this again!
I have done quite a bit or reading on this, and have read a number of other tickets that have been raised (not least http://bugs.jquery.com/ticket/8171). In my opinion it warrants another ticket, but apologies if this can be resloved by other means.
The issue occurs when using the jQuery UI 1.8.9. I've rasied the issue against jQuery core as other previous issues seem to have been realted to this.
When refreshing the tabs, a memory leak is observed:
$("#tabContent").tabs("load", $("#tabContent").tabs('option', 'selected'));
I've managed to reproduce this issue in this jsfidde: http://jsfiddle.net/nxDUN/2/
To view the issue, open task manager, and repeatedly click the 'test' button. For me, the memory started climbing. On occasions it would stop. Waiting 10+ seconds and the restarting clicking showed the same issue.
Doubtful that it might be some leak in JSFiddle, I created a test suite with the same code. I've uploaded it to me site at: http://www.jameswiseman.com/blog/wp-content/uploads/2011/02/testsuite.zip
Hope this is useful. Feel free to contact me regarding this.
Attachments (0)
Change History (4)
Changed February 21, 2011 03:37PM UTC by comment:1
| owner: | → James Wiseman <jwiseman@scottishlife.co.uk> |
|---|---|
| status: | new → pending |
Changed February 22, 2011 04:01PM UTC by comment:2
| status: | pending → new |
|---|
Thanks for your quick response.
In response to the issues you raised
- The test case was really to try to isolate the bones of the problem. Not being sure whether the issue I perceived was in jQuery or jQuery UI, I opted for the most common historical cause of memory leaks. For most of the reproduction, I used to code in the test ZIP, anyway.
- Strange, because I can, very reliably. Moreover, I'm now seeing it in Crome now as well (I've switched over to 1.5.1rc1, but am not attributing it to this). Captured in the below images:
http://www.jameswiseman.com/blog/wp-content/uploads/2011/02/chromestart.jpg
http://www.jameswiseman.com/blog/wp-content/uploads/2011/02/chromend.jpg
All I am doing is frenetically clicking the 'test' button and observing a memory gain. After a short period of time the gain reaches a plateau. After another interval, the clicking brings about another memory gain. The memory doesn’t get released.
- Thanks for the IEJSLeaksDetector reference - that’s a useful resource that can be absorbed into our development community. Interestingly enough, testing on the code in the zip file in the same way also yields the same memory gain symptoms.
- JSFiddle doesn’t let you select 1.8.9. I observe the same symptoms with both the fiddle and the test zip, however.
- As mentioned above, I have switched to 1.5.1r1, and the same issue prevails.
It all seems a bit strange if you ask me. Other developers in our company working on a different app with an independent infrastructure are reporting the same issues. I’ll suggest they add to this report.
Changed February 23, 2011 12:40AM UTC by comment:3
| keywords: | → needsreview |
|---|
Changed March 30, 2011 06:08PM UTC by comment:4
| resolution: | → worksforme |
|---|---|
| status: | new → closed |
We can't repro this.
Thanks for taking the time to contribute to the jQuery project by writing a bug report and providing a test case!
There are several issues with your report:
btw. you can try http://jsfiddle.net/nxDUN/2/show to rule out the possibility of this being a jsfiddle leak.