On 05/28/2013 12:26 PM, Boris Zbarsky wrote:
On 5/27/13 2:25 PM, Josh Matthews wrote:
The following script (test_hammer_layout.html) can now run in Servo
This yields 1239493.
That's nanoseconds per iteration. As in, 1.2ms (!). This is not really
ok. ;)
Is this with builds that do the "e
On 5/27/13 9:19 PM, Robert O'Callahan wrote:
I'm pretty sure that Gecko short-circuits setting of an attribute to the
value it already has, in the DOM code.
And the spec requires that behavior, fwiw, in various ways.
-Boris
___
dev-servo mailing list
On 5/27/13 2:25 PM, Josh Matthews wrote:
The following script (test_hammer_layout.html) can now run in Servo
This yields 1239493.
That's nanoseconds per iteration. As in, 1.2ms (!). This is not really
ok. ;)
Is this with builds that do the "eagerly kick off layout from any DOM
change"
I'm pretty sure that Gecko short-circuits setting of an attribute to the
value it already has, in the DOM code. So this may not be testing what you
want.
Rob
--
q“qIqfq qyqoquq qlqoqvqeq qtqhqoqsqeq qwqhqoq qlqoqvqeq qyqoquq,q qwqhqaqtq
qcqrqeqdqiqtq qiqsq qtqhqaqtq qtqoq qyqoquq?q qEqvqeqnq qsqi
That appears to be the wrong conclusion; when I remove the code that
triggers the reflow from set_attr, the number drops by two orders of
magnitude. The layout task appears to be eating into the script task's
performance.
On 05/27/2013 04:18 PM, Patrick Walton wrote:
So I guess the problem is
So I guess the problem is that setAttribute() is slow?
Josh Matthews wrote:
>On 05/27/2013 03:45 PM, Josh Matthews wrote:
>> On 05/27/2013 02:25 PM, Josh Matthews wrote:
>>> The following script (test_hammer_layout.html) can now run in Servo,
>so
>>> I've finally been able to begin obtaining bas
On 05/27/2013 03:45 PM, Josh Matthews wrote:
On 05/27/2013 02:25 PM, Josh Matthews wrote:
The following script (test_hammer_layout.html) can now run in Servo, so
I've finally been able to begin obtaining baseline performance data.
Note that this is with the exiting Rust scheduler, not Brian's ne
On 05/27/2013 02:25 PM, Josh Matthews wrote:
The following script (test_hammer_layout.html) can now run in Servo, so
I've finally been able to begin obtaining baseline performance data.
Note that this is with the exiting Rust scheduler, not Brian's new one:
>var divs = document.getElementsByTag
8 matches
Mail list logo
