Click on the "Start the window scroll time test" button to measure performance of scrolling with the testcase with the large background image. I notice when so far is scrolled down that the background image gets repeated, then suddenly scrolling becomes much slower. This doesn't happen in Google Chrome, there scrolling stays just as fast. Firefox11beta: 35297ms Trunk time this took is: 28687ms Google Chrome: 8351ms

I can reproduce with HWA off. However, I cannot reproduce with HWA on. http://hg.mozilla.org/mozilla-central/rev/373c710112e6 Mozilla/5.0 (Windows NT 6.1; WOW64; rv:13.0) Gecko/20120222 Firefox/13.0a1 ID:20120222031223 And This is reproducible since Firefox3.5. Regression #1: Scroll fast: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.1a2pre) Gecko/20080825031951 Shiretoko/3.1a2pre ID:20080825031951 Scroll slow all Image: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.1a2pre) Gecko/20080826034105 Shiretoko/3.1a2pre ID:20080826034105 Pushlog: http://bonsai.mozilla.org/cvsquery.cgi?treeid=default&module=all&branch=HEAD&branchtype=match&dir=&file=&filetype=match&who=&whotype=match&sortby=Date&hours=2&date=explicit&mindate=2008-08-25+00%3A00%3A00&maxdate=2008-08-26+03%3A41%3A05&cvsroot=%2Fcvsroot Regression #2: Scroll slow all Image: http://hg.mozilla.org/mozilla-central/rev/4f1383af8f10 Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.1b2pre) Gecko/20081101 Shiretoko/3.1b2pre ID:20081101032525 Scroll slow except first Image: http://hg.mozilla.org/mozilla-central/rev/30a69fb5df8e Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.1b2pre) Gecko/20081102 Shiretoko/3.1b2pre ID:20081102034109 Pushlog: http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=4f1383af8f10&tochange=30a69fb5df8e

Sorry,regression #2 in comment#1 is wrong. Regression #2: Scroll slow all Image: http://hg.mozilla.org/mozilla-central/rev/fbae114d6133 Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.1b2pre) Gecko/20081104 Shiretoko/3.1b2pre ID:20081104040308 Scroll slow except first Image: http://hg.mozilla.org/mozilla-central/rev/dcec193ba5d7 Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.1b2pre) Gecko/20081105 Shiretoko/3.1b2pre ID:20081105032350 Pushlog: http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=fbae114d6133&tochange=dcec193ba5d7

Thanks! I guess this is a regression from bug 458487 then. I noticed in about:support that I have layers.acceleration.disabled=true on this computer (Toshiba Satellite L505D-S5983)

Mozilla/5.0 (Windows NT 6.1; WOW64; rv:16.0) Gecko/16.0 Firefox/16.0a1 ID:20120617030532: time this took is: 10383ms Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.1 (KHTML, like Gecko) Chrome/21.0.1177.0 Safari/537.1: time this took is: 8377ms Seems to be independent on HWA on/off for me. Cleopatra Profile: http://people.mozilla.com/~bgirard/cleopatra/?report=AMIfv94_pr6vaoTiPrDbKcmCFairKu0BxnPw7827xbE-hMVleUQwgQ_3b9qU7AcBs4PKnFbr0IAQ2uQaG-pLK7pRM7iTttC5ADOD9B1NvMfbgF2E7mCT11HNAL8Iy7e7hKU0doQpUjH6ReBiZ7fqesxWni2Rc0q-Aw (against http://hg.mozilla.org/projects/profiling/rev/4f05b58b9925)

I got 9s on Nightly, and 8s on Chrome 31 and IE 11. So it seems it's very close, and a lot better than in the past.