100833 2012-10-31 02:22:39 -0700 [chromium] Two touchadjustment tests are failing on mac 2012-11-06 10:40:56 -0800 1 1 1 Unclassified WebKit Tools / Tests 528+ (Nightly build) Unspecified Unspecified RESOLVED FIXED P2 Normal --- 1 podivilov rbyers allan.jensen kevers tonikitoo webkit.review.bot oldest_to_newest 755072 0 podivilov 2012-10-31 02:22:39 -0700 touchadjustment/html-label.html and touchadjustment/nested-touch.html started to fail on mac after r132929. FAIL adjusted node should be A#mylink. Was LABEL#mylabel. 755235 1 rbyers 2012-10-31 06:36:24 -0700 Shoot, not sure how that happened - I'll look into it ASAP 755341 2 rbyers 2012-10-31 08:46:43 -0700 Just to clarify, this is chromium mac. touchadjustment tests aren't run at all for the mac port of WebKit. I see the failures, eg. here: http://build.webkit.org/results/Chromium%20Mac%20Release%20%28Tests%29/r132992%20%2827175%29/touchadjustment/html-label-diff.txt The revision range here is huge though (r132751 to r132992) - just out of curiosity (not that I'm really doubting my change is what broke this), do we have more fine grained test results somewhere? Why do we have a gap of >240 commits? 755431 3 podivilov 2012-10-31 10:21:34 -0700 (In reply to comment #2) > Just to clarify, this is chromium mac. touchadjustment tests aren't run at all for the mac port of WebKit. Yes, it's chromium, see the bug title. > > I see the failures, eg. here: http://build.webkit.org/results/Chromium%20Mac%20Release%20%28Tests%29/r132992%20%2827175%29/touchadjustment/html-label-diff.txt > > The revision range here is huge though (r132751 to r132992) - just out of curiosity (not that I'm really doubting my change is what broke this), do we have more fine grained test results somewhere? Why do we have a gap of >240 commits? The best way to track layout test failures is to use our flakiness dashboard: http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=touchadjustment%2Fhtml-label.html%2Ctouchadjustment%2Fnested-touch.html. It shows results from chrome.webkit canary bots (http://build.chromium.org/p/chromium.webkit/console). Click on the rightmost red bar for e.g. mac 10.6 to get the WebKit blamelist for the failure (http://trac.webkit.org/log/?verbose=on&rev=132930&stop_rev=132929). Not sure about webkit.org chromium bots, me guess is most people just ignore them. 755453 4 rbyers 2012-10-31 10:45:34 -0700 (In reply to comment #3) > (In reply to comment #2) > > Just to clarify, this is chromium mac. touchadjustment tests aren't run at all for the mac port of WebKit. > > Yes, it's chromium, see the bug title. Yeah, I just added [chromium] (wasn't present initially). > > > > I see the failures, eg. here: http://build.webkit.org/results/Chromium%20Mac%20Release%20%28Tests%29/r132992%20%2827175%29/touchadjustment/html-label-diff.txt > > > > The revision range here is huge though (r132751 to r132992) - just out of curiosity (not that I'm really doubting my change is what broke this), do we have more fine grained test results somewhere? Why do we have a gap of >240 commits? > > The best way to track layout test failures is to use our flakiness dashboard: > http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=touchadjustment%2Fhtml-label.html%2Ctouchadjustment%2Fnested-touch.html. > > It shows results from chrome.webkit canary bots (http://build.chromium.org/p/chromium.webkit/console). > > Click on the rightmost red bar for e.g. mac 10.6 to get the WebKit blamelist for the failure (http://trac.webkit.org/log/?verbose=on&rev=132930&stop_rev=132929). > > Not sure about webkit.org chromium bots, me guess is most people just ignore them. (In reply to comment #3) > (In reply to comment #2) > > Just to clarify, this is chromium mac. touchadjustment tests aren't run at all for the mac port of WebKit. > > Yes, it's chromium, see the bug title. > > > > > I see the failures, eg. here: http://build.webkit.org/results/Chromium%20Mac%20Release%20%28Tests%29/r132992%20%2827175%29/touchadjustment/html-label-diff.txt > > > > The revision range here is huge though (r132751 to r132992) - just out of curiosity (not that I'm really doubting my change is what broke this), do we have more fine grained test results somewhere? Why do we have a gap of >240 commits? > > The best way to track layout test failures is to use our flakiness dashboard: > http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=touchadjustment%2Fhtml-label.html%2Ctouchadjustment%2Fnested-touch.html. > > It shows results from chrome.webkit canary bots (http://build.chromium.org/p/chromium.webkit/console). > > Click on the rightmost red bar for e.g. mac 10.6 to get the WebKit blamelist for the failure (http://trac.webkit.org/log/?verbose=on&rev=132930&stop_rev=132929). > > Not sure about webkit.org chromium bots, me guess is most people just ignore them. Perfect, thank you for the details! 757523 5 rbyers 2012-11-02 13:22:20 -0700 As far as I can tell these tests are legitimately failing due to a problem with touch adjustment. The failures we're previously masked - before my fix the code relied on clientWidth/clientHeight which are always 0 for these inline elements. So rather than touch the center of the element as desired, we were always touching the top-left corner. Now that we're actually touching the center, we're much closer to the node below and apparently the touch adjustment score is high enough that we're adjusting to it. I don't know the touch adjustment algorithm well enough myself to say whether or not it's reasonable that we're adjusting to this other node. In this case it's the smaller node and so we have a higher percentage overlap. As for why we're getting different results on Mac and Linux: I suspect this is due to the differences in text metrics. In particular, for html-label on Mac the A#mylink node is 18px tall, with 2px space below it before the top of the LABEL#mylabel node. On Linux it's 19px tall with 4px space. This means that on Mac the distance from the touch to the mylabel element is 2.5 pixels shorter, giving it a higher touch adjustment score. 757951 6 rbyers 2012-11-03 06:13:30 -0700 With the fix to bug 101046 the behavior of html-label is now different. The direct hit is succeeding (which is an improvement), but the indirect hit on the same element is now failing. It looks to me like we're now preferring the larger link above due to the greater overlap. This seems reasonable to me since the label is nice and wide so the potential overlap is high. So now that bug 101046 is fixed, the right thing to do is just to adjust the test. 757952 7 rbyers 2012-11-03 06:25:03 -0700 As for the nested-touch test, this started failing with my change because using getBoundingClientRect to get the size of the div included the border (1px on all edges), where the previous approach (clientWidth/clientHeight) didn't. So what was already a very borderline case (touching 10 pixels below the box) got pushed over the edge by being slightly lower. Again, the right change is to adjust the test to compensate. 757953 8 172221 rbyers 2012-11-03 06:41:17 -0700 Created attachment 172221 Patch 757955 9 allan.jensen 2012-11-03 07:00:07 -0700 (In reply to comment #5) > As for why we're getting different results on Mac and Linux: I suspect this is due to the differences in text metrics. In particular, for html-label on Mac the A#mylink node is 18px tall, with 2px space below it before the top of the LABEL#mylabel node. On Linux it's 19px tall with 4px space. This means that on Mac the distance from the touch to the mylabel element is 2.5 pixels shorter, giving it a higher touch adjustment score. Perhaps a more robust solution would be changing the font to Ahem, so the tests are always the same between platforms? I did that to a few of the other touch adjustment tests to get around similar problems. 758113 10 172244 rbyers 2012-11-04 04:41:06 -0800 Created attachment 172244 Patch 758114 11 rbyers 2012-11-04 04:51:47 -0800 (In reply to comment #9) > (In reply to comment #5) > > As for why we're getting different results on Mac and Linux: I suspect this is due to the differences in text metrics. In particular, for html-label on Mac the A#mylink node is 18px tall, with 2px space below it before the top of the LABEL#mylabel node. On Linux it's 19px tall with 4px space. This means that on Mac the distance from the touch to the mylabel element is 2.5 pixels shorter, giving it a higher touch adjustment score. > > Perhaps a more robust solution would be changing the font to Ahem, so the tests are always the same between platforms? I did that to a few of the other touch adjustment tests to get around similar problems. Good call, I've done that now. The new metrics made the targets even harder to hit (eg. the html-label indirect label case went from working at up to 4px outside the node to working at up to only 1px outside. I saw that you are using line-height: 20px in some of the other tests. I initially did that too (it would work up to 8px outside the node), but I decided that was a bad idea. Our biggest problem with touch adjustment in chrome has been when something that would have worked fine without adjustment no longer works (like the plugin boundary case kevers recently fixed). So top priority for me is that we have good coverage of the direct hit cases in realistic scenarios. So I'd rather not add artificial padding between the lines - reducing the coverage of the direct-hit tests. The fact that touching just outside this label will fuzz up to the link when it's more than 1px to the right of the label seems fine to me. So I'll just test that (the key thing is that it's not fuzzing up until the center point is off of the label). 758115 12 172245 rbyers 2012-11-04 04:55:35 -0800 Created attachment 172245 Patch 758116 13 allan.jensen 2012-11-04 04:58:21 -0800 (In reply to comment #11) > (In reply to comment #9) > > (In reply to comment #5) > > > As for why we're getting different results on Mac and Linux: I suspect this is due to the differences in text metrics. In particular, for html-label on Mac the A#mylink node is 18px tall, with 2px space below it before the top of the LABEL#mylabel node. On Linux it's 19px tall with 4px space. This means that on Mac the distance from the touch to the mylabel element is 2.5 pixels shorter, giving it a higher touch adjustment score. > > > > Perhaps a more robust solution would be changing the font to Ahem, so the tests are always the same between platforms? I did that to a few of the other touch adjustment tests to get around similar problems. > > Good call, I've done that now. The new metrics made the targets even harder to hit (eg. the html-label indirect label case went from working at up to 4px outside the node to working at up to only 1px outside. > > I saw that you are using line-height: 20px in some of the other tests. I initially did that too (it would work up to 8px outside the node), but I decided that was a bad idea. Our biggest problem with touch adjustment in chrome has been when something that would have worked fine without adjustment no longer works (like the plugin boundary case kevers recently fixed). So top priority for me is that we have good coverage of the direct hit cases in realistic scenarios. Most of the tests have been made to test specific new improvements or discovered bugs, we might actually be missing good test coverage of realistic 'standard' cases. 758119 14 172245 allan.jensen 2012-11-04 05:08:27 -0800 Comment on attachment 172245 Patch Looks good to me (as a non-reviewer). 758494 15 rbyers 2012-11-05 07:01:12 -0800 (In reply to comment #13) > (In reply to comment #11) > > (In reply to comment #9) > > > (In reply to comment #5) > > > > As for why we're getting different results on Mac and Linux: I suspect this is due to the differences in text metrics. In particular, for html-label on Mac the A#mylink node is 18px tall, with 2px space below it before the top of the LABEL#mylabel node. On Linux it's 19px tall with 4px space. This means that on Mac the distance from the touch to the mylabel element is 2.5 pixels shorter, giving it a higher touch adjustment score. > > > > > > Perhaps a more robust solution would be changing the font to Ahem, so the tests are always the same between platforms? I did that to a few of the other touch adjustment tests to get around similar problems. > > > > Good call, I've done that now. The new metrics made the targets even harder to hit (eg. the html-label indirect label case went from working at up to 4px outside the node to working at up to only 1px outside. > > > > I saw that you are using line-height: 20px in some of the other tests. I initially did that too (it would work up to 8px outside the node), but I decided that was a bad idea. Our biggest problem with touch adjustment in chrome has been when something that would have worked fine without adjustment no longer works (like the plugin boundary case kevers recently fixed). So top priority for me is that we have good coverage of the direct hit cases in realistic scenarios. > > Most of the tests have been made to test specific new improvements or discovered bugs, we might actually be missing good test coverage of realistic 'standard' cases. Makes sense. Kevers has been talking about maybe doing some work here, and I'd also love to see us use our metrics system in chrome to track the effectiveness of touch adjustment in the wild (eg. how often users appear to refine their touch by touching a second time). I'm not sure what the right cost-benefit trade off is though - it works quite well in practice most of the time :-) 758495 16 rbyers 2012-11-05 07:01:36 -0800 (In reply to comment #14) > (From update of attachment 172245 [details]) > Looks good to me (as a non-reviewer). Thanks! Antonio? 759848 17 172245 webkit.review.bot 2012-11-06 10:40:52 -0800 Comment on attachment 172245 Patch Clearing flags on attachment: 172245 Committed r133636: <http://trac.webkit.org/changeset/133636> 759850 18 webkit.review.bot 2012-11-06 10:40:56 -0800 All reviewed patches have been landed. Closing bug. 172221 2012-11-03 06:41:17 -0700 2012-11-04 04:41:02 -0800 Patch bug-100833-20121103093940.patch text/plain 3834 rbyers U3VidmVyc2lvbiBSZXZpc2lvbjogMTMzMzM0CmRpZmYgLS1naXQgYS9MYXlvdXRUZXN0cy9DaGFu Z2VMb2cgYi9MYXlvdXRUZXN0cy9DaGFuZ2VMb2cKaW5kZXggNzhhNTg0ZTI2M2I3NWE2MWZhNzJh MWNmODNjNWYzYjU4MWE5OTBiNS4uZWIyZmNkYjlkMWVmZmM5MGI4NmYzZTExZDNjZjU3YTM4Y2Yy MmZhYiAxMDA2NDQKLS0tIGEvTGF5b3V0VGVzdHMvQ2hhbmdlTG9nCisrKyBiL0xheW91dFRlc3Rz L0NoYW5nZUxvZwpAQCAtMSwzICsxLDE4IEBACisyMDEyLTExLTAzICBSaWNrIEJ5ZXJzICA8cmJ5 ZXJzQGNocm9taXVtLm9yZz4KKworICAgICAgICBbY2hyb21pdW1dIFR3byB0b3VjaGFkanVzdG1l bnQgdGVzdHMgYXJlIGZhaWxpbmcgb24gbWFjCisgICAgICAgIGh0dHBzOi8vYnVncy53ZWJraXQu b3JnL3Nob3dfYnVnLmNnaT9pZD0xMDA4MzMKKworICAgICAgICBSZXZpZXdlZCBieSBOT0JPRFkg KE9PUFMhKS4KKworICAgICAgICBSZS1lbmFibGUgdGhlc2UgdHdvIHRlc3RzIHdpdGggYXBwcm9w cmlhdGUgdHdlYWtzLgorCisgICAgICAgICogcGxhdGZvcm0vY2hyb21pdW0vVGVzdEV4cGVjdGF0 aW9uczogUmVtb3ZlIHRoZSBmYWlsdXJlIGV4cGVjdGF0aW9ucy4KKyAgICAgICAgKiB0b3VjaGFk anVzdG1lbnQvaHRtbC1sYWJlbC5odG1sOiBUd2VhayBvbmUgdmVyeSBib3JkZXJsaW5lIGNhc2Ug dG8gYmUgY2xvc2VyIHRvIHBhc3MKKyAgICAgICAgcmVsaWFibHkuICBQcmV2aW91c2x5IHRoaXMg cGFzc2VkIG9ubHkgZHVlIHRvIGEgYnVnIGluIHRoZSB0ZXN0IGluZnJhc3RydWN0dXJlIGZpeGVk IGluIHIxMzI5MjkuCisgICAgICAgICogdG91Y2hhZGp1c3RtZW50L25lc3RlZC10b3VjaC5odG1s OiBUd2VhayBib3JkZXJsaW5lIGNhc2UgdG8gY29tcGVuc2F0ZSBmb3IgaW5jbHVzaW9uIG9mCisg ICAgICAgIHRoZSBib3JkZXIgaW4gdGhlIG1lYXN1cmVtZW50IG9mIHRoZSBzaXplIG9mIHRoZSBk aXYuCisKIDIwMTItMTEtMDIgIElhbiBWb2xsaWNrICA8dm9sbGlja0BjaHJvbWl1bS5vcmc+CiAK ICAgICAgICAgU3VwcG9ydCBpbnZhbGlkYXRpb24gdHJhY2tpbmcgZm9yIGNvbXBvc2l0ZWQgbGF5 ZXJzCmRpZmYgLS1naXQgYS9MYXlvdXRUZXN0cy9wbGF0Zm9ybS9jaHJvbWl1bS9UZXN0RXhwZWN0 YXRpb25zIGIvTGF5b3V0VGVzdHMvcGxhdGZvcm0vY2hyb21pdW0vVGVzdEV4cGVjdGF0aW9ucwpp bmRleCA3ZDUxMDU3NWRlNWNjZjFjZWI4MTAzODQ2Y2JmMDEwMTFmZDZkMDJmLi5jM2QxYjJkNDI0 ZTg2MDc5MTNkMjM3M2FlZGY3NDBiYTk3ZmE4MWI5IDEwMDY0NAotLS0gYS9MYXlvdXRUZXN0cy9w bGF0Zm9ybS9jaHJvbWl1bS9UZXN0RXhwZWN0YXRpb25zCisrKyBiL0xheW91dFRlc3RzL3BsYXRm b3JtL2Nocm9taXVtL1Rlc3RFeHBlY3RhdGlvbnMKQEAgLTQxNjIsOCArNDE2Miw1IEBAIHdlYmtp dC5vcmcvYi8xMDA4MDYgWyBXaW4gXSBmYXN0L2Nzcy9mb250LWZhY2UtZG93bmxvYWQtZXJyb3Iu aHRtbCBbIFRleHQgXQogIyBTZWVtcyB0byBoYXZlIHRoZSB3cm9uZyBmb250IGluIHRoZSBTVkcg cGFydCBvZiB0aGUgdGVzdCAoQ1NTIGlzIG9rKS4KIHdlYmtpdC5vcmcvYi8xMDA4MDYgWyBXaW4g XSBzdmcvY3VzdG9tL2ZvbnQtZmFjZS1zaW1wbGUuc3ZnIFsgRmFpbHVyZSBdCiAKLXdlYmtpdC5v cmcvYi8xMDA4MzMgWyBNYWMgXSB0b3VjaGFkanVzdG1lbnQvaHRtbC1sYWJlbC5odG1sIFsgVGV4 dCBdCi13ZWJraXQub3JnL2IvMTAwODMzIFsgTWFjIF0gdG91Y2hhZGp1c3RtZW50L25lc3RlZC10 b3VjaC5odG1sIFsgVGV4dCBdCi0KICMgVGhpcyB0ZXN0IGlzIGNvbnNpc3RlbnRseSBsZWFraW5n IHN0YXRlIGludG8gdGhlIG5leHQgdGVzdCAob2JqZWN0LXNyYy11cmwtYWxsb3dlZC5odG1sKSBh ZnRlciByMTMzMDY5Lgogd2Via2l0Lm9yZy9iLzEwMDk1NSBodHRwL3Rlc3RzL3NlY3VyaXR5L2Nv bnRlbnRTZWN1cml0eVBvbGljeS9vYmplY3Qtc3JjLW5vbmUtYmxvY2tlZC5odG1sIFsgU2tpcCBd CmRpZmYgLS1naXQgYS9MYXlvdXRUZXN0cy90b3VjaGFkanVzdG1lbnQvaHRtbC1sYWJlbC5odG1s IGIvTGF5b3V0VGVzdHMvdG91Y2hhZGp1c3RtZW50L2h0bWwtbGFiZWwuaHRtbAppbmRleCAzYzEx NmE1NWJhYTBkNWViNDNmMzUxMDExNDZhZjlkZWM4MzkyZDljLi43ZGUxZGM2ZThkMDQ1YzdiOWU2 N2Q4NDk3ZDE1OTdhMmJlMzFlMDNlIDEwMDY0NAotLS0gYS9MYXlvdXRUZXN0cy90b3VjaGFkanVz dG1lbnQvaHRtbC1sYWJlbC5odG1sCisrKyBiL0xheW91dFRlc3RzL3RvdWNoYWRqdXN0bWVudC9o dG1sLWxhYmVsLmh0bWwKQEAgLTQyLDEwICs0MiwxMCBAQAogICAgICAgICB0ZXN0VG91Y2hQb2lu dCh0b3VjaHBvaW50LCBlbGVtZW50KTsKICAgICB9CiAKLSAgICBmdW5jdGlvbiB0ZXN0SW5kaXJl Y3RUb3VjaChlbGVtZW50KQorICAgIGZ1bmN0aW9uIHRlc3RJbmRpcmVjdFRvdWNoKGVsZW1lbnQs IG9mZnNldCkKICAgICB7CiAgICAgICAgIC8vIFRvdWNoIGp1c3QgcmlnaHQgb2YgdGhlIGVsZW1l bnQuCi0gICAgICAgIHZhciB0b3VjaHBvaW50ID0gb2Zmc2V0VG91Y2hQb2ludChmaW5kQWJzb2x1 dGVCb3VuZHMoZWxlbWVudCksICdyaWdodCcsIDEwLCAzMCwgMjApOworICAgICAgICB2YXIgdG91 Y2hwb2ludCA9IG9mZnNldFRvdWNoUG9pbnQoZmluZEFic29sdXRlQm91bmRzKGVsZW1lbnQpLCAn cmlnaHQnLCBvZmZzZXQsIDMwLCAyMCk7CiAgICAgICAgIHRlc3RUb3VjaFBvaW50KHRvdWNocG9p bnQsIGVsZW1lbnQpOwogICAgIH0KIApAQCAtNjIsOSArNjIsOSBAQAogICAgIHsKICAgICAgICAg ZGVidWcoJ1Rlc3RpbmcgaW5kaXJlY3QgaGl0cy4nKTsKIAotICAgICAgICB0ZXN0SW5kaXJlY3RU b3VjaChlLm15bGluayk7Ci0gICAgICAgIHRlc3RJbmRpcmVjdFRvdWNoKGUubXlsYWJlbCk7Ci0g ICAgICAgIHRlc3RJbmRpcmVjdFRvdWNoKGUubXlpbnB1dCk7CisgICAgICAgIHRlc3RJbmRpcmVj dFRvdWNoKGUubXlsaW5rLCAxMCk7CisgICAgICAgIHRlc3RJbmRpcmVjdFRvdWNoKGUubXlsYWJl bCwgNCk7CisgICAgICAgIHRlc3RJbmRpcmVjdFRvdWNoKGUubXlpbnB1dCwgMTApOwogICAgIH0K IAogICAgIGZ1bmN0aW9uIHJ1blRlc3RzKCkKZGlmZiAtLWdpdCBhL0xheW91dFRlc3RzL3RvdWNo YWRqdXN0bWVudC9uZXN0ZWQtdG91Y2guaHRtbCBiL0xheW91dFRlc3RzL3RvdWNoYWRqdXN0bWVu dC9uZXN0ZWQtdG91Y2guaHRtbAppbmRleCAyNjRlODg3YTY5ZWJmYmEwNzQyYWEyYWFiZTNiMzFh ZmMzZjZlMTU1Li5lYjY0Mzc3OGViN2VlMTQzMzhlMjNlODk1ODg2OWVhYWE4ZDQyOTMyIDEwMDY0 NAotLS0gYS9MYXlvdXRUZXN0cy90b3VjaGFkanVzdG1lbnQvbmVzdGVkLXRvdWNoLmh0bWwKKysr IGIvTGF5b3V0VGVzdHMvdG91Y2hhZGp1c3RtZW50L25lc3RlZC10b3VjaC5odG1sCkBAIC03Niw3 ICs3Niw3IEBACiAgICAgICAgIC8vIEZpcnN0IHRlc3Qgd2FzIGNlbnRlcmVkLCBub3cgbW92ZSB0 aGUgdGVzdCBjbG9zZXIgdG8gdGhlIHdyb25nIG5vZGUsIGFuZCBlbnN1cmUgd2Ugc3RpbGwgZ2V0 IHRoZSBwcmVmZXJlZCBub2RlLgogICAgICAgICB0ZXN0VG91Y2hQb2ludCh0b3VjaFBvaW50RG91 YmxlVG91Y2goZS5teWJveDEsIGUubXlib3gyLCA1KSwgZS5teWJveDEpOwogCi0gICAgICAgIHRl c3RUb3VjaFBvaW50KHRvdWNoUG9pbnREb3VibGVUb3VjaChlLm15Ym94MSwgZS5teWJveDIsIDEw KSwgZS5teWJveDEpOworICAgICAgICB0ZXN0VG91Y2hQb2ludCh0b3VjaFBvaW50RG91YmxlVG91 Y2goZS5teWJveDEsIGUubXlib3gyLCA5KSwgZS5teWJveDEpOwogICAgIH0KIAogICAgIGZ1bmN0 aW9uIHJ1blRlc3RzKCkK 172244 2012-11-04 04:41:06 -0800 2012-11-04 04:55:31 -0800 Patch bug-100833-20121104073931.patch text/plain 4444 rbyers U3VidmVyc2lvbiBSZXZpc2lvbjogMTMzMzM0CmRpZmYgLS1naXQgYS9MYXlvdXRUZXN0cy9DaGFu Z2VMb2cgYi9MYXlvdXRUZXN0cy9DaGFuZ2VMb2cKaW5kZXggNzhhNTg0ZTI2M2I3NWE2MWZhNzJh MWNmODNjNWYzYjU4MWE5OTBiNS4uYTFjNjc3OGNlMzM0MjFjMTI2MzFmNTBkOWMzODA0OGQ0NTZh YWU0NCAxMDA2NDQKLS0tIGEvTGF5b3V0VGVzdHMvQ2hhbmdlTG9nCisrKyBiL0xheW91dFRlc3Rz L0NoYW5nZUxvZwpAQCAtMSwzICsxLDE5IEBACisyMDEyLTExLTAzICBSaWNrIEJ5ZXJzICA8cmJ5 ZXJzQGNocm9taXVtLm9yZz4KKworICAgICAgICBbY2hyb21pdW1dIFR3byB0b3VjaGFkanVzdG1l bnQgdGVzdHMgYXJlIGZhaWxpbmcgb24gbWFjCisgICAgICAgIGh0dHBzOi8vYnVncy53ZWJraXQu b3JnL3Nob3dfYnVnLmNnaT9pZD0xMDA4MzMKKworICAgICAgICBSZXZpZXdlZCBieSBOT0JPRFkg KE9PUFMhKS4KKworICAgICAgICBSZS1lbmFibGUgdGhlc2UgdHdvIHRlc3RzIHdpdGggYXBwcm9w cmlhdGUgdHdlYWtzLiAgQWxzbyBhdm9pZCB0aGUgcGxhdGZvcm0gZGVwZW5kYW50CisgICAgICAg IGJlaGF2aW9yIGJ5IHN3aXRjaGluZyB0byB0aGUgQWhlbSBmb250LgorCisgICAgICAgICogcGxh dGZvcm0vY2hyb21pdW0vVGVzdEV4cGVjdGF0aW9uczogUmVtb3ZlIHRoZSBmYWlsdXJlIGV4cGVj dGF0aW9ucy4KKyAgICAgICAgKiB0b3VjaGFkanVzdG1lbnQvaHRtbC1sYWJlbC5odG1sOiBUd2Vh ayBvbmUgdmVyeSBib3JkZXJsaW5lIGNhc2UgdG8gYmUgY2xvc2VyIHRvIHBhc3MKKyAgICAgICAg cmVsaWFibHkuICBQcmV2aW91c2x5IHRoaXMgcGFzc2VkIG9ubHkgZHVlIHRvIGEgYnVnIGluIHRo ZSB0ZXN0IGluZnJhc3RydWN0dXJlIGZpeGVkIGluIHIxMzI5MjkuCisgICAgICAgICogdG91Y2hh ZGp1c3RtZW50L25lc3RlZC10b3VjaC5odG1sOiBUd2VhayBib3JkZXJsaW5lIGNhc2UgdG8gY29t cGVuc2F0ZSBmb3IgaW5jbHVzaW9uIG9mCisgICAgICAgIHRoZSBib3JkZXIgaW4gdGhlIG1lYXN1 cmVtZW50IG9mIHRoZSBzaXplIG9mIHRoZSBkaXYuCisKIDIwMTItMTEtMDIgIElhbiBWb2xsaWNr ICA8dm9sbGlja0BjaHJvbWl1bS5vcmc+CiAKICAgICAgICAgU3VwcG9ydCBpbnZhbGlkYXRpb24g dHJhY2tpbmcgZm9yIGNvbXBvc2l0ZWQgbGF5ZXJzCmRpZmYgLS1naXQgYS9MYXlvdXRUZXN0cy9w bGF0Zm9ybS9jaHJvbWl1bS9UZXN0RXhwZWN0YXRpb25zIGIvTGF5b3V0VGVzdHMvcGxhdGZvcm0v Y2hyb21pdW0vVGVzdEV4cGVjdGF0aW9ucwppbmRleCA3ZDUxMDU3NWRlNWNjZjFjZWI4MTAzODQ2 Y2JmMDEwMTFmZDZkMDJmLi5jM2QxYjJkNDI0ZTg2MDc5MTNkMjM3M2FlZGY3NDBiYTk3ZmE4MWI5 IDEwMDY0NAotLS0gYS9MYXlvdXRUZXN0cy9wbGF0Zm9ybS9jaHJvbWl1bS9UZXN0RXhwZWN0YXRp b25zCisrKyBiL0xheW91dFRlc3RzL3BsYXRmb3JtL2Nocm9taXVtL1Rlc3RFeHBlY3RhdGlvbnMK QEAgLTQxNjIsOCArNDE2Miw1IEBAIHdlYmtpdC5vcmcvYi8xMDA4MDYgWyBXaW4gXSBmYXN0L2Nz cy9mb250LWZhY2UtZG93bmxvYWQtZXJyb3IuaHRtbCBbIFRleHQgXQogIyBTZWVtcyB0byBoYXZl IHRoZSB3cm9uZyBmb250IGluIHRoZSBTVkcgcGFydCBvZiB0aGUgdGVzdCAoQ1NTIGlzIG9rKS4K IHdlYmtpdC5vcmcvYi8xMDA4MDYgWyBXaW4gXSBzdmcvY3VzdG9tL2ZvbnQtZmFjZS1zaW1wbGUu c3ZnIFsgRmFpbHVyZSBdCiAKLXdlYmtpdC5vcmcvYi8xMDA4MzMgWyBNYWMgXSB0b3VjaGFkanVz dG1lbnQvaHRtbC1sYWJlbC5odG1sIFsgVGV4dCBdCi13ZWJraXQub3JnL2IvMTAwODMzIFsgTWFj IF0gdG91Y2hhZGp1c3RtZW50L25lc3RlZC10b3VjaC5odG1sIFsgVGV4dCBdCi0KICMgVGhpcyB0 ZXN0IGlzIGNvbnNpc3RlbnRseSBsZWFraW5nIHN0YXRlIGludG8gdGhlIG5leHQgdGVzdCAob2Jq ZWN0LXNyYy11cmwtYWxsb3dlZC5odG1sKSBhZnRlciByMTMzMDY5Lgogd2Via2l0Lm9yZy9iLzEw MDk1NSBodHRwL3Rlc3RzL3NlY3VyaXR5L2NvbnRlbnRTZWN1cml0eVBvbGljeS9vYmplY3Qtc3Jj LW5vbmUtYmxvY2tlZC5odG1sIFsgU2tpcCBdCmRpZmYgLS1naXQgYS9MYXlvdXRUZXN0cy90b3Vj aGFkanVzdG1lbnQvaHRtbC1sYWJlbC5odG1sIGIvTGF5b3V0VGVzdHMvdG91Y2hhZGp1c3RtZW50 L2h0bWwtbGFiZWwuaHRtbAppbmRleCAzYzExNmE1NWJhYTBkNWViNDNmMzUxMDExNDZhZjlkZWM4 MzkyZDljLi4yZTFiZjBiMTUzYTJmOTBkYWRiMmFlMzc1ZTI1OGU5NTQzZWQyOWZjIDEwMDY0NAot LS0gYS9MYXlvdXRUZXN0cy90b3VjaGFkanVzdG1lbnQvaHRtbC1sYWJlbC5odG1sCisrKyBiL0xh eW91dFRlc3RzL3RvdWNoYWRqdXN0bWVudC9odG1sLWxhYmVsLmh0bWwKQEAgLTksNiArOSw4IEBA CiAgICAgICAgICAgICBwb3NpdGlvbjogYWJzb2x1dGU7CiAgICAgICAgICAgICBsZWZ0OiAwcHg7 CiAgICAgICAgICAgICB0b3A6IDBweDsKKyAgICAgICAgICAgIGZvbnQ6IDE2cHggQWhlbTsKKyAg ICAgICAgICAgIGxpbmUtaGVpZ2h0OiAyMHB4OwogICAgICAgICB9CiAgICAgPC9zdHlsZT4KIDwv aGVhZD4KQEAgLTQyLDEwICs0NCwxMCBAQAogICAgICAgICB0ZXN0VG91Y2hQb2ludCh0b3VjaHBv aW50LCBlbGVtZW50KTsKICAgICB9CiAKLSAgICBmdW5jdGlvbiB0ZXN0SW5kaXJlY3RUb3VjaChl bGVtZW50KQorICAgIGZ1bmN0aW9uIHRlc3RJbmRpcmVjdFRvdWNoKGVsZW1lbnQsIG9mZnNldCkK ICAgICB7CiAgICAgICAgIC8vIFRvdWNoIGp1c3QgcmlnaHQgb2YgdGhlIGVsZW1lbnQuCi0gICAg ICAgIHZhciB0b3VjaHBvaW50ID0gb2Zmc2V0VG91Y2hQb2ludChmaW5kQWJzb2x1dGVCb3VuZHMo ZWxlbWVudCksICdyaWdodCcsIDEwLCAzMCwgMjApOworICAgICAgICB2YXIgdG91Y2hwb2ludCA9 IG9mZnNldFRvdWNoUG9pbnQoZmluZEFic29sdXRlQm91bmRzKGVsZW1lbnQpLCAncmlnaHQnLCBv ZmZzZXQsIDMwLCAyMCk7CiAgICAgICAgIHRlc3RUb3VjaFBvaW50KHRvdWNocG9pbnQsIGVsZW1l bnQpOwogICAgIH0KIApAQCAtNjIsOSArNjQsOSBAQAogICAgIHsKICAgICAgICAgZGVidWcoJ1Rl c3RpbmcgaW5kaXJlY3QgaGl0cy4nKTsKIAotICAgICAgICB0ZXN0SW5kaXJlY3RUb3VjaChlLm15 bGluayk7Ci0gICAgICAgIHRlc3RJbmRpcmVjdFRvdWNoKGUubXlsYWJlbCk7Ci0gICAgICAgIHRl c3RJbmRpcmVjdFRvdWNoKGUubXlpbnB1dCk7CisgICAgICAgIHRlc3RJbmRpcmVjdFRvdWNoKGUu bXlsaW5rLCAxMCk7CisgICAgICAgIHRlc3RJbmRpcmVjdFRvdWNoKGUubXlsYWJlbCwgOCk7Cisg ICAgICAgIHRlc3RJbmRpcmVjdFRvdWNoKGUubXlpbnB1dCwgMTApOwogICAgIH0KIAogICAgIGZ1 bmN0aW9uIHJ1blRlc3RzKCkKZGlmZiAtLWdpdCBhL0xheW91dFRlc3RzL3RvdWNoYWRqdXN0bWVu dC9uZXN0ZWQtdG91Y2guaHRtbCBiL0xheW91dFRlc3RzL3RvdWNoYWRqdXN0bWVudC9uZXN0ZWQt dG91Y2guaHRtbAppbmRleCAyNjRlODg3YTY5ZWJmYmEwNzQyYWEyYWFiZTNiMzFhZmMzZjZlMTU1 Li43ZTdiYzI2OWViOThiNzQ1NDlhNGEyNGZjZDM4ZGVjZWJiYjgwY2JjIDEwMDY0NAotLS0gYS9M YXlvdXRUZXN0cy90b3VjaGFkanVzdG1lbnQvbmVzdGVkLXRvdWNoLmh0bWwKKysrIGIvTGF5b3V0 VGVzdHMvdG91Y2hhZGp1c3RtZW50L25lc3RlZC10b3VjaC5odG1sCkBAIC01LDcgKzUsNyBAQAog ICAgIDxzY3JpcHQgc3JjPSIuLi9mYXN0L2pzL3Jlc291cmNlcy9qcy10ZXN0LXByZS5qcyI+PC9z Y3JpcHQ+CiAgICAgPHNjcmlwdCBzcmM9InJlc291cmNlcy90b3VjaGFkanVzdG1lbnQuanMiPjwv c2NyaXB0PgogICAgIDxzdHlsZT4KLSAgICAgICAgI3NhbmRib3ggeyB9CisgICAgICAgICNzYW5k Ym94IHsgZm9udDogMTZweCBBaGVtOyB9CiAgICAgICAgIC5ib3ggeyBib3JkZXI6IDFweCBzb2xp ZCBibGFjazsgYm9yZGVyLXJhZGl1czogNXB4IDVweDsgbWFyZ2luOiAxMnB4OyBtYXgtd2lkdGg6 IDQwZW07IH0KICAgICA8L3N0eWxlPgogPC9oZWFkPgpAQCAtNzYsNyArNzYsNyBAQAogICAgICAg ICAvLyBGaXJzdCB0ZXN0IHdhcyBjZW50ZXJlZCwgbm93IG1vdmUgdGhlIHRlc3QgY2xvc2VyIHRv IHRoZSB3cm9uZyBub2RlLCBhbmQgZW5zdXJlIHdlIHN0aWxsIGdldCB0aGUgcHJlZmVyZWQgbm9k ZS4KICAgICAgICAgdGVzdFRvdWNoUG9pbnQodG91Y2hQb2ludERvdWJsZVRvdWNoKGUubXlib3gx LCBlLm15Ym94MiwgNSksIGUubXlib3gxKTsKIAotICAgICAgICB0ZXN0VG91Y2hQb2ludCh0b3Vj aFBvaW50RG91YmxlVG91Y2goZS5teWJveDEsIGUubXlib3gyLCAxMCksIGUubXlib3gxKTsKKyAg ICAgICAgdGVzdFRvdWNoUG9pbnQodG91Y2hQb2ludERvdWJsZVRvdWNoKGUubXlib3gxLCBlLm15 Ym94MiwgOCksIGUubXlib3gxKTsKICAgICB9CiAKICAgICBmdW5jdGlvbiBydW5UZXN0cygpCg== 172245 2012-11-04 04:55:35 -0800 2012-11-06 10:40:52 -0800 Patch bug-100833-20121104075401.patch text/plain 4626 rbyers U3VidmVyc2lvbiBSZXZpc2lvbjogMTMzMzM0CmRpZmYgLS1naXQgYS9MYXlvdXRUZXN0cy9DaGFu Z2VMb2cgYi9MYXlvdXRUZXN0cy9DaGFuZ2VMb2cKaW5kZXggNzhhNTg0ZTI2M2I3NWE2MWZhNzJh MWNmODNjNWYzYjU4MWE5OTBiNS4uYTFjNjc3OGNlMzM0MjFjMTI2MzFmNTBkOWMzODA0OGQ0NTZh YWU0NCAxMDA2NDQKLS0tIGEvTGF5b3V0VGVzdHMvQ2hhbmdlTG9nCisrKyBiL0xheW91dFRlc3Rz L0NoYW5nZUxvZwpAQCAtMSwzICsxLDE5IEBACisyMDEyLTExLTAzICBSaWNrIEJ5ZXJzICA8cmJ5 ZXJzQGNocm9taXVtLm9yZz4KKworICAgICAgICBbY2hyb21pdW1dIFR3byB0b3VjaGFkanVzdG1l bnQgdGVzdHMgYXJlIGZhaWxpbmcgb24gbWFjCisgICAgICAgIGh0dHBzOi8vYnVncy53ZWJraXQu b3JnL3Nob3dfYnVnLmNnaT9pZD0xMDA4MzMKKworICAgICAgICBSZXZpZXdlZCBieSBOT0JPRFkg KE9PUFMhKS4KKworICAgICAgICBSZS1lbmFibGUgdGhlc2UgdHdvIHRlc3RzIHdpdGggYXBwcm9w cmlhdGUgdHdlYWtzLiAgQWxzbyBhdm9pZCB0aGUgcGxhdGZvcm0gZGVwZW5kYW50CisgICAgICAg IGJlaGF2aW9yIGJ5IHN3aXRjaGluZyB0byB0aGUgQWhlbSBmb250LgorCisgICAgICAgICogcGxh dGZvcm0vY2hyb21pdW0vVGVzdEV4cGVjdGF0aW9uczogUmVtb3ZlIHRoZSBmYWlsdXJlIGV4cGVj dGF0aW9ucy4KKyAgICAgICAgKiB0b3VjaGFkanVzdG1lbnQvaHRtbC1sYWJlbC5odG1sOiBUd2Vh ayBvbmUgdmVyeSBib3JkZXJsaW5lIGNhc2UgdG8gYmUgY2xvc2VyIHRvIHBhc3MKKyAgICAgICAg cmVsaWFibHkuICBQcmV2aW91c2x5IHRoaXMgcGFzc2VkIG9ubHkgZHVlIHRvIGEgYnVnIGluIHRo ZSB0ZXN0IGluZnJhc3RydWN0dXJlIGZpeGVkIGluIHIxMzI5MjkuCisgICAgICAgICogdG91Y2hh ZGp1c3RtZW50L25lc3RlZC10b3VjaC5odG1sOiBUd2VhayBib3JkZXJsaW5lIGNhc2UgdG8gY29t cGVuc2F0ZSBmb3IgaW5jbHVzaW9uIG9mCisgICAgICAgIHRoZSBib3JkZXIgaW4gdGhlIG1lYXN1 cmVtZW50IG9mIHRoZSBzaXplIG9mIHRoZSBkaXYuCisKIDIwMTItMTEtMDIgIElhbiBWb2xsaWNr ICA8dm9sbGlja0BjaHJvbWl1bS5vcmc+CiAKICAgICAgICAgU3VwcG9ydCBpbnZhbGlkYXRpb24g dHJhY2tpbmcgZm9yIGNvbXBvc2l0ZWQgbGF5ZXJzCmRpZmYgLS1naXQgYS9MYXlvdXRUZXN0cy9w bGF0Zm9ybS9jaHJvbWl1bS9UZXN0RXhwZWN0YXRpb25zIGIvTGF5b3V0VGVzdHMvcGxhdGZvcm0v Y2hyb21pdW0vVGVzdEV4cGVjdGF0aW9ucwppbmRleCA3ZDUxMDU3NWRlNWNjZjFjZWI4MTAzODQ2 Y2JmMDEwMTFmZDZkMDJmLi5jM2QxYjJkNDI0ZTg2MDc5MTNkMjM3M2FlZGY3NDBiYTk3ZmE4MWI5 IDEwMDY0NAotLS0gYS9MYXlvdXRUZXN0cy9wbGF0Zm9ybS9jaHJvbWl1bS9UZXN0RXhwZWN0YXRp b25zCisrKyBiL0xheW91dFRlc3RzL3BsYXRmb3JtL2Nocm9taXVtL1Rlc3RFeHBlY3RhdGlvbnMK QEAgLTQxNjIsOCArNDE2Miw1IEBAIHdlYmtpdC5vcmcvYi8xMDA4MDYgWyBXaW4gXSBmYXN0L2Nz cy9mb250LWZhY2UtZG93bmxvYWQtZXJyb3IuaHRtbCBbIFRleHQgXQogIyBTZWVtcyB0byBoYXZl IHRoZSB3cm9uZyBmb250IGluIHRoZSBTVkcgcGFydCBvZiB0aGUgdGVzdCAoQ1NTIGlzIG9rKS4K IHdlYmtpdC5vcmcvYi8xMDA4MDYgWyBXaW4gXSBzdmcvY3VzdG9tL2ZvbnQtZmFjZS1zaW1wbGUu c3ZnIFsgRmFpbHVyZSBdCiAKLXdlYmtpdC5vcmcvYi8xMDA4MzMgWyBNYWMgXSB0b3VjaGFkanVz dG1lbnQvaHRtbC1sYWJlbC5odG1sIFsgVGV4dCBdCi13ZWJraXQub3JnL2IvMTAwODMzIFsgTWFj IF0gdG91Y2hhZGp1c3RtZW50L25lc3RlZC10b3VjaC5odG1sIFsgVGV4dCBdCi0KICMgVGhpcyB0 ZXN0IGlzIGNvbnNpc3RlbnRseSBsZWFraW5nIHN0YXRlIGludG8gdGhlIG5leHQgdGVzdCAob2Jq ZWN0LXNyYy11cmwtYWxsb3dlZC5odG1sKSBhZnRlciByMTMzMDY5Lgogd2Via2l0Lm9yZy9iLzEw MDk1NSBodHRwL3Rlc3RzL3NlY3VyaXR5L2NvbnRlbnRTZWN1cml0eVBvbGljeS9vYmplY3Qtc3Jj LW5vbmUtYmxvY2tlZC5odG1sIFsgU2tpcCBdCmRpZmYgLS1naXQgYS9MYXlvdXRUZXN0cy90b3Vj aGFkanVzdG1lbnQvaHRtbC1sYWJlbC5odG1sIGIvTGF5b3V0VGVzdHMvdG91Y2hhZGp1c3RtZW50 L2h0bWwtbGFiZWwuaHRtbAppbmRleCAzYzExNmE1NWJhYTBkNWViNDNmMzUxMDExNDZhZjlkZWM4 MzkyZDljLi5kNTQ0MTMxMWYxNzliMTJkYWMzYzMwMDRjNDFlYmYzOTYzNDQyOTI2IDEwMDY0NAot LS0gYS9MYXlvdXRUZXN0cy90b3VjaGFkanVzdG1lbnQvaHRtbC1sYWJlbC5odG1sCisrKyBiL0xh eW91dFRlc3RzL3RvdWNoYWRqdXN0bWVudC9odG1sLWxhYmVsLmh0bWwKQEAgLTksNiArOSw3IEBA CiAgICAgICAgICAgICBwb3NpdGlvbjogYWJzb2x1dGU7CiAgICAgICAgICAgICBsZWZ0OiAwcHg7 CiAgICAgICAgICAgICB0b3A6IDBweDsKKyAgICAgICAgICAgIGZvbnQ6IDE2cHggQWhlbTsKICAg ICAgICAgfQogICAgIDwvc3R5bGU+CiA8L2hlYWQ+CkBAIC00MiwxMCArNDMsMTAgQEAKICAgICAg ICAgdGVzdFRvdWNoUG9pbnQodG91Y2hwb2ludCwgZWxlbWVudCk7CiAgICAgfQogCi0gICAgZnVu Y3Rpb24gdGVzdEluZGlyZWN0VG91Y2goZWxlbWVudCkKKyAgICBmdW5jdGlvbiB0ZXN0SW5kaXJl Y3RUb3VjaChlbGVtZW50LCBvZmZzZXQpCiAgICAgewogICAgICAgICAvLyBUb3VjaCBqdXN0IHJp Z2h0IG9mIHRoZSBlbGVtZW50LgotICAgICAgICB2YXIgdG91Y2hwb2ludCA9IG9mZnNldFRvdWNo UG9pbnQoZmluZEFic29sdXRlQm91bmRzKGVsZW1lbnQpLCAncmlnaHQnLCAxMCwgMzAsIDIwKTsK KyAgICAgICAgdmFyIHRvdWNocG9pbnQgPSBvZmZzZXRUb3VjaFBvaW50KGZpbmRBYnNvbHV0ZUJv dW5kcyhlbGVtZW50KSwgJ3JpZ2h0Jywgb2Zmc2V0LCAzMCwgMjApOwogICAgICAgICB0ZXN0VG91 Y2hQb2ludCh0b3VjaHBvaW50LCBlbGVtZW50KTsKICAgICB9CiAKQEAgLTYyLDkgKzYzLDE0IEBA CiAgICAgewogICAgICAgICBkZWJ1ZygnVGVzdGluZyBpbmRpcmVjdCBoaXRzLicpOwogCi0gICAg ICAgIHRlc3RJbmRpcmVjdFRvdWNoKGUubXlsaW5rKTsKLSAgICAgICAgdGVzdEluZGlyZWN0VG91 Y2goZS5teWxhYmVsKTsKLSAgICAgICAgdGVzdEluZGlyZWN0VG91Y2goZS5teWlucHV0KTsKKyAg ICAgICAgdGVzdEluZGlyZWN0VG91Y2goZS5teWxpbmssIDEwKTsKKworICAgICAgICAvLyBUaGlz IGNhc2UgaXMgdmVyeSBib3JkZXJsaW5lIC0gdGhlcmUncyBub3QgbXVjaCByZWFzb24gdG8gYWRq dXN0CisgICAgICAgIC8vIGxlZnQgdG8gdGhlIGxhYmVsIGluc3RlYWQgb2YgdXAgdG8gdGhlIGxp bmsgKGl0J3MgY2xvc2VyIHRvCisgICAgICAgIC8vIHRoZSBsYWJlbCwgYnV0IG92ZXJsYXBzIHdp dGggbW9yZSBvZiB0aGUgbGluaykuCisgICAgICAgIHRlc3RJbmRpcmVjdFRvdWNoKGUubXlsYWJl bCwgMSk7CisKKyAgICAgICAgdGVzdEluZGlyZWN0VG91Y2goZS5teWlucHV0LCAxMCk7CiAgICAg fQogCiAgICAgZnVuY3Rpb24gcnVuVGVzdHMoKQpkaWZmIC0tZ2l0IGEvTGF5b3V0VGVzdHMvdG91 Y2hhZGp1c3RtZW50L25lc3RlZC10b3VjaC5odG1sIGIvTGF5b3V0VGVzdHMvdG91Y2hhZGp1c3Rt ZW50L25lc3RlZC10b3VjaC5odG1sCmluZGV4IDI2NGU4ODdhNjllYmZiYTA3NDJhYTJhYWJlM2Iz MWFmYzNmNmUxNTUuLjdlN2JjMjY5ZWI5OGI3NDU0OWE0YTI0ZmNkMzhkZWNlYmJiODBjYmMgMTAw NjQ0Ci0tLSBhL0xheW91dFRlc3RzL3RvdWNoYWRqdXN0bWVudC9uZXN0ZWQtdG91Y2guaHRtbAor KysgYi9MYXlvdXRUZXN0cy90b3VjaGFkanVzdG1lbnQvbmVzdGVkLXRvdWNoLmh0bWwKQEAgLTUs NyArNSw3IEBACiAgICAgPHNjcmlwdCBzcmM9Ii4uL2Zhc3QvanMvcmVzb3VyY2VzL2pzLXRlc3Qt cHJlLmpzIj48L3NjcmlwdD4KICAgICA8c2NyaXB0IHNyYz0icmVzb3VyY2VzL3RvdWNoYWRqdXN0 bWVudC5qcyI+PC9zY3JpcHQ+CiAgICAgPHN0eWxlPgotICAgICAgICAjc2FuZGJveCB7IH0KKyAg ICAgICAgI3NhbmRib3ggeyBmb250OiAxNnB4IEFoZW07IH0KICAgICAgICAgLmJveCB7IGJvcmRl cjogMXB4IHNvbGlkIGJsYWNrOyBib3JkZXItcmFkaXVzOiA1cHggNXB4OyBtYXJnaW46IDEycHg7 IG1heC13aWR0aDogNDBlbTsgfQogICAgIDwvc3R5bGU+CiA8L2hlYWQ+CkBAIC03Niw3ICs3Niw3 IEBACiAgICAgICAgIC8vIEZpcnN0IHRlc3Qgd2FzIGNlbnRlcmVkLCBub3cgbW92ZSB0aGUgdGVz dCBjbG9zZXIgdG8gdGhlIHdyb25nIG5vZGUsIGFuZCBlbnN1cmUgd2Ugc3RpbGwgZ2V0IHRoZSBw cmVmZXJlZCBub2RlLgogICAgICAgICB0ZXN0VG91Y2hQb2ludCh0b3VjaFBvaW50RG91YmxlVG91 Y2goZS5teWJveDEsIGUubXlib3gyLCA1KSwgZS5teWJveDEpOwogCi0gICAgICAgIHRlc3RUb3Vj aFBvaW50KHRvdWNoUG9pbnREb3VibGVUb3VjaChlLm15Ym94MSwgZS5teWJveDIsIDEwKSwgZS5t eWJveDEpOworICAgICAgICB0ZXN0VG91Y2hQb2ludCh0b3VjaFBvaW50RG91YmxlVG91Y2goZS5t eWJveDEsIGUubXlib3gyLCA4KSwgZS5teWJveDEpOwogICAgIH0KIAogICAgIGZ1bmN0aW9uIHJ1 blRlc3RzKCkK