193108 2019-01-03 08:48:05 -0800 [ Mojave WK1 ] Some Image tests are flakey failures and are failing in tandem with zoomed in or blank image results 2019-07-15 19:32:50 -0700 1 1 1 Unclassified WebKit Tools / Tests WebKit Nightly Build Unspecified Unspecified RESOLVED FIXED https://bugs.webkit.org/show_bug.cgi?id=193107 https://bugs.webkit.org/show_bug.cgi?id=193091 https://bugs.webkit.org/show_bug.cgi?id=192881 https://bugs.webkit.org/show_bug.cgi?id=193116 https://bugs.webkit.org/show_bug.cgi?id=190383 InRadar P2 Normal --- 1 tsavell sabouhallawa commit-queue jlewis3 lforschler ryanhaddad sabouhallawa simon.fraser sroberts thorton webkit-bug-importer oldest_to_newest 1491432 0 tsavell 2019-01-03 08:48:05 -0800 There seems to be a fair amount of image tests that are flakey failures due to a zoomed in or blank rendered result. These tests also appear to be failing in tandem in small groups. These tests are not all part of the same directory but are displaying very similar failures. Bugs: https://bugs.webkit.org/show_bug.cgi?id=193107 https://bugs.webkit.org/show_bug.cgi?id=193091 https://bugs.webkit.org/show_bug.cgi?id=192881 1491481 1 tsavell 2019-01-03 11:39:05 -0800 another set of tests appear to exhibiting this issue. https://bugs.webkit.org/show_bug.cgi?id=193116 1491672 2 tsavell 2019-01-04 09:28:33 -0800 It appears that Image tests on Mojave WK1 in general have this zooming or blank issue. Here is a run with 15 failing image tests with many of them showing a zoomed in result: https://build.webkit.org/results/Apple%20Mojave%20Release%20WK1%20(Tests)/r239610%20(1925)/results.html 1491673 3 webkit-bug-importer 2019-01-04 09:32:48 -0800 <rdar://problem/47049667> 1503369 4 sroberts 2019-02-06 10:06:27 -0800 Another set of tests exhibiting this issue. https://build.webkit.org/results/Apple%20Mojave%20Release%20WK1%20(Tests)/r241018%20(2754)/results.html 1503370 5 sroberts 2019-02-06 10:08:25 -0800 Also: https://build.webkit.org/results/Apple%20Mojave%20Release%20WK1%20(Tests)/r241021%20(2756)/results.html 1536147 6 tsavell 2019-05-14 11:12:36 -0700 Set of 12 newer failure: https://build.webkit.org/results/Apple%20Mojave%20Release%20WK1%20(Tests)/r245285%20(4955)/results.html Set of 5: https://build.webkit.org/results/Apple%20Mojave%20Release%20WK1%20(Tests)/r245210%20(4919)/results.html Set of 10: https://build.webkit.org/results/Apple%20Mojave%20Debug%20WK1%20(Tests)/r245219%20(3433)/results.html 1552988 7 sabouhallawa 2019-07-15 15:15:21 -0700 This bug is happening because of a bug in CGWindowListCreateImage() in which it returns the wrong size snapshot when changing the window deviceScaleFactor from 2 to 1. When running DRT, the bug happens only when running a hidpi test followed by a composite layout test. To fake the hipdi the deviceScaleFactor is set manually to 2. Before running a new test, the deviceScaleFactor has to be set back to 1. But this does not make CGWindowListCreateImage() works correctly in this case. When createBitmapContextFromWebView() calls CGContextDrawImage(), only the bottom left quarter snapshot is drawn from the snapshot image to the bitmapContext. Most of the layout tests do not draw in this part of the WebView. So the result is seen as blank. Like what we did in WebViewImpl::takeViewSnapshot() as a workaround for <rdar://problem/17084993>, we have to do similar workaround for DRT function createBitmapContextFromWebView(). 1552989 8 374154 sabouhallawa 2019-07-15 15:22:32 -0700 Created attachment 374154 Patch 1553078 9 374154 commit-queue 2019-07-15 19:32:48 -0700 Comment on attachment 374154 Patch Clearing flags on attachment: 374154 Committed r247466: <https://trac.webkit.org/changeset/247466> 1553079 10 commit-queue 2019-07-15 19:32:50 -0700 All reviewed patches have been landed. Closing bug. 374154 2019-07-15 15:22:32 -0700 2019-07-15 19:32:48 -0700 Patch bug-193108-20190715152231.patch text/plain 3539 sabouhallawa U3VidmVyc2lvbiBSZXZpc2lvbjogMjQ3MzczCmRpZmYgLS1naXQgYS9Ub29scy9DaGFuZ2VMb2cg Yi9Ub29scy9DaGFuZ2VMb2cKaW5kZXggMzAzN2E5YzkwNmNlNDFjMDcwYWMyOTMxNzVlMmQzNTkx ZDhhMWQ0OS4uMTYxNjliMjhiNTQ4OGRlYmQ5YjEzMDQ2ZjIyNWZmZDU5YjRlODk4YSAxMDA2NDQK LS0tIGEvVG9vbHMvQ2hhbmdlTG9nCisrKyBiL1Rvb2xzL0NoYW5nZUxvZwpAQCAtMSwzICsxLDE4 IEBACisyMDE5LTA3LTE1ICBTYWlkIEFib3UtSGFsbGF3YSAgPHNhYm91aGFsbGF3YUBhcHBsZS5j b20+CisKKyAgICAgICAgWyBNb2phdmUgV0sxIF0gU29tZSBJbWFnZSB0ZXN0cyBhcmUgZmxha2V5 IGZhaWx1cmVzIGFuZCBhcmUgZmFpbGluZyBpbiB0YW5kZW0gd2l0aCB6b29tZWQgaW4gb3IgYmxh bmsgaW1hZ2UgcmVzdWx0cworICAgICAgICBodHRwczovL2J1Z3Mud2Via2l0Lm9yZy9zaG93X2J1 Zy5jZ2k/aWQ9MTkzMTA4CisKKyAgICAgICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChPT1BTISkuCisK KyAgICAgICAgQWRkIGEgd29ya2Fyb3VuZCBmb3IgPHJkYXI6Ly9wcm9ibGVtLzE3MDg0OTkzPiBp biBjcmVhdGVCaXRtYXBDb250ZXh0RnJvbVdlYlZpZXcoKS4KKyAgICAgICAgUmUtcmVxdWVzdCB0 aGUgc25hcHNob3QgYXQga0NHV2luZG93SW1hZ2VOb21pbmFsUmVzb2x1dGlvbiBpZiBpdCB3YXMg Y2FwdHVyZWQKKyAgICAgICAgYXQgdGhlIHdyb25nIHNjYWxlLgorCisgICAgICAgICogRHVtcFJl bmRlclRyZWUvbWFjL1BpeGVsRHVtcFN1cHBvcnRNYWMubW06CisgICAgICAgICh0YWtlV2luZG93 U25hcHNob3QpOgorICAgICAgICAoY3JlYXRlQml0bWFwQ29udGV4dEZyb21XZWJWaWV3KToKKwog MjAxOS0wNy0xMSAgQWFrYXNoIEphaW4gIDxhYWthc2hfamFpbkBhcHBsZS5jb20+CiAKICAgICAg ICAgVW5yZXZpZXdlZCBmb2xsb3ctdXAgZml4IHRvIHIyNDczNjQuCmRpZmYgLS1naXQgYS9Ub29s cy9EdW1wUmVuZGVyVHJlZS9tYWMvUGl4ZWxEdW1wU3VwcG9ydE1hYy5tbSBiL1Rvb2xzL0R1bXBS ZW5kZXJUcmVlL21hYy9QaXhlbER1bXBTdXBwb3J0TWFjLm1tCmluZGV4IDBiMmE5N2M0NmJjY2Q1 Y2NhYjcyZWYyZWI4NjBiZmYzMGEwZDY1YzIuLjdjYWJiMTI1NzI4NGMzNDEwNDMxNWQwZjk3YmIy ZDgxZDkzMmEyNTkgMTAwNjQ0Ci0tLSBhL1Rvb2xzL0R1bXBSZW5kZXJUcmVlL21hYy9QaXhlbER1 bXBTdXBwb3J0TWFjLm1tCisrKyBiL1Rvb2xzL0R1bXBSZW5kZXJUcmVlL21hYy9QaXhlbER1bXBT dXBwb3J0TWFjLm1tCkBAIC0zNiw2ICszNiw3IEBACiAjaW1wb3J0ICJUZXN0UnVubmVyLmgiCiAj aW1wb3J0IDxDb3JlR3JhcGhpY3MvQ0dCaXRtYXBDb250ZXh0Lmg+CiAjaW1wb3J0IDxRdWFydHpD b3JlL1F1YXJ0ekNvcmUuaD4KKyNpbXBvcnQgPHBhbC9zcGkvY2cvQ29yZUdyYXBoaWNzU1BJLmg+ CiAjaW1wb3J0IDx3dGYvQXNzZXJ0aW9ucy5oPgogI2ltcG9ydCA8d3RmL1JlZlB0ci5oPgogCkBA IC04Miw2ICs4MywxMiBAQCBzdGF0aWMgdm9pZCBwYWludFJlcGFpbnRSZWN0T3ZlcmxheShXZWJW aWV3KiB3ZWJWaWV3LCBDR0NvbnRleHRSZWYgY29udGV4dCkKICAgICBDR0NvbnRleHRSZXN0b3Jl R1N0YXRlKGNvbnRleHQpOwogfQogCitzdGF0aWMgQ0dJbWFnZVJlZiB0YWtlV2luZG93U25hcHNo b3QoQ0dTV2luZG93SUQgd2luZG93SUQsIENHV2luZG93SW1hZ2VPcHRpb24gaW1hZ2VPcHRpb25z KQoreworICAgIGltYWdlT3B0aW9ucyB8PSBrQ0dXaW5kb3dJbWFnZUJvdW5kc0lnbm9yZUZyYW1p bmcgfCBrQ0dXaW5kb3dJbWFnZVNob3VsZEJlT3BhcXVlOworICAgIHJldHVybiBDR1dpbmRvd0xp c3RDcmVhdGVJbWFnZShDR1JlY3ROdWxsLCBrQ0dXaW5kb3dMaXN0T3B0aW9uSW5jbHVkaW5nV2lu ZG93LCB3aW5kb3dJRCwgaW1hZ2VPcHRpb25zKTsKK30KKwogUmVmUHRyPEJpdG1hcENvbnRleHQ+ IGNyZWF0ZUJpdG1hcENvbnRleHRGcm9tV2ViVmlldyhib29sIG9uc2NyZWVuLCBib29sIGluY3Jl bWVudGFsUmVwYWludCwgYm9vbCBzd2VlcEhvcml6b250YWxseSwgYm9vbCBkcmF3U2VsZWN0aW9u UmVjdCkKIHsKICAgICBXZWJWaWV3KiB2aWV3ID0gW21haW5GcmFtZSB3ZWJWaWV3XTsKQEAgLTEy OCw3ICsxMzUsMTkgQEAgUmVmUHRyPEJpdG1hcENvbnRleHQ+IGNyZWF0ZUJpdG1hcENvbnRleHRG cm9tV2ViVmlldyhib29sIG9uc2NyZWVuLCBib29sIGluY3JlbWUKIAogICAgICAgICAgICAgLy8g QXNrIHRoZSB3aW5kb3cgc2VydmVyIHRvIHByb3ZpZGUgdXMgYSBjb21wb3NpdGVkIHZlcnNpb24g b2YgdGhlICpyZWFsKiB3aW5kb3cgY29udGVudCBpbmNsdWRpbmcgc3VyZmFjZXMgKGkuZS4gT3Bl bkdMIGNvbnRlbnQpCiAgICAgICAgICAgICAvLyBOb3RlIHRoYXQgdGhlIHJldHVybmVkIGltYWdl IG1pZ2h0IGRpZmZlciB2ZXJ5IHNsaWdodGx5IGZyb20gdGhlIHdpbmRvdyBiYWNraW5nIGJlY2F1 c2Ugb2YgZGl0aGVyaW5nIGFydGlmYWN0cyBpbiB0aGUgd2luZG93IHNlcnZlciBjb21wb3NpdG9y LgotICAgICAgICAgICAgQ0dJbWFnZVJlZiBpbWFnZSA9IENHV2luZG93TGlzdENyZWF0ZUltYWdl KENHUmVjdE51bGwsIGtDR1dpbmRvd0xpc3RPcHRpb25JbmNsdWRpbmdXaW5kb3csIFtbdmlldyB3 aW5kb3ddIHdpbmRvd051bWJlcl0sIGtDR1dpbmRvd0ltYWdlQm91bmRzSWdub3JlRnJhbWluZyB8 IGtDR1dpbmRvd0ltYWdlU2hvdWxkQmVPcGFxdWUpOworICAgICAgICAgICAgTlNXaW5kb3cgKndp bmRvdyA9IFt2aWV3IHdpbmRvd107CisgICAgICAgICAgICBDR0ltYWdlUmVmIGltYWdlID0gdGFr ZVdpbmRvd1NuYXBzaG90KFt3aW5kb3cgd2luZG93TnVtYmVyXSwga0NHV2luZG93SW1hZ2VEZWZh dWx0KTsKKworICAgICAgICAgICAgaWYgKGltYWdlKSB7CisgICAgICAgICAgICAgICAgLy8gV29y ayBhcm91bmQgPHJkYXI6Ly9wcm9ibGVtLzE3MDg0OTkzPjsgcmUtcmVxdWVzdCB0aGUgc25hcHNo b3QgYXQga0NHV2luZG93SW1hZ2VOb21pbmFsUmVzb2x1dGlvbiBpZiBpdCB3YXMgY2FwdHVyZWQg YXQgdGhlIHdyb25nIHNjYWxlLgorICAgICAgICAgICAgICAgIENHRmxvYXQgZGVzaXJlZFNuYXBz aG90V2lkdGggPSB3aW5kb3cuZnJhbWUuc2l6ZS53aWR0aCAqIGRldmljZVNjYWxlRmFjdG9yOwor ICAgICAgICAgICAgICAgIGlmIChDR0ltYWdlR2V0V2lkdGgoaW1hZ2UpICE9IGRlc2lyZWRTbmFw c2hvdFdpZHRoKQorICAgICAgICAgICAgICAgICAgICBpbWFnZSA9IHRha2VXaW5kb3dTbmFwc2hv dChbd2luZG93IHdpbmRvd051bWJlcl0sIGtDR1dpbmRvd0ltYWdlTm9taW5hbFJlc29sdXRpb24p OworICAgICAgICAgICAgfQorCisgICAgICAgICAgICBpZiAoIWltYWdlKQorICAgICAgICAgICAg ICAgIHJldHVybiBudWxscHRyOworCiAgICAgICAgICAgICBDR0NvbnRleHREcmF3SW1hZ2UoY29u dGV4dCwgQ0dSZWN0TWFrZSgwLCAwLCBDR0ltYWdlR2V0V2lkdGgoaW1hZ2UpLCBDR0ltYWdlR2V0 SGVpZ2h0KGltYWdlKSksIGltYWdlKTsKICAgICAgICAgICAgIENHSW1hZ2VSZWxlYXNlKGltYWdl KTsKIAo=