83051 2012-04-03 11:19:06 -0700 [Cairo] SVG shadows show through opaque objects 2022-07-15 15:07:44 -0700 1 1 1 Unclassified WebKit SVG 528+ (Nightly build) Unspecified Unspecified NEW DoNotImportToRadar P2 Normal --- 1 pdr webkit-unassigned bfulgham bugs-noreply krit mrobinson thorton zimmermann oldest_to_newest 594666 0 pdr 2012-04-03 11:19:06 -0700 SVG shadows show through opaque objects, causing the test "svg/css/group-with-shadow.svg" to fail on skia platforms. For a repro, just open LayoutTests/svg/css/group-with-shadow.svg and notice the faint grey shadow on top of the purple box. It's much easier to see if you change the testcase to use a white box, and I would recommend that this testcase be changed so busy sheriffs don't make the same mistake again and check in incorrect expectations. 594687 1 135375 pdr 2012-04-03 11:34:51 -0700 Created attachment 135375 Update test expectations so this test fails instead of silently passing 683833 2 135375 pkasting 2012-07-31 20:48:08 -0700 Comment on attachment 135375 Update test expectations so this test fails instead of silently passing This patch landed long ago. 1008944 3 krit 2014-05-13 08:02:50 -0700 Looking at the implementation, Cairo might still be affected. CC: mrobinson 1009096 4 mrobinson 2014-05-13 15:04:18 -0700 The original bug seems to be about the Skia backend. What about the Cairo implementation looks broken to you? 1009192 5 krit 2014-05-13 22:51:06 -0700 (In reply to comment #4) > The original bug seems to be about the Skia backend. What about the Cairo implementation looks broken to you? It was implemented by Aryeh in a platform independent way for Cairo and Qt. Here is the link to the test: http://trac.webkit.org/export/168780/trunk/LayoutTests/svg/css/group-with-shadow.svg Compare it to http://trac.webkit.org/export/168780/trunk/LayoutTests/platform/mac/svg/css/group-with-shadow-expected.png Also, try a tests where elements are overlapping and apply the shadow on a group element: http://jsfiddle.net/rTPaG/ Is the shadow drawn for the group (as it should be) or for each element individually? 1009635 6 mrobinson 2014-05-15 15:09:46 -0700 (In reply to comment #5) > Is the shadow drawn for the group (as it should be) or for each element individually? In the jsfiddle example the shadow is drawn for each element individually. 1009637 7 mrobinson 2014-05-15 15:12:17 -0700 (In reply to comment #5) > Here is the link to the test: http://trac.webkit.org/export/168780/trunk/LayoutTests/svg/css/group-with-shadow.svg Compare it to http://trac.webkit.org/export/168780/trunk/LayoutTests/platform/mac/svg/css/group-with-shadow-expected.png The faint shadow also appears in the Mac expected results, so is this also an issue with the CG backend? 135375 2012-04-03 11:34:51 -0700 2012-07-31 20:48:07 -0700 Update test expectations so this test fails instead of silently passing 83051.1.patch text/plain 3685 pdr SW5kZXg6IExheW91dFRlc3RzL0NoYW5nZUxvZwo9PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBMYXlvdXRUZXN0cy9D aGFuZ2VMb2cJKHJldmlzaW9uIDExMzA2NSkKKysrIExheW91dFRlc3RzL0NoYW5nZUxvZwkod29y a2luZyBjb3B5KQpAQCAtMSwzICsxLDE4IEBACisyMDEyLTA0LTAzICBQaGlsaXAgUm9nZXJzICA8 cGRyQGdvb2dsZS5jb20+CisKKyAgICAgICAgVXBkYXRlIHRlc3QgZXhwZWN0YXRpb25zIHNvIHRo aXMgdGVzdCBmYWlscyBpbnN0ZWFkIG9mIHNpbGVudGx5IHBhc3NpbmcKKyAgICAgICAgaHR0cHM6 Ly9idWdzLndlYmtpdC5vcmcvc2hvd19idWcuY2dpP2lkPTgzMDUxCisKKyAgICAgICAgVW5yZXZp ZXdlZCB0ZXN0IGV4cGVjdGF0aW9ucyB1cGRhdGUgY29ycmVjdGluZyBhbiBpbmNvcnJlY3RseSBj aGVja2VkIGluIGV4cGVjdGFpdG9uLgorCisgICAgICAgICogcGxhdGZvcm0vY2hyb21pdW0tbGlu dXgvc3ZnL2Nzcy9ncm91cC13aXRoLXNoYWRvdy1leHBlY3RlZC5wbmc6IFJlbW92ZWQuCisgICAg ICAgICogcGxhdGZvcm0vY2hyb21pdW0tbWFjLWxlb3BhcmQvc3ZnL2Nzcy9ncm91cC13aXRoLXNo YWRvdy1leHBlY3RlZC5wbmc6IFJlbW92ZWQuCisgICAgICAgICogcGxhdGZvcm0vY2hyb21pdW0t d2luLXZpc3RhL3N2Zy9jc3MvZ3JvdXAtd2l0aC1zaGFkb3ctZXhwZWN0ZWQucG5nOiBSZW1vdmVk LgorICAgICAgICAqIHBsYXRmb3JtL2Nocm9taXVtLXdpbi14cC9zdmcvY3NzL2dyb3VwLXdpdGgt c2hhZG93LWV4cGVjdGVkLnBuZzogUmVtb3ZlZC4KKyAgICAgICAgKiBwbGF0Zm9ybS9jaHJvbWl1 bS13aW4vc3ZnL2Nzcy9ncm91cC13aXRoLXNoYWRvdy1leHBlY3RlZC5wbmc6IFJlbW92ZWQuCisg ICAgICAgICogcGxhdGZvcm0vY2hyb21pdW0vdGVzdF9leHBlY3RhdGlvbnMudHh0OgorICAgICAg ICAqIHBsYXRmb3JtL21hYy1zbm93bGVvcGFyZC9zdmcvY3NzL2dyb3VwLXdpdGgtc2hhZG93LWV4 cGVjdGVkLnBuZzogUmVtb3ZlZC4KKwogMjAxMi0wNC0wMyAgQW1pIEZpc2NobWFuICA8ZmlzY2ht YW5AY2hyb21pdW0ub3JnPgogCiAgICAgICAgIExheW91dCBUZXN0IG1lZGlhL3ZpZGVvLXNyYy1z b3VyY2UuaHRtbCBpcyBmbGFreQpJbmRleDogTGF5b3V0VGVzdHMvcGxhdGZvcm0vY2hyb21pdW0t bGludXgvc3ZnL2Nzcy9ncm91cC13aXRoLXNoYWRvdy1leHBlY3RlZC5wbmcKPT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQpD YW5ub3QgZGlzcGxheTogZmlsZSBtYXJrZWQgYXMgYSBiaW5hcnkgdHlwZS4Kc3ZuOm1pbWUtdHlw ZSA9IGltYWdlL3BuZwpJbmRleDogTGF5b3V0VGVzdHMvcGxhdGZvcm0vY2hyb21pdW0tbWFjLWxl b3BhcmQvc3ZnL2Nzcy9ncm91cC13aXRoLXNoYWRvdy1leHBlY3RlZC5wbmcKPT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQpD YW5ub3QgZGlzcGxheTogZmlsZSBtYXJrZWQgYXMgYSBiaW5hcnkgdHlwZS4Kc3ZuOm1pbWUtdHlw ZSA9IGltYWdlL3BuZwpJbmRleDogTGF5b3V0VGVzdHMvcGxhdGZvcm0vY2hyb21pdW0td2luLXZp c3RhL3N2Zy9jc3MvZ3JvdXAtd2l0aC1zaGFkb3ctZXhwZWN0ZWQucG5nCj09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0KQ2Fu bm90IGRpc3BsYXk6IGZpbGUgbWFya2VkIGFzIGEgYmluYXJ5IHR5cGUuCnN2bjptaW1lLXR5cGUg PSBhcHBsaWNhdGlvbi9vY3RldC1zdHJlYW0KSW5kZXg6IExheW91dFRlc3RzL3BsYXRmb3JtL2No cm9taXVtLXdpbi14cC9zdmcvY3NzL2dyb3VwLXdpdGgtc2hhZG93LWV4cGVjdGVkLnBuZwo9PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09CkNhbm5vdCBkaXNwbGF5OiBmaWxlIG1hcmtlZCBhcyBhIGJpbmFyeSB0eXBlLgpzdm46 bWltZS10eXBlID0gaW1hZ2UvcG5nCkluZGV4OiBMYXlvdXRUZXN0cy9wbGF0Zm9ybS9jaHJvbWl1 bS13aW4vc3ZnL2Nzcy9ncm91cC13aXRoLXNoYWRvdy1leHBlY3RlZC5wbmcKPT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQpD YW5ub3QgZGlzcGxheTogZmlsZSBtYXJrZWQgYXMgYSBiaW5hcnkgdHlwZS4Kc3ZuOm1pbWUtdHlw ZSA9IGltYWdlL3BuZwpJbmRleDogTGF5b3V0VGVzdHMvcGxhdGZvcm0vY2hyb21pdW0vdGVzdF9l eHBlY3RhdGlvbnMudHh0Cj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIExheW91dFRlc3RzL3BsYXRmb3JtL2Nocm9t aXVtL3Rlc3RfZXhwZWN0YXRpb25zLnR4dAkocmV2aXNpb24gMTEzMDM3KQorKysgTGF5b3V0VGVz dHMvcGxhdGZvcm0vY2hyb21pdW0vdGVzdF9leHBlY3RhdGlvbnMudHh0CSh3b3JraW5nIGNvcHkp CkBAIC0xMTc2LDcgKzExNzYsNiBAQCBCVUdXSzc5NTY4IDogc3ZnL1czQy1TVkctMS4xL3N0cnVj dC11c2UtCiBCVUdXSzc1OTY4IFdJTiBMRU9QQVJEIFNOT1dMRU9QQVJEIDogc3ZnL1czQy1TVkct MS4xL2FuaW1hdGUtZWxlbS0zOS10LnN2ZyA9IFRFWFQKIEJVR1dLNzU5NjggV0lOIExFT1BBUkQg U05PV0xFT1BBUkQgOiBzdmcvVzNDLVNWRy0xLjEvYW5pbWF0ZS1lbGVtLTQ2LXQuc3ZnID0gSU1B R0UrVEVYVCBURVhUCiBCVUdXSzc1OTY4IDogc3ZnL1czQy1TVkctMS4xL2ZpbHRlcnMtbW9ycGgt MDEtZi5zdmcgPSBJTUFHRStURVhUIElNQUdFCi1CVUdXSzc1OTY4IExJT04gOiBzdmcvY3NzL2dy b3VwLXdpdGgtc2hhZG93LnN2ZyA9IElNQUdFCiAKIC8vIEZsYWt5IG9uIE1BQwogQlVHV0s4Mjkx MSBNQUMgOiBzdmcvVzNDLVNWRy0xLjEvYW5pbWF0ZS1lbGVtLTgwLXQuc3ZnID0gVEVYVCBJTUFH RStURVhUIElNQUdFIFBBU1MKQEAgLTExODksNiArMTE4OCw4IEBAIEJVR1dLODI5NTQgU05PV0xF T1BBUkQgOiBzdmcvVzNDLVNWRy0xLjEKIC8vIEZsYWt5IG9uIExpbnV4IGFmdGVyIGh0dHA6Ly90 cmFjLndlYmtpdC5vcmcvY2hhbmdlc2V0LzExMjAyMgogQlVHV0s4MjIxNyBMSU5VWCA6IHN2Zy96 b29tL3RleHQvem9vbS1mb3JlaWduT2JqZWN0LnN2ZyA9IElNQUdFIFBBU1MKIAorLy8gRmFpbGlu ZyBvbiBTa2lhIHBsYXRmb3JtcworQlVHV0s4MzA1MSBXSU4gTElOVVggTElPTiA6IHN2Zy9jc3Mv Z3JvdXAtd2l0aC1zaGFkb3cuc3ZnID0gSU1BR0UgUEFTUyAKIAogLy8gV2luZG93cyBhbmQgTWFj IG5lZWQgcmViYXNlbGluaW5nIGFmdGVyIDUzMzc4IHBhdGNoLgogQlVHV0s1MzM3OCBXSU4gTUFD IDogc3ZnL2NsaXAtcGF0aC9jbGlwLWluLW1hc2suc3ZnID0gSU1BR0UKSW5kZXg6IExheW91dFRl c3RzL3BsYXRmb3JtL21hYy1zbm93bGVvcGFyZC9zdmcvY3NzL2dyb3VwLXdpdGgtc2hhZG93LWV4 cGVjdGVkLnBuZwo9PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09CkNhbm5vdCBkaXNwbGF5OiBmaWxlIG1hcmtlZCBhcyBhIGJp bmFyeSB0eXBlLgpzdm46bWltZS10eXBlID0gaW1hZ2UvcG5nCg==