149752 2015-10-02 09:35:51 -0700 [Cairo] fast/canvas/canvas-imageSmoothingFoo tests failed after r190383. 2015-10-06 09:35:37 -0700 1 1 1 Unclassified WebKit Platform WebKit Nightly Build Unspecified Unspecified RESOLVED FIXED P2 Normal --- 1 hs85.jeong hs85.jeong cgarcia commit-queue esprehn+autocc gyuyoung.kim mrobinson oldest_to_newest 1130248 0 hs85.jeong 2015-10-02 09:35:51 -0700 https://build.webkit.org/builders/GTK%20Linux%2064-bit%20Debug%20%28Tests%29/builds/5838/steps/layout-test/logs/stdio fast/canvas/canvas-imageSmoothingEnabled.html [ Failure ] fast/canvas/canvas-imageSmoothingQuality.html [ Failure ] 1130249 1 262339 hs85.jeong 2015-10-02 09:41:35 -0700 Created attachment 262339 Patch 1130563 2 262339 mrobinson 2015-10-04 23:42:41 -0700 Comment on attachment 262339 Patch I wonder if it is possible to have Cairo more closely match the behavior of CG here? What do you think? 1130567 3 cgarcia 2015-10-05 00:13:35 -0700 (In reply to comment #2) > Comment on attachment 262339 [details] > Patch > > I wonder if it is possible to have Cairo more closely match the behavior of > CG here? What do you think? I don't think it's possible, because CG has more options than cairo, so there's always an option we can't match. In cairo Low and None are the same, which is CAIRO_FILTER_FAST, medium and default are also the same CAIRO_FILTER_GOOD and high is CAIRO_FILTER_BEST. A possible change could be: - None -> FAST - Low, Default, Medium -> GOOD - High -> BEST But there are cases where Low is used for performance reasons (scrolling) and we really want to use FAST there. What we currently do in HTMLCanvasElement.h is defining a DefaultInterpolationQuality #if USE(CG) #define DefaultInterpolationQuality InterpolationLow #else #define DefaultInterpolationQuality InterpolationDefault #endif I think we could do the same for SmoothingQuality, and use the default value in the constructor. 1130837 4 hs85.jeong 2015-10-05 19:41:38 -0700 (In reply to comment #3) > (In reply to comment #2) > > Comment on attachment 262339 [details] > > Patch > > > > I wonder if it is possible to have Cairo more closely match the behavior of > > CG here? What do you think? > > I don't think it's possible, because CG has more options than cairo, so > there's always an option we can't match. In cairo Low and None are the same, > which is CAIRO_FILTER_FAST, medium and default are also the same > CAIRO_FILTER_GOOD and high is CAIRO_FILTER_BEST. A possible change could be: > > - None -> FAST > - Low, Default, Medium -> GOOD > - High -> BEST > You mean, we change the match : switch (m_state->m_imageInterpolationQuality) { case InterpolationNone: cairo_pattern_set_filter(pattern.get(), CAIRO_FILTER_FAST); break; case InterpolationLow: case InterpolationMedium: case InterpolationDefault: cairo_pattern_set_filter(pattern.get(), CAIRO_FILTER_GOOD); break; case InterpolationHigh: cairo_pattern_set_filter(pattern.get(), CAIRO_FILTER_BEST); break; } ? > But there are cases where Low is used for performance reasons (scrolling) > and we really want to use FAST there. What we currently do in > HTMLCanvasElement.h is defining a DefaultInterpolationQuality > > #if USE(CG) > #define DefaultInterpolationQuality InterpolationLow > #else > #define DefaultInterpolationQuality InterpolationDefault > #endif > > I think we could do the same for SmoothingQuality, and use the default value > in the constructor. and use the 'InterpolationLow' like as CG in HTMLCanvasElement? 1130877 5 cgarcia 2015-10-05 22:50:09 -0700 (In reply to comment #4) > (In reply to comment #3) > > (In reply to comment #2) > > > Comment on attachment 262339 [details] > > > Patch > > > > > > I wonder if it is possible to have Cairo more closely match the behavior of > > > CG here? What do you think? > > > > I don't think it's possible, because CG has more options than cairo, so > > there's always an option we can't match. In cairo Low and None are the same, > > which is CAIRO_FILTER_FAST, medium and default are also the same > > CAIRO_FILTER_GOOD and high is CAIRO_FILTER_BEST. A possible change could be: > > > > - None -> FAST > > - Low, Default, Medium -> GOOD > > - High -> BEST > > > You mean, we change the match : > > switch (m_state->m_imageInterpolationQuality) { > case InterpolationNone: > cairo_pattern_set_filter(pattern.get(), CAIRO_FILTER_FAST); > break; > case InterpolationLow: > case InterpolationMedium: > case InterpolationDefault: > cairo_pattern_set_filter(pattern.get(), CAIRO_FILTER_GOOD); > break; > case InterpolationHigh: > cairo_pattern_set_filter(pattern.get(), CAIRO_FILTER_BEST); > break; > } > ? No, as I said below, this will cause performance regression in some situations. > > But there are cases where Low is used for performance reasons (scrolling) > > and we really want to use FAST there. What we currently do in > > HTMLCanvasElement.h is defining a DefaultInterpolationQuality > > > > #if USE(CG) > > #define DefaultInterpolationQuality InterpolationLow > > #else > > #define DefaultInterpolationQuality InterpolationDefault > > #endif > > > > I think we could do the same for SmoothingQuality, and use the default value > > in the constructor. > > and use the 'InterpolationLow' like as CG in HTMLCanvasElement? Define a DefaultSmoothingQuality there and use it in the constructor 1130884 6 262503 hs85.jeong 2015-10-05 23:01:42 -0700 Created attachment 262503 Patch 1130885 7 hs85.jeong 2015-10-05 23:04:08 -0700 (In reply to comment #5) > (In reply to comment #4) > > (In reply to comment #3) > > > (In reply to comment #2) > > > > Comment on attachment 262339 [details] > > > > Patch > > > > > > > > I wonder if it is possible to have Cairo more closely match the behavior of > > > > CG here? What do you think? > > > > > > I don't think it's possible, because CG has more options than cairo, so > > > there's always an option we can't match. In cairo Low and None are the same, > > > which is CAIRO_FILTER_FAST, medium and default are also the same > > > CAIRO_FILTER_GOOD and high is CAIRO_FILTER_BEST. A possible change could be: > > > > > > - None -> FAST > > > - Low, Default, Medium -> GOOD > > > - High -> BEST > > > > > You mean, we change the match : > > > > switch (m_state->m_imageInterpolationQuality) { > > case InterpolationNone: > > cairo_pattern_set_filter(pattern.get(), CAIRO_FILTER_FAST); > > break; > > case InterpolationLow: > > case InterpolationMedium: > > case InterpolationDefault: > > cairo_pattern_set_filter(pattern.get(), CAIRO_FILTER_GOOD); > > break; > > case InterpolationHigh: > > cairo_pattern_set_filter(pattern.get(), CAIRO_FILTER_BEST); > > break; > > } > > ? > > No, as I said below, this will cause performance regression in some > situations. > > > > But there are cases where Low is used for performance reasons (scrolling) > > > and we really want to use FAST there. What we currently do in > > > HTMLCanvasElement.h is defining a DefaultInterpolationQuality > > > > > > #if USE(CG) > > > #define DefaultInterpolationQuality InterpolationLow > > > #else > > > #define DefaultInterpolationQuality InterpolationDefault > > > #endif > > > > > > I think we could do the same for SmoothingQuality, and use the default value > > > in the constructor. > > > > and use the 'InterpolationLow' like as CG in HTMLCanvasElement? > > Define a DefaultSmoothingQuality there and use it in the constructor Oh..I see. But I got a fail message after applying this patch in fast/canvas/canvas-imageSmoothingQuality.html. PASS defaultContext.imageSmoothingQuality is 'low' FAIL defaultContext.imageSmoothingQuality should be low. Was medium. 1130888 8 hs85.jeong 2015-10-05 23:07:23 -0700 (In reply to comment #7) > (In reply to comment #5) > > (In reply to comment #4) > > > (In reply to comment #3) > > > > (In reply to comment #2) > > > > > Comment on attachment 262339 [details] > > > > > Patch > > > > > > > > > > I wonder if it is possible to have Cairo more closely match the behavior of > > > > > CG here? What do you think? > > > > > > > > I don't think it's possible, because CG has more options than cairo, so > > > > there's always an option we can't match. In cairo Low and None are the same, > > > > which is CAIRO_FILTER_FAST, medium and default are also the same > > > > CAIRO_FILTER_GOOD and high is CAIRO_FILTER_BEST. A possible change could be: > > > > > > > > - None -> FAST > > > > - Low, Default, Medium -> GOOD > > > > - High -> BEST > > > > > > > You mean, we change the match : > > > > > > switch (m_state->m_imageInterpolationQuality) { > > > case InterpolationNone: > > > cairo_pattern_set_filter(pattern.get(), CAIRO_FILTER_FAST); > > > break; > > > case InterpolationLow: > > > case InterpolationMedium: > > > case InterpolationDefault: > > > cairo_pattern_set_filter(pattern.get(), CAIRO_FILTER_GOOD); > > > break; > > > case InterpolationHigh: > > > cairo_pattern_set_filter(pattern.get(), CAIRO_FILTER_BEST); > > > break; > > > } > > > ? > > > > No, as I said below, this will cause performance regression in some > > situations. > > > > > > But there are cases where Low is used for performance reasons (scrolling) > > > > and we really want to use FAST there. What we currently do in > > > > HTMLCanvasElement.h is defining a DefaultInterpolationQuality > > > > > > > > #if USE(CG) > > > > #define DefaultInterpolationQuality InterpolationLow > > > > #else > > > > #define DefaultInterpolationQuality InterpolationDefault > > > > #endif > > > > > > > > I think we could do the same for SmoothingQuality, and use the default value > > > > in the constructor. > > > > > > and use the 'InterpolationLow' like as CG in HTMLCanvasElement? > > > > Define a DefaultSmoothingQuality there and use it in the constructor > > Oh..I see. But I got a fail message after applying this patch in > fast/canvas/canvas-imageSmoothingQuality.html. > > PASS defaultContext.imageSmoothingQuality is 'low' > FAIL defaultContext.imageSmoothingQuality should be low. Was medium. fase/canvas/canvas-imageSmoothingQuality test checks the default value of imageSmoothingQuality whether is low. 1130895 9 cgarcia 2015-10-05 23:13:36 -0700 So, we need platform specific test result, then 1130905 10 262505 hs85.jeong 2015-10-05 23:49:05 -0700 Created attachment 262505 Patch 1130906 11 hs85.jeong 2015-10-05 23:51:11 -0700 (In reply to comment #9) > So, we need platform specific test result, then Do I add the platform specific test expected file with FAIL result? 1130912 12 cgarcia 2015-10-06 00:44:43 -0700 (In reply to comment #11) > (In reply to comment #9) > > So, we need platform specific test result, then > > Do I add the platform specific test expected file with FAIL result? Yes, that's what we expect :-) Thanks 1130989 13 262505 commit-queue 2015-10-06 09:35:32 -0700 Comment on attachment 262505 Patch Clearing flags on attachment: 262505 Committed r190618: <http://trac.webkit.org/changeset/190618> 1130990 14 commit-queue 2015-10-06 09:35:37 -0700 All reviewed patches have been landed. Closing bug. 262339 2015-10-02 09:41:35 -0700 2015-10-05 23:01:34 -0700 Patch bug-149752-20151003103914.patch text/plain 1613 hs85.jeong U3VidmVyc2lvbiBSZXZpc2lvbjogMTkwNDg0CmRpZmYgLS1naXQgYS9Tb3VyY2UvV2ViQ29yZS9D aGFuZ2VMb2cgYi9Tb3VyY2UvV2ViQ29yZS9DaGFuZ2VMb2cKaW5kZXggMjdmOWIwNGE5Yzg5ODBi YjkxZWMxMjQ0OThlMzY1OTg3ZTE3YWEwYi4uYWRkY2U2NWE1OWVkYjU5NmYzNmI1NGFhYzk2YjFi ZTJjZjgzNTA1MyAxMDA2NDQKLS0tIGEvU291cmNlL1dlYkNvcmUvQ2hhbmdlTG9nCisrKyBiL1Nv dXJjZS9XZWJDb3JlL0NoYW5nZUxvZwpAQCAtMSwzICsxLDE1IEBACisyMDE1LTEwLTAyICBIdW5z ZW9wIEplb25nICA8aHM4NS5qZW9uZ0BzYW1zdW5nLmNvbT4KKworICAgICAgICBbQ2Fpcm9dIGZh c3QvY2FudmFzL2NhbnZhcy1pbWFnZVNtb290aGluZ0ZvbyB0ZXN0cyBmYWlsZWQgYWZ0ZXIgcjE5 MDM4My4KKyAgICAgICAgaHR0cHM6Ly9idWdzLndlYmtpdC5vcmcvc2hvd19idWcuY2dpP2lkPTE0 OTc1MgorCisgICAgICAgIFJldmlld2VkIGJ5IE5PQk9EWSAoT09QUyEpLgorCisgICAgICAgIENH J3MgbG93IGludGVycG9sYXRpb24gcXVhbGl0eSBzZXR0aW5nIGlzIGVxdWl2YWxlbnQgdG8gbW9z dCBvdGhlciBicm93c2VycyBkZWZhdWx0IG9yIGhpZ2ggc2V0dGluZ3MuCisKKyAgICAgICAgKiBo dG1sL2NhbnZhcy9DYW52YXNSZW5kZXJpbmdDb250ZXh0MkQuY3BwOgorICAgICAgICAoV2ViQ29y ZTo6Q2FudmFzUmVuZGVyaW5nQ29udGV4dDJEOjpTdGF0ZTo6U3RhdGUpOgorCiAyMDE1LTEwLTAy ICBKYXZpZXIgRmVybmFuZGV6ICA8amZlcm5hbmRlekBpZ2FsaWEuY29tPgogCiAgICAgICAgIFtD U1MgR3JpZCBMYXlvdXRdIFN1cHBvcnQgZm9yIENvbnRlbnQgQWxpZ25tZW50IGluIGdyaWQgbGF5 b3V0CmRpZmYgLS1naXQgYS9Tb3VyY2UvV2ViQ29yZS9odG1sL2NhbnZhcy9DYW52YXNSZW5kZXJp bmdDb250ZXh0MkQuY3BwIGIvU291cmNlL1dlYkNvcmUvaHRtbC9jYW52YXMvQ2FudmFzUmVuZGVy aW5nQ29udGV4dDJELmNwcAppbmRleCA1NWY1YTM3MWFkMjY5N2E4NzMzNWI4NDY1OTE3YjQ0NzU0 OTdjYjU1Li45NTdiZGRiOGUxYjJiNmU4NjdjYjkyZWQ5YjFiYzY2YmIyZmQ4ZmM0IDEwMDY0NAot LS0gYS9Tb3VyY2UvV2ViQ29yZS9odG1sL2NhbnZhcy9DYW52YXNSZW5kZXJpbmdDb250ZXh0MkQu Y3BwCisrKyBiL1NvdXJjZS9XZWJDb3JlL2h0bWwvY2FudmFzL0NhbnZhc1JlbmRlcmluZ0NvbnRl eHQyRC5jcHAKQEAgLTE3Miw3ICsxNzIsMTEgQEAgQ2FudmFzUmVuZGVyaW5nQ29udGV4dDJEOjpT dGF0ZTo6U3RhdGUoKQogICAgICwgaGFzSW52ZXJ0aWJsZVRyYW5zZm9ybSh0cnVlKQogICAgICwg bGluZURhc2hPZmZzZXQoMCkKICAgICAsIGltYWdlU21vb3RoaW5nRW5hYmxlZCh0cnVlKQorI2lm IFVTRShDRykKICAgICAsIGltYWdlU21vb3RoaW5nUXVhbGl0eShTbW9vdGhpbmdRdWFsaXR5OjpM b3cpCisjZWxzZQorICAgICwgaW1hZ2VTbW9vdGhpbmdRdWFsaXR5KFNtb290aGluZ1F1YWxpdHk6 Ok1lZGl1bSkKKyNlbmRpZgogICAgICwgdGV4dEFsaWduKFN0YXJ0VGV4dEFsaWduKQogICAgICwg dGV4dEJhc2VsaW5lKEFscGhhYmV0aWNUZXh0QmFzZWxpbmUpCiAgICAgLCBkaXJlY3Rpb24oRGly ZWN0aW9uOjpJbmhlcml0KQo= 262503 2015-10-05 23:01:42 -0700 2015-10-05 23:48:57 -0700 Patch bug-149752-20151006150111.patch text/plain 1668 hs85.jeong SW5kZXg6IFNvdXJjZS9XZWJDb3JlL0NoYW5nZUxvZwo9PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBTb3VyY2UvV2Vi Q29yZS9DaGFuZ2VMb2cJKHJldmlzaW9uIDE5MDYwNikKKysrIFNvdXJjZS9XZWJDb3JlL0NoYW5n ZUxvZwkod29ya2luZyBjb3B5KQpAQCAtMSwzICsxLDE1IEBACisyMDE1LTEwLTA1ICBIdW5zZW9w IEplb25nICA8aHM4NS5qZW9uZ0BzYW1zdW5nLmNvbT4KKworICAgICAgICBbQ2Fpcm9dIGZhc3Qv Y2FudmFzL2NhbnZhcy1pbWFnZVNtb290aGluZ0ZvbyB0ZXN0cyBmYWlsZWQgYWZ0ZXIgcjE5MDM4 My4KKyAgICAgICAgaHR0cHM6Ly9idWdzLndlYmtpdC5vcmcvc2hvd19idWcuY2dpP2lkPTE0OTc1 MgorCisgICAgICAgIFJldmlld2VkIGJ5IE5PQk9EWSAoT09QUyEpLgorCisgICAgICAgIENHJ3Mg bG93IGludGVycG9sYXRpb24gcXVhbGl0eSBzZXR0aW5nIGlzIGVxdWl2YWxlbnQgdG8gbW9zdCBv dGhlciBicm93c2VycyBkZWZhdWx0IG9yIGhpZ2ggc2V0dGluZ3MuCisKKyAgICAgICAgKiBodG1s L2NhbnZhcy9DYW52YXNSZW5kZXJpbmdDb250ZXh0MkQuY3BwOgorICAgICAgICAoV2ViQ29yZTo6 Q2FudmFzUmVuZGVyaW5nQ29udGV4dDJEOjpTdGF0ZTo6U3RhdGUpOgorCiAyMDE1LTEwLTA1ICBD aHJpcyBEdW1leiAgPGNkdW1lekBhcHBsZS5jb20+CiAKICAgICAgICAgZGF0YTogVVJMcyBzaG91 bGQgbm90IGJlIHByZWxvYWRlZApJbmRleDogU291cmNlL1dlYkNvcmUvaHRtbC9jYW52YXMvQ2Fu dmFzUmVuZGVyaW5nQ29udGV4dDJELmNwcAo9PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBTb3VyY2UvV2ViQ29yZS9o dG1sL2NhbnZhcy9DYW52YXNSZW5kZXJpbmdDb250ZXh0MkQuY3BwCShyZXZpc2lvbiAxOTA1NTYp CisrKyBTb3VyY2UvV2ViQ29yZS9odG1sL2NhbnZhcy9DYW52YXNSZW5kZXJpbmdDb250ZXh0MkQu Y3BwCSh3b3JraW5nIGNvcHkpCkBAIC03MCw2ICs3MCwxMiBAQAogI2luY2x1ZGUgIlNldHRpbmdz LmgiCiAjZW5kaWYKIAorI2lmIFVTRShDRykKKyNkZWZpbmUgRGVmYXVsdFNtb290aGluZ1F1YWxp dHkgU21vb3RoaW5nUXVhbGl0eTo6TG93CisjZWxzZQorI2RlZmluZSBEZWZhdWx0U21vb3RoaW5n UXVhbGl0eSBTbW9vdGhpbmdRdWFsaXR5OjpNZWRpdW0KKyNlbmRpZgorCiBuYW1lc3BhY2UgV2Vi Q29yZSB7CiAKIHVzaW5nIG5hbWVzcGFjZSBIVE1MTmFtZXM7CkBAIC0xNzIsNyArMTc4LDcgQEAg Q2FudmFzUmVuZGVyaW5nQ29udGV4dDJEOjpTdGF0ZTo6U3RhdGUoKQogICAgICwgaGFzSW52ZXJ0 aWJsZVRyYW5zZm9ybSh0cnVlKQogICAgICwgbGluZURhc2hPZmZzZXQoMCkKICAgICAsIGltYWdl U21vb3RoaW5nRW5hYmxlZCh0cnVlKQotICAgICwgaW1hZ2VTbW9vdGhpbmdRdWFsaXR5KFNtb290 aGluZ1F1YWxpdHk6OkxvdykKKyAgICAsIGltYWdlU21vb3RoaW5nUXVhbGl0eShEZWZhdWx0U21v b3RoaW5nUXVhbGl0eSkKICAgICAsIHRleHRBbGlnbihTdGFydFRleHRBbGlnbikKICAgICAsIHRl eHRCYXNlbGluZShBbHBoYWJldGljVGV4dEJhc2VsaW5lKQogICAgICwgZGlyZWN0aW9uKERpcmVj dGlvbjo6SW5oZXJpdCkK 262505 2015-10-05 23:49:05 -0700 2015-10-06 09:35:32 -0700 Patch bug-149752-20151006154833.patch text/plain 6429 hs85.jeong SW5kZXg6IFNvdXJjZS9XZWJDb3JlL0NoYW5nZUxvZwo9PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBTb3VyY2UvV2Vi Q29yZS9DaGFuZ2VMb2cJKHJldmlzaW9uIDE5MDYwOCkKKysrIFNvdXJjZS9XZWJDb3JlL0NoYW5n ZUxvZwkod29ya2luZyBjb3B5KQpAQCAtMSwzICsxLDE1IEBACisyMDE1LTEwLTA1ICBIdW5zZW9w IEplb25nICA8aHM4NS5qZW9uZ0BzYW1zdW5nLmNvbT4KKworICAgICAgICBbQ2Fpcm9dIGZhc3Qv Y2FudmFzL2NhbnZhcy1pbWFnZVNtb290aGluZ0ZvbyB0ZXN0cyBmYWlsZWQgYWZ0ZXIgcjE5MDM4 My4KKyAgICAgICAgaHR0cHM6Ly9idWdzLndlYmtpdC5vcmcvc2hvd19idWcuY2dpP2lkPTE0OTc1 MgorCisgICAgICAgIFJldmlld2VkIGJ5IE5PQk9EWSAoT09QUyEpLgorCisgICAgICAgIENHJ3Mg bG93IGludGVycG9sYXRpb24gcXVhbGl0eSBzZXR0aW5nIGlzIGVxdWl2YWxlbnQgdG8gbW9zdCBv dGhlciBicm93c2VycyBkZWZhdWx0IG9yIGhpZ2ggc2V0dGluZ3MuCisKKyAgICAgICAgKiBodG1s L2NhbnZhcy9DYW52YXNSZW5kZXJpbmdDb250ZXh0MkQuY3BwOgorICAgICAgICAoV2ViQ29yZTo6 Q2FudmFzUmVuZGVyaW5nQ29udGV4dDJEOjpTdGF0ZTo6U3RhdGUpOgorCiAyMDE1LTEwLTA1ICBZ b3Vlbm4gRmFibGV0ICA8eW91ZW5uLmZhYmxldEBjcmYuY2Fub24uZnI+CiAKICAgICAgICAgTWln cmF0ZSBzdHJlYW1zIEFQSSB0byBKUyBCdWlsdGlucwpJbmRleDogU291cmNlL1dlYkNvcmUvaHRt bC9jYW52YXMvQ2FudmFzUmVuZGVyaW5nQ29udGV4dDJELmNwcAo9PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBTb3Vy Y2UvV2ViQ29yZS9odG1sL2NhbnZhcy9DYW52YXNSZW5kZXJpbmdDb250ZXh0MkQuY3BwCShyZXZp c2lvbiAxOTA1NTYpCisrKyBTb3VyY2UvV2ViQ29yZS9odG1sL2NhbnZhcy9DYW52YXNSZW5kZXJp bmdDb250ZXh0MkQuY3BwCSh3b3JraW5nIGNvcHkpCkBAIC03MCw2ICs3MCwxMiBAQAogI2luY2x1 ZGUgIlNldHRpbmdzLmgiCiAjZW5kaWYKIAorI2lmIFVTRShDRykKKyNkZWZpbmUgRGVmYXVsdFNt b290aGluZ1F1YWxpdHkgU21vb3RoaW5nUXVhbGl0eTo6TG93CisjZWxzZQorI2RlZmluZSBEZWZh dWx0U21vb3RoaW5nUXVhbGl0eSBTbW9vdGhpbmdRdWFsaXR5OjpNZWRpdW0KKyNlbmRpZgorCiBu YW1lc3BhY2UgV2ViQ29yZSB7CiAKIHVzaW5nIG5hbWVzcGFjZSBIVE1MTmFtZXM7CkBAIC0xNzIs NyArMTc4LDcgQEAgQ2FudmFzUmVuZGVyaW5nQ29udGV4dDJEOjpTdGF0ZTo6U3RhdGUoKQogICAg ICwgaGFzSW52ZXJ0aWJsZVRyYW5zZm9ybSh0cnVlKQogICAgICwgbGluZURhc2hPZmZzZXQoMCkK ICAgICAsIGltYWdlU21vb3RoaW5nRW5hYmxlZCh0cnVlKQotICAgICwgaW1hZ2VTbW9vdGhpbmdR dWFsaXR5KFNtb290aGluZ1F1YWxpdHk6OkxvdykKKyAgICAsIGltYWdlU21vb3RoaW5nUXVhbGl0 eShEZWZhdWx0U21vb3RoaW5nUXVhbGl0eSkKICAgICAsIHRleHRBbGlnbihTdGFydFRleHRBbGln bikKICAgICAsIHRleHRCYXNlbGluZShBbHBoYWJldGljVGV4dEJhc2VsaW5lKQogICAgICwgZGly ZWN0aW9uKERpcmVjdGlvbjo6SW5oZXJpdCkKSW5kZXg6IExheW91dFRlc3RzL0NoYW5nZUxvZwo9 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09Ci0tLSBMYXlvdXRUZXN0cy9DaGFuZ2VMb2cJKHJldmlzaW9uIDE5MDU1NikKKysr IExheW91dFRlc3RzL0NoYW5nZUxvZwkod29ya2luZyBjb3B5KQpAQCAtMSwzICsxLDE1IEBACisy MDE1LTEwLTA1ICBIdW5zZW9wIEplb25nICA8aHM4NS5qZW9uZ0BzYW1zdW5nLmNvbT4KKworICAg ICAgICBbQ2Fpcm9dIGZhc3QvY2FudmFzL2NhbnZhcy1pbWFnZVNtb290aGluZ0ZvbyB0ZXN0cyBm YWlsZWQgYWZ0ZXIgcjE5MDM4My4KKyAgICAgICAgaHR0cHM6Ly9idWdzLndlYmtpdC5vcmcvc2hv d19idWcuY2dpP2lkPTE0OTc1MgorCisgICAgICAgIFJldmlld2VkIGJ5IE5PQk9EWSAoT09QUyEp LgorCisgICAgICAgIENHJ3MgbG93IGludGVycG9sYXRpb24gcXVhbGl0eSBzZXR0aW5nIGlzIGVx dWl2YWxlbnQgdG8gbW9zdCBvdGhlciBicm93c2VycyBkZWZhdWx0IG9yIGhpZ2ggc2V0dGluZ3Mu CisKKyAgICAgICAgKiBwbGF0Zm9ybS9lZmwvZmFzdC9jYW52YXMvY2FudmFzLWltYWdlU21vb3Ro aW5nUXVhbGl0eS1leHBlY3RlZC50eHQ6IEFkZGVkLgorICAgICAgICAqIHBsYXRmb3JtL2d0ay9m YXN0L2NhbnZhcy9jYW52YXMtaW1hZ2VTbW9vdGhpbmdRdWFsaXR5LWV4cGVjdGVkLnR4dDogQWRk ZWQuCisKIDIwMTUtMTAtMDQgIEFsZXhleSBQcm9za3VyeWFrb3YgIDxhcEBhcHBsZS5jb20+CiAK ICAgICAgICAgTWFyayB0ZXN0cyBhcyBmbGFreSBmb3IgCkluZGV4OiBMYXlvdXRUZXN0cy9wbGF0 Zm9ybS9lZmwvZmFzdC9jYW52YXMvY2FudmFzLWltYWdlU21vb3RoaW5nUXVhbGl0eS1leHBlY3Rl ZC50eHQKPT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PQotLS0gTGF5b3V0VGVzdHMvcGxhdGZvcm0vZWZsL2Zhc3QvY2FudmFz L2NhbnZhcy1pbWFnZVNtb290aGluZ1F1YWxpdHktZXhwZWN0ZWQudHh0CShyZXZpc2lvbiAwKQor KysgTGF5b3V0VGVzdHMvcGxhdGZvcm0vZWZsL2Zhc3QvY2FudmFzL2NhbnZhcy1pbWFnZVNtb290 aGluZ1F1YWxpdHktZXhwZWN0ZWQudHh0CSh3b3JraW5nIGNvcHkpCkBAIC0wLDAgKzEsNDAgQEAK K1Rlc3RzIGZvciB0aGUgaW1hZ2VTbW9vdGhpbmdRdWFsaXR5IGF0dHJpYnV0ZS4KKworT24gc3Vj Y2VzcywgeW91IHdpbGwgc2VlIGEgc2VyaWVzIG9mICJQQVNTIiBtZXNzYWdlcywgZm9sbG93ZWQg YnkgIlRFU1QgQ09NUExFVEUiLgorCisKKworUEFTUyBsb3dDb250ZXh0LmltYWdlU21vb3RoaW5n UXVhbGl0eSBpcyAnbG93JworUEFTUyBtZWRpdW1Db250ZXh0LmltYWdlU21vb3RoaW5nUXVhbGl0 eSBpcyAnbWVkaXVtJworUEFTUyBoaWdoQ29udGV4dC5pbWFnZVNtb290aGluZ1F1YWxpdHkgaXMg J2hpZ2gnCisKK1BBU1MgbG93RGF0YSBpcyBub3QgbWVkaXVtRGF0YQorUEFTUyBtZWRpdW1EYXRh IGlzIG5vdCBoaWdoRGF0YQorUEFTUyBsb3dEYXRhIGlzIG5vdCBoaWdoRGF0YQorCitQQVNTIHNh bXBsZUFscGhhKGxvd0RhdGEpIGlzID49IHNhbXBsZUFscGhhKG1lZGl1bURhdGEpCitQQVNTIHNh bXBsZUFscGhhKG1lZGl1bURhdGEpIGlzID49IHNhbXBsZUFscGhhKGhpZ2hEYXRhKQorCitGQUlM IGRlZmF1bHRDb250ZXh0LmltYWdlU21vb3RoaW5nUXVhbGl0eSBzaG91bGQgYmUgbG93LiBXYXMg bWVkaXVtLgorCitQQVNTIGhpZ2hDb250ZXh0LmltYWdlU21vb3RoaW5nRW5hYmxlZCA9IGZhbHNl OyBoaWdoQ29udGV4dC5pbWFnZVNtb290aGluZ1F1YWxpdHkgaXMgJ2hpZ2gnCitQQVNTIGhpZ2hD b250ZXh0LmltYWdlU21vb3RoaW5nUXVhbGl0eSA9ICdtZWRpdW0nOyBoaWdoQ29udGV4dC5pbWFn ZVNtb290aGluZ1F1YWxpdHkgaXMgJ21lZGl1bScKKworaGlnaENvbnRleHQuaW1hZ2VTbW9vdGhp bmdFbmFibGVkID0gdHJ1ZTsgaGlnaENvbnRleHQuaW1hZ2VTbW9vdGhpbmdRdWFsaXR5ID0gJ2hp Z2gnOworUEFTUyBzY2FsZUltYWdlRGF0YShoaWdoQ2FudmFzLCAnMzIyMycpIGRpZCBub3QgdGhy b3cgZXhjZXB0aW9uLgorUEFTUyBoaWdoQ29udGV4dC5pbWFnZVNtb290aGluZ1F1YWxpdHkgaXMg J2hpZ2gnCitQQVNTIHNjYWxlSW1hZ2VEYXRhKGhpZ2hDYW52YXMsICdiYWRfaW5wdXQnKSBkaWQg bm90IHRocm93IGV4Y2VwdGlvbi4KK1BBU1MgaGlnaENvbnRleHQuaW1hZ2VTbW9vdGhpbmdRdWFs aXR5IGlzICdoaWdoJworUEFTUyBzY2FsZUltYWdlRGF0YShoaWdoQ2FudmFzLCAnTE9XJykgZGlk IG5vdCB0aHJvdyBleGNlcHRpb24uCitQQVNTIGhpZ2hDb250ZXh0LmltYWdlU21vb3RoaW5nUXVh bGl0eSBpcyAnaGlnaCcKK1BBU1Mgc2NhbGVJbWFnZURhdGEoaGlnaENhbnZhcywgJ01lZGl1bScp IGRpZCBub3QgdGhyb3cgZXhjZXB0aW9uLgorUEFTUyBoaWdoQ29udGV4dC5pbWFnZVNtb290aGlu Z1F1YWxpdHkgaXMgJ2hpZ2gnCisKK2hpZ2hDb250ZXh0LnNhdmUoKTsgaGlnaENvbnRleHQuaW1h Z2VTbW9vdGhpbmdRdWFsaXR5ID0gJ21lZGl1bSc7CitQQVNTIGhpZ2hDb250ZXh0LnJlc3RvcmUo KTsgaGlnaENvbnRleHQuaW1hZ2VTbW9vdGhpbmdRdWFsaXR5IGlzICdoaWdoJworUEFTUyBzY2Fs ZUltYWdlRGF0YShoaWdoQ2FudmFzLCBoaWdoQ2FudmFzLmltYWdlU21vb3RoaW5nUXVhbGl0eSk7 IGlzIGhpZ2hEYXRhCisKK1BBU1Mgc3VjY2Vzc2Z1bGx5UGFyc2VkIGlzIHRydWUKKworVEVTVCBD T01QTEVURQorIApJbmRleDogTGF5b3V0VGVzdHMvcGxhdGZvcm0vZ3RrL2Zhc3QvY2FudmFzL2Nh bnZhcy1pbWFnZVNtb290aGluZ1F1YWxpdHktZXhwZWN0ZWQudHh0Cj09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIExh eW91dFRlc3RzL3BsYXRmb3JtL2d0ay9mYXN0L2NhbnZhcy9jYW52YXMtaW1hZ2VTbW9vdGhpbmdR dWFsaXR5LWV4cGVjdGVkLnR4dAkocmV2aXNpb24gMCkKKysrIExheW91dFRlc3RzL3BsYXRmb3Jt L2d0ay9mYXN0L2NhbnZhcy9jYW52YXMtaW1hZ2VTbW9vdGhpbmdRdWFsaXR5LWV4cGVjdGVkLnR4 dAkod29ya2luZyBjb3B5KQpAQCAtMCwwICsxLDQwIEBACitUZXN0cyBmb3IgdGhlIGltYWdlU21v b3RoaW5nUXVhbGl0eSBhdHRyaWJ1dGUuCisKK09uIHN1Y2Nlc3MsIHlvdSB3aWxsIHNlZSBhIHNl cmllcyBvZiAiUEFTUyIgbWVzc2FnZXMsIGZvbGxvd2VkIGJ5ICJURVNUIENPTVBMRVRFIi4KKwor CisKK1BBU1MgbG93Q29udGV4dC5pbWFnZVNtb290aGluZ1F1YWxpdHkgaXMgJ2xvdycKK1BBU1Mg bWVkaXVtQ29udGV4dC5pbWFnZVNtb290aGluZ1F1YWxpdHkgaXMgJ21lZGl1bScKK1BBU1MgaGln aENvbnRleHQuaW1hZ2VTbW9vdGhpbmdRdWFsaXR5IGlzICdoaWdoJworCitQQVNTIGxvd0RhdGEg aXMgbm90IG1lZGl1bURhdGEKK1BBU1MgbWVkaXVtRGF0YSBpcyBub3QgaGlnaERhdGEKK1BBU1Mg bG93RGF0YSBpcyBub3QgaGlnaERhdGEKKworUEFTUyBzYW1wbGVBbHBoYShsb3dEYXRhKSBpcyA+ PSBzYW1wbGVBbHBoYShtZWRpdW1EYXRhKQorUEFTUyBzYW1wbGVBbHBoYShtZWRpdW1EYXRhKSBp cyA+PSBzYW1wbGVBbHBoYShoaWdoRGF0YSkKKworRkFJTCBkZWZhdWx0Q29udGV4dC5pbWFnZVNt b290aGluZ1F1YWxpdHkgc2hvdWxkIGJlIGxvdy4gV2FzIG1lZGl1bS4KKworUEFTUyBoaWdoQ29u dGV4dC5pbWFnZVNtb290aGluZ0VuYWJsZWQgPSBmYWxzZTsgaGlnaENvbnRleHQuaW1hZ2VTbW9v dGhpbmdRdWFsaXR5IGlzICdoaWdoJworUEFTUyBoaWdoQ29udGV4dC5pbWFnZVNtb290aGluZ1F1 YWxpdHkgPSAnbWVkaXVtJzsgaGlnaENvbnRleHQuaW1hZ2VTbW9vdGhpbmdRdWFsaXR5IGlzICdt ZWRpdW0nCisKK2hpZ2hDb250ZXh0LmltYWdlU21vb3RoaW5nRW5hYmxlZCA9IHRydWU7IGhpZ2hD b250ZXh0LmltYWdlU21vb3RoaW5nUXVhbGl0eSA9ICdoaWdoJzsKK1BBU1Mgc2NhbGVJbWFnZURh dGEoaGlnaENhbnZhcywgJzMyMjMnKSBkaWQgbm90IHRocm93IGV4Y2VwdGlvbi4KK1BBU1MgaGln aENvbnRleHQuaW1hZ2VTbW9vdGhpbmdRdWFsaXR5IGlzICdoaWdoJworUEFTUyBzY2FsZUltYWdl RGF0YShoaWdoQ2FudmFzLCAnYmFkX2lucHV0JykgZGlkIG5vdCB0aHJvdyBleGNlcHRpb24uCitQ QVNTIGhpZ2hDb250ZXh0LmltYWdlU21vb3RoaW5nUXVhbGl0eSBpcyAnaGlnaCcKK1BBU1Mgc2Nh bGVJbWFnZURhdGEoaGlnaENhbnZhcywgJ0xPVycpIGRpZCBub3QgdGhyb3cgZXhjZXB0aW9uLgor UEFTUyBoaWdoQ29udGV4dC5pbWFnZVNtb290aGluZ1F1YWxpdHkgaXMgJ2hpZ2gnCitQQVNTIHNj YWxlSW1hZ2VEYXRhKGhpZ2hDYW52YXMsICdNZWRpdW0nKSBkaWQgbm90IHRocm93IGV4Y2VwdGlv bi4KK1BBU1MgaGlnaENvbnRleHQuaW1hZ2VTbW9vdGhpbmdRdWFsaXR5IGlzICdoaWdoJworCito aWdoQ29udGV4dC5zYXZlKCk7IGhpZ2hDb250ZXh0LmltYWdlU21vb3RoaW5nUXVhbGl0eSA9ICdt ZWRpdW0nOworUEFTUyBoaWdoQ29udGV4dC5yZXN0b3JlKCk7IGhpZ2hDb250ZXh0LmltYWdlU21v b3RoaW5nUXVhbGl0eSBpcyAnaGlnaCcKK1BBU1Mgc2NhbGVJbWFnZURhdGEoaGlnaENhbnZhcywg aGlnaENhbnZhcy5pbWFnZVNtb290aGluZ1F1YWxpdHkpOyBpcyBoaWdoRGF0YQorCitQQVNTIHN1 Y2Nlc3NmdWxseVBhcnNlZCBpcyB0cnVlCisKK1RFU1QgQ09NUExFVEUKKyAK