searching for reviewer(seth)
8de973bd95adf8b7106bb196343ebb7bf420aba8: Bug 1284350. Reland on top of backouts "Bug 1282710 - part 2 - Plumb the visibility event from nsIFrame to nsIDOMMediaElemnt; r=seth" r=kaku
Kaku Kuo <tkuo@mozilla.com> - Wed, 29 Jun 2016 17:36:24 +0800 - rev 313497
Bug 1284350. Reland on top of backouts "Bug 1282710 - part 2 - Plumb the visibility event from nsIFrame to nsIDOMMediaElemnt; r=seth" r=kaku
f6bd48b932a5546aa2606fcdefaf81f113897706: Bug 1284350. Backed out changeset 103dc4eddacf (Bug 1282710 - part 2 - Plumb the visibility event from nsIFrame to nsIDOMMediaElemnt; r=seth)
Timothy Nikkel <tnikkel@gmail.com> - Tue, 26 Jul 2016 17:43:58 -0500 - rev 313483
Bug 1284350. Backed out changeset 103dc4eddacf (Bug 1282710 - part 2 - Plumb the visibility event from nsIFrame to nsIDOMMediaElemnt; r=seth)
85cfd0bd6eb0c321ccfa105f49e2f607d88b79f9: Bug 1260324. Don't draw garbage to the screen if an image doesn't happen to be decoded. r=seth
Timothy Nikkel <tnikkel@gmail.com> - Mon, 22 Aug 2016 21:15:38 -0500 - rev 310966
Bug 1260324. Don't draw garbage to the screen if an image doesn't happen to be decoded. r=seth Layout has been using imgIContainer::IsOpaque to determine if the image will draw opaquely to all pixels it covers, and doing culling based on this. However imgIContainer::IsOpaque doesn't guarantee anything. It only describes if the image, when in a decoded state, has all opaque pixels. So if the image doesn't have fully decoded frames around (because they got discarded) it may not draw opaquely to all of its pixels. So we create a new function that first checks if there is a fully decoded frame.
54933b5b96f1673a94a6463baf2c9cc643973e2d: Bug 1260324. Don't draw garbage to the screen if an image doesn't happen to be decoded. r=seth
Timothy Nikkel <tnikkel@gmail.com> - Mon, 22 Aug 2016 21:15:38 -0500 - rev 310752
Bug 1260324. Don't draw garbage to the screen if an image doesn't happen to be decoded. r=seth Layout has been using imgIContainer::IsOpaque to determine if the image will draw opaquely to all pixels it covers, and doing culling based on this. However imgIContainer::IsOpaque doesn't guarantee anything. It only describes if the image, when in a decoded state, has all opaque pixels. So if the image doesn't have fully decoded frames around (because they got discarded) it may not draw opaquely to all of its pixels. So we create a new function that first checks if there is a fully decoded frame.
3eade32e18c4446960aebf28039c6d116d63e255: Bug 1293577 - Part 2. Add mutex to MetadataDecodingTask to protect from decoding races. r=seth
Andrew Osmond <aosmond@mozilla.com> - Thu, 18 Aug 2016 10:23:42 -0400 - rev 310124
Bug 1293577 - Part 2. Add mutex to MetadataDecodingTask to protect from decoding races. r=seth
1d4993c450667a9fbe3acfaaba4c094b6a20bf23: Bug 1291986 - Part 01 (libpng) Update embedded libpng to version 1.6.24. r=seth
Glenn Randers-Pehrson <glennrp+bmo@gmail.com> - Thu, 04 Aug 2016 20:10:00 +0200 - rev 308923
Bug 1291986 - Part 01 (libpng) Update embedded libpng to version 1.6.24. r=seth
64863b1fcf6470de45a6f3979dd3b1709e4b10e8: Bug 1293577 - Protect the image decoding task path with a mutex to avoid race conditions. r=seth
Andrew Osmond <aosmond@mozilla.com> - Wed, 10 Aug 2016 07:35:07 -0400 - rev 308883
Bug 1293577 - Protect the image decoding task path with a mutex to avoid race conditions. r=seth
ae34bd5397a98933a18eb538a5e41eb2393bdf74: Bug 1270680 - Part 2: Tests that make sure image cache respect originAttributes. r=jdm,seth
Jonathan Hao <jhao@mozilla.com> - Fri, 05 Aug 2016 00:10:00 +0200 - rev 308674
Bug 1270680 - Part 2: Tests that make sure image cache respect originAttributes. r=jdm,seth MozReview-Commit-ID: LbGy6MfZGEL
d289b7a8ad1f3374897bff0d2d59c4b1179f81cd: Bug 1292632 part 3: Include <algorithm> in image/SourceBuffer.h, to provide std::min. r=seth
Daniel Holbert <dholbert@cs.stanford.edu> - Fri, 05 Aug 2016 17:09:48 -0700 - rev 308460
Bug 1292632 part 3: Include <algorithm> in image/SourceBuffer.h, to provide std::min. r=seth MozReview-Commit-ID: DJvb7Sz4efF
461d7103db3a6dc9d9f9ea4b54c48eef1202cdef: Bug 1292632 part 2: Add missing "using" decl for IntRect in IDecodingTask.cpp. r=seth
Daniel Holbert <dholbert@cs.stanford.edu> - Fri, 05 Aug 2016 17:09:47 -0700 - rev 308459
Bug 1292632 part 2: Add missing "using" decl for IntRect in IDecodingTask.cpp. r=seth MozReview-Commit-ID: 6OeckqKu7k7
27c0dd5afde663dd19e094c022e8fb2149177940: Bug 1292632 part 1: Remove a stale/unused #include from SurfaceCache.cpp. r=seth
Daniel Holbert <dholbert@cs.stanford.edu> - Fri, 05 Aug 2016 17:09:45 -0700 - rev 308458
Bug 1292632 part 1: Remove a stale/unused #include from SurfaceCache.cpp. r=seth MozReview-Commit-ID: feZgLQ1R7u
457ebb2b73adfed9dc9ff8c1ccaa38723471b59e: Bug 664717 - Part 3: Include file promise when copying an image in Windows. r=seth
Hector Zhao <bzhao@mozilla.com> - Thu, 30 Jun 2016 16:23:30 +0800 - rev 307813
Bug 664717 - Part 3: Include file promise when copying an image in Windows. r=seth MozReview-Commit-ID: ANSey2vGwaG
7841d3498cee851b47971938669a199b8be6c724: Bug 1283961 - Part 2: Remove limits on PNG image dimensions (nsPNGDecoder). r=seth
Glenn Randers-Pehrson <glennrp+bmo@gmail.com> - Thu, 28 Jul 2016 07:37:00 -0400 - rev 307154
Bug 1283961 - Part 2: Remove limits on PNG image dimensions (nsPNGDecoder). r=seth
0e271858ba5e975617d82b8ebfe7b775258171b1: Bug 1283961 - Part 1: Remove limits on PNG image dimensions (libpng). r=seth
Glenn Randers-Pehrson <glennrp+bmo@gmail.com> - Thu, 21 Jul 2016 11:13:00 -0400 - rev 307153
Bug 1283961 - Part 1: Remove limits on PNG image dimensions (libpng). r=seth
343a4eced34bdb2c610da8d439da348f1e159eaa: Bug 1283462 - Fix race condition when notifying on image decoding progress. r=seth
Andrew Osmond <aosmond@mozilla.com> - Fri, 22 Jul 2016 11:04:09 -0400 - rev 306984
Bug 1283462 - Fix race condition when notifying on image decoding progress. r=seth
8e868621be2f70ebe0846ee7b5a66dce010140f0: Bug 1273455: Let CSS Image surface caches expire, rather than immediately discarding. r=seth
Jet Villegas <jvillegas@mozilla.com> - Tue, 19 Jul 2016 13:29:21 +0800 - rev 305678
Bug 1273455: Let CSS Image surface caches expire, rather than immediately discarding. r=seth
385feddefee4bb63b7899d5c87bb3ef3b7bb38ba: Bug 1285006 - Remove IMAGE_MAX_DECODE_COUNT Telemetry r=gfritzsche, seth
Chris H-C <chutten@mozilla.com> - Wed, 13 Jul 2016 11:53:00 +0200 - rev 305032
Bug 1285006 - Remove IMAGE_MAX_DECODE_COUNT Telemetry r=gfritzsche, seth It clears a histogram on the child process which is no longer allowed (see bug 1218576) MozReview-Commit-ID: GXWN7qpalAc
103dc4eddacf4d128ae5bd6516485ee53d61c967: Bug 1282710 - part 2 - Plumb the visibility event from nsIFrame to nsIDOMMediaElemnt; r=seth
Kaku Kuo <tkuo@mozilla.com> - Wed, 29 Jun 2016 17:36:24 +0800 - rev 303635
Bug 1282710 - part 2 - Plumb the visibility event from nsIFrame to nsIDOMMediaElemnt; r=seth MozReview-Commit-ID: DnlTghYdrug
13cd8e7c973a86d612e04819118d2ca1234059ef: Bug 1282496 - memset unoptimized images with RGBX surface format to 0xFF. r=seth
Mason Chang <mchang@mozilla.com> - Thu, 30 Jun 2016 18:59:29 -0700 - rev 303336
Bug 1282496 - memset unoptimized images with RGBX surface format to 0xFF. r=seth
7339ed5e03791159421bc91995ebc7160d4976b2: Bug 1282670, part 3: Make several imagelib .cpp files #inclue their corresponding .h file as their very first #include, per Gecko convention. r=seth
Daniel Holbert <dholbert@cs.stanford.edu> - Thu, 30 Jun 2016 10:27:03 -0700 - rev 303313
Bug 1282670, part 3: Make several imagelib .cpp files #inclue their corresponding .h file as their very first #include, per Gecko convention. r=seth (In some cases, I've left "ImageLogging.h" being included before the corresponding .h file, because I ran across a warning comment saying that it needs to be included before any IPDL-generated files & anything that includes prlog.h; and it seems possible that Foo.cpp's corresponding Foo.h file could include such headers now or in the future.) MozReview-Commit-ID: HPvUVj8YuKc
02fe9f94bde1e323cfabcc2e679ba431fdb9dea8: Bug 1282670, part 2: Remove unnecessary/redundant string-related #include & "using" declarations in imagelib. r=seth
Daniel Holbert <dholbert@cs.stanford.edu> - Thu, 30 Jun 2016 10:25:35 -0700 - rev 303312
Bug 1282670, part 2: Remove unnecessary/redundant string-related #include & "using" declarations in imagelib. r=seth MozReview-Commit-ID: 1hgnUosiOFU
a7d76de3b61befc41c027e64178af57a21efba23: Bug 1282670, part 1: Add missing #include & "using" declarations in imagelib. r=seth
Daniel Holbert <dholbert@cs.stanford.edu> - Thu, 30 Jun 2016 10:24:27 -0700 - rev 303311
Bug 1282670, part 1: Add missing #include & "using" declarations in imagelib. r=seth I found these issues locally by moving all of imagelib's .cpp files to SOURCES instead of UNIFIED_SOURCES. (That change isn't part of this patch, though.) MozReview-Commit-ID: 97Xpfu8eFE6
269a48e6757917fdad3780f1f48dc99f4b2044ca: Bug 1282496 - memset unoptimized images with RGBX surface format to 0xFF when using the skia backend. r=seth
Mason Chang <mchang@mozilla.com> - Mon, 27 Jun 2016 17:45:55 -0700 - rev 302815
Bug 1282496 - memset unoptimized images with RGBX surface format to 0xFF when using the skia backend. r=seth
a21989d264ee66319e274891a12c390482dd9345: Bug 1233086, Part 4 - Gecko style fixes. r=seth
Kit Cambridge <kcambridge@mozilla.com> - Fri, 04 Mar 2016 10:20:26 -0800 - rev 302785
Bug 1233086, Part 4 - Gecko style fixes. r=seth MozReview-Commit-ID: 8QtO5TDODFD
e6fc6996103a59b11dcaa3a6c454eb41cdc810f1: Bug 1233086, Part 1 - Unify alert image loading. r=seth
Kit Cambridge <kcambridge@mozilla.com> - Mon, 27 Jun 2016 08:27:11 -0700 - rev 302782
Bug 1233086, Part 1 - Unify alert image loading. r=seth MozReview-Commit-ID: DSbJfH2FjYX
45fb2585696d35cf628486eac7022aaed16d9939: Bug 1281419 - ignore negative icon size in URI. r=seth
Lee Salzman <lsalzman@mozilla.com> - Thu, 23 Jun 2016 10:47:36 -0400 - rev 302494
Bug 1281419 - ignore negative icon size in URI. r=seth
c84be565e4e86d51e30b51a03db8980deb13e92d: Bug 1275901 - Update in-tree libpng to version 1.6.23 (part 2, old-configure.in). r=seth
Glenn Randers-Pehrson <glennrp+bmo@gmail.com> - Thu, 09 Jun 2016 12:12:00 -0400 - rev 302133
Bug 1275901 - Update in-tree libpng to version 1.6.23 (part 2, old-configure.in). r=seth
7f9d9dfa09d2d3b27bd3f3cc45aae160cf52fb5d: Bug 1275901 - Update in-tree libpng to version 1.6.23. r=seth
Glenn Randers-Pehrson <glennrp+bmo@gmail.com> - Thu, 09 Jun 2016 12:05:00 -0400 - rev 302132
Bug 1275901 - Update in-tree libpng to version 1.6.23. r=seth
b64db3b38fd8fa0bad48acfc0dc4de52a23f9e5b: Bug 1278927 - TSan: data race image/imgFrame.cpp on hasCheckedOptimize. r=seth.
Julian Seward <jseward@acm.org> - Mon, 20 Jun 2016 15:46:00 +0200 - rev 302087
Bug 1278927 - TSan: data race image/imgFrame.cpp on hasCheckedOptimize. r=seth.
8d51d4e1930f5df3d99e714513400b7e76a90306: Bug 1274516 - Add a helper to iterate over all ContentParents and use it to clear the image cache in both the parent and child processes. r=billm r=seth
Blake Kaplan <mrbkap@gmail.com> - Mon, 06 Jun 2016 15:23:43 -0700 - rev 301587
Bug 1274516 - Add a helper to iterate over all ContentParents and use it to clear the image cache in both the parent and child processes. r=billm r=seth
026cf6432f4473549aa6193272764a06be33a1c3: Bug 1204392 - Use StreamingLexer in the GIF decoder. r=njn,seth
Seth Fowler <mark.seth.fowler@gmail.com> - Wed, 08 Jun 2016 12:17:17 -0700 - rev 301185
Bug 1204392 - Use StreamingLexer in the GIF decoder. r=njn,seth
20ab9d3966135fb8e3b37660a8c85bf394ebdbda: Bug 1270680 - Part 2: Tests that make sure image cache respect originAttributes. r=jdm,seth
Jonathan Hao <jhao@mozilla.com> - Mon, 06 Jun 2016 03:39:00 +0200 - rev 300803
Bug 1270680 - Part 2: Tests that make sure image cache respect originAttributes. r=jdm,seth MozReview-Commit-ID: LbGy6MfZGEL
d08891438e5835be6bdf3bbb6de932c845a73604: Bug 1270680 - Part 1: Double-key the image cache by origin attribute. r=jdm,seth
Jonathan Hao <jhao@mozilla.com> - Wed, 01 Jun 2016 23:07:00 +0200 - rev 300802
Bug 1270680 - Part 1: Double-key the image cache by origin attribute. r=jdm,seth
50e32be6552151fd71dd828b30b24749b5fbf4d4: Bug 1277122 - Add missing null checks for pixman_image_create_bits(). r=seth,lsalzman.
Nicholas Nethercote <nnethercote@mozilla.com> - Wed, 01 Jun 2016 14:27:30 +1000 - rev 300051
Bug 1277122 - Add missing null checks for pixman_image_create_bits(). r=seth,lsalzman.
2e2085781ab223642bb266ff729a7569fdaf1858: Bug 1261964 part 2: Add reftest for downscaling a moz-icon image. r=seth
Daniel Holbert <dholbert@cs.stanford.edu> - Fri, 27 May 2016 14:03:35 -0700 - rev 299409
Bug 1261964 part 2: Add reftest for downscaling a moz-icon image. r=seth MozReview-Commit-ID: 61m5k5mYSWN
459f06fccd1c628221cfd6676363a9a3788c9628: Bug 1262066 - GFX: 2D: Use ConvolveHorizontally1_LS3. r=seth
Heiher <r@hev.cc> - Mon, 23 May 2016 10:30:39 +0800 - rev 298460
Bug 1262066 - GFX: 2D: Use ConvolveHorizontally1_LS3. r=seth --- gfx/2d/convolver.cpp | 4 ++++ 1 file changed, 4 insertions(+)
962463926b903521f4872a4e810da4eb20d7ad6c: Bug 1271002. Notify right away after getting the first frame of an animated image. r=seth
Timothy Nikkel <tnikkel@gmail.com> - Fri, 13 May 2016 21:32:21 -0500 - rev 297445
Bug 1271002. Notify right away after getting the first frame of an animated image. r=seth The decoding loop in Decoder::Decode only pauses to report progress when it runs out of bytes to decode. So for long animated images where the network is keeping up with decoding it will be a relatively long time until we deliver the first frame complete notification and corresponding invalidation. In most cases this shouldn't be too expensive as it is just dispatching a runnable to the main thread from the decoding thread.
2ac62de7365d1557e3f2f53900f587cf4bc68cd9: Bug 1271002. Notify right away after getting the first frame of an animated image. r=seth
Timothy Nikkel <tnikkel@gmail.com> - Fri, 13 May 2016 16:19:55 -0500 - rev 297419
Bug 1271002. Notify right away after getting the first frame of an animated image. r=seth The decoding loop in Decoder::Decode only pauses to report progress when it runs out of bytes to decode. So for long animated images where the network is keeping up with decoding it will be a relatively long time until we deliver the first frame complete notification and corresponding invalidation. In most cases this shouldn't be too expensive as it is just dispatching a runnable to the main thread from the decoding thread.
1db2431203b3d2f31dd323efe6cdfb404375ca9a: Bug 1267260 - Change CanvasImageCache to lookup cache images based on imgIContainer instead of imgIRequest. r=seth
Mason Chang <mchang@mozilla.com> - Tue, 10 May 2016 08:00:06 -0700 - rev 296910
Bug 1267260 - Change CanvasImageCache to lookup cache images based on imgIContainer instead of imgIRequest. r=seth
8723a2bd8ecf33230cdfc8259c5501cb60bae8a1: Bug 1251644 - Snap object-contain fit if we were going to reduce the size by less than one css pixel. r=seth
Matt Woodrow <mwoodrow@mozilla.com> - Thu, 07 Apr 2016 18:51:45 +1200 - rev 296841
Bug 1251644 - Snap object-contain fit if we were going to reduce the size by less than one css pixel. r=seth
51922d1ebfa13a4932ac4c6dfd6010a39d145812: Bug 1251644 - Snap object-contain fit if we were going to reduce the size by less than one css pixel. r=seth
Matt Woodrow <mwoodrow@mozilla.com> - Tue, 10 May 2016 13:25:20 +1200 - rev 296743
Bug 1251644 - Snap object-contain fit if we were going to reduce the size by less than one css pixel. r=seth
830de7d2dc769f8f20a7ad96acb5a6f169fc0a09: Bug 1265053 - don't needlessly construct nsAutoCString temporaries in image gtests; r=seth
Nathan Froyd <froydnj@gmail.com> - Mon, 09 May 2016 13:10:09 -0400 - rev 296681
Bug 1265053 - don't needlessly construct nsAutoCString temporaries in image gtests; r=seth Passing nsDependentCString gives us the same effect with less memory traffic.
be4a6303512e8062c3bfb0a33ffc7440af203ae3: Bug 1269271 - Remove SEC_NORMAL from image/decoders (r=seth)
Christoph Kerschbaumer <ckerschb@christophkerschbaumer.com> - Sun, 08 May 2016 20:05:51 +0200 - rev 296560
Bug 1269271 - Remove SEC_NORMAL from image/decoders (r=seth)
45acc6a022912f3be4c79f13efa584365d848888: Bug 1265744 - Make the "RasterImage::Init failed"/"VectorImage::Init failed" messages include the failing URI. r=seth
Jonathan Watt <jwatt@jwatt.org> - Tue, 03 May 2016 17:16:31 +0100 - rev 296557
Bug 1265744 - Make the "RasterImage::Init failed"/"VectorImage::Init failed" messages include the failing URI. r=seth
01b755909dc18b700bd88eb616862e4872a6daf9: Bug 1267272 - NULL pointer mozilla::dom::SVGSVGElement::GetIntrinsicHeight. r=seth
Robert Longson <longsonr@gmail.com> - Sun, 08 May 2016 11:17:46 +0100 - rev 296555
Bug 1267272 - NULL pointer mozilla::dom::SVGSVGElement::GetIntrinsicHeight. r=seth
8fa8077e9c5f8610cf55f084930b8650b3710b1f: Bug 1270997. Send invalidations for the first image frame properly. r=seth
Timothy Nikkel <tnikkel@gmail.com> - Sat, 07 May 2016 15:54:39 -0500 - rev 296544
Bug 1270997. Send invalidations for the first image frame properly. r=seth This was caused by http://hg.mozilla.org/mozilla-central/rev/167ceb965079 (bug 1194059). Before that changeset mIsAnimated meant "we currently have more than one frame". After that changeset mIsAnimated was replaced with HasAnimation(). HasAnimation() just looks at the metadata to see if the image is animated. That changeset had the effect of always detected if an image is animated during the metadata decode. Therefore during a full decode we always know the image is animated, even before we've decoded two or more frames. The fix is to go back to using the actual current frame count to manage invalidations.
a6e3503205e87001cfb0b53b457bdaa595fd9521: Bug 1206961 - Use channel->AsyncOpen2() in image/imgLoader.cpp; removing security checks from the callsite reveals that we have to pass the accurate contentPolicyType to ValidateEntry (r=seth,bz)
Christoph Kerschbaumer <mozilla@christophkerschbaumer.com> - Wed, 27 Apr 2016 19:40:56 +0200 - rev 295188
Bug 1206961 - Use channel->AsyncOpen2() in image/imgLoader.cpp; removing security checks from the callsite reveals that we have to pass the accurate contentPolicyType to ValidateEntry (r=seth,bz)
3e12df7cefe4cf2e0b8dc11867e1008b21aa135a: Bug 1253995 - Display stale image in nsImageFrame if we have a new src but haven't decoded it yet - r=seth
Edwin Flores <eflores@mozilla.com> - Fri, 22 Apr 2016 14:08:25 +0100 - rev 294546
Bug 1253995 - Display stale image in nsImageFrame if we have a new src but haven't decoded it yet - r=seth
1258adc301a8a61b1cb401c28144202ba35825ea: Bug 1256999 - Use nsIDocument for ImageCacheKey. r=bz r=seth
Christoph Kerschbaumer <mozilla@christophkerschbaumer.com> - Mon, 11 Apr 2016 11:00:03 +0200 - rev 292589
Bug 1256999 - Use nsIDocument for ImageCacheKey. r=bz r=seth
3b53ecaae8f278470d5551c377d3580ff8af9d05: Bug 1256999 - Pass the right context to new channels for image loads. r=bz r=seth
Christoph Kerschbaumer <mozilla@christophkerschbaumer.com> - Mon, 11 Apr 2016 10:58:03 +0200 - rev 292588
Bug 1256999 - Pass the right context to new channels for image loads. r=bz r=seth