After allow Flash plugin command flash content plays only in new tab or after browser restart
Reported by
ivankosi...@gmail.com,
Jun 22 2017
|
||||||||||||
Issue descriptionUserAgent: Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.109 Safari/537.36 Example URL: http://www.dhs.state.il.us/accessibility/tests/flash/video.html Steps to reproduce the problem: 1. Option chrome://settings/content/flash Ask First have to be enabled 2. Open page with flash (e.g. http://www.dhs.state.il.us/accessibility/tests/flash/video.html ) 3. Click on link in frase "Please install or enable Adobe Flash Player to view this content." (Flash player was already installed in system) 4. You will see "Run flash" dialog. Click Allow. Page will be reloaded. But no flash content. What is the expected behavior? After reload of page, flash content have to be displayed. What went wrong? No flash content after page reload. We can see the flash content only if we will open new tab with the same URL. Or if we will restart the browser. Did this work before? Yes 60.0.3110.0 Is it a problem with Flash or HTML5? Flash Does this work in other browsers? N/A Chrome version: 61.0.3126.0 Channel: dev OS Version: Windows 7 Flash Version: 26.0.0.131 Contents of chrome://gpu: Graphics Feature Status Canvas: Hardware accelerated CheckerImaging: Enabled Flash: Hardware accelerated Flash Stage3D: Hardware accelerated Flash Stage3D Baseline profile: Hardware accelerated Compositing: Hardware accelerated Multiple Raster Threads: Enabled Native GpuMemoryBuffers: Software only. Hardware acceleration disabled Rasterization: Hardware accelerated Video Decode: Hardware accelerated Video Encode: Hardware accelerated WebGL: Hardware accelerated WebGL2: Hardware accelerated Driver Bug Workarounds clear_uniforms_before_first_program_use decode_encode_srgb_for_generatemipmap disable_accelerated_vpx_decode disable_direct_composition disable_discard_framebuffer disable_dxgi_zero_copy_video disable_framebuffer_cmaa exit_on_context_lost force_cube_complete scalarize_vec_and_mat_constructor_args texsubimage_faster_than_teximage Problems Detected Some drivers are unable to reset the D3D device in the GPU process sandbox Applied Workarounds: exit_on_context_lost TexSubImage is faster for full uploads on ANGLE Applied Workarounds: texsubimage_faster_than_teximage Clear uniforms before first program use on all platforms: 124764, 349137 Applied Workarounds: clear_uniforms_before_first_program_use Always rewrite vec/mat constructors to be consistent: 398694 Applied Workarounds: scalarize_vec_and_mat_constructor_args ANGLE crash on glReadPixels from incomplete cube map texture: 518889 Applied Workarounds: force_cube_complete Framebuffer discarding can hurt performance on non-tilers: 570897 Applied Workarounds: disable_discard_framebuffer Direct composition flashes black initially on Win <10: 588588 Applied Workarounds: disable_direct_composition Zero copy DXGI video hangs on shutdown on Win < 8.1: 621190 Applied Workarounds: disable_dxgi_zero_copy_video Limited enabling of Chromium GL_INTEL_framebuffer_CMAA: 535198 Applied Workarounds: disable_framebuffer_cmaa Disable KHR_blend_equation_advanced until cc shaders are updated: 661715 Decode and Encode before generateMipmap for srgb format textures on Windows: 634519 Applied Workarounds: decode_encode_srgb_for_generatemipmap Zero-copy DXGI video hangs or displays incorrect colors on AMD drivers: 623029 Applied Workarounds: disable_dxgi_zero_copy_video VPx decoding isn't supported well before Windows 10 creators update.: 616318, 667532 Applied Workarounds: disable_accelerated_vpx_decode Native GpuMemoryBuffers have been disabled, either via about:flags or command line. Disabled Features: native_gpu_memory_buffers
,
Jun 22 2017
,
Jun 22 2017
i can't repro this on windows chrome 60.0.3112.32 (beta 64-bit).
,
Jun 23 2017
Tested this issue on Ubuntu 14.04, Windows-10 and Mac OS 10.12.5 using chrome latest dev 61.0.3138.0. This issue is not able to reproduce consistently, Able to reproduce this issue after upgrading the chrome to latest canary version but not able to reproduce on the same build after installing the market dev builds manually. Note: Attaching screen-cast of both the behavior on same version for reference. Due to inconsistent repro unable to find the regression range to provide bisect information. So, removing bisect label please feel free to add it back if it's required. Thanks!
,
Jun 23 2017
-RB, since its not reproducible.
,
Jun 30 2017
The bug is easily reproducible by removing your website domain from the flash-allowed list (parameters > content parameter > flash) and reloading chrome. The "navigator.plugins" list is not updated when flash is activated (there should be an item "Shockwave Flash something" in the list, like in previous versions of chrome). Since most (all) of flash detection plugins check this "navigator.plugins" entry, it always fail after enabling flash in chrome 61 if the browser is not reloaded.
,
Jun 30 2017
I am not able to repro the bug on windows + chrome 60.0.3112.40. However, I was able to repro this on M61.0.3145.0: [STARTING STATE] -Flash set to ASK in chrome://settings/content/flash -no exceptions listed [REPRO] 1. browse to www.dhs.state.il.us/accessibility/tests/flash/video.html 2. check flash not in navigator.plugins - it's not. 3. click on link to download flash -> "Run Flash" prompt shown 4. click Allow [RESULT] -page appears to refresh. -settings exception is logged now for ALLOW on www.dhs.state.il.us -but no navigator.plugins listing for "Shockwave Flash" -and clicking on Flash link is not intercepted anymore - i navigate to flash player download. As mentioned by frederic in c#6 : you have to close the tab and browse back to the site in new tab. (using existing session no longer works). repro'd this on crunchyroll.com too. @tommycli : something pretty wonky regressed between 60 & 61. over to you to investigate.
,
Jun 30 2017
requesting bisect : repro's on 61.0.3145.0, does not on 60.0.3112.40
,
Jul 5 2017
Bisect information: -------------------- Good build: 61.0.3118.0 Bad build: 61.0.3119.0 Unable to perform per revision bisect due to different plugin settings, After navigating to chrome://settings/content/flash "Ask first" option is not seen to enable it. So providing manual CL below. Change Log from Omahaproxy: ------------------------------ https://chromium.googlesource.com/chromium/src/+log/61.0.3118.0..61.0.3119.0?pretty=fuller&n=10000 Unable to find the actual suspect from the above log, Could any one help us in assigning the issue to the concerned owner. Thanks!
,
Jul 5 2017
Looking at the range, it may be related to Lucas's change to navigator.plugins return value behavior. Would it be possible to get a narrower bisect w/ Chromium build and a system installation of Flash Player? https://chromium.googlesource.com/chromium/src/+/819697cf41f496d914415cc88a39e39afe4dfe29
,
Jul 5 2017
,
Jul 5 2017
,
Jul 5 2017
This is because I removed the code that clears the plugin data of *all* pages in the renderer process inside Page::RefreshPlugins(). The spec says that "The PluginArray objects created by a user agent must not be live. The set of plugins represented by the objects must not change once an object is created, except when it is updated by the refresh() method." This piece of code used to change the live object inside all pages in the renderer process, not just the affected page that's currently being reloaded by the flash reload mechanism. I'll put up a fix for this, but to remain spec-compliant I'll need to force a reload of all the affected pages when the setting changes. Let me know if anyone has any concerns about this plan.
,
Jul 5 2017
lfg: Hey Lucas, sorry i did not chime in on your CL sooner. There's no way we can break the HBD behavior (too many parties rely on it). Can we schedule a VC so we can find a way to preserve the HBD behavior while still achieving your goal of better spec conformance? Thanks, Tommy
,
Jul 5 2017
To update the bug -- I've chatted with Tommy about this and we decided to revert to the previous behavior, even though it's out-of-spec; it only happens when the user explicitly changes the permissions, and it's better than forcing a reload, which could cause other issues (reloading many unneeded tabs, potential loss of form data).
,
Jul 7 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/a22efb271e62974dc0f1016fbb2d00730079eff9 commit a22efb271e62974dc0f1016fbb2d00730079eff9 Author: Lucas Gadani <lfg@chromium.org> Date: Fri Jul 07 16:25:37 2017 Invalidate navigator.plugins when permissions change. This fixes reloading of flash plugins when content settings change. Bug: 735854 Change-Id: I1f6db54c73436223ecbbcf713034a3a590c2e67d Reviewed-on: https://chromium-review.googlesource.com/559710 Reviewed-by: Rick Byers <rbyers@chromium.org> Reviewed-by: Jeremy Roman <jbroman@chromium.org> Commit-Queue: Lucas Gadani <lfg@chromium.org> Cr-Commit-Position: refs/heads/master@{#484949} [modify] https://crrev.com/a22efb271e62974dc0f1016fbb2d00730079eff9/third_party/WebKit/Source/core/page/Page.cpp [modify] https://crrev.com/a22efb271e62974dc0f1016fbb2d00730079eff9/third_party/WebKit/Source/core/page/Page.h [modify] https://crrev.com/a22efb271e62974dc0f1016fbb2d00730079eff9/third_party/WebKit/Source/web/WebKit.cpp [modify] https://crrev.com/a22efb271e62974dc0f1016fbb2d00730079eff9/third_party/WebKit/public/web/WebKit.h
,
Jul 10 2017
Verified this issue on Ubuntu 14.04, Windows-10 and Mac OS 10.12.5 using chrome latest dev #61.0.3153.0. After navigating the flash site www.dhs.state.il.us/accessibility/tests/flash/video.html no "Run flash" dialog has observed, but able to play the video directly. Considering the above fix is working as intended and marking it as TE-Verified label. Thanks!
,
Jul 10 2017
|
||||||||||||
►
Sign in to add a comment |
||||||||||||
Comment 1 by pbomm...@chromium.org
, Jun 22 2017Components: -Internals>Media Internals>Plugins>Flash
Labels: ReleaseBlock-Dev M-61 Needs-Bisect OS-Chrome OS-Linux OS-Mac
Status: Untriaged (was: Unconfirmed)