1159495 - [EME] Make navigator.requestMediaKeySystemAccess() only report Adobe EME is available if it's likely to work

Status: RESOLVED FIXED

(Core :: Audio/Video, defect, P1)

Description

[Chris Pearce [:cpearce (Not reading bugmail)]]

Currently our navigator.requestMediaKeySystemAccess() implementation reports that Adobe EME works when it can't. This can happen if we're running on Vista or Windows Server and the Windows Updates that install the H.264/AAC decoders is not available. We can also fail if we don't have a plugin-container voucher, which happens in a local build, or a downstream build.

It doesn't make much sense to advertise that Adobe EME will work when it actually won't, so we should check for these conditions and fail in navigator.requestMediaKeySystemAccess() when they occur.

Comment 1

[Chris Pearce [:cpearce (Not reading bugmail)]]

Attachment: Patch 1: Only report Adobe EME supported if required WMF codecs are installed

Only report Adobe EME supported if required WMF codecs are installed.

Comment 2

[Chris Pearce [:cpearce (Not reading bugmail)]]

Attachment: Patch 2: Only report that Adobe EME is available if we have a plugin-container voucher

Only report that Adobe EME is available if we have a plugin-container voucher.

This means local dev-builds and downstream builds won't report that Adobe EME is available, as it won't work there anyway.

Comment 3

[Chris Pearce [:cpearce (Not reading bugmail)]]

Comment on attachment 8598962 [details] [diff] [review]
Patch 1: Only report Adobe EME supported if required WMF codecs are installed

Approval Request Comment
[Feature/regressing bug #]: EME
[User impact if declined]: Without this patch, Firefox on some Vista and Windows Server/Enterprise machines that lack the Windows Media Foundation service packs installed will report that EME will work when really it won't, and sites will try to use EME on these machines will fail. This affects our EME launch partner.
[Describe test coverage new/current, TreeHerder]: Local testing only. This only affects Adobe EME, which we can't test in our automation yet.
[Risks and why]: Low, this uses a well tested code path to detect whether EME is going to work, and make us report that it won't work when it won't work.
[String/UUID change made/needed]: None.

Comment 4

[Chris Pearce [:cpearce (Not reading bugmail)]]

Comment on attachment 8598964 [details] [diff] [review]
Patch 2: Only report that Adobe EME is available if we have a plugin-container voucher

Approval Request Comment
[Feature/regressing bug #]: EME
[User impact if declined]: Without this patch, Firefox on builds built by local developers and by downstream projects that lack a signed EME plugin container/sandbox will think they can support Adobe EME, where really such builds cannot. Without the signed "voucher" for the plugin-container, Adobe's plugin will refuse to run. They will report the sites using EME that they support Adobe EME, and try to use it, and fail. This failure is showing up in our EME launch partner's telemetry, pushing up our error rate.
[Describe test coverage new/current, TreeHerder]: Local testing only. This only affects Adobe EME, which we can't test in our automation yet.
[Risks and why]: Low, this basically just checks if a file exists in our "do we support Adobe EME" logic.
[String/UUID change made/needed]: None.

Comment 5

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/e3864c540bed
https://hg.mozilla.org/integration/mozilla-inbound/rev/cb9c0ddc2e22

Comment 6

[Chris Pearce [:cpearce (Not reading bugmail)]]

Attachment: [Aurora & Release rebase] Patch 1: Only report Adobe EME supported if required WMF codecs are installed

Rebased, applies cleanly on aurora and release.

Comment 7

[Chris Pearce [:cpearce (Not reading bugmail)]]

Attachment: [Aurora & Release rebase] Patch 2: Only report that Adobe EME is available if we have a plugin-container voucher

Rebased onto aurora and release.

Comment 8

[Carsten Book [:Tomcat]]

sorry had to back this out for bustage in nexus builds like https://treeherder.mozilla.org/logviewer.html#?job_id=9399464&repo=mozilla-inbound

Comment 9

[Pulsebot]

Backout:
https://hg.mozilla.org/integration/mozilla-inbound/rev/d889fd14f2d9
https://hg.mozilla.org/integration/mozilla-inbound/rev/9a5ae8694c6b

Comment 10

[Chris Pearce [:cpearce (Not reading bugmail)]]

Looks like bustage because we don't build with --enable-eme or MOZ_FMP4 on L Emulator.

 06:19:48 INFO - /builds/slave/b2g_m-in_n5-l_eng_dep-00000000/build/gecko/dom/media/gmp/GMPProcessParent.cpp:8:30: fatal error: mozilla/EMEUtils.h: No such file or directory
55730 06:19:48 ERROR - make[6]: *** [Unified_cpp_dom_media_gmp0.o] Error 1

Comment 11

[Chris Pearce [:cpearce (Not reading bugmail)]]

Attachment: Patch 2v2: Only report that Adobe EME is available if we have a plugin-container voucher

Patch 2 with --disable-eme build failure fixed.

Approval Request Comment
[Feature/regressing bug #]: EME
[User impact if declined]: Without this patch, Firefox on builds built by local developers and by downstream projects that lack a signed EME plugin container/sandbox will think they can support Adobe EME, where really such builds cannot. Without the signed "voucher" for the plugin-container, Adobe's plugin will refuse to run. They will report the sites using EME that they support Adobe EME, and try to use it, and fail. This failure is showing up in our EME launch partner's telemetry, pushing up our error rate.
[Describe test coverage new/current, TreeHerder]: Local testing only. This only affects Adobe EME, which we can't test in our automation yet.
[Risks and why]: Low, this basically just checks if a file exists in our "do we support Adobe EME" logic.
[String/UUID change made/needed]: None.

Comment 12

[Chris Pearce [:cpearce (Not reading bugmail)]]

Attachment: [Aurora rebase] Patch 2v2: Only report that Adobe EME is available if we have a plugin-container voucher

Working version of Patch 2 rebased on the aurora/beta.

Comment 13

[Chris Pearce [:cpearce (Not reading bugmail)]]

Comment on attachment 8599614 [details] [diff] [review]
[Aurora rebase] Patch 2v2: Only report that Adobe EME is available if we have a plugin-container voucher

This patch only builds on Aurora, not release...

Comment 14

[Chris Pearce [:cpearce (Not reading bugmail)]]

Attachment: [Release rebase] Patch 2v2: Only report that Adobe EME is available if we have a plugin-container voucher

Patch 2 rebased and building on Release.

Comment 15

[Chris Pearce [:cpearce (Not reading bugmail)]]

Release try push:
https://treeherder.mozilla.org/#/jobs?repo=try&revision=9efb5d870437

Aurora try push:
https://treeherder.mozilla.org/#/jobs?repo=try&revision=e524c9068462

Comment 16

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/2a4864545b47
https://hg.mozilla.org/integration/mozilla-inbound/rev/469ae46092fc

Comment 17

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/2a4864545b47
https://hg.mozilla.org/mozilla-central/rev/469ae46092fc

Comment 18

[Sylvestre Ledru [:Sylvestre]]

Comment on attachment 8599599 [details] [diff] [review]
Patch 2v2:  Only report that Adobe EME is available if we have a plugin-container voucher

Should be in 38 rc 1

Comment 19

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-aurora/rev/02609ffc2453
https://hg.mozilla.org/releases/mozilla-aurora/rev/3f5b3cfd2071

Comment 20

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-release/rev/60555feb4888
https://hg.mozilla.org/releases/mozilla-release/rev/6e95db92c8d4

Comment 21

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-esr38/rev/60555feb4888
https://hg.mozilla.org/releases/mozilla-esr38/rev/6e95db92c8d4

Comment 22

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-beta/rev/60555feb4888
https://hg.mozilla.org/releases/mozilla-beta/rev/6e95db92c8d4

Comment 23

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/34d3296e6980

Comment 24

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/34d3296e6980