1027953 - Fix assumptions about int64_t in MediaResourceServer

Status: RESOLVED FIXED

(Core :: WebRTC, defect)

Description

[Jed Davis [:jld] ⟨⏰|UTC-7⟩ ⟦he/him⟧]

int64_t isn't necessarily long long int; on x86_64 Linux it's long int, and even though long long int is also 64 bits, the two types are treated as distinct by the C/C++ compiler.

This is currently breaking the --enable-media-plugins build, because a call to std::max in content/media/plugins/MediaResourceServer.cpp fails to unify.

Also, there are some misuses of "%lld" to printf an int64_t in the same file, but those are only warnings (despite --enable-warnings-as-errors).

Comment 1

[Jed Davis [:jld] ⟨⏰|UTC-7⟩ ⟦he/him⟧]

Attachment: bug1027953-gmp-64bit.diff

Trying: https://tbpl.mozilla.org/?tree=Try&rev=cae982acb686

Note that some of the whitespace is significant: "%"PRId64 is *not* the same as "%" PRId64 in C++11.

Also, I'm not sure if this is the right component or the right reviewer; [[Modules/All]] is not maximally helpful for this file.

Comment 2

[Randell Jesup [:jesup] (needinfo me)]

Comment on attachment 8443186 [details] [diff] [review]
bug1027953-gmp-64bit.diff

Review of attachment 8443186 [details] [diff] [review]:
-----------------------------------------------------------------

-> cpearce; I have no contact with MediaResourceServer

Comment 3

[Jed Davis [:jld] ⟨⏰|UTC-7⟩ ⟦he/him⟧]

c-n note: This is currently NPOTB on desktop, and that probably explains why the Try run in comment #1 has Windows builds busted in code I didn't touch, but here's another run if more assurance is desired: https://tbpl.mozilla.org/?tree=Try&rev=8439a57db1c6

Comment 4

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/5bfbe2cc3247

Comment 5

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/5bfbe2cc3247