[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [tor-bugs] #22548 [Applications/Tor Browser]: Firefox downgrades VP9 videos to VP8 when measured performance is not enough
#22548: Firefox downgrades VP9 videos to VP8 when measured performance is not
enough
-------------------------------------------------+-------------------------
Reporter: cypherpunks | Owner: tbb-
| team
Type: defect | Status: closed
Priority: Medium | Milestone:
Component: Applications/Tor Browser | Version:
Severity: Normal | Resolution: fixed
Keywords: tbb-fingerprinting, | Actual Points:
TorBrowserTeam201801R |
Parent ID: | Points:
Reviewer: | Sponsor:
-------------------------------------------------+-------------------------
Comment (by gk):
So, the patch in `Benchmark.cpp` does not even compile actually. Here is
the output with GCC 5.4.0:
{{{
/var/tmp/build/firefox-f0b673dbac3a/dom/media/Benchmark.cpp: In static
member function 'static bool mozilla::VP9Benchmark::IsVP9DecodeFast()':
/var/tmp/build/firefox-f0b673dbac3a/dom/media/Benchmark.cpp:83:60: error:
call of overloaded 'GetUint(const char [30], int)' is ambiguous
Preferences::GetUint("media.benchmark.vp9.threshold", 0);
^
In file included from /var/tmp/build/firefox-f0b673dbac3a/obj-x86_64-pc-
linux-gnu/dist/include/xpcpublic.h:28:0,
from /var/tmp/build/firefox-
f0b673dbac3a/dom/base/nsJSEnvironment.h:19,
from /var/tmp/build/firefox-f0b673dbac3a/obj-x86_64-pc-
linux-gnu/dist/include/mozilla/dom/CallbackObject.h:31,
from /var/tmp/build/firefox-f0b673dbac3a/obj-x86_64-pc-
linux-gnu/dist/include/mozilla/dom/BindingUtils.h:20,
from /var/tmp/build/firefox-f0b673dbac3a/obj-x86_64-pc-
linux-gnu/dist/include/mozilla/dom/ToJSValue.h:12,
from /var/tmp/build/firefox-f0b673dbac3a/obj-x86_64-pc-
linux-gnu/dist/include/mozilla/dom/IterableIterator.h:34,
from /var/tmp/build/firefox-f0b673dbac3a/obj-x86_64-pc-
linux-gnu/dist/include/mozilla/dom/MediaKeyStatusMapBinding.h:11,
from /var/tmp/build/firefox-f0b673dbac3a/obj-x86_64-pc-
linux-gnu/dist/include/mozilla/CDMCaps.h:18,
from /var/tmp/build/firefox-f0b673dbac3a/obj-x86_64-pc-
linux-gnu/dist/include/mozilla/CDMProxy.h:10,
from /var/tmp/build/firefox-
f0b673dbac3a/dom/media/MediaDecoder.h:11,
from /var/tmp/build/firefox-
f0b673dbac3a/dom/media/ADTSDecoder.h:10,
from /var/tmp/build/firefox-
f0b673dbac3a/dom/media/ADTSDecoder.cpp:7,
from /var/tmp/build/firefox-f0b673dbac3a/obj-x86_64-pc-
linux-gnu/dom/media/Unified_cpp_dom_media0.cpp:2:
/var/tmp/build/firefox-f0b673dbac3a/obj-x86_64-pc-linux-
gnu/dist/include/mozilla/Preferences.h:122:19: note: candidate: static
uint32_t mozilla::Preferences::GetUint(const char*, uint32_t)
static uint32_t GetUint(const char* aPref, uint32_t aDefault = 0)
^
/var/tmp/build/firefox-f0b673dbac3a/obj-x86_64-pc-linux-
gnu/dist/include/mozilla/Preferences.h:173:19: note: candidate: static
nsresult mozilla::Preferences::GetUint(const char*, uint32_t*)
static nsresult GetUint(const char* aPref, uint32_t* aResult)
^
}}}
Now, if you look closer at the function in questions you'll see:
{{{
uint32_t hadRecentUpdate = Preferences::GetUint(sBenchmarkFpsVersionCheck,
0U)
}}}
And, yes, replacing `0` with `0U` in the patch I merged fixes the
compilation. I've pushed a fixup commit to `tor-browser-52.6.0esr-8.0-1`
(commit d93d0469d2b21b67869369063e17074563a49abf).
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/22548#comment:22>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online
_______________________________________________
tor-bugs mailing list
tor-bugs@xxxxxxxxxxxxxxxxxxxx
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-bugs