Henrik Boström cbf5122333 Avoid signaling requested_resolution back to the adapting source.
When requested_resolution uses a different aspect ratio than the source
the encoder will restrict the frame without changing its aspect ratio,
e.g. a 60x30 input frame that is restricted to 30x30 results in 30x15,
not 30x30.

While this logic works correctly in isolation, if the source also adapts
the frame size based on the sink_wants.requested_resolution that is
signaled back to the source, then the source will produce stretched
30x30 prior to the encoder which happily sends 30x30 not knowing any
wiser.

This is incompatible with the spec[1] and makes this WPT[2] fail. The
correct behavior is to NOT signal the requested_resolution back to the
source, the encoder already configures the correct resolution so this
isn't actually needed and the source shouldn't need to know this API.

In order not to break downstream projects, the new behavior is landed
behind a flag and both behaviors are tested with TEST_P.

This unblocks launching scaleResolutionDownTo API on Web. Migrating
from old to new code path and deleting the flag is a follow-up AI:
webrtc:366284861.

[1] https://w3c.github.io/webrtc-extensions/#dom-rtcrtpencodingparameters-scaleresolutiondownto
[2] https://chromium-review.googlesource.com/c/chromium/src/+/5853944

# Relying on previous green runs for confidence due to purple bots atm,
# see b/367211396
NOTRY=True
NOPRESUBMIT=True

Bug: webrtc:366067962, webrtc:366284861
Change-Id: I7fd1016e9cc6f0b0b9b8c23b0708e521f8e12642
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/362541
Reviewed-by: Henrik Andreassson <henrika@webrtc.org>
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#43024}
2024-09-16 11:00:13 +00:00
..
2024-06-19 06:19:20 +00:00
2024-06-19 06:19:20 +00:00
2024-08-07 07:00:06 +00:00
2024-08-18 06:55:23 +00:00
2024-06-19 06:19:20 +00:00
2024-09-10 08:29:26 +00:00
2023-03-27 17:06:33 +00:00
2023-09-07 10:41:49 +00:00

How to write code in the api/ directory

Mostly, just follow the regular style guide, but:

  • Note that api/ code is not exempt from the “.h and .cc files come in pairs” rule, so if you declare something in api/path/to/foo.h, it should be defined in api/path/to/foo.cc.
  • Headers in api/ should, if possible, not #include headers outside api/. Its not always possible to avoid this, but be aware that it adds to a small mountain of technical debt that were trying to shrink.
  • .cc files in api/, on the other hand, are free to #include headers outside api/.
  • Avoid structs in api, prefer classes.

The preferred way for api/ code to access non-api/ code is to call it from a .cc file, so that users of our API headers wont transitively #include non-public headers.

For headers in api/ that need to refer to non-public types, forward declarations are often a lesser evil than including non-public header files. The usual rules still apply, though.

.cc files in api/ should preferably be kept reasonably small. If a substantial implementation is needed, consider putting it with our non-public code, and just call it from the api/ .cc file.

Avoid defining api with structs as it makes harder for the api to evolve. Your struct may gain invariant, or change how it represents data. Evolving struct from the api is particular challenging as it is designed to be used in other code bases and thus needs to be updated independetly from its usage. Class with accessors and setters makes such migration safer. See Google C++ style guide for more.

If you need to evolve existent struct in api, prefer first to convert it into a class.