[core] Improve resource and api logging.

[jimblandy]

Marked as draft, because this depends on #5802.

• Improve logging in StatelessTracker::remove_abandoned to show the
  outcome of the call.

• Add similar logging to BufferTracker and TextureTracker.

• Let device_create_buffer's log only the new buffer's label, id,
  and whether it's mapped at creation. It used to show the whole
  descriptor, which is too much detail.

• Have queue_submit log the submission id, and have device_poll
  log what it was waiting for, and what actually got done.

• Have Device::drop log the destruction of the raw device when it
  actually happens, so it's properly ordered with respect to logging
  from other parts of the device, like Device::command_allocator.

[JMS55]

While you're here, maybe you feel like adding a cargo feature flag to remove logs statically? Iirc they have a fair amount of overhead just for checking if logging is enabled or not.

[jimblandy]

While you're here, maybe you feel like adding a cargo feature flag to remove logs statically? Iirc they have a fair amount of overhead just for checking if logging is enabled or not.

Sorry, I can't - this was just a drive-by.

[Wumpf]

tbh I doubt that the checks on the log flags are significant at all given all the other things that happen in those methods, but if this actually turns out to be a measurable overhead the macros/feature flags here can be expanded accordingly
https://github.com/jimblandy/wgpu/blob/c7a16b36b16be7484c2306d1155705f3f6241bf4/wgpu-core/src/lib.rs#L309
(not in this pr obviously ;-))

[nical]

While you're here, maybe you feel like adding a cargo feature flag to remove logs statically? Iirc they have a fair amount of overhead just for checking if logging is enabled or not.

api_logs are trace-level (the most verbose level) by default. Apps that depend on wgpu can filter out the logs at compile time by setting the appropriate feature flags on the log crate.