Allow tracing categories to default to disabled.

Allow tracing categories to default to disabled.

This CL allows the prefixing of category names with disabled- to have them
not traced by default. Only if an explicit disabled-foo is provided in the
category filter do we enable the category.

Providing a disabled-foo in the category filter does not act like an include
category in that all other includes have to be specified. The disables are
treated as their own group.

BUG=232109
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=198787

[dsinclair, 2013-05-01 04:07:29]

nduca, PTAL.

I think I have all the various cases covered in the tests, but please let me
know if you can think of any I'm missing that should be covered.

[dsinclair, 2013-05-01 14:50:20]

I took a closer look at the way categories were being merged and realized it
wouldn't work for disabled categories.  I've changed the merge strategy slightly
with this CL.

Previously:
(foo,-bar) Merged (-baz) would become (*)

Now:
(foo,-bar) Merged (-baz) would become (-bar,-baz)

Essentially, in the old system if either of the filters that was being merged
did not have included categories we'd clear both the included and excluded. I
don't think that's right as I believe we'd still want the excludes in the
filter. So now, I clear the includes while the disables and excludes get merged
as expected.

[nduca, 2013-05-02 06:11:23]

This looks good. LGTM but some quick followups....

Do you think we could tolerate the more verbose:
disabled-by-default-cc

And make a macro for this?
TRACE_EVENT(TRACE_DISABLED_BY_DEFAULT("cc'), "myevent");

[dsinclair, 2013-05-02 14:45:47]

On 2013/05/02 06:11:23, nduca wrote:
> This looks good. LGTM but some quick followups....
> 
> Do you think we could tolerate the more verbose:
> disabled-by-default-cc
> 
> And make a macro for this?
> TRACE_EVENT(TRACE_DISABLED_BY_DEFAULT("cc'), "myevent");

Thought of the macro last night, it's now implemented. I've also switched to
disabled-by-default- as I think it's a lot clearer as to the purpose.

[dsinclair, 2013-05-02 14:46:37]

jar, PTAL for base/ OWNERS.

Thanks,
dan

[nduca, 2013-05-02 20:36:48]

very lgtm :)

[jar (doing other things), 2013-05-03 19:48:00]

rubber stamp base LGTM

[commit-bot: I haz the power, 2013-05-07 14:12:42]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/dsinclair@chromium.org/14748003/29001

[commit-bot: I haz the power, 2013-05-07 20:17:33]

Change committed as 198787