[chromium] force NRWT to use DRT

• Component: New Bugs
• Status: RESOLVED FIXED

[Tony Chang]

[chromium] force NRWT to use DRT

[Tony Chang]

Created attachment 75333 [details]
Patch

[Tony Chang]

This is intentionally small so when the canaries break, I can roll back.  If it sticks, I'll make a follow up patch to remove the flag (and to remove the flag from the bots).

[Dirk Pranke]

Heh. LGTM!

[Tony Chang]

Just an fyi for victorw, the current gardener.

[Tony Chang]

Committed r73164: <http://trac.webkit.org/changeset/73164>

[Tony Chang]

Reverted r73164 for reason:

broke debugger tests on win

Committed r73207: <http://trac.webkit.org/changeset/73207>

[Tony Chang]

Committed r73456: <http://trac.webkit.org/changeset/73456>

[Dirk Pranke]

Okay, I should have caught this before, but realizing this now ... changing the default this way doesn't give us a way to run w/ TestShell anymore. We need to add a --no-use-drt flag (or a --use-test-shell or something). It's not hard to hack the RWT code to work around this, obviously, and we don't really want to encourage people to use this code, but as long as TestShell is being supported, I'd prefer we have a way to run this without local mods.

Thoughts?

Also, any thought as to when we might actually stop supporting TestShell and remove the code from the Tree?

[Ojan Vafai]

This was the most expedient way we could think of making this change. The plan, if this sticks for the next day or two, is to fully remove all TestShell support.

[Tony Chang]

I'm going to write a patch today for --use-test-shell (jamesr's recommendation for testing for DRT specific failures).

As mentioned in comment #2, this patch was intentionally small.

[Dirk Pranke]

(In reply to comment #10)
> I'm going to write a patch today for --use-test-shell (jamesr's recommendation for testing for DRT specific failures).
> 
> As mentioned in comment #2, this patch was intentionally small.

Sure, I think keeping the patch small was a good thing to. Just a bit too small ;)