Bug 74612

Summary: Each worker separately logs its complaint about the directory I am in
Product: WebKit Reporter: Darin Adler <darin>
Component: Tools / TestsAssignee: Nobody <webkit-unassigned>
Status: RESOLVED FIXED    
Severity: Normal CC: dpranke, eric
Priority: P2 Keywords: NRWT
Version: 528+ (Nightly build)   
Hardware: Unspecified   
OS: Unspecified   

Description Darin Adler 2011-12-15 09:01:23 PST
When I run tests, I get this:

The current directory (/Users/darin/Safari) is not a WebKit checkout, using /Users/darin/Safari/OpenSource
Starting 8 workers ...The current directory (/Users/darin/Safari) is not a WebKit checkout, using /Users/darin/Safari/OpenSource
The current directory (/Users/darin/Safari) is not a WebKit checkout, using /Users/darin/Safari/OpenSource
The current directory (/Users/darin/Safari) is not a WebKit checkout, using /Users/darin/Safari/OpenSource
The current directory (/Users/darin/Safari) is not a WebKit checkout, using /Users/darin/Safari/OpenSource
The current directory (/Users/darin/Safari) is not a WebKit checkout, using /Users/darin/Safari/OpenSource
The current directory (/Users/darin/Safari) is not a WebKit checkout, using /Users/darin/Safari/OpenSource
The current directory (/Users/darin/Safari) is not a WebKit checkout, using /Users/darin/Safari/OpenSource
The current directory (/Users/darin/Safari) is not a WebKit checkout, using /Users/darin/Safari/OpenSource

I don’t need to see that log message 9 times. In fact, I’d prefer not to see it even once.
Comment 1 Eric Seidel (no email) 2011-12-15 10:06:52 PST
Sure thing.  Easy to fix.  Just an unintented consequence of re-using single-process code in our new multi-process architecture. :)
Comment 2 Eric Seidel (no email) 2011-12-15 10:15:51 PST
You'll likely want this when running webkit-patch apply-from-bug outside of a webkit directory, but not when running new-run-webkit-tests.  Or at least not 7 times. :)

Then again, it's possible it's not even useful when running webkit-patch from outside of a webkit directory.
Comment 3 Dirk Pranke 2011-12-15 12:38:18 PST
I agree w/ Darin; new-run-webkit-tests shouldn't be complaining about this at all by default, and it should definitely only be complaining once. Can we change scm/detection.py to provide a am_inside_the_checkout() function so that webkit-patch can call this separately and log the warning?
Comment 4 Eric Seidel (no email) 2011-12-15 12:43:19 PST
(In reply to comment #3)
> I agree w/ Darin; new-run-webkit-tests shouldn't be complaining about this at all by default, and it should definitely only be complaining once. Can we change scm/detection.py to provide a am_inside_the_checkout() function so that webkit-patch can call this separately and log the warning?

Sounds like a good approach.

I agree.  webkit-patch should be specifically opting into this warning, instead of all SCM clients causing it.
Comment 5 Dirk Pranke 2012-06-19 21:10:01 PDT
I think this was fixed in r120646. Please let me know if you're still seeing an issue.