Fix 10.6 problem where killall doesn't work as expected so lighttpd...

Fix 10.6 problem where killall doesn't work as expected so lighttpd
doesn't get killed.

BUG=http://crbug.com/23919

TEST=The following command lines will now work and give no errors on 10.6:
python webkit/tools/layout_tests/layout_package/http_server.py --server start
python webkit/tools/layout_tests/layout_package/http_server.py --server stop



Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=28259

[TVL, 2009-10-06 11:51:33]

lg or use the suggested change.

http://codereview.chromium.org/257059/diff/1/2
File webkit/tools/layout_tests/layout_package/platform_utils_mac.py (right):

http://codereview.chromium.org/257059/diff/1/2#newcode109
Line 109: subprocess.call(['killall', '-u', os.getenv('USER'), '-TERM',
'lighttpd'])
Doing a quick test, killall -TERM -u USER lighttpd works on both os versions. 
use that instead?

[John Grabowski, 2009-10-07 16:36:47]

On 2009/10/06 11:51:33, TVL wrote:
> lg or use the suggested change.
> 
> http://codereview.chromium.org/257059/diff/1/2
> File webkit/tools/layout_tests/layout_package/platform_utils_mac.py (right):
> 
> http://codereview.chromium.org/257059/diff/1/2#newcode109
> Line 109: subprocess.call(['killall', '-u', os.getenv('USER'), '-TERM',
> 'lighttpd'])
> Doing a quick test, killall -TERM -u USER lighttpd works on both os versions. 
> use that instead?

Using your idea.  New diff up.

[Mark Mentovai, 2009-10-07 16:39:14]

LGTM