Enable 3 webkit layout tests

Enabled tests:
LayoutTests/fast/block/float/crash-on-absolute-positioning.html
  I upstreamed new results to WebKit in r41874:
  http://trac.webkit.org/changeset/41874

LayoutTests/http/tests/xmlhttprequest/redirect-cross-origin-post.html
  Crash is fixed in CL: http://codereview.chromium.org/50077.
  Further failure is fixed in CL: http://codereview.chromium.org/53070

LayoutTests/http/tests/xmlhttprequest/simple-cross-origin-progress-events.html
  First timeout was fixed in http://codereview.chromium.org/53070
  This test is still time consuming in nature, it barely finishes in 10s.
  If we can set timeout to 20s for this test we can fully enable it.


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=12527

[Pam (message me for reviews), 2009-03-25 22:59:45]

One of these days maybe we'll implement per-test timeouts. (That would only help
a handful of tests, so it's not very high priority.) In the meantime, can we
mark the TIMEOUT for that one test as DEBUG only, or does it sometimes time out
in Release too?

[Alpha Left Google, 2009-03-25 23:19:54]

On 2009/03/25 22:59:45, Pam wrote:
> One of these days maybe we'll implement per-test timeouts. (That would only
help
> a handful of tests, so it's not very high priority.) In the meantime, can we
> mark the TIMEOUT for that one test as DEBUG only, or does it sometimes time
out
> in Release too?

It constantly passes in my machine, I'll enable it to see if it would fail in
buildbot.

[Pam (message me for reviews), 2009-03-26 00:07:31]

On 2009/03/25 23:19:54, Alpha wrote:
> On 2009/03/25 22:59:45, Pam wrote:
> > One of these days maybe we'll implement per-test timeouts. (That would only
> help
> > a handful of tests, so it's not very high priority.) In the meantime, can we
> > mark the TIMEOUT for that one test as DEBUG only, or does it sometimes time
> out
> > in Release too?
> 
> It constantly passes in my machine, I'll enable it to see if it would fail in
> buildbot.


OK, LGTM.

- Pam