[CC]: Reduce required precision for TransformNormal test.

[CC]: Reduce required precision for TransformNormal test.

This CL changes the precision requirements for
DrawPolygonTransformTest.TransformNormal.  This is required since it currently
fails on Arm64 due to gcc's use of a fused multiple instruction on mapMScalars.

Details of this issue are available on the associated bug.

BUG=401117
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=287871

[rmcilroy, 2014-08-06 16:38:04]

Sami, could you please take a look, thanks.

[Sami, 2014-08-06 17:09:24]

https://codereview.chromium.org/443903004/diff/1/cc/quads/draw_polygon_unitte...
File cc/quads/draw_polygon_unittest.cc (right):

https://codereview.chromium.org/443903004/diff/1/cc/quads/draw_polygon_unitte...
cc/quads/draw_polygon_unittest.cc:20: EXPECT_TRUE(std::abs(a - b) <
std::numeric_limits<float>::epsilon());
I think in general the epsilon should be scaled up like shown here:
http://en.cppreference.com/w/cpp/types/numeric_limits/epsilon

However that wouldn't work with zero like we talked before, so maybe we should
just throw a std::max(1.f, ...) in there.

Alternatively we could rename this macro to only work with comparisons with <=
1.

[Sami, 2014-08-06 17:14:23]

Ack, thanks for pointing out that the macro name already does what I suggested
:) lgtm.

[rmcilroy, 2014-08-06 17:37:25]

The CQ bit was checked by rmcilroy@chromium.org

[commit-bot: I haz the power, 2014-08-06 17:37:48]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/rmcilroy@chromium.org/443903004/1

[commit-bot: I haz the power, 2014-08-06 21:07:53]

FYI, CQ is re-trying this CL (attempt #1).
The failing builders are:
  mac_chromium_compile_dbg on tryserver.chromium.mac
(http://build.chromium.org/p/tryserver.chromium.mac/builders/mac_chromium_comp...)

[commit-bot: I haz the power, 2014-08-06 22:09:19]

Change committed as 287871