x87: fix the use of CheckFloatEq and CheckDoubleEq in test.

x87: fix the use of CheckFloatEq and CheckDoubleEq in test.

Instead of using CheckFloatEq and CheckDoubleEq directly, I introduced
a macro which first stores the expected result in a volatile variable.

Here are some comments of previous CLs:

The reason is same as the CL #31808 (issue 1430943002, X87: Change the test case for X87 float operations), please refer: https://codereview.chromium.org/1430943002/.

  Here is the key comments from CL #31808
  Some new test cases use CheckFloatEq(...) and CheckDoubleEq(...) function for result check. When GCC compiling the CheckFloatEq() and CheckDoubleEq() function,
  those inlined functions has different behavior comparing with GCC ia32 build and x87 build.
  The major difference is sse float register still has single precision rounding semantic. While X87 register has no such rounding precsion semantic when directly use register value.
  The V8 turbofan JITTed has exactly same result in both X87 and IA32 port.

  So we add the following sentence to do type cast to keep the same precision for RunCallInt64ToFloat32/RunCallInt64ToFloat64. Such as: volatile double expect = static_cast<float>(*i).

R=titzer@chromium.org, weiliang.lin@intel.com

Committed: https://crrev.com/a5d41888493694e7fa49f5b9c1661936a54a6101
Cr-Commit-Position: refs/heads/master@{#34534}

[titzer, 2016-03-07 10:38:12]

https://codereview.chromium.org/1773513002/diff/1/test/cctest/compiler/value-...
File test/cctest/compiler/value-helper.h (right):

https://codereview.chromium.org/1773513002/diff/1/test/cctest/compiler/value-...
test/cctest/compiler/value-helper.h:324: #define CHECK_FLOAT_EQ(lhs, rhs) \
You should use a do { .. } while(false) to be a good macro citizen.

[ahaas, 2016-03-07 11:59:06]

On 2016/03/07 at 10:38:12, titzer wrote:
>
https://codereview.chromium.org/1773513002/diff/1/test/cctest/compiler/value-...
> File test/cctest/compiler/value-helper.h (right):
> 
>
https://codereview.chromium.org/1773513002/diff/1/test/cctest/compiler/value-...
> test/cctest/compiler/value-helper.h:324: #define CHECK_FLOAT_EQ(lhs, rhs) \
> You should use a do { .. } while(false) to be a good macro citizen.

Done.

[titzer, 2016-03-07 12:01:18]

lgtm

[ahaas, 2016-03-07 12:13:45]

The CQ bit was checked by ahaas@chromium.org

[commit-bot: I haz the power, 2016-03-07 12:13:50]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1773513002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1773513002/20001

[commit-bot: I haz the power, 2016-03-07 12:29:13]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-03-07 12:30:28]

Description was changed from

==========
x87: fix the use of CheckFloatEq and CheckDoubleEq in test.

Instead of using CheckFloatEq and CheckDoubleEq directly, I introduced
a macro which first stores the expected result in a volatile variable.

Here are some comments of previous CLs:

The reason is same as the CL #31808 (issue 1430943002, X87: Change the test case
for X87 float operations), please refer:
https://codereview.chromium.org/1430943002/.

  Here is the key comments from CL #31808
  Some new test cases use CheckFloatEq(...) and CheckDoubleEq(...) function for
result check. When GCC compiling the CheckFloatEq() and CheckDoubleEq()
function,
  those inlined functions has different behavior comparing with GCC ia32 build
and x87 build.
  The major difference is sse float register still has single precision rounding
semantic. While X87 register has no such rounding precsion semantic when
directly use register value.
  The V8 turbofan JITTed has exactly same result in both X87 and IA32 port.

  So we add the following sentence to do type cast to keep the same precision
for RunCallInt64ToFloat32/RunCallInt64ToFloat64. Such as: volatile double expect
= static_cast<float>(*i).

R=titzer@chromium.org, weiliang.lin@intel.com
==========

to

==========
x87: fix the use of CheckFloatEq and CheckDoubleEq in test.

Instead of using CheckFloatEq and CheckDoubleEq directly, I introduced
a macro which first stores the expected result in a volatile variable.

Here are some comments of previous CLs:

The reason is same as the CL #31808 (issue 1430943002, X87: Change the test case
for X87 float operations), please refer:
https://codereview.chromium.org/1430943002/.

  Here is the key comments from CL #31808
  Some new test cases use CheckFloatEq(...) and CheckDoubleEq(...) function for
result check. When GCC compiling the CheckFloatEq() and CheckDoubleEq()
function,
  those inlined functions has different behavior comparing with GCC ia32 build
and x87 build.
  The major difference is sse float register still has single precision rounding
semantic. While X87 register has no such rounding precsion semantic when
directly use register value.
  The V8 turbofan JITTed has exactly same result in both X87 and IA32 port.

  So we add the following sentence to do type cast to keep the same precision
for RunCallInt64ToFloat32/RunCallInt64ToFloat64. Such as: volatile double expect
= static_cast<float>(*i).

R=titzer@chromium.org, weiliang.lin@intel.com

Committed: https://crrev.com/a5d41888493694e7fa49f5b9c1661936a54a6101
Cr-Commit-Position: refs/heads/master@{#34534}
==========

[commit-bot: I haz the power, 2016-03-07 12:30:29]

Patchset 2 (id:??) landed as
https://crrev.com/a5d41888493694e7fa49f5b9c1661936a54a6101
Cr-Commit-Position: refs/heads/master@{#34534}