test: Add proper f{32,64}_add tests #467
Conversation
chfast
force-pushed
the
float-comparison
branch
from
d15f8b8 to
8b963e4
Compare
Codecov Report
@@ Coverage Diff @@
## master #467 +/- ##
========================================
Coverage 99.53% 99.53%
========================================
Files 54 54
Lines 15821 15926 +105
========================================
+ Hits 15747 15852 +105
Misses 74 74 |
chfast
force-pushed
the
float-comparison
branch
from
3524d8e to
abc3358
Compare
chfast
force-pushed
the
fp_add
branch
6 times, most recently
from
017c219 to
9e55d85
Compare
chfast
force-pushed
the
fp_add
branch
2 times, most recently
from
eb547f4 to
1a740b6
Compare
| @@ -121,9 +92,21 @@ template <typename T> | |||
| class execute_floating_point_types : public testing::Test | |||
| { | |||
| protected: | |||
| using L = typename FP<T>::Limits; | |||
|
|
|||
| // The q is any positive floating-point value without zeros, infinities and NaNs. | |||
There was a problem hiding this comment.
The meta variable q is used to range over (signless) rational magnitudes, excluding nan or ∞.
https://webassembly.github.io/spec/core/exec/numerics.html#numerics
There was a problem hiding this comment.
That includes zeroes, but it's not so important
There was a problem hiding this comment.
I would call these two maybe positive_special_values and ordered_special_values
| EXPECT_THAT(exec(-TypeParam{0.0}, Limits::infinity()), Result(Limits::infinity())); | ||
| EXPECT_THAT(exec(-TypeParam{0.0}, -Limits::infinity()), Result(-Limits::infinity())); | ||
|
|
||
| for (const auto q : this->q_values) |
There was a problem hiding this comment.
| for (const auto q : this->q_values) | |
| for (const auto q : q_values) |
There was a problem hiding this comment.
Does not work because of whatever is under GTest macros. You need this-> or execute_floating_point_types::.
| class execute_floating_point_types : public testing::Test | ||
| { | ||
| protected: | ||
| using L = typename FP<T>::Limits; |
There was a problem hiding this comment.
I would prefer this to be Limits
There was a problem hiding this comment.
It cannot be because it will conflict of other using Limits in the tests.
There was a problem hiding this comment.
Why can't it use this one in the tests?
There was a problem hiding this comment.
Because each TYPED_TEST inherits a template execute_floating_point_types<TypeParam> you have to access it as typename execute_floating_point_types<TypeParam>::Limits. Not very short.
chfast
force-pushed
the
fp_add
branch
3 times, most recently
from
50f007b to
403e427
Compare
It adds/fixes some testing helpers for floating point instructions.
Then it adds wasm spec-driver tests for
fadd, i.e. it adds checks for these points from the definition: