Implement f{32,64}.neg instructions #477
Conversation
gumb0
commented
Aug 12, 2020
•
gumb0
commented
Codecov Report
@@ Coverage Diff @@
## master #477 +/- ##
=======================================
Coverage 99.54% 99.54%
=======================================
Files 54 54
Lines 16196 16226 +30
=======================================
+ Hits 16122 16152 +30
Misses 74 74 |
gumb0
force-pushed
the
fabs
branch
2 times, most recently
from
3f0373e
to
64ca4c3
Compare
| { | ||
| // fneg(+-p) = -+p | ||
| EXPECT_THAT(exec(p), Result(-p)); | ||
| EXPECT_THAT(exec(-p), Result(p)); |
There was a problem hiding this comment.
Aren't in these tests practically trusting the compiler compared to the others where we have explicit constants?
There was a problem hiding this comment.
We trust that negation work everywhere.
There was a problem hiding this comment.
If we wanted to be pedantic we'd always test against the encoded representation.
There was a problem hiding this comment.
Possible in the future... i.e. we would need to extend FP to do some software floating-point operations.
There was a problem hiding this comment.
Something like wast2wasm4cpp script could work, which evaluates the C++ code and puts in the hard constants. We could manually verify that when committed and at every change.
