-
Notifications
You must be signed in to change notification settings - Fork 1.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
cranelift: Add MinGW fma
regression tests
#4517
cranelift: Add MinGW fma
regression tests
#4517
Conversation
Since this is just for the interpreter one possibility would be to use |
libm's I've filed rust-lang/libm#263 to follow up with libm. |
@alexcrichton since there is no way to know how long this is going to take with libm. What do you think about disabling the interpreter on the An alternative could be to have a reduced set of inputs that does not fail in a separate file. That way we can resolve this now and move ahead with the CI upgrades. |
The interpreter can run `fma.clif` on most platforms, however on `x86_64-pc-windows-gnu` we use libm which has issues with some inputs. We should delete `fma-interpreter.clif` and enable the interpreter on the main `fma.clif` file once those are fixed.
1ff2113
to
f157967
Compare
Sounds reasonable to me, thanks for helping to push on this @afonso360! |
This resolves an issue with incorrect fmaf on the x86_64-pc-windows-gnu target under some inputs. See: bytecodealliance#4517
* cranelift: Upgrade libm to 0.2.4 This resolves an issue with incorrect fmaf on the x86_64-pc-windows-gnu target under some inputs. See: #4517 * supply-chain: Vet `libm` 0.2.4
See #4512 for context.
It looks like MinGW has an incorrect implementation of
fma
. This did not show up in the currentfma
test suite, but it did show when @alexcrichton tried to upgrade towindows-latest
on CI.In this PR we just add the test cases that were submitted to the original bug report of MinGW.
Opening this as a draft, and eventually we'll add a fix for this as well.
cc: @cfallin @alexcrichton