fpu: round Zfa values directly to integral floating-point values - #302
Open
carlosqwqqwq wants to merge 1 commit into
Open
fpu: round Zfa values directly to integral floating-point values#302carlosqwqqwq wants to merge 1 commit into
carlosqwqqwq wants to merge 1 commit into
Conversation
carlosqwqqwq
force-pushed
the
fix/zfa-fround-integral
branch
from
September 5, 2026 12:20
b550f80 to
ad9d6c2
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
fpu: round Zfa values directly to integral floating-point values
Fixes #296
Commit message
Description
The current
fround/froundnximplementation rounds through a signed 64-bit integer and converts back. That saturates huge, infinity, and NaN inputs, raises flags from the conversion path, and makesfroundsetNX. Use the existing bit-exact integral-rounding helpers directly, resolveRM_DYNto the guestfrmat the caller, raiseNXonly for thefroundnxvariant, and retainNVfor signaling NaNs.Validation
qemu-riscv64, Zfa) preserves huge/inf/NaN inputs unmodified, setsNXonly forfroundnx, and raises onlyNVfor signaling NaNs; the patched build matches on the single- and double-precision witness matrix.rm=dynwithfrm=RNE,RDN,RUP, andRMM; the helpers receive the resolved mode rather thanRM_DYN.