Closed michaeljclark closed 6 years ago
These two patches were dropped because softfloat was under change in upstream: https://github.com/riscv/riscv-qemu/commit/74d08bbe01036c601c838f6c5c0196739e26394c https://github.com/riscv/riscv-qemu/commit/19861c3396695fc69076e4bd0542df5a9e15b69b
The change to the minmax macro needs to be forward ported to the latest softfloat code which now may be a static inline function instead of a macro. This will fix fmin/fmax failures in riscv-tests.
Closing this as riscv-tests fmin/fmax pass in upstream qemu due to qNaN/sNaN changes in the existing minnum/maxnum
These two patches were dropped because softfloat was under change in upstream: https://github.com/riscv/riscv-qemu/commit/74d08bbe01036c601c838f6c5c0196739e26394c https://github.com/riscv/riscv-qemu/commit/19861c3396695fc69076e4bd0542df5a9e15b69b
The change to the minmax macro needs to be forward ported to the latest softfloat code which now may be a static inline function instead of a macro. This will fix fmin/fmax failures in riscv-tests.