-
Type:
Bug
-
Resolution: Unresolved
-
Priority:
Low
-
Code Generation Tools
-
CODEGEN-10343
-
-
default
The attached source file has these statements ...
float A = 1.002002954483032F; // as U32 -> 0x3F8041A2
float B = 0.999000012874603F; // as U32 -> 0x3F7FBE77
float C = A/B;
The result computed into C is hex: 0x3f80627f float: 1.003005862236023 . When the same code is run on a typical x86 machine/compiler, the result is hex: 0x3f806280 float: 1.003005981445313.
The same result is seen whether the DIVF32 instruction, or the RTS function FS$$DIV is used.
Here is a typical build command ...
"C:/ti/compilers/ti-cgt-c2000_22.6.0.LTS/bin/cl2000" -v28 -ml -mt --cla_support=cla2 --float_support=fpu32 --tmu_support=tmu0 --vcu_support=vcu0 --fp_mode=strict --include_path="C:/ti/compilers/ti-cgt-c2000_22.6.0.LTS/include" --include_path="C:/ti/compilers/ti-cgt-c2000_22.6.0.LTS/lib/src" -g --diag_warning=225 --diag_wrap=off --display_error_number --abi=coffabi --src_interlist --preproc_with_compile --preproc_dependency="main.d_raw" "../main.c"
