Ti compiler (15.12.0.LTS) ) options --float_operations_allowed and --fp_mode gives compile error in math.h

My guess is that you code calls sqrt when it should call sqrtf instead.  Is that it?

Thanks and regards,

-George

Hello George,

sorry but it does not seem to be that simple.

I definitely used the sqrtf function / also tried the compiler Intrinsics __sqrtf call.

Even when completely removing the call to sqrtf I get the error messages just by including math.h.

A simple main..c like this also fails:

#include <stdio.h>
#include <math.h>
/*
 * main.c
 */
int main(void)
{
	printf("Hello World!\n");
	
	return 0;
}

Thanks and regards,

Josef

I continue to be unable to reproduce the same error messages. Please show all the build options exactly as the compiler sees them.

Thanks and regards,

-George

George Mock said:

I continue to be unable to reproduce the same error messages. Please show all the build options exactly as the compiler sees them.

#include <stdio.h>
#include <math.h>
/*
 * main.c
 */
int main(void)
{
    printf("Hello World!\n");

    return 0;
}

The CCS console build output is:

**** Build of configuration Debug for project RM48_fp_mode ****

"C:\\ti_ccs6_1\\ccsv6\\utils\\bin\\gmake" -k all 
'Building file: ../main.c'
'Invoking: ARM Compiler'
"C:/ti_ccs6_1/ccsv6/tools/compiler/ti-cgt-arm_15.12.1.LTS/bin/armcl" -mv7R4 --code_state=32 --float_support=VFPv3D16 -me --fp_mode=relaxed --include_path="C:/ti_ccs6_1/ccsv6/tools/compiler/ti-cgt-arm_15.12.1.LTS/include" -g --float_operations_allowed=32 --diag_wrap=off --diag_warning=225 --display_error_number --enum_type=packed --abi=eabi --preproc_with_compile --preproc_dependency="main.pp"  "../main.c"
"C:/ti_ccs6_1/ccsv6/tools/compiler/ti-cgt-arm_15.12.1.LTS/include/math.h", line 200 (col. 17): error #1558-D: 64-bit floating point operations are not allowed
"C:/ti_ccs6_1/ccsv6/tools/compiler/ti-cgt-arm_15.12.1.LTS/include/math.h", line 200 (col. 17): error #1558-D: 64-bit floating point operations are not allowed
"C:/ti_ccs6_1/ccsv6/tools/compiler/ti-cgt-arm_15.12.1.LTS/include/math.h", line 202 (col. 18): error #1558-D: 64-bit floating point operations are not allowed
"C:/ti_ccs6_1/ccsv6/tools/compiler/ti-cgt-arm_15.12.1.LTS/include/math.h", line 202 (col. 18): error #1558-D: 64-bit floating point operations are not allowed
"C:/ti_ccs6_1/ccsv6/tools/compiler/ti-cgt-arm_15.12.1.LTS/include/math.h", line 206 (col. 22): error #1558-D: 64-bit floating point operations are not allowed
"C:/ti_ccs6_1/ccsv6/tools/compiler/ti-cgt-arm_15.12.1.LTS/include/math.h", line 206 (col. 22): error #1558-D: 64-bit floating point operations are not allowed
"C:/ti_ccs6_1/ccsv6/tools/compiler/ti-cgt-arm_15.12.1.LTS/include/math.h", line 208 (col. 18): error #1558-D: 64-bit floating point operations are not allowed
"C:/ti_ccs6_1/ccsv6/tools/compiler/ti-cgt-arm_15.12.1.LTS/include/math.h", line 208 (col. 18): error #1558-D: 64-bit floating point operations are not allowed
"C:/ti_ccs6_1/ccsv6/tools/compiler/ti-cgt-arm_15.12.1.LTS/include/math.h", line 208 (col. 18): error #1558-D: 64-bit floating point operations are not allowed
"C:/ti_ccs6_1/ccsv6/tools/compiler/ti-cgt-arm_15.12.1.LTS/include/math.h", line 208 (col. 19): error #1558-D: 64-bit floating point operations are not allowed
10 errors detected in the compilation of "../main.c".

>> Compilation failure
gmake: *** [main.obj] Error 1
gmake: Target `all' not remade because of errors.

**** Build Finished ****

The errors are coming from inline functions in math.h, such as the following:

/*---------------------------------------------------------------------------*/
/* The FPv4SP supported on Cortex-M4 does not have double precision hardware */
/* so avoid using the intrinsic.                                             */
/*---------------------------------------------------------------------------*/
#if !defined(__TI_FPv4SPD16_SUPPORT__)
#define sqrt  __relaxed_sqrt
__inline double __relaxed_sqrt(double x)
{
    return __sqrt(x);
}