topic how to use float operation in uboot codes in i.MX Processors

how to use float operation in uboot codes

yibhan — Tue, 07 Jul 2020 01:11:44 GMT

Hi guys,

I am porting some of functions those are using float operation, after build, there are some of link error.

My kernel version is 4.14.98 and u-boot version is kernel's package.

link error is below:

"undefined reference to '__aeabi_fmul'"

Should I modify compile and link option? if so, where can i modify and some of points.

Please share me.

yibhan — Tue, 07 Jul 2020 01:15:11 GMT

My psudo codes is below:

float pdiff = (float) (int temp);

float ratio = (float) (int temp1(/ (int temp2);

int tmpval = (int) ( int temp3 + pdiff * ratio);

Yuri — Tue, 07 Jul 2020 05:31:11 GMT

Hello,

U-boot is not intended to support floating point; it should provide minimal

options just to boot a Linux system. You may look at Chapter 3 (Porting U-Boot)

of ''i.MX_Porting_Guide.pdf" in NXP Linux documentation and try to modify

building options in make file, but, again, floating point support has been never

tested and guaranteed.

Regards,

Yuri.