Runtime does not compile when Has_FPU is set to False

[NicoPy]

Hi,

To do some tests, I need to compile the runtime (ravenscar full) with soft Float instead of hard Float. Here is what I did :

• set _HasFPU to False in s-bbpara__xxxx.ads.
• make _hasfpu() property return False (cortexm.py)
• make _has_single_precisionfpu() property return False in (cortexm.py)
• make _has_double_precisionfpu() property return False in (cortexm.py)
• remove the -mfpu= switch in _compilerswitches() (cortexm.py)
• add the -msoft-float switch in _compilerswitches() (cortexm.py)

Setting _HasFPU to False makes the runtime compilation fail :

C:\Users\Nicolas\AppData\Local\Temp\cctSKb2B.s: Assembler messages:
C:\Users\Nicolas\AppData\Local\Temp\cctSKb2B.s:83: Error: selected FPU does not support instruction -- `vstmdbeq r12!,{s16-s31}'
C:\Users\Nicolas\AppData\Local\Temp\cctSKb2B.s:94: Error: selected FPU does not support instruction -- `vldmiane r12!,{s16-s31}'

This error is not a surprise since I instructed the compiler to not use hard float instructions (-msoft-float switch).

However, setting _HasFPU to False makes the runtime compilation fail :

s-statxd.adb:396:22: construct not allowed in configurable run-time mode
s-statxd.adb:396:22: file s-fatflt.ads not found
s-statxd.adb:396:22: entity "System.Fat_Flt.Attr_Float" not available

When I fix this error by manually disabling Streams, the same error rises again somewhere else.

How can I fix this problem ?

[NicoPy]

I have understood how the FPU configuration works (for ARM architecture). Here it is for reference.

cortexm.py is used to configure the runtime sources/configuration :

• _hasfpu() property must always be set to True else the runtime does not compile.

• To use software floating point calculations, _has_single_precisionfpu() and _has_double_precisionfpu() properties must return False. You also have to set -msoft-float switch in _compilerswitches().

• To use hardware floating point calculations, _has_single_precisionfpu() or _has_double_precisionfpu() or both must return True. You also have to set -mfpu=xxx switch to the correct value in _compilerswitches().

In _s-bbparaxxxx.ads_ the constant _HasFPU is indirectly related to the runtime configuration. When _HasFPU is True, the floating point context is saved/restored in _Pend_SVHandler() (_s-bcpcstpendsv.adb_).

• To use software floating point calculations, set _HasFPU to False.
• To use hardware floating point calculations, set _HasFPU to True.

To be more complete :	_has_double_precisionfpu()	_has_single_precisionfpu()	-->		compiler switch	_HasFPU
False	False	-->	soft float	-msoft-float	False
False	True	-->	single precision : hard double precision : soft	-mfpu=xxx	True
True	False	-->	single precision : soft double precision : hard	-mfpu=xxx	True
True	True	-->	single precision : hard double precision : hard	-mfpu=xxx	True

[Fabien-Chouteau]

Hi @NicoPy, I am glad you found the answer :)