Hello,
Can somebody point me to some general places, where I can look on how FPU
support is implemented. I understand that there is not one place, where things
done, the only I know, that there should be places for detecting FPU hardware
and places with assembly routines for pushing around floating point registers,
assembly/C-hybrid code for handling floating point exceptions.
I'm interesting mostly in x86 (386) things, not other architectures.
If you know some good docs, please point me to them.
P.S. Sorry if it's a wrong list to ask such questions.
--
Best regards, Evgeniy.
Key Fingerprint: F316B5A1F6D2054FCD18B74A95400ABB1FE567A3
On 03/25/2009 12:09 PM, Evgeniy Ivanov wrote:
> Hello,
>
> Can somebody point me to some general places, where I can look on how FPU
> support is implemented. I understand that there is not one place, where things
> done, the only I know, that there should be places for detecting FPU hardware
> and places with assembly routines for pushing around floating point registers,
I think you want to start with
arch/x86/kernel/i387.c
arch/x86/math-emu/
> assembly/C-hybrid code for handling floating point exceptions.
and arch/x86/kernel/traps.c