On 2/16/07, Robert P. J. Day <rpjday@xxxxxxxxxxxxxx> wrote:
i'm curious about the rationale for why a header file would be placed in include/asm-generic as opposed to include/linux. as an example, consider include/asm-generic/cputime.h. if you check that file, there's nothing in it that seems tied to assembler, and it looks like it would fit just fine under include/linux.
asm, in this context, isn't ASseMbly, it's Architecture-Specific Module (IIRC). The idea is some architectures may have fanciness that allows for special optimizations (or just plain different ways of doing things). Then, we can have generic code #include <asm/somefile.h> and the arch-specific file can have special code if it needs it, or just #include <asm-generic/somefile.h>, it it doesn't. In cputime.h's case, only powerpc & s390 have special stuff, afaics. So every other arch uses asm-generic/cputime.h
so what determines where a header file ends up? thanks.
Whether there are architecture specific bits or not. Thanks, Nish -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx Please read the FAQ at http://kernelnewbies.org/FAQ
