Re: Compiler options when building for an Avanti alpha computer

From: Rafael E. Herrera (raffo@neuronet.pitt.edu)
Date: Tue Jul 11 2000 - 19:37:45 EST


David Huggins-Daines wrote:
>
> "Rafael E. Herrera" <raffo@neuronet.pitt.edu> writes:
>
> > One of the oddities trying to build them has been the compiler options,
> > the look like this:
> >
> > gcc -D__KERNEL__ -I/usr/src/linux/include -Wall -Wstrict-prototypes -O2
> > -fomit-frame-pointer -Werror -fno-strict-aliasing -pipe -mno-fp-regs
> > -ffixed-8 -mcpu=ev4 -Wa,-mev6 -DEXPORT_SYMTAB -c ...
> >
> > my question is about the option, should that be -Wa,-mev64?
>
> Read the GCC info pages and the comments in arch/alpha/Makefile. The
> -Wa,-mev6 is an option passed to the assembler which tells it to
> assemble for an EV6 target. This is needed for building generic
> kernels because otherwise the assembler might treat BWX (byte/word
> I/O) instructions as macros and emulate them in code that is specific
> to EV56 and above (the generic kernel is built with -mcpu=ev5).
>
> For other targets it does nothing, because when the compiler is given
> it will never generate those instructions.
>
> Anyway EV64 doesn't exist - IIRC that would mean an EV6 core on an EV4
> (.5 micron) process :-)

-Wa,-mev64 was a typo, I meant -Wa,-mev4. My question was geared to what
happens when you don't have a ev6 machine. My AS400 4/233 is an ev4
machine. Will the generated code work well?

Every test kernel I've built so far fails to mount the root file system,
fails to open a console, or just stops after the step in which it
releases memory (init, kflushd, kupdate, and one other are running, but
the boot sequence fail)

-- 
Rafael

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Sat Jul 15 2000 - 21:00:13 EST