about RTEMS toolchain target mips32 little endian

Ralf Corsepius ralf.corsepius@rtems.org
Mon Jan 7 15:26:00 GMT 2013


On 01/05/2013 03:11 AM, Devin wrote:
>
> Thanks Ralf ,
> I have removed line "MULTILIB_EXCEPTIONS += mips32/EL*" from "t-rtems".
> and then configure&make&make install  GCC .
> lastly , I compile RTEMS successfully .
>
> But why gcc add "mips32/El*" to MULTILIB_EXCEPTIONS ?
I regret, but I don't recall the details.

Normally, there are 2 reasons for us to remove a multilib:
- Support for a multilib-variant is missing in RTEMS
- Nobody uses a multilib-variant.

IIRC, both had applied in this case.

> Maybe  it will causes
> some problems or errors ?
c.f. above.

The initial, minimum requirements for me to add the mips32/EL multilib 
would be to know "gcc/newlib and rtems compiles" and to know people are 
actually using it.

> I hope  RTEMS  compiled  in mips32 EL model  will run stably.
I have no idea, whether RTEMS even compiles for mips32/EL nor do I know 
if the results are usable. I can try to activate mips32/EL multilib in 
my next toolchain spin and watch for what will happen ;)

Ralf



--
For unsubscribe information see http://sourceware.org/lists.html#faq



More information about the crossgcc mailing list