On 17 August 2011 20:19, Bryce Lanham <blan...@gmail.com> wrote:
> I have been working on adding NeXT black hardware support to QEMU for
> Google Summer of Code. While only a base level of hardware is
> implemented, I need to have the code in upstream before the end of the
> program. I based my work off of Laurent Vivier's work on the m68k
> core, so that work would also need to be upstreamed as well.
You've messed something up in your patch generation process here:
there are a lot of changes which have no relation to NeXT support,
like:
> diff --git bswap.h bswap.h
> index f41bebe..13fd5cf 100644
> --- bswap.h
> +++ bswap.h
> @@ -533,7 +533,7 @@ static inline int lduw_be_p(const void *ptr)
> : "m" (*(uint16_t *)ptr));
> return val;
> #else
> - const uint8_t *b = ptr;
> + const uint8_t *b = (const uint8_t *)ptr;
> return ((b[0] << 8) | b[1]);
> #endif
> }
and:
> diff --git fpu/softfloat.c fpu/softfloat.c
> index 7951a0e..dbce683 100644
> --- fpu/softfloat.c
> +++ fpu/softfloat.c
> @@ -768,7 +768,9 @@ static floatx80
> ) {
> return packFloatx80( zSign, 0x7FFE, ~ roundMask );
> }
> - return packFloatx80( zSign, 0x7FFF, LIT64( 0x8000000000000000 )
> );
> + return packFloatx80( zSign,
> + floatx80_default_inf_high,
> + floatx80_default_inf_low );
> }
> if ( zExp <= 0 ) {
> isTiny =
just to pick a few early examples.
Please fix this and try again; this isn't reviewable as it stands,
I'm afraid.
Large new features like a new platform and devices should
also ideally be split into coherent logical patches for ease of
review. http://wiki.qemu.org/Contribute/SubmitAPatch has a few
further remarks. I'd also suggest using git-format-patch / git-send-email
to get the patch mail formatting right.
-- PMM
