On Tue, 7 Jun 2011, Richard Henderson wrote:
> On 06/06/2011 04:07 PM, malc wrote:
> > Depends on how bswap_16 is defined. If it is __builtin_bswap16
> > then 4.5.0 and 4.6.0 generate byte reversed loads, and previous
> > versions lack that builtin, so i don't think this generic code
> > should go
On 06/06/2011 04:07 PM, malc wrote:
> Depends on how bswap_16 is defined. If it is __builtin_bswap16
> then 4.5.0 and 4.6.0 generate byte reversed loads, and previous
> versions lack that builtin, so i don't think this generic code
> should go in.
It would continue to be defined as-is, without dir
On Mon, 6 Jun 2011, Richard Henderson wrote:
> Patches 1-3:
> Reviewed-by: Richard Henderson
>
> That said,
>
> On 06/06/2011 09:25 AM, Paolo Bonzini wrote:
> > +/* conservative code for little endian unaligned accesses */
> > +static inline int lduw_le_p(const void *ptr)
> > +{
> > +#ifdef _AR
On 06/06/2011 01:27 PM, Paolo Bonzini wrote:
> On 06/06/2011 09:56 PM, Richard Henderson wrote:
>> Can we add a patch 4/3 that removes this sort of hard-coded
>> assembly stuff in favour of generic gcc code.
>
> Then it's also possible to commit patch 2 from this submission, and
> wait while I rew
On 06/06/2011 09:56 PM, Richard Henderson wrote:
Can we add a patch 4/3 that removes this sort of hard-coded
assembly stuff in favour of generic gcc code.
Then it's also possible to commit patch 2 from this submission, and wait
while I rework the other two to use generic GCC code.
Paolo
Patches 1-3:
Reviewed-by: Richard Henderson
That said,
On 06/06/2011 09:25 AM, Paolo Bonzini wrote:
> +/* conservative code for little endian unaligned accesses */
> +static inline int lduw_le_p(const void *ptr)
> +{
> +#ifdef _ARCH_PPC
> +int val;
> +__asm__ __volatile__ ("lhbrx %0,0,%1
This is just code movement, and moving the fpu/ include path from
target-dependent to target-independent Make variables.
Signed-off-by: Paolo Bonzini
---
Makefile.hw |2 +-
bswap.h | 474 +++
configure |3 +-
cpu-all.h | 4
