Re: [libav-devel] [PATCH 3/5] configure: Factorize check_64_bit()

2017-11-10 Thread Luca Barbato

On 10/11/2017 19:54, Diego Biurrun wrote:

On Fri, Nov 10, 2017 at 07:47:44PM +0100, Luca Barbato wrote:

On 03/11/2017 17:43, Diego Biurrun wrote:

---

I'm assuming that the change for MIPS is safe.


 From what I recall of the many many ABIs of MIPS it should not.


Can you elaborate, I'm not following ..



Check how _MIPS_SIM is set, I'm afraid there is at least one abi for 
which _MIPS_SIM is 0 while sizeof(ptr) > 4.


lu


___
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel

Re: [libav-devel] [PATCH 3/5] configure: Factorize check_64_bit()

2017-11-10 Thread Diego Biurrun
On Fri, Nov 10, 2017 at 07:47:44PM +0100, Luca Barbato wrote:
> On 03/11/2017 17:43, Diego Biurrun wrote:
> > ---
> > 
> > I'm assuming that the change for MIPS is safe.
> 
> From what I recall of the many many ABIs of MIPS it should not.

Can you elaborate, I'm not following ..

Diego
___
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel

Re: [libav-devel] [PATCH 3/5] configure: Factorize check_64_bit()

2017-11-10 Thread Luca Barbato

On 03/11/2017 17:43, Diego Biurrun wrote:

---

I'm assuming that the change for MIPS is safe.


From what I recall of the many many ABIs of MIPS it should not.

lu

___
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel

Re: [libav-devel] [PATCH 3/5] configure: Factorize check_64_bit()

2017-11-10 Thread James Almer
On 11/3/2017 1:43 PM, Diego Biurrun wrote:
> ---
> 
> I'm assuming that the change for MIPS is safe.

If you can't test it on a MIPS target, then it would be best to ask
someone who can.
It probably has its own special check for a reason.

> 
>  configure | 15 +++
>  1 file changed, 7 insertions(+), 8 deletions(-)
> 
> diff --git a/configure b/configure
> index 169a6fa33f..21f24e8e0f 100755
> --- a/configure
> +++ b/configure
> @@ -3764,8 +3764,7 @@ check_host_cflags $host_cflags_speed
>  check_64bit(){
>  arch32=$1
>  arch64=$2
> -expr=$3
> -check_code cc "" "int test[2*($expr) - 1]" &&
> +check_code cc "" "int test[2*(sizeof(void *) > 4) - 1]" &&
>  subarch=$arch64 || subarch=$arch32
>  enable $subarch
>  }
> @@ -3775,27 +3774,27 @@ case "$arch" in
>  enabled shared && enable_weak pic
>  ;;
>  mips)
> -check_64bit mips mips64 '_MIPS_SIM > 1'
> +check_64bit mips mips64
>  enabled shared && enable_weak pic
>  ;;
>  parisc)
> -check_64bit parisc parisc64 'sizeof(void *) > 4'
> +check_64bit parisc parisc64
>  enabled shared && enable_weak pic
>  ;;
>  ppc)
> -check_64bit ppc ppc64 'sizeof(void *) > 4'
> +check_64bit ppc ppc64
>  enabled shared && enable_weak pic
>  ;;
>  s390)
> -check_64bit s390 s390x 'sizeof(void *) > 4'
> +check_64bit s390 s390x
>  enabled shared && enable_weak pic
>  ;;
>  sparc)
> -check_64bit sparc sparc64 'sizeof(void *) > 4'
> +check_64bit sparc sparc64
>  enabled shared && enable_weak pic
>  ;;
>  x86)
> -check_64bit x86_32 x86_64 'sizeof(void *) > 4'
> +check_64bit x86_32 x86_64
>  if enabled x86_64; then
>  enabled shared && enable_weak pic
>  objformat=elf64
> 

Should be good if MIPS64 is confirmed to not break with it.
___
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel