On Tue, Jul 5, 2011 at 11:14 AM, Uros Bizjak <ubizjak@gmail.com> wrote:
> On Tue, Jul 5, 2011 at 7:54 PM, H.J. Lu <hjl.tools@gmail.com> wrote:
>
>>>> I'd like to start submitting a series of patches to enable x32:
>>>>
>>>> https://sites.google.com/site/x32abi/
>>>>
>>>> The GCC x32 branch is very stable. There are no unexpected failures in
>>>> C, C++, Fortran and Objective C testsuites.  SPEC CPU 2K/2006 compile
>>>> and run correctly at -O2 and -O3.
>>>>
>>>> More than 90% of changes are in x86 backend.  I have submitted non-x86
>>>> backend patches.  Most of them have been reviewed and checked in.  Only
>>>> 4 patches are pending for review/approval.
>>>>
>>>> This is the first x86 backend patch to support x32.  By default, x32 is
>>>> disabled and x32 run-time support isn't required.  OK for trunk?
>>>
>>> Please strip out --enable-ia32 stuff, it complicates things ATM.  I
>>> assume that --enable-x32 applies only to 64bit targets, so this part
>
>
> I think that better name of the file would be "t-linux64-x32".
>
> @@ -2631,6 +2640,7 @@ esac
>  case ${target} in
>  i[34567]86-*-linux* | x86_64-*-linux*)
>        tmake_file="${tmake_file} i386/t-pmm_malloc i386/t-i386"
> +       libgcc_tm_file="${libgcc_tm_file} i386/value-unwind.h"
>
> Not yet.

Removed.

> @@ -58,25 +58,31 @@ see the files COPYING3 and COPYING.RUNTIME
> respectively.  If not, see
>
>  #if TARGET_64BIT_DEFAULT
>  #define SPEC_32 "m32"
> -#define SPEC_64 "!m32"
> +#define SPEC_64 "m32|mx32:;"
> +#define SPEC_X32 "mx32"
>  #else
> -#define SPEC_32 "!m64"
> +#define SPEC_32 "m64|mx32:;"
>  #define SPEC_64 "m64"
> +#define SPEC_X32 "mx32"
>  #endif
>
> I really think that "!(m64|mx32)" is more descriptive and clear...

I don't know if gcc spec supports it.  I just followed mips backend.

>  #undef LINK_SPEC
>  #define LINK_SPEC "%{" SPEC_64 ":-m " GNU_USER_LINK_EMULATION64 "} \
>                    %{" SPEC_32 ":-m " GNU_USER_LINK_EMULATION32 "} \
> +                   %{" SPEC_X32 ":-m " GNU_USER_LINK_EMULATIONX32 "} \
>   %{shared:-shared} \
>   %{!shared: \
>     %{!static: \
>       %{rdynamic:-export-dynamic} \
>       %{" SPEC_32 ":-dynamic-linker " GNU_USER_DYNAMIC_LINKER32 "} \
> -      %{" SPEC_64 ":-dynamic-linker " GNU_USER_DYNAMIC_LINKER64 "}} \
> +      %{" SPEC_64 ":-dynamic-linker " GNU_USER_DYNAMIC_LINKER64 "} \
> +      %{" SPEC_X32 ":-dynamic-linker " GNU_USER_DYNAMIC_LINKERX32 "}} \
>     %{static:-static}}"
>
> On the border of bikesheding, GNU_USER_LINK_EMULATION64_X32 and
> GNU_USER_DYNAMIC_LINKER64_X32 sounds better to me.
>
> Same with the below:
>
> +#define GLIBC_DYNAMIC_LINKERX32 "/libx32/ld-linux-x32.so.2"
> +#define UCLIBC_DYNAMIC_LINKERX32 "/lib/ldx32-uClibc.so.0"
> +#define BIONIC_DYNAMIC_LINKERX32 "/system/bin/linkerx32"
>
> +++ b/gcc/config/i386/t-linux-x32
>
> Please rename above file to t-linux64-x32.

X32 is the name of the psABI:

https://sites.google.com/site/x32abi/

We have -mx32, -m32 and -m64 command line options and
There are macros like TARGET_X32. I'd like to be consistent
and avoid "64" when referring to x32 if possible. But I won't
insist.  Please let me know that you really won't like x32 without
64.

> With above small changes, the patch looks OK to me. Please also wait
> for build and options maintainer (CC'd) approvals.
>
> Thanks,
> Uros.
>

Thanks.


-- 
H.J.