rs6000: Don't use TFmode for 128 bits fp constant in toc [PR110011]

Hi,

As PR110011 shows, when encoding 128 bits fp constant into
toc, we adopts REAL_VALUE_TO_TARGET_LONG_DOUBLE which is
to find the first float mode with LONG_DOUBLE_TYPE_SIZE
bits of precision, it would be TFmode here.  But the 128
bits fp constant can be with mode IFmode or KFmode, which
doesn't necessarily have the same underlying float format
as the one of TFmode, like this PR exposes, with option
-mabi=ibmlongdouble TFmode has ibm_extended_format while
KFmode has ieee_quad_format, mixing up the formats (the
encoding/decoding ways) would cause unexpected results.

This patch is to make it use constant's own mode instead
of TFmode for real_to_target call.

Bootstrapped and regtested on powerpc64-linux-gnu P7/P8/P9 and
powerpc64le-linux-gnu P9 and P10.

I'll push this next week if no objections.

BR,
Kewen
-----
	PR target/110011

gcc/ChangeLog:

	* config/rs6000/rs6000.cc (output_toc): Use its own mode of the
	128-bit float constant for real_to_target call.

gcc/testsuite/ChangeLog:

	* gcc.target/powerpc/pr110011.c: New test.
---
 gcc/config/rs6000/rs6000.cc                 |  2 +-
 gcc/testsuite/gcc.target/powerpc/pr110011.c | 42 +++++++++++++++++++++
 2 files changed, 43 insertions(+), 1 deletion(-)
 create mode 100644 gcc/testsuite/gcc.target/powerpc/pr110011.c

--
2.31.1

On Tue, Jun 6, 2023 at 5:20 AM Kewen.Lin <linkw@linux.ibm.com> wrote:

> Hi,
>
> As PR110011 shows, when encoding 128 bits fp constant into
> toc, we adopts REAL_VALUE_TO_TARGET_LONG_DOUBLE which is
> to find the first float mode with LONG_DOUBLE_TYPE_SIZE
> bits of precision, it would be TFmode here.  But the 128
> bits fp constant can be with mode IFmode or KFmode, which
> doesn't necessarily have the same underlying float format
> as the one of TFmode, like this PR exposes, with option
> -mabi=ibmlongdouble TFmode has ibm_extended_format while
> KFmode has ieee_quad_format, mixing up the formats (the
> encoding/decoding ways) would cause unexpected results.
>
> This patch is to make it use constant's own mode instead
> of TFmode for real_to_target call.
>
> Bootstrapped and regtested on powerpc64-linux-gnu P7/P8/P9 and
> powerpc64le-linux-gnu P9 and P10.
>
> I'll push this next week if no objections.
>
> BR,
> Kewen
> -----
>         PR target/110011
>
> gcc/ChangeLog:
>
>         * config/rs6000/rs6000.cc (output_toc): Use its own mode of the
>         128-bit float constant for real_to_target call.
>

The comment wording can be worded better.  Maybe

Use the mode of the 128-bit floating constant itself for real_to_target
call.

This is okay.

Thanks, David


> gcc/testsuite/ChangeLog:
>
>         * gcc.target/powerpc/pr110011.c: New test.
> ---
>  gcc/config/rs6000/rs6000.cc                 |  2 +-
>  gcc/testsuite/gcc.target/powerpc/pr110011.c | 42 +++++++++++++++++++++
>  2 files changed, 43 insertions(+), 1 deletion(-)
>  create mode 100644 gcc/testsuite/gcc.target/powerpc/pr110011.c
>
> diff --git a/gcc/config/rs6000/rs6000.cc b/gcc/config/rs6000/rs6000.cc
> index 3f129ea37d2..330c6a6fa5f 100644
> --- a/gcc/config/rs6000/rs6000.cc
> +++ b/gcc/config/rs6000/rs6000.cc
> @@ -17314,7 +17314,7 @@ output_toc (FILE *file, rtx x, int labelno,
> machine_mode mode)
>        if (DECIMAL_FLOAT_MODE_P (GET_MODE (x)))
>         REAL_VALUE_TO_TARGET_DECIMAL128 (*CONST_DOUBLE_REAL_VALUE (x), k);
>        else
> -       REAL_VALUE_TO_TARGET_LONG_DOUBLE (*CONST_DOUBLE_REAL_VALUE (x), k);
> +       real_to_target (k, CONST_DOUBLE_REAL_VALUE (x), GET_MODE (x));
>
>        if (TARGET_64BIT)
>         {
> diff --git a/gcc/testsuite/gcc.target/powerpc/pr110011.c
> b/gcc/testsuite/gcc.target/powerpc/pr110011.c
> new file mode 100644
> index 00000000000..5b04d3e298a
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/powerpc/pr110011.c
> @@ -0,0 +1,42 @@
> +/* { dg-do run } */
> +/* { dg-require-effective-target float128_runtime } */
> +/* Force long double to be with IBM format here, to verify
> +   _Float128 constant still uses its own format (IEEE) for
> +   encoding rather than IBM format.  */
> +/* { dg-options "-mfp-in-toc -mabi=ibmlongdouble" } */
> +/* { dg-add-options float128 } */
> +
> +#define MPFR_FLOAT128_MAX 0x1.ffffffffffffffffffffffffffffp+16383f128
> +
> +__attribute__ ((noipa))
> +_Float128 f128_max ()
> +{
> +  return MPFR_FLOAT128_MAX;
> +}
> +
> +typedef union
> +{
> +  int w[4];
> +  _Float128 f128;
> +} U;
> +
> +int main ()
> +{
> +
> +  U umax;
> +  umax.f128 = f128_max ();
> +  /* ieee float128 max:
> +     7ffeffff ffffffff ffffffff ffffffff.  */
> +  if (umax.w[1] != 0xffffffff || umax.w[2] != 0xffffffff)
> +    __builtin_abort ();
> +#ifdef __LITTLE_ENDIAN__
> +  if (umax.w[0] != 0xffffffff || umax.w[3] != 0x7ffeffff)
> +    __builtin_abort ();
> +#else
> +  if (umax.w[3] != 0xffffffff || umax.w[0] != 0x7ffeffff)
> +    __builtin_abort ();
> +#endif
> +
> +  return 0;
> +}
> +
> --
> 2.31.1
>

on 2023/6/11 10:04, David Edelsohn wrote:
> On Tue, Jun 6, 2023 at 5: 20 AM Kewen. Lin <linkw@ linux. ibm. com> wrote: Hi, As PR110011 shows, when encoding 128 bits fp constant into toc, we adopts REAL_VALUE_TO_TARGET_LONG_DOUBLE which is to find the first float mode with LONG_DOUBLE_TYPE_SIZE
> ZjQcmQRYFpfptBannerStart
> This Message Is From an External Sender
> This message came from outside your organization.
>  
> ZjQcmQRYFpfptBannerEnd
> On Tue, Jun 6, 2023 at 5:20 AM Kewen.Lin <linkw@linux.ibm.com <mailto:linkw@linux.ibm.com>> wrote:
> 
>     Hi,
> 
>     As PR110011 shows, when encoding 128 bits fp constant into
>     toc, we adopts REAL_VALUE_TO_TARGET_LONG_DOUBLE which is
>     to find the first float mode with LONG_DOUBLE_TYPE_SIZE
>     bits of precision, it would be TFmode here.  But the 128
>     bits fp constant can be with mode IFmode or KFmode, which
>     doesn't necessarily have the same underlying float format
>     as the one of TFmode, like this PR exposes, with option
>     -mabi=ibmlongdouble TFmode has ibm_extended_format while
>     KFmode has ieee_quad_format, mixing up the formats (the
>     encoding/decoding ways) would cause unexpected results.
> 
>     This patch is to make it use constant's own mode instead
>     of TFmode for real_to_target call.
> 
>     Bootstrapped and regtested on powerpc64-linux-gnu P7/P8/P9 and
>     powerpc64le-linux-gnu P9 and P10.
> 
>     I'll push this next week if no objections.
> 
>     BR,
>     Kewen
>     -----
>             PR target/110011
> 
>     gcc/ChangeLog:
> 
>             * config/rs6000/rs6000.cc (output_toc): Use its own mode of the
>             128-bit float constant for real_to_target call.
> 
> 
> The comment wording can be worded better.  Maybe 
> 
> Use the mode of the 128-bit floating constant itself for real_to_target call.
> 
> This is okay.

Thanks David, I've updated changelog as you suggested and pushed as r14-1703.
Will backport after burn-in time.

BR,
Kewen

> 
> Thanks, David
> 
> 
>     gcc/testsuite/ChangeLog:
> 
>             * gcc.target/powerpc/pr110011.c: New test.
>     ---
>      gcc/config/rs6000/rs6000.cc                 |  2 +-
>      gcc/testsuite/gcc.target/powerpc/pr110011.c | 42 +++++++++++++++++++++
>      2 files changed, 43 insertions(+), 1 deletion(-)
>      create mode 100644 gcc/testsuite/gcc.target/powerpc/pr110011.c
> 
>     diff --git a/gcc/config/rs6000/rs6000.cc b/gcc/config/rs6000/rs6000.cc
>     index 3f129ea37d2..330c6a6fa5f 100644
>     --- a/gcc/config/rs6000/rs6000.cc
>     +++ b/gcc/config/rs6000/rs6000.cc
>     @@ -17314,7 +17314,7 @@ output_toc (FILE *file, rtx x, int labelno, machine_mode mode)
>            if (DECIMAL_FLOAT_MODE_P (GET_MODE (x)))
>             REAL_VALUE_TO_TARGET_DECIMAL128 (*CONST_DOUBLE_REAL_VALUE (x), k);
>            else
>     -       REAL_VALUE_TO_TARGET_LONG_DOUBLE (*CONST_DOUBLE_REAL_VALUE (x), k);
>     +       real_to_target (k, CONST_DOUBLE_REAL_VALUE (x), GET_MODE (x));
> 
>            if (TARGET_64BIT)
>             {
>     diff --git a/gcc/testsuite/gcc.target/powerpc/pr110011.c b/gcc/testsuite/gcc.target/powerpc/pr110011.c
>     new file mode 100644
>     index 00000000000..5b04d3e298a
>     --- /dev/null
>     +++ b/gcc/testsuite/gcc.target/powerpc/pr110011.c
>     @@ -0,0 +1,42 @@
>     +/* { dg-do run } */
>     +/* { dg-require-effective-target float128_runtime } */
>     +/* Force long double to be with IBM format here, to verify
>     +   _Float128 constant still uses its own format (IEEE) for
>     +   encoding rather than IBM format.  */
>     +/* { dg-options "-mfp-in-toc -mabi=ibmlongdouble" } */
>     +/* { dg-add-options float128 } */
>     +
>     +#define MPFR_FLOAT128_MAX 0x1.ffffffffffffffffffffffffffffp+16383f128
>     +
>     +__attribute__ ((noipa))
>     +_Float128 f128_max ()
>     +{
>     +  return MPFR_FLOAT128_MAX;
>     +}
>     +
>     +typedef union
>     +{
>     +  int w[4];
>     +  _Float128 f128;
>     +} U;
>     +
>     +int main ()
>     +{
>     +
>     +  U umax;
>     +  umax.f128 = f128_max ();
>     +  /* ieee float128 max:
>     +     7ffeffff ffffffff ffffffff ffffffff.  */
>     +  if (umax.w[1] != 0xffffffff || umax.w[2] != 0xffffffff)
>     +    __builtin_abort ();
>     +#ifdef __LITTLE_ENDIAN__
>     +  if (umax.w[0] != 0xffffffff || umax.w[3] != 0x7ffeffff)
>     +    __builtin_abort ();
>     +#else
>     +  if (umax.w[3] != 0xffffffff || umax.w[0] != 0x7ffeffff)
>     +    __builtin_abort ();
>     +#endif
>     +
>     +  return 0;
>     +}
>     +
>     --
>     2.31.1
>