From: Richard Biener <rguenther@suse.de>
To: Jakub Jelinek <jakub@redhat.com>
Cc: gcc-patches@gcc.gnu.org
Subject: Re: [PATCH] call-cdce: Add missing BUILT_IN_*F{32,64}X handling and improve BUILT_IN_*L [PR113993]
Date: Thu, 22 Feb 2024 10:14:45 +0100 (CET) [thread overview]
Message-ID: <p650s89o-8564-07r0-4r5p-0qs88o697o57@fhfr.qr> (raw)
In-Reply-To: <ZdW2rkXeVr77plE+@tucnak>
On Wed, 21 Feb 2024, Jakub Jelinek wrote:
> Hi!
>
> The following testcase ICEs, because can_test_argument_range
> returns true for BUILT_IN_{COSH,SINH,EXP{,M1,2}}{F32X,F64X}
> among many other builtins, but get_no_error_domain doesn't handle
> those.
>
> float32x_type_node when supported in GCC always has DFmode, so that
> case is easy (and call-cdce assumes that SFmode is IEEE float and DFmode
> is IEEE double). So *F32X is simply handled by adding those cases
> next to *F64.
> float64x_type_node when supported in GCC by definition has a mode
> with larger precision and exponent range than DFmode, so it can be XFmode,
> TFmode or KFmode. I went through all the l/f128 suffixed builtins and
> verified that the float128x_type_node no error domain range is actually
> identical to the Intel extended long double no error domain range; it isn't
> that surprising, both IEEE quad and Intel/Motorola extended have the same
> exponent range [-16381, 16384] (well, Motorola -16382 probably because of
> different behavior for denormals, but that has nothing to do with
> get_no_error_domain which is about large inputs overflowing into +-Inf
> or triggering NaN, denormals could in theory do something solely for sqrt
> and even that is fine). In theory some target could have different larger
> type, so for *F64X the code verifies that
> REAL_MODE_FORMAT (TYPE_MODE (float64x_type_node))->emax == 16384
> and if so, uses the *F128 domains, otherwise falls back to the non-suffixed
> ones (aka *F64), that is certainly the conservative minimum.
> While at it, the patch also changes the *L suffixed cases to do pretty much
> the same, the comment said that the function just assumes for *L
> the *F64 ranges, but that is unnecessarily conservative.
> All we currently have for long double is:
> 1) IEEE quad (emax 16384, *F128 ranges)
> 2) XFmode Intel/Motorola extended (emax 16384, same as *F128 ranges)
> 3) IBM extended (double double, emax 1024, the extra precision doesn't
> really help and the domains are the same as for *F64)
> 4) same as double (*F64 again)
> So, the patch uses also for *L
> REAL_MODE_FORMAT (TYPE_MODE (long_double_type_node))->emax == 16384
> checks and either tail recurses into the *F128 case for that or to
> non-suffixed (aka *F64) case otherwise.
> BUILT_IN_*F128X not handled because no target has those and it doesn't
> seem something is on the horizon and who knows what would be used for that.
> Thus, all we get this wrong for are probably VAX floats or something
> similar, no intent from me to look at that, that is preexisting issue.
>
> BTW, I'm surprised we don't have BUILT_IN_EXP10F{16,32,64,128,32X,64X,128X}
> builtins, seems glibc has those (sure, I think except *16 and *128x).
>
> Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?
OK.
Thanks,
Richard.
> 2024-02-21 Jakub Jelinek <jakub@redhat.com>
>
> PR tree-optimization/113993
> * tree-call-cdce.cc (get_no_error_domain): Handle
> BUILT_IN_{COSH,SINH,EXP{,M1,2}}{F32X,F64X}. Handle
> BUILT_IN_{COSH,SINH,EXP{,M1,2}}L for
> REAL_MODE_FORMAT (TYPE_MODE (long_double_type_node))->emax == 16384
> the as the F128 suffixed cases, otherwise as non-suffixed ones.
> Handle BUILT_IN_{EXP,POW}10L for
> REAL_MODE_FORMAT (TYPE_MODE (long_double_type_node))->emax == 16384
> as (-inf, 4932).
>
> * gcc.dg/tree-ssa/pr113993.c: New test.
>
> --- gcc/tree-call-cdce.cc.jj 2024-01-03 11:51:37.654646209 +0100
> +++ gcc/tree-call-cdce.cc 2024-02-20 09:19:24.432837856 +0100
> @@ -677,14 +677,14 @@ gen_conditions_for_pow (gcall *pow_call,
> Since IEEE only sets minimum requirements for long double format,
> different long double formats exist under different implementations
> (e.g, 64 bit double precision (DF), 80 bit double-extended
> - precision (XF), and 128 bit quad precision (QF) ). For simplicity,
> + precision (XF), and 128 bit quad precision (TF) ). For simplicity,
> in this implementation, the computed bounds for long double assume
> - 64 bit format (DF), and are therefore conservative. Another
> - assumption is that single precision float type is always SF mode,
> - and double type is DF mode. This function is quite
> - implementation specific, so it may not be suitable to be part of
> - builtins.cc. This needs to be revisited later to see if it can
> - be leveraged in x87 assembly expansion. */
> + 64 bit format (DF) except when it is IEEE quad or extended with the same
> + emax, and are therefore sometimes conservative. Another assumption is
> + that single precision float type is always SF mode, and double type is DF
> + mode. This function is quite implementation specific, so it may not be
> + suitable to be part of builtins.cc. This needs to be revisited later
> + to see if it can be leveraged in x87 assembly expansion. */
>
> static inp_domain
> get_no_error_domain (enum built_in_function fnc)
> @@ -723,10 +723,10 @@ get_no_error_domain (enum built_in_funct
> 89, true, false);
> case BUILT_IN_COSH:
> case BUILT_IN_SINH:
> - case BUILT_IN_COSHL:
> - case BUILT_IN_SINHL:
> case BUILT_IN_COSHF64:
> case BUILT_IN_SINHF64:
> + case BUILT_IN_COSHF32X:
> + case BUILT_IN_SINHF32X:
> /* cosh: (-710, +710) */
> return get_domain (-710, true, false,
> 710, true, false);
> @@ -735,6 +735,16 @@ get_no_error_domain (enum built_in_funct
> /* coshf128: (-11357, +11357) */
> return get_domain (-11357, true, false,
> 11357, true, false);
> + case BUILT_IN_COSHL:
> + case BUILT_IN_SINHL:
> + if (REAL_MODE_FORMAT (TYPE_MODE (long_double_type_node))->emax == 16384)
> + return get_no_error_domain (BUILT_IN_COSHF128);
> + return get_no_error_domain (BUILT_IN_COSH);
> + case BUILT_IN_COSHF64X:
> + case BUILT_IN_SINHF64X:
> + if (REAL_MODE_FORMAT (TYPE_MODE (float64x_type_node))->emax == 16384)
> + return get_no_error_domain (BUILT_IN_COSHF128);
> + return get_no_error_domain (BUILT_IN_COSH);
> /* Log functions: (0, +inf) */
> CASE_FLT_FN (BUILT_IN_LOG):
> CASE_FLT_FN_FLOATN_NX (BUILT_IN_LOG):
> @@ -751,7 +761,7 @@ get_no_error_domain (enum built_in_funct
> /* Exp functions. */
> case BUILT_IN_EXPF16:
> case BUILT_IN_EXPM1F16:
> - /* expf: (-inf, 11) */
> + /* expf16: (-inf, 11) */
> return get_domain (-1, false, false,
> 11, true, false);
> case BUILT_IN_EXPF:
> @@ -763,10 +773,10 @@ get_no_error_domain (enum built_in_funct
> 88, true, false);
> case BUILT_IN_EXP:
> case BUILT_IN_EXPM1:
> - case BUILT_IN_EXPL:
> - case BUILT_IN_EXPM1L:
> case BUILT_IN_EXPF64:
> case BUILT_IN_EXPM1F64:
> + case BUILT_IN_EXPF32X:
> + case BUILT_IN_EXPM1F32X:
> /* exp: (-inf, 709) */
> return get_domain (-1, false, false,
> 709, true, false);
> @@ -775,6 +785,16 @@ get_no_error_domain (enum built_in_funct
> /* expf128: (-inf, 11356) */
> return get_domain (-1, false, false,
> 11356, true, false);
> + case BUILT_IN_EXPL:
> + case BUILT_IN_EXPM1L:
> + if (REAL_MODE_FORMAT (TYPE_MODE (long_double_type_node))->emax == 16384)
> + return get_no_error_domain (BUILT_IN_EXPF128);
> + return get_no_error_domain (BUILT_IN_EXP);
> + case BUILT_IN_EXPF64X:
> + case BUILT_IN_EXPM1F64X:
> + if (REAL_MODE_FORMAT (TYPE_MODE (float64x_type_node))->emax == 16384)
> + return get_no_error_domain (BUILT_IN_EXPF128);
> + return get_no_error_domain (BUILT_IN_EXP);
> case BUILT_IN_EXP2F16:
> /* exp2f16: (-inf, 16) */
> return get_domain (-1, false, false,
> @@ -785,8 +805,8 @@ get_no_error_domain (enum built_in_funct
> return get_domain (-1, false, false,
> 128, true, false);
> case BUILT_IN_EXP2:
> - case BUILT_IN_EXP2L:
> case BUILT_IN_EXP2F64:
> + case BUILT_IN_EXP2F32X:
> /* exp2: (-inf, 1024) */
> return get_domain (-1, false, false,
> 1024, true, false);
> @@ -794,6 +814,14 @@ get_no_error_domain (enum built_in_funct
> /* exp2f128: (-inf, 16384) */
> return get_domain (-1, false, false,
> 16384, true, false);
> + case BUILT_IN_EXP2L:
> + if (REAL_MODE_FORMAT (TYPE_MODE (long_double_type_node))->emax == 16384)
> + return get_no_error_domain (BUILT_IN_EXP2F128);
> + return get_no_error_domain (BUILT_IN_EXP2);
> + case BUILT_IN_EXP2F64X:
> + if (REAL_MODE_FORMAT (TYPE_MODE (float64x_type_node))->emax == 16384)
> + return get_no_error_domain (BUILT_IN_EXP2F128);
> + return get_no_error_domain (BUILT_IN_EXP2);
> case BUILT_IN_EXP10F:
> case BUILT_IN_POW10F:
> /* exp10f: (-inf, 38) */
> @@ -801,11 +829,16 @@ get_no_error_domain (enum built_in_funct
> 38, true, false);
> case BUILT_IN_EXP10:
> case BUILT_IN_POW10:
> - case BUILT_IN_EXP10L:
> - case BUILT_IN_POW10L:
> /* exp10: (-inf, 308) */
> return get_domain (-1, false, false,
> 308, true, false);
> + case BUILT_IN_EXP10L:
> + case BUILT_IN_POW10L:
> + if (REAL_MODE_FORMAT (TYPE_MODE (long_double_type_node))->emax == 16384)
> + /* exp10l: (-inf, 4932) */
> + return get_domain (-1, false, false,
> + 4932, true, false);
> + return get_no_error_domain (BUILT_IN_EXP10);
> /* sqrt: [0, +inf) */
> CASE_FLT_FN (BUILT_IN_SQRT):
> CASE_FLT_FN_FLOATN_NX (BUILT_IN_SQRT):
> --- gcc/testsuite/gcc.dg/tree-ssa/pr113993.c.jj 2024-02-20 09:51:59.755613591 +0100
> +++ gcc/testsuite/gcc.dg/tree-ssa/pr113993.c 2024-02-20 09:52:28.815210185 +0100
> @@ -0,0 +1,299 @@
> +/* PR tree-optimization/113993 */
> +/* { dg-do compile } */
> +/* { dg-options "-O2 -fdump-tree-optimized" } */
> +/* { dg-add-options float32 } */
> +/* { dg-add-options float64 } */
> +/* { dg-add-options float128 } */
> +/* { dg-add-options float32x } */
> +/* { dg-add-options float64x } */
> +/* { dg-final { scan-tree-dump-not "__builtin_\[a-z0-9\] \\\(\[^\n\r\]\\\);" "optimized" } } */
> +
> +void
> +flt (float f1, float f2, float f3, float f4, float f5,
> + float f6, float f7, float f8, float f9, float f10)
> +{
> + if (!(f1 >= -1.0f && f1 <= 1.0f)) __builtin_unreachable ();
> + __builtin_acosf (f1);
> + __builtin_asinf (f1);
> + if (!(f2 >= 1.0f && f2 <= __builtin_inff ())) __builtin_unreachable ();
> + __builtin_acoshf (f2);
> + if (!(f3 > -1.0f && f3 < 1.0f)) __builtin_unreachable ();
> + __builtin_atanhf (f3);
> + if (!(f4 > 0.0f && f4 < __builtin_inff ())) __builtin_unreachable ();
> + __builtin_logf (f4);
> + __builtin_log2f (f4);
> + __builtin_log10f (f4);
> + if (!(f5 > -1.0f && f5 < __builtin_inff ())) __builtin_unreachable ();
> + __builtin_log1pf (f5);
> + if (!(f6 >= 0.0f && f6 < __builtin_inff ())) __builtin_unreachable ();
> + __builtin_sqrtf (f6);
> +#if __FLT_MANT_DIG__ == __FLT32_MANT_DIG__ && __FLT_MAX_EXP__ == __FLT32_MAX_EXP__
> + if (!(f7 > -89.0f && f7 < 89.0f)) __builtin_unreachable ();
> + __builtin_coshf (f7);
> + __builtin_sinhf (f7);
> + if (!(f8 > -__builtin_inff () && f8 < 88.0f)) __builtin_unreachable ();
> + __builtin_expf (f8);
> + if (!(f9 > -__builtin_inff () && f9 < 128.0f)) __builtin_unreachable ();
> + __builtin_exp2f (f9);
> + if (!(f10 > -__builtin_inff () && f10 < 38.0f)) __builtin_unreachable ();
> + __builtin_exp10f (f10);
> +#endif
> +}
> +
> +#if defined(__FLT16_MANT_DIG__) && 0 /* No library routines here, these don't actually fold away. */
> +void
> +flt16 (_Float16 f1, _Float16 f2, _Float16 f3, _Float16 f4, _Float16 f5,
> + _Float16 f6, _Float16 f7, _Float16 f8, _Float16 f9)
> +{
> + if (!(f1 >= -1.0f16 && f1 <= 1.0f16)) __builtin_unreachable ();
> + __builtin_acosf16 (f1);
> + __builtin_asinf16 (f1);
> + if (!(f2 >= 1.0f16 && f2 <= __builtin_inff16 ())) __builtin_unreachable ();
> + __builtin_acoshf16 (f2);
> + if (!(f3 > -1.0f16 && f3 < 1.0f16)) __builtin_unreachable ();
> + __builtin_atanhf16 (f3);
> + if (!(f4 > 0.0f16 && f4 < __builtin_inff16 ())) __builtin_unreachable ();
> + __builtin_logf16 (f4);
> + __builtin_log2f16 (f4);
> + __builtin_log10f16 (f4);
> + if (!(f5 > -1.0f16 && f5 < __builtin_inff16 ())) __builtin_unreachable ();
> + __builtin_log1pf16 (f5);
> + if (!(f6 >= 0.0f16 && f6 < __builtin_inff16 ())) __builtin_unreachable ();
> + __builtin_sqrtf16 (f6);
> + if (!(f7 > -11.0f16 && f7 < 11.0f16)) __builtin_unreachable ();
> + __builtin_coshf16 (f7);
> + __builtin_sinhf16 (f7);
> + if (!(f8 > -__builtin_inff16 () && f8 < 11.0f16)) __builtin_unreachable ();
> + __builtin_expf16 (f8);
> + if (!(f9 > -__builtin_inff16 () && f9 < 16.0f16)) __builtin_unreachable ();
> + __builtin_exp2f16 (f9);
> +}
> +#endif
> +
> +#ifdef __FLT32_MANT_DIG__
> +void
> +flt32 (_Float32 f1, _Float32 f2, _Float32 f3, _Float32 f4, _Float32 f5,
> + _Float32 f6, _Float32 f7, _Float32 f8, _Float32 f9)
> +{
> + if (!(f1 >= -1.0f32 && f1 <= 1.0f32)) __builtin_unreachable ();
> + __builtin_acosf32 (f1);
> + __builtin_asinf32 (f1);
> + if (!(f2 >= 1.0f32 && f2 <= __builtin_inff32 ())) __builtin_unreachable ();
> + __builtin_acoshf32 (f2);
> + if (!(f3 > -1.0f32 && f3 < 1.0f32)) __builtin_unreachable ();
> + __builtin_atanhf32 (f3);
> + if (!(f4 > 0.0f32 && f4 < __builtin_inff32 ())) __builtin_unreachable ();
> + __builtin_logf32 (f4);
> + __builtin_log2f32 (f4);
> + __builtin_log10f32 (f4);
> + if (!(f5 > -1.0f32 && f5 < __builtin_inff32 ())) __builtin_unreachable ();
> + __builtin_log1pf32 (f5);
> + if (!(f6 >= 0.0f32 && f6 < __builtin_inff32 ())) __builtin_unreachable ();
> + __builtin_sqrtf32 (f6);
> + if (!(f7 > -89.0f32 && f7 < 89.0f32)) __builtin_unreachable ();
> + __builtin_coshf32 (f7);
> + __builtin_sinhf32 (f7);
> + if (!(f8 > -__builtin_inff32 () && f8 < 88.0f32)) __builtin_unreachable ();
> + __builtin_expf32 (f8);
> + if (!(f9 > -__builtin_inff32 () && f9 < 128.0f32)) __builtin_unreachable ();
> + __builtin_exp2f32 (f9);
> +}
> +#endif
> +
> +void
> +dbl (double f1, double f2, double f3, double f4, double f5,
> + double f6, double f7, double f8, double f9, double f10)
> +{
> + if (!(f1 >= -1.0 && f1 <= 1.0)) __builtin_unreachable ();
> + __builtin_acos (f1);
> + __builtin_asin (f1);
> + if (!(f2 >= 1.0 && f2 <= __builtin_inf ())) __builtin_unreachable ();
> + __builtin_acosh (f2);
> + if (!(f3 > -1.0 && f3 < 1.0)) __builtin_unreachable ();
> + __builtin_atanh (f3);
> + if (!(f4 > 0.0 && f4 < __builtin_inf ())) __builtin_unreachable ();
> + __builtin_log (f4);
> + __builtin_log2 (f4);
> + __builtin_log10 (f4);
> + if (!(f5 > -1.0 && f5 < __builtin_inf ())) __builtin_unreachable ();
> + __builtin_log1p (f5);
> + if (!(f6 >= 0.0 && f6 < __builtin_inf ())) __builtin_unreachable ();
> + __builtin_sqrt (f6);
> +#if __DBL_MANT_DIG__ == __FLT64_MANT_DIG__ && __DBL_MAX_EXP__ == __FLT64_MAX_EXP__
> + if (!(f7 > -710.0 && f7 < 710.0)) __builtin_unreachable ();
> + __builtin_cosh (f7);
> + __builtin_sinh (f7);
> + if (!(f8 > -__builtin_inf () && f8 < 709.0)) __builtin_unreachable ();
> + __builtin_exp (f8);
> + if (!(f9 > -__builtin_inf () && f9 < 1024.0)) __builtin_unreachable ();
> + __builtin_exp2 (f9);
> + if (!(f10 > -__builtin_inf () && f10 < 308.0)) __builtin_unreachable ();
> + __builtin_exp10 (f10);
> +#endif
> +}
> +
> +#ifdef __FLT64_MANT_DIG__
> +void
> +flt64 (_Float64 f1, _Float64 f2, _Float64 f3, _Float64 f4, _Float64 f5,
> + _Float64 f6, _Float64 f7, _Float64 f8, _Float64 f9)
> +{
> + if (!(f1 >= -1.0f64 && f1 <= 1.0f64)) __builtin_unreachable ();
> + __builtin_acosf64 (f1);
> + __builtin_asinf64 (f1);
> + if (!(f2 >= 1.0f64 && f2 <= __builtin_inff64 ())) __builtin_unreachable ();
> + __builtin_acoshf64 (f2);
> + if (!(f3 > -1.0f64 && f3 < 1.0f64)) __builtin_unreachable ();
> + __builtin_atanhf64 (f3);
> + if (!(f4 > 0.0f64 && f4 < __builtin_inff64 ())) __builtin_unreachable ();
> + __builtin_logf64 (f4);
> + __builtin_log2f64 (f4);
> + __builtin_log10f64 (f4);
> + if (!(f5 > -1.0f64 && f5 < __builtin_inff64 ())) __builtin_unreachable ();
> + __builtin_log1pf64 (f5);
> + if (!(f6 >= 0.0f64 && f6 < __builtin_inff64 ())) __builtin_unreachable ();
> + __builtin_sqrtf64 (f6);
> + if (!(f7 > -710.0f64 && f7 < 710.0f64)) __builtin_unreachable ();
> + __builtin_coshf64 (f7);
> + __builtin_sinhf64 (f7);
> + if (!(f8 > -__builtin_inff64 () && f8 < 709.0f64)) __builtin_unreachable ();
> + __builtin_expf64 (f8);
> + if (!(f9 > -__builtin_inff64 () && f9 < 1024.0f64)) __builtin_unreachable ();
> + __builtin_exp2f64 (f9);
> +}
> +#endif
> +
> +#ifdef __FLT32X_MANT_DIG__
> +void
> +flt32x (_Float32x f1, _Float32x f2, _Float32x f3, _Float32x f4, _Float32x f5,
> + _Float32x f6, _Float32x f7, _Float32x f8, _Float32x f9)
> +{
> + if (!(f1 >= -1.0f32x && f1 <= 1.0f32x)) __builtin_unreachable ();
> + __builtin_acosf32x (f1);
> + __builtin_asinf32x (f1);
> + if (!(f2 >= 1.0f32x && f2 <= __builtin_inff32x ())) __builtin_unreachable ();
> + __builtin_acoshf32x (f2);
> + if (!(f3 > -1.0f32x && f3 < 1.0f32x)) __builtin_unreachable ();
> + __builtin_atanhf32x (f3);
> + if (!(f4 > 0.0f32x && f4 < __builtin_inff32x ())) __builtin_unreachable ();
> + __builtin_logf32x (f4);
> + __builtin_log2f32x (f4);
> + __builtin_log10f32x (f4);
> + if (!(f5 > -1.0f32x && f5 < __builtin_inff32x ())) __builtin_unreachable ();
> + __builtin_log1pf32x (f5);
> + if (!(f6 >= 0.0f32x && f6 < __builtin_inff32x ())) __builtin_unreachable ();
> + __builtin_sqrtf32x (f6);
> +#if __FLT32X_MANT_DIG__ == __FLT64_MANT_DIG__ && __FLT32X_MAX_EXP__ == __FLT64_MAX_EXP__
> + if (!(f7 > -710.0f32x && f7 < 710.0f32x)) __builtin_unreachable ();
> + __builtin_coshf32x (f7);
> + __builtin_sinhf32x (f7);
> + if (!(f8 > -__builtin_inff32x () && f8 < 709.0f32x)) __builtin_unreachable ();
> + __builtin_expf32x (f8);
> + if (!(f9 > -__builtin_inff32x () && f9 < 1024.0f32x)) __builtin_unreachable ();
> + __builtin_exp2f32x (f9);
> +#endif
> +}
> +#endif
> +
> +void
> +ldbl (long double f1, long double f2, long double f3, long double f4, long double f5,
> + long double f6, long double f7, long double f8, long double f9, long double f10)
> +{
> + if (!(f1 >= -1.0L && f1 <= 1.0L)) __builtin_unreachable ();
> + __builtin_acosl (f1);
> + __builtin_asinl (f1);
> + if (!(f2 >= 1.0L && f2 <= __builtin_infl ())) __builtin_unreachable ();
> + __builtin_acoshl (f2);
> + if (!(f3 > -1.0L && f3 < 1.0L)) __builtin_unreachable ();
> + __builtin_atanhl (f3);
> + if (!(f4 > 0.0L && f4 < __builtin_infl ())) __builtin_unreachable ();
> + __builtin_logl (f4);
> + __builtin_log2l (f4);
> + __builtin_log10l (f4);
> + if (!(f5 > -1.0L && f5 < __builtin_infl ())) __builtin_unreachable ();
> + __builtin_log1pl (f5);
> + if (!(f6 >= 0.0L && f6 < __builtin_infl ())) __builtin_unreachable ();
> + __builtin_sqrtl (f6);
> +#if __LDBL_MAX_EXP__ == 16384
> + if (!(f7 > -11357.0L && f7 < 11357.0L)) __builtin_unreachable ();
> + __builtin_coshl (f7);
> + __builtin_sinhl (f7);
> + if (!(f8 > -__builtin_infl () && f8 < 11356.0L)) __builtin_unreachable ();
> + __builtin_expl (f8);
> + if (!(f9 > -__builtin_infl () && f9 < 16384.0L)) __builtin_unreachable ();
> + __builtin_exp2l (f9);
> + if (!(f10 > -__builtin_infl () && f10 < 4932.0L)) __builtin_unreachable ();
> + __builtin_exp10l (f10);
> +#elif __LDBL_MANT_DIG__ == __FLT64_MANT_DIG__ && __LDBL_MAX_EXP__ == __FLT64_MAX_EXP__
> + if (!(f7 > -710.0L && f7 < 710.0L)) __builtin_unreachable ();
> + __builtin_coshl (f7);
> + __builtin_sinhl (f7);
> + if (!(f8 > -__builtin_infl () && f8 < 709.0L)) __builtin_unreachable ();
> + __builtin_expl (f8);
> + if (!(f9 > -__builtin_infl () && f9 < 1024.0L)) __builtin_unreachable ();
> + __builtin_exp2l (f9);
> + if (!(f10 > -__builtin_infl () && f10 < 308.0L)) __builtin_unreachable ();
> + __builtin_exp10l (f10);
> +#endif
> +}
> +
> +#ifdef __FLT128_MANT_DIG__
> +void
> +flt128 (_Float128 f1, _Float128 f2, _Float128 f3, _Float128 f4, _Float128 f5,
> + _Float128 f6, _Float128 f7, _Float128 f8, _Float128 f9)
> +{
> + if (!(f1 >= -1.0f128 && f1 <= 1.0f128)) __builtin_unreachable ();
> + __builtin_acosf128 (f1);
> + __builtin_asinf128 (f1);
> + if (!(f2 >= 1.0f128 && f2 <= __builtin_inff128 ())) __builtin_unreachable ();
> + __builtin_acoshf128 (f2);
> + if (!(f3 > -1.0f128 && f3 < 1.0f128)) __builtin_unreachable ();
> + __builtin_atanhf128 (f3);
> + if (!(f4 > 0.0f128 && f4 < __builtin_inff128 ())) __builtin_unreachable ();
> + __builtin_logf128 (f4);
> + __builtin_log2f128 (f4);
> + __builtin_log10f128 (f4);
> + if (!(f5 > -1.0f128 && f5 < __builtin_inff128 ())) __builtin_unreachable ();
> + __builtin_log1pf128 (f5);
> + if (!(f6 >= 0.0f128 && f6 < __builtin_inff128 ())) __builtin_unreachable ();
> + __builtin_sqrtf128 (f6);
> + if (!(f7 > -11357.0f128 && f7 < 11357.0f128)) __builtin_unreachable ();
> + __builtin_coshf128 (f7);
> + __builtin_sinhf128 (f7);
> + if (!(f8 > -__builtin_inff128 () && f8 < 11356.0f128)) __builtin_unreachable ();
> + __builtin_expf128 (f8);
> + if (!(f9 > -__builtin_inff128 () && f9 < 16384.0f128)) __builtin_unreachable ();
> + __builtin_exp2f128 (f9);
> +}
> +#endif
> +
> +#ifdef __FLT64X_MANT_DIG__
> +void
> +flt64x (_Float64x f1, _Float64x f2, _Float64x f3, _Float64x f4, _Float64x f5,
> + _Float64x f6, _Float64x f7, _Float64x f8, _Float64x f9)
> +{
> + if (!(f1 >= -1.0f64x && f1 <= 1.0f64x)) __builtin_unreachable ();
> + __builtin_acosf64x (f1);
> + __builtin_asinf64x (f1);
> + if (!(f2 >= 1.0f64x && f2 <= __builtin_inff64x ())) __builtin_unreachable ();
> + __builtin_acoshf64x (f2);
> + if (!(f3 > -1.0f64x && f3 < 1.0f64x)) __builtin_unreachable ();
> + __builtin_atanhf64x (f3);
> + if (!(f4 > 0.0f64x && f4 < __builtin_inff64x ())) __builtin_unreachable ();
> + __builtin_logf64x (f4);
> + __builtin_log2f64x (f4);
> + __builtin_log10f64x (f4);
> + if (!(f5 > -1.0f64x && f5 < __builtin_inff64x ())) __builtin_unreachable ();
> + __builtin_log1pf64x (f5);
> + if (!(f6 >= 0.0f64x && f6 < __builtin_inff64x ())) __builtin_unreachable ();
> + __builtin_sqrtf64x (f6);
> +#if __FLT64X_MAX_EXP__ == 16384
> + if (!(f7 > -11357.0f64x && f7 < 11357.0f64x)) __builtin_unreachable ();
> + __builtin_coshf64x (f7);
> + __builtin_sinhf64x (f7);
> + if (!(f8 > -__builtin_inff64x () && f8 < 11356.0f64x)) __builtin_unreachable ();
> + __builtin_expf64x (f8);
> + if (!(f9 > -__builtin_inff64x () && f9 < 16384.0f64x)) __builtin_unreachable ();
> + __builtin_exp2f64x (f9);
> +#endif
> +}
> +#endif
>
> Jakub
>
>
--
Richard Biener <rguenther@suse.de>
SUSE Software Solutions Germany GmbH,
Frankenstrasse 146, 90461 Nuernberg, Germany;
GF: Ivo Totev, Andrew McDonald, Werner Knoblich; (HRB 36809, AG Nuernberg)
prev parent reply other threads:[~2024-02-22 9:14 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-02-21 8:39 Jakub Jelinek
2024-02-22 9:14 ` Richard Biener [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=p650s89o-8564-07r0-4r5p-0qs88o697o57@fhfr.qr \
--to=rguenther@suse.de \
--cc=gcc-patches@gcc.gnu.org \
--cc=jakub@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).