[PATCH 1/1] LoongArch: Add soft float support.

caiyinyu caiyinyu@loongson.cn
Wed Aug 31 12:00:43 GMT 2022


diff --git a/sysdeps/loongarch/Makefile b/sysdeps/loongarch/Makefile
index 41c3449670..746d4c2c8f 100644
--- a/sysdeps/loongarch/Makefile
+++ b/sysdeps/loongarch/Makefile
@@ -5,11 +5,3 @@ endif
  # LoongArch's assembler also needs to know about PIC as it changes the
  # definition of some assembler macros.
  ASFLAGS-.os += $(pic-ccflag)
-
-abi-variants := lp64
-
-ifeq (,$(filter $(default-abi),$(abi-variants)))
-$(error Unknown ABI $(default-abi), must be one of $(abi-variants))
-endif
-
-abi-lp64-condition    := defined __loongarch_lp64

Why move this to be Linux specific? It won't matter in the end though.

The rest looks ok.


I did this to follow the example of other architectures (like riscv arc

aarch64 arm mips...) though there seems to be no difference.

I need help to include this patch since there are some changes in common

file build-many-glibcs.py.



在 2022/8/30 下午9:00, Adhemerval Zanella Netto 写道:
>
> On 30/08/22 08:02, caiyinyu wrote:
>> ---
>>   scripts/build-many-glibcs.py                  |  4 ++
>>   sysdeps/loongarch/Makefile                    |  8 ---
>>   sysdeps/loongarch/__longjmp.S                 |  2 +
>>   sysdeps/loongarch/bits/setjmp.h               |  3 ++
>>   sysdeps/loongarch/dl-trampoline.S             | 11 +++-
>>   sysdeps/loongarch/fpu_control.h               | 13 +++++
>>   sysdeps/loongarch/nofpu/Implies               |  1 +
>>   .../loongarch/nofpu/math-tests-exceptions.h   | 28 ++++++++++
>>   sysdeps/loongarch/nofpu/math-tests-rounding.h | 27 ++++++++++
>>   sysdeps/loongarch/preconfigure                |  1 -
>>   sysdeps/loongarch/preconfigure.ac             |  1 -
>>   sysdeps/loongarch/setjmp.S                    |  2 +
>>   sysdeps/unix/sysv/linux/loongarch/Makefile    |  9 ++++
>>   sysdeps/unix/sysv/linux/loongarch/configure   | 51 +++++++++++++++++--
>>   .../unix/sysv/linux/loongarch/configure.ac    | 22 ++++++--
>>   .../unix/sysv/linux/loongarch/ldd-rewrite.sed |  2 +-
>>   .../unix/sysv/linux/loongarch/shlib-versions  |  2 +
>>   17 files changed, 168 insertions(+), 19 deletions(-)
>>   create mode 100644 sysdeps/loongarch/nofpu/Implies
>>   create mode 100644 sysdeps/loongarch/nofpu/math-tests-exceptions.h
>>   create mode 100644 sysdeps/loongarch/nofpu/math-tests-rounding.h
>>
>> diff --git a/scripts/build-many-glibcs.py b/scripts/build-many-glibcs.py
>> index 98f02104a2..db632b18a3 100755
>> --- a/scripts/build-many-glibcs.py
>> +++ b/scripts/build-many-glibcs.py
>> @@ -222,6 +222,10 @@ class Context(object):
>>                           os_name='linux-gnu',
>>                           variant='lp64d',
>>                           gcc_cfg=['--with-abi=lp64d','--disable-multilib'])
>> +        self.add_config(arch='loongarch64',
>> +                        os_name='linux-gnu',
>> +                        variant='lp64s',
>> +                        gcc_cfg=['--with-abi=lp64s','--disable-multilib'])
>>           self.add_config(arch='m68k',
>>                           os_name='linux-gnu',
>>                           gcc_cfg=['--disable-multilib'])
>> diff --git a/sysdeps/loongarch/Makefile b/sysdeps/loongarch/Makefile
>> index 41c3449670..746d4c2c8f 100644
>> --- a/sysdeps/loongarch/Makefile
>> +++ b/sysdeps/loongarch/Makefile
>> @@ -5,11 +5,3 @@ endif
>>   # LoongArch's assembler also needs to know about PIC as it changes the
>>   # definition of some assembler macros.
>>   ASFLAGS-.os += $(pic-ccflag)
>> -
>> -abi-variants := lp64
>> -
>> -ifeq (,$(filter $(default-abi),$(abi-variants)))
>> -$(error Unknown ABI $(default-abi), must be one of $(abi-variants))
>> -endif
>> -
>> -abi-lp64-condition    := defined __loongarch_lp64
> Why move this to be Linux specific? It won't matter in the end though.
>
> The rest looks ok.
>
>> diff --git a/sysdeps/loongarch/__longjmp.S b/sysdeps/loongarch/__longjmp.S
>> index 4207376f5e..d6a99fcbc8 100644
>> --- a/sysdeps/loongarch/__longjmp.S
>> +++ b/sysdeps/loongarch/__longjmp.S
>> @@ -41,6 +41,7 @@ ENTRY (__longjmp)
>>   	REG_L s7, a0, 11*SZREG
>>   	REG_L s8, a0, 12*SZREG
>>   
>> +#ifndef __loongarch_soft_float
>>   	FREG_L $f24, a0, 13*SZREG + 0*SZFREG
>>   	FREG_L $f25, a0, 13*SZREG + 1*SZFREG
>>   	FREG_L $f26, a0, 13*SZREG + 2*SZFREG
>> @@ -49,6 +50,7 @@ ENTRY (__longjmp)
>>   	FREG_L $f29, a0, 13*SZREG + 5*SZFREG
>>   	FREG_L $f30, a0, 13*SZREG + 6*SZFREG
>>   	FREG_L $f31, a0, 13*SZREG + 7*SZFREG
>> +#endif
>>   
>>   	sltui	a0,a1,1
>>   	ADD	a0, a0, a1	 # a0 = (a1 == 0) ? 1 : a1
>> diff --git a/sysdeps/loongarch/bits/setjmp.h b/sysdeps/loongarch/bits/setjmp.h
>> index 42f8fa7657..8b323ad2b6 100644
>> --- a/sysdeps/loongarch/bits/setjmp.h
>> +++ b/sysdeps/loongarch/bits/setjmp.h
>> @@ -31,8 +31,11 @@ typedef struct __jmp_buf_internal_tag
>>     long int __fp;
>>     /* Callee-saved registers.  */
>>     long int __regs[9];
>> +
>> +#ifndef __loongarch_soft_float
>>     /* Callee-saved floating point registers.  */
>>     double __fpregs[8];
>> +#endif
>>   
>>   } __jmp_buf[1];
>>   
>> diff --git a/sysdeps/loongarch/dl-trampoline.S b/sysdeps/loongarch/dl-trampoline.S
>> index ad8ab0fda7..9a8f580971 100644
>> --- a/sysdeps/loongarch/dl-trampoline.S
>> +++ b/sysdeps/loongarch/dl-trampoline.S
>> @@ -21,8 +21,11 @@
>>   
>>   /* Assembler veneer called from the PLT header code for lazy loading.
>>      The PLT header passes its own args in t0-t2.  */
>> -
>> -# define FRAME_SIZE (-((-10 * SZREG - 8 * SZFREG) & ALMASK))
>> +#ifdef __loongarch_soft_float
>> +#define FRAME_SIZE (-((-10 * SZREG) & ALMASK))
>> +#else
>> +#define FRAME_SIZE (-((-10 * SZREG - 8 * SZFREG) & ALMASK))
>> +#endif
>>   
>>   ENTRY (_dl_runtime_resolve)
>>   
>> @@ -39,6 +42,7 @@ ENTRY (_dl_runtime_resolve)
>>   	REG_S	a6, sp, 7*SZREG
>>   	REG_S	a7, sp, 8*SZREG
>>   
>> +#ifndef __loongarch_soft_float
>>   	FREG_S	fa0, sp, 10*SZREG + 0*SZFREG
>>   	FREG_S	fa1, sp, 10*SZREG + 1*SZFREG
>>   	FREG_S	fa2, sp, 10*SZREG + 2*SZFREG
>> @@ -47,6 +51,7 @@ ENTRY (_dl_runtime_resolve)
>>   	FREG_S	fa5, sp, 10*SZREG + 5*SZFREG
>>   	FREG_S	fa6, sp, 10*SZREG + 6*SZFREG
>>   	FREG_S	fa7, sp, 10*SZREG + 7*SZFREG
>> +#endif
>>   
>>   	/* Update .got.plt and obtain runtime address of callee */
>>   	SLLI	a1, t1, 1
>> @@ -67,6 +72,7 @@ ENTRY (_dl_runtime_resolve)
>>   	REG_L	a6, sp, 7*SZREG
>>   	REG_L	a7, sp, 8*SZREG
>>   
>> +#ifndef __loongarch_soft_float
>>   	FREG_L	fa0, sp, 10*SZREG + 0*SZFREG
>>   	FREG_L	fa1, sp, 10*SZREG + 1*SZFREG
>>   	FREG_L	fa2, sp, 10*SZREG + 2*SZFREG
>> @@ -75,6 +81,7 @@ ENTRY (_dl_runtime_resolve)
>>   	FREG_L	fa5, sp, 10*SZREG + 5*SZFREG
>>   	FREG_L	fa6, sp, 10*SZREG + 6*SZFREG
>>   	FREG_L	fa7, sp, 10*SZREG + 7*SZFREG
>> +#endif
>>   
>>   	ADDI	sp, sp, FRAME_SIZE
>>   
>> diff --git a/sysdeps/loongarch/fpu_control.h b/sysdeps/loongarch/fpu_control.h
>> index 1cccc933a5..f482395dba 100644
>> --- a/sysdeps/loongarch/fpu_control.h
>> +++ b/sysdeps/loongarch/fpu_control.h
>> @@ -51,6 +51,17 @@
>>   
>>   #include <features.h>
>>   
>> +#ifdef __loongarch_soft_float
>> +
>> +#define _FPU_RESERVED 0xffffffff
>> +#define _FPU_DEFAULT 0x00000000
>> +typedef unsigned int fpu_control_t;
>> +#define _FPU_GETCW(cw) (cw) = 0
>> +#define _FPU_SETCW(cw) (void) (cw)
>> +extern fpu_control_t __fpu_control;
>> +
>> +#else /* __loongarch_soft_float */
>> +
>>   /* Masks for interrupts.  */
>>   #define _FPU_MASK_V 0x10 /* Invalid operation */
>>   #define _FPU_MASK_Z 0x08 /* Division by zero  */
>> @@ -86,4 +97,6 @@ extern void __loongarch_fpu_setcw (fpu_control_t) __THROW;
>>   /* Default control word set at startup.  */
>>   extern fpu_control_t __fpu_control;
>>   
>> +#endif /* __loongarch_soft_float */
>> +
>>   #endif /* fpu_control.h */
>> diff --git a/sysdeps/loongarch/nofpu/Implies b/sysdeps/loongarch/nofpu/Implies
>> new file mode 100644
>> index 0000000000..abcbadb25f
>> --- /dev/null
>> +++ b/sysdeps/loongarch/nofpu/Implies
>> @@ -0,0 +1 @@
>> +ieee754/soft-fp
>> diff --git a/sysdeps/loongarch/nofpu/math-tests-exceptions.h b/sysdeps/loongarch/nofpu/math-tests-exceptions.h
>> new file mode 100644
>> index 0000000000..b22bb01d0e
>> --- /dev/null
>> +++ b/sysdeps/loongarch/nofpu/math-tests-exceptions.h
>> @@ -0,0 +1,28 @@
>> +/* Configuration for math tests: support for exceptions.
>> +   Copyright (C) 2022 Free Software Foundation, Inc.
>> +   This file is part of the GNU C Library.
>> +
>> +   The GNU C Library is free software; you can redistribute it and/or
>> +   modify it under the terms of the GNU Lesser General Public
>> +   License as published by the Free Software Foundation; either
>> +   version 2.1 of the License, or (at your option) any later version.
>> +
>> +   The GNU C Library is distributed in the hope that it will be useful,
>> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
>> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
>> +   Lesser General Public License for more details.
>> +
>> +   You should have received a copy of the GNU Lesser General Public
>> +   License along with the GNU C Library; if not, see
>> +   <https://www.gnu.org/licenses/>.  */
>> +
>> +#ifndef LOONGARCH_NOFPU_MATH_TESTS_EXCEPTIONS_H
>> +#define LOONGARCH_NOFPU_MATH_TESTS_EXCEPTIONS_H 1
>> +
>> +/* We support setting floating-point exception flags on hard-float
>> +   targets.  These are not supported on soft-float targets.  */
>> +#define EXCEPTION_TESTS_float 0
>> +#define EXCEPTION_TESTS_double        0
>> +#define EXCEPTION_TESTS_long_double   0
>> +
>> +#endif /* math-tests-exceptions.h.  */
>> diff --git a/sysdeps/loongarch/nofpu/math-tests-rounding.h b/sysdeps/loongarch/nofpu/math-tests-rounding.h
>> new file mode 100644
>> index 0000000000..5322e481b9
>> --- /dev/null
>> +++ b/sysdeps/loongarch/nofpu/math-tests-rounding.h
>> @@ -0,0 +1,27 @@
>> +/* Configuration for math tests: rounding mode support.
>> +   Copyright (C) 2022 Free Software Foundation, Inc.
>> +   This file is part of the GNU C Library.
>> +
>> +   The GNU C Library is free software; you can redistribute it and/or
>> +   modify it under the terms of the GNU Lesser General Public
>> +   License as published by the Free Software Foundation; either
>> +   version 2.1 of the License, or (at your option) any later version.
>> +
>> +   The GNU C Library is distributed in the hope that it will be useful,
>> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
>> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
>> +   Lesser General Public License for more details.
>> +
>> +   You should have received a copy of the GNU Lesser General Public
>> +   License along with the GNU C Library; if not, see
>> +   <https://www.gnu.org/licenses/>.  */
>> +
>> +#ifndef LOONGARCH_NOFPU_MATH_TESTS_ROUNDING_H
>> +#define LOONGARCH_NOFPU_MATH_TESTS_ROUNDING_H 1
>> +
>> +/* On soft-float targets we only support the "to nearest" rounding mode.  */
>> +#define ROUNDING_TESTS_float(MODE)		((MODE) == FE_TONEAREST)
>> +#define ROUNDING_TESTS_double(MODE)		((MODE) == FE_TONEAREST)
>> +#define ROUNDING_TESTS_long_double(MODE)	((MODE) == FE_TONEAREST)
>> +
>> +#endif /* math-tests-rounding.h.  */
>> diff --git a/sysdeps/loongarch/preconfigure b/sysdeps/loongarch/preconfigure
>> index 118963cda6..f2d1a0d87b 100644
>> --- a/sysdeps/loongarch/preconfigure
>> +++ b/sysdeps/loongarch/preconfigure
>> @@ -12,7 +12,6 @@ loongarch*)
>>       case "$float_abi" in
>>       soft)
>>   	abi_flen=0
>> -	as_fn_error 1 "loongarch does not yet support soft floating-point  ABI!!" "$LINENO" 5
>>   	;;
>>       single)
>>   	as_fn_error 1 "glibc does not yet support the single floating-point ABI!!" "$LINENO" 5
>> diff --git a/sysdeps/loongarch/preconfigure.ac b/sysdeps/loongarch/preconfigure.ac
>> index 1aba743c15..67e4357013 100644
>> --- a/sysdeps/loongarch/preconfigure.ac
>> +++ b/sysdeps/loongarch/preconfigure.ac
>> @@ -12,7 +12,6 @@ loongarch*)
>>       case "$float_abi" in
>>       soft)
>>   	abi_flen=0
>> -	AC_MSG_ERROR([loongarch does not yet support soft floating-point  ABI!!], 1)
>>   	;;
>>       single)
>>   	AC_MSG_ERROR([glibc does not yet support the single floating-point ABI!!], 1)
>> diff --git a/sysdeps/loongarch/setjmp.S b/sysdeps/loongarch/setjmp.S
>> index 298bb02a82..9b1cdea48c 100644
>> --- a/sysdeps/loongarch/setjmp.S
>> +++ b/sysdeps/loongarch/setjmp.S
>> @@ -50,6 +50,7 @@ ENTRY (__sigsetjmp)
>>   	REG_S s7, a0, 11*SZREG
>>   	REG_S s8, a0, 12*SZREG
>>   
>> +#ifndef __loongarch_soft_float
>>   	FREG_S $f24, a0, 13*SZREG + 0*SZFREG
>>   	FREG_S $f25, a0, 13*SZREG + 1*SZFREG
>>   	FREG_S $f26, a0, 13*SZREG + 2*SZFREG
>> @@ -58,6 +59,7 @@ ENTRY (__sigsetjmp)
>>   	FREG_S $f29, a0, 13*SZREG + 5*SZFREG
>>   	FREG_S $f30, a0, 13*SZREG + 6*SZFREG
>>   	FREG_S $f31, a0, 13*SZREG + 7*SZFREG
>> +#endif
>>   
>>   #if !IS_IN (libc) && IS_IN(rtld)
>>   	li.w		v0, 0
>> diff --git a/sysdeps/unix/sysv/linux/loongarch/Makefile b/sysdeps/unix/sysv/linux/loongarch/Makefile
>> index 91bd35800a..c84a1762ed 100644
>> --- a/sysdeps/unix/sysv/linux/loongarch/Makefile
>> +++ b/sysdeps/unix/sysv/linux/loongarch/Makefile
>> @@ -1,3 +1,12 @@
>>   ifeq ($(subdir),stdlib)
>>   gen-as-const-headers += ucontext_i.sym
>>   endif
>> +
>> +abi-variants := lp64s lp64d
>> +
>> +ifeq (,$(filter $(default-abi),$(abi-variants)))
>> +$(error Unknown ABI $(default-abi), must be one of $(abi-variants))
>> +endif
>> +
>> +abi-lp64s-condition	:= __WORDSIZE == 64 && defined __loongarch_soft_float
>> +abi-lp64d-condition	:= __WORDSIZE == 64 && defined __loongarch_double_float
>> diff --git a/sysdeps/unix/sysv/linux/loongarch/configure b/sysdeps/unix/sysv/linux/loongarch/configure
>> index 60a410303e..28216c165e 100644
>> --- a/sysdeps/unix/sysv/linux/loongarch/configure
>> +++ b/sysdeps/unix/sysv/linux/loongarch/configure
>> @@ -151,11 +151,56 @@ if test $libc_cv_loongarch_int_abi = no; then
>>     as_fn_error $? "Unable to determine integer ABI" "$LINENO" 5
>>   fi
>>   
>> +libc_cv_loongarch_float_abi=no
>> +cat confdefs.h - <<_ACEOF >conftest.$ac_ext
>> +/* end confdefs.h.  */
>> +#ifdef __loongarch_double_float
>> +		   yes
>> +		   #endif
>> +
>> +_ACEOF
>> +if (eval "$ac_cpp conftest.$ac_ext") 2>&5 |
>> +  $EGREP "yes" >/dev/null 2>&1; then :
>> +  libc_cv_loongarch_float_abi=d
>> +fi
>> +rm -f conftest*
>> +
>> +cat confdefs.h - <<_ACEOF >conftest.$ac_ext
>> +/* end confdefs.h.  */
>> +#ifdef __loongarch_soft_float
>> +		   yes
>> +		   #endif
>> +
>> +_ACEOF
>> +if (eval "$ac_cpp conftest.$ac_ext") 2>&5 |
>> +  $EGREP "yes" >/dev/null 2>&1; then :
>> +  libc_cv_loongarch_float_abi=s
>> +fi
>> +rm -f conftest*
>> +
>> +if test "$libc_cv_loongarch_float_abi" = no; then
>> +  as_fn_error $? "Unable to determine floating-point ABI" "$LINENO" 5
>> +fi
>> +
>>   config_vars="$config_vars
>> -default-abi = $libc_cv_loongarch_int_abi"
>> +default-abi = $libc_cv_loongarch_int_abi$libc_cv_loongarch_float_abi"
>>   
>> -case $libc_cv_loongarch_int_abi in
>> -lp64)
>> +case $libc_cv_loongarch_int_abi$libc_cv_loongarch_float_abi in
>> +lp64s)
>> +  test -n "$libc_cv_slibdir" ||
>> +case "$prefix" in
>> +/usr | /usr/)
>> +  libc_cv_slibdir='/lib64/sf'
>> +  libc_cv_rtlddir='/lib64'
>> +  if test "$libdir" = '${exec_prefix}/lib'; then
>> +    libdir='${exec_prefix}/lib64/sf';
>> +    # Locale data can be shared between 32-bit and 64-bit libraries.
>> +    libc_cv_complocaledir='${exec_prefix}/lib/locale'
>> +  fi
>> +  ;;
>> +esac
>> +  ;;
>> +lp64d)
>>     test -n "$libc_cv_slibdir" ||
>>   case "$prefix" in
>>   /usr | /usr/)
>> diff --git a/sysdeps/unix/sysv/linux/loongarch/configure.ac b/sysdeps/unix/sysv/linux/loongarch/configure.ac
>> index 7de1e95ff6..04e9150a9b 100644
>> --- a/sysdeps/unix/sysv/linux/loongarch/configure.ac
>> +++ b/sysdeps/unix/sysv/linux/loongarch/configure.ac
>> @@ -11,10 +11,26 @@ if test $libc_cv_loongarch_int_abi = no; then
>>     AC_MSG_ERROR([Unable to determine integer ABI])
>>   fi
>>   
>> -LIBC_CONFIG_VAR([default-abi], [$libc_cv_loongarch_int_abi])
>> +libc_cv_loongarch_float_abi=no
>> +AC_EGREP_CPP(yes, [#ifdef __loongarch_double_float
>> +		   yes
>> +		   #endif
>> +  ],libc_cv_loongarch_float_abi=d)
>> +AC_EGREP_CPP(yes, [#ifdef __loongarch_soft_float
>> +		   yes
>> +		   #endif
>> +  ],libc_cv_loongarch_float_abi=s)
>> +if test "$libc_cv_loongarch_float_abi" = no; then
>> +  AC_MSG_ERROR([Unable to determine floating-point ABI])
>> +fi
>> +
>> +LIBC_CONFIG_VAR([default-abi], [$libc_cv_loongarch_int_abi$libc_cv_loongarch_float_abi])
>>   
>> -case $libc_cv_loongarch_int_abi in
>> -lp64)
>> +case $libc_cv_loongarch_int_abi$libc_cv_loongarch_float_abi in
>> +lp64s)
>> +  LIBC_SLIBDIR_RTLDDIR([lib64/sf], [lib64])
>> +  ;;
>> +lp64d)
>>     LIBC_SLIBDIR_RTLDDIR([lib64], [lib64])
>>     ;;
>>   esac
>> diff --git a/sysdeps/unix/sysv/linux/loongarch/ldd-rewrite.sed b/sysdeps/unix/sysv/linux/loongarch/ldd-rewrite.sed
>> index f8976fd239..cdbe5c3dc5 100644
>> --- a/sysdeps/unix/sysv/linux/loongarch/ldd-rewrite.sed
>> +++ b/sysdeps/unix/sysv/linux/loongarch/ldd-rewrite.sed
>> @@ -1 +1 @@
>> -s_^\(RTLDLIST=\)\(.*lib\)\(\|64\|32\)\(/[^/]*\)\(-loongarch\)\(64\|32\)\(\.so\.[0-9.]*\)[[:blank:]]*$_\1"\2\4\7 \264\4-loongarch64\7 \232\4-loongarch32\7"_
>> +s_^\(RTLDLIST=\)\(.*lib\)\(\|64\|32\)\(/ld-linux-loongarch-\)\(lp\|ilp\)\(64\|32\)\(d\|s\)\(\.so\.[0-9.]*\)[[:blank:]]*$_\1"\264\4lp64d\8 \264\4lp64s\8"_
>> diff --git a/sysdeps/unix/sysv/linux/loongarch/shlib-versions b/sysdeps/unix/sysv/linux/loongarch/shlib-versions
>> index dc2220b4be..5f40e7f5ec 100644
>> --- a/sysdeps/unix/sysv/linux/loongarch/shlib-versions
>> +++ b/sysdeps/unix/sysv/linux/loongarch/shlib-versions
>> @@ -2,6 +2,8 @@ DEFAULT		GLIBC_2.36
>>   
>>   %if LOONGARCH_ABI_GRLEN == 64 && LOONGARCH_ABI_FRLEN == 64
>>   ld=ld-linux-loongarch-lp64d.so.1
>> +%elif LOONGARCH_ABI_GRLEN == 64 && LOONGARCH_ABI_FRLEN == 0
>> +ld=ld-linux-loongarch-lp64s.so.1
>>   %else
>>   %error cannot determine ABI
>>   %endif



More information about the Libc-alpha mailing list