* [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions
@ 2024-10-02 15:40 J. Neuschäfer
2024-10-02 15:40 ` [musl] [PATCH MCM 1/3] add gcc 12.4 hash and patches J. Neuschäfer
` (3 more replies)
0 siblings, 4 replies; 9+ messages in thread
From: J. Neuschäfer @ 2024-10-02 15:40 UTC (permalink / raw)
To: musl; +Cc: J. Neuschäfer
This patchset adds newer GCC versions to MCM. Patches are copied from
GCC 11 and only very lightly modifed (which can be verified by diffing
the patch directories).
I haven't tested the resulting toolchains, so there might be some bugs
lurking.
Signed-off-by: J. Neuschäfer <j.neuschaefer@gmx.net>
---
J. Neuschäfer (3):
add gcc 12.4 hash and patches
add gcc 13.3 hash and patches
add gcc 14.2 hash and patches
hashes/gcc-12.4.0.tar.xz.sha1 | 1 +
hashes/gcc-13.3.0.tar.xz.sha1 | 1 +
hashes/gcc-14.2.0.tar.xz.sha1 | 1 +
patches/gcc-12.4.0/0001-ssp_nonshared.diff | 14 +
patches/gcc-12.4.0/0002-posix_memalign.diff | 30 ++
patches/gcc-12.4.0/0003-j2.diff | 346 +++++++++++++++++++++
patches/gcc-12.4.0/0004-static-pie.diff | 92 ++++++
patches/gcc-12.4.0/0005-m68k-sqrt.diff | 20 ++
patches/gcc-12.4.0/0006-cow-libstdc++v3.diff | 8 +
patches/gcc-12.4.0/0007-fdpic-unwind.diff | 11 +
.../gcc-12.4.0/0008-fdpic-crtstuff-pr114158.diff | 38 +++
patches/gcc-12.4.0/0009-sh-fdpic-pr114641.diff | 12 +
patches/gcc-13.3.0/0001-ssp_nonshared.diff | 14 +
patches/gcc-13.3.0/0002-posix_memalign.diff | 30 ++
patches/gcc-13.3.0/0003-j2.diff | 346 +++++++++++++++++++++
patches/gcc-13.3.0/0004-static-pie.diff | 92 ++++++
patches/gcc-13.3.0/0005-m68k-sqrt.diff | 20 ++
patches/gcc-13.3.0/0006-cow-libstdc++v3.diff | 8 +
patches/gcc-13.3.0/0007-fdpic-unwind.diff | 11 +
.../gcc-13.3.0/0008-fdpic-crtstuff-pr114158.diff | 38 +++
patches/gcc-13.3.0/0009-sh-fdpic-pr114641.diff | 12 +
patches/gcc-14.2.0/0001-ssp_nonshared.diff | 14 +
patches/gcc-14.2.0/0002-posix_memalign.diff | 30 ++
patches/gcc-14.2.0/0003-j2.diff | 346 +++++++++++++++++++++
patches/gcc-14.2.0/0004-static-pie.diff | 92 ++++++
patches/gcc-14.2.0/0005-m68k-sqrt.diff | 20 ++
patches/gcc-14.2.0/0006-cow-libstdc++v3.diff | 8 +
patches/gcc-14.2.0/0007-fdpic-unwind.diff | 11 +
.../gcc-14.2.0/0008-fdpic-crtstuff-pr114158.diff | 38 +++
patches/gcc-14.2.0/0009-sh-fdpic-pr114641.diff | 12 +
30 files changed, 1716 insertions(+)
---
base-commit: fd6be58297ee21fcba89216ccd0d4aca1e3f1c5c
change-id: 20241002-gcc-bump-1d99a65a2422
Best regards,
--
J. Neuschäfer <j.neuschaefer@gmx.net>
^ permalink raw reply [flat|nested] 9+ messages in thread* [musl] [PATCH MCM 1/3] add gcc 12.4 hash and patches 2024-10-02 15:40 [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions J. Neuschäfer @ 2024-10-02 15:40 ` J. Neuschäfer 2024-10-02 15:40 ` [musl] [PATCH MCM 2/3] add gcc 13.3 " J. Neuschäfer ` (2 subsequent siblings) 3 siblings, 0 replies; 9+ messages in thread From: J. Neuschäfer @ 2024-10-02 15:40 UTC (permalink / raw) To: musl; +Cc: J. Neuschäfer The patches have been copied from 11.4 and slightly adjusted to account for renamed files (.c -> .cc). --- hashes/gcc-12.4.0.tar.xz.sha1 | 1 + patches/gcc-12.4.0/0001-ssp_nonshared.diff | 14 + patches/gcc-12.4.0/0002-posix_memalign.diff | 30 ++ patches/gcc-12.4.0/0003-j2.diff | 346 +++++++++++++++++++++ patches/gcc-12.4.0/0004-static-pie.diff | 92 ++++++ patches/gcc-12.4.0/0005-m68k-sqrt.diff | 20 ++ patches/gcc-12.4.0/0006-cow-libstdc++v3.diff | 8 + patches/gcc-12.4.0/0007-fdpic-unwind.diff | 11 + .../gcc-12.4.0/0008-fdpic-crtstuff-pr114158.diff | 38 +++ patches/gcc-12.4.0/0009-sh-fdpic-pr114641.diff | 12 + 10 files changed, 572 insertions(+) diff --git a/hashes/gcc-12.4.0.tar.xz.sha1 b/hashes/gcc-12.4.0.tar.xz.sha1 new file mode 100644 index 0000000..0c9d653 --- /dev/null +++ b/hashes/gcc-12.4.0.tar.xz.sha1 @@ -0,0 +1 @@ +b373d4ac29bb06ca64d288621906cbf63ab5a1f5 gcc-12.4.0.tar.xz diff --git a/patches/gcc-12.4.0/0001-ssp_nonshared.diff b/patches/gcc-12.4.0/0001-ssp_nonshared.diff new file mode 100644 index 0000000..b250f15 --- /dev/null +++ b/patches/gcc-12.4.0/0001-ssp_nonshared.diff @@ -0,0 +1,14 @@ +diff --git a/gcc/gcc.cc b/gcc/gcc.cc +index 7837553958b..3c81c5798d8 100644 +--- a/gcc/gcc.cc ++++ b/gcc/gcc.cc +@@ -980,7 +980,8 @@ proper position among the other output files. */ + #ifndef LINK_SSP_SPEC + #ifdef TARGET_LIBC_PROVIDES_SSP + #define LINK_SSP_SPEC "%{fstack-protector|fstack-protector-all" \ +- "|fstack-protector-strong|fstack-protector-explicit:}" ++ "|fstack-protector-strong|fstack-protector-explicit" \ ++ ":-lssp_nonshared}" + #else + #define LINK_SSP_SPEC "%{fstack-protector|fstack-protector-all" \ + "|fstack-protector-strong|fstack-protector-explicit" \ diff --git a/patches/gcc-12.4.0/0002-posix_memalign.diff b/patches/gcc-12.4.0/0002-posix_memalign.diff new file mode 100644 index 0000000..1d80096 --- /dev/null +++ b/patches/gcc-12.4.0/0002-posix_memalign.diff @@ -0,0 +1,30 @@ +diff --git a/gcc/config/i386/pmm_malloc.h b/gcc/config/i386/pmm_malloc.h +index 1b0bfe37852..d7b2b19bb3c 100644 +--- a/gcc/config/i386/pmm_malloc.h ++++ b/gcc/config/i386/pmm_malloc.h +@@ -27,12 +27,13 @@ + #include <stdlib.h> + + /* We can't depend on <stdlib.h> since the prototype of posix_memalign +- may not be visible. */ ++ may not be visible and we can't pollute the namespace either. */ + #ifndef __cplusplus +-extern int posix_memalign (void **, size_t, size_t); ++extern int _mm_posix_memalign (void **, size_t, size_t) + #else +-extern "C" int posix_memalign (void **, size_t, size_t) throw (); ++extern "C" int _mm_posix_memalign (void **, size_t, size_t) throw () + #endif ++__asm__("posix_memalign"); + + static __inline void * + _mm_malloc (size_t __size, size_t __alignment) +@@ -42,7 +43,7 @@ _mm_malloc (size_t __size, size_t __alignment) + return malloc (__size); + if (__alignment == 2 || (sizeof (void *) == 8 && __alignment == 4)) + __alignment = sizeof (void *); +- if (posix_memalign (&__ptr, __alignment, __size) == 0) ++ if (_mm_posix_memalign (&__ptr, __alignment, __size) == 0) + return __ptr; + else + return NULL; diff --git a/patches/gcc-12.4.0/0003-j2.diff b/patches/gcc-12.4.0/0003-j2.diff new file mode 100644 index 0000000..c378645 --- /dev/null +++ b/patches/gcc-12.4.0/0003-j2.diff @@ -0,0 +1,346 @@ +diff --git a/gcc/config.gcc b/gcc/config.gcc +index 357b0bed067..528add999f2 100644 +--- a/gcc/config.gcc ++++ b/gcc/config.gcc +@@ -556,7 +556,7 @@ s390*-*-*) + extra_headers="s390intrin.h htmintrin.h htmxlintrin.h vecintrin.h" + ;; + # Note the 'l'; we need to be able to match e.g. "shle" or "shl". +-sh[123456789lbe]*-*-* | sh-*-*) ++sh[123456789lbej]*-*-* | sh-*-*) + cpu_type=sh + extra_options="${extra_options} fused-madd.opt" + extra_objs="${extra_objs} sh_treg_combine.o sh-mem.o sh_optimize_sett_clrt.o" +@@ -3202,18 +3202,18 @@ s390x-ibm-tpf*) + extra_options="${extra_options} s390/tpf.opt" + tmake_file="${tmake_file} s390/t-s390" + ;; +-sh-*-elf* | sh[12346l]*-*-elf* | \ +- sh-*-linux* | sh[2346lbe]*-*-linux* | \ ++sh-*-elf* | sh[12346lj]*-*-elf* | \ ++ sh-*-linux* | sh[2346lbej]*-*-linux* | \ + sh-*-netbsdelf* | shl*-*-netbsdelf*) + tmake_file="${tmake_file} sh/t-sh sh/t-elf" + if test x${with_endian} = x; then + case ${target} in +- sh[1234]*be-*-* | sh[1234]*eb-*-*) with_endian=big ;; ++ sh[j1234]*be-*-* | sh[j1234]*eb-*-*) with_endian=big ;; + shbe-*-* | sheb-*-*) with_endian=big,little ;; + sh[1234]l* | sh[34]*-*-linux*) with_endian=little ;; + shl* | sh*-*-linux* | \ + sh-superh-elf) with_endian=little,big ;; +- sh[1234]*-*-*) with_endian=big ;; ++ sh[j1234]*-*-*) with_endian=big ;; + *) with_endian=big,little ;; + esac + fi +@@ -3280,6 +3280,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + sh2a_nofpu*) sh_cpu_target=sh2a-nofpu ;; + sh2a*) sh_cpu_target=sh2a ;; + sh2e*) sh_cpu_target=sh2e ;; ++ shj2*) sh_cpu_target=shj2;; + sh2*) sh_cpu_target=sh2 ;; + *) sh_cpu_target=sh1 ;; + esac +@@ -3301,7 +3302,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + sh2a-single-only | sh2a-single | sh2a-nofpu | sh2a | \ + sh4a-single-only | sh4a-single | sh4a-nofpu | sh4a | sh4al | \ + sh4-single-only | sh4-single | sh4-nofpu | sh4 | sh4-300 | \ +- sh3e | sh3 | sh2e | sh2 | sh1) ;; ++ sh3e | sh3 | sh2e | sh2 | sh1 | shj2) ;; + "") sh_cpu_default=${sh_cpu_target} ;; + *) echo "with_cpu=$with_cpu not supported"; exit 1 ;; + esac +@@ -3310,9 +3311,9 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + case ${target} in + sh[1234]*) sh_multilibs=${sh_cpu_target} ;; + sh-superh-*) sh_multilibs=m4,m4-single,m4-single-only,m4-nofpu ;; +- sh*-*-linux*) sh_multilibs=m1,m2,m2a,m3e,m4 ;; ++ sh*-*-linux*) sh_multilibs=m1,m2,m2a,m3e,m4,mj2 ;; + sh*-*-netbsd*) sh_multilibs=m3,m3e,m4 ;; +- *) sh_multilibs=m1,m2,m2e,m4,m4-single,m4-single-only,m2a,m2a-single ;; ++ *) sh_multilibs=m1,m2,m2e,m4,m4-single,m4-single-only,m2a,m2a-single,mj2 ;; + esac + if test x$with_fp = xno; then + sh_multilibs="`echo $sh_multilibs|sed -e s/m4/sh4-nofpu/ -e s/,m4-[^,]*//g -e s/,m[23]e// -e s/m2a,m2a-single/m2a-nofpu/ -e s/m5-..m....,//g`" +@@ -3327,7 +3328,8 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + m1 | m2 | m2e | m3 | m3e | \ + m4 | m4-single | m4-single-only | m4-nofpu | m4-300 |\ + m4a | m4a-single | m4a-single-only | m4a-nofpu | m4al | \ +- m2a | m2a-single | m2a-single-only | m2a-nofpu) ++ m2a | m2a-single | m2a-single-only | m2a-nofpu | \ ++ mj2) + # TM_MULTILIB_CONFIG is used by t-sh for the non-endian multilib definition + # It is passed to MULTIILIB_OPTIONS verbatim. + TM_MULTILIB_CONFIG="${TM_MULTILIB_CONFIG}/${sh_multilib}" +@@ -3344,7 +3346,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + done + TM_MULTILIB_CONFIG=`echo $TM_MULTILIB_CONFIG | sed 's:^/::'` + if test x${enable_incomplete_targets} = xyes ; then +- tm_defines="$tm_defines SUPPORT_SH1=1 SUPPORT_SH2E=1 SUPPORT_SH4=1 SUPPORT_SH4_SINGLE=1 SUPPORT_SH2A=1 SUPPORT_SH2A_SINGLE=1" ++ tm_defines="$tm_defines SUPPORT_SH1=1 SUPPORT_SH2E=1 SUPPORT_SH4=1 SUPPORT_SH4_SINGLE=1 SUPPORT_SH2A=1 SUPPORT_SH2A_SINGLE=1 SUPPORT_SHJ2=1" + fi + tm_file="$tm_file ./sysroot-suffix.h" + tmake_file="$tmake_file t-sysroot-suffix" +@@ -5175,6 +5177,8 @@ case "${target}" in + ;; + m4a | m4a-single | m4a-single-only | m4a-nofpu | m4al) + ;; ++ mj2) ++ ;; + *) + echo "Unknown CPU used in --with-cpu=$with_cpu, known values:" 1>&2 + echo "m1 m2 m2e m3 m3e m4 m4-single m4-single-only m4-nofpu" 1>&2 +@@ -5385,7 +5389,7 @@ case ${target} in + tmake_file="${cpu_type}/t-${cpu_type} ${tmake_file}" + ;; + +- sh[123456ble]*-*-* | sh-*-*) ++ sh[123456blej]*-*-* | sh-*-*) + c_target_objs="${c_target_objs} sh-c.o" + cxx_target_objs="${cxx_target_objs} sh-c.o" + ;; +diff --git a/gcc/config/sh/sh.cc b/gcc/config/sh/sh.cc +index 1564109c942..798c1c1c1a3 100644 +--- a/gcc/config/sh/sh.cc ++++ b/gcc/config/sh/sh.cc +@@ -686,6 +686,7 @@ parse_validate_atomic_model_option (const char* str) + model_names[sh_atomic_model::hard_llcs] = "hard-llcs"; + model_names[sh_atomic_model::soft_tcb] = "soft-tcb"; + model_names[sh_atomic_model::soft_imask] = "soft-imask"; ++ model_names[sh_atomic_model::hard_cas] = "hard-cas"; + + const char* model_cdef_names[sh_atomic_model::num_models]; + model_cdef_names[sh_atomic_model::none] = "NONE"; +@@ -693,6 +694,7 @@ parse_validate_atomic_model_option (const char* str) + model_cdef_names[sh_atomic_model::hard_llcs] = "HARD_LLCS"; + model_cdef_names[sh_atomic_model::soft_tcb] = "SOFT_TCB"; + model_cdef_names[sh_atomic_model::soft_imask] = "SOFT_IMASK"; ++ model_cdef_names[sh_atomic_model::hard_cas] = "HARD_CAS"; + + sh_atomic_model ret; + ret.type = sh_atomic_model::none; +@@ -771,6 +773,9 @@ got_mode_name:; + if (ret.type == sh_atomic_model::soft_imask && TARGET_USERMODE) + err_ret ("cannot use atomic model %s in user mode", ret.name); + ++ if (ret.type == sh_atomic_model::hard_cas && !TARGET_SHJ2) ++ err_ret ("atomic model %s is only available J2 targets", ret.name); ++ + return ret; + + #undef err_ret +@@ -827,6 +832,8 @@ sh_option_override (void) + sh_cpu = PROCESSOR_SH2E; + if (TARGET_SH2A) + sh_cpu = PROCESSOR_SH2A; ++ if (TARGET_SHJ2) ++ sh_cpu = PROCESSOR_SHJ2; + if (TARGET_SH3) + sh_cpu = PROCESSOR_SH3; + if (TARGET_SH3E) +diff --git a/gcc/config/sh/sh.h b/gcc/config/sh/sh.h +index d2280e2ffe6..3a54a896721 100644 +--- a/gcc/config/sh/sh.h ++++ b/gcc/config/sh/sh.h +@@ -85,6 +85,7 @@ extern int code_for_indirect_jump_scratch; + #define SUPPORT_SH4_SINGLE 1 + #define SUPPORT_SH2A 1 + #define SUPPORT_SH2A_SINGLE 1 ++#define SUPPORT_SHJ2 1 + #endif + + #define TARGET_DIVIDE_CALL_DIV1 (sh_div_strategy == SH_DIV_CALL_DIV1) +@@ -117,6 +118,7 @@ extern int code_for_indirect_jump_scratch; + #define SELECT_SH4A_SINGLE_ONLY (MASK_SH4A | SELECT_SH4_SINGLE_ONLY) + #define SELECT_SH4A (MASK_SH4A | SELECT_SH4) + #define SELECT_SH4A_SINGLE (MASK_SH4A | SELECT_SH4_SINGLE) ++#define SELECT_SHJ2 (MASK_SHJ2 | SELECT_SH2) + + #if SUPPORT_SH1 + #define SUPPORT_SH2 1 +@@ -124,6 +126,7 @@ extern int code_for_indirect_jump_scratch; + #if SUPPORT_SH2 + #define SUPPORT_SH3 1 + #define SUPPORT_SH2A_NOFPU 1 ++#define SUPPORT_SHJ2 1 + #endif + #if SUPPORT_SH3 + #define SUPPORT_SH4_NOFPU 1 +@@ -156,7 +159,7 @@ extern int code_for_indirect_jump_scratch; + #define MASK_ARCH (MASK_SH1 | MASK_SH2 | MASK_SH3 | MASK_SH_E | MASK_SH4 \ + | MASK_HARD_SH2A | MASK_HARD_SH2A_DOUBLE | MASK_SH4A \ + | MASK_HARD_SH4 | MASK_FPU_SINGLE \ +- | MASK_FPU_SINGLE_ONLY) ++ | MASK_FPU_SINGLE_ONLY | MASK_SHJ2) + + /* This defaults us to big-endian. */ + #ifndef TARGET_ENDIAN_DEFAULT +@@ -231,7 +234,8 @@ extern int code_for_indirect_jump_scratch; + %{m2a-single:--isa=sh2a} \ + %{m2a-single-only:--isa=sh2a} \ + %{m2a-nofpu:--isa=sh2a-nofpu} \ +-%{m4al:-dsp}" ++%{m4al:-dsp} \ ++%{mj2:-isa=j2}" + + #define ASM_SPEC SH_ASM_SPEC + +@@ -347,6 +351,7 @@ struct sh_atomic_model + hard_llcs, + soft_tcb, + soft_imask, ++ hard_cas, + + num_models + }; +@@ -390,6 +395,9 @@ extern const sh_atomic_model& selected_atomic_model (void); + #define TARGET_ATOMIC_SOFT_IMASK \ + (selected_atomic_model ().type == sh_atomic_model::soft_imask) + ++#define TARGET_ATOMIC_HARD_CAS \ ++ (selected_atomic_model ().type == sh_atomic_model::hard_cas) ++ + #endif // __cplusplus + + #define SUBTARGET_OVERRIDE_OPTIONS (void) 0 +@@ -1484,7 +1492,7 @@ extern bool current_function_interrupt; + + /* Nonzero if the target supports dynamic shift instructions + like shad and shld. */ +-#define TARGET_DYNSHIFT (TARGET_SH3 || TARGET_SH2A) ++#define TARGET_DYNSHIFT (TARGET_SH3 || TARGET_SH2A || TARGET_SHJ2) + + /* The cost of using the dynamic shift insns (shad, shld) are the same + if they are available. If they are not available a library function will +@@ -1747,6 +1755,7 @@ enum processor_type { + PROCESSOR_SH2, + PROCESSOR_SH2E, + PROCESSOR_SH2A, ++ PROCESSOR_SHJ2, + PROCESSOR_SH3, + PROCESSOR_SH3E, + PROCESSOR_SH4, +diff --git a/gcc/config/sh/sh.opt b/gcc/config/sh/sh.opt +index b4755a812f3..0989a1c18da 100644 +--- a/gcc/config/sh/sh.opt ++++ b/gcc/config/sh/sh.opt +@@ -65,6 +65,10 @@ m2e + Target RejectNegative Condition(SUPPORT_SH2E) + Generate SH2e code. + ++mj2 ++Target RejectNegative Mask(SHJ2) Condition(SUPPORT_SHJ2) ++Generate J2 code. ++ + m3 + Target RejectNegative Mask(SH3) Condition(SUPPORT_SH3) + Generate SH3 code. +diff --git a/gcc/config/sh/sync.md b/gcc/config/sh/sync.md +index 2b43f8edb86..118fc5d06db 100644 +--- a/gcc/config/sh/sync.md ++++ b/gcc/config/sh/sync.md +@@ -240,6 +240,9 @@ + || (TARGET_SH4A && <MODE>mode == SImode && !TARGET_ATOMIC_STRICT)) + atomic_insn = gen_atomic_compare_and_swap<mode>_hard (old_val, mem, + exp_val, new_val); ++ else if (TARGET_ATOMIC_HARD_CAS && <MODE>mode == SImode) ++ atomic_insn = gen_atomic_compare_and_swap<mode>_cas (old_val, mem, ++ exp_val, new_val); + else if (TARGET_ATOMIC_SOFT_GUSA) + atomic_insn = gen_atomic_compare_and_swap<mode>_soft_gusa (old_val, mem, + exp_val, new_val); +@@ -306,6 +309,57 @@ + } + [(set_attr "length" "14")]) + ++(define_expand "atomic_compare_and_swapsi_cas" ++ [(set (match_operand:SI 0 "register_operand" "=r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "register_operand" "r") ++ (match_operand:SI 3 "register_operand" "r")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{ ++ rtx mem = gen_rtx_REG (SImode, 0); ++ emit_move_insn (mem, force_reg (SImode, XEXP (operands[1], 0))); ++ emit_insn (gen_shj2_cas (operands[0], mem, operands[2], operands[3])); ++ DONE; ++}) ++ ++(define_insn "shj2_cas" ++ [(set (match_operand:SI 0 "register_operand" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "register_operand" "=r") ++ (match_operand:SI 2 "register_operand" "r") ++ (match_operand:SI 3 "register_operand" "0")] ++ UNSPECV_CMPXCHG_1)) ++ (set (reg:SI T_REG) ++ (unspec_volatile:SI [(const_int 0)] UNSPECV_CMPXCHG_3))] ++ "TARGET_ATOMIC_HARD_CAS" ++ "cas.l %2,%0,@%1" ++ [(set_attr "length" "2")] ++) ++ ++(define_expand "atomic_compare_and_swapqi_cas" ++ [(set (match_operand:SI 0 "arith_reg_dest" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "arith_operand" "rI08") ++ (match_operand:SI 3 "arith_operand" "rI08")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{FAIL;} ++) ++ ++(define_expand "atomic_compare_and_swaphi_cas" ++ [(set (match_operand:SI 0 "arith_reg_dest" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "arith_operand" "rI08") ++ (match_operand:SI 3 "arith_operand" "rI08")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{FAIL;} ++) ++ + ;; The QIHImode llcs patterns modify the address register of the memory + ;; operand. In order to express that, we have to open code the memory + ;; operand. Initially the insn is expanded like every other atomic insn +diff --git a/gcc/config/sh/t-sh b/gcc/config/sh/t-sh +index 888f8ff7f25..29fd6ae45fd 100644 +--- a/gcc/config/sh/t-sh ++++ b/gcc/config/sh/t-sh +@@ -50,7 +50,8 @@ MULTILIB_MATCHES = $(shell \ + m2e,m3e,m4-single-only,m4-100-single-only,m4-200-single-only,m4-300-single-only,m4a-single-only \ + m2a-single,m2a-single-only \ + m4-single,m4-100-single,m4-200-single,m4-300-single,m4a-single \ +- m4,m4-100,m4-200,m4-300,m4a; do \ ++ m4,m4-100,m4-200,m4-300,m4a \ ++ mj2; do \ + subst= ; \ + for lib in `echo $$abi|tr , ' '` ; do \ + if test "`echo $$multilibs|sed s/$$lib//`" != "$$multilibs"; then \ +@@ -63,9 +64,9 @@ MULTILIB_MATCHES = $(shell \ + + # SH1 and SH2A support big endian only. + ifeq ($(DEFAULT_ENDIAN),ml) +-MULTILIB_EXCEPTIONS = m1 ml/m1 m2a* ml/m2a* $(TM_MULTILIB_EXCEPTIONS_CONFIG) ++MULTILIB_EXCEPTIONS = m1 ml/m1 m2a* ml/m2a* ml/mj2 $(TM_MULTILIB_EXCEPTIONS_CONFIG) + else +-MULTILIB_EXCEPTIONS = ml/m1 ml/m2a* $(TM_MULTILIB_EXCEPTIONS_CONFIG) ++MULTILIB_EXCEPTIONS = ml/m1 ml/m2a* ml/mj2 $(TM_MULTILIB_EXCEPTIONS_CONFIG) + endif + + MULTILIB_OSDIRNAMES = \ +@@ -87,7 +88,8 @@ MULTILIB_OSDIRNAMES = \ + m4a-single-only=!m4a-single-only $(OTHER_ENDIAN)/m4a-single-only=!$(OTHER_ENDIAN)/m4a-single-only \ + m4a-single=!m4a-single $(OTHER_ENDIAN)/m4a-single=!$(OTHER_ENDIAN)/m4a-single \ + m4a=!m4a $(OTHER_ENDIAN)/m4a=!$(OTHER_ENDIAN)/m4a \ +- m4al=!m4al $(OTHER_ENDIAN)/m4al=!$(OTHER_ENDIAN)/m4al ++ m4al=!m4al $(OTHER_ENDIAN)/m4al=!$(OTHER_ENDIAN)/m4al \ ++ mj2=!j2 + + $(out_object_file): gt-sh.h + gt-sh.h : s-gtype ; @true diff --git a/patches/gcc-12.4.0/0004-static-pie.diff b/patches/gcc-12.4.0/0004-static-pie.diff new file mode 100644 index 0000000..a85a3d7 --- /dev/null +++ b/patches/gcc-12.4.0/0004-static-pie.diff @@ -0,0 +1,92 @@ +diff --git a/gcc/common.opt b/gcc/common.opt +index a75b44ee47e..7c564818b49 100644 +--- a/gcc/common.opt ++++ b/gcc/common.opt +@@ -3473,11 +3473,11 @@ Driver + + no-pie + Driver RejectNegative Negative(shared) +-Don't create a dynamically linked position independent executable. ++Don't create a position independent executable. + + pie + Driver RejectNegative Negative(no-pie) +-Create a dynamically linked position independent executable. ++Create a position independent executable. + + static-pie + Driver RejectNegative Negative(pie) +diff --git a/gcc/config/gnu-user.h b/gcc/config/gnu-user.h +index 5ebbf42a13d..bb907d8e89a 100644 +--- a/gcc/config/gnu-user.h ++++ b/gcc/config/gnu-user.h +@@ -51,13 +51,12 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + #define GNU_USER_TARGET_STARTFILE_SPEC \ + "%{shared:; \ + pg|p|profile:%{static-pie:grcrt1.o%s;:gcrt1.o%s}; \ +- static:crt1.o%s; \ +- static-pie:rcrt1.o%s; \ ++ static|static-pie:%{" PIE_SPEC ":rcrt1.o%s;:crt1.o%s}; \ + " PIE_SPEC ":Scrt1.o%s; \ + :crt1.o%s} " \ + GNU_USER_TARGET_CRTI " \ +- %{static:crtbeginT.o%s; \ +- shared|static-pie|" PIE_SPEC ":crtbeginS.o%s; \ ++ %{shared|" PIE_SPEC ":crtbeginS.o%s; \ ++ static:crtbeginT.o%s; \ + :crtbegin.o%s} \ + %{fvtable-verify=none:%s; \ + fvtable-verify=preinit:vtv_start_preinit.o%s; \ +@@ -73,11 +72,11 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + GNU userspace "finalizer" file, `crtn.o'. */ + + #define GNU_USER_TARGET_ENDFILE_SPEC \ +- "%{!static:%{fvtable-verify=none:%s; \ ++ "%{static|static-pie:; \ ++ fvtable-verify=none:%s; \ + fvtable-verify=preinit:vtv_end_preinit.o%s; \ +- fvtable-verify=std:vtv_end.o%s}} \ +- %{static:crtend.o%s; \ +- shared|static-pie|" PIE_SPEC ":crtendS.o%s; \ ++ fvtable-verify=std:vtv_end.o%s} \ ++ %{shared|" PIE_SPEC ":crtendS.o%s; \ + :crtend.o%s} " \ + GNU_USER_TARGET_CRTN " " \ + CRTOFFLOADEND +@@ -106,7 +105,7 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + #define LIB_SPEC GNU_USER_TARGET_LIB_SPEC + + #if defined(HAVE_LD_EH_FRAME_HDR) +-#define LINK_EH_SPEC "%{!static|static-pie:--eh-frame-hdr} " ++#define LINK_EH_SPEC "%{!static|" PIE_SPEC ":--eh-frame-hdr} " + #endif + + #define GNU_USER_TARGET_LINK_GCC_C_SEQUENCE_SPEC \ +diff --git a/gcc/gcc.cc b/gcc/gcc.cc +index 3c81c5798d8..cd96eac5d12 100644 +--- a/gcc/gcc.cc ++++ b/gcc/gcc.cc +@@ -1010,7 +1010,7 @@ proper position among the other output files. */ + #define NO_FPIE_AND_FPIC_SPEC NO_FPIE_SPEC "|" NO_FPIC_SPEC + #define FPIE_OR_FPIC_SPEC NO_FPIE_AND_FPIC_SPEC ":;" + #else +-#define PIE_SPEC "pie" ++#define PIE_SPEC "pie|static-pie" + #define FPIE1_SPEC "fpie" + #define NO_FPIE1_SPEC FPIE1_SPEC ":;" + #define FPIE2_SPEC "fPIE" +@@ -1034,12 +1034,12 @@ proper position among the other output files. */ + #ifndef LINK_PIE_SPEC + #ifdef HAVE_LD_PIE + #ifndef LD_PIE_SPEC +-#define LD_PIE_SPEC "-pie" ++#define LD_PIE_SPEC "-pie %{static|static-pie:--no-dynamic-linker -z text -Bsymbolic}" + #endif + #else + #define LD_PIE_SPEC "" + #endif +-#define LINK_PIE_SPEC "%{static|shared|r:;" PIE_SPEC ":" LD_PIE_SPEC "} " ++#define LINK_PIE_SPEC "%{shared|r:;" PIE_SPEC ":" LD_PIE_SPEC "} " + #endif + + #ifndef LINK_BUILDID_SPEC diff --git a/patches/gcc-12.4.0/0005-m68k-sqrt.diff b/patches/gcc-12.4.0/0005-m68k-sqrt.diff new file mode 100644 index 0000000..652efb1 --- /dev/null +++ b/patches/gcc-12.4.0/0005-m68k-sqrt.diff @@ -0,0 +1,20 @@ +diff --git a/gcc/config/m68k/m68k.md b/gcc/config/m68k/m68k.md +index 59a456cd496..dbfddea41bd 100644 +--- a/gcc/config/m68k/m68k.md ++++ b/gcc/config/m68k/m68k.md +@@ -4174,13 +4174,13 @@ + (define_expand "sqrt<mode>2" + [(set (match_operand:FP 0 "nonimmediate_operand" "") + (sqrt:FP (match_operand:FP 1 "general_operand" "")))] +- "TARGET_HARD_FLOAT" ++ "(TARGET_68881 && TARGET_68040) || TARGET_COLDFIRE_FPU" + "") + + (define_insn "sqrt<mode>2_68881" + [(set (match_operand:FP 0 "nonimmediate_operand" "=f") + (sqrt:FP (match_operand:FP 1 "general_operand" "f<FP:dreg>m")))] +- "TARGET_68881" ++ "TARGET_68881 && TARGET_68040" + { + if (FP_REG_P (operands[1])) + return "f<FP:round>sqrt%.x %1,%0"; diff --git a/patches/gcc-12.4.0/0006-cow-libstdc++v3.diff b/patches/gcc-12.4.0/0006-cow-libstdc++v3.diff new file mode 100644 index 0000000..b49b22c --- /dev/null +++ b/patches/gcc-12.4.0/0006-cow-libstdc++v3.diff @@ -0,0 +1,8 @@ +--- a/libstdc++-v3/ChangeLog 2024-03-13 18:04:21.050801063 -0400 ++++ b/libstdc++-v3/ChangeLog 2024-03-13 18:04:25.665836804 -0400 +@@ -4068,4 +4068,4 @@ + + Copying and distribution of this file, with or without modification, + are permitted in any medium without royalty provided the copyright +-notice and this notice are preserved. ++notice and this notice are preserved. diff --git a/patches/gcc-12.4.0/0007-fdpic-unwind.diff b/patches/gcc-12.4.0/0007-fdpic-unwind.diff new file mode 100644 index 0000000..61d3479 --- /dev/null +++ b/patches/gcc-12.4.0/0007-fdpic-unwind.diff @@ -0,0 +1,11 @@ +--- a/libgcc/unwind-pe.h 2024-03-14 05:59:53.754073149 +0900 ++++ b/libgcc/unwind-pe.h 2024-03-14 06:00:41.226074492 +0900 +@@ -262,7 +262,7 @@ + + if (result != 0) + { +-#if __FDPIC__ ++#if __FDPIC__ && __arm__ + /* FDPIC relative addresses imply taking the GOT address + into account. */ + if ((encoding & DW_EH_PE_pcrel) && (encoding & DW_EH_PE_indirect)) diff --git a/patches/gcc-12.4.0/0008-fdpic-crtstuff-pr114158.diff b/patches/gcc-12.4.0/0008-fdpic-crtstuff-pr114158.diff new file mode 100644 index 0000000..3986eed --- /dev/null +++ b/patches/gcc-12.4.0/0008-fdpic-crtstuff-pr114158.diff @@ -0,0 +1,38 @@ +--- a/libgcc/crtstuff.c 2023-05-29 17:46:32.000000000 +0900 ++++ b/libgcc/crtstuff.c 2024-03-14 06:03:42.398079615 +0900 +@@ -441,17 +441,9 @@ + #ifdef FINI_SECTION_ASM_OP + CRT_CALL_STATIC_FUNCTION (FINI_SECTION_ASM_OP, __do_global_dtors_aux) + #elif defined (FINI_ARRAY_SECTION_ASM_OP) +-#if defined(__FDPIC__) +-__asm__("\t.equ\t__do_global_dtors_aux_alias, __do_global_dtors_aux\n"); +-extern char __do_global_dtors_aux_alias; +-static void *__do_global_dtors_aux_fini_array_entry[] +-__attribute__ ((__used__, section(".fini_array"), aligned(sizeof(void *)))) +- = { &__do_global_dtors_aux_alias }; +-#else /* defined(__FDPIC__) */ + static func_ptr __do_global_dtors_aux_fini_array_entry[] + __attribute__ ((__used__, section(".fini_array"), + aligned(__alignof__(func_ptr)))) = { __do_global_dtors_aux }; +-#endif /* defined(__FDPIC__) */ + #else /* !FINI_SECTION_ASM_OP && !FINI_ARRAY_SECTION_ASM_OP */ + static void __attribute__((used)) + __do_global_dtors_aux_1 (void) +@@ -494,17 +486,9 @@ + #ifdef __LIBGCC_INIT_SECTION_ASM_OP__ + CRT_CALL_STATIC_FUNCTION (__LIBGCC_INIT_SECTION_ASM_OP__, frame_dummy) + #else /* defined(__LIBGCC_INIT_SECTION_ASM_OP__) */ +-#if defined(__FDPIC__) +-__asm__("\t.equ\t__frame_dummy_alias, frame_dummy\n"); +-extern char __frame_dummy_alias; +-static void *__frame_dummy_init_array_entry[] +-__attribute__ ((__used__, section(".init_array"), aligned(sizeof(void *)))) +- = { &__frame_dummy_alias }; +-#else /* defined(__FDPIC__) */ + static func_ptr __frame_dummy_init_array_entry[] + __attribute__ ((__used__, section(".init_array"), + aligned(__alignof__(func_ptr)))) = { frame_dummy }; +-#endif /* defined(__FDPIC__) */ + #endif /* !defined(__LIBGCC_INIT_SECTION_ASM_OP__) */ + #endif /* USE_EH_FRAME_REGISTRY || USE_TM_CLONE_REGISTRY */ + diff --git a/patches/gcc-12.4.0/0009-sh-fdpic-pr114641.diff b/patches/gcc-12.4.0/0009-sh-fdpic-pr114641.diff new file mode 100644 index 0000000..020422c --- /dev/null +++ b/patches/gcc-12.4.0/0009-sh-fdpic-pr114641.diff @@ -0,0 +1,12 @@ +--- gcc-11.4.0/gcc/config/sh/sh.cc.orig 2024-04-04 05:52:42.125373614 +0900 ++++ gcc-11.4.0/gcc/config/sh/sh.cc 2024-04-04 22:54:01.875106654 +0900 +@@ -9147,7 +9147,7 @@ + { + /* Weak functions may be NULL which doesn't work with + GOTOFFFUNCDESC because the runtime offset is not known. */ +- if (SYMBOL_REF_WEAK (orig)) ++ if (SYMBOL_REF_WEAK (orig) || (TREE_PUBLIC(SYMBOL_REF_DECL(orig)) && DECL_VISIBILITY (SYMBOL_REF_DECL(orig)) != VISIBILITY_HIDDEN)) + emit_insn (gen_symGOTFUNCDESC2reg (reg, orig)); + else + emit_insn (gen_symGOTOFFFUNCDESC2reg (reg, orig)); + -- 2.45.2 ^ permalink raw reply [flat|nested] 9+ messages in thread
* [musl] [PATCH MCM 2/3] add gcc 13.3 hash and patches 2024-10-02 15:40 [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions J. Neuschäfer 2024-10-02 15:40 ` [musl] [PATCH MCM 1/3] add gcc 12.4 hash and patches J. Neuschäfer @ 2024-10-02 15:40 ` J. Neuschäfer 2024-10-02 15:40 ` [musl] [PATCH MCM 3/3] add gcc 14.2 " J. Neuschäfer 2024-10-22 21:52 ` [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions Rich Felker 3 siblings, 0 replies; 9+ messages in thread From: J. Neuschäfer @ 2024-10-02 15:40 UTC (permalink / raw) To: musl; +Cc: J. Neuschäfer The patches have been copied from 12.4. --- hashes/gcc-13.3.0.tar.xz.sha1 | 1 + patches/gcc-13.3.0/0001-ssp_nonshared.diff | 14 + patches/gcc-13.3.0/0002-posix_memalign.diff | 30 ++ patches/gcc-13.3.0/0003-j2.diff | 346 +++++++++++++++++++++ patches/gcc-13.3.0/0004-static-pie.diff | 92 ++++++ patches/gcc-13.3.0/0005-m68k-sqrt.diff | 20 ++ patches/gcc-13.3.0/0006-cow-libstdc++v3.diff | 8 + patches/gcc-13.3.0/0007-fdpic-unwind.diff | 11 + .../gcc-13.3.0/0008-fdpic-crtstuff-pr114158.diff | 38 +++ patches/gcc-13.3.0/0009-sh-fdpic-pr114641.diff | 12 + 10 files changed, 572 insertions(+) diff --git a/hashes/gcc-13.3.0.tar.xz.sha1 b/hashes/gcc-13.3.0.tar.xz.sha1 new file mode 100644 index 0000000..4dc0f16 --- /dev/null +++ b/hashes/gcc-13.3.0.tar.xz.sha1 @@ -0,0 +1 @@ +6501872415823c95d48be28853ce3ebd6c1040c4 gcc-13.3.0.tar.xz diff --git a/patches/gcc-13.3.0/0001-ssp_nonshared.diff b/patches/gcc-13.3.0/0001-ssp_nonshared.diff new file mode 100644 index 0000000..b250f15 --- /dev/null +++ b/patches/gcc-13.3.0/0001-ssp_nonshared.diff @@ -0,0 +1,14 @@ +diff --git a/gcc/gcc.cc b/gcc/gcc.cc +index 7837553958b..3c81c5798d8 100644 +--- a/gcc/gcc.cc ++++ b/gcc/gcc.cc +@@ -980,7 +980,8 @@ proper position among the other output files. */ + #ifndef LINK_SSP_SPEC + #ifdef TARGET_LIBC_PROVIDES_SSP + #define LINK_SSP_SPEC "%{fstack-protector|fstack-protector-all" \ +- "|fstack-protector-strong|fstack-protector-explicit:}" ++ "|fstack-protector-strong|fstack-protector-explicit" \ ++ ":-lssp_nonshared}" + #else + #define LINK_SSP_SPEC "%{fstack-protector|fstack-protector-all" \ + "|fstack-protector-strong|fstack-protector-explicit" \ diff --git a/patches/gcc-13.3.0/0002-posix_memalign.diff b/patches/gcc-13.3.0/0002-posix_memalign.diff new file mode 100644 index 0000000..1d80096 --- /dev/null +++ b/patches/gcc-13.3.0/0002-posix_memalign.diff @@ -0,0 +1,30 @@ +diff --git a/gcc/config/i386/pmm_malloc.h b/gcc/config/i386/pmm_malloc.h +index 1b0bfe37852..d7b2b19bb3c 100644 +--- a/gcc/config/i386/pmm_malloc.h ++++ b/gcc/config/i386/pmm_malloc.h +@@ -27,12 +27,13 @@ + #include <stdlib.h> + + /* We can't depend on <stdlib.h> since the prototype of posix_memalign +- may not be visible. */ ++ may not be visible and we can't pollute the namespace either. */ + #ifndef __cplusplus +-extern int posix_memalign (void **, size_t, size_t); ++extern int _mm_posix_memalign (void **, size_t, size_t) + #else +-extern "C" int posix_memalign (void **, size_t, size_t) throw (); ++extern "C" int _mm_posix_memalign (void **, size_t, size_t) throw () + #endif ++__asm__("posix_memalign"); + + static __inline void * + _mm_malloc (size_t __size, size_t __alignment) +@@ -42,7 +43,7 @@ _mm_malloc (size_t __size, size_t __alignment) + return malloc (__size); + if (__alignment == 2 || (sizeof (void *) == 8 && __alignment == 4)) + __alignment = sizeof (void *); +- if (posix_memalign (&__ptr, __alignment, __size) == 0) ++ if (_mm_posix_memalign (&__ptr, __alignment, __size) == 0) + return __ptr; + else + return NULL; diff --git a/patches/gcc-13.3.0/0003-j2.diff b/patches/gcc-13.3.0/0003-j2.diff new file mode 100644 index 0000000..c378645 --- /dev/null +++ b/patches/gcc-13.3.0/0003-j2.diff @@ -0,0 +1,346 @@ +diff --git a/gcc/config.gcc b/gcc/config.gcc +index 357b0bed067..528add999f2 100644 +--- a/gcc/config.gcc ++++ b/gcc/config.gcc +@@ -556,7 +556,7 @@ s390*-*-*) + extra_headers="s390intrin.h htmintrin.h htmxlintrin.h vecintrin.h" + ;; + # Note the 'l'; we need to be able to match e.g. "shle" or "shl". +-sh[123456789lbe]*-*-* | sh-*-*) ++sh[123456789lbej]*-*-* | sh-*-*) + cpu_type=sh + extra_options="${extra_options} fused-madd.opt" + extra_objs="${extra_objs} sh_treg_combine.o sh-mem.o sh_optimize_sett_clrt.o" +@@ -3202,18 +3202,18 @@ s390x-ibm-tpf*) + extra_options="${extra_options} s390/tpf.opt" + tmake_file="${tmake_file} s390/t-s390" + ;; +-sh-*-elf* | sh[12346l]*-*-elf* | \ +- sh-*-linux* | sh[2346lbe]*-*-linux* | \ ++sh-*-elf* | sh[12346lj]*-*-elf* | \ ++ sh-*-linux* | sh[2346lbej]*-*-linux* | \ + sh-*-netbsdelf* | shl*-*-netbsdelf*) + tmake_file="${tmake_file} sh/t-sh sh/t-elf" + if test x${with_endian} = x; then + case ${target} in +- sh[1234]*be-*-* | sh[1234]*eb-*-*) with_endian=big ;; ++ sh[j1234]*be-*-* | sh[j1234]*eb-*-*) with_endian=big ;; + shbe-*-* | sheb-*-*) with_endian=big,little ;; + sh[1234]l* | sh[34]*-*-linux*) with_endian=little ;; + shl* | sh*-*-linux* | \ + sh-superh-elf) with_endian=little,big ;; +- sh[1234]*-*-*) with_endian=big ;; ++ sh[j1234]*-*-*) with_endian=big ;; + *) with_endian=big,little ;; + esac + fi +@@ -3280,6 +3280,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + sh2a_nofpu*) sh_cpu_target=sh2a-nofpu ;; + sh2a*) sh_cpu_target=sh2a ;; + sh2e*) sh_cpu_target=sh2e ;; ++ shj2*) sh_cpu_target=shj2;; + sh2*) sh_cpu_target=sh2 ;; + *) sh_cpu_target=sh1 ;; + esac +@@ -3301,7 +3302,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + sh2a-single-only | sh2a-single | sh2a-nofpu | sh2a | \ + sh4a-single-only | sh4a-single | sh4a-nofpu | sh4a | sh4al | \ + sh4-single-only | sh4-single | sh4-nofpu | sh4 | sh4-300 | \ +- sh3e | sh3 | sh2e | sh2 | sh1) ;; ++ sh3e | sh3 | sh2e | sh2 | sh1 | shj2) ;; + "") sh_cpu_default=${sh_cpu_target} ;; + *) echo "with_cpu=$with_cpu not supported"; exit 1 ;; + esac +@@ -3310,9 +3311,9 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + case ${target} in + sh[1234]*) sh_multilibs=${sh_cpu_target} ;; + sh-superh-*) sh_multilibs=m4,m4-single,m4-single-only,m4-nofpu ;; +- sh*-*-linux*) sh_multilibs=m1,m2,m2a,m3e,m4 ;; ++ sh*-*-linux*) sh_multilibs=m1,m2,m2a,m3e,m4,mj2 ;; + sh*-*-netbsd*) sh_multilibs=m3,m3e,m4 ;; +- *) sh_multilibs=m1,m2,m2e,m4,m4-single,m4-single-only,m2a,m2a-single ;; ++ *) sh_multilibs=m1,m2,m2e,m4,m4-single,m4-single-only,m2a,m2a-single,mj2 ;; + esac + if test x$with_fp = xno; then + sh_multilibs="`echo $sh_multilibs|sed -e s/m4/sh4-nofpu/ -e s/,m4-[^,]*//g -e s/,m[23]e// -e s/m2a,m2a-single/m2a-nofpu/ -e s/m5-..m....,//g`" +@@ -3327,7 +3328,8 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + m1 | m2 | m2e | m3 | m3e | \ + m4 | m4-single | m4-single-only | m4-nofpu | m4-300 |\ + m4a | m4a-single | m4a-single-only | m4a-nofpu | m4al | \ +- m2a | m2a-single | m2a-single-only | m2a-nofpu) ++ m2a | m2a-single | m2a-single-only | m2a-nofpu | \ ++ mj2) + # TM_MULTILIB_CONFIG is used by t-sh for the non-endian multilib definition + # It is passed to MULTIILIB_OPTIONS verbatim. + TM_MULTILIB_CONFIG="${TM_MULTILIB_CONFIG}/${sh_multilib}" +@@ -3344,7 +3346,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + done + TM_MULTILIB_CONFIG=`echo $TM_MULTILIB_CONFIG | sed 's:^/::'` + if test x${enable_incomplete_targets} = xyes ; then +- tm_defines="$tm_defines SUPPORT_SH1=1 SUPPORT_SH2E=1 SUPPORT_SH4=1 SUPPORT_SH4_SINGLE=1 SUPPORT_SH2A=1 SUPPORT_SH2A_SINGLE=1" ++ tm_defines="$tm_defines SUPPORT_SH1=1 SUPPORT_SH2E=1 SUPPORT_SH4=1 SUPPORT_SH4_SINGLE=1 SUPPORT_SH2A=1 SUPPORT_SH2A_SINGLE=1 SUPPORT_SHJ2=1" + fi + tm_file="$tm_file ./sysroot-suffix.h" + tmake_file="$tmake_file t-sysroot-suffix" +@@ -5175,6 +5177,8 @@ case "${target}" in + ;; + m4a | m4a-single | m4a-single-only | m4a-nofpu | m4al) + ;; ++ mj2) ++ ;; + *) + echo "Unknown CPU used in --with-cpu=$with_cpu, known values:" 1>&2 + echo "m1 m2 m2e m3 m3e m4 m4-single m4-single-only m4-nofpu" 1>&2 +@@ -5385,7 +5389,7 @@ case ${target} in + tmake_file="${cpu_type}/t-${cpu_type} ${tmake_file}" + ;; + +- sh[123456ble]*-*-* | sh-*-*) ++ sh[123456blej]*-*-* | sh-*-*) + c_target_objs="${c_target_objs} sh-c.o" + cxx_target_objs="${cxx_target_objs} sh-c.o" + ;; +diff --git a/gcc/config/sh/sh.cc b/gcc/config/sh/sh.cc +index 1564109c942..798c1c1c1a3 100644 +--- a/gcc/config/sh/sh.cc ++++ b/gcc/config/sh/sh.cc +@@ -686,6 +686,7 @@ parse_validate_atomic_model_option (const char* str) + model_names[sh_atomic_model::hard_llcs] = "hard-llcs"; + model_names[sh_atomic_model::soft_tcb] = "soft-tcb"; + model_names[sh_atomic_model::soft_imask] = "soft-imask"; ++ model_names[sh_atomic_model::hard_cas] = "hard-cas"; + + const char* model_cdef_names[sh_atomic_model::num_models]; + model_cdef_names[sh_atomic_model::none] = "NONE"; +@@ -693,6 +694,7 @@ parse_validate_atomic_model_option (const char* str) + model_cdef_names[sh_atomic_model::hard_llcs] = "HARD_LLCS"; + model_cdef_names[sh_atomic_model::soft_tcb] = "SOFT_TCB"; + model_cdef_names[sh_atomic_model::soft_imask] = "SOFT_IMASK"; ++ model_cdef_names[sh_atomic_model::hard_cas] = "HARD_CAS"; + + sh_atomic_model ret; + ret.type = sh_atomic_model::none; +@@ -771,6 +773,9 @@ got_mode_name:; + if (ret.type == sh_atomic_model::soft_imask && TARGET_USERMODE) + err_ret ("cannot use atomic model %s in user mode", ret.name); + ++ if (ret.type == sh_atomic_model::hard_cas && !TARGET_SHJ2) ++ err_ret ("atomic model %s is only available J2 targets", ret.name); ++ + return ret; + + #undef err_ret +@@ -827,6 +832,8 @@ sh_option_override (void) + sh_cpu = PROCESSOR_SH2E; + if (TARGET_SH2A) + sh_cpu = PROCESSOR_SH2A; ++ if (TARGET_SHJ2) ++ sh_cpu = PROCESSOR_SHJ2; + if (TARGET_SH3) + sh_cpu = PROCESSOR_SH3; + if (TARGET_SH3E) +diff --git a/gcc/config/sh/sh.h b/gcc/config/sh/sh.h +index d2280e2ffe6..3a54a896721 100644 +--- a/gcc/config/sh/sh.h ++++ b/gcc/config/sh/sh.h +@@ -85,6 +85,7 @@ extern int code_for_indirect_jump_scratch; + #define SUPPORT_SH4_SINGLE 1 + #define SUPPORT_SH2A 1 + #define SUPPORT_SH2A_SINGLE 1 ++#define SUPPORT_SHJ2 1 + #endif + + #define TARGET_DIVIDE_CALL_DIV1 (sh_div_strategy == SH_DIV_CALL_DIV1) +@@ -117,6 +118,7 @@ extern int code_for_indirect_jump_scratch; + #define SELECT_SH4A_SINGLE_ONLY (MASK_SH4A | SELECT_SH4_SINGLE_ONLY) + #define SELECT_SH4A (MASK_SH4A | SELECT_SH4) + #define SELECT_SH4A_SINGLE (MASK_SH4A | SELECT_SH4_SINGLE) ++#define SELECT_SHJ2 (MASK_SHJ2 | SELECT_SH2) + + #if SUPPORT_SH1 + #define SUPPORT_SH2 1 +@@ -124,6 +126,7 @@ extern int code_for_indirect_jump_scratch; + #if SUPPORT_SH2 + #define SUPPORT_SH3 1 + #define SUPPORT_SH2A_NOFPU 1 ++#define SUPPORT_SHJ2 1 + #endif + #if SUPPORT_SH3 + #define SUPPORT_SH4_NOFPU 1 +@@ -156,7 +159,7 @@ extern int code_for_indirect_jump_scratch; + #define MASK_ARCH (MASK_SH1 | MASK_SH2 | MASK_SH3 | MASK_SH_E | MASK_SH4 \ + | MASK_HARD_SH2A | MASK_HARD_SH2A_DOUBLE | MASK_SH4A \ + | MASK_HARD_SH4 | MASK_FPU_SINGLE \ +- | MASK_FPU_SINGLE_ONLY) ++ | MASK_FPU_SINGLE_ONLY | MASK_SHJ2) + + /* This defaults us to big-endian. */ + #ifndef TARGET_ENDIAN_DEFAULT +@@ -231,7 +234,8 @@ extern int code_for_indirect_jump_scratch; + %{m2a-single:--isa=sh2a} \ + %{m2a-single-only:--isa=sh2a} \ + %{m2a-nofpu:--isa=sh2a-nofpu} \ +-%{m4al:-dsp}" ++%{m4al:-dsp} \ ++%{mj2:-isa=j2}" + + #define ASM_SPEC SH_ASM_SPEC + +@@ -347,6 +351,7 @@ struct sh_atomic_model + hard_llcs, + soft_tcb, + soft_imask, ++ hard_cas, + + num_models + }; +@@ -390,6 +395,9 @@ extern const sh_atomic_model& selected_atomic_model (void); + #define TARGET_ATOMIC_SOFT_IMASK \ + (selected_atomic_model ().type == sh_atomic_model::soft_imask) + ++#define TARGET_ATOMIC_HARD_CAS \ ++ (selected_atomic_model ().type == sh_atomic_model::hard_cas) ++ + #endif // __cplusplus + + #define SUBTARGET_OVERRIDE_OPTIONS (void) 0 +@@ -1484,7 +1492,7 @@ extern bool current_function_interrupt; + + /* Nonzero if the target supports dynamic shift instructions + like shad and shld. */ +-#define TARGET_DYNSHIFT (TARGET_SH3 || TARGET_SH2A) ++#define TARGET_DYNSHIFT (TARGET_SH3 || TARGET_SH2A || TARGET_SHJ2) + + /* The cost of using the dynamic shift insns (shad, shld) are the same + if they are available. If they are not available a library function will +@@ -1747,6 +1755,7 @@ enum processor_type { + PROCESSOR_SH2, + PROCESSOR_SH2E, + PROCESSOR_SH2A, ++ PROCESSOR_SHJ2, + PROCESSOR_SH3, + PROCESSOR_SH3E, + PROCESSOR_SH4, +diff --git a/gcc/config/sh/sh.opt b/gcc/config/sh/sh.opt +index b4755a812f3..0989a1c18da 100644 +--- a/gcc/config/sh/sh.opt ++++ b/gcc/config/sh/sh.opt +@@ -65,6 +65,10 @@ m2e + Target RejectNegative Condition(SUPPORT_SH2E) + Generate SH2e code. + ++mj2 ++Target RejectNegative Mask(SHJ2) Condition(SUPPORT_SHJ2) ++Generate J2 code. ++ + m3 + Target RejectNegative Mask(SH3) Condition(SUPPORT_SH3) + Generate SH3 code. +diff --git a/gcc/config/sh/sync.md b/gcc/config/sh/sync.md +index 2b43f8edb86..118fc5d06db 100644 +--- a/gcc/config/sh/sync.md ++++ b/gcc/config/sh/sync.md +@@ -240,6 +240,9 @@ + || (TARGET_SH4A && <MODE>mode == SImode && !TARGET_ATOMIC_STRICT)) + atomic_insn = gen_atomic_compare_and_swap<mode>_hard (old_val, mem, + exp_val, new_val); ++ else if (TARGET_ATOMIC_HARD_CAS && <MODE>mode == SImode) ++ atomic_insn = gen_atomic_compare_and_swap<mode>_cas (old_val, mem, ++ exp_val, new_val); + else if (TARGET_ATOMIC_SOFT_GUSA) + atomic_insn = gen_atomic_compare_and_swap<mode>_soft_gusa (old_val, mem, + exp_val, new_val); +@@ -306,6 +309,57 @@ + } + [(set_attr "length" "14")]) + ++(define_expand "atomic_compare_and_swapsi_cas" ++ [(set (match_operand:SI 0 "register_operand" "=r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "register_operand" "r") ++ (match_operand:SI 3 "register_operand" "r")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{ ++ rtx mem = gen_rtx_REG (SImode, 0); ++ emit_move_insn (mem, force_reg (SImode, XEXP (operands[1], 0))); ++ emit_insn (gen_shj2_cas (operands[0], mem, operands[2], operands[3])); ++ DONE; ++}) ++ ++(define_insn "shj2_cas" ++ [(set (match_operand:SI 0 "register_operand" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "register_operand" "=r") ++ (match_operand:SI 2 "register_operand" "r") ++ (match_operand:SI 3 "register_operand" "0")] ++ UNSPECV_CMPXCHG_1)) ++ (set (reg:SI T_REG) ++ (unspec_volatile:SI [(const_int 0)] UNSPECV_CMPXCHG_3))] ++ "TARGET_ATOMIC_HARD_CAS" ++ "cas.l %2,%0,@%1" ++ [(set_attr "length" "2")] ++) ++ ++(define_expand "atomic_compare_and_swapqi_cas" ++ [(set (match_operand:SI 0 "arith_reg_dest" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "arith_operand" "rI08") ++ (match_operand:SI 3 "arith_operand" "rI08")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{FAIL;} ++) ++ ++(define_expand "atomic_compare_and_swaphi_cas" ++ [(set (match_operand:SI 0 "arith_reg_dest" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "arith_operand" "rI08") ++ (match_operand:SI 3 "arith_operand" "rI08")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{FAIL;} ++) ++ + ;; The QIHImode llcs patterns modify the address register of the memory + ;; operand. In order to express that, we have to open code the memory + ;; operand. Initially the insn is expanded like every other atomic insn +diff --git a/gcc/config/sh/t-sh b/gcc/config/sh/t-sh +index 888f8ff7f25..29fd6ae45fd 100644 +--- a/gcc/config/sh/t-sh ++++ b/gcc/config/sh/t-sh +@@ -50,7 +50,8 @@ MULTILIB_MATCHES = $(shell \ + m2e,m3e,m4-single-only,m4-100-single-only,m4-200-single-only,m4-300-single-only,m4a-single-only \ + m2a-single,m2a-single-only \ + m4-single,m4-100-single,m4-200-single,m4-300-single,m4a-single \ +- m4,m4-100,m4-200,m4-300,m4a; do \ ++ m4,m4-100,m4-200,m4-300,m4a \ ++ mj2; do \ + subst= ; \ + for lib in `echo $$abi|tr , ' '` ; do \ + if test "`echo $$multilibs|sed s/$$lib//`" != "$$multilibs"; then \ +@@ -63,9 +64,9 @@ MULTILIB_MATCHES = $(shell \ + + # SH1 and SH2A support big endian only. + ifeq ($(DEFAULT_ENDIAN),ml) +-MULTILIB_EXCEPTIONS = m1 ml/m1 m2a* ml/m2a* $(TM_MULTILIB_EXCEPTIONS_CONFIG) ++MULTILIB_EXCEPTIONS = m1 ml/m1 m2a* ml/m2a* ml/mj2 $(TM_MULTILIB_EXCEPTIONS_CONFIG) + else +-MULTILIB_EXCEPTIONS = ml/m1 ml/m2a* $(TM_MULTILIB_EXCEPTIONS_CONFIG) ++MULTILIB_EXCEPTIONS = ml/m1 ml/m2a* ml/mj2 $(TM_MULTILIB_EXCEPTIONS_CONFIG) + endif + + MULTILIB_OSDIRNAMES = \ +@@ -87,7 +88,8 @@ MULTILIB_OSDIRNAMES = \ + m4a-single-only=!m4a-single-only $(OTHER_ENDIAN)/m4a-single-only=!$(OTHER_ENDIAN)/m4a-single-only \ + m4a-single=!m4a-single $(OTHER_ENDIAN)/m4a-single=!$(OTHER_ENDIAN)/m4a-single \ + m4a=!m4a $(OTHER_ENDIAN)/m4a=!$(OTHER_ENDIAN)/m4a \ +- m4al=!m4al $(OTHER_ENDIAN)/m4al=!$(OTHER_ENDIAN)/m4al ++ m4al=!m4al $(OTHER_ENDIAN)/m4al=!$(OTHER_ENDIAN)/m4al \ ++ mj2=!j2 + + $(out_object_file): gt-sh.h + gt-sh.h : s-gtype ; @true diff --git a/patches/gcc-13.3.0/0004-static-pie.diff b/patches/gcc-13.3.0/0004-static-pie.diff new file mode 100644 index 0000000..a85a3d7 --- /dev/null +++ b/patches/gcc-13.3.0/0004-static-pie.diff @@ -0,0 +1,92 @@ +diff --git a/gcc/common.opt b/gcc/common.opt +index a75b44ee47e..7c564818b49 100644 +--- a/gcc/common.opt ++++ b/gcc/common.opt +@@ -3473,11 +3473,11 @@ Driver + + no-pie + Driver RejectNegative Negative(shared) +-Don't create a dynamically linked position independent executable. ++Don't create a position independent executable. + + pie + Driver RejectNegative Negative(no-pie) +-Create a dynamically linked position independent executable. ++Create a position independent executable. + + static-pie + Driver RejectNegative Negative(pie) +diff --git a/gcc/config/gnu-user.h b/gcc/config/gnu-user.h +index 5ebbf42a13d..bb907d8e89a 100644 +--- a/gcc/config/gnu-user.h ++++ b/gcc/config/gnu-user.h +@@ -51,13 +51,12 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + #define GNU_USER_TARGET_STARTFILE_SPEC \ + "%{shared:; \ + pg|p|profile:%{static-pie:grcrt1.o%s;:gcrt1.o%s}; \ +- static:crt1.o%s; \ +- static-pie:rcrt1.o%s; \ ++ static|static-pie:%{" PIE_SPEC ":rcrt1.o%s;:crt1.o%s}; \ + " PIE_SPEC ":Scrt1.o%s; \ + :crt1.o%s} " \ + GNU_USER_TARGET_CRTI " \ +- %{static:crtbeginT.o%s; \ +- shared|static-pie|" PIE_SPEC ":crtbeginS.o%s; \ ++ %{shared|" PIE_SPEC ":crtbeginS.o%s; \ ++ static:crtbeginT.o%s; \ + :crtbegin.o%s} \ + %{fvtable-verify=none:%s; \ + fvtable-verify=preinit:vtv_start_preinit.o%s; \ +@@ -73,11 +72,11 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + GNU userspace "finalizer" file, `crtn.o'. */ + + #define GNU_USER_TARGET_ENDFILE_SPEC \ +- "%{!static:%{fvtable-verify=none:%s; \ ++ "%{static|static-pie:; \ ++ fvtable-verify=none:%s; \ + fvtable-verify=preinit:vtv_end_preinit.o%s; \ +- fvtable-verify=std:vtv_end.o%s}} \ +- %{static:crtend.o%s; \ +- shared|static-pie|" PIE_SPEC ":crtendS.o%s; \ ++ fvtable-verify=std:vtv_end.o%s} \ ++ %{shared|" PIE_SPEC ":crtendS.o%s; \ + :crtend.o%s} " \ + GNU_USER_TARGET_CRTN " " \ + CRTOFFLOADEND +@@ -106,7 +105,7 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + #define LIB_SPEC GNU_USER_TARGET_LIB_SPEC + + #if defined(HAVE_LD_EH_FRAME_HDR) +-#define LINK_EH_SPEC "%{!static|static-pie:--eh-frame-hdr} " ++#define LINK_EH_SPEC "%{!static|" PIE_SPEC ":--eh-frame-hdr} " + #endif + + #define GNU_USER_TARGET_LINK_GCC_C_SEQUENCE_SPEC \ +diff --git a/gcc/gcc.cc b/gcc/gcc.cc +index 3c81c5798d8..cd96eac5d12 100644 +--- a/gcc/gcc.cc ++++ b/gcc/gcc.cc +@@ -1010,7 +1010,7 @@ proper position among the other output files. */ + #define NO_FPIE_AND_FPIC_SPEC NO_FPIE_SPEC "|" NO_FPIC_SPEC + #define FPIE_OR_FPIC_SPEC NO_FPIE_AND_FPIC_SPEC ":;" + #else +-#define PIE_SPEC "pie" ++#define PIE_SPEC "pie|static-pie" + #define FPIE1_SPEC "fpie" + #define NO_FPIE1_SPEC FPIE1_SPEC ":;" + #define FPIE2_SPEC "fPIE" +@@ -1034,12 +1034,12 @@ proper position among the other output files. */ + #ifndef LINK_PIE_SPEC + #ifdef HAVE_LD_PIE + #ifndef LD_PIE_SPEC +-#define LD_PIE_SPEC "-pie" ++#define LD_PIE_SPEC "-pie %{static|static-pie:--no-dynamic-linker -z text -Bsymbolic}" + #endif + #else + #define LD_PIE_SPEC "" + #endif +-#define LINK_PIE_SPEC "%{static|shared|r:;" PIE_SPEC ":" LD_PIE_SPEC "} " ++#define LINK_PIE_SPEC "%{shared|r:;" PIE_SPEC ":" LD_PIE_SPEC "} " + #endif + + #ifndef LINK_BUILDID_SPEC diff --git a/patches/gcc-13.3.0/0005-m68k-sqrt.diff b/patches/gcc-13.3.0/0005-m68k-sqrt.diff new file mode 100644 index 0000000..652efb1 --- /dev/null +++ b/patches/gcc-13.3.0/0005-m68k-sqrt.diff @@ -0,0 +1,20 @@ +diff --git a/gcc/config/m68k/m68k.md b/gcc/config/m68k/m68k.md +index 59a456cd496..dbfddea41bd 100644 +--- a/gcc/config/m68k/m68k.md ++++ b/gcc/config/m68k/m68k.md +@@ -4174,13 +4174,13 @@ + (define_expand "sqrt<mode>2" + [(set (match_operand:FP 0 "nonimmediate_operand" "") + (sqrt:FP (match_operand:FP 1 "general_operand" "")))] +- "TARGET_HARD_FLOAT" ++ "(TARGET_68881 && TARGET_68040) || TARGET_COLDFIRE_FPU" + "") + + (define_insn "sqrt<mode>2_68881" + [(set (match_operand:FP 0 "nonimmediate_operand" "=f") + (sqrt:FP (match_operand:FP 1 "general_operand" "f<FP:dreg>m")))] +- "TARGET_68881" ++ "TARGET_68881 && TARGET_68040" + { + if (FP_REG_P (operands[1])) + return "f<FP:round>sqrt%.x %1,%0"; diff --git a/patches/gcc-13.3.0/0006-cow-libstdc++v3.diff b/patches/gcc-13.3.0/0006-cow-libstdc++v3.diff new file mode 100644 index 0000000..b49b22c --- /dev/null +++ b/patches/gcc-13.3.0/0006-cow-libstdc++v3.diff @@ -0,0 +1,8 @@ +--- a/libstdc++-v3/ChangeLog 2024-03-13 18:04:21.050801063 -0400 ++++ b/libstdc++-v3/ChangeLog 2024-03-13 18:04:25.665836804 -0400 +@@ -4068,4 +4068,4 @@ + + Copying and distribution of this file, with or without modification, + are permitted in any medium without royalty provided the copyright +-notice and this notice are preserved. ++notice and this notice are preserved. diff --git a/patches/gcc-13.3.0/0007-fdpic-unwind.diff b/patches/gcc-13.3.0/0007-fdpic-unwind.diff new file mode 100644 index 0000000..61d3479 --- /dev/null +++ b/patches/gcc-13.3.0/0007-fdpic-unwind.diff @@ -0,0 +1,11 @@ +--- a/libgcc/unwind-pe.h 2024-03-14 05:59:53.754073149 +0900 ++++ b/libgcc/unwind-pe.h 2024-03-14 06:00:41.226074492 +0900 +@@ -262,7 +262,7 @@ + + if (result != 0) + { +-#if __FDPIC__ ++#if __FDPIC__ && __arm__ + /* FDPIC relative addresses imply taking the GOT address + into account. */ + if ((encoding & DW_EH_PE_pcrel) && (encoding & DW_EH_PE_indirect)) diff --git a/patches/gcc-13.3.0/0008-fdpic-crtstuff-pr114158.diff b/patches/gcc-13.3.0/0008-fdpic-crtstuff-pr114158.diff new file mode 100644 index 0000000..3986eed --- /dev/null +++ b/patches/gcc-13.3.0/0008-fdpic-crtstuff-pr114158.diff @@ -0,0 +1,38 @@ +--- a/libgcc/crtstuff.c 2023-05-29 17:46:32.000000000 +0900 ++++ b/libgcc/crtstuff.c 2024-03-14 06:03:42.398079615 +0900 +@@ -441,17 +441,9 @@ + #ifdef FINI_SECTION_ASM_OP + CRT_CALL_STATIC_FUNCTION (FINI_SECTION_ASM_OP, __do_global_dtors_aux) + #elif defined (FINI_ARRAY_SECTION_ASM_OP) +-#if defined(__FDPIC__) +-__asm__("\t.equ\t__do_global_dtors_aux_alias, __do_global_dtors_aux\n"); +-extern char __do_global_dtors_aux_alias; +-static void *__do_global_dtors_aux_fini_array_entry[] +-__attribute__ ((__used__, section(".fini_array"), aligned(sizeof(void *)))) +- = { &__do_global_dtors_aux_alias }; +-#else /* defined(__FDPIC__) */ + static func_ptr __do_global_dtors_aux_fini_array_entry[] + __attribute__ ((__used__, section(".fini_array"), + aligned(__alignof__(func_ptr)))) = { __do_global_dtors_aux }; +-#endif /* defined(__FDPIC__) */ + #else /* !FINI_SECTION_ASM_OP && !FINI_ARRAY_SECTION_ASM_OP */ + static void __attribute__((used)) + __do_global_dtors_aux_1 (void) +@@ -494,17 +486,9 @@ + #ifdef __LIBGCC_INIT_SECTION_ASM_OP__ + CRT_CALL_STATIC_FUNCTION (__LIBGCC_INIT_SECTION_ASM_OP__, frame_dummy) + #else /* defined(__LIBGCC_INIT_SECTION_ASM_OP__) */ +-#if defined(__FDPIC__) +-__asm__("\t.equ\t__frame_dummy_alias, frame_dummy\n"); +-extern char __frame_dummy_alias; +-static void *__frame_dummy_init_array_entry[] +-__attribute__ ((__used__, section(".init_array"), aligned(sizeof(void *)))) +- = { &__frame_dummy_alias }; +-#else /* defined(__FDPIC__) */ + static func_ptr __frame_dummy_init_array_entry[] + __attribute__ ((__used__, section(".init_array"), + aligned(__alignof__(func_ptr)))) = { frame_dummy }; +-#endif /* defined(__FDPIC__) */ + #endif /* !defined(__LIBGCC_INIT_SECTION_ASM_OP__) */ + #endif /* USE_EH_FRAME_REGISTRY || USE_TM_CLONE_REGISTRY */ + diff --git a/patches/gcc-13.3.0/0009-sh-fdpic-pr114641.diff b/patches/gcc-13.3.0/0009-sh-fdpic-pr114641.diff new file mode 100644 index 0000000..020422c --- /dev/null +++ b/patches/gcc-13.3.0/0009-sh-fdpic-pr114641.diff @@ -0,0 +1,12 @@ +--- gcc-11.4.0/gcc/config/sh/sh.cc.orig 2024-04-04 05:52:42.125373614 +0900 ++++ gcc-11.4.0/gcc/config/sh/sh.cc 2024-04-04 22:54:01.875106654 +0900 +@@ -9147,7 +9147,7 @@ + { + /* Weak functions may be NULL which doesn't work with + GOTOFFFUNCDESC because the runtime offset is not known. */ +- if (SYMBOL_REF_WEAK (orig)) ++ if (SYMBOL_REF_WEAK (orig) || (TREE_PUBLIC(SYMBOL_REF_DECL(orig)) && DECL_VISIBILITY (SYMBOL_REF_DECL(orig)) != VISIBILITY_HIDDEN)) + emit_insn (gen_symGOTFUNCDESC2reg (reg, orig)); + else + emit_insn (gen_symGOTOFFFUNCDESC2reg (reg, orig)); + -- 2.45.2 ^ permalink raw reply [flat|nested] 9+ messages in thread
* [musl] [PATCH MCM 3/3] add gcc 14.2 hash and patches 2024-10-02 15:40 [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions J. Neuschäfer 2024-10-02 15:40 ` [musl] [PATCH MCM 1/3] add gcc 12.4 hash and patches J. Neuschäfer 2024-10-02 15:40 ` [musl] [PATCH MCM 2/3] add gcc 13.3 " J. Neuschäfer @ 2024-10-02 15:40 ` J. Neuschäfer 2024-10-22 21:52 ` [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions Rich Felker 3 siblings, 0 replies; 9+ messages in thread From: J. Neuschäfer @ 2024-10-02 15:40 UTC (permalink / raw) To: musl; +Cc: J. Neuschäfer The patches have been copied from 12.4. --- hashes/gcc-14.2.0.tar.xz.sha1 | 1 + patches/gcc-14.2.0/0001-ssp_nonshared.diff | 14 + patches/gcc-14.2.0/0002-posix_memalign.diff | 30 ++ patches/gcc-14.2.0/0003-j2.diff | 346 +++++++++++++++++++++ patches/gcc-14.2.0/0004-static-pie.diff | 92 ++++++ patches/gcc-14.2.0/0005-m68k-sqrt.diff | 20 ++ patches/gcc-14.2.0/0006-cow-libstdc++v3.diff | 8 + patches/gcc-14.2.0/0007-fdpic-unwind.diff | 11 + .../gcc-14.2.0/0008-fdpic-crtstuff-pr114158.diff | 38 +++ patches/gcc-14.2.0/0009-sh-fdpic-pr114641.diff | 12 + 10 files changed, 572 insertions(+) diff --git a/hashes/gcc-14.2.0.tar.xz.sha1 b/hashes/gcc-14.2.0.tar.xz.sha1 new file mode 100644 index 0000000..0b9c02e --- /dev/null +++ b/hashes/gcc-14.2.0.tar.xz.sha1 @@ -0,0 +1 @@ +d91ecc3d20ce6298bd95f9b09cc51dc6d3c73ae3 gcc-14.2.0.tar.xz diff --git a/patches/gcc-14.2.0/0001-ssp_nonshared.diff b/patches/gcc-14.2.0/0001-ssp_nonshared.diff new file mode 100644 index 0000000..b250f15 --- /dev/null +++ b/patches/gcc-14.2.0/0001-ssp_nonshared.diff @@ -0,0 +1,14 @@ +diff --git a/gcc/gcc.cc b/gcc/gcc.cc +index 7837553958b..3c81c5798d8 100644 +--- a/gcc/gcc.cc ++++ b/gcc/gcc.cc +@@ -980,7 +980,8 @@ proper position among the other output files. */ + #ifndef LINK_SSP_SPEC + #ifdef TARGET_LIBC_PROVIDES_SSP + #define LINK_SSP_SPEC "%{fstack-protector|fstack-protector-all" \ +- "|fstack-protector-strong|fstack-protector-explicit:}" ++ "|fstack-protector-strong|fstack-protector-explicit" \ ++ ":-lssp_nonshared}" + #else + #define LINK_SSP_SPEC "%{fstack-protector|fstack-protector-all" \ + "|fstack-protector-strong|fstack-protector-explicit" \ diff --git a/patches/gcc-14.2.0/0002-posix_memalign.diff b/patches/gcc-14.2.0/0002-posix_memalign.diff new file mode 100644 index 0000000..1d80096 --- /dev/null +++ b/patches/gcc-14.2.0/0002-posix_memalign.diff @@ -0,0 +1,30 @@ +diff --git a/gcc/config/i386/pmm_malloc.h b/gcc/config/i386/pmm_malloc.h +index 1b0bfe37852..d7b2b19bb3c 100644 +--- a/gcc/config/i386/pmm_malloc.h ++++ b/gcc/config/i386/pmm_malloc.h +@@ -27,12 +27,13 @@ + #include <stdlib.h> + + /* We can't depend on <stdlib.h> since the prototype of posix_memalign +- may not be visible. */ ++ may not be visible and we can't pollute the namespace either. */ + #ifndef __cplusplus +-extern int posix_memalign (void **, size_t, size_t); ++extern int _mm_posix_memalign (void **, size_t, size_t) + #else +-extern "C" int posix_memalign (void **, size_t, size_t) throw (); ++extern "C" int _mm_posix_memalign (void **, size_t, size_t) throw () + #endif ++__asm__("posix_memalign"); + + static __inline void * + _mm_malloc (size_t __size, size_t __alignment) +@@ -42,7 +43,7 @@ _mm_malloc (size_t __size, size_t __alignment) + return malloc (__size); + if (__alignment == 2 || (sizeof (void *) == 8 && __alignment == 4)) + __alignment = sizeof (void *); +- if (posix_memalign (&__ptr, __alignment, __size) == 0) ++ if (_mm_posix_memalign (&__ptr, __alignment, __size) == 0) + return __ptr; + else + return NULL; diff --git a/patches/gcc-14.2.0/0003-j2.diff b/patches/gcc-14.2.0/0003-j2.diff new file mode 100644 index 0000000..c378645 --- /dev/null +++ b/patches/gcc-14.2.0/0003-j2.diff @@ -0,0 +1,346 @@ +diff --git a/gcc/config.gcc b/gcc/config.gcc +index 357b0bed067..528add999f2 100644 +--- a/gcc/config.gcc ++++ b/gcc/config.gcc +@@ -556,7 +556,7 @@ s390*-*-*) + extra_headers="s390intrin.h htmintrin.h htmxlintrin.h vecintrin.h" + ;; + # Note the 'l'; we need to be able to match e.g. "shle" or "shl". +-sh[123456789lbe]*-*-* | sh-*-*) ++sh[123456789lbej]*-*-* | sh-*-*) + cpu_type=sh + extra_options="${extra_options} fused-madd.opt" + extra_objs="${extra_objs} sh_treg_combine.o sh-mem.o sh_optimize_sett_clrt.o" +@@ -3202,18 +3202,18 @@ s390x-ibm-tpf*) + extra_options="${extra_options} s390/tpf.opt" + tmake_file="${tmake_file} s390/t-s390" + ;; +-sh-*-elf* | sh[12346l]*-*-elf* | \ +- sh-*-linux* | sh[2346lbe]*-*-linux* | \ ++sh-*-elf* | sh[12346lj]*-*-elf* | \ ++ sh-*-linux* | sh[2346lbej]*-*-linux* | \ + sh-*-netbsdelf* | shl*-*-netbsdelf*) + tmake_file="${tmake_file} sh/t-sh sh/t-elf" + if test x${with_endian} = x; then + case ${target} in +- sh[1234]*be-*-* | sh[1234]*eb-*-*) with_endian=big ;; ++ sh[j1234]*be-*-* | sh[j1234]*eb-*-*) with_endian=big ;; + shbe-*-* | sheb-*-*) with_endian=big,little ;; + sh[1234]l* | sh[34]*-*-linux*) with_endian=little ;; + shl* | sh*-*-linux* | \ + sh-superh-elf) with_endian=little,big ;; +- sh[1234]*-*-*) with_endian=big ;; ++ sh[j1234]*-*-*) with_endian=big ;; + *) with_endian=big,little ;; + esac + fi +@@ -3280,6 +3280,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + sh2a_nofpu*) sh_cpu_target=sh2a-nofpu ;; + sh2a*) sh_cpu_target=sh2a ;; + sh2e*) sh_cpu_target=sh2e ;; ++ shj2*) sh_cpu_target=shj2;; + sh2*) sh_cpu_target=sh2 ;; + *) sh_cpu_target=sh1 ;; + esac +@@ -3301,7 +3302,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + sh2a-single-only | sh2a-single | sh2a-nofpu | sh2a | \ + sh4a-single-only | sh4a-single | sh4a-nofpu | sh4a | sh4al | \ + sh4-single-only | sh4-single | sh4-nofpu | sh4 | sh4-300 | \ +- sh3e | sh3 | sh2e | sh2 | sh1) ;; ++ sh3e | sh3 | sh2e | sh2 | sh1 | shj2) ;; + "") sh_cpu_default=${sh_cpu_target} ;; + *) echo "with_cpu=$with_cpu not supported"; exit 1 ;; + esac +@@ -3310,9 +3311,9 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + case ${target} in + sh[1234]*) sh_multilibs=${sh_cpu_target} ;; + sh-superh-*) sh_multilibs=m4,m4-single,m4-single-only,m4-nofpu ;; +- sh*-*-linux*) sh_multilibs=m1,m2,m2a,m3e,m4 ;; ++ sh*-*-linux*) sh_multilibs=m1,m2,m2a,m3e,m4,mj2 ;; + sh*-*-netbsd*) sh_multilibs=m3,m3e,m4 ;; +- *) sh_multilibs=m1,m2,m2e,m4,m4-single,m4-single-only,m2a,m2a-single ;; ++ *) sh_multilibs=m1,m2,m2e,m4,m4-single,m4-single-only,m2a,m2a-single,mj2 ;; + esac + if test x$with_fp = xno; then + sh_multilibs="`echo $sh_multilibs|sed -e s/m4/sh4-nofpu/ -e s/,m4-[^,]*//g -e s/,m[23]e// -e s/m2a,m2a-single/m2a-nofpu/ -e s/m5-..m....,//g`" +@@ -3327,7 +3328,8 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + m1 | m2 | m2e | m3 | m3e | \ + m4 | m4-single | m4-single-only | m4-nofpu | m4-300 |\ + m4a | m4a-single | m4a-single-only | m4a-nofpu | m4al | \ +- m2a | m2a-single | m2a-single-only | m2a-nofpu) ++ m2a | m2a-single | m2a-single-only | m2a-nofpu | \ ++ mj2) + # TM_MULTILIB_CONFIG is used by t-sh for the non-endian multilib definition + # It is passed to MULTIILIB_OPTIONS verbatim. + TM_MULTILIB_CONFIG="${TM_MULTILIB_CONFIG}/${sh_multilib}" +@@ -3344,7 +3346,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + done + TM_MULTILIB_CONFIG=`echo $TM_MULTILIB_CONFIG | sed 's:^/::'` + if test x${enable_incomplete_targets} = xyes ; then +- tm_defines="$tm_defines SUPPORT_SH1=1 SUPPORT_SH2E=1 SUPPORT_SH4=1 SUPPORT_SH4_SINGLE=1 SUPPORT_SH2A=1 SUPPORT_SH2A_SINGLE=1" ++ tm_defines="$tm_defines SUPPORT_SH1=1 SUPPORT_SH2E=1 SUPPORT_SH4=1 SUPPORT_SH4_SINGLE=1 SUPPORT_SH2A=1 SUPPORT_SH2A_SINGLE=1 SUPPORT_SHJ2=1" + fi + tm_file="$tm_file ./sysroot-suffix.h" + tmake_file="$tmake_file t-sysroot-suffix" +@@ -5175,6 +5177,8 @@ case "${target}" in + ;; + m4a | m4a-single | m4a-single-only | m4a-nofpu | m4al) + ;; ++ mj2) ++ ;; + *) + echo "Unknown CPU used in --with-cpu=$with_cpu, known values:" 1>&2 + echo "m1 m2 m2e m3 m3e m4 m4-single m4-single-only m4-nofpu" 1>&2 +@@ -5385,7 +5389,7 @@ case ${target} in + tmake_file="${cpu_type}/t-${cpu_type} ${tmake_file}" + ;; + +- sh[123456ble]*-*-* | sh-*-*) ++ sh[123456blej]*-*-* | sh-*-*) + c_target_objs="${c_target_objs} sh-c.o" + cxx_target_objs="${cxx_target_objs} sh-c.o" + ;; +diff --git a/gcc/config/sh/sh.cc b/gcc/config/sh/sh.cc +index 1564109c942..798c1c1c1a3 100644 +--- a/gcc/config/sh/sh.cc ++++ b/gcc/config/sh/sh.cc +@@ -686,6 +686,7 @@ parse_validate_atomic_model_option (const char* str) + model_names[sh_atomic_model::hard_llcs] = "hard-llcs"; + model_names[sh_atomic_model::soft_tcb] = "soft-tcb"; + model_names[sh_atomic_model::soft_imask] = "soft-imask"; ++ model_names[sh_atomic_model::hard_cas] = "hard-cas"; + + const char* model_cdef_names[sh_atomic_model::num_models]; + model_cdef_names[sh_atomic_model::none] = "NONE"; +@@ -693,6 +694,7 @@ parse_validate_atomic_model_option (const char* str) + model_cdef_names[sh_atomic_model::hard_llcs] = "HARD_LLCS"; + model_cdef_names[sh_atomic_model::soft_tcb] = "SOFT_TCB"; + model_cdef_names[sh_atomic_model::soft_imask] = "SOFT_IMASK"; ++ model_cdef_names[sh_atomic_model::hard_cas] = "HARD_CAS"; + + sh_atomic_model ret; + ret.type = sh_atomic_model::none; +@@ -771,6 +773,9 @@ got_mode_name:; + if (ret.type == sh_atomic_model::soft_imask && TARGET_USERMODE) + err_ret ("cannot use atomic model %s in user mode", ret.name); + ++ if (ret.type == sh_atomic_model::hard_cas && !TARGET_SHJ2) ++ err_ret ("atomic model %s is only available J2 targets", ret.name); ++ + return ret; + + #undef err_ret +@@ -827,6 +832,8 @@ sh_option_override (void) + sh_cpu = PROCESSOR_SH2E; + if (TARGET_SH2A) + sh_cpu = PROCESSOR_SH2A; ++ if (TARGET_SHJ2) ++ sh_cpu = PROCESSOR_SHJ2; + if (TARGET_SH3) + sh_cpu = PROCESSOR_SH3; + if (TARGET_SH3E) +diff --git a/gcc/config/sh/sh.h b/gcc/config/sh/sh.h +index d2280e2ffe6..3a54a896721 100644 +--- a/gcc/config/sh/sh.h ++++ b/gcc/config/sh/sh.h +@@ -85,6 +85,7 @@ extern int code_for_indirect_jump_scratch; + #define SUPPORT_SH4_SINGLE 1 + #define SUPPORT_SH2A 1 + #define SUPPORT_SH2A_SINGLE 1 ++#define SUPPORT_SHJ2 1 + #endif + + #define TARGET_DIVIDE_CALL_DIV1 (sh_div_strategy == SH_DIV_CALL_DIV1) +@@ -117,6 +118,7 @@ extern int code_for_indirect_jump_scratch; + #define SELECT_SH4A_SINGLE_ONLY (MASK_SH4A | SELECT_SH4_SINGLE_ONLY) + #define SELECT_SH4A (MASK_SH4A | SELECT_SH4) + #define SELECT_SH4A_SINGLE (MASK_SH4A | SELECT_SH4_SINGLE) ++#define SELECT_SHJ2 (MASK_SHJ2 | SELECT_SH2) + + #if SUPPORT_SH1 + #define SUPPORT_SH2 1 +@@ -124,6 +126,7 @@ extern int code_for_indirect_jump_scratch; + #if SUPPORT_SH2 + #define SUPPORT_SH3 1 + #define SUPPORT_SH2A_NOFPU 1 ++#define SUPPORT_SHJ2 1 + #endif + #if SUPPORT_SH3 + #define SUPPORT_SH4_NOFPU 1 +@@ -156,7 +159,7 @@ extern int code_for_indirect_jump_scratch; + #define MASK_ARCH (MASK_SH1 | MASK_SH2 | MASK_SH3 | MASK_SH_E | MASK_SH4 \ + | MASK_HARD_SH2A | MASK_HARD_SH2A_DOUBLE | MASK_SH4A \ + | MASK_HARD_SH4 | MASK_FPU_SINGLE \ +- | MASK_FPU_SINGLE_ONLY) ++ | MASK_FPU_SINGLE_ONLY | MASK_SHJ2) + + /* This defaults us to big-endian. */ + #ifndef TARGET_ENDIAN_DEFAULT +@@ -231,7 +234,8 @@ extern int code_for_indirect_jump_scratch; + %{m2a-single:--isa=sh2a} \ + %{m2a-single-only:--isa=sh2a} \ + %{m2a-nofpu:--isa=sh2a-nofpu} \ +-%{m4al:-dsp}" ++%{m4al:-dsp} \ ++%{mj2:-isa=j2}" + + #define ASM_SPEC SH_ASM_SPEC + +@@ -347,6 +351,7 @@ struct sh_atomic_model + hard_llcs, + soft_tcb, + soft_imask, ++ hard_cas, + + num_models + }; +@@ -390,6 +395,9 @@ extern const sh_atomic_model& selected_atomic_model (void); + #define TARGET_ATOMIC_SOFT_IMASK \ + (selected_atomic_model ().type == sh_atomic_model::soft_imask) + ++#define TARGET_ATOMIC_HARD_CAS \ ++ (selected_atomic_model ().type == sh_atomic_model::hard_cas) ++ + #endif // __cplusplus + + #define SUBTARGET_OVERRIDE_OPTIONS (void) 0 +@@ -1484,7 +1492,7 @@ extern bool current_function_interrupt; + + /* Nonzero if the target supports dynamic shift instructions + like shad and shld. */ +-#define TARGET_DYNSHIFT (TARGET_SH3 || TARGET_SH2A) ++#define TARGET_DYNSHIFT (TARGET_SH3 || TARGET_SH2A || TARGET_SHJ2) + + /* The cost of using the dynamic shift insns (shad, shld) are the same + if they are available. If they are not available a library function will +@@ -1747,6 +1755,7 @@ enum processor_type { + PROCESSOR_SH2, + PROCESSOR_SH2E, + PROCESSOR_SH2A, ++ PROCESSOR_SHJ2, + PROCESSOR_SH3, + PROCESSOR_SH3E, + PROCESSOR_SH4, +diff --git a/gcc/config/sh/sh.opt b/gcc/config/sh/sh.opt +index b4755a812f3..0989a1c18da 100644 +--- a/gcc/config/sh/sh.opt ++++ b/gcc/config/sh/sh.opt +@@ -65,6 +65,10 @@ m2e + Target RejectNegative Condition(SUPPORT_SH2E) + Generate SH2e code. + ++mj2 ++Target RejectNegative Mask(SHJ2) Condition(SUPPORT_SHJ2) ++Generate J2 code. ++ + m3 + Target RejectNegative Mask(SH3) Condition(SUPPORT_SH3) + Generate SH3 code. +diff --git a/gcc/config/sh/sync.md b/gcc/config/sh/sync.md +index 2b43f8edb86..118fc5d06db 100644 +--- a/gcc/config/sh/sync.md ++++ b/gcc/config/sh/sync.md +@@ -240,6 +240,9 @@ + || (TARGET_SH4A && <MODE>mode == SImode && !TARGET_ATOMIC_STRICT)) + atomic_insn = gen_atomic_compare_and_swap<mode>_hard (old_val, mem, + exp_val, new_val); ++ else if (TARGET_ATOMIC_HARD_CAS && <MODE>mode == SImode) ++ atomic_insn = gen_atomic_compare_and_swap<mode>_cas (old_val, mem, ++ exp_val, new_val); + else if (TARGET_ATOMIC_SOFT_GUSA) + atomic_insn = gen_atomic_compare_and_swap<mode>_soft_gusa (old_val, mem, + exp_val, new_val); +@@ -306,6 +309,57 @@ + } + [(set_attr "length" "14")]) + ++(define_expand "atomic_compare_and_swapsi_cas" ++ [(set (match_operand:SI 0 "register_operand" "=r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "register_operand" "r") ++ (match_operand:SI 3 "register_operand" "r")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{ ++ rtx mem = gen_rtx_REG (SImode, 0); ++ emit_move_insn (mem, force_reg (SImode, XEXP (operands[1], 0))); ++ emit_insn (gen_shj2_cas (operands[0], mem, operands[2], operands[3])); ++ DONE; ++}) ++ ++(define_insn "shj2_cas" ++ [(set (match_operand:SI 0 "register_operand" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "register_operand" "=r") ++ (match_operand:SI 2 "register_operand" "r") ++ (match_operand:SI 3 "register_operand" "0")] ++ UNSPECV_CMPXCHG_1)) ++ (set (reg:SI T_REG) ++ (unspec_volatile:SI [(const_int 0)] UNSPECV_CMPXCHG_3))] ++ "TARGET_ATOMIC_HARD_CAS" ++ "cas.l %2,%0,@%1" ++ [(set_attr "length" "2")] ++) ++ ++(define_expand "atomic_compare_and_swapqi_cas" ++ [(set (match_operand:SI 0 "arith_reg_dest" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "arith_operand" "rI08") ++ (match_operand:SI 3 "arith_operand" "rI08")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{FAIL;} ++) ++ ++(define_expand "atomic_compare_and_swaphi_cas" ++ [(set (match_operand:SI 0 "arith_reg_dest" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "arith_operand" "rI08") ++ (match_operand:SI 3 "arith_operand" "rI08")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{FAIL;} ++) ++ + ;; The QIHImode llcs patterns modify the address register of the memory + ;; operand. In order to express that, we have to open code the memory + ;; operand. Initially the insn is expanded like every other atomic insn +diff --git a/gcc/config/sh/t-sh b/gcc/config/sh/t-sh +index 888f8ff7f25..29fd6ae45fd 100644 +--- a/gcc/config/sh/t-sh ++++ b/gcc/config/sh/t-sh +@@ -50,7 +50,8 @@ MULTILIB_MATCHES = $(shell \ + m2e,m3e,m4-single-only,m4-100-single-only,m4-200-single-only,m4-300-single-only,m4a-single-only \ + m2a-single,m2a-single-only \ + m4-single,m4-100-single,m4-200-single,m4-300-single,m4a-single \ +- m4,m4-100,m4-200,m4-300,m4a; do \ ++ m4,m4-100,m4-200,m4-300,m4a \ ++ mj2; do \ + subst= ; \ + for lib in `echo $$abi|tr , ' '` ; do \ + if test "`echo $$multilibs|sed s/$$lib//`" != "$$multilibs"; then \ +@@ -63,9 +64,9 @@ MULTILIB_MATCHES = $(shell \ + + # SH1 and SH2A support big endian only. + ifeq ($(DEFAULT_ENDIAN),ml) +-MULTILIB_EXCEPTIONS = m1 ml/m1 m2a* ml/m2a* $(TM_MULTILIB_EXCEPTIONS_CONFIG) ++MULTILIB_EXCEPTIONS = m1 ml/m1 m2a* ml/m2a* ml/mj2 $(TM_MULTILIB_EXCEPTIONS_CONFIG) + else +-MULTILIB_EXCEPTIONS = ml/m1 ml/m2a* $(TM_MULTILIB_EXCEPTIONS_CONFIG) ++MULTILIB_EXCEPTIONS = ml/m1 ml/m2a* ml/mj2 $(TM_MULTILIB_EXCEPTIONS_CONFIG) + endif + + MULTILIB_OSDIRNAMES = \ +@@ -87,7 +88,8 @@ MULTILIB_OSDIRNAMES = \ + m4a-single-only=!m4a-single-only $(OTHER_ENDIAN)/m4a-single-only=!$(OTHER_ENDIAN)/m4a-single-only \ + m4a-single=!m4a-single $(OTHER_ENDIAN)/m4a-single=!$(OTHER_ENDIAN)/m4a-single \ + m4a=!m4a $(OTHER_ENDIAN)/m4a=!$(OTHER_ENDIAN)/m4a \ +- m4al=!m4al $(OTHER_ENDIAN)/m4al=!$(OTHER_ENDIAN)/m4al ++ m4al=!m4al $(OTHER_ENDIAN)/m4al=!$(OTHER_ENDIAN)/m4al \ ++ mj2=!j2 + + $(out_object_file): gt-sh.h + gt-sh.h : s-gtype ; @true diff --git a/patches/gcc-14.2.0/0004-static-pie.diff b/patches/gcc-14.2.0/0004-static-pie.diff new file mode 100644 index 0000000..a85a3d7 --- /dev/null +++ b/patches/gcc-14.2.0/0004-static-pie.diff @@ -0,0 +1,92 @@ +diff --git a/gcc/common.opt b/gcc/common.opt +index a75b44ee47e..7c564818b49 100644 +--- a/gcc/common.opt ++++ b/gcc/common.opt +@@ -3473,11 +3473,11 @@ Driver + + no-pie + Driver RejectNegative Negative(shared) +-Don't create a dynamically linked position independent executable. ++Don't create a position independent executable. + + pie + Driver RejectNegative Negative(no-pie) +-Create a dynamically linked position independent executable. ++Create a position independent executable. + + static-pie + Driver RejectNegative Negative(pie) +diff --git a/gcc/config/gnu-user.h b/gcc/config/gnu-user.h +index 5ebbf42a13d..bb907d8e89a 100644 +--- a/gcc/config/gnu-user.h ++++ b/gcc/config/gnu-user.h +@@ -51,13 +51,12 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + #define GNU_USER_TARGET_STARTFILE_SPEC \ + "%{shared:; \ + pg|p|profile:%{static-pie:grcrt1.o%s;:gcrt1.o%s}; \ +- static:crt1.o%s; \ +- static-pie:rcrt1.o%s; \ ++ static|static-pie:%{" PIE_SPEC ":rcrt1.o%s;:crt1.o%s}; \ + " PIE_SPEC ":Scrt1.o%s; \ + :crt1.o%s} " \ + GNU_USER_TARGET_CRTI " \ +- %{static:crtbeginT.o%s; \ +- shared|static-pie|" PIE_SPEC ":crtbeginS.o%s; \ ++ %{shared|" PIE_SPEC ":crtbeginS.o%s; \ ++ static:crtbeginT.o%s; \ + :crtbegin.o%s} \ + %{fvtable-verify=none:%s; \ + fvtable-verify=preinit:vtv_start_preinit.o%s; \ +@@ -73,11 +72,11 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + GNU userspace "finalizer" file, `crtn.o'. */ + + #define GNU_USER_TARGET_ENDFILE_SPEC \ +- "%{!static:%{fvtable-verify=none:%s; \ ++ "%{static|static-pie:; \ ++ fvtable-verify=none:%s; \ + fvtable-verify=preinit:vtv_end_preinit.o%s; \ +- fvtable-verify=std:vtv_end.o%s}} \ +- %{static:crtend.o%s; \ +- shared|static-pie|" PIE_SPEC ":crtendS.o%s; \ ++ fvtable-verify=std:vtv_end.o%s} \ ++ %{shared|" PIE_SPEC ":crtendS.o%s; \ + :crtend.o%s} " \ + GNU_USER_TARGET_CRTN " " \ + CRTOFFLOADEND +@@ -106,7 +105,7 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + #define LIB_SPEC GNU_USER_TARGET_LIB_SPEC + + #if defined(HAVE_LD_EH_FRAME_HDR) +-#define LINK_EH_SPEC "%{!static|static-pie:--eh-frame-hdr} " ++#define LINK_EH_SPEC "%{!static|" PIE_SPEC ":--eh-frame-hdr} " + #endif + + #define GNU_USER_TARGET_LINK_GCC_C_SEQUENCE_SPEC \ +diff --git a/gcc/gcc.cc b/gcc/gcc.cc +index 3c81c5798d8..cd96eac5d12 100644 +--- a/gcc/gcc.cc ++++ b/gcc/gcc.cc +@@ -1010,7 +1010,7 @@ proper position among the other output files. */ + #define NO_FPIE_AND_FPIC_SPEC NO_FPIE_SPEC "|" NO_FPIC_SPEC + #define FPIE_OR_FPIC_SPEC NO_FPIE_AND_FPIC_SPEC ":;" + #else +-#define PIE_SPEC "pie" ++#define PIE_SPEC "pie|static-pie" + #define FPIE1_SPEC "fpie" + #define NO_FPIE1_SPEC FPIE1_SPEC ":;" + #define FPIE2_SPEC "fPIE" +@@ -1034,12 +1034,12 @@ proper position among the other output files. */ + #ifndef LINK_PIE_SPEC + #ifdef HAVE_LD_PIE + #ifndef LD_PIE_SPEC +-#define LD_PIE_SPEC "-pie" ++#define LD_PIE_SPEC "-pie %{static|static-pie:--no-dynamic-linker -z text -Bsymbolic}" + #endif + #else + #define LD_PIE_SPEC "" + #endif +-#define LINK_PIE_SPEC "%{static|shared|r:;" PIE_SPEC ":" LD_PIE_SPEC "} " ++#define LINK_PIE_SPEC "%{shared|r:;" PIE_SPEC ":" LD_PIE_SPEC "} " + #endif + + #ifndef LINK_BUILDID_SPEC diff --git a/patches/gcc-14.2.0/0005-m68k-sqrt.diff b/patches/gcc-14.2.0/0005-m68k-sqrt.diff new file mode 100644 index 0000000..652efb1 --- /dev/null +++ b/patches/gcc-14.2.0/0005-m68k-sqrt.diff @@ -0,0 +1,20 @@ +diff --git a/gcc/config/m68k/m68k.md b/gcc/config/m68k/m68k.md +index 59a456cd496..dbfddea41bd 100644 +--- a/gcc/config/m68k/m68k.md ++++ b/gcc/config/m68k/m68k.md +@@ -4174,13 +4174,13 @@ + (define_expand "sqrt<mode>2" + [(set (match_operand:FP 0 "nonimmediate_operand" "") + (sqrt:FP (match_operand:FP 1 "general_operand" "")))] +- "TARGET_HARD_FLOAT" ++ "(TARGET_68881 && TARGET_68040) || TARGET_COLDFIRE_FPU" + "") + + (define_insn "sqrt<mode>2_68881" + [(set (match_operand:FP 0 "nonimmediate_operand" "=f") + (sqrt:FP (match_operand:FP 1 "general_operand" "f<FP:dreg>m")))] +- "TARGET_68881" ++ "TARGET_68881 && TARGET_68040" + { + if (FP_REG_P (operands[1])) + return "f<FP:round>sqrt%.x %1,%0"; diff --git a/patches/gcc-14.2.0/0006-cow-libstdc++v3.diff b/patches/gcc-14.2.0/0006-cow-libstdc++v3.diff new file mode 100644 index 0000000..b49b22c --- /dev/null +++ b/patches/gcc-14.2.0/0006-cow-libstdc++v3.diff @@ -0,0 +1,8 @@ +--- a/libstdc++-v3/ChangeLog 2024-03-13 18:04:21.050801063 -0400 ++++ b/libstdc++-v3/ChangeLog 2024-03-13 18:04:25.665836804 -0400 +@@ -4068,4 +4068,4 @@ + + Copying and distribution of this file, with or without modification, + are permitted in any medium without royalty provided the copyright +-notice and this notice are preserved. ++notice and this notice are preserved. diff --git a/patches/gcc-14.2.0/0007-fdpic-unwind.diff b/patches/gcc-14.2.0/0007-fdpic-unwind.diff new file mode 100644 index 0000000..61d3479 --- /dev/null +++ b/patches/gcc-14.2.0/0007-fdpic-unwind.diff @@ -0,0 +1,11 @@ +--- a/libgcc/unwind-pe.h 2024-03-14 05:59:53.754073149 +0900 ++++ b/libgcc/unwind-pe.h 2024-03-14 06:00:41.226074492 +0900 +@@ -262,7 +262,7 @@ + + if (result != 0) + { +-#if __FDPIC__ ++#if __FDPIC__ && __arm__ + /* FDPIC relative addresses imply taking the GOT address + into account. */ + if ((encoding & DW_EH_PE_pcrel) && (encoding & DW_EH_PE_indirect)) diff --git a/patches/gcc-14.2.0/0008-fdpic-crtstuff-pr114158.diff b/patches/gcc-14.2.0/0008-fdpic-crtstuff-pr114158.diff new file mode 100644 index 0000000..3986eed --- /dev/null +++ b/patches/gcc-14.2.0/0008-fdpic-crtstuff-pr114158.diff @@ -0,0 +1,38 @@ +--- a/libgcc/crtstuff.c 2023-05-29 17:46:32.000000000 +0900 ++++ b/libgcc/crtstuff.c 2024-03-14 06:03:42.398079615 +0900 +@@ -441,17 +441,9 @@ + #ifdef FINI_SECTION_ASM_OP + CRT_CALL_STATIC_FUNCTION (FINI_SECTION_ASM_OP, __do_global_dtors_aux) + #elif defined (FINI_ARRAY_SECTION_ASM_OP) +-#if defined(__FDPIC__) +-__asm__("\t.equ\t__do_global_dtors_aux_alias, __do_global_dtors_aux\n"); +-extern char __do_global_dtors_aux_alias; +-static void *__do_global_dtors_aux_fini_array_entry[] +-__attribute__ ((__used__, section(".fini_array"), aligned(sizeof(void *)))) +- = { &__do_global_dtors_aux_alias }; +-#else /* defined(__FDPIC__) */ + static func_ptr __do_global_dtors_aux_fini_array_entry[] + __attribute__ ((__used__, section(".fini_array"), + aligned(__alignof__(func_ptr)))) = { __do_global_dtors_aux }; +-#endif /* defined(__FDPIC__) */ + #else /* !FINI_SECTION_ASM_OP && !FINI_ARRAY_SECTION_ASM_OP */ + static void __attribute__((used)) + __do_global_dtors_aux_1 (void) +@@ -494,17 +486,9 @@ + #ifdef __LIBGCC_INIT_SECTION_ASM_OP__ + CRT_CALL_STATIC_FUNCTION (__LIBGCC_INIT_SECTION_ASM_OP__, frame_dummy) + #else /* defined(__LIBGCC_INIT_SECTION_ASM_OP__) */ +-#if defined(__FDPIC__) +-__asm__("\t.equ\t__frame_dummy_alias, frame_dummy\n"); +-extern char __frame_dummy_alias; +-static void *__frame_dummy_init_array_entry[] +-__attribute__ ((__used__, section(".init_array"), aligned(sizeof(void *)))) +- = { &__frame_dummy_alias }; +-#else /* defined(__FDPIC__) */ + static func_ptr __frame_dummy_init_array_entry[] + __attribute__ ((__used__, section(".init_array"), + aligned(__alignof__(func_ptr)))) = { frame_dummy }; +-#endif /* defined(__FDPIC__) */ + #endif /* !defined(__LIBGCC_INIT_SECTION_ASM_OP__) */ + #endif /* USE_EH_FRAME_REGISTRY || USE_TM_CLONE_REGISTRY */ + diff --git a/patches/gcc-14.2.0/0009-sh-fdpic-pr114641.diff b/patches/gcc-14.2.0/0009-sh-fdpic-pr114641.diff new file mode 100644 index 0000000..020422c --- /dev/null +++ b/patches/gcc-14.2.0/0009-sh-fdpic-pr114641.diff @@ -0,0 +1,12 @@ +--- gcc-11.4.0/gcc/config/sh/sh.cc.orig 2024-04-04 05:52:42.125373614 +0900 ++++ gcc-11.4.0/gcc/config/sh/sh.cc 2024-04-04 22:54:01.875106654 +0900 +@@ -9147,7 +9147,7 @@ + { + /* Weak functions may be NULL which doesn't work with + GOTOFFFUNCDESC because the runtime offset is not known. */ +- if (SYMBOL_REF_WEAK (orig)) ++ if (SYMBOL_REF_WEAK (orig) || (TREE_PUBLIC(SYMBOL_REF_DECL(orig)) && DECL_VISIBILITY (SYMBOL_REF_DECL(orig)) != VISIBILITY_HIDDEN)) + emit_insn (gen_symGOTFUNCDESC2reg (reg, orig)); + else + emit_insn (gen_symGOTOFFFUNCDESC2reg (reg, orig)); + -- 2.45.2 ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions 2024-10-02 15:40 [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions J. Neuschäfer ` (2 preceding siblings ...) 2024-10-02 15:40 ` [musl] [PATCH MCM 3/3] add gcc 14.2 " J. Neuschäfer @ 2024-10-22 21:52 ` Rich Felker 2024-10-23 12:29 ` J. Neuschäfer 3 siblings, 1 reply; 9+ messages in thread From: Rich Felker @ 2024-10-22 21:52 UTC (permalink / raw) To: J. Neuschäfer; +Cc: musl On Wed, Oct 02, 2024 at 05:40:03PM +0200, J. Neuschäfer wrote: > This patchset adds newer GCC versions to MCM. Patches are copied from > GCC 11 and only very lightly modifed (which can be verified by diffing > the patch directories). > > I haven't tested the resulting toolchains, so there might be some bugs > lurking. Could you resend these as attachments instead of lkml-thread-style? Something got messed up with the whitespace and git am is applying them but giving results that have gratuitous whitespace diffs, and I'd rather not track down whether any of that is breaking. Rich ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions 2024-10-22 21:52 ` [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions Rich Felker @ 2024-10-23 12:29 ` J. Neuschäfer 2024-10-24 13:11 ` Rich Felker 0 siblings, 1 reply; 9+ messages in thread From: J. Neuschäfer @ 2024-10-23 12:29 UTC (permalink / raw) To: musl; +Cc: J. Neuschäfer [-- Attachment #1: Type: text/plain, Size: 717 bytes --] On Tue, Oct 22, 2024 at 05:52:22PM -0400, Rich Felker wrote: > On Wed, Oct 02, 2024 at 05:40:03PM +0200, J. Neuschäfer wrote: > > This patchset adds newer GCC versions to MCM. Patches are copied from > > GCC 11 and only very lightly modifed (which can be verified by diffing > > the patch directories). > > > > I haven't tested the resulting toolchains, so there might be some bugs > > lurking. > > Could you resend these as attachments instead of lkml-thread-style? > Something got messed up with the whitespace and git am is applying > them but giving results that have gratuitous whitespace diffs, and I'd > rather not track down whether any of that is breaking. > > Rich Here they are. jn [-- Attachment #2: 0001-add-gcc-12-4-hash-and-patches.eml --] [-- Type: message/rfc822, Size: 27607 bytes --] From: "J. Neuschäfer" <j.neuschaefer@gmx.net> To: musl@lists.openwall.com Cc: "J. Neuschäfer" <j.neuschaefer@gmx.net> Subject: [PATCH MCM 1/3] add gcc 12.4 hash and patches Date: Wed, 23 Oct 2024 14:23:00 +0200 Message-ID: <20241023-gcc-bump-v1-1-0760a1580713@gmx.net> The patches have been copied from 11.4 and slightly adjusted to account for renamed files (.c -> .cc). --- hashes/gcc-12.4.0.tar.xz.sha1 | 1 + patches/gcc-12.4.0/0001-ssp_nonshared.diff | 14 + patches/gcc-12.4.0/0002-posix_memalign.diff | 30 ++ patches/gcc-12.4.0/0003-j2.diff | 346 +++++++++++++++++++++ patches/gcc-12.4.0/0004-static-pie.diff | 92 ++++++ patches/gcc-12.4.0/0005-m68k-sqrt.diff | 20 ++ patches/gcc-12.4.0/0006-cow-libstdc++v3.diff | 8 + patches/gcc-12.4.0/0007-fdpic-unwind.diff | 11 + .../gcc-12.4.0/0008-fdpic-crtstuff-pr114158.diff | 38 +++ patches/gcc-12.4.0/0009-sh-fdpic-pr114641.diff | 12 + 10 files changed, 572 insertions(+) diff --git a/hashes/gcc-12.4.0.tar.xz.sha1 b/hashes/gcc-12.4.0.tar.xz.sha1 new file mode 100644 index 0000000000000000000000000000000000000000..0c9d653ee9c7df0e7454f48d37ba4d99775ec3d1 --- /dev/null +++ b/hashes/gcc-12.4.0.tar.xz.sha1 @@ -0,0 +1 @@ +b373d4ac29bb06ca64d288621906cbf63ab5a1f5 gcc-12.4.0.tar.xz diff --git a/patches/gcc-12.4.0/0001-ssp_nonshared.diff b/patches/gcc-12.4.0/0001-ssp_nonshared.diff new file mode 100644 index 0000000000000000000000000000000000000000..b250f15833de6ae5bd732935f9bd8a924bbd783e --- /dev/null +++ b/patches/gcc-12.4.0/0001-ssp_nonshared.diff @@ -0,0 +1,14 @@ +diff --git a/gcc/gcc.cc b/gcc/gcc.cc +index 7837553958b..3c81c5798d8 100644 +--- a/gcc/gcc.cc ++++ b/gcc/gcc.cc +@@ -980,7 +980,8 @@ proper position among the other output files. */ + #ifndef LINK_SSP_SPEC + #ifdef TARGET_LIBC_PROVIDES_SSP + #define LINK_SSP_SPEC "%{fstack-protector|fstack-protector-all" \ +- "|fstack-protector-strong|fstack-protector-explicit:}" ++ "|fstack-protector-strong|fstack-protector-explicit" \ ++ ":-lssp_nonshared}" + #else + #define LINK_SSP_SPEC "%{fstack-protector|fstack-protector-all" \ + "|fstack-protector-strong|fstack-protector-explicit" \ diff --git a/patches/gcc-12.4.0/0002-posix_memalign.diff b/patches/gcc-12.4.0/0002-posix_memalign.diff new file mode 100644 index 0000000000000000000000000000000000000000..1d80096318d43bac22bb885a56fbbd9dd7fd3a74 --- /dev/null +++ b/patches/gcc-12.4.0/0002-posix_memalign.diff @@ -0,0 +1,30 @@ +diff --git a/gcc/config/i386/pmm_malloc.h b/gcc/config/i386/pmm_malloc.h +index 1b0bfe37852..d7b2b19bb3c 100644 +--- a/gcc/config/i386/pmm_malloc.h ++++ b/gcc/config/i386/pmm_malloc.h +@@ -27,12 +27,13 @@ + #include <stdlib.h> + + /* We can't depend on <stdlib.h> since the prototype of posix_memalign +- may not be visible. */ ++ may not be visible and we can't pollute the namespace either. */ + #ifndef __cplusplus +-extern int posix_memalign (void **, size_t, size_t); ++extern int _mm_posix_memalign (void **, size_t, size_t) + #else +-extern "C" int posix_memalign (void **, size_t, size_t) throw (); ++extern "C" int _mm_posix_memalign (void **, size_t, size_t) throw () + #endif ++__asm__("posix_memalign"); + + static __inline void * + _mm_malloc (size_t __size, size_t __alignment) +@@ -42,7 +43,7 @@ _mm_malloc (size_t __size, size_t __alignment) + return malloc (__size); + if (__alignment == 2 || (sizeof (void *) == 8 && __alignment == 4)) + __alignment = sizeof (void *); +- if (posix_memalign (&__ptr, __alignment, __size) == 0) ++ if (_mm_posix_memalign (&__ptr, __alignment, __size) == 0) + return __ptr; + else + return NULL; diff --git a/patches/gcc-12.4.0/0003-j2.diff b/patches/gcc-12.4.0/0003-j2.diff new file mode 100644 index 0000000000000000000000000000000000000000..c3786456c958668a1e08dd0a2b1e8280ea50fd7a --- /dev/null +++ b/patches/gcc-12.4.0/0003-j2.diff @@ -0,0 +1,346 @@ +diff --git a/gcc/config.gcc b/gcc/config.gcc +index 357b0bed067..528add999f2 100644 +--- a/gcc/config.gcc ++++ b/gcc/config.gcc +@@ -556,7 +556,7 @@ s390*-*-*) + extra_headers="s390intrin.h htmintrin.h htmxlintrin.h vecintrin.h" + ;; + # Note the 'l'; we need to be able to match e.g. "shle" or "shl". +-sh[123456789lbe]*-*-* | sh-*-*) ++sh[123456789lbej]*-*-* | sh-*-*) + cpu_type=sh + extra_options="${extra_options} fused-madd.opt" + extra_objs="${extra_objs} sh_treg_combine.o sh-mem.o sh_optimize_sett_clrt.o" +@@ -3202,18 +3202,18 @@ s390x-ibm-tpf*) + extra_options="${extra_options} s390/tpf.opt" + tmake_file="${tmake_file} s390/t-s390" + ;; +-sh-*-elf* | sh[12346l]*-*-elf* | \ +- sh-*-linux* | sh[2346lbe]*-*-linux* | \ ++sh-*-elf* | sh[12346lj]*-*-elf* | \ ++ sh-*-linux* | sh[2346lbej]*-*-linux* | \ + sh-*-netbsdelf* | shl*-*-netbsdelf*) + tmake_file="${tmake_file} sh/t-sh sh/t-elf" + if test x${with_endian} = x; then + case ${target} in +- sh[1234]*be-*-* | sh[1234]*eb-*-*) with_endian=big ;; ++ sh[j1234]*be-*-* | sh[j1234]*eb-*-*) with_endian=big ;; + shbe-*-* | sheb-*-*) with_endian=big,little ;; + sh[1234]l* | sh[34]*-*-linux*) with_endian=little ;; + shl* | sh*-*-linux* | \ + sh-superh-elf) with_endian=little,big ;; +- sh[1234]*-*-*) with_endian=big ;; ++ sh[j1234]*-*-*) with_endian=big ;; + *) with_endian=big,little ;; + esac + fi +@@ -3280,6 +3280,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + sh2a_nofpu*) sh_cpu_target=sh2a-nofpu ;; + sh2a*) sh_cpu_target=sh2a ;; + sh2e*) sh_cpu_target=sh2e ;; ++ shj2*) sh_cpu_target=shj2;; + sh2*) sh_cpu_target=sh2 ;; + *) sh_cpu_target=sh1 ;; + esac +@@ -3301,7 +3302,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + sh2a-single-only | sh2a-single | sh2a-nofpu | sh2a | \ + sh4a-single-only | sh4a-single | sh4a-nofpu | sh4a | sh4al | \ + sh4-single-only | sh4-single | sh4-nofpu | sh4 | sh4-300 | \ +- sh3e | sh3 | sh2e | sh2 | sh1) ;; ++ sh3e | sh3 | sh2e | sh2 | sh1 | shj2) ;; + "") sh_cpu_default=${sh_cpu_target} ;; + *) echo "with_cpu=$with_cpu not supported"; exit 1 ;; + esac +@@ -3310,9 +3311,9 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + case ${target} in + sh[1234]*) sh_multilibs=${sh_cpu_target} ;; + sh-superh-*) sh_multilibs=m4,m4-single,m4-single-only,m4-nofpu ;; +- sh*-*-linux*) sh_multilibs=m1,m2,m2a,m3e,m4 ;; ++ sh*-*-linux*) sh_multilibs=m1,m2,m2a,m3e,m4,mj2 ;; + sh*-*-netbsd*) sh_multilibs=m3,m3e,m4 ;; +- *) sh_multilibs=m1,m2,m2e,m4,m4-single,m4-single-only,m2a,m2a-single ;; ++ *) sh_multilibs=m1,m2,m2e,m4,m4-single,m4-single-only,m2a,m2a-single,mj2 ;; + esac + if test x$with_fp = xno; then + sh_multilibs="`echo $sh_multilibs|sed -e s/m4/sh4-nofpu/ -e s/,m4-[^,]*//g -e s/,m[23]e// -e s/m2a,m2a-single/m2a-nofpu/ -e s/m5-..m....,//g`" +@@ -3327,7 +3328,8 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + m1 | m2 | m2e | m3 | m3e | \ + m4 | m4-single | m4-single-only | m4-nofpu | m4-300 |\ + m4a | m4a-single | m4a-single-only | m4a-nofpu | m4al | \ +- m2a | m2a-single | m2a-single-only | m2a-nofpu) ++ m2a | m2a-single | m2a-single-only | m2a-nofpu | \ ++ mj2) + # TM_MULTILIB_CONFIG is used by t-sh for the non-endian multilib definition + # It is passed to MULTIILIB_OPTIONS verbatim. + TM_MULTILIB_CONFIG="${TM_MULTILIB_CONFIG}/${sh_multilib}" +@@ -3344,7 +3346,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + done + TM_MULTILIB_CONFIG=`echo $TM_MULTILIB_CONFIG | sed 's:^/::'` + if test x${enable_incomplete_targets} = xyes ; then +- tm_defines="$tm_defines SUPPORT_SH1=1 SUPPORT_SH2E=1 SUPPORT_SH4=1 SUPPORT_SH4_SINGLE=1 SUPPORT_SH2A=1 SUPPORT_SH2A_SINGLE=1" ++ tm_defines="$tm_defines SUPPORT_SH1=1 SUPPORT_SH2E=1 SUPPORT_SH4=1 SUPPORT_SH4_SINGLE=1 SUPPORT_SH2A=1 SUPPORT_SH2A_SINGLE=1 SUPPORT_SHJ2=1" + fi + tm_file="$tm_file ./sysroot-suffix.h" + tmake_file="$tmake_file t-sysroot-suffix" +@@ -5175,6 +5177,8 @@ case "${target}" in + ;; + m4a | m4a-single | m4a-single-only | m4a-nofpu | m4al) + ;; ++ mj2) ++ ;; + *) + echo "Unknown CPU used in --with-cpu=$with_cpu, known values:" 1>&2 + echo "m1 m2 m2e m3 m3e m4 m4-single m4-single-only m4-nofpu" 1>&2 +@@ -5385,7 +5389,7 @@ case ${target} in + tmake_file="${cpu_type}/t-${cpu_type} ${tmake_file}" + ;; + +- sh[123456ble]*-*-* | sh-*-*) ++ sh[123456blej]*-*-* | sh-*-*) + c_target_objs="${c_target_objs} sh-c.o" + cxx_target_objs="${cxx_target_objs} sh-c.o" + ;; +diff --git a/gcc/config/sh/sh.cc b/gcc/config/sh/sh.cc +index 1564109c942..798c1c1c1a3 100644 +--- a/gcc/config/sh/sh.cc ++++ b/gcc/config/sh/sh.cc +@@ -686,6 +686,7 @@ parse_validate_atomic_model_option (const char* str) + model_names[sh_atomic_model::hard_llcs] = "hard-llcs"; + model_names[sh_atomic_model::soft_tcb] = "soft-tcb"; + model_names[sh_atomic_model::soft_imask] = "soft-imask"; ++ model_names[sh_atomic_model::hard_cas] = "hard-cas"; + + const char* model_cdef_names[sh_atomic_model::num_models]; + model_cdef_names[sh_atomic_model::none] = "NONE"; +@@ -693,6 +694,7 @@ parse_validate_atomic_model_option (const char* str) + model_cdef_names[sh_atomic_model::hard_llcs] = "HARD_LLCS"; + model_cdef_names[sh_atomic_model::soft_tcb] = "SOFT_TCB"; + model_cdef_names[sh_atomic_model::soft_imask] = "SOFT_IMASK"; ++ model_cdef_names[sh_atomic_model::hard_cas] = "HARD_CAS"; + + sh_atomic_model ret; + ret.type = sh_atomic_model::none; +@@ -771,6 +773,9 @@ got_mode_name:; + if (ret.type == sh_atomic_model::soft_imask && TARGET_USERMODE) + err_ret ("cannot use atomic model %s in user mode", ret.name); + ++ if (ret.type == sh_atomic_model::hard_cas && !TARGET_SHJ2) ++ err_ret ("atomic model %s is only available J2 targets", ret.name); ++ + return ret; + + #undef err_ret +@@ -827,6 +832,8 @@ sh_option_override (void) + sh_cpu = PROCESSOR_SH2E; + if (TARGET_SH2A) + sh_cpu = PROCESSOR_SH2A; ++ if (TARGET_SHJ2) ++ sh_cpu = PROCESSOR_SHJ2; + if (TARGET_SH3) + sh_cpu = PROCESSOR_SH3; + if (TARGET_SH3E) +diff --git a/gcc/config/sh/sh.h b/gcc/config/sh/sh.h +index d2280e2ffe6..3a54a896721 100644 +--- a/gcc/config/sh/sh.h ++++ b/gcc/config/sh/sh.h +@@ -85,6 +85,7 @@ extern int code_for_indirect_jump_scratch; + #define SUPPORT_SH4_SINGLE 1 + #define SUPPORT_SH2A 1 + #define SUPPORT_SH2A_SINGLE 1 ++#define SUPPORT_SHJ2 1 + #endif + + #define TARGET_DIVIDE_CALL_DIV1 (sh_div_strategy == SH_DIV_CALL_DIV1) +@@ -117,6 +118,7 @@ extern int code_for_indirect_jump_scratch; + #define SELECT_SH4A_SINGLE_ONLY (MASK_SH4A | SELECT_SH4_SINGLE_ONLY) + #define SELECT_SH4A (MASK_SH4A | SELECT_SH4) + #define SELECT_SH4A_SINGLE (MASK_SH4A | SELECT_SH4_SINGLE) ++#define SELECT_SHJ2 (MASK_SHJ2 | SELECT_SH2) + + #if SUPPORT_SH1 + #define SUPPORT_SH2 1 +@@ -124,6 +126,7 @@ extern int code_for_indirect_jump_scratch; + #if SUPPORT_SH2 + #define SUPPORT_SH3 1 + #define SUPPORT_SH2A_NOFPU 1 ++#define SUPPORT_SHJ2 1 + #endif + #if SUPPORT_SH3 + #define SUPPORT_SH4_NOFPU 1 +@@ -156,7 +159,7 @@ extern int code_for_indirect_jump_scratch; + #define MASK_ARCH (MASK_SH1 | MASK_SH2 | MASK_SH3 | MASK_SH_E | MASK_SH4 \ + | MASK_HARD_SH2A | MASK_HARD_SH2A_DOUBLE | MASK_SH4A \ + | MASK_HARD_SH4 | MASK_FPU_SINGLE \ +- | MASK_FPU_SINGLE_ONLY) ++ | MASK_FPU_SINGLE_ONLY | MASK_SHJ2) + + /* This defaults us to big-endian. */ + #ifndef TARGET_ENDIAN_DEFAULT +@@ -231,7 +234,8 @@ extern int code_for_indirect_jump_scratch; + %{m2a-single:--isa=sh2a} \ + %{m2a-single-only:--isa=sh2a} \ + %{m2a-nofpu:--isa=sh2a-nofpu} \ +-%{m4al:-dsp}" ++%{m4al:-dsp} \ ++%{mj2:-isa=j2}" + + #define ASM_SPEC SH_ASM_SPEC + +@@ -347,6 +351,7 @@ struct sh_atomic_model + hard_llcs, + soft_tcb, + soft_imask, ++ hard_cas, + + num_models + }; +@@ -390,6 +395,9 @@ extern const sh_atomic_model& selected_atomic_model (void); + #define TARGET_ATOMIC_SOFT_IMASK \ + (selected_atomic_model ().type == sh_atomic_model::soft_imask) + ++#define TARGET_ATOMIC_HARD_CAS \ ++ (selected_atomic_model ().type == sh_atomic_model::hard_cas) ++ + #endif // __cplusplus + + #define SUBTARGET_OVERRIDE_OPTIONS (void) 0 +@@ -1484,7 +1492,7 @@ extern bool current_function_interrupt; + + /* Nonzero if the target supports dynamic shift instructions + like shad and shld. */ +-#define TARGET_DYNSHIFT (TARGET_SH3 || TARGET_SH2A) ++#define TARGET_DYNSHIFT (TARGET_SH3 || TARGET_SH2A || TARGET_SHJ2) + + /* The cost of using the dynamic shift insns (shad, shld) are the same + if they are available. If they are not available a library function will +@@ -1747,6 +1755,7 @@ enum processor_type { + PROCESSOR_SH2, + PROCESSOR_SH2E, + PROCESSOR_SH2A, ++ PROCESSOR_SHJ2, + PROCESSOR_SH3, + PROCESSOR_SH3E, + PROCESSOR_SH4, +diff --git a/gcc/config/sh/sh.opt b/gcc/config/sh/sh.opt +index b4755a812f3..0989a1c18da 100644 +--- a/gcc/config/sh/sh.opt ++++ b/gcc/config/sh/sh.opt +@@ -65,6 +65,10 @@ m2e + Target RejectNegative Condition(SUPPORT_SH2E) + Generate SH2e code. + ++mj2 ++Target RejectNegative Mask(SHJ2) Condition(SUPPORT_SHJ2) ++Generate J2 code. ++ + m3 + Target RejectNegative Mask(SH3) Condition(SUPPORT_SH3) + Generate SH3 code. +diff --git a/gcc/config/sh/sync.md b/gcc/config/sh/sync.md +index 2b43f8edb86..118fc5d06db 100644 +--- a/gcc/config/sh/sync.md ++++ b/gcc/config/sh/sync.md +@@ -240,6 +240,9 @@ + || (TARGET_SH4A && <MODE>mode == SImode && !TARGET_ATOMIC_STRICT)) + atomic_insn = gen_atomic_compare_and_swap<mode>_hard (old_val, mem, + exp_val, new_val); ++ else if (TARGET_ATOMIC_HARD_CAS && <MODE>mode == SImode) ++ atomic_insn = gen_atomic_compare_and_swap<mode>_cas (old_val, mem, ++ exp_val, new_val); + else if (TARGET_ATOMIC_SOFT_GUSA) + atomic_insn = gen_atomic_compare_and_swap<mode>_soft_gusa (old_val, mem, + exp_val, new_val); +@@ -306,6 +309,57 @@ + } + [(set_attr "length" "14")]) + ++(define_expand "atomic_compare_and_swapsi_cas" ++ [(set (match_operand:SI 0 "register_operand" "=r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "register_operand" "r") ++ (match_operand:SI 3 "register_operand" "r")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{ ++ rtx mem = gen_rtx_REG (SImode, 0); ++ emit_move_insn (mem, force_reg (SImode, XEXP (operands[1], 0))); ++ emit_insn (gen_shj2_cas (operands[0], mem, operands[2], operands[3])); ++ DONE; ++}) ++ ++(define_insn "shj2_cas" ++ [(set (match_operand:SI 0 "register_operand" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "register_operand" "=r") ++ (match_operand:SI 2 "register_operand" "r") ++ (match_operand:SI 3 "register_operand" "0")] ++ UNSPECV_CMPXCHG_1)) ++ (set (reg:SI T_REG) ++ (unspec_volatile:SI [(const_int 0)] UNSPECV_CMPXCHG_3))] ++ "TARGET_ATOMIC_HARD_CAS" ++ "cas.l %2,%0,@%1" ++ [(set_attr "length" "2")] ++) ++ ++(define_expand "atomic_compare_and_swapqi_cas" ++ [(set (match_operand:SI 0 "arith_reg_dest" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "arith_operand" "rI08") ++ (match_operand:SI 3 "arith_operand" "rI08")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{FAIL;} ++) ++ ++(define_expand "atomic_compare_and_swaphi_cas" ++ [(set (match_operand:SI 0 "arith_reg_dest" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "arith_operand" "rI08") ++ (match_operand:SI 3 "arith_operand" "rI08")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{FAIL;} ++) ++ + ;; The QIHImode llcs patterns modify the address register of the memory + ;; operand. In order to express that, we have to open code the memory + ;; operand. Initially the insn is expanded like every other atomic insn +diff --git a/gcc/config/sh/t-sh b/gcc/config/sh/t-sh +index 888f8ff7f25..29fd6ae45fd 100644 +--- a/gcc/config/sh/t-sh ++++ b/gcc/config/sh/t-sh +@@ -50,7 +50,8 @@ MULTILIB_MATCHES = $(shell \ + m2e,m3e,m4-single-only,m4-100-single-only,m4-200-single-only,m4-300-single-only,m4a-single-only \ + m2a-single,m2a-single-only \ + m4-single,m4-100-single,m4-200-single,m4-300-single,m4a-single \ +- m4,m4-100,m4-200,m4-300,m4a; do \ ++ m4,m4-100,m4-200,m4-300,m4a \ ++ mj2; do \ + subst= ; \ + for lib in `echo $$abi|tr , ' '` ; do \ + if test "`echo $$multilibs|sed s/$$lib//`" != "$$multilibs"; then \ +@@ -63,9 +64,9 @@ MULTILIB_MATCHES = $(shell \ + + # SH1 and SH2A support big endian only. + ifeq ($(DEFAULT_ENDIAN),ml) +-MULTILIB_EXCEPTIONS = m1 ml/m1 m2a* ml/m2a* $(TM_MULTILIB_EXCEPTIONS_CONFIG) ++MULTILIB_EXCEPTIONS = m1 ml/m1 m2a* ml/m2a* ml/mj2 $(TM_MULTILIB_EXCEPTIONS_CONFIG) + else +-MULTILIB_EXCEPTIONS = ml/m1 ml/m2a* $(TM_MULTILIB_EXCEPTIONS_CONFIG) ++MULTILIB_EXCEPTIONS = ml/m1 ml/m2a* ml/mj2 $(TM_MULTILIB_EXCEPTIONS_CONFIG) + endif + + MULTILIB_OSDIRNAMES = \ +@@ -87,7 +88,8 @@ MULTILIB_OSDIRNAMES = \ + m4a-single-only=!m4a-single-only $(OTHER_ENDIAN)/m4a-single-only=!$(OTHER_ENDIAN)/m4a-single-only \ + m4a-single=!m4a-single $(OTHER_ENDIAN)/m4a-single=!$(OTHER_ENDIAN)/m4a-single \ + m4a=!m4a $(OTHER_ENDIAN)/m4a=!$(OTHER_ENDIAN)/m4a \ +- m4al=!m4al $(OTHER_ENDIAN)/m4al=!$(OTHER_ENDIAN)/m4al ++ m4al=!m4al $(OTHER_ENDIAN)/m4al=!$(OTHER_ENDIAN)/m4al \ ++ mj2=!j2 + + $(out_object_file): gt-sh.h + gt-sh.h : s-gtype ; @true diff --git a/patches/gcc-12.4.0/0004-static-pie.diff b/patches/gcc-12.4.0/0004-static-pie.diff new file mode 100644 index 0000000000000000000000000000000000000000..a85a3d738ce88537da7a31f4076d3ad9c0e59ae9 --- /dev/null +++ b/patches/gcc-12.4.0/0004-static-pie.diff @@ -0,0 +1,92 @@ +diff --git a/gcc/common.opt b/gcc/common.opt +index a75b44ee47e..7c564818b49 100644 +--- a/gcc/common.opt ++++ b/gcc/common.opt +@@ -3473,11 +3473,11 @@ Driver + + no-pie + Driver RejectNegative Negative(shared) +-Don't create a dynamically linked position independent executable. ++Don't create a position independent executable. + + pie + Driver RejectNegative Negative(no-pie) +-Create a dynamically linked position independent executable. ++Create a position independent executable. + + static-pie + Driver RejectNegative Negative(pie) +diff --git a/gcc/config/gnu-user.h b/gcc/config/gnu-user.h +index 5ebbf42a13d..bb907d8e89a 100644 +--- a/gcc/config/gnu-user.h ++++ b/gcc/config/gnu-user.h +@@ -51,13 +51,12 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + #define GNU_USER_TARGET_STARTFILE_SPEC \ + "%{shared:; \ + pg|p|profile:%{static-pie:grcrt1.o%s;:gcrt1.o%s}; \ +- static:crt1.o%s; \ +- static-pie:rcrt1.o%s; \ ++ static|static-pie:%{" PIE_SPEC ":rcrt1.o%s;:crt1.o%s}; \ + " PIE_SPEC ":Scrt1.o%s; \ + :crt1.o%s} " \ + GNU_USER_TARGET_CRTI " \ +- %{static:crtbeginT.o%s; \ +- shared|static-pie|" PIE_SPEC ":crtbeginS.o%s; \ ++ %{shared|" PIE_SPEC ":crtbeginS.o%s; \ ++ static:crtbeginT.o%s; \ + :crtbegin.o%s} \ + %{fvtable-verify=none:%s; \ + fvtable-verify=preinit:vtv_start_preinit.o%s; \ +@@ -73,11 +72,11 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + GNU userspace "finalizer" file, `crtn.o'. */ + + #define GNU_USER_TARGET_ENDFILE_SPEC \ +- "%{!static:%{fvtable-verify=none:%s; \ ++ "%{static|static-pie:; \ ++ fvtable-verify=none:%s; \ + fvtable-verify=preinit:vtv_end_preinit.o%s; \ +- fvtable-verify=std:vtv_end.o%s}} \ +- %{static:crtend.o%s; \ +- shared|static-pie|" PIE_SPEC ":crtendS.o%s; \ ++ fvtable-verify=std:vtv_end.o%s} \ ++ %{shared|" PIE_SPEC ":crtendS.o%s; \ + :crtend.o%s} " \ + GNU_USER_TARGET_CRTN " " \ + CRTOFFLOADEND +@@ -106,7 +105,7 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + #define LIB_SPEC GNU_USER_TARGET_LIB_SPEC + + #if defined(HAVE_LD_EH_FRAME_HDR) +-#define LINK_EH_SPEC "%{!static|static-pie:--eh-frame-hdr} " ++#define LINK_EH_SPEC "%{!static|" PIE_SPEC ":--eh-frame-hdr} " + #endif + + #define GNU_USER_TARGET_LINK_GCC_C_SEQUENCE_SPEC \ +diff --git a/gcc/gcc.cc b/gcc/gcc.cc +index 3c81c5798d8..cd96eac5d12 100644 +--- a/gcc/gcc.cc ++++ b/gcc/gcc.cc +@@ -1010,7 +1010,7 @@ proper position among the other output files. */ + #define NO_FPIE_AND_FPIC_SPEC NO_FPIE_SPEC "|" NO_FPIC_SPEC + #define FPIE_OR_FPIC_SPEC NO_FPIE_AND_FPIC_SPEC ":;" + #else +-#define PIE_SPEC "pie" ++#define PIE_SPEC "pie|static-pie" + #define FPIE1_SPEC "fpie" + #define NO_FPIE1_SPEC FPIE1_SPEC ":;" + #define FPIE2_SPEC "fPIE" +@@ -1034,12 +1034,12 @@ proper position among the other output files. */ + #ifndef LINK_PIE_SPEC + #ifdef HAVE_LD_PIE + #ifndef LD_PIE_SPEC +-#define LD_PIE_SPEC "-pie" ++#define LD_PIE_SPEC "-pie %{static|static-pie:--no-dynamic-linker -z text -Bsymbolic}" + #endif + #else + #define LD_PIE_SPEC "" + #endif +-#define LINK_PIE_SPEC "%{static|shared|r:;" PIE_SPEC ":" LD_PIE_SPEC "} " ++#define LINK_PIE_SPEC "%{shared|r:;" PIE_SPEC ":" LD_PIE_SPEC "} " + #endif + + #ifndef LINK_BUILDID_SPEC diff --git a/patches/gcc-12.4.0/0005-m68k-sqrt.diff b/patches/gcc-12.4.0/0005-m68k-sqrt.diff new file mode 100644 index 0000000000000000000000000000000000000000..652efb15fc38c9198626ce8744e3ea932820ce7e --- /dev/null +++ b/patches/gcc-12.4.0/0005-m68k-sqrt.diff @@ -0,0 +1,20 @@ +diff --git a/gcc/config/m68k/m68k.md b/gcc/config/m68k/m68k.md +index 59a456cd496..dbfddea41bd 100644 +--- a/gcc/config/m68k/m68k.md ++++ b/gcc/config/m68k/m68k.md +@@ -4174,13 +4174,13 @@ + (define_expand "sqrt<mode>2" + [(set (match_operand:FP 0 "nonimmediate_operand" "") + (sqrt:FP (match_operand:FP 1 "general_operand" "")))] +- "TARGET_HARD_FLOAT" ++ "(TARGET_68881 && TARGET_68040) || TARGET_COLDFIRE_FPU" + "") + + (define_insn "sqrt<mode>2_68881" + [(set (match_operand:FP 0 "nonimmediate_operand" "=f") + (sqrt:FP (match_operand:FP 1 "general_operand" "f<FP:dreg>m")))] +- "TARGET_68881" ++ "TARGET_68881 && TARGET_68040" + { + if (FP_REG_P (operands[1])) + return "f<FP:round>sqrt%.x %1,%0"; diff --git a/patches/gcc-12.4.0/0006-cow-libstdc++v3.diff b/patches/gcc-12.4.0/0006-cow-libstdc++v3.diff new file mode 100644 index 0000000000000000000000000000000000000000..b49b22c247dd8cb7459143dfdc0e2c32548e5a1b --- /dev/null +++ b/patches/gcc-12.4.0/0006-cow-libstdc++v3.diff @@ -0,0 +1,8 @@ +--- a/libstdc++-v3/ChangeLog 2024-03-13 18:04:21.050801063 -0400 ++++ b/libstdc++-v3/ChangeLog 2024-03-13 18:04:25.665836804 -0400 +@@ -4068,4 +4068,4 @@ + + Copying and distribution of this file, with or without modification, + are permitted in any medium without royalty provided the copyright +-notice and this notice are preserved. ++notice and this notice are preserved. diff --git a/patches/gcc-12.4.0/0007-fdpic-unwind.diff b/patches/gcc-12.4.0/0007-fdpic-unwind.diff new file mode 100644 index 0000000000000000000000000000000000000000..61d3479514991fc2823c0d5f1c76dca48761a114 --- /dev/null +++ b/patches/gcc-12.4.0/0007-fdpic-unwind.diff @@ -0,0 +1,11 @@ +--- a/libgcc/unwind-pe.h 2024-03-14 05:59:53.754073149 +0900 ++++ b/libgcc/unwind-pe.h 2024-03-14 06:00:41.226074492 +0900 +@@ -262,7 +262,7 @@ + + if (result != 0) + { +-#if __FDPIC__ ++#if __FDPIC__ && __arm__ + /* FDPIC relative addresses imply taking the GOT address + into account. */ + if ((encoding & DW_EH_PE_pcrel) && (encoding & DW_EH_PE_indirect)) diff --git a/patches/gcc-12.4.0/0008-fdpic-crtstuff-pr114158.diff b/patches/gcc-12.4.0/0008-fdpic-crtstuff-pr114158.diff new file mode 100644 index 0000000000000000000000000000000000000000..3986eed9be2275c5b942d47e9760de674df20f39 --- /dev/null +++ b/patches/gcc-12.4.0/0008-fdpic-crtstuff-pr114158.diff @@ -0,0 +1,38 @@ +--- a/libgcc/crtstuff.c 2023-05-29 17:46:32.000000000 +0900 ++++ b/libgcc/crtstuff.c 2024-03-14 06:03:42.398079615 +0900 +@@ -441,17 +441,9 @@ + #ifdef FINI_SECTION_ASM_OP + CRT_CALL_STATIC_FUNCTION (FINI_SECTION_ASM_OP, __do_global_dtors_aux) + #elif defined (FINI_ARRAY_SECTION_ASM_OP) +-#if defined(__FDPIC__) +-__asm__("\t.equ\t__do_global_dtors_aux_alias, __do_global_dtors_aux\n"); +-extern char __do_global_dtors_aux_alias; +-static void *__do_global_dtors_aux_fini_array_entry[] +-__attribute__ ((__used__, section(".fini_array"), aligned(sizeof(void *)))) +- = { &__do_global_dtors_aux_alias }; +-#else /* defined(__FDPIC__) */ + static func_ptr __do_global_dtors_aux_fini_array_entry[] + __attribute__ ((__used__, section(".fini_array"), + aligned(__alignof__(func_ptr)))) = { __do_global_dtors_aux }; +-#endif /* defined(__FDPIC__) */ + #else /* !FINI_SECTION_ASM_OP && !FINI_ARRAY_SECTION_ASM_OP */ + static void __attribute__((used)) + __do_global_dtors_aux_1 (void) +@@ -494,17 +486,9 @@ + #ifdef __LIBGCC_INIT_SECTION_ASM_OP__ + CRT_CALL_STATIC_FUNCTION (__LIBGCC_INIT_SECTION_ASM_OP__, frame_dummy) + #else /* defined(__LIBGCC_INIT_SECTION_ASM_OP__) */ +-#if defined(__FDPIC__) +-__asm__("\t.equ\t__frame_dummy_alias, frame_dummy\n"); +-extern char __frame_dummy_alias; +-static void *__frame_dummy_init_array_entry[] +-__attribute__ ((__used__, section(".init_array"), aligned(sizeof(void *)))) +- = { &__frame_dummy_alias }; +-#else /* defined(__FDPIC__) */ + static func_ptr __frame_dummy_init_array_entry[] + __attribute__ ((__used__, section(".init_array"), + aligned(__alignof__(func_ptr)))) = { frame_dummy }; +-#endif /* defined(__FDPIC__) */ + #endif /* !defined(__LIBGCC_INIT_SECTION_ASM_OP__) */ + #endif /* USE_EH_FRAME_REGISTRY || USE_TM_CLONE_REGISTRY */ + diff --git a/patches/gcc-12.4.0/0009-sh-fdpic-pr114641.diff b/patches/gcc-12.4.0/0009-sh-fdpic-pr114641.diff new file mode 100644 index 0000000000000000000000000000000000000000..020422c30e8d16052967ca3ec9375b9232f39932 --- /dev/null +++ b/patches/gcc-12.4.0/0009-sh-fdpic-pr114641.diff @@ -0,0 +1,12 @@ +--- gcc-11.4.0/gcc/config/sh/sh.cc.orig 2024-04-04 05:52:42.125373614 +0900 ++++ gcc-11.4.0/gcc/config/sh/sh.cc 2024-04-04 22:54:01.875106654 +0900 +@@ -9147,7 +9147,7 @@ + { + /* Weak functions may be NULL which doesn't work with + GOTOFFFUNCDESC because the runtime offset is not known. */ +- if (SYMBOL_REF_WEAK (orig)) ++ if (SYMBOL_REF_WEAK (orig) || (TREE_PUBLIC(SYMBOL_REF_DECL(orig)) && DECL_VISIBILITY (SYMBOL_REF_DECL(orig)) != VISIBILITY_HIDDEN)) + emit_insn (gen_symGOTFUNCDESC2reg (reg, orig)); + else + emit_insn (gen_symGOTOFFFUNCDESC2reg (reg, orig)); + -- 2.45.2 [-- Attachment #3: 0002-add-gcc-13-3-hash-and-patches.eml --] [-- Type: message/rfc822, Size: 27544 bytes --] From: "J. Neuschäfer" <j.neuschaefer@gmx.net> To: musl@lists.openwall.com Cc: "J. Neuschäfer" <j.neuschaefer@gmx.net> Subject: [PATCH MCM 2/3] add gcc 13.3 hash and patches Date: Wed, 23 Oct 2024 14:23:01 +0200 Message-ID: <20241023-gcc-bump-v1-2-0760a1580713@gmx.net> The patches have been copied from 12.4. --- hashes/gcc-13.3.0.tar.xz.sha1 | 1 + patches/gcc-13.3.0/0001-ssp_nonshared.diff | 14 + patches/gcc-13.3.0/0002-posix_memalign.diff | 30 ++ patches/gcc-13.3.0/0003-j2.diff | 346 +++++++++++++++++++++ patches/gcc-13.3.0/0004-static-pie.diff | 92 ++++++ patches/gcc-13.3.0/0005-m68k-sqrt.diff | 20 ++ patches/gcc-13.3.0/0006-cow-libstdc++v3.diff | 8 + patches/gcc-13.3.0/0007-fdpic-unwind.diff | 11 + .../gcc-13.3.0/0008-fdpic-crtstuff-pr114158.diff | 38 +++ patches/gcc-13.3.0/0009-sh-fdpic-pr114641.diff | 12 + 10 files changed, 572 insertions(+) diff --git a/hashes/gcc-13.3.0.tar.xz.sha1 b/hashes/gcc-13.3.0.tar.xz.sha1 new file mode 100644 index 0000000000000000000000000000000000000000..4dc0f167a3f1b401bff30cd64d1fe0745c3da0cf --- /dev/null +++ b/hashes/gcc-13.3.0.tar.xz.sha1 @@ -0,0 +1 @@ +6501872415823c95d48be28853ce3ebd6c1040c4 gcc-13.3.0.tar.xz diff --git a/patches/gcc-13.3.0/0001-ssp_nonshared.diff b/patches/gcc-13.3.0/0001-ssp_nonshared.diff new file mode 100644 index 0000000000000000000000000000000000000000..b250f15833de6ae5bd732935f9bd8a924bbd783e --- /dev/null +++ b/patches/gcc-13.3.0/0001-ssp_nonshared.diff @@ -0,0 +1,14 @@ +diff --git a/gcc/gcc.cc b/gcc/gcc.cc +index 7837553958b..3c81c5798d8 100644 +--- a/gcc/gcc.cc ++++ b/gcc/gcc.cc +@@ -980,7 +980,8 @@ proper position among the other output files. */ + #ifndef LINK_SSP_SPEC + #ifdef TARGET_LIBC_PROVIDES_SSP + #define LINK_SSP_SPEC "%{fstack-protector|fstack-protector-all" \ +- "|fstack-protector-strong|fstack-protector-explicit:}" ++ "|fstack-protector-strong|fstack-protector-explicit" \ ++ ":-lssp_nonshared}" + #else + #define LINK_SSP_SPEC "%{fstack-protector|fstack-protector-all" \ + "|fstack-protector-strong|fstack-protector-explicit" \ diff --git a/patches/gcc-13.3.0/0002-posix_memalign.diff b/patches/gcc-13.3.0/0002-posix_memalign.diff new file mode 100644 index 0000000000000000000000000000000000000000..1d80096318d43bac22bb885a56fbbd9dd7fd3a74 --- /dev/null +++ b/patches/gcc-13.3.0/0002-posix_memalign.diff @@ -0,0 +1,30 @@ +diff --git a/gcc/config/i386/pmm_malloc.h b/gcc/config/i386/pmm_malloc.h +index 1b0bfe37852..d7b2b19bb3c 100644 +--- a/gcc/config/i386/pmm_malloc.h ++++ b/gcc/config/i386/pmm_malloc.h +@@ -27,12 +27,13 @@ + #include <stdlib.h> + + /* We can't depend on <stdlib.h> since the prototype of posix_memalign +- may not be visible. */ ++ may not be visible and we can't pollute the namespace either. */ + #ifndef __cplusplus +-extern int posix_memalign (void **, size_t, size_t); ++extern int _mm_posix_memalign (void **, size_t, size_t) + #else +-extern "C" int posix_memalign (void **, size_t, size_t) throw (); ++extern "C" int _mm_posix_memalign (void **, size_t, size_t) throw () + #endif ++__asm__("posix_memalign"); + + static __inline void * + _mm_malloc (size_t __size, size_t __alignment) +@@ -42,7 +43,7 @@ _mm_malloc (size_t __size, size_t __alignment) + return malloc (__size); + if (__alignment == 2 || (sizeof (void *) == 8 && __alignment == 4)) + __alignment = sizeof (void *); +- if (posix_memalign (&__ptr, __alignment, __size) == 0) ++ if (_mm_posix_memalign (&__ptr, __alignment, __size) == 0) + return __ptr; + else + return NULL; diff --git a/patches/gcc-13.3.0/0003-j2.diff b/patches/gcc-13.3.0/0003-j2.diff new file mode 100644 index 0000000000000000000000000000000000000000..c3786456c958668a1e08dd0a2b1e8280ea50fd7a --- /dev/null +++ b/patches/gcc-13.3.0/0003-j2.diff @@ -0,0 +1,346 @@ +diff --git a/gcc/config.gcc b/gcc/config.gcc +index 357b0bed067..528add999f2 100644 +--- a/gcc/config.gcc ++++ b/gcc/config.gcc +@@ -556,7 +556,7 @@ s390*-*-*) + extra_headers="s390intrin.h htmintrin.h htmxlintrin.h vecintrin.h" + ;; + # Note the 'l'; we need to be able to match e.g. "shle" or "shl". +-sh[123456789lbe]*-*-* | sh-*-*) ++sh[123456789lbej]*-*-* | sh-*-*) + cpu_type=sh + extra_options="${extra_options} fused-madd.opt" + extra_objs="${extra_objs} sh_treg_combine.o sh-mem.o sh_optimize_sett_clrt.o" +@@ -3202,18 +3202,18 @@ s390x-ibm-tpf*) + extra_options="${extra_options} s390/tpf.opt" + tmake_file="${tmake_file} s390/t-s390" + ;; +-sh-*-elf* | sh[12346l]*-*-elf* | \ +- sh-*-linux* | sh[2346lbe]*-*-linux* | \ ++sh-*-elf* | sh[12346lj]*-*-elf* | \ ++ sh-*-linux* | sh[2346lbej]*-*-linux* | \ + sh-*-netbsdelf* | shl*-*-netbsdelf*) + tmake_file="${tmake_file} sh/t-sh sh/t-elf" + if test x${with_endian} = x; then + case ${target} in +- sh[1234]*be-*-* | sh[1234]*eb-*-*) with_endian=big ;; ++ sh[j1234]*be-*-* | sh[j1234]*eb-*-*) with_endian=big ;; + shbe-*-* | sheb-*-*) with_endian=big,little ;; + sh[1234]l* | sh[34]*-*-linux*) with_endian=little ;; + shl* | sh*-*-linux* | \ + sh-superh-elf) with_endian=little,big ;; +- sh[1234]*-*-*) with_endian=big ;; ++ sh[j1234]*-*-*) with_endian=big ;; + *) with_endian=big,little ;; + esac + fi +@@ -3280,6 +3280,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + sh2a_nofpu*) sh_cpu_target=sh2a-nofpu ;; + sh2a*) sh_cpu_target=sh2a ;; + sh2e*) sh_cpu_target=sh2e ;; ++ shj2*) sh_cpu_target=shj2;; + sh2*) sh_cpu_target=sh2 ;; + *) sh_cpu_target=sh1 ;; + esac +@@ -3301,7 +3302,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + sh2a-single-only | sh2a-single | sh2a-nofpu | sh2a | \ + sh4a-single-only | sh4a-single | sh4a-nofpu | sh4a | sh4al | \ + sh4-single-only | sh4-single | sh4-nofpu | sh4 | sh4-300 | \ +- sh3e | sh3 | sh2e | sh2 | sh1) ;; ++ sh3e | sh3 | sh2e | sh2 | sh1 | shj2) ;; + "") sh_cpu_default=${sh_cpu_target} ;; + *) echo "with_cpu=$with_cpu not supported"; exit 1 ;; + esac +@@ -3310,9 +3311,9 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + case ${target} in + sh[1234]*) sh_multilibs=${sh_cpu_target} ;; + sh-superh-*) sh_multilibs=m4,m4-single,m4-single-only,m4-nofpu ;; +- sh*-*-linux*) sh_multilibs=m1,m2,m2a,m3e,m4 ;; ++ sh*-*-linux*) sh_multilibs=m1,m2,m2a,m3e,m4,mj2 ;; + sh*-*-netbsd*) sh_multilibs=m3,m3e,m4 ;; +- *) sh_multilibs=m1,m2,m2e,m4,m4-single,m4-single-only,m2a,m2a-single ;; ++ *) sh_multilibs=m1,m2,m2e,m4,m4-single,m4-single-only,m2a,m2a-single,mj2 ;; + esac + if test x$with_fp = xno; then + sh_multilibs="`echo $sh_multilibs|sed -e s/m4/sh4-nofpu/ -e s/,m4-[^,]*//g -e s/,m[23]e// -e s/m2a,m2a-single/m2a-nofpu/ -e s/m5-..m....,//g`" +@@ -3327,7 +3328,8 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + m1 | m2 | m2e | m3 | m3e | \ + m4 | m4-single | m4-single-only | m4-nofpu | m4-300 |\ + m4a | m4a-single | m4a-single-only | m4a-nofpu | m4al | \ +- m2a | m2a-single | m2a-single-only | m2a-nofpu) ++ m2a | m2a-single | m2a-single-only | m2a-nofpu | \ ++ mj2) + # TM_MULTILIB_CONFIG is used by t-sh for the non-endian multilib definition + # It is passed to MULTIILIB_OPTIONS verbatim. + TM_MULTILIB_CONFIG="${TM_MULTILIB_CONFIG}/${sh_multilib}" +@@ -3344,7 +3346,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + done + TM_MULTILIB_CONFIG=`echo $TM_MULTILIB_CONFIG | sed 's:^/::'` + if test x${enable_incomplete_targets} = xyes ; then +- tm_defines="$tm_defines SUPPORT_SH1=1 SUPPORT_SH2E=1 SUPPORT_SH4=1 SUPPORT_SH4_SINGLE=1 SUPPORT_SH2A=1 SUPPORT_SH2A_SINGLE=1" ++ tm_defines="$tm_defines SUPPORT_SH1=1 SUPPORT_SH2E=1 SUPPORT_SH4=1 SUPPORT_SH4_SINGLE=1 SUPPORT_SH2A=1 SUPPORT_SH2A_SINGLE=1 SUPPORT_SHJ2=1" + fi + tm_file="$tm_file ./sysroot-suffix.h" + tmake_file="$tmake_file t-sysroot-suffix" +@@ -5175,6 +5177,8 @@ case "${target}" in + ;; + m4a | m4a-single | m4a-single-only | m4a-nofpu | m4al) + ;; ++ mj2) ++ ;; + *) + echo "Unknown CPU used in --with-cpu=$with_cpu, known values:" 1>&2 + echo "m1 m2 m2e m3 m3e m4 m4-single m4-single-only m4-nofpu" 1>&2 +@@ -5385,7 +5389,7 @@ case ${target} in + tmake_file="${cpu_type}/t-${cpu_type} ${tmake_file}" + ;; + +- sh[123456ble]*-*-* | sh-*-*) ++ sh[123456blej]*-*-* | sh-*-*) + c_target_objs="${c_target_objs} sh-c.o" + cxx_target_objs="${cxx_target_objs} sh-c.o" + ;; +diff --git a/gcc/config/sh/sh.cc b/gcc/config/sh/sh.cc +index 1564109c942..798c1c1c1a3 100644 +--- a/gcc/config/sh/sh.cc ++++ b/gcc/config/sh/sh.cc +@@ -686,6 +686,7 @@ parse_validate_atomic_model_option (const char* str) + model_names[sh_atomic_model::hard_llcs] = "hard-llcs"; + model_names[sh_atomic_model::soft_tcb] = "soft-tcb"; + model_names[sh_atomic_model::soft_imask] = "soft-imask"; ++ model_names[sh_atomic_model::hard_cas] = "hard-cas"; + + const char* model_cdef_names[sh_atomic_model::num_models]; + model_cdef_names[sh_atomic_model::none] = "NONE"; +@@ -693,6 +694,7 @@ parse_validate_atomic_model_option (const char* str) + model_cdef_names[sh_atomic_model::hard_llcs] = "HARD_LLCS"; + model_cdef_names[sh_atomic_model::soft_tcb] = "SOFT_TCB"; + model_cdef_names[sh_atomic_model::soft_imask] = "SOFT_IMASK"; ++ model_cdef_names[sh_atomic_model::hard_cas] = "HARD_CAS"; + + sh_atomic_model ret; + ret.type = sh_atomic_model::none; +@@ -771,6 +773,9 @@ got_mode_name:; + if (ret.type == sh_atomic_model::soft_imask && TARGET_USERMODE) + err_ret ("cannot use atomic model %s in user mode", ret.name); + ++ if (ret.type == sh_atomic_model::hard_cas && !TARGET_SHJ2) ++ err_ret ("atomic model %s is only available J2 targets", ret.name); ++ + return ret; + + #undef err_ret +@@ -827,6 +832,8 @@ sh_option_override (void) + sh_cpu = PROCESSOR_SH2E; + if (TARGET_SH2A) + sh_cpu = PROCESSOR_SH2A; ++ if (TARGET_SHJ2) ++ sh_cpu = PROCESSOR_SHJ2; + if (TARGET_SH3) + sh_cpu = PROCESSOR_SH3; + if (TARGET_SH3E) +diff --git a/gcc/config/sh/sh.h b/gcc/config/sh/sh.h +index d2280e2ffe6..3a54a896721 100644 +--- a/gcc/config/sh/sh.h ++++ b/gcc/config/sh/sh.h +@@ -85,6 +85,7 @@ extern int code_for_indirect_jump_scratch; + #define SUPPORT_SH4_SINGLE 1 + #define SUPPORT_SH2A 1 + #define SUPPORT_SH2A_SINGLE 1 ++#define SUPPORT_SHJ2 1 + #endif + + #define TARGET_DIVIDE_CALL_DIV1 (sh_div_strategy == SH_DIV_CALL_DIV1) +@@ -117,6 +118,7 @@ extern int code_for_indirect_jump_scratch; + #define SELECT_SH4A_SINGLE_ONLY (MASK_SH4A | SELECT_SH4_SINGLE_ONLY) + #define SELECT_SH4A (MASK_SH4A | SELECT_SH4) + #define SELECT_SH4A_SINGLE (MASK_SH4A | SELECT_SH4_SINGLE) ++#define SELECT_SHJ2 (MASK_SHJ2 | SELECT_SH2) + + #if SUPPORT_SH1 + #define SUPPORT_SH2 1 +@@ -124,6 +126,7 @@ extern int code_for_indirect_jump_scratch; + #if SUPPORT_SH2 + #define SUPPORT_SH3 1 + #define SUPPORT_SH2A_NOFPU 1 ++#define SUPPORT_SHJ2 1 + #endif + #if SUPPORT_SH3 + #define SUPPORT_SH4_NOFPU 1 +@@ -156,7 +159,7 @@ extern int code_for_indirect_jump_scratch; + #define MASK_ARCH (MASK_SH1 | MASK_SH2 | MASK_SH3 | MASK_SH_E | MASK_SH4 \ + | MASK_HARD_SH2A | MASK_HARD_SH2A_DOUBLE | MASK_SH4A \ + | MASK_HARD_SH4 | MASK_FPU_SINGLE \ +- | MASK_FPU_SINGLE_ONLY) ++ | MASK_FPU_SINGLE_ONLY | MASK_SHJ2) + + /* This defaults us to big-endian. */ + #ifndef TARGET_ENDIAN_DEFAULT +@@ -231,7 +234,8 @@ extern int code_for_indirect_jump_scratch; + %{m2a-single:--isa=sh2a} \ + %{m2a-single-only:--isa=sh2a} \ + %{m2a-nofpu:--isa=sh2a-nofpu} \ +-%{m4al:-dsp}" ++%{m4al:-dsp} \ ++%{mj2:-isa=j2}" + + #define ASM_SPEC SH_ASM_SPEC + +@@ -347,6 +351,7 @@ struct sh_atomic_model + hard_llcs, + soft_tcb, + soft_imask, ++ hard_cas, + + num_models + }; +@@ -390,6 +395,9 @@ extern const sh_atomic_model& selected_atomic_model (void); + #define TARGET_ATOMIC_SOFT_IMASK \ + (selected_atomic_model ().type == sh_atomic_model::soft_imask) + ++#define TARGET_ATOMIC_HARD_CAS \ ++ (selected_atomic_model ().type == sh_atomic_model::hard_cas) ++ + #endif // __cplusplus + + #define SUBTARGET_OVERRIDE_OPTIONS (void) 0 +@@ -1484,7 +1492,7 @@ extern bool current_function_interrupt; + + /* Nonzero if the target supports dynamic shift instructions + like shad and shld. */ +-#define TARGET_DYNSHIFT (TARGET_SH3 || TARGET_SH2A) ++#define TARGET_DYNSHIFT (TARGET_SH3 || TARGET_SH2A || TARGET_SHJ2) + + /* The cost of using the dynamic shift insns (shad, shld) are the same + if they are available. If they are not available a library function will +@@ -1747,6 +1755,7 @@ enum processor_type { + PROCESSOR_SH2, + PROCESSOR_SH2E, + PROCESSOR_SH2A, ++ PROCESSOR_SHJ2, + PROCESSOR_SH3, + PROCESSOR_SH3E, + PROCESSOR_SH4, +diff --git a/gcc/config/sh/sh.opt b/gcc/config/sh/sh.opt +index b4755a812f3..0989a1c18da 100644 +--- a/gcc/config/sh/sh.opt ++++ b/gcc/config/sh/sh.opt +@@ -65,6 +65,10 @@ m2e + Target RejectNegative Condition(SUPPORT_SH2E) + Generate SH2e code. + ++mj2 ++Target RejectNegative Mask(SHJ2) Condition(SUPPORT_SHJ2) ++Generate J2 code. ++ + m3 + Target RejectNegative Mask(SH3) Condition(SUPPORT_SH3) + Generate SH3 code. +diff --git a/gcc/config/sh/sync.md b/gcc/config/sh/sync.md +index 2b43f8edb86..118fc5d06db 100644 +--- a/gcc/config/sh/sync.md ++++ b/gcc/config/sh/sync.md +@@ -240,6 +240,9 @@ + || (TARGET_SH4A && <MODE>mode == SImode && !TARGET_ATOMIC_STRICT)) + atomic_insn = gen_atomic_compare_and_swap<mode>_hard (old_val, mem, + exp_val, new_val); ++ else if (TARGET_ATOMIC_HARD_CAS && <MODE>mode == SImode) ++ atomic_insn = gen_atomic_compare_and_swap<mode>_cas (old_val, mem, ++ exp_val, new_val); + else if (TARGET_ATOMIC_SOFT_GUSA) + atomic_insn = gen_atomic_compare_and_swap<mode>_soft_gusa (old_val, mem, + exp_val, new_val); +@@ -306,6 +309,57 @@ + } + [(set_attr "length" "14")]) + ++(define_expand "atomic_compare_and_swapsi_cas" ++ [(set (match_operand:SI 0 "register_operand" "=r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "register_operand" "r") ++ (match_operand:SI 3 "register_operand" "r")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{ ++ rtx mem = gen_rtx_REG (SImode, 0); ++ emit_move_insn (mem, force_reg (SImode, XEXP (operands[1], 0))); ++ emit_insn (gen_shj2_cas (operands[0], mem, operands[2], operands[3])); ++ DONE; ++}) ++ ++(define_insn "shj2_cas" ++ [(set (match_operand:SI 0 "register_operand" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "register_operand" "=r") ++ (match_operand:SI 2 "register_operand" "r") ++ (match_operand:SI 3 "register_operand" "0")] ++ UNSPECV_CMPXCHG_1)) ++ (set (reg:SI T_REG) ++ (unspec_volatile:SI [(const_int 0)] UNSPECV_CMPXCHG_3))] ++ "TARGET_ATOMIC_HARD_CAS" ++ "cas.l %2,%0,@%1" ++ [(set_attr "length" "2")] ++) ++ ++(define_expand "atomic_compare_and_swapqi_cas" ++ [(set (match_operand:SI 0 "arith_reg_dest" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "arith_operand" "rI08") ++ (match_operand:SI 3 "arith_operand" "rI08")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{FAIL;} ++) ++ ++(define_expand "atomic_compare_and_swaphi_cas" ++ [(set (match_operand:SI 0 "arith_reg_dest" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "arith_operand" "rI08") ++ (match_operand:SI 3 "arith_operand" "rI08")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{FAIL;} ++) ++ + ;; The QIHImode llcs patterns modify the address register of the memory + ;; operand. In order to express that, we have to open code the memory + ;; operand. Initially the insn is expanded like every other atomic insn +diff --git a/gcc/config/sh/t-sh b/gcc/config/sh/t-sh +index 888f8ff7f25..29fd6ae45fd 100644 +--- a/gcc/config/sh/t-sh ++++ b/gcc/config/sh/t-sh +@@ -50,7 +50,8 @@ MULTILIB_MATCHES = $(shell \ + m2e,m3e,m4-single-only,m4-100-single-only,m4-200-single-only,m4-300-single-only,m4a-single-only \ + m2a-single,m2a-single-only \ + m4-single,m4-100-single,m4-200-single,m4-300-single,m4a-single \ +- m4,m4-100,m4-200,m4-300,m4a; do \ ++ m4,m4-100,m4-200,m4-300,m4a \ ++ mj2; do \ + subst= ; \ + for lib in `echo $$abi|tr , ' '` ; do \ + if test "`echo $$multilibs|sed s/$$lib//`" != "$$multilibs"; then \ +@@ -63,9 +64,9 @@ MULTILIB_MATCHES = $(shell \ + + # SH1 and SH2A support big endian only. + ifeq ($(DEFAULT_ENDIAN),ml) +-MULTILIB_EXCEPTIONS = m1 ml/m1 m2a* ml/m2a* $(TM_MULTILIB_EXCEPTIONS_CONFIG) ++MULTILIB_EXCEPTIONS = m1 ml/m1 m2a* ml/m2a* ml/mj2 $(TM_MULTILIB_EXCEPTIONS_CONFIG) + else +-MULTILIB_EXCEPTIONS = ml/m1 ml/m2a* $(TM_MULTILIB_EXCEPTIONS_CONFIG) ++MULTILIB_EXCEPTIONS = ml/m1 ml/m2a* ml/mj2 $(TM_MULTILIB_EXCEPTIONS_CONFIG) + endif + + MULTILIB_OSDIRNAMES = \ +@@ -87,7 +88,8 @@ MULTILIB_OSDIRNAMES = \ + m4a-single-only=!m4a-single-only $(OTHER_ENDIAN)/m4a-single-only=!$(OTHER_ENDIAN)/m4a-single-only \ + m4a-single=!m4a-single $(OTHER_ENDIAN)/m4a-single=!$(OTHER_ENDIAN)/m4a-single \ + m4a=!m4a $(OTHER_ENDIAN)/m4a=!$(OTHER_ENDIAN)/m4a \ +- m4al=!m4al $(OTHER_ENDIAN)/m4al=!$(OTHER_ENDIAN)/m4al ++ m4al=!m4al $(OTHER_ENDIAN)/m4al=!$(OTHER_ENDIAN)/m4al \ ++ mj2=!j2 + + $(out_object_file): gt-sh.h + gt-sh.h : s-gtype ; @true diff --git a/patches/gcc-13.3.0/0004-static-pie.diff b/patches/gcc-13.3.0/0004-static-pie.diff new file mode 100644 index 0000000000000000000000000000000000000000..a85a3d738ce88537da7a31f4076d3ad9c0e59ae9 --- /dev/null +++ b/patches/gcc-13.3.0/0004-static-pie.diff @@ -0,0 +1,92 @@ +diff --git a/gcc/common.opt b/gcc/common.opt +index a75b44ee47e..7c564818b49 100644 +--- a/gcc/common.opt ++++ b/gcc/common.opt +@@ -3473,11 +3473,11 @@ Driver + + no-pie + Driver RejectNegative Negative(shared) +-Don't create a dynamically linked position independent executable. ++Don't create a position independent executable. + + pie + Driver RejectNegative Negative(no-pie) +-Create a dynamically linked position independent executable. ++Create a position independent executable. + + static-pie + Driver RejectNegative Negative(pie) +diff --git a/gcc/config/gnu-user.h b/gcc/config/gnu-user.h +index 5ebbf42a13d..bb907d8e89a 100644 +--- a/gcc/config/gnu-user.h ++++ b/gcc/config/gnu-user.h +@@ -51,13 +51,12 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + #define GNU_USER_TARGET_STARTFILE_SPEC \ + "%{shared:; \ + pg|p|profile:%{static-pie:grcrt1.o%s;:gcrt1.o%s}; \ +- static:crt1.o%s; \ +- static-pie:rcrt1.o%s; \ ++ static|static-pie:%{" PIE_SPEC ":rcrt1.o%s;:crt1.o%s}; \ + " PIE_SPEC ":Scrt1.o%s; \ + :crt1.o%s} " \ + GNU_USER_TARGET_CRTI " \ +- %{static:crtbeginT.o%s; \ +- shared|static-pie|" PIE_SPEC ":crtbeginS.o%s; \ ++ %{shared|" PIE_SPEC ":crtbeginS.o%s; \ ++ static:crtbeginT.o%s; \ + :crtbegin.o%s} \ + %{fvtable-verify=none:%s; \ + fvtable-verify=preinit:vtv_start_preinit.o%s; \ +@@ -73,11 +72,11 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + GNU userspace "finalizer" file, `crtn.o'. */ + + #define GNU_USER_TARGET_ENDFILE_SPEC \ +- "%{!static:%{fvtable-verify=none:%s; \ ++ "%{static|static-pie:; \ ++ fvtable-verify=none:%s; \ + fvtable-verify=preinit:vtv_end_preinit.o%s; \ +- fvtable-verify=std:vtv_end.o%s}} \ +- %{static:crtend.o%s; \ +- shared|static-pie|" PIE_SPEC ":crtendS.o%s; \ ++ fvtable-verify=std:vtv_end.o%s} \ ++ %{shared|" PIE_SPEC ":crtendS.o%s; \ + :crtend.o%s} " \ + GNU_USER_TARGET_CRTN " " \ + CRTOFFLOADEND +@@ -106,7 +105,7 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + #define LIB_SPEC GNU_USER_TARGET_LIB_SPEC + + #if defined(HAVE_LD_EH_FRAME_HDR) +-#define LINK_EH_SPEC "%{!static|static-pie:--eh-frame-hdr} " ++#define LINK_EH_SPEC "%{!static|" PIE_SPEC ":--eh-frame-hdr} " + #endif + + #define GNU_USER_TARGET_LINK_GCC_C_SEQUENCE_SPEC \ +diff --git a/gcc/gcc.cc b/gcc/gcc.cc +index 3c81c5798d8..cd96eac5d12 100644 +--- a/gcc/gcc.cc ++++ b/gcc/gcc.cc +@@ -1010,7 +1010,7 @@ proper position among the other output files. */ + #define NO_FPIE_AND_FPIC_SPEC NO_FPIE_SPEC "|" NO_FPIC_SPEC + #define FPIE_OR_FPIC_SPEC NO_FPIE_AND_FPIC_SPEC ":;" + #else +-#define PIE_SPEC "pie" ++#define PIE_SPEC "pie|static-pie" + #define FPIE1_SPEC "fpie" + #define NO_FPIE1_SPEC FPIE1_SPEC ":;" + #define FPIE2_SPEC "fPIE" +@@ -1034,12 +1034,12 @@ proper position among the other output files. */ + #ifndef LINK_PIE_SPEC + #ifdef HAVE_LD_PIE + #ifndef LD_PIE_SPEC +-#define LD_PIE_SPEC "-pie" ++#define LD_PIE_SPEC "-pie %{static|static-pie:--no-dynamic-linker -z text -Bsymbolic}" + #endif + #else + #define LD_PIE_SPEC "" + #endif +-#define LINK_PIE_SPEC "%{static|shared|r:;" PIE_SPEC ":" LD_PIE_SPEC "} " ++#define LINK_PIE_SPEC "%{shared|r:;" PIE_SPEC ":" LD_PIE_SPEC "} " + #endif + + #ifndef LINK_BUILDID_SPEC diff --git a/patches/gcc-13.3.0/0005-m68k-sqrt.diff b/patches/gcc-13.3.0/0005-m68k-sqrt.diff new file mode 100644 index 0000000000000000000000000000000000000000..652efb15fc38c9198626ce8744e3ea932820ce7e --- /dev/null +++ b/patches/gcc-13.3.0/0005-m68k-sqrt.diff @@ -0,0 +1,20 @@ +diff --git a/gcc/config/m68k/m68k.md b/gcc/config/m68k/m68k.md +index 59a456cd496..dbfddea41bd 100644 +--- a/gcc/config/m68k/m68k.md ++++ b/gcc/config/m68k/m68k.md +@@ -4174,13 +4174,13 @@ + (define_expand "sqrt<mode>2" + [(set (match_operand:FP 0 "nonimmediate_operand" "") + (sqrt:FP (match_operand:FP 1 "general_operand" "")))] +- "TARGET_HARD_FLOAT" ++ "(TARGET_68881 && TARGET_68040) || TARGET_COLDFIRE_FPU" + "") + + (define_insn "sqrt<mode>2_68881" + [(set (match_operand:FP 0 "nonimmediate_operand" "=f") + (sqrt:FP (match_operand:FP 1 "general_operand" "f<FP:dreg>m")))] +- "TARGET_68881" ++ "TARGET_68881 && TARGET_68040" + { + if (FP_REG_P (operands[1])) + return "f<FP:round>sqrt%.x %1,%0"; diff --git a/patches/gcc-13.3.0/0006-cow-libstdc++v3.diff b/patches/gcc-13.3.0/0006-cow-libstdc++v3.diff new file mode 100644 index 0000000000000000000000000000000000000000..b49b22c247dd8cb7459143dfdc0e2c32548e5a1b --- /dev/null +++ b/patches/gcc-13.3.0/0006-cow-libstdc++v3.diff @@ -0,0 +1,8 @@ +--- a/libstdc++-v3/ChangeLog 2024-03-13 18:04:21.050801063 -0400 ++++ b/libstdc++-v3/ChangeLog 2024-03-13 18:04:25.665836804 -0400 +@@ -4068,4 +4068,4 @@ + + Copying and distribution of this file, with or without modification, + are permitted in any medium without royalty provided the copyright +-notice and this notice are preserved. ++notice and this notice are preserved. diff --git a/patches/gcc-13.3.0/0007-fdpic-unwind.diff b/patches/gcc-13.3.0/0007-fdpic-unwind.diff new file mode 100644 index 0000000000000000000000000000000000000000..61d3479514991fc2823c0d5f1c76dca48761a114 --- /dev/null +++ b/patches/gcc-13.3.0/0007-fdpic-unwind.diff @@ -0,0 +1,11 @@ +--- a/libgcc/unwind-pe.h 2024-03-14 05:59:53.754073149 +0900 ++++ b/libgcc/unwind-pe.h 2024-03-14 06:00:41.226074492 +0900 +@@ -262,7 +262,7 @@ + + if (result != 0) + { +-#if __FDPIC__ ++#if __FDPIC__ && __arm__ + /* FDPIC relative addresses imply taking the GOT address + into account. */ + if ((encoding & DW_EH_PE_pcrel) && (encoding & DW_EH_PE_indirect)) diff --git a/patches/gcc-13.3.0/0008-fdpic-crtstuff-pr114158.diff b/patches/gcc-13.3.0/0008-fdpic-crtstuff-pr114158.diff new file mode 100644 index 0000000000000000000000000000000000000000..3986eed9be2275c5b942d47e9760de674df20f39 --- /dev/null +++ b/patches/gcc-13.3.0/0008-fdpic-crtstuff-pr114158.diff @@ -0,0 +1,38 @@ +--- a/libgcc/crtstuff.c 2023-05-29 17:46:32.000000000 +0900 ++++ b/libgcc/crtstuff.c 2024-03-14 06:03:42.398079615 +0900 +@@ -441,17 +441,9 @@ + #ifdef FINI_SECTION_ASM_OP + CRT_CALL_STATIC_FUNCTION (FINI_SECTION_ASM_OP, __do_global_dtors_aux) + #elif defined (FINI_ARRAY_SECTION_ASM_OP) +-#if defined(__FDPIC__) +-__asm__("\t.equ\t__do_global_dtors_aux_alias, __do_global_dtors_aux\n"); +-extern char __do_global_dtors_aux_alias; +-static void *__do_global_dtors_aux_fini_array_entry[] +-__attribute__ ((__used__, section(".fini_array"), aligned(sizeof(void *)))) +- = { &__do_global_dtors_aux_alias }; +-#else /* defined(__FDPIC__) */ + static func_ptr __do_global_dtors_aux_fini_array_entry[] + __attribute__ ((__used__, section(".fini_array"), + aligned(__alignof__(func_ptr)))) = { __do_global_dtors_aux }; +-#endif /* defined(__FDPIC__) */ + #else /* !FINI_SECTION_ASM_OP && !FINI_ARRAY_SECTION_ASM_OP */ + static void __attribute__((used)) + __do_global_dtors_aux_1 (void) +@@ -494,17 +486,9 @@ + #ifdef __LIBGCC_INIT_SECTION_ASM_OP__ + CRT_CALL_STATIC_FUNCTION (__LIBGCC_INIT_SECTION_ASM_OP__, frame_dummy) + #else /* defined(__LIBGCC_INIT_SECTION_ASM_OP__) */ +-#if defined(__FDPIC__) +-__asm__("\t.equ\t__frame_dummy_alias, frame_dummy\n"); +-extern char __frame_dummy_alias; +-static void *__frame_dummy_init_array_entry[] +-__attribute__ ((__used__, section(".init_array"), aligned(sizeof(void *)))) +- = { &__frame_dummy_alias }; +-#else /* defined(__FDPIC__) */ + static func_ptr __frame_dummy_init_array_entry[] + __attribute__ ((__used__, section(".init_array"), + aligned(__alignof__(func_ptr)))) = { frame_dummy }; +-#endif /* defined(__FDPIC__) */ + #endif /* !defined(__LIBGCC_INIT_SECTION_ASM_OP__) */ + #endif /* USE_EH_FRAME_REGISTRY || USE_TM_CLONE_REGISTRY */ + diff --git a/patches/gcc-13.3.0/0009-sh-fdpic-pr114641.diff b/patches/gcc-13.3.0/0009-sh-fdpic-pr114641.diff new file mode 100644 index 0000000000000000000000000000000000000000..020422c30e8d16052967ca3ec9375b9232f39932 --- /dev/null +++ b/patches/gcc-13.3.0/0009-sh-fdpic-pr114641.diff @@ -0,0 +1,12 @@ +--- gcc-11.4.0/gcc/config/sh/sh.cc.orig 2024-04-04 05:52:42.125373614 +0900 ++++ gcc-11.4.0/gcc/config/sh/sh.cc 2024-04-04 22:54:01.875106654 +0900 +@@ -9147,7 +9147,7 @@ + { + /* Weak functions may be NULL which doesn't work with + GOTOFFFUNCDESC because the runtime offset is not known. */ +- if (SYMBOL_REF_WEAK (orig)) ++ if (SYMBOL_REF_WEAK (orig) || (TREE_PUBLIC(SYMBOL_REF_DECL(orig)) && DECL_VISIBILITY (SYMBOL_REF_DECL(orig)) != VISIBILITY_HIDDEN)) + emit_insn (gen_symGOTFUNCDESC2reg (reg, orig)); + else + emit_insn (gen_symGOTOFFFUNCDESC2reg (reg, orig)); + -- 2.45.2 [-- Attachment #4: 0003-add-gcc-14-2-hash-and-patches.eml --] [-- Type: message/rfc822, Size: 27544 bytes --] From: "J. Neuschäfer" <j.neuschaefer@gmx.net> To: musl@lists.openwall.com Cc: "J. Neuschäfer" <j.neuschaefer@gmx.net> Subject: [PATCH MCM 3/3] add gcc 14.2 hash and patches Date: Wed, 23 Oct 2024 14:23:02 +0200 Message-ID: <20241023-gcc-bump-v1-3-0760a1580713@gmx.net> The patches have been copied from 12.4. --- hashes/gcc-14.2.0.tar.xz.sha1 | 1 + patches/gcc-14.2.0/0001-ssp_nonshared.diff | 14 + patches/gcc-14.2.0/0002-posix_memalign.diff | 30 ++ patches/gcc-14.2.0/0003-j2.diff | 346 +++++++++++++++++++++ patches/gcc-14.2.0/0004-static-pie.diff | 92 ++++++ patches/gcc-14.2.0/0005-m68k-sqrt.diff | 20 ++ patches/gcc-14.2.0/0006-cow-libstdc++v3.diff | 8 + patches/gcc-14.2.0/0007-fdpic-unwind.diff | 11 + .../gcc-14.2.0/0008-fdpic-crtstuff-pr114158.diff | 38 +++ patches/gcc-14.2.0/0009-sh-fdpic-pr114641.diff | 12 + 10 files changed, 572 insertions(+) diff --git a/hashes/gcc-14.2.0.tar.xz.sha1 b/hashes/gcc-14.2.0.tar.xz.sha1 new file mode 100644 index 0000000000000000000000000000000000000000..0b9c02e6f546dc23e3e9b77f6bac85f0c4315f28 --- /dev/null +++ b/hashes/gcc-14.2.0.tar.xz.sha1 @@ -0,0 +1 @@ +d91ecc3d20ce6298bd95f9b09cc51dc6d3c73ae3 gcc-14.2.0.tar.xz diff --git a/patches/gcc-14.2.0/0001-ssp_nonshared.diff b/patches/gcc-14.2.0/0001-ssp_nonshared.diff new file mode 100644 index 0000000000000000000000000000000000000000..b250f15833de6ae5bd732935f9bd8a924bbd783e --- /dev/null +++ b/patches/gcc-14.2.0/0001-ssp_nonshared.diff @@ -0,0 +1,14 @@ +diff --git a/gcc/gcc.cc b/gcc/gcc.cc +index 7837553958b..3c81c5798d8 100644 +--- a/gcc/gcc.cc ++++ b/gcc/gcc.cc +@@ -980,7 +980,8 @@ proper position among the other output files. */ + #ifndef LINK_SSP_SPEC + #ifdef TARGET_LIBC_PROVIDES_SSP + #define LINK_SSP_SPEC "%{fstack-protector|fstack-protector-all" \ +- "|fstack-protector-strong|fstack-protector-explicit:}" ++ "|fstack-protector-strong|fstack-protector-explicit" \ ++ ":-lssp_nonshared}" + #else + #define LINK_SSP_SPEC "%{fstack-protector|fstack-protector-all" \ + "|fstack-protector-strong|fstack-protector-explicit" \ diff --git a/patches/gcc-14.2.0/0002-posix_memalign.diff b/patches/gcc-14.2.0/0002-posix_memalign.diff new file mode 100644 index 0000000000000000000000000000000000000000..1d80096318d43bac22bb885a56fbbd9dd7fd3a74 --- /dev/null +++ b/patches/gcc-14.2.0/0002-posix_memalign.diff @@ -0,0 +1,30 @@ +diff --git a/gcc/config/i386/pmm_malloc.h b/gcc/config/i386/pmm_malloc.h +index 1b0bfe37852..d7b2b19bb3c 100644 +--- a/gcc/config/i386/pmm_malloc.h ++++ b/gcc/config/i386/pmm_malloc.h +@@ -27,12 +27,13 @@ + #include <stdlib.h> + + /* We can't depend on <stdlib.h> since the prototype of posix_memalign +- may not be visible. */ ++ may not be visible and we can't pollute the namespace either. */ + #ifndef __cplusplus +-extern int posix_memalign (void **, size_t, size_t); ++extern int _mm_posix_memalign (void **, size_t, size_t) + #else +-extern "C" int posix_memalign (void **, size_t, size_t) throw (); ++extern "C" int _mm_posix_memalign (void **, size_t, size_t) throw () + #endif ++__asm__("posix_memalign"); + + static __inline void * + _mm_malloc (size_t __size, size_t __alignment) +@@ -42,7 +43,7 @@ _mm_malloc (size_t __size, size_t __alignment) + return malloc (__size); + if (__alignment == 2 || (sizeof (void *) == 8 && __alignment == 4)) + __alignment = sizeof (void *); +- if (posix_memalign (&__ptr, __alignment, __size) == 0) ++ if (_mm_posix_memalign (&__ptr, __alignment, __size) == 0) + return __ptr; + else + return NULL; diff --git a/patches/gcc-14.2.0/0003-j2.diff b/patches/gcc-14.2.0/0003-j2.diff new file mode 100644 index 0000000000000000000000000000000000000000..c3786456c958668a1e08dd0a2b1e8280ea50fd7a --- /dev/null +++ b/patches/gcc-14.2.0/0003-j2.diff @@ -0,0 +1,346 @@ +diff --git a/gcc/config.gcc b/gcc/config.gcc +index 357b0bed067..528add999f2 100644 +--- a/gcc/config.gcc ++++ b/gcc/config.gcc +@@ -556,7 +556,7 @@ s390*-*-*) + extra_headers="s390intrin.h htmintrin.h htmxlintrin.h vecintrin.h" + ;; + # Note the 'l'; we need to be able to match e.g. "shle" or "shl". +-sh[123456789lbe]*-*-* | sh-*-*) ++sh[123456789lbej]*-*-* | sh-*-*) + cpu_type=sh + extra_options="${extra_options} fused-madd.opt" + extra_objs="${extra_objs} sh_treg_combine.o sh-mem.o sh_optimize_sett_clrt.o" +@@ -3202,18 +3202,18 @@ s390x-ibm-tpf*) + extra_options="${extra_options} s390/tpf.opt" + tmake_file="${tmake_file} s390/t-s390" + ;; +-sh-*-elf* | sh[12346l]*-*-elf* | \ +- sh-*-linux* | sh[2346lbe]*-*-linux* | \ ++sh-*-elf* | sh[12346lj]*-*-elf* | \ ++ sh-*-linux* | sh[2346lbej]*-*-linux* | \ + sh-*-netbsdelf* | shl*-*-netbsdelf*) + tmake_file="${tmake_file} sh/t-sh sh/t-elf" + if test x${with_endian} = x; then + case ${target} in +- sh[1234]*be-*-* | sh[1234]*eb-*-*) with_endian=big ;; ++ sh[j1234]*be-*-* | sh[j1234]*eb-*-*) with_endian=big ;; + shbe-*-* | sheb-*-*) with_endian=big,little ;; + sh[1234]l* | sh[34]*-*-linux*) with_endian=little ;; + shl* | sh*-*-linux* | \ + sh-superh-elf) with_endian=little,big ;; +- sh[1234]*-*-*) with_endian=big ;; ++ sh[j1234]*-*-*) with_endian=big ;; + *) with_endian=big,little ;; + esac + fi +@@ -3280,6 +3280,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + sh2a_nofpu*) sh_cpu_target=sh2a-nofpu ;; + sh2a*) sh_cpu_target=sh2a ;; + sh2e*) sh_cpu_target=sh2e ;; ++ shj2*) sh_cpu_target=shj2;; + sh2*) sh_cpu_target=sh2 ;; + *) sh_cpu_target=sh1 ;; + esac +@@ -3301,7 +3302,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + sh2a-single-only | sh2a-single | sh2a-nofpu | sh2a | \ + sh4a-single-only | sh4a-single | sh4a-nofpu | sh4a | sh4al | \ + sh4-single-only | sh4-single | sh4-nofpu | sh4 | sh4-300 | \ +- sh3e | sh3 | sh2e | sh2 | sh1) ;; ++ sh3e | sh3 | sh2e | sh2 | sh1 | shj2) ;; + "") sh_cpu_default=${sh_cpu_target} ;; + *) echo "with_cpu=$with_cpu not supported"; exit 1 ;; + esac +@@ -3310,9 +3311,9 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + case ${target} in + sh[1234]*) sh_multilibs=${sh_cpu_target} ;; + sh-superh-*) sh_multilibs=m4,m4-single,m4-single-only,m4-nofpu ;; +- sh*-*-linux*) sh_multilibs=m1,m2,m2a,m3e,m4 ;; ++ sh*-*-linux*) sh_multilibs=m1,m2,m2a,m3e,m4,mj2 ;; + sh*-*-netbsd*) sh_multilibs=m3,m3e,m4 ;; +- *) sh_multilibs=m1,m2,m2e,m4,m4-single,m4-single-only,m2a,m2a-single ;; ++ *) sh_multilibs=m1,m2,m2e,m4,m4-single,m4-single-only,m2a,m2a-single,mj2 ;; + esac + if test x$with_fp = xno; then + sh_multilibs="`echo $sh_multilibs|sed -e s/m4/sh4-nofpu/ -e s/,m4-[^,]*//g -e s/,m[23]e// -e s/m2a,m2a-single/m2a-nofpu/ -e s/m5-..m....,//g`" +@@ -3327,7 +3328,8 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + m1 | m2 | m2e | m3 | m3e | \ + m4 | m4-single | m4-single-only | m4-nofpu | m4-300 |\ + m4a | m4a-single | m4a-single-only | m4a-nofpu | m4al | \ +- m2a | m2a-single | m2a-single-only | m2a-nofpu) ++ m2a | m2a-single | m2a-single-only | m2a-nofpu | \ ++ mj2) + # TM_MULTILIB_CONFIG is used by t-sh for the non-endian multilib definition + # It is passed to MULTIILIB_OPTIONS verbatim. + TM_MULTILIB_CONFIG="${TM_MULTILIB_CONFIG}/${sh_multilib}" +@@ -3344,7 +3346,7 @@ sh-*-elf* | sh[12346l]*-*-elf* | \ + done + TM_MULTILIB_CONFIG=`echo $TM_MULTILIB_CONFIG | sed 's:^/::'` + if test x${enable_incomplete_targets} = xyes ; then +- tm_defines="$tm_defines SUPPORT_SH1=1 SUPPORT_SH2E=1 SUPPORT_SH4=1 SUPPORT_SH4_SINGLE=1 SUPPORT_SH2A=1 SUPPORT_SH2A_SINGLE=1" ++ tm_defines="$tm_defines SUPPORT_SH1=1 SUPPORT_SH2E=1 SUPPORT_SH4=1 SUPPORT_SH4_SINGLE=1 SUPPORT_SH2A=1 SUPPORT_SH2A_SINGLE=1 SUPPORT_SHJ2=1" + fi + tm_file="$tm_file ./sysroot-suffix.h" + tmake_file="$tmake_file t-sysroot-suffix" +@@ -5175,6 +5177,8 @@ case "${target}" in + ;; + m4a | m4a-single | m4a-single-only | m4a-nofpu | m4al) + ;; ++ mj2) ++ ;; + *) + echo "Unknown CPU used in --with-cpu=$with_cpu, known values:" 1>&2 + echo "m1 m2 m2e m3 m3e m4 m4-single m4-single-only m4-nofpu" 1>&2 +@@ -5385,7 +5389,7 @@ case ${target} in + tmake_file="${cpu_type}/t-${cpu_type} ${tmake_file}" + ;; + +- sh[123456ble]*-*-* | sh-*-*) ++ sh[123456blej]*-*-* | sh-*-*) + c_target_objs="${c_target_objs} sh-c.o" + cxx_target_objs="${cxx_target_objs} sh-c.o" + ;; +diff --git a/gcc/config/sh/sh.cc b/gcc/config/sh/sh.cc +index 1564109c942..798c1c1c1a3 100644 +--- a/gcc/config/sh/sh.cc ++++ b/gcc/config/sh/sh.cc +@@ -686,6 +686,7 @@ parse_validate_atomic_model_option (const char* str) + model_names[sh_atomic_model::hard_llcs] = "hard-llcs"; + model_names[sh_atomic_model::soft_tcb] = "soft-tcb"; + model_names[sh_atomic_model::soft_imask] = "soft-imask"; ++ model_names[sh_atomic_model::hard_cas] = "hard-cas"; + + const char* model_cdef_names[sh_atomic_model::num_models]; + model_cdef_names[sh_atomic_model::none] = "NONE"; +@@ -693,6 +694,7 @@ parse_validate_atomic_model_option (const char* str) + model_cdef_names[sh_atomic_model::hard_llcs] = "HARD_LLCS"; + model_cdef_names[sh_atomic_model::soft_tcb] = "SOFT_TCB"; + model_cdef_names[sh_atomic_model::soft_imask] = "SOFT_IMASK"; ++ model_cdef_names[sh_atomic_model::hard_cas] = "HARD_CAS"; + + sh_atomic_model ret; + ret.type = sh_atomic_model::none; +@@ -771,6 +773,9 @@ got_mode_name:; + if (ret.type == sh_atomic_model::soft_imask && TARGET_USERMODE) + err_ret ("cannot use atomic model %s in user mode", ret.name); + ++ if (ret.type == sh_atomic_model::hard_cas && !TARGET_SHJ2) ++ err_ret ("atomic model %s is only available J2 targets", ret.name); ++ + return ret; + + #undef err_ret +@@ -827,6 +832,8 @@ sh_option_override (void) + sh_cpu = PROCESSOR_SH2E; + if (TARGET_SH2A) + sh_cpu = PROCESSOR_SH2A; ++ if (TARGET_SHJ2) ++ sh_cpu = PROCESSOR_SHJ2; + if (TARGET_SH3) + sh_cpu = PROCESSOR_SH3; + if (TARGET_SH3E) +diff --git a/gcc/config/sh/sh.h b/gcc/config/sh/sh.h +index d2280e2ffe6..3a54a896721 100644 +--- a/gcc/config/sh/sh.h ++++ b/gcc/config/sh/sh.h +@@ -85,6 +85,7 @@ extern int code_for_indirect_jump_scratch; + #define SUPPORT_SH4_SINGLE 1 + #define SUPPORT_SH2A 1 + #define SUPPORT_SH2A_SINGLE 1 ++#define SUPPORT_SHJ2 1 + #endif + + #define TARGET_DIVIDE_CALL_DIV1 (sh_div_strategy == SH_DIV_CALL_DIV1) +@@ -117,6 +118,7 @@ extern int code_for_indirect_jump_scratch; + #define SELECT_SH4A_SINGLE_ONLY (MASK_SH4A | SELECT_SH4_SINGLE_ONLY) + #define SELECT_SH4A (MASK_SH4A | SELECT_SH4) + #define SELECT_SH4A_SINGLE (MASK_SH4A | SELECT_SH4_SINGLE) ++#define SELECT_SHJ2 (MASK_SHJ2 | SELECT_SH2) + + #if SUPPORT_SH1 + #define SUPPORT_SH2 1 +@@ -124,6 +126,7 @@ extern int code_for_indirect_jump_scratch; + #if SUPPORT_SH2 + #define SUPPORT_SH3 1 + #define SUPPORT_SH2A_NOFPU 1 ++#define SUPPORT_SHJ2 1 + #endif + #if SUPPORT_SH3 + #define SUPPORT_SH4_NOFPU 1 +@@ -156,7 +159,7 @@ extern int code_for_indirect_jump_scratch; + #define MASK_ARCH (MASK_SH1 | MASK_SH2 | MASK_SH3 | MASK_SH_E | MASK_SH4 \ + | MASK_HARD_SH2A | MASK_HARD_SH2A_DOUBLE | MASK_SH4A \ + | MASK_HARD_SH4 | MASK_FPU_SINGLE \ +- | MASK_FPU_SINGLE_ONLY) ++ | MASK_FPU_SINGLE_ONLY | MASK_SHJ2) + + /* This defaults us to big-endian. */ + #ifndef TARGET_ENDIAN_DEFAULT +@@ -231,7 +234,8 @@ extern int code_for_indirect_jump_scratch; + %{m2a-single:--isa=sh2a} \ + %{m2a-single-only:--isa=sh2a} \ + %{m2a-nofpu:--isa=sh2a-nofpu} \ +-%{m4al:-dsp}" ++%{m4al:-dsp} \ ++%{mj2:-isa=j2}" + + #define ASM_SPEC SH_ASM_SPEC + +@@ -347,6 +351,7 @@ struct sh_atomic_model + hard_llcs, + soft_tcb, + soft_imask, ++ hard_cas, + + num_models + }; +@@ -390,6 +395,9 @@ extern const sh_atomic_model& selected_atomic_model (void); + #define TARGET_ATOMIC_SOFT_IMASK \ + (selected_atomic_model ().type == sh_atomic_model::soft_imask) + ++#define TARGET_ATOMIC_HARD_CAS \ ++ (selected_atomic_model ().type == sh_atomic_model::hard_cas) ++ + #endif // __cplusplus + + #define SUBTARGET_OVERRIDE_OPTIONS (void) 0 +@@ -1484,7 +1492,7 @@ extern bool current_function_interrupt; + + /* Nonzero if the target supports dynamic shift instructions + like shad and shld. */ +-#define TARGET_DYNSHIFT (TARGET_SH3 || TARGET_SH2A) ++#define TARGET_DYNSHIFT (TARGET_SH3 || TARGET_SH2A || TARGET_SHJ2) + + /* The cost of using the dynamic shift insns (shad, shld) are the same + if they are available. If they are not available a library function will +@@ -1747,6 +1755,7 @@ enum processor_type { + PROCESSOR_SH2, + PROCESSOR_SH2E, + PROCESSOR_SH2A, ++ PROCESSOR_SHJ2, + PROCESSOR_SH3, + PROCESSOR_SH3E, + PROCESSOR_SH4, +diff --git a/gcc/config/sh/sh.opt b/gcc/config/sh/sh.opt +index b4755a812f3..0989a1c18da 100644 +--- a/gcc/config/sh/sh.opt ++++ b/gcc/config/sh/sh.opt +@@ -65,6 +65,10 @@ m2e + Target RejectNegative Condition(SUPPORT_SH2E) + Generate SH2e code. + ++mj2 ++Target RejectNegative Mask(SHJ2) Condition(SUPPORT_SHJ2) ++Generate J2 code. ++ + m3 + Target RejectNegative Mask(SH3) Condition(SUPPORT_SH3) + Generate SH3 code. +diff --git a/gcc/config/sh/sync.md b/gcc/config/sh/sync.md +index 2b43f8edb86..118fc5d06db 100644 +--- a/gcc/config/sh/sync.md ++++ b/gcc/config/sh/sync.md +@@ -240,6 +240,9 @@ + || (TARGET_SH4A && <MODE>mode == SImode && !TARGET_ATOMIC_STRICT)) + atomic_insn = gen_atomic_compare_and_swap<mode>_hard (old_val, mem, + exp_val, new_val); ++ else if (TARGET_ATOMIC_HARD_CAS && <MODE>mode == SImode) ++ atomic_insn = gen_atomic_compare_and_swap<mode>_cas (old_val, mem, ++ exp_val, new_val); + else if (TARGET_ATOMIC_SOFT_GUSA) + atomic_insn = gen_atomic_compare_and_swap<mode>_soft_gusa (old_val, mem, + exp_val, new_val); +@@ -306,6 +309,57 @@ + } + [(set_attr "length" "14")]) + ++(define_expand "atomic_compare_and_swapsi_cas" ++ [(set (match_operand:SI 0 "register_operand" "=r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "register_operand" "r") ++ (match_operand:SI 3 "register_operand" "r")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{ ++ rtx mem = gen_rtx_REG (SImode, 0); ++ emit_move_insn (mem, force_reg (SImode, XEXP (operands[1], 0))); ++ emit_insn (gen_shj2_cas (operands[0], mem, operands[2], operands[3])); ++ DONE; ++}) ++ ++(define_insn "shj2_cas" ++ [(set (match_operand:SI 0 "register_operand" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "register_operand" "=r") ++ (match_operand:SI 2 "register_operand" "r") ++ (match_operand:SI 3 "register_operand" "0")] ++ UNSPECV_CMPXCHG_1)) ++ (set (reg:SI T_REG) ++ (unspec_volatile:SI [(const_int 0)] UNSPECV_CMPXCHG_3))] ++ "TARGET_ATOMIC_HARD_CAS" ++ "cas.l %2,%0,@%1" ++ [(set_attr "length" "2")] ++) ++ ++(define_expand "atomic_compare_and_swapqi_cas" ++ [(set (match_operand:SI 0 "arith_reg_dest" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "arith_operand" "rI08") ++ (match_operand:SI 3 "arith_operand" "rI08")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{FAIL;} ++) ++ ++(define_expand "atomic_compare_and_swaphi_cas" ++ [(set (match_operand:SI 0 "arith_reg_dest" "=&r") ++ (unspec_volatile:SI ++ [(match_operand:SI 1 "atomic_mem_operand_0" "=Sra") ++ (match_operand:SI 2 "arith_operand" "rI08") ++ (match_operand:SI 3 "arith_operand" "rI08")] ++ UNSPECV_CMPXCHG_1))] ++ "TARGET_ATOMIC_HARD_CAS" ++{FAIL;} ++) ++ + ;; The QIHImode llcs patterns modify the address register of the memory + ;; operand. In order to express that, we have to open code the memory + ;; operand. Initially the insn is expanded like every other atomic insn +diff --git a/gcc/config/sh/t-sh b/gcc/config/sh/t-sh +index 888f8ff7f25..29fd6ae45fd 100644 +--- a/gcc/config/sh/t-sh ++++ b/gcc/config/sh/t-sh +@@ -50,7 +50,8 @@ MULTILIB_MATCHES = $(shell \ + m2e,m3e,m4-single-only,m4-100-single-only,m4-200-single-only,m4-300-single-only,m4a-single-only \ + m2a-single,m2a-single-only \ + m4-single,m4-100-single,m4-200-single,m4-300-single,m4a-single \ +- m4,m4-100,m4-200,m4-300,m4a; do \ ++ m4,m4-100,m4-200,m4-300,m4a \ ++ mj2; do \ + subst= ; \ + for lib in `echo $$abi|tr , ' '` ; do \ + if test "`echo $$multilibs|sed s/$$lib//`" != "$$multilibs"; then \ +@@ -63,9 +64,9 @@ MULTILIB_MATCHES = $(shell \ + + # SH1 and SH2A support big endian only. + ifeq ($(DEFAULT_ENDIAN),ml) +-MULTILIB_EXCEPTIONS = m1 ml/m1 m2a* ml/m2a* $(TM_MULTILIB_EXCEPTIONS_CONFIG) ++MULTILIB_EXCEPTIONS = m1 ml/m1 m2a* ml/m2a* ml/mj2 $(TM_MULTILIB_EXCEPTIONS_CONFIG) + else +-MULTILIB_EXCEPTIONS = ml/m1 ml/m2a* $(TM_MULTILIB_EXCEPTIONS_CONFIG) ++MULTILIB_EXCEPTIONS = ml/m1 ml/m2a* ml/mj2 $(TM_MULTILIB_EXCEPTIONS_CONFIG) + endif + + MULTILIB_OSDIRNAMES = \ +@@ -87,7 +88,8 @@ MULTILIB_OSDIRNAMES = \ + m4a-single-only=!m4a-single-only $(OTHER_ENDIAN)/m4a-single-only=!$(OTHER_ENDIAN)/m4a-single-only \ + m4a-single=!m4a-single $(OTHER_ENDIAN)/m4a-single=!$(OTHER_ENDIAN)/m4a-single \ + m4a=!m4a $(OTHER_ENDIAN)/m4a=!$(OTHER_ENDIAN)/m4a \ +- m4al=!m4al $(OTHER_ENDIAN)/m4al=!$(OTHER_ENDIAN)/m4al ++ m4al=!m4al $(OTHER_ENDIAN)/m4al=!$(OTHER_ENDIAN)/m4al \ ++ mj2=!j2 + + $(out_object_file): gt-sh.h + gt-sh.h : s-gtype ; @true diff --git a/patches/gcc-14.2.0/0004-static-pie.diff b/patches/gcc-14.2.0/0004-static-pie.diff new file mode 100644 index 0000000000000000000000000000000000000000..a85a3d738ce88537da7a31f4076d3ad9c0e59ae9 --- /dev/null +++ b/patches/gcc-14.2.0/0004-static-pie.diff @@ -0,0 +1,92 @@ +diff --git a/gcc/common.opt b/gcc/common.opt +index a75b44ee47e..7c564818b49 100644 +--- a/gcc/common.opt ++++ b/gcc/common.opt +@@ -3473,11 +3473,11 @@ Driver + + no-pie + Driver RejectNegative Negative(shared) +-Don't create a dynamically linked position independent executable. ++Don't create a position independent executable. + + pie + Driver RejectNegative Negative(no-pie) +-Create a dynamically linked position independent executable. ++Create a position independent executable. + + static-pie + Driver RejectNegative Negative(pie) +diff --git a/gcc/config/gnu-user.h b/gcc/config/gnu-user.h +index 5ebbf42a13d..bb907d8e89a 100644 +--- a/gcc/config/gnu-user.h ++++ b/gcc/config/gnu-user.h +@@ -51,13 +51,12 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + #define GNU_USER_TARGET_STARTFILE_SPEC \ + "%{shared:; \ + pg|p|profile:%{static-pie:grcrt1.o%s;:gcrt1.o%s}; \ +- static:crt1.o%s; \ +- static-pie:rcrt1.o%s; \ ++ static|static-pie:%{" PIE_SPEC ":rcrt1.o%s;:crt1.o%s}; \ + " PIE_SPEC ":Scrt1.o%s; \ + :crt1.o%s} " \ + GNU_USER_TARGET_CRTI " \ +- %{static:crtbeginT.o%s; \ +- shared|static-pie|" PIE_SPEC ":crtbeginS.o%s; \ ++ %{shared|" PIE_SPEC ":crtbeginS.o%s; \ ++ static:crtbeginT.o%s; \ + :crtbegin.o%s} \ + %{fvtable-verify=none:%s; \ + fvtable-verify=preinit:vtv_start_preinit.o%s; \ +@@ -73,11 +72,11 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + GNU userspace "finalizer" file, `crtn.o'. */ + + #define GNU_USER_TARGET_ENDFILE_SPEC \ +- "%{!static:%{fvtable-verify=none:%s; \ ++ "%{static|static-pie:; \ ++ fvtable-verify=none:%s; \ + fvtable-verify=preinit:vtv_end_preinit.o%s; \ +- fvtable-verify=std:vtv_end.o%s}} \ +- %{static:crtend.o%s; \ +- shared|static-pie|" PIE_SPEC ":crtendS.o%s; \ ++ fvtable-verify=std:vtv_end.o%s} \ ++ %{shared|" PIE_SPEC ":crtendS.o%s; \ + :crtend.o%s} " \ + GNU_USER_TARGET_CRTN " " \ + CRTOFFLOADEND +@@ -106,7 +105,7 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see + #define LIB_SPEC GNU_USER_TARGET_LIB_SPEC + + #if defined(HAVE_LD_EH_FRAME_HDR) +-#define LINK_EH_SPEC "%{!static|static-pie:--eh-frame-hdr} " ++#define LINK_EH_SPEC "%{!static|" PIE_SPEC ":--eh-frame-hdr} " + #endif + + #define GNU_USER_TARGET_LINK_GCC_C_SEQUENCE_SPEC \ +diff --git a/gcc/gcc.cc b/gcc/gcc.cc +index 3c81c5798d8..cd96eac5d12 100644 +--- a/gcc/gcc.cc ++++ b/gcc/gcc.cc +@@ -1010,7 +1010,7 @@ proper position among the other output files. */ + #define NO_FPIE_AND_FPIC_SPEC NO_FPIE_SPEC "|" NO_FPIC_SPEC + #define FPIE_OR_FPIC_SPEC NO_FPIE_AND_FPIC_SPEC ":;" + #else +-#define PIE_SPEC "pie" ++#define PIE_SPEC "pie|static-pie" + #define FPIE1_SPEC "fpie" + #define NO_FPIE1_SPEC FPIE1_SPEC ":;" + #define FPIE2_SPEC "fPIE" +@@ -1034,12 +1034,12 @@ proper position among the other output files. */ + #ifndef LINK_PIE_SPEC + #ifdef HAVE_LD_PIE + #ifndef LD_PIE_SPEC +-#define LD_PIE_SPEC "-pie" ++#define LD_PIE_SPEC "-pie %{static|static-pie:--no-dynamic-linker -z text -Bsymbolic}" + #endif + #else + #define LD_PIE_SPEC "" + #endif +-#define LINK_PIE_SPEC "%{static|shared|r:;" PIE_SPEC ":" LD_PIE_SPEC "} " ++#define LINK_PIE_SPEC "%{shared|r:;" PIE_SPEC ":" LD_PIE_SPEC "} " + #endif + + #ifndef LINK_BUILDID_SPEC diff --git a/patches/gcc-14.2.0/0005-m68k-sqrt.diff b/patches/gcc-14.2.0/0005-m68k-sqrt.diff new file mode 100644 index 0000000000000000000000000000000000000000..652efb15fc38c9198626ce8744e3ea932820ce7e --- /dev/null +++ b/patches/gcc-14.2.0/0005-m68k-sqrt.diff @@ -0,0 +1,20 @@ +diff --git a/gcc/config/m68k/m68k.md b/gcc/config/m68k/m68k.md +index 59a456cd496..dbfddea41bd 100644 +--- a/gcc/config/m68k/m68k.md ++++ b/gcc/config/m68k/m68k.md +@@ -4174,13 +4174,13 @@ + (define_expand "sqrt<mode>2" + [(set (match_operand:FP 0 "nonimmediate_operand" "") + (sqrt:FP (match_operand:FP 1 "general_operand" "")))] +- "TARGET_HARD_FLOAT" ++ "(TARGET_68881 && TARGET_68040) || TARGET_COLDFIRE_FPU" + "") + + (define_insn "sqrt<mode>2_68881" + [(set (match_operand:FP 0 "nonimmediate_operand" "=f") + (sqrt:FP (match_operand:FP 1 "general_operand" "f<FP:dreg>m")))] +- "TARGET_68881" ++ "TARGET_68881 && TARGET_68040" + { + if (FP_REG_P (operands[1])) + return "f<FP:round>sqrt%.x %1,%0"; diff --git a/patches/gcc-14.2.0/0006-cow-libstdc++v3.diff b/patches/gcc-14.2.0/0006-cow-libstdc++v3.diff new file mode 100644 index 0000000000000000000000000000000000000000..b49b22c247dd8cb7459143dfdc0e2c32548e5a1b --- /dev/null +++ b/patches/gcc-14.2.0/0006-cow-libstdc++v3.diff @@ -0,0 +1,8 @@ +--- a/libstdc++-v3/ChangeLog 2024-03-13 18:04:21.050801063 -0400 ++++ b/libstdc++-v3/ChangeLog 2024-03-13 18:04:25.665836804 -0400 +@@ -4068,4 +4068,4 @@ + + Copying and distribution of this file, with or without modification, + are permitted in any medium without royalty provided the copyright +-notice and this notice are preserved. ++notice and this notice are preserved. diff --git a/patches/gcc-14.2.0/0007-fdpic-unwind.diff b/patches/gcc-14.2.0/0007-fdpic-unwind.diff new file mode 100644 index 0000000000000000000000000000000000000000..61d3479514991fc2823c0d5f1c76dca48761a114 --- /dev/null +++ b/patches/gcc-14.2.0/0007-fdpic-unwind.diff @@ -0,0 +1,11 @@ +--- a/libgcc/unwind-pe.h 2024-03-14 05:59:53.754073149 +0900 ++++ b/libgcc/unwind-pe.h 2024-03-14 06:00:41.226074492 +0900 +@@ -262,7 +262,7 @@ + + if (result != 0) + { +-#if __FDPIC__ ++#if __FDPIC__ && __arm__ + /* FDPIC relative addresses imply taking the GOT address + into account. */ + if ((encoding & DW_EH_PE_pcrel) && (encoding & DW_EH_PE_indirect)) diff --git a/patches/gcc-14.2.0/0008-fdpic-crtstuff-pr114158.diff b/patches/gcc-14.2.0/0008-fdpic-crtstuff-pr114158.diff new file mode 100644 index 0000000000000000000000000000000000000000..3986eed9be2275c5b942d47e9760de674df20f39 --- /dev/null +++ b/patches/gcc-14.2.0/0008-fdpic-crtstuff-pr114158.diff @@ -0,0 +1,38 @@ +--- a/libgcc/crtstuff.c 2023-05-29 17:46:32.000000000 +0900 ++++ b/libgcc/crtstuff.c 2024-03-14 06:03:42.398079615 +0900 +@@ -441,17 +441,9 @@ + #ifdef FINI_SECTION_ASM_OP + CRT_CALL_STATIC_FUNCTION (FINI_SECTION_ASM_OP, __do_global_dtors_aux) + #elif defined (FINI_ARRAY_SECTION_ASM_OP) +-#if defined(__FDPIC__) +-__asm__("\t.equ\t__do_global_dtors_aux_alias, __do_global_dtors_aux\n"); +-extern char __do_global_dtors_aux_alias; +-static void *__do_global_dtors_aux_fini_array_entry[] +-__attribute__ ((__used__, section(".fini_array"), aligned(sizeof(void *)))) +- = { &__do_global_dtors_aux_alias }; +-#else /* defined(__FDPIC__) */ + static func_ptr __do_global_dtors_aux_fini_array_entry[] + __attribute__ ((__used__, section(".fini_array"), + aligned(__alignof__(func_ptr)))) = { __do_global_dtors_aux }; +-#endif /* defined(__FDPIC__) */ + #else /* !FINI_SECTION_ASM_OP && !FINI_ARRAY_SECTION_ASM_OP */ + static void __attribute__((used)) + __do_global_dtors_aux_1 (void) +@@ -494,17 +486,9 @@ + #ifdef __LIBGCC_INIT_SECTION_ASM_OP__ + CRT_CALL_STATIC_FUNCTION (__LIBGCC_INIT_SECTION_ASM_OP__, frame_dummy) + #else /* defined(__LIBGCC_INIT_SECTION_ASM_OP__) */ +-#if defined(__FDPIC__) +-__asm__("\t.equ\t__frame_dummy_alias, frame_dummy\n"); +-extern char __frame_dummy_alias; +-static void *__frame_dummy_init_array_entry[] +-__attribute__ ((__used__, section(".init_array"), aligned(sizeof(void *)))) +- = { &__frame_dummy_alias }; +-#else /* defined(__FDPIC__) */ + static func_ptr __frame_dummy_init_array_entry[] + __attribute__ ((__used__, section(".init_array"), + aligned(__alignof__(func_ptr)))) = { frame_dummy }; +-#endif /* defined(__FDPIC__) */ + #endif /* !defined(__LIBGCC_INIT_SECTION_ASM_OP__) */ + #endif /* USE_EH_FRAME_REGISTRY || USE_TM_CLONE_REGISTRY */ + diff --git a/patches/gcc-14.2.0/0009-sh-fdpic-pr114641.diff b/patches/gcc-14.2.0/0009-sh-fdpic-pr114641.diff new file mode 100644 index 0000000000000000000000000000000000000000..020422c30e8d16052967ca3ec9375b9232f39932 --- /dev/null +++ b/patches/gcc-14.2.0/0009-sh-fdpic-pr114641.diff @@ -0,0 +1,12 @@ +--- gcc-11.4.0/gcc/config/sh/sh.cc.orig 2024-04-04 05:52:42.125373614 +0900 ++++ gcc-11.4.0/gcc/config/sh/sh.cc 2024-04-04 22:54:01.875106654 +0900 +@@ -9147,7 +9147,7 @@ + { + /* Weak functions may be NULL which doesn't work with + GOTOFFFUNCDESC because the runtime offset is not known. */ +- if (SYMBOL_REF_WEAK (orig)) ++ if (SYMBOL_REF_WEAK (orig) || (TREE_PUBLIC(SYMBOL_REF_DECL(orig)) && DECL_VISIBILITY (SYMBOL_REF_DECL(orig)) != VISIBILITY_HIDDEN)) + emit_insn (gen_symGOTFUNCDESC2reg (reg, orig)); + else + emit_insn (gen_symGOTOFFFUNCDESC2reg (reg, orig)); + -- 2.45.2 ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions 2024-10-23 12:29 ` J. Neuschäfer @ 2024-10-24 13:11 ` Rich Felker 2024-10-24 13:36 ` Rich Felker 0 siblings, 1 reply; 9+ messages in thread From: Rich Felker @ 2024-10-24 13:11 UTC (permalink / raw) To: J. Neuschäfer; +Cc: musl On Wed, Oct 23, 2024 at 02:29:13PM +0200, J. Neuschäfer wrote: > On Tue, Oct 22, 2024 at 05:52:22PM -0400, Rich Felker wrote: > > On Wed, Oct 02, 2024 at 05:40:03PM +0200, J. Neuschäfer wrote: > > > This patchset adds newer GCC versions to MCM. Patches are copied from > > > GCC 11 and only very lightly modifed (which can be verified by diffing > > > the patch directories). > > > > > > I haven't tested the resulting toolchains, so there might be some bugs > > > lurking. > > > > Could you resend these as attachments instead of lkml-thread-style? > > Something got messed up with the whitespace and git am is applying > > them but giving results that have gratuitous whitespace diffs, and I'd > > rather not track down whether any of that is breaking. > > > > Rich > > Here they are. Weird, I'm still getting problems. git-am gives me: .git/rebase-apply/patch:40: space before tab in indent. "|fstack-protector-strong|fstack-protector-explicit" \ .git/rebase-apply/patch:88: space before tab in indent. extra_headers="s390intrin.h htmintrin.h htmxlintrin.h vecintrin.h" .git/rebase-apply/patch:89: space before tab in indent. ;; .git/rebase-apply/patch:93: space before tab in indent. cpu_type=sh .git/rebase-apply/patch:94: space before tab in indent. extra_options="${extra_options} fused-madd.opt" warning: squelched 79 whitespace errors warning: 84 lines add whitespace errors. and the applied patches have context lines that should be an empty line with the one-space diff prefix replaced to empty lines (no leading diff space). Can you verify whether this is expected or if something is getting munged in the mail? Rich ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions 2024-10-24 13:11 ` Rich Felker @ 2024-10-24 13:36 ` Rich Felker 2024-10-25 11:39 ` J. Neuschäfer 0 siblings, 1 reply; 9+ messages in thread From: Rich Felker @ 2024-10-24 13:36 UTC (permalink / raw) To: J. Neuschäfer; +Cc: musl On Thu, Oct 24, 2024 at 09:11:42AM -0400, Rich Felker wrote: > On Wed, Oct 23, 2024 at 02:29:13PM +0200, J. Neuschäfer wrote: > > On Tue, Oct 22, 2024 at 05:52:22PM -0400, Rich Felker wrote: > > > On Wed, Oct 02, 2024 at 05:40:03PM +0200, J. Neuschäfer wrote: > > > > This patchset adds newer GCC versions to MCM. Patches are copied from > > > > GCC 11 and only very lightly modifed (which can be verified by diffing > > > > the patch directories). > > > > > > > > I haven't tested the resulting toolchains, so there might be some bugs > > > > lurking. > > > > > > Could you resend these as attachments instead of lkml-thread-style? > > > Something got messed up with the whitespace and git am is applying > > > them but giving results that have gratuitous whitespace diffs, and I'd > > > rather not track down whether any of that is breaking. > > > > > > Rich > > > > Here they are. > > Weird, I'm still getting problems. git-am gives me: > > ..git/rebase-apply/patch:40: space before tab in indent. > "|fstack-protector-strong|fstack-protector-explicit" \ > ..git/rebase-apply/patch:88: space before tab in indent. > extra_headers="s390intrin.h htmintrin.h htmxlintrin.h vecintrin.h" > ..git/rebase-apply/patch:89: space before tab in indent. > ;; > ..git/rebase-apply/patch:93: space before tab in indent. > cpu_type=sh > ..git/rebase-apply/patch:94: space before tab in indent. > extra_options="${extra_options} fused-madd.opt" > warning: squelched 79 whitespace errors > warning: 84 lines add whitespace errors. > > and the applied patches have context lines that should be an empty > line with the one-space diff prefix replaced to empty lines (no > leading diff space). > > Can you verify whether this is expected or if something is getting > munged in the mail? It looks like your mailer munged them sending the attachments as message/rfc822 mime, which is lossy with respect to whitespace, rather than text/plain or something that would be safely preserved. The "trailing whitespace" on the context lines is missing in the actual message I received. Rich ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions 2024-10-24 13:36 ` Rich Felker @ 2024-10-25 11:39 ` J. Neuschäfer 0 siblings, 0 replies; 9+ messages in thread From: J. Neuschäfer @ 2024-10-25 11:39 UTC (permalink / raw) To: musl; +Cc: J. Neuschäfer [-- Attachment #1: Type: text/plain, Size: 1849 bytes --] On Thu, Oct 24, 2024 at 09:36:20AM -0400, Rich Felker wrote: > On Thu, Oct 24, 2024 at 09:11:42AM -0400, Rich Felker wrote: [...] > > Weird, I'm still getting problems. git-am gives me: > > > > ..git/rebase-apply/patch:40: space before tab in indent. > > "|fstack-protector-strong|fstack-protector-explicit" \ > > ..git/rebase-apply/patch:88: space before tab in indent. > > extra_headers="s390intrin.h htmintrin.h htmxlintrin.h vecintrin.h" > > ..git/rebase-apply/patch:89: space before tab in indent. > > ;; > > ..git/rebase-apply/patch:93: space before tab in indent. > > cpu_type=sh > > ..git/rebase-apply/patch:94: space before tab in indent. > > extra_options="${extra_options} fused-madd.opt" > > warning: squelched 79 whitespace errors > > warning: 84 lines add whitespace errors. These are normal, AFAICT, and I also get them with the original patches, which apply fine. > > and the applied patches have context lines that should be an empty > > line with the one-space diff prefix replaced to empty lines (no > > leading diff space). Ah, that's quite unfortunate > > > > Can you verify whether this is expected or if something is getting > > munged in the mail? I can confirm that this happens after the email round-trip, but not before. > It looks like your mailer munged them sending the attachments as > message/rfc822 mime, which is lossy with respect to whitespace, rather > than text/plain or something that would be safely preserved. The > "trailing whitespace" on the context lines is missing in the actual > message I received. I'm sending them in a tarball now, in the hope that no mail software looks inside. Alternatively, you may: git pull https://codeberg.org/neuschaefer/musl-cross-make bump Best regards, jn [-- Attachment #2: patches.tgz --] [-- Type: application/x-gtar-compressed, Size: 16390 bytes --] ^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2024-10-25 11:39 UTC | newest] Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2024-10-02 15:40 [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions J. Neuschäfer 2024-10-02 15:40 ` [musl] [PATCH MCM 1/3] add gcc 12.4 hash and patches J. Neuschäfer 2024-10-02 15:40 ` [musl] [PATCH MCM 2/3] add gcc 13.3 " J. Neuschäfer 2024-10-02 15:40 ` [musl] [PATCH MCM 3/3] add gcc 14.2 " J. Neuschäfer 2024-10-22 21:52 ` [musl] [PATCH MCM 0/3] musl-cross-make: add newer GCC versions Rich Felker 2024-10-23 12:29 ` J. Neuschäfer 2024-10-24 13:11 ` Rich Felker 2024-10-24 13:36 ` Rich Felker 2024-10-25 11:39 ` J. Neuschäfer
Code repositories for project(s) associated with this public inbox https://git.vuxu.org/mirror/musl/ 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 NNTP newsgroup(s).