Commit a1e553aa authored by Alexandre Julliard's avatar Alexandre Julliard

vcomp: Move the fork wrapper to a separate file.

And build it as x86-64 code on ARM64EC.
parent 6af17eec
MODULE = vcomp.dll MODULE = vcomp.dll
SOURCES = \ SOURCES = \
fork.c \
main.c main.c
/*
* vcomp fork implementation
*
* Copyright 2012 Dan Kegel
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library; if not, write to the Free Software
* Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA
*/
#if 0
#pragma makedep arm64ec_x64
#endif
#include <stdarg.h>
#include "windef.h"
#include "wine/asm.h"
#ifdef __i386__
__ASM_GLOBAL_FUNC( _vcomp_fork_call_wrapper,
"pushl %ebp\n\t"
__ASM_CFI(".cfi_adjust_cfa_offset 4\n\t")
__ASM_CFI(".cfi_rel_offset %ebp,0\n\t")
"movl %esp,%ebp\n\t"
__ASM_CFI(".cfi_def_cfa_register %ebp\n\t")
"pushl %esi\n\t"
__ASM_CFI(".cfi_rel_offset %esi,-4\n\t")
"pushl %edi\n\t"
__ASM_CFI(".cfi_rel_offset %edi,-8\n\t")
"movl 12(%ebp),%edx\n\t"
"movl %esp,%edi\n\t"
"shll $2,%edx\n\t"
"jz 1f\n\t"
"subl %edx,%edi\n\t"
"andl $~15,%edi\n\t"
"movl %edi,%esp\n\t"
"movl 12(%ebp),%ecx\n\t"
"movl 16(%ebp),%esi\n\t"
"cld\n\t"
"rep; movsl\n"
"1:\tcall *8(%ebp)\n\t"
"leal -8(%ebp),%esp\n\t"
"popl %edi\n\t"
__ASM_CFI(".cfi_same_value %edi\n\t")
"popl %esi\n\t"
__ASM_CFI(".cfi_same_value %esi\n\t")
"popl %ebp\n\t"
__ASM_CFI(".cfi_def_cfa %esp,4\n\t")
__ASM_CFI(".cfi_same_value %ebp\n\t")
"ret" )
#elif defined __x86_64__
__ASM_GLOBAL_FUNC( _vcomp_fork_call_wrapper,
"pushq %rbp\n\t"
__ASM_SEH(".seh_pushreg %rbp\n\t")
__ASM_CFI(".cfi_adjust_cfa_offset 8\n\t")
__ASM_CFI(".cfi_rel_offset %rbp,0\n\t")
"movq %rsp,%rbp\n\t"
__ASM_SEH(".seh_setframe %rbp,0\n\t")
__ASM_CFI(".cfi_def_cfa_register %rbp\n\t")
"pushq %rsi\n\t"
__ASM_SEH(".seh_pushreg %rsi\n\t")
__ASM_CFI(".cfi_rel_offset %rsi,-8\n\t")
"pushq %rdi\n\t"
__ASM_SEH(".seh_pushreg %rdi\n\t")
__ASM_SEH(".seh_endprologue\n\t")
__ASM_CFI(".cfi_rel_offset %rdi,-16\n\t")
"movq %rcx,%rax\n\t"
"movq $4,%rcx\n\t"
"cmp %rcx,%rdx\n\t"
"cmovgq %rdx,%rcx\n\t"
"leaq 0(,%rcx,8),%rdx\n\t"
"subq %rdx,%rsp\n\t"
"andq $~15,%rsp\n\t"
"movq %rsp,%rdi\n\t"
"movq %r8,%rsi\n\t"
"rep; movsq\n\t"
"movq 0(%rsp),%rcx\n\t"
"movq 8(%rsp),%rdx\n\t"
"movq 16(%rsp),%r8\n\t"
"movq 24(%rsp),%r9\n\t"
"callq *%rax\n\t"
"leaq -16(%rbp),%rsp\n\t"
"popq %rdi\n\t"
__ASM_CFI(".cfi_same_value %rdi\n\t")
"popq %rsi\n\t"
__ASM_CFI(".cfi_same_value %rsi\n\t")
__ASM_CFI(".cfi_def_cfa_register %rsp\n\t")
"popq %rbp\n\t"
__ASM_CFI(".cfi_adjust_cfa_offset -8\n\t")
__ASM_CFI(".cfi_same_value %rbp\n\t")
"ret" )
#elif defined __arm__
__ASM_GLOBAL_FUNC( _vcomp_fork_call_wrapper,
"push {r4, r5, LR}\n\t"
"mov r4, r0\n\t"
"mov r5, SP\n\t"
"lsl r3, r1, #2\n\t"
"cmp r3, #0\n\t"
"beq 5f\n\t"
"sub SP, SP, r3\n\t"
"tst r1, #1\n\t"
"it eq\n\t"
"subeq SP, SP, #4\n\t"
"1:\tsub r3, r3, #4\n\t"
"ldr r0, [r2, r3]\n\t"
"str r0, [SP, r3]\n\t"
"cmp r3, #0\n\t"
"bgt 1b\n\t"
"cmp r1, #1\n\t"
"bgt 2f\n\t"
"pop {r0}\n\t"
"b 5f\n\t"
"2:\tcmp r1, #2\n\t"
"bgt 3f\n\t"
"pop {r0-r1}\n\t"
"b 5f\n\t"
"3:\tcmp r1, #3\n\t"
"bgt 4f\n\t"
"pop {r0-r2}\n\t"
"b 5f\n\t"
"4:\tpop {r0-r3}\n\t"
"5:\tblx r4\n\t"
"mov SP, r5\n\t"
"pop {r4, r5, PC}" )
#elif defined __aarch64__
__ASM_GLOBAL_FUNC( _vcomp_fork_call_wrapper,
"stp x29, x30, [SP,#-16]!\n\t"
".seh_save_fplr_x 16\n\t"
"mov x29, SP\n\t"
".seh_set_fp\n\t"
".seh_endprologue\n\t"
"mov x9, x0\n\t"
"cbz w1, 4f\n\t"
"lsl w8, w1, #3\n\t"
"cmp w8, #64\n\t"
"b.ge 1f\n\t"
"mov w8, #64\n"
"1:\ttbz w8, #3, 2f\n\t"
"add w8, w8, #8\n"
"2:\tsub x10, x29, x8\n\t"
"mov sp, x10\n"
"3:\tldr x0, [x2], #8\n\t"
"str x0, [x10], #8\n\t"
"subs w1, w1, #1\n\t"
"b.ne 3b\n\t"
"ldp x0, x1, [sp], #16\n\t"
"ldp x2, x3, [sp], #16\n\t"
"ldp x4, x5, [sp], #16\n\t"
"ldp x6, x7, [sp], #16\n"
"4:\tblr x9\n\t"
"mov SP, x29\n\t"
"ldp x29, x30, [SP], #16\n\t"
"ret" )
#endif
...@@ -120,6 +120,8 @@ struct vcomp_task_data ...@@ -120,6 +120,8 @@ struct vcomp_task_data
unsigned int dynamic_chunksize; unsigned int dynamic_chunksize;
}; };
extern void CDECL _vcomp_fork_call_wrapper(void *wrapper, int nargs, void **args);
static void **ptr_from_va_list(va_list valist) static void **ptr_from_va_list(va_list valist)
{ {
return *(void ***)&valist; return *(void ***)&valist;
...@@ -133,162 +135,6 @@ static void copy_va_list_data(void **args, va_list valist, int args_count) ...@@ -133,162 +135,6 @@ static void copy_va_list_data(void **args, va_list valist, int args_count)
args[i] = va_arg(valist, void *); args[i] = va_arg(valist, void *);
} }
#if defined(__i386__)
extern void CDECL _vcomp_fork_call_wrapper(void *wrapper, int nargs, void **args);
__ASM_GLOBAL_FUNC( _vcomp_fork_call_wrapper,
"pushl %ebp\n\t"
__ASM_CFI(".cfi_adjust_cfa_offset 4\n\t")
__ASM_CFI(".cfi_rel_offset %ebp,0\n\t")
"movl %esp,%ebp\n\t"
__ASM_CFI(".cfi_def_cfa_register %ebp\n\t")
"pushl %esi\n\t"
__ASM_CFI(".cfi_rel_offset %esi,-4\n\t")
"pushl %edi\n\t"
__ASM_CFI(".cfi_rel_offset %edi,-8\n\t")
"movl 12(%ebp),%edx\n\t"
"movl %esp,%edi\n\t"
"shll $2,%edx\n\t"
"jz 1f\n\t"
"subl %edx,%edi\n\t"
"andl $~15,%edi\n\t"
"movl %edi,%esp\n\t"
"movl 12(%ebp),%ecx\n\t"
"movl 16(%ebp),%esi\n\t"
"cld\n\t"
"rep; movsl\n"
"1:\tcall *8(%ebp)\n\t"
"leal -8(%ebp),%esp\n\t"
"popl %edi\n\t"
__ASM_CFI(".cfi_same_value %edi\n\t")
"popl %esi\n\t"
__ASM_CFI(".cfi_same_value %esi\n\t")
"popl %ebp\n\t"
__ASM_CFI(".cfi_def_cfa %esp,4\n\t")
__ASM_CFI(".cfi_same_value %ebp\n\t")
"ret" )
#elif defined(__x86_64__)
extern void CDECL _vcomp_fork_call_wrapper(void *wrapper, int nargs, void **args);
__ASM_GLOBAL_FUNC( _vcomp_fork_call_wrapper,
"pushq %rbp\n\t"
__ASM_SEH(".seh_pushreg %rbp\n\t")
__ASM_CFI(".cfi_adjust_cfa_offset 8\n\t")
__ASM_CFI(".cfi_rel_offset %rbp,0\n\t")
"movq %rsp,%rbp\n\t"
__ASM_SEH(".seh_setframe %rbp,0\n\t")
__ASM_CFI(".cfi_def_cfa_register %rbp\n\t")
"pushq %rsi\n\t"
__ASM_SEH(".seh_pushreg %rsi\n\t")
__ASM_CFI(".cfi_rel_offset %rsi,-8\n\t")
"pushq %rdi\n\t"
__ASM_SEH(".seh_pushreg %rdi\n\t")
__ASM_SEH(".seh_endprologue\n\t")
__ASM_CFI(".cfi_rel_offset %rdi,-16\n\t")
"movq %rcx,%rax\n\t"
"movq $4,%rcx\n\t"
"cmp %rcx,%rdx\n\t"
"cmovgq %rdx,%rcx\n\t"
"leaq 0(,%rcx,8),%rdx\n\t"
"subq %rdx,%rsp\n\t"
"andq $~15,%rsp\n\t"
"movq %rsp,%rdi\n\t"
"movq %r8,%rsi\n\t"
"rep; movsq\n\t"
"movq 0(%rsp),%rcx\n\t"
"movq 8(%rsp),%rdx\n\t"
"movq 16(%rsp),%r8\n\t"
"movq 24(%rsp),%r9\n\t"
"callq *%rax\n\t"
"leaq -16(%rbp),%rsp\n\t"
"popq %rdi\n\t"
__ASM_CFI(".cfi_same_value %rdi\n\t")
"popq %rsi\n\t"
__ASM_CFI(".cfi_same_value %rsi\n\t")
__ASM_CFI(".cfi_def_cfa_register %rsp\n\t")
"popq %rbp\n\t"
__ASM_CFI(".cfi_adjust_cfa_offset -8\n\t")
__ASM_CFI(".cfi_same_value %rbp\n\t")
"ret")
#elif defined(__arm__)
extern void CDECL _vcomp_fork_call_wrapper(void *wrapper, int nargs, void **args);
__ASM_GLOBAL_FUNC( _vcomp_fork_call_wrapper,
"push {r4, r5, LR}\n\t"
"mov r4, r0\n\t"
"mov r5, SP\n\t"
"lsl r3, r1, #2\n\t"
"cmp r3, #0\n\t"
"beq 5f\n\t"
"sub SP, SP, r3\n\t"
"tst r1, #1\n\t"
"it eq\n\t"
"subeq SP, SP, #4\n\t"
"1:\tsub r3, r3, #4\n\t"
"ldr r0, [r2, r3]\n\t"
"str r0, [SP, r3]\n\t"
"cmp r3, #0\n\t"
"bgt 1b\n\t"
"cmp r1, #1\n\t"
"bgt 2f\n\t"
"pop {r0}\n\t"
"b 5f\n\t"
"2:\tcmp r1, #2\n\t"
"bgt 3f\n\t"
"pop {r0-r1}\n\t"
"b 5f\n\t"
"3:\tcmp r1, #3\n\t"
"bgt 4f\n\t"
"pop {r0-r2}\n\t"
"b 5f\n\t"
"4:\tpop {r0-r3}\n\t"
"5:\tblx r4\n\t"
"mov SP, r5\n\t"
"pop {r4, r5, PC}" )
#elif defined(__aarch64__)
extern void CDECL _vcomp_fork_call_wrapper(void *wrapper, int nargs, void **args);
__ASM_GLOBAL_FUNC( _vcomp_fork_call_wrapper,
"stp x29, x30, [SP,#-16]!\n\t"
__ASM_SEH(".seh_save_fplr_x 16\n\t")
"mov x29, SP\n\t"
__ASM_SEH(".seh_set_fp\n\t")
__ASM_SEH(".seh_endprologue\n\t")
"mov x9, x0\n\t"
"cbz w1, 4f\n\t"
"lsl w8, w1, #3\n\t"
"cmp w8, #64\n\t"
"b.ge 1f\n\t"
"mov w8, #64\n"
"1:\ttbz w8, #3, 2f\n\t"
"add w8, w8, #8\n"
"2:\tsub x10, x29, x8\n\t"
"mov sp, x10\n"
"3:\tldr x0, [x2], #8\n\t"
"str x0, [x10], #8\n\t"
"subs w1, w1, #1\n\t"
"b.ne 3b\n\t"
"ldp x0, x1, [sp], #16\n\t"
"ldp x2, x3, [sp], #16\n\t"
"ldp x4, x5, [sp], #16\n\t"
"ldp x6, x7, [sp], #16\n"
"4:\tblr x9\n\t"
"mov SP, x29\n\t"
"ldp x29, x30, [SP], #16\n\t"
"ret\n" )
#else
static void CDECL _vcomp_fork_call_wrapper(void *wrapper, int nargs, void **args)
{
ERR("Not implemented for this architecture\n");
}
#endif
#if defined(__GNUC__) && (defined(__i386__) || defined(__x86_64__)) #if defined(__GNUC__) && (defined(__i386__) || defined(__x86_64__))
static inline char interlocked_cmpxchg8(char *dest, char xchg, char compare) static inline char interlocked_cmpxchg8(char *dest, char xchg, char compare)
......
...@@ -2,4 +2,5 @@ MODULE = vcomp100.dll ...@@ -2,4 +2,5 @@ MODULE = vcomp100.dll
PARENTSRC = ../vcomp PARENTSRC = ../vcomp
SOURCES = \ SOURCES = \
fork.c \
main.c main.c
...@@ -2,4 +2,5 @@ MODULE = vcomp110.dll ...@@ -2,4 +2,5 @@ MODULE = vcomp110.dll
PARENTSRC = ../vcomp PARENTSRC = ../vcomp
SOURCES = \ SOURCES = \
fork.c \
main.c main.c
...@@ -2,4 +2,5 @@ MODULE = vcomp120.dll ...@@ -2,4 +2,5 @@ MODULE = vcomp120.dll
PARENTSRC = ../vcomp PARENTSRC = ../vcomp
SOURCES = \ SOURCES = \
fork.c \
main.c main.c
...@@ -2,4 +2,5 @@ MODULE = vcomp140.dll ...@@ -2,4 +2,5 @@ MODULE = vcomp140.dll
PARENTSRC = ../vcomp PARENTSRC = ../vcomp
SOURCES = \ SOURCES = \
fork.c \
main.c main.c
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment