#define _LARGEFILE_SOURCE 1
#define _FILE_OFFSET_BITS 64
#define _GNU_SOURCE         /* See feature_test_macros(7) */
#include <stdlib.h>
#include <stdio.h>
#include <stddef.h>
#include <string.h>
#include <wchar.h>
#include <dlfcn.h>
#include <signal.h>
#include <errno.h>
#include <err.h>
#include <stdarg.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/select.h>
#include <unistd.h>
#include <fcntl.h>
#include <glob.h>
#include <ctype.h>
#include <dirent.h>
#include <search.h>
#include <sys/types.h>
#include <poll.h>
#include <sys/epoll.h>
#include <ftw.h>
#include <sys/syscall.h> 
#include <sys/socket.h>
#include <sys/utsname.h>
#include <sys/mman.h>
#include <sys/types.h>
#include <sys/ipc.h>
#include <sys/sem.h>
#include <setjmp.h>
#include <sys/vfs.h>
#include <spawn.h>
#include <fts.h>
#include <syslog.h>
#include <malloc.h>
#include <getopt.h>
#undef LOG_INFO
#undef LOG_DEBUG

#include "wrappedlibs.h"

#include "box64stack.h"
#include "x64emu.h"
#include "debug.h"
#include "wrapper.h"
#include "bridge.h"
#include "callback.h"
#include "librarian.h"
#include "librarian/library_private.h"
#include "emu/x64emu_private.h"
#include "box64context.h"
#include "myalign.h"
#include "signals.h"
#include "fileutils.h"
#include "auxval.h"
#include "elfloader.h"
#include "bridge.h"
#include "globalsymbols.h"

#define LIBNAME libc
const char* libcName = "libc.so.6";

typedef int (*iFi_t)(int);
typedef int (*iFL_t)(unsigned long);
typedef void (*vFpp_t)(void*, void*);
typedef void (*vFipp_t)(int32_t, void*, void*);
typedef int32_t (*iFpi_t)(void*, int32_t);
typedef int32_t (*iFpp_t)(void*, void*);
typedef int32_t (*iFpL_t)(void*, size_t);
typedef int32_t (*iFiip_t)(int32_t, int32_t, void*);
typedef int32_t (*iFipp_t)(int32_t, void*, void*);
typedef int32_t (*iFppi_t)(void*, void*, int32_t);
typedef int32_t (*iFpup_t)(void*, uint32_t, void*);
typedef int32_t (*iFpuu_t)(void*, uint32_t, uint32_t);
typedef int32_t (*iFiiII_t)(int, int, int64_t, int64_t);
typedef int32_t (*iFiiiV_t)(int, int, int, ...);
typedef int32_t (*iFippi_t)(int32_t, void*, void*, int32_t);
typedef int32_t (*iFpppp_t)(void*, void*, void*, void*);
typedef int32_t (*iFpipp_t)(void*, int32_t, void*, void*);
typedef int32_t (*iFppii_t)(void*, void*, int32_t, int32_t);
typedef int32_t (*iFipuu_t)(int32_t, void*, uint32_t, uint32_t);
typedef int32_t (*iFipiI_t)(int32_t, void*, int32_t, int64_t);
typedef int32_t (*iFipuup_t)(int32_t, void*, uint32_t, uint32_t, void*);
typedef int32_t (*iFiiV_t)(int32_t, int32_t, ...);
typedef void* (*pFp_t)(void*);

#define SUPER() \
    GO(_ITM_addUserCommitAction, iFpup_t)   \
    GO(_IO_file_stat, iFpp_t)


#include "wrappercallback.h"

static int regs_abi[] = {_DI, _SI, _DX, _CX, _R8, _R9};
void* getVargN(x64emu_t *emu, int n)
{
    if(n<6)
        return (void*)emu->regs[regs_abi[n]].q[0];
    return ((void**)R_RSP)[1+n-6];
}

// utility functions
#define SUPER() \
GO(0)   \
GO(1)   \
GO(2)   \
GO(3)   \
GO(4)   \
GO(5)   \
GO(6)   \
GO(7)   \
GO(8)   \
GO(9)   \
GO(10)  \
GO(11)  \
GO(12)  \
GO(13)  \
GO(14)  \
GO(15)

// compare
#define GO(A)   \
static uintptr_t my_compare_fct_##A = 0;        \
static int my_compare_##A(void* a, void* b)     \
{                                               \
    return (int)RunFunction(my_context, my_compare_fct_##A, 2, a, b);\
}
SUPER()
#undef GO
static void* findcompareFct(void* fct)
{
    if(!fct) return NULL;
    void* p;
    if((p = GetNativeFnc((uintptr_t)fct))) return p;
    #define GO(A) if(my_compare_fct_##A == (uintptr_t)fct) return my_compare_##A;
    SUPER()
    #undef GO
    #define GO(A) if(my_compare_fct_##A == 0) {my_compare_fct_##A = (uintptr_t)fct; return my_compare_##A; }
    SUPER()
    #undef GO
    printf_log(LOG_NONE, "Warning, no more slot for libc compare callback\n");
    return NULL;
}

// ftw64
#define GO(A)   \
static uintptr_t my_ftw64_fct_##A = 0;                      \
static int my_ftw64_##A(void* fpath, void* sb, int flag)    \
{                                                           \
    struct x64_stat64 x64st;                                \
    UnalignStat64(sb, &x64st);                              \
    return (int)RunFunction(my_context, my_ftw64_fct_##A, 3, fpath, &x64st, flag);  \
}
SUPER()
#undef GO
static void* findftw64Fct(void* fct)
{
    if(!fct) return NULL;
    #define GO(A) if(my_ftw64_fct_##A == (uintptr_t)fct) return my_ftw64_##A;
    SUPER()
    #undef GO
    #define GO(A) if(my_ftw64_fct_##A == 0) {my_ftw64_fct_##A = (uintptr_t)fct; return my_ftw64_##A; }
    SUPER()
    #undef GO
    printf_log(LOG_NONE, "Warning, no more slot for libc ftw64 callback\n");
    return NULL;
}

// nftw64
#define GO(A)   \
static uintptr_t my_nftw64_fct_##A = 0;                                     \
static int my_nftw64_##A(void* fpath, void* sb, int flag, void* ftwbuff)    \
{                                                                           \
    struct x64_stat64 x64st;                                              \
    UnalignStat64(sb, &x64st);                                             \
    return (int)RunFunction(my_context, my_nftw64_fct_##A, 4, fpath, &x64st, flag, ftwbuff);   \
}
SUPER()
#undef GO
static void* findnftw64Fct(void* fct)
{
    if(!fct) return NULL;
    #define GO(A) if(my_nftw64_fct_##A == (uintptr_t)fct) return my_nftw64_##A;
    SUPER()
    #undef GO
    #define GO(A) if(my_nftw64_fct_##A == 0) {my_nftw64_fct_##A = (uintptr_t)fct; return my_nftw64_##A; }
    SUPER()
    #undef GO
    printf_log(LOG_NONE, "Warning, no more slot for libc nftw64 callback\n");
    return NULL;
}
// globerr
#define GO(A)   \
static uintptr_t my_globerr_fct_##A = 0;                                        \
static int my_globerr_##A(void* epath, int eerrno)                              \
{                                                                               \
    return (int)RunFunction(my_context, my_globerr_fct_##A, 2, epath, eerrno);  \
}
SUPER()
#undef GO
static void* findgloberrFct(void* fct)
{
    if(!fct) return NULL;
    void* p;
    if((p = GetNativeFnc((uintptr_t)fct))) return p;
    #define GO(A) if(my_globerr_fct_##A == (uintptr_t)fct) return my_globerr_##A;
    SUPER()
    #undef GO
    #define GO(A) if(my_globerr_fct_##A == 0) {my_globerr_fct_##A = (uintptr_t)fct; return my_globerr_##A; }
    SUPER()
    #undef GO
    printf_log(LOG_NONE, "Warning, no more slot for libc globerr callback\n");
    return NULL;
}
// free
#define GO(A)   \
static uintptr_t my_free_fct_##A = 0;               \
static void my_free_##A(void* p)                    \
{                                                   \
    RunFunction(my_context, my_free_fct_##A, 1, p); \
}
SUPER()
#undef GO
static void* findfreeFct(void* fct)
{
    if(!fct) return NULL;
    void* p;
    if((p = GetNativeFnc((uintptr_t)fct))) return p;
    #define GO(A) if(my_free_fct_##A == (uintptr_t)fct) return my_free_##A;
    SUPER()
    #undef GO
    #define GO(A) if(my_free_fct_##A == 0) {my_free_fct_##A = (uintptr_t)fct; return my_free_##A; }
    SUPER()
    #undef GO
    printf_log(LOG_NONE, "Warning, no more slot for libc free callback\n");
    return NULL;
}

#if 0
#undef dirent
// filter_dir
#define GO(A)   \
static uintptr_t my_filter_dir_fct_##A = 0;                               \
static int my_filter_dir_##A(const struct dirent* a)                    \
{                                                                       \
    return (int)RunFunction(my_context, my_filter_dir_fct_##A, 1, a);     \
}
SUPER()
#undef GO
static void* findfilter_dirFct(void* fct)
{
    if(!fct) return NULL;
    void* p;
    if((p = GetNativeFnc((uintptr_t)fct))) return p;
    #define GO(A) if(my_filter_dir_fct_##A == (uintptr_t)fct) return my_filter_dir_##A;
    SUPER()
    #undef GO
    #define GO(A) if(my_filter_dir_fct_##A == 0) {my_filter_dir_fct_##A = (uintptr_t)fct; return my_filter_dir_##A; }
    SUPER()
    #undef GO
    printf_log(LOG_NONE, "Warning, no more slot for libc filter_dir callback\n");
    return NULL;
}
// compare_dir
#define GO(A)   \
static uintptr_t my_compare_dir_fct_##A = 0;                                  \
static int my_compare_dir_##A(const struct dirent* a, const struct dirent* b)    \
{                                                                           \
    return (int)RunFunction(my_context, my_compare_dir_fct_##A, 2, a, b);     \
}
SUPER()
#undef GO
static void* findcompare_dirFct(void* fct)
{
    if(!fct) return NULL;
    void* p;
    if((p = GetNativeFnc((uintptr_t)fct))) return p;
    #define GO(A) if(my_compare_dir_fct_##A == (uintptr_t)fct) return my_compare_dir_##A;
    SUPER()
    #undef GO
    #define GO(A) if(my_compare_dir_fct_##A == 0) {my_compare_dir_fct_##A = (uintptr_t)fct; return my_compare_dir_##A; }
    SUPER()
    #undef GO
    printf_log(LOG_NONE, "Warning, no more slot for libc compare_dir callback\n");
    return NULL;
}
#endif

// filter64
#define GO(A)   \
static uintptr_t my_filter64_fct_##A = 0;                               \
static int my_filter64_##A(const struct dirent64* a)                    \
{                                                                       \
    return (int)RunFunction(my_context, my_filter64_fct_##A, 1, a);     \
}
SUPER()
#undef GO
static void* findfilter64Fct(void* fct)
{
    if(!fct) return NULL;
    void* p;
    if((p = GetNativeFnc((uintptr_t)fct))) return p;
    #define GO(A) if(my_filter64_fct_##A == (uintptr_t)fct) return my_filter64_##A;
    SUPER()
    #undef GO
    #define GO(A) if(my_filter64_fct_##A == 0) {my_filter64_fct_##A = (uintptr_t)fct; return my_filter64_##A; }
    SUPER()
    #undef GO
    printf_log(LOG_NONE, "Warning, no more slot for libc filter64 callback\n");
    return NULL;
}
// compare64
#define GO(A)   \
static uintptr_t my_compare64_fct_##A = 0;                                      \
static int my_compare64_##A(const struct dirent64* a, const struct dirent64* b) \
{                                                                               \
    return (int)RunFunction(my_context, my_compare64_fct_##A, 2, a, b);         \
}
SUPER()
#undef GO
static void* findcompare64Fct(void* fct)
{
    if(!fct) return NULL;
    void* p;
    if((p = GetNativeFnc((uintptr_t)fct))) return p;
    #define GO(A) if(my_compare64_fct_##A == (uintptr_t)fct) return my_compare64_##A;
    SUPER()
    #undef GO
    #define GO(A) if(my_compare64_fct_##A == 0) {my_compare64_fct_##A = (uintptr_t)fct; return my_compare64_##A; }
    SUPER()
    #undef GO
    printf_log(LOG_NONE, "Warning, no more slot for libc compare64 callback\n");
    return NULL;
}

#undef SUPER

// some my_XXX declare and defines
int32_t my___libc_start_main(x64emu_t* emu, int *(main) (int, char * *, char * *), 
    int argc, char * * ubp_av, void (*init) (void), void (*fini) (void), 
    void (*rtld_fini) (void), void (* stack_end)); // implemented in x64run_private.c
EXPORT void my___libc_init_first(x64emu_t* emu, int argc, char* arg0, char** b)
{
    // do nothing specific for now
    (void)emu; (void)argc; (void)arg0; (void)b;
    return;
}
uintptr_t my_syscall(x64emu_t *emu); // implemented in x64syscall.c
void EXPORT my___stack_chk_fail(x64emu_t* emu)
{
    char buff[200];
    #ifdef HAVE_TRACE
    sprintf(buff, "%p: Stack is corrupted, aborting (prev IP=%p)\n", (void*)emu->old_ip, (void*)emu->prev2_ip);
    #else
    sprintf(buff, "%p: Stack is corrupted, aborting\n", (void*)emu->old_ip);
    #endif
    StopEmu(emu, buff);
}
void EXPORT my___gmon_start__(x64emu_t *emu)
{
    (void)emu;
    printf_log(LOG_DEBUG, "__gmon_start__ called (dummy call)\n");
}

int EXPORT my___cxa_atexit(x64emu_t* emu, void* p, void* a, void* dso_handle)
{
    (void)dso_handle;
    AddCleanup1Arg(emu, p, a);
    return 0;
}
void EXPORT my___cxa_finalize(x64emu_t* emu, void* p)
{
    if(!p) {
        // p is null, call (and remove) all Cleanup functions
        CallAllCleanup(emu);
        return;
    }
    CallCleanup(emu, p);
}
int EXPORT my_atexit(x64emu_t* emu, void *p)
{
    AddCleanup(emu, p);
    return 0;
}

int my_getcontext(x64emu_t* emu, void* ucp);
int my_setcontext(x64emu_t* emu, void* ucp);
int my_makecontext(x64emu_t* emu, void* ucp, void* fnc, int32_t argc, void* argv);
int my_swapcontext(x64emu_t* emu, void* ucp1, void* ucp2);

// All signal and context functions defined in signals.c

// All fts function defined in myfts.c

// getauxval implemented in auxval.c


// this one is defined in elfloader.c
int my_dl_iterate_phdr(x64emu_t *emu, void* F, void *data);

pid_t EXPORT my_fork(x64emu_t* emu)
{
    // execute atforks prepare functions, in reverse order
    for (int i=my_context->atfork_sz-1; i>=0; --i)
        if(my_context->atforks[i].prepare)
            RunFunctionWithEmu(emu, 0, my_context->atforks[i].prepare, 0);
    int type = emu->type;
    pid_t v;
    v = fork();
    if(type == EMUTYPE_MAIN)
        thread_set_emu(emu);
    if(v<0) {
        printf_log(LOG_NONE, "BOX64: Warning, fork errored... (%d)\n", v);
        // error...
    } else if(v>0) {  
        // execute atforks parent functions
        for (int i=0; i<my_context->atfork_sz; --i)
            if(my_context->atforks[i].parent)
                RunFunctionWithEmu(emu, 0, my_context->atforks[i].parent, 0);

    } else /*if(v==0)*/ {
        // execute atforks child functions
        for (int i=0; i<my_context->atfork_sz; --i)
            if(my_context->atforks[i].child)
                RunFunctionWithEmu(emu, 0, my_context->atforks[i].child, 0);
    }
    return v;
}
pid_t EXPORT my___fork(x64emu_t* emu) __attribute__((alias("my_fork")));
pid_t EXPORT my_vfork(x64emu_t* emu)
{
    #if 1
    emu->quit = 1;
    emu->fork = 1;  // use regular fork...
    return 0;
    #else
    return 0;
    #endif
}

int EXPORT my_uname(struct utsname *buf)
{
    //TODO: check sizeof(struct utsname) == 390
    int ret = uname(buf);
    strcpy(buf->machine, "x86_64");
    return ret;
}

// X86_O_RDONLY 0x00
#define X86_O_WRONLY       0x01     // octal     01
#define X86_O_RDWR         0x02     // octal     02
#define X86_O_CREAT        0x40     // octal     0100
#define X86_O_EXCL         0x80     // octal     0200
#define X86_O_NOCTTY       0x100    // octal     0400
#define X86_O_TRUNC        0x200    // octal    01000
#define X86_O_APPEND       0x400    // octal    02000
#define X86_O_NONBLOCK     0x800    // octal    04000
#define X86_O_SYNC         0x101000 // octal 04010000
#define X86_O_DSYNC        0x1000   // octal   010000
#define X86_O_RSYNC        O_SYNC
#define X86_FASYNC         020000
#define X86_O_DIRECT       040000
#define X86_O_LARGEFILE    0100000
#define X86_O_DIRECTORY    0200000
#define X86_O_NOFOLLOW     0400000
#define X86_O_NOATIME      01000000
#define X86_O_CLOEXEC      02000000
#define X86_O_PATH         010000000
#define X86_O_TMPFILE      020200000

#ifndef O_TMPFILE
#define O_TMPFILE (020000000 | O_DIRECTORY)
#endif

#define SUPER()     \
    GO(O_WRONLY)    \
    GO(O_RDWR)      \
    GO(O_CREAT)     \
    GO(O_EXCL)      \
    GO(O_NOCTTY)    \
    GO(O_TRUNC)     \
    GO(O_APPEND)    \
    GO(O_NONBLOCK)  \
    GO(O_SYNC)      \
    GO(O_DSYNC)     \
    GO(O_RSYNC)     \
    GO(FASYNC)      \
    GO(O_DIRECT)    \
    GO(O_LARGEFILE) \
    GO(O_TMPFILE)   \
    GO(O_DIRECTORY) \
    GO(O_NOFOLLOW)  \
    GO(O_NOATIME)   \
    GO(O_CLOEXEC)   \
    GO(O_PATH)      \

// x86->arm
int of_convert(int a)
{
    if(!a || a==-1) return a;
    int b=0;
    #define GO(A) if((a&(X86_##A))==(X86_##A)) {a&=~(X86_##A); b|=(A);}
    SUPER();
    #undef GO
    if(a) {
        printf_log(LOG_NONE, "Warning, of_convert(...) left over 0x%x, converted 0x%x\n", a, b);
    }
    return a|b;
}

// arm->x86
int of_unconvert(int a)
{
    if(!a || a==-1) return a;
    int b=0;
    #define GO(A) if((a&(A))==(A)) {a&=~(A); b|=(X86_##A);}
    SUPER();
    #undef GO
    // flags 0x20000 unknown?!
    if(a && (a&~0x20000)) {
        printf_log(LOG_NONE, "Warning, of_unconvert(...) left over 0x%x, converted 0x%x\n", a, b);
    }
    return a|b;
}
#undef SUPER

EXPORT void* my__ZGTtnaX (size_t a) { (void)a; printf("warning _ZGTtnaX called\n"); return NULL; }
EXPORT void my__ZGTtdlPv (void* a) { (void)a; printf("warning _ZGTtdlPv called\n"); }
EXPORT uint8_t my__ITM_RU1(const uint8_t * a) { (void)a; printf("warning _ITM_RU1 called\n"); return 0; }
EXPORT uint32_t my__ITM_RU4(const uint32_t * a) { (void)a; printf("warning _ITM_RU4 called\n"); return 0; }
EXPORT uint64_t my__ITM_RU8(const uint64_t * a) { (void)a; printf("warning _ITM_RU8 called\n"); return 0; }
EXPORT void my__ITM_memcpyRtWn(void * a, const void * b, size_t c) { (void)a; (void)b; (void)c; printf("warning _ITM_memcpyRtWn called\n"); }
EXPORT void my__ITM_memcpyRnWt(void * a, const void * b, size_t c) { (void)a; (void)b; (void)c; printf("warning _ITM_memcpyRnWt called\n"); }

EXPORT void my_longjmp(x64emu_t* emu, /*struct __jmp_buf_tag __env[1]*/void *p, int32_t __val);
EXPORT void my__longjmp(x64emu_t* emu, /*struct __jmp_buf_tag __env[1]*/void *p, int32_t __val) __attribute__((alias("my_longjmp")));
EXPORT void my_siglongjmp(x64emu_t* emu, /*struct __jmp_buf_tag __env[1]*/void *p, int32_t __val) __attribute__((alias("my_longjmp")));
EXPORT void my___longjmp_chk(x64emu_t* emu, /*struct __jmp_buf_tag __env[1]*/void *p, int32_t __val) __attribute__((alias("my_longjmp")));

//EXPORT int32_t my_setjmp(x64emu_t* emu, /*struct __jmp_buf_tag __env[1]*/void *p);
//EXPORT int32_t my__setjmp(x64emu_t* emu, /*struct __jmp_buf_tag __env[1]*/void *p) __attribute__((alias("my_setjmp")));
//EXPORT int32_t my___sigsetjmp(x64emu_t* emu, /*struct __jmp_buf_tag __env[1]*/void *p) __attribute__((alias("my_setjmp")));

EXPORT int my_printf(x64emu_t *emu, void* fmt, void* b) {
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 1);
    PREPARE_VALIST;
    return vprintf((const char*)fmt, VARARGS);
}
EXPORT int my___printf_chk(x64emu_t *emu, int chk, void* fmt, void* b)
{
    (void)chk;
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 2);
    PREPARE_VALIST;
    return vprintf((const char*)fmt, VARARGS);
}

EXPORT int my_vprintf(x64emu_t *emu, void* fmt, x64_va_list_t b) {
    (void)emu;
    #ifdef CONVERT_VALIST
    CONVERT_VALIST(b);
    #else
    myStackAlignValist(emu, (const char*)fmt, emu->scratch, b);
    PREPARE_VALIST;
    #endif
    return vprintf(fmt, VARARGS);
}
EXPORT int my___vprintf_chk(x64emu_t *emu, void* fmt, x64_va_list_t b) __attribute__((alias("my_vprintf")));

EXPORT int my_vfprintf(x64emu_t *emu, void* F, void* fmt, x64_va_list_t b) {
    (void)emu;
    #ifdef CONVERT_VALIST
    CONVERT_VALIST(b);
    #else
    myStackAlignValist(emu, (const char*)fmt, emu->scratch, b);
    PREPARE_VALIST;
    #endif
    return vfprintf(F, fmt, VARARGS);
}
EXPORT int my___vfprintf_chk(x64emu_t *emu, void* F, void* fmt, x64_va_list_t b) __attribute__((alias("my_vfprintf")));
EXPORT int my__IO_vfprintf(x64emu_t *emu, void* F, void* fmt, x64_va_list_t b) __attribute__((alias("my_vfprintf")));

EXPORT int my_fprintf(x64emu_t *emu, void* F, void* fmt, void* b)  {
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 2);
    PREPARE_VALIST;
    return vfprintf(F, fmt, VARARGS);
}
EXPORT int my___fprintf_chk(x64emu_t *emu, void* F, int flag, void* fmt, void* b)  {
    (void)flag;
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 3);
    PREPARE_VALIST;
    return vfprintf(F, fmt, VARARGS);
}

#if 0
EXPORT int my_wprintf(x64emu_t *emu, void* fmt, void* b, va_list V) {
    #ifndef NOALIGN
    // need to align on arm
    myStackAlignW((const char*)fmt, b, emu->scratch);
    PREPARE_VALIST;
    void* f = vwprintf;
    return ((iFpp_t)f)(fmt, VARARGS);
    #else
    // other platform don't need that
    return vwprintf((const wchar_t*)fmt, V);
    #endif
}
EXPORT int my___wprintf_chk(x64emu_t *emu, int flag, void* fmt, void* b, va_list V) {
    #ifndef NOALIGN
    // need to align on arm
    myStackAlignW((const char*)fmt, b, emu->scratch);
    PREPARE_VALIST;
    void* f = vwprintf;
    return ((iFpp_t)f)(fmt, VARARGS);
    #else
    // other platform don't need that
    return vwprintf((const wchar_t*)fmt, V);
    #endif
}
#endif

EXPORT int my_fwprintf(x64emu_t *emu, void* F, void* fmt, void* b)  {
    myStackAlignW(emu, (const char*)fmt, b, emu->scratch, R_EAX, 2);
    PREPARE_VALIST;
    return vfwprintf(F, fmt, VARARGS);
}

#if 0
EXPORT int my___fwprintf_chk(x64emu_t *emu, void* F, void* fmt, void* b, va_list V) __attribute__((alias("my_fwprintf")));

EXPORT int my_vfwprintf(x64emu_t *emu, void* F, void* fmt, void* b) {
    #ifndef NOALIGN
    myStackAlignW((const char*)fmt, b, emu->scratch);
    PREPARE_VALIST;
    void* f = vfwprintf;
    return ((iFppp_t)f)(F, fmt, VARARGS);
    #else
    void* f = vfwprintf;
    return ((iFppp_t)f)(F, fmt, b);
    #endif
}

EXPORT int my_vwprintf(x64emu_t *emu, void* fmt, void* b) {
    #ifndef NOALIGN
    myStackAlignW((const char*)fmt, b, emu->scratch);
    PREPARE_VALIST;
    void* f = vwprintf;
    return ((iFpp_t)f)(fmt, VARARGS);
    #else
    void* f = vwprintf;
    return ((iFpp_t)f)(fmt, b);
    #endif
}

EXPORT void *my_div(void *result, int numerator, int denominator) {
    *(div_t *)result = div(numerator, denominator);
    return result;
}
#endif

EXPORT int my_snprintf(x64emu_t* emu, void* buff, size_t s, void * fmt, uint64_t * b) {
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 3);
    PREPARE_VALIST;
    int r = vsnprintf(buff, s, fmt, VARARGS);
    return r;
}
EXPORT int my___snprintf(x64emu_t* emu, void* buff, size_t s, void * fmt, uint64_t * b) __attribute__((alias("my_snprintf")));
EXPORT int my___snprintf_chk(x64emu_t* emu, void* buff, size_t s, int flags, size_t maxlen, void * fmt, uint64_t * b)
{
    (void)flags; (void)maxlen;
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 5);
    PREPARE_VALIST;
    int r = vsnprintf(buff, s, fmt, VARARGS);
    return r;
}

EXPORT int my_sprintf(x64emu_t* emu, void* buff, void * fmt, void * b) {
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 2);
    PREPARE_VALIST;
    return vsprintf(buff, (const char*)fmt, VARARGS);
}
EXPORT int my___sprintf_chk(x64emu_t* emu, void* buff, int flag, size_t l, void * fmt, void * b) {
    (void)flag; (void)l;
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 4);
    PREPARE_VALIST;
    return vsprintf(buff, (const char*)fmt, VARARGS);
}

EXPORT int my_asprintf(x64emu_t* emu, void** buff, void * fmt, uint64_t * b) {
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 2);
    PREPARE_VALIST;
    return vasprintf((char**)buff, (char*)fmt, VARARGS);
}
EXPORT int my___asprintf(x64emu_t* emu, void** buff, void * fmt, uint64_t * b) __attribute__((alias("my_asprintf")));

EXPORT int my_vasprintf(x64emu_t* emu, char** buff, void* fmt, x64_va_list_t b) {
    (void)emu;
    #ifdef CONVERT_VALIST
    CONVERT_VALIST(b);
    #else
    myStackAlignValist(emu, (const char*)fmt, emu->scratch, b);
    PREPARE_VALIST;
    #endif
    return vasprintf(buff, fmt, VARARGS);
}

EXPORT int my_vsprintf(x64emu_t* emu, void* buff,  void * fmt, x64_va_list_t b) {
    (void)emu;
    #ifdef CONVERT_VALIST
    CONVERT_VALIST(b);
    #else
    myStackAlignValist(emu, (const char*)fmt, emu->scratch, b);
    PREPARE_VALIST;
    #endif
    return vsprintf(buff, fmt, VARARGS);
}
EXPORT int my___vsprintf_chk(x64emu_t* emu, void* buff, void * fmt, x64_va_list_t b) __attribute__((alias("my_vsprintf")));

EXPORT int my_vfscanf(x64emu_t* emu, void* stream, void* fmt, x64_va_list_t b)
{
    (void)emu;
    #ifdef CONVERT_VALIST
    CONVERT_VALIST(b);
    #else
    myStackAlignScanfValist(emu, (const char*)fmt, emu->scratch, b);
    PREPARE_VALIST;
    #endif
    return vfscanf(stream, fmt, VARARGS);
}

EXPORT int my_vsscanf(x64emu_t* emu, void* stream, void* fmt, x64_va_list_t b)
{
    (void)emu;
    #ifdef CONVERT_VALIST
    CONVERT_VALIST(b);
    #else
    myStackAlignScanfValist(emu, (const char*)fmt, emu->scratch, b);
    PREPARE_VALIST;
    #endif
    return vsscanf(stream, fmt, VARARGS);
}

EXPORT int my__vsscanf(x64emu_t* emu, void* stream, void* fmt, void* b) __attribute__((alias("my_vsscanf")));

EXPORT int my_vswscanf(x64emu_t* emu, void* stream, void* fmt, x64_va_list_t b)
{
    (void)emu;
    #ifdef CONVERT_VALIST
    CONVERT_VALIST(b);
    #else
    myStackAlignScanfWValist(emu, (const char*)fmt, emu->scratch, b);
    PREPARE_VALIST;
    #endif
    return vswscanf(stream, fmt, VARARGS);
}

EXPORT int my_sscanf(x64emu_t* emu, void* stream, void* fmt, uint64_t* b)
{
    myStackAlignScanf(emu, (const char*)fmt, b, emu->scratch, 2);
    PREPARE_VALIST;

    return vsscanf(stream, fmt, VARARGS);
}
EXPORT int my__IO_vfscanf(x64emu_t* emu, void* stream, void* fmt, void* b) __attribute__((alias("my_vfscanf")));
EXPORT int my___isoc99_vsscanf(x64emu_t* emu, void* stream, void* fmt, void* b) __attribute__((alias("my_vsscanf")));
EXPORT int my___isoc99_vfscanf(x64emu_t* emu, void* stream, void* fmt, void* b) __attribute__((alias("my_vfscanf")));

EXPORT int my___isoc99_fscanf(x64emu_t* emu, void* stream, void* fmt, uint64_t* b)
{
  myStackAlignScanf(emu, (const char*)fmt, b, emu->scratch, 2);
  PREPARE_VALIST;

  return vfscanf(stream, fmt, VARARGS);
}
EXPORT int my_fscanf(x64emu_t* emu, void* stream, void* fmt, uint64_t* b) __attribute__((alias("my___isoc99_fscanf")));

EXPORT int my___isoc99_scanf(x64emu_t* emu, void* fmt, uint64_t* b)
{
  myStackAlignScanf(emu, (const char*)fmt, b, emu->scratch, 1);
  PREPARE_VALIST;

  return vscanf(fmt, VARARGS);
}

EXPORT int my___isoc99_sscanf(x64emu_t* emu, void* stream, void* fmt, uint64_t* b)
{
  myStackAlignScanf(emu, (const char*)fmt, b, emu->scratch, 2);
  PREPARE_VALIST;

  return vsscanf(stream, fmt, VARARGS);
}

EXPORT int my_vsnprintf(x64emu_t* emu, void* buff, size_t s, void * fmt, x64_va_list_t b) {
    (void)emu;
    #ifdef CONVERT_VALIST
    CONVERT_VALIST(b);
    #else
    myStackAlignValist(emu, (const char*)fmt, emu->scratch, b);
    PREPARE_VALIST;
    #endif
    int r = vsnprintf(buff, s, fmt, VARARGS);
    return r;
}
EXPORT int my___vsnprintf(x64emu_t* emu, void* buff, size_t s, void * fmt, x64_va_list_t b) __attribute__((alias("my_vsnprintf")));
EXPORT int my___vsnprintf_chk(x64emu_t* emu, void* buff, size_t s, void * fmt, x64_va_list_t b) __attribute__((alias("my_vsnprintf")));
#if 0
EXPORT int my_vasprintf(x64emu_t* emu, void* strp, void* fmt, void* b, va_list V)
{
    #ifndef NOALIGN
    // need to align on arm
    myStackAlign((const char*)fmt, (uint32_t*)b, emu->scratch);
    PREPARE_VALIST;
    void* f = vasprintf;
    int r = ((iFppp_t)f)(strp, fmt, VARARGS);
    return r;
    #else
    void* f = vasprintf;
    int r = ((iFppp_t)f)(strp, fmt, (uint32_t*)b);
    return r;
    #endif
}
#endif
EXPORT int my___vasprintf_chk(x64emu_t* emu, void* buff, int flags, void* fmt, x64_va_list_t b)
{
    (void)emu; (void)flags;
    #ifdef CONVERT_VALIST
    CONVERT_VALIST(b);
    #else
    myStackAlignValist(emu, (const char*)fmt, emu->scratch, b);
    PREPARE_VALIST;
    #endif
    int r = vasprintf(buff, fmt, VARARGS);
    return r;
}
EXPORT int my___asprintf_chk(x64emu_t* emu, void* result_ptr, int flags, void* fmt, void* b)
{
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 3);
    PREPARE_VALIST;
    return vasprintf((char**)result_ptr, (char*)fmt, VARARGS);
}
EXPORT int my_vswprintf(x64emu_t* emu, void* buff, size_t s, void * fmt, x64_va_list_t b) {
    (void)emu;
    #ifdef CONVERT_VALIST
    CONVERT_VALIST(b);
    #else
    myStackAlignWValist(emu, (const char*)fmt, emu->scratch, b);
    PREPARE_VALIST;
    #endif
    int r = vswprintf(buff, s, fmt, VARARGS);
    return r;
}
EXPORT int my___vswprintf(x64emu_t* emu, void* buff, size_t s, void * fmt, x64_va_list_t b) __attribute__((alias("my_vswprintf")));
EXPORT int my___vswprintf_chk(x64emu_t* emu, void* buff, size_t s, void * fmt, x64_va_list_t b) __attribute__((alias("my_vswprintf")));

EXPORT int my_swscanf(x64emu_t* emu, void* stream, void* fmt, uint64_t* b)
{
    myStackAlignScanfW(emu, (const char*)fmt, b, emu->scratch, 2);
    PREPARE_VALIST;

    return vswscanf(stream, fmt, VARARGS);
}

#if 0
EXPORT void my_verr(x64emu_t* emu, int eval, void* fmt, void* b) {
    #ifndef NOALIGN
    myStackAlignW((const char*)fmt, (uint32_t*)b, emu->scratch);
    PREPARE_VALIST;
    void* f = verr;
    ((vFipp_t)f)(eval, fmt, VARARGS);
    #else
    void* f = verr;
    ((vFipp_t)f)(eval, fmt, (uint32_t*)b);
    #endif
}

EXPORT void my_vwarn(x64emu_t* emu, void* fmt, void* b) {
    #ifndef NOALIGN
    myStackAlignW((const char*)fmt, (uint32_t*)b, emu->scratch);
    PREPARE_VALIST;
    void* f = vwarn;
    ((vFpp_t)f)(fmt, VARARGS);
    #else
    void* f = vwarn;
    ((vFpp_t)f)(fmt, (uint32_t*)b);
    #endif
}
#endif
EXPORT void my_err(x64emu_t *emu, int eval, void* fmt, void* b) {
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 2);
    PREPARE_VALIST;
    verr(eval, (const char*)fmt, VARARGS);
}
EXPORT void my_errx(x64emu_t *emu, int eval, void* fmt, void* b) {
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 2);
    PREPARE_VALIST;
    verrx(eval, (const char*)fmt, VARARGS);
}
EXPORT void my_warn(x64emu_t *emu, void* fmt, void* b) {
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 1);
    PREPARE_VALIST;
    vwarn((const char*)fmt, VARARGS);
}
EXPORT void my_warnx(x64emu_t *emu, void* fmt, void* b) {
    myStackAlign(emu, (const char*)fmt, b, emu->scratch, R_EAX, 1);
    PREPARE_VALIST;
    vwarnx((const char*)fmt, VARARGS);
}

EXPORT void my_syslog(x64emu_t* emu, int priority, const char* fmt, uint64_t* b)
{
    myStackAlign(emu, fmt, b, emu->scratch, R_EAX, 2);
    PREPARE_VALIST;
    return vsyslog(priority, fmt, VARARGS);
}
EXPORT void my___syslog_chk(x64emu_t* emu, int priority, int flags, const char* fmt, uint64_t* b)
{
    (void)flags;
    myStackAlign(emu, fmt, b, emu->scratch, R_EAX, 3);
    PREPARE_VALIST;
    return vsyslog(priority, fmt, VARARGS);
}
EXPORT void my_vsyslog(x64emu_t* emu, int priority, const char* fmt, x64_va_list_t b)
{
    (void)emu;
    #ifdef CONVERT_VALIST
    CONVERT_VALIST(b);
    #else
    myStackAlignValist(emu, (const char*)fmt, emu->scratch, b);
    PREPARE_VALIST;
    #endif
    return vsyslog(priority, fmt, VARARGS);
}
EXPORT void my___vsyslog_chk(x64emu_t* emu, int priority, int flag, const char* fmt, x64_va_list_t b)
{
    (void)emu;
    #ifdef CONVERT_VALIST
    CONVERT_VALIST(b);
    #else
    myStackAlignValist(emu, (const char*)fmt, emu->scratch, b);
    PREPARE_VALIST;
    #endif
    return vsyslog(priority, fmt, VARARGS);
}

EXPORT int my___swprintf_chk(x64emu_t* emu, void* s, size_t n, int32_t flag, size_t slen, void* fmt, uint64_t* b)
{
    (void)flag;
    (void)slen;
    myStackAlignW(emu, (const char*)fmt, b, emu->scratch, R_EAX, 5);
    PREPARE_VALIST;
    return vswprintf(s, n, (const wchar_t*)fmt, VARARGS);
}
EXPORT int my_swprintf(x64emu_t* emu, void* s, size_t n, void* fmt, uint64_t* b)
{
    myStackAlignW(emu, (const char*)fmt, b, emu->scratch, R_EAX, 3);
    PREPARE_VALIST;
    return vswprintf(s, n, (const wchar_t*)fmt, VARARGS);
}

EXPORT void my__ITM_addUserCommitAction(x64emu_t* emu, void* cb, uint32_t b, void* c)
{
    // disabled for now... Are all this _ITM_ stuff really mendatory?
    #if 0
    // quick and dirty... Should store the callback to be removed later....
    libc_my_t *my = (libc_my_t *)emu->context->libclib->priv.w.p2;
    x64emu_t *cbemu = AddCallback(emu, (uintptr_t)cb, 1, c, NULL, NULL, NULL);
    my->_ITM_addUserCommitAction(libc1ArgCallback, b, cbemu);
    // should keep track of cbemu to remove at some point...
    #else
    (void)emu; (void)cb; (void)b; (void)c;
    printf("warning _ITM_addUserCommitAction called\n");
    #endif
}
EXPORT void my__ITM_registerTMCloneTable(x64emu_t* emu, void* p, uint32_t s) { (void)emu; (void)p; (void)s; }
EXPORT void my__ITM_deregisterTMCloneTable(x64emu_t* emu, void* p) { (void)emu; (void)p; }


EXPORT int my___fxstat(x64emu_t *emu, int vers, int fd, void* buf)
{
    (void)emu; (void)vers;
    struct stat64 st;
    int r = fstat64(fd, &st);
    UnalignStat64(&st, buf);
    return r;
}

EXPORT int my___fxstat64(x64emu_t *emu, int vers, int fd, void* buf)
{
    (void)emu; (void)vers;
    struct stat64 st;
    int r = fstat64(fd, &st);
    //int r = syscall(__NR_stat64, fd, &st);
    UnalignStat64(&st, buf);
    return r;
}

EXPORT int my___xstat(x64emu_t* emu, int v, void* path, void* buf)
{
    (void)emu; (void)v;
    struct stat64 st;
    int r = stat64((const char*)path, &st);
    UnalignStat64(&st, buf);
    return r;
}

EXPORT int my___xstat64(x64emu_t* emu, int v, void* path, void* buf)
{
    (void)emu; (void)v;
    struct stat64 st;
    int r = stat64((const char*)path, &st);
    UnalignStat64(&st, buf);
    return r;
}

EXPORT int my___lxstat(x64emu_t* emu, int v, void* name, void* buf)
{
    (void)emu; (void)v;
    struct stat64 st;
    int r = lstat64((const char*)name, &st);
    UnalignStat64(&st, buf);
    return r;
}

EXPORT int my___lxstat64(x64emu_t* emu, int v, void* name, void* buf)
{
    (void)emu; (void)v;
    struct stat64 st;
    int r = lstat64((const char*)name, &st);
    UnalignStat64(&st, buf);
    return r;
}

EXPORT int my___fxstatat(x64emu_t* emu, int v, int d, void* path, void* buf, int flags)
{
    (void)emu; (void)v;
    struct  stat64 st;
    int r = fstatat64(d, path, &st, flags);
    UnalignStat64(&st, buf);
    return r;
}

EXPORT int my___fxstatat64(x64emu_t* emu, int v, int d, void* path, void* buf, int flags)
{
    (void)emu; (void)v;
    struct  stat64 st;
    int r = fstatat64(d, path, &st, flags);
    UnalignStat64(&st, buf);
    return r;
}

EXPORT int my_stat(x64emu_t *emu, void* filename, void* buf)
{
    (void)emu;
    struct stat64 st;
    int r = stat(filename, (struct stat*)&st);
    UnalignStat64(&st, buf);
    return r;
}

EXPORT int my_lstat(x64emu_t *emu, void* filename, void* buf)
{
    (void)emu;
    struct stat64 st;
    int r = lstat(filename, (struct stat*)&st);
    UnalignStat64(&st, buf);
    return r;
}

EXPORT int my_fstat(x64emu_t *emu, int fd, void* buf)
{
    (void)emu;
    struct stat64 st;
    int r = fstat(fd, (struct stat*)&st);
    UnalignStat64(&st, buf);
    return r;
}

EXPORT int my_fstatat(x64emu_t *emu, int fd, const char* path, void* buf, int flags)
{
    (void)emu;
    struct stat64 st;
    int r = fstatat(fd, path, (struct stat*)&st, flags);
    UnalignStat64(&st, buf);
    return r;
}

EXPORT int my__IO_file_stat(x64emu_t* emu, void* f, void* buf)
{
    struct stat64 st;
    int r = my->_IO_file_stat(f, &st);
    UnalignStat64(&st, buf);
    return r;
}

#if 0
EXPORT int my_fstatfs64(int fd, void* buf)
{
    struct statfs64 st;
    int r = fstatfs64(fd, &st);
    UnalignStatFS64(&st, buf);
    return r;
}

EXPORT int my_statfs64(const char* path, void* buf)
{
    struct statfs64 st;
    int r = statfs64(path, &st);
    UnalignStatFS64(&st, buf);
    return r;
}
#endif

typedef struct compare_r_s {
    x64emu_t* emu;
    uintptr_t f;
    void*     data;
    int       r;
} compare_r_t;

static int my_compare_r_cb(void* a, void* b, compare_r_t* arg)
{
    return (int)RunFunctionWithEmu(arg->emu, 0, arg->f, 2+arg->r, a, b, arg->data);
}
EXPORT void my_qsort(x64emu_t* emu, void* base, size_t nmemb, size_t size, void* fnc)
{
    compare_r_t args;
    args.emu = emu; args.f = (uintptr_t)fnc; args.r = 0; args.data = NULL;
    qsort_r(base, nmemb, size, (__compar_d_fn_t)my_compare_r_cb, &args);
}
EXPORT void my_qsort_r(x64emu_t* emu, void* base, size_t nmemb, size_t size, void* fnc, void* data)
{
    compare_r_t args;
    args.emu = emu; args.f = (uintptr_t)fnc; args.r = 1; args.data = data;
    qsort_r(base, nmemb, size, (__compar_d_fn_t)my_compare_r_cb, &args);
}
EXPORT void* my_bsearch(x64emu_t* emu, void* key, void* base, size_t nmemb, size_t size, void* fnc)
{
    (void)emu;
    return bsearch(key, base, nmemb, size, findcompareFct(fnc));
}

EXPORT void* my_lsearch(x64emu_t* emu, void* key, void* base, size_t* nmemb, size_t size, void* fnc)
{
    (void)emu;
    return lsearch(key, base, nmemb, size, findcompareFct(fnc));
}

EXPORT void* my_tsearch(x64emu_t* emu, void* key, void* root, void* fnc)
{
    (void)emu;
    return tsearch(key, root, findcompareFct(fnc));
}
EXPORT void my_tdestroy(x64emu_t* emu, void* root, void* fnc)
{
    (void)emu;
    return tdestroy(root, findfreeFct(fnc));
}
EXPORT void* my_lfind(x64emu_t* emu, void* key, void* base, size_t* nmemb, size_t size, void* fnc)
{
    (void)emu;
    return lfind(key, base, nmemb, size, findcompareFct(fnc));
}

EXPORT void* my_fts_open(x64emu_t* emu, void* path, int options, void* c)
{
    (void)emu;
    return fts_open(path, options, findcompareFct(c));
}

#if 0
struct i386_dirent {
    uint32_t d_ino;
    int32_t  d_off;
    uint16_t d_reclen;
    uint8_t  d_type;
    char     d_name[256];
};

EXPORT void* my_readdir(x64emu_t* emu, void* dirp)
{
    if (fix_64bit_inodes)
    {
        struct dirent64 *dp64 = readdir64((DIR *)dirp);
        if (!dp64) return NULL;
        uint32_t ino32 = dp64->d_ino ^ (dp64->d_ino >> 32);
        int32_t off32 = dp64->d_off;
        struct i386_dirent *dp32 = (struct i386_dirent *)&(dp64->d_off);
        dp32->d_ino = ino32;
        dp32->d_off = off32;
        dp32->d_reclen -= 8;
        return dp32;
    }
    else
    {
        static pFp_t f = NULL;
        if(!f) {
            library_t* lib = my_lib;
            if(!lib) return NULL;
            f = (pFp_t)dlsym(lib->priv.w.lib, "readdir");
        }

        return f(dirp);
    }
}

EXPORT int32_t my_readdir_r(x64emu_t* emu, void* dirp, void* entry, void** result)
{
    struct dirent64 d64, *dp64;
    if (fix_64bit_inodes && (sizeof(d64.d_name) > 1))
    {
        static iFppp_t f = NULL;
        if(!f) {
            library_t* lib = my_lib;
            if(!lib)
            {
                *result = NULL;
                return 0;
            }
            f = (iFppp_t)dlsym(lib->priv.w.lib, "readdir64_r");
        }

        int r = f(dirp, &d64, &dp64);
        if (r || !dp64 || !entry)
        {
            *result = NULL;
            return r;
        }

        struct i386_dirent *dp32 = (struct i386_dirent *)entry;
        int namelen = dp64->d_reclen - offsetof(struct dirent64, d_name);
        if (namelen > sizeof(dp32->d_name))
        {
            *result = NULL;
            return ENAMETOOLONG;
        }

        dp32->d_ino = dp64->d_ino ^ (dp64->d_ino >> 32);
        dp32->d_off = dp64->d_off;
        dp32->d_reclen = namelen + offsetof(struct i386_dirent, d_name);
        dp32->d_type = dp64->d_type;
        memcpy(dp32->d_name, dp64->d_name, namelen);
        *result = dp32;
        return 0;
    }
    else
    {
        static iFppp_t f = NULL;
        if(!f) {
            library_t* lib = my_lib;
            if(!lib)
            {
                *result = NULL;
                return 0;
            }
            f = (iFppp_t)dlsym(lib->priv.w.lib, "readdir_r");
        }

        return f(dirp, entry, result);
    }
}
#endif

static int isProcSelf(const char *path, const char* w)
{
    if(strncmp(path, "/proc/", 6)==0) {
        char tmp[64];
        // check if self ....
        sprintf(tmp, "/proc/self/%s", w);
        if(strcmp((const char*)path, tmp)==0)
            return 1;
        // check if self PID ....
        pid_t pid = getpid();
        sprintf(tmp, "/proc/%d/%s", pid, w);
        if(strcmp((const char*)path, tmp)==0)
            return 1;
    }
    return 0;
}

EXPORT ssize_t my_readlink(x64emu_t* emu, void* path, void* buf, size_t sz)
{
    if(isProcSelf((const char*)path, "exe")) {
        // special case for self...
        return strlen(strncpy((char*)buf, emu->context->fullpath, sz));
    }
    return readlink((const char*)path, (char*)buf, sz);
}

static int nCPU = 0;
static double bogoMips = 100.;

void grabNCpu() {
    nCPU = 1;  // default number of CPU to 1
    FILE *f = fopen("/proc/cpuinfo", "r");
    size_t dummy;
    if(f) {
        nCPU = 0;
        size_t len = 0;
        char* line = NULL;
        while ((dummy = getline(&line, &len, f)) != -1) {
            if(!strncmp(line, "processor\t", strlen("processor\t")))
                ++nCPU;
            if(!nCPU && !strncmp(line, "BogoMIPS\t", strlen("BogoMIPS\t"))) {
                // grab 1st BogoMIPS
                float tmp;
                if(sscanf(line, "BogoMIPS\t: %g", &tmp)==1)
                    bogoMips = tmp;
            }
        }
        if(line) free(line);
        fclose(f);
        if(!nCPU) nCPU=1;
    }
}
int getNCpu()
{
    if(!nCPU)
        grabNCpu();
    return nCPU;
}
#ifndef NOALIGN


void CreateCPUInfoFile(int fd)
{
    size_t dummy;
    char buff[600];
    double freq = 600.0; // default to 600 MHz
    // try to get actual ARM max speed:
    FILE *f = fopen("/sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_max_freq", "r");
    if(f) {
        int r;
        if(1==fscanf(f, "%d", &r))
            freq = r/1000.;
        fclose(f);
    }
    int n = getNCpu();
    // generate fake CPUINFO
    int gigahertz=(freq>=1000.);
    #define P \
    dummy = write(fd, buff, strlen(buff))
    for (int i=0; i<n; ++i) {
        sprintf(buff, "processor\t: %d\n", i);
        P;
        sprintf(buff, "vendor_id\t: GenuineIntel\n");
        P;
        sprintf(buff, "cpu family\t: 6\n");
        P;
        sprintf(buff, "model\t\t: 1\n");
        P;
        sprintf(buff, "model name\t: Intel Pentium IV @ %g%cHz\n", gigahertz?(freq/1000.):freq, gigahertz?'G':'M');
        P;
        sprintf(buff, "stepping\t: 1\nmicrocode\t: 0x10\n");
        P;
        sprintf(buff, "cpu MHz\t\t: %g\n", freq);
        P;
        sprintf(buff, "cache size\t: %d\n", 4096);
        P;
        sprintf(buff, "physical id\t: %d\nsiblings\t: %d\n", i, n);
        P;
        sprintf(buff, "core id\t\t:%d\ncpu cores\t: %d\n", i, 1);
        P;
        sprintf(buff, "bogomips\t: %g\n", bogoMips);
        P;
        sprintf(buff, "flags\t\t: fpu cx8 sep cmov clflush mmx sse sse2 syscall tsc lahf_lm ssse3 ht tm lm fma fxsr cpuid cx16 aes movbe pni\n");
        P;
        sprintf(buff, "address sizes\t: 46 bits physical, 48 bits virtual\n");
        P;
        sprintf(buff, "\n");
        P;
    }
    (void)dummy;
    #undef P
}

#define TMP_CPUINFO "box64_tmpcpuinfo"
#define TMP_CPUTOPO "box64_tmpcputopo%d"
#endif
#define TMP_MEMMAP  "box64_tmpmemmap"
#define TMP_CMDLINE "box64_tmpcmdline"
EXPORT int32_t my_open(x64emu_t* emu, void* pathname, int32_t flags, uint32_t mode)
{
    if(isProcSelf((const char*) pathname, "cmdline")) {
        // special case for self command line...
        #if 0
        char tmpcmdline[200] = {0};
        char tmpbuff[100] = {0};
        sprintf(tmpbuff, "%s/cmdlineXXXXXX", getenv("TMP")?getenv("TMP"):".");
        int tmp = mkstemp(tmpbuff);
        int dummy;
        if(tmp<0) return open(pathname, flags, mode);
        dummy = write(tmp, emu->context->fullpath, strlen(emu->context->fullpath)+1);
        for (int i=1; i<emu->context->argc; ++i)
            dummy = write(tmp, emu->context->argv[i], strlen(emu->context->argv[i])+1);
        lseek(tmp, 0, SEEK_SET);
        #else
        int tmp = shm_open(TMP_CMDLINE, O_RDWR | O_CREAT, S_IRWXU);
        if(tmp<0) return open(pathname, flags, mode);
        shm_unlink(TMP_CMDLINE);    // remove the shm file, but it will still exist because it's currently in use
        int dummy = write(tmp, emu->context->fullpath, strlen(emu->context->fullpath)+1);
        (void)dummy;
        for (int i=1; i<emu->context->argc; ++i)
            dummy = write(tmp, emu->context->argv[i], strlen(emu->context->argv[i])+1);
        lseek(tmp, 0, SEEK_SET);
        #endif
        return tmp;
    }
    if(isProcSelf((const char*)pathname, "exe")) {
        return open(emu->context->fullpath, flags, mode);
    }
    #ifndef NOALIGN
    if(strcmp((const char*)pathname, "/proc/cpuinfo")==0) {
        // special case for cpuinfo
        int tmp = shm_open(TMP_CPUINFO, O_RDWR | O_CREAT, S_IRWXU);
        if(tmp<0) return open(pathname, flags, mode); // error fallback
        shm_unlink(TMP_CPUINFO);    // remove the shm file, but it will still exist because it's currently in use
        CreateCPUInfoFile(tmp);
        lseek(tmp, 0, SEEK_SET);
        return tmp;
    }
    #endif
    int ret = open(pathname, flags, mode);
    return ret;
}
EXPORT int32_t my___open(x64emu_t* emu, void* pathname, int32_t flags, uint32_t mode) __attribute__((alias("my_open")));

//#ifdef DYNAREC
//static int hasDBFromAddress(uintptr_t addr)
//{
//    int idx = (addr>>DYNAMAP_SHIFT);
//    return getDB(idx)?1:0;
//}
//#endif

//EXPORT int32_t my_read(int fd, void* buf, uint32_t count)
//{
//    int ret = read(fd, buf, count);
//#ifdef DYNAREC
//    if(ret!=count && ret>0) {
//        // continue reading...
//        void* p = buf+ret;
//        if(hasDBFromAddress((uintptr_t)p)) {
//            // allow writing the whole block (this happens with HalfLife, libMiles load code directly from .mix and other file like that)
//            unprotectDB((uintptr_t)p, count-ret);
//            int l;
//            do {
//                l = read(fd, p, count-ret); 
//                if(l>0) {
//                    p+=l; ret+=l;
//                }
//            } while(l>0);
//        }
//    }
//#endif
//    return ret;
//}

EXPORT int32_t my_open64(x64emu_t* emu, void* pathname, int32_t flags, uint32_t mode)
{
    if(isProcSelf((const char*)pathname, "cmdline")) {
        // special case for self command line...
        #if 0
        char tmpcmdline[200] = {0};
        char tmpbuff[100] = {0};
        sprintf(tmpbuff, "%s/cmdlineXXXXXX", getenv("TMP")?getenv("TMP"):".");
        int tmp = mkstemp64(tmpbuff);
        int dummy;
        if(tmp<0) return open64(pathname, flags, mode);
        dummy = write(tmp, emu->context->fullpath, strlen(emu->context->fullpath)+1);
        for (int i=1; i<emu->context->argc; ++i)
            dummy = write(tmp, emu->context->argv[i], strlen(emu->context->argv[i])+1);
        lseek64(tmp, 0, SEEK_SET);
        #else
        int tmp = shm_open(TMP_CMDLINE, O_RDWR | O_CREAT, S_IRWXU);
        if(tmp<0) return open64(pathname, flags, mode);
        shm_unlink(TMP_CMDLINE);    // remove the shm file, but it will still exist because it's currently in use
        int dummy = write(tmp, emu->context->fullpath, strlen(emu->context->fullpath)+1);
        (void)dummy;
        for (int i=1; i<emu->context->argc; ++i)
            dummy = write(tmp, emu->context->argv[i], strlen(emu->context->argv[i])+1);
        lseek(tmp, 0, SEEK_SET);
        #endif
        return tmp;
    }
    if(isProcSelf((const char*)pathname, "exe")) {
        return open64(emu->context->fullpath, flags, mode);
    }
    #ifndef NOALIGN
    if(strcmp((const char*)pathname, "/proc/cpuinfo")==0) {
        // special case for cpuinfo
        int tmp = shm_open(TMP_CPUINFO, O_RDWR | O_CREAT, S_IRWXU);
        if(tmp<0) return open64(pathname, flags, mode); // error fallback
        shm_unlink(TMP_CPUINFO);    // remove the shm file, but it will still exist because it's currently in use
        CreateCPUInfoFile(tmp);
        lseek(tmp, 0, SEEK_SET);
        return tmp;
    }
    #endif
    return open64(pathname, flags, mode);
}

EXPORT FILE* my_fopen(x64emu_t* emu, const char* path, const char* mode)
{
    if(isProcSelf(path, "maps")) {
        // special case for self memory map
        int tmp = shm_open(TMP_MEMMAP, O_RDWR | O_CREAT, S_IRWXU);
        if(tmp<0) return fopen(path, mode); // error fallback
        shm_unlink(TMP_MEMMAP);    // remove the shm file, but it will still exist because it's currently in use
        CreateMemorymapFile(emu->context, tmp);
        lseek(tmp, 0, SEEK_SET);
        return fdopen(tmp, mode);
    }
    #ifndef NOALIGN
    if(strcmp(path, "/proc/cpuinfo")==0) {
        // special case for cpuinfo
        int tmp = shm_open(TMP_CPUINFO, O_RDWR | O_CREAT, S_IRWXU);
        if(tmp<0) return fopen(path, mode); // error fallback
        shm_unlink(TMP_CPUINFO);    // remove the shm file, but it will still exist because it's currently in use
        CreateCPUInfoFile(tmp);
        lseek(tmp, 0, SEEK_SET);
        return fdopen(tmp, mode);
    }
    #endif
    if(isProcSelf(path, "exe")) {
        return fopen(emu->context->fullpath, mode);
    }
    return fopen(path, mode);
}

EXPORT FILE* my_fopen64(x64emu_t* emu, const char* path, const char* mode)
{
    if(isProcSelf(path, "maps")) {
        // special case for self memory map
        int tmp = shm_open(TMP_MEMMAP, O_RDWR | O_CREAT, S_IRWXU);
        if(tmp<0) return fopen64(path, mode); // error fallback
        shm_unlink(TMP_MEMMAP);    // remove the shm file, but it will still exist because it's currently in use
        CreateMemorymapFile(emu->context, tmp);
        lseek(tmp, 0, SEEK_SET);
        return fdopen(tmp, mode);
    }
    #ifndef NOALIGN
    if(strcmp(path, "/proc/cpuinfo")==0) {
        // special case for cpuinfo
        int tmp = shm_open(TMP_CPUINFO, O_RDWR | O_CREAT, S_IRWXU);
        if(tmp<0) return fopen64(path, mode); // error fallback
        shm_unlink(TMP_CPUINFO);    // remove the shm file, but it will still exist because it's currently in use
        CreateCPUInfoFile(tmp);
        lseek(tmp, 0, SEEK_SET);
        return fdopen(tmp, mode);
    }
    #endif
    if(isProcSelf(path, "exe")) {
        return fopen64(emu->context->fullpath, mode);
    }
    return fopen64(path, mode);
}

#if 0
EXPORT int32_t my_ftw(x64emu_t* emu, void* pathname, void* B, int32_t nopenfd)
{
    static iFppi_t f = NULL;
    if(!f) {
        library_t* lib = my_lib;
        if(!lib) return 0;
        f = (iFppi_t)dlsym(lib->priv.w.lib, "ftw");
    }

    return f(pathname, findftwFct(B), nopenfd);
}

EXPORT int32_t my_nftw(x64emu_t* emu, void* pathname, void* B, int32_t nopenfd, int32_t flags)
{
    static iFppii_t f = NULL;
    if(!f) {
        library_t* lib = my_lib;
        if(!lib) return 0;
        f = (iFppii_t)dlsym(lib->priv.w.lib, "nftw");
    }

    return f(pathname, findnftwFct(B), nopenfd, flags);
}

EXPORT void* my_ldiv(x64emu_t* emu, void* p, int32_t num, int32_t den)
{
    *((ldiv_t*)p) = ldiv(num, den);
    return p;
}
#endif

#ifndef NOALIGN
EXPORT int32_t my_epoll_ctl(x64emu_t* emu, int32_t epfd, int32_t op, int32_t fd, void* event)
{
    struct epoll_event _event[1] = {0};
    if(event && (op!=EPOLL_CTL_DEL))
        AlignEpollEvent(_event, event, 1);
    return epoll_ctl(epfd, op, fd, event?_event:event);
}
EXPORT int32_t my_epoll_wait(x64emu_t* emu, int32_t epfd, void* events, int32_t maxevents, int32_t timeout)
{
    struct epoll_event _events[maxevents];
    //AlignEpollEvent(_events, events, maxevents);
    int32_t ret = epoll_wait(epfd, events?_events:NULL, maxevents, timeout);
    if(ret>0)
        UnalignEpollEvent(events, _events, ret);
    return ret;
}
EXPORT int32_t my_epoll_pwait(x64emu_t* emu, int32_t epfd, void* events, int32_t maxevents, int32_t timeout, const sigset_t *sigmask)
{
    struct epoll_event _events[maxevents];
    //AlignEpollEvent(_events, events, maxevents);
    int32_t ret = epoll_pwait(epfd, events?_events:NULL, maxevents, timeout, sigmask);
    if(ret>0)
        UnalignEpollEvent(events, _events, ret);
    return ret;
}
#endif

EXPORT int32_t my_glob64(x64emu_t *emu, void* pat, int32_t flags, void* errfnc, void* pglob)
{
    (void)emu;
    return glob64(pat, flags, findgloberrFct(errfnc), pglob);
}
EXPORT int32_t my_glob(x64emu_t *emu, void* pat, int32_t flags, void* errfnc, void* pglob) __attribute__((alias("my_glob64")));

EXPORT int my_scandir64(x64emu_t *emu, void* dir, void* namelist, void* sel, void* comp)
{
    (void)emu;
    return scandir64(dir, namelist, findfilter64Fct(sel), findcompare64Fct(comp));
}
EXPORT int my_scandir(x64emu_t *emu, void* dir, void* namelist, void* sel, void* comp) __attribute__((alias("my_scandir64")));

EXPORT int my_scandirat(x64emu_t *emu, int dirfd, void* dirp, void* namelist, void* sel, void* comp)
{
    (void)emu;
    return scandirat(dirfd, dirp, namelist, findfilter64Fct(sel), findcompare64Fct(comp));
}

EXPORT int my_ftw64(x64emu_t* emu, void* filename, void* func, int descriptors)
{
    (void)emu;
    return ftw64(filename, findftw64Fct(func), descriptors);
}
EXPORT int my_ftw(x64emu_t* emu, void* filename, void* func, int descriptors) __attribute__((alias("my_ftw64")));

EXPORT int32_t my_nftw64(x64emu_t* emu, void* pathname, void* B, int32_t nopenfd, int32_t flags)
{
    (void)emu;
    return nftw64(pathname, findnftw64Fct(B), nopenfd, flags);
}

EXPORT int32_t my_execv(x64emu_t* emu, const char* path, char* const argv[])
{
    int self = isProcSelf(path, "exe");
    int x64 = FileIsX64ELF(path);
    int x86 = my_context->box86path?FileIsX86ELF(path):0;
    printf_log(LOG_DEBUG, "execv(\"%s\", %p) is x64=%d x86=%d\n", path, argv, x64, x86);
    #if 1
    if (x64 || x86 || self) {
        int skip_first = 0;
        if(strlen(path)>=strlen("wine64-preloader") && strcmp(path+strlen(path)-strlen("wine64-preloader"), "wine64-preloader")==0)
            skip_first++;
        // count argv...
        int n=skip_first;
        while(argv[n]) ++n;
        const char** newargv = (const char**)calloc(n+2, sizeof(char*));
        newargv[0] = x86?emu->context->box86path:emu->context->box64path;
        memcpy(newargv+1, argv+skip_first, sizeof(char*)*(n+1));
        if(self) newargv[1] = emu->context->fullpath;
        printf_log(LOG_DEBUG, " => execv(\"%s\", %p [\"%s\", \"%s\", \"%s\"...:%d])\n", newargv[0], newargv, newargv[0], n?newargv[1]:"", (n>1)?newargv[2]:"",n);
        int ret = execv(newargv[0], (char* const*)newargv);
        free(newargv);
        return ret;
    }
    #endif
    return execv(path, argv);
}

EXPORT int32_t my_execve(x64emu_t* emu, const char* path, char* const argv[], char* const envp[])
{
    int self = isProcSelf(path, "exe");
    int x64 = FileIsX64ELF(path);
    int x86 = my_context->box86path?FileIsX86ELF(path):0;
    // hack to update the environ var if needed
    if(envp == my_context->envv && environ) {
        envp = environ;
    }
    printf_log(LOG_DEBUG, "execve(\"%s\", %p, %p) is x64=%d x86=%d\n", path, argv, envp, x64, x86);
    #if 1
    if (x64 || x86 || self) {
        int skip_first = 0;
        if(strlen(path)>=strlen("wine64-preloader") && strcmp(path+strlen(path)-strlen("wine64-preloader"), "wine64-preloader")==0)
            skip_first++;
        // count argv...
        int n=skip_first;
        while(argv[n]) ++n;
        const char** newargv = (const char**)calloc(n+2, sizeof(char*));
        newargv[0] = x86?emu->context->box86path:emu->context->box64path;
        memcpy(newargv+1, argv+skip_first, sizeof(char*)*(n+1));
        if(self) newargv[1] = emu->context->fullpath;
        printf_log(LOG_DEBUG, " => execve(\"%s\", %p [\"%s\", \"%s\", \"%s\"...:%d], %p)\n", newargv[0], newargv, newargv[0], n?newargv[1]:"", (n>1)?newargv[2]:"",n, envp);
        int ret = execve(newargv[0], (char* const*)newargv, envp);
        free(newargv);
        return ret;
    }
    #endif
    if(!strcmp(path + strlen(path) - strlen("/uname"), "/uname")
     && argv[1] && (!strcmp(argv[1], "-m") || !strcmp(argv[1], "-p") || !strcmp(argv[1], "-i"))
     && !argv[2]) {
        // uname -m is redirected to box64 -m
        path = my_context->box64path;
        char *argv2[3] = { my_context->box64path, argv[1], NULL };
        return execve(path, argv2, envp);
    }

    return execve(path, argv, envp);
}

// execvp should use PATH to search for the program first
EXPORT int32_t my_execvp(x64emu_t* emu, const char* path, char* const argv[])
{
    // need to use BOX64_PATH / PATH here...
    char* fullpath = ResolveFile(path, &my_context->box64_path);
    // use fullpath...
    int self = isProcSelf(fullpath, "exe");
    int x64 = FileIsX64ELF(fullpath);
    int x86 = my_context->box86path?FileIsX86ELF(path):0;
    printf_log(LOG_DEBUG, "execvp(\"%s\", %p), IsX86=%d / fullpath=\"%s\"\n", path, argv, x64, fullpath);
    free(fullpath);
    if (x64 || x86 || self) {
        // count argv...
        int i=0;
        while(argv[i]) ++i;
        char** newargv = (char**)calloc(i+2, sizeof(char*));
        newargv[0] = x86?emu->context->box86path:emu->context->box64path;
        for (int j=0; j<i; ++j)
            newargv[j+1] = argv[j];
        if(self) newargv[1] = emu->context->fullpath;
        printf_log(LOG_DEBUG, " => execvp(\"%s\", %p [\"%s\", \"%s\"...:%d])\n", newargv[0], newargv, newargv[1], i?newargv[2]:"", i);
        int ret = execvp(newargv[0], newargv);
        free(newargv);
        return ret;
    }
    if((!strcmp(path + strlen(path) - strlen("/uname"), "/uname") || !strcmp(path, "uname"))
     && argv[1] && (!strcmp(argv[1], "-m") || !strcmp(argv[1], "-p") || !strcmp(argv[1], "-i"))
     && !argv[2]) {
        // uname -m is redirected to box64 -m
        path = my_context->box64path;
        char *argv2[3] = { my_context->box64path, argv[1], NULL };
        return execvp(path, argv2);
    }

    // fullpath is gone, so the search will only be on PATH, not on BOX64_PATH (is that an issue?)
    return execvp(path, argv);
}

EXPORT int32_t my_execl(x64emu_t* emu, const char* path)
{
    int self = isProcSelf(path, "exe");
    int x64 = FileIsX64ELF(path);
    int x86 = my_context->box86path?FileIsX86ELF(path):0;
    printf_log(LOG_DEBUG, "execl(\"%s\", ...), IsX86=%d, self=%d\n", path, x64, self);
    // count argv...
    int i=0;
    while(getVargN(emu, i+1)) ++i;
    char** newargv = (char**)calloc(i+((x64 || self)?2:1), sizeof(char*));
    int j=0;
    if ((x64 || x86 || self))
        newargv[j++] = x86?emu->context->box86path:emu->context->box64path;
    for (int k=0; k<i; ++k)
        newargv[j++] = getVargN(emu, k+1);
    if(self) newargv[1] = emu->context->fullpath;
    printf_log(LOG_DEBUG, " => execl(\"%s\", %p [\"%s\", \"%s\"...:%d])\n", newargv[0], newargv, newargv[1], i?newargv[2]:"", i);
    int ret = execv(newargv[0], newargv);
    free(newargv);
    return ret;
}

EXPORT int32_t my_execlp(x64emu_t* emu, const char* path)
{
    // need to use BOX64_PATH / PATH here...
    char* fullpath = ResolveFile(path, &my_context->box64_path);
    // use fullpath...
    int self = isProcSelf(fullpath, "exe");
    int x64 = FileIsX64ELF(fullpath);
    int x86 = my_context->box86path?FileIsX86ELF(path):0;
    printf_log(LOG_DEBUG, "execlp(\"%s\", ...), IsX86=%d / fullpath=\"%s\"\n", path, x64, fullpath);
    free(fullpath);
    // count argv...
    int i=0;
    while(getVargN(emu, i+1)) ++i;
    char** newargv = (char**)calloc(i+((x64 || self)?2:1), sizeof(char*));
    int j=0;
    if ((x64 || x86 || self))
        newargv[j++] = x86?emu->context->box86path:emu->context->box64path;
    for (int k=0; k<i; ++k)
        newargv[j++] = getVargN(emu, k+1);
    if(self) newargv[1] = emu->context->fullpath;
    printf_log(LOG_DEBUG, " => execlp(\"%s\", %p [\"%s\", \"%s\"...:%d])\n", newargv[0], newargv, newargv[1], i?newargv[2]:"", i);
    int ret = execvp(newargv[0], newargv);
    free(newargv);
    return ret;
}

EXPORT int32_t my_posix_spawn(x64emu_t* emu, pid_t* pid, const char* fullpath, 
    const posix_spawn_file_actions_t *actions, const posix_spawnattr_t* attrp,  char* const argv[], char* const envp[])
{
    int self = isProcSelf(fullpath, "exe");
    int x64 = FileIsX64ELF(fullpath);
    int x86 = my_context->box86path?FileIsX86ELF(fullpath):0;
    printf_log(/*LOG_DEBUG*/LOG_INFO, "posix_spawn(%p, \"%s\", %p, %p, %p, %p), IsX86=%d\n", pid, fullpath, actions, attrp, argv, envp, x64);
    if (x64 || x86 || self) {
        // count argv...
        int i=0;
        while(argv[i]) ++i;
        char** newargv = (char**)calloc(i+2, sizeof(char*));
        newargv[0] = x86?emu->context->box86path:emu->context->box64path;
        for (int j=0; j<i; ++j)
            newargv[j+1] = argv[j];
        if(self) newargv[1] = emu->context->fullpath;
        printf_log(/*LOG_DEBUG*/LOG_INFO, " => posix_spawn(%p, \"%s\", %p, %p, %p [\"%s\", \"%s\"...:%d], %p)\n", pid, newargv[0], actions, attrp, newargv, newargv[1], i?newargv[2]:"", i, envp);
        int ret = posix_spawn(pid, newargv[0], actions, attrp, newargv, envp);
        printf_log(/*LOG_DEBUG*/LOG_INFO, "posix_spawn returned %d\n", ret);
        //free(newargv);
        return ret;
    }
    // fullpath is gone, so the search will only be on PATH, not on BOX64_PATH (is that an issue?)
    return posix_spawn(pid, fullpath, actions, attrp, argv, envp);
}

// execvp should use PATH to search for the program first
EXPORT int32_t my_posix_spawnp(x64emu_t* emu, pid_t* pid, const char* path, 
    const posix_spawn_file_actions_t *actions, const posix_spawnattr_t* attrp,  char* const argv[], char* const envp[])
{
    // need to use BOX64_PATH / PATH here...
    char* fullpath = ResolveFile(path, &my_context->box64_path);
    // use fullpath...
    int self = isProcSelf(fullpath, "exe");
    int x64 = FileIsX64ELF(fullpath);
    int x86 = my_context->box86path?FileIsX86ELF(path):0;
    printf_log(/*LOG_DEBUG*/LOG_INFO, "posix_spawnp(%p, \"%s\", %p, %p, %p, %p), IsX86=%d / fullpath=\"%s\"\n", pid, path, actions, attrp, argv, envp, x64, fullpath);
    free(fullpath);
    if (x64 || x86 || self) {
        // count argv...
        int i=0;
        while(argv[i]) ++i;
        char** newargv = (char**)calloc(i+2, sizeof(char*));
        newargv[0] = x86?emu->context->box86path:emu->context->box64path;
        for (int j=0; j<i; ++j)
            newargv[j+1] = argv[j];
        if(self) newargv[1] = emu->context->fullpath;
        printf_log(/*LOG_DEBUG*/LOG_INFO, " => posix_spawnp(%p, \"%s\", %p, %p, %p [\"%s\", \"%s\"...:%d], %p)\n", pid, newargv[0], actions, attrp, newargv, newargv[1], i?newargv[2]:"", i, envp);
        int ret = posix_spawnp(pid, newargv[0], actions, attrp, newargv, envp);
        printf_log(/*LOG_DEBUG*/LOG_INFO, "posix_spawnp returned %d\n", ret);
        //free(newargv);
        return ret;
    }
    // fullpath is gone, so the search will only be on PATH, not on BOX64_PATH (is that an issue?)
    return posix_spawnp(pid, path, actions, attrp, argv, envp);
}

EXPORT void my__Jv_RegisterClasses() {}

EXPORT int32_t my___cxa_thread_atexit_impl(x64emu_t* emu, void* dtor, void* obj, void* dso)
{
    (void)emu;
    printf_log(LOG_INFO, "Warning, call to __cxa_thread_atexit_impl(%p, %p, %p) ignored\n", dtor, obj, dso);
    return 0;
}

EXPORT int32_t my___register_atfork(x64emu_t *emu, void* prepare, void* parent, void* child, void* handle)
{
    (void)emu;
    // this is partly incorrect, because the emulated funcionts should be executed by actual fork and not by my_atfork...
    if(my_context->atfork_sz==my_context->atfork_cap) {
        my_context->atfork_cap += 4;
        my_context->atforks = (atfork_fnc_t*)realloc(my_context->atforks, my_context->atfork_cap*sizeof(atfork_fnc_t));
    }
    my_context->atforks[my_context->atfork_sz].prepare = (uintptr_t)prepare;
    my_context->atforks[my_context->atfork_sz].parent = (uintptr_t)parent;
    my_context->atforks[my_context->atfork_sz].child = (uintptr_t)child;
    my_context->atforks[my_context->atfork_sz].handle = handle;
    return 0;
}

#if 0
EXPORT uint64_t my___umoddi3(uint64_t a, uint64_t b)
{
    return a%b;
}
EXPORT uint64_t my___udivdi3(uint64_t a, uint64_t b)
{
    return a/b;
}
EXPORT int64_t my___divdi3(int64_t a, int64_t b)
{
    return a/b;
}

EXPORT int32_t my___poll_chk(void* a, uint32_t b, int c, int l)
{
    return poll(a, b, c);   // no check...
}
#endif

EXPORT int32_t my_fcntl64(x64emu_t* emu, int32_t a, int32_t b, void* c)
{
    (void)emu;
    if(b==F_SETFL)
        c = (void*)(uintptr_t)of_convert((intptr_t)c);
    #if 0
    if(b==F_GETLK64 || b==F_SETLK64 || b==F_SETLKW64)
    {
        my_flock64_t fl;
        AlignFlock64(&fl, c);
        int ret = fcntl(a, b, &fl);
        UnalignFlock64(c, &fl);
        return ret;
    }
    #endif
    int ret = fcntl(a, b, c);
    if(b==F_GETFL && ret!=-1)
        ret = of_unconvert(ret);

    return ret;
}

EXPORT int32_t my_fcntl(x64emu_t* emu, int32_t a, int32_t b, void* c)
{
    (void)emu;
    if(b==F_SETFL && (intptr_t)c==0xFFFFF7FF) {
        // special case for ~O_NONBLOCK...
        int flags = fcntl(a, F_GETFL);
        if(flags&O_NONBLOCK) {
            flags &= ~O_NONBLOCK;
            return fcntl(a, b, flags);
        }
        return 0;
    }
    if(b==F_SETFL)
        c = (void*)(uintptr_t)of_convert((intptr_t)c);
    #if 0
    if(b==F_GETLK64 || b==F_SETLK64 || b==F_SETLKW64)
    {
        my_flock64_t fl;
        AlignFlock64(&fl, c);
        int ret = fcntl(a, b, &fl);
        UnalignFlock64(c, &fl);
        return ret;
    }
    #endif
    int ret = fcntl(a, b, c);
    if(b==F_GETFL && ret!=-1)
        ret = of_unconvert(ret);
    
    return ret;    
}
EXPORT int32_t my___fcntl(x64emu_t* emu, int32_t a, int32_t b, void* c) __attribute__((alias("my_fcntl")));

#if 0
EXPORT int32_t my_preadv64(x64emu_t* emu, int32_t fd, void* v, int32_t c, int64_t o)
{
    library_t* lib = my_lib;
    if(!lib) return 0;
    void* f = dlsym(lib->priv.w.lib, "preadv64");
    if(f)
        return ((iFipiI_t)f)(fd, v, c, o);
    return syscall(__NR_preadv, fd, v, c,(uint32_t)(o&0xffffffff), (uint32_t)((o>>32)&0xffffffff));
}

EXPORT int32_t my_pwritev64(x64emu_t* emu, int32_t fd, void* v, int32_t c, int64_t o)
{
    library_t* lib = my_lib;
    if(!lib) return 0;
    void* f = dlsym(lib->priv.w.lib, "pwritev64");
    if(f)
        return ((iFipiI_t)f)(fd, v, c, o);
    #ifdef __arm__
    return syscall(__NR_pwritev, fd, v, c, 0, (uint32_t)(o&0xffffffff), (uint32_t)((o>>32)&0xffffffff));
    // on arm, 64bits args needs to be on even/odd register, so need to put a 0 for aligment
    #else
    return syscall(__NR_pwritev, fd, v, c,(uint32_t)(o&0xffffffff), (uint32_t)((o>>32)&0xffffffff));
    #endif
}

EXPORT int32_t my_accept4(x64emu_t* emu, int32_t fd, void* a, void* l, int32_t flags)
{
    library_t* lib = my_lib;
    if(!lib) return 0;
    void* f = dlsym(lib->priv.w.lib, "accept4");
    if(f)
        return ((iFippi_t)f)(fd, a, l, flags);
    if(!flags)
        return accept(fd, a, l);
    return syscall(__NR_accept4, fd, a, l, flags);
}

EXPORT  int32_t my_fallocate64(int fd, int mode, int64_t offs, int64_t len)
{
    iFiiII_t f = NULL;
    static int done = 0;
    if(!done) {
        library_t* lib = my_lib;
        f = (iFiiII_t)dlsym(lib->priv.w.lib, "fallocate64");
        done = 1;
    }
    if(f)
        return f(fd, mode, offs, len);
    else
        return syscall(__NR_fallocate, fd, mode, (uint32_t)(offs&0xffffffff), (uint32_t)((offs>>32)&0xffffffff), (uint32_t)(len&0xffffffff), (uint32_t)((len>>32)&0xffffffff));
        //return posix_fallocate64(fd, offs, len);
}

EXPORT struct __processor_model
{
  unsigned int __cpu_vendor;
  unsigned int __cpu_type;
  unsigned int __cpu_subtype;
  unsigned int __cpu_features[1];
} my___cpu_model;

#include "cpu_info.h"
void InitCpuModel()
{
    // some pseudo random cpu info...
    my___cpu_model.__cpu_vendor = VENDOR_INTEL;
    my___cpu_model.__cpu_type = INTEL_PENTIUM_M;
    my___cpu_model.__cpu_subtype = 0; // N/A
    my___cpu_model.__cpu_features[0] = (1<<FEATURE_CMOV) 
                                     | (1<<FEATURE_MMX) 
                                     | (1<<FEATURE_SSE) 
                                     | (1<<FEATURE_SSE2) 
                                     | (1<<FEATURE_SSE3)
                                     | (1<<FEATURE_SSSE3)
                                     | (1<<FEATURE_MOVBE)
                                     | (1<<FEATURE_ADX);
}
#endif

EXPORT const unsigned short int *my___ctype_b;
EXPORT const int32_t *my___ctype_tolower;
EXPORT const int32_t *my___ctype_toupper;

void ctSetup()
{
    my___ctype_b = *(__ctype_b_loc());
    my___ctype_toupper = *(__ctype_toupper_loc());
    my___ctype_tolower = *(__ctype_tolower_loc());
}

EXPORT void* my___libc_stack_end;
void stSetup(box64context_t* context)
{
    my___libc_stack_end = context->stack;   // is this the end, or should I add stasz?
}

EXPORT void my___register_frame_info(void* a, void* b)
{
    // nothing
    (void)a; (void)b;
}
EXPORT void* my___deregister_frame_info(void* a)
{
    (void)a;
    return NULL;
}

EXPORT void* my____brk_addr = NULL;

// longjmp / setjmp
typedef struct jump_buff_x64_s {
    uint64_t save_rbx;
    uint64_t save_rbp;
    uint64_t save_r12;
    uint64_t save_r13;
    uint64_t save_r14;
    uint64_t save_r15;
    uint64_t save_rsp;
    uint64_t save_rip;
} jump_buff_x64_t;

typedef struct __jmp_buf_tag_s {
    jump_buff_x64_t __jmpbuf;
    int              __mask_was_saved;
    __sigset_t       __saved_mask;
} __jmp_buf_tag_t;

void EXPORT my_longjmp(x64emu_t* emu, /*struct __jmp_buf_tag __env[1]*/void *p, int32_t __val)
{
    jump_buff_x64_t *jpbuff = &((__jmp_buf_tag_t*)p)->__jmpbuf;
    //restore  regs
    R_RBX = jpbuff->save_rbx;
    R_RBP = jpbuff->save_rbp;
    R_R12 = jpbuff->save_r12;
    R_R13 = jpbuff->save_r13;
    R_R14 = jpbuff->save_r14;
    R_R15 = jpbuff->save_r15;
    R_RSP = jpbuff->save_rsp;
    // jmp to saved location, plus restore val to rax
    R_RAX = __val;
    R_RIP = jpbuff->save_rip;
    if(((__jmp_buf_tag_t*)p)->__mask_was_saved) {
        sigprocmask(SIG_SETMASK, &((__jmp_buf_tag_t*)p)->__saved_mask, NULL);
    }
    if(emu->quitonlongjmp) {
        emu->longjmp = 1;
        emu->quit = 1;
    }
}

EXPORT int32_t my___sigsetjmp(x64emu_t* emu, /*struct __jmp_buf_tag __env[1]*/void *p, int savesigs)
{
    jump_buff_x64_t *jpbuff = &((__jmp_buf_tag_t*)p)->__jmpbuf;
    // save the buffer
    jpbuff->save_rbx = R_RBX;
    jpbuff->save_rbp = R_RBP;
    jpbuff->save_r12 = R_R12;
    jpbuff->save_r13 = R_R13;
    jpbuff->save_r14 = R_R14;
    jpbuff->save_r15 = R_R15;
    jpbuff->save_rsp = R_RSP+sizeof(uintptr_t); // include "return address"
    jpbuff->save_rip = *(uintptr_t*)(R_RSP);
    if(savesigs) {
        if(sigprocmask(SIG_SETMASK, NULL, &((__jmp_buf_tag_t*)p)->__saved_mask))
            ((__jmp_buf_tag_t*)p)->__mask_was_saved = 0;
        else
            ((__jmp_buf_tag_t*)p)->__mask_was_saved = 1;
    } else
        ((__jmp_buf_tag_t*)p)->__mask_was_saved = 0;
    return 0;
}

EXPORT int32_t my__setjmp(x64emu_t* emu, /*struct __jmp_buf_tag __env[1]*/void *p)
{
    return  my___sigsetjmp(emu, p, 0);
}
EXPORT int32_t my_setjmp(x64emu_t* emu, /*struct __jmp_buf_tag __env[1]*/void *p)
{
    return  my___sigsetjmp(emu, p, 1);
}

EXPORT void my___explicit_bzero_chk(x64emu_t* emu, void* dst, uint32_t len, uint32_t dstlen)
{
    (void)emu; (void)dstlen;
    memset(dst, 0, len);
}

EXPORT void* my_realpath(x64emu_t* emu, void* path, void* resolved_path)
{
    if(isProcSelf(path, "exe")) {
        return realpath(emu->context->fullpath, resolved_path);
    }
    return realpath(path, resolved_path);
}

EXPORT int my_readlinkat(x64emu_t* emu, int fd, void* path, void* buf, size_t bufsize)
{
    if(isProcSelf(path, "exe")) {
        strncpy(buf, emu->context->fullpath, bufsize);
        size_t l = strlen(emu->context->fullpath);
        return (l>bufsize)?bufsize:(l+1);
    }
    return readlinkat(fd, path, buf, bufsize);
}


EXPORT void* my_mmap64(x64emu_t* emu, void *addr, unsigned long length, int prot, int flags, int fd, int64_t offset)
{
    (void)emu;
    if(prot&PROT_WRITE) 
        prot|=PROT_READ;    // PROT_READ is implicit with PROT_WRITE on i386
    if(box64_log<LOG_DEBUG) {dynarec_log(LOG_DEBUG, "mmap64(%p, %lu, 0x%x, 0x%x, %d, %ld) => ", addr, length, prot, flags, fd, offset);}
    #ifndef NOALIGN
    void* old_addr = addr;
    if(flags&0x40) {
        // 0x40 is MAP_32BIT, wich only exist on x86_64!
        //flags &= ~0x40;   // let the flags in?
        addr = findBlockNearHint(addr, length);
    } else if (box64_wine) {
        if(!addr)
            addr = find47bitBlock(length);
    }
    #endif
    void* ret = mmap64(addr, length, prot, flags, fd, offset);
    #ifndef NOALIGN
    if((ret!=(void*)-1) && (flags&0x40) && 
      (((uintptr_t)ret>0xffffffffLL) || (box64_wine && ((uintptr_t)ret&0xffff)))) {
        printf_log(LOG_DEBUG, "Warning, mmap on 32bits didn't worked, ask %p, got %p ", addr, ret);
        munmap(ret, length);
        loadProtectionFromMap();    // reload map, because something went wrong previously
        addr = findBlockNearHint(old_addr, length); // is this the best way?
        ret = mmap64(addr, length, prot, flags, fd, offset);
        printf_log(LOG_DEBUG, " tried again with %p, got %p\n", addr, ret);
    } else if((ret!=(void*)-1) && !(flags&MAP_FIXED) && (box64_wine) && (old_addr) && (addr!=ret) &&
             (((uintptr_t)ret>0x7fffffffffffLL) || ((uintptr_t)ret&~0xffff))) {
        printf_log(LOG_DEBUG, "Warning, mmap on 47bits didn't worked, ask %p, got %p ", addr, ret);
        munmap(ret, length);
        loadProtectionFromMap();    // reload map, because something went wrong previously
        addr = find47bitBlockNearHint(old_addr, length); // is this the best way?
        ret = mmap64(addr, length, prot, flags, fd, offset);
        printf_log(LOG_DEBUG, " tried again with %p, got %p\n", addr, ret);
    }
    #endif
    if(box64_log<LOG_DEBUG) {dynarec_log(LOG_DEBUG, "%p\n", ret);}
    #ifdef DYNAREC
    if(box64_dynarec && ret!=(void*)-1) {
        if(flags&0x100000 && addr!=ret)
        {
            // program used MAP_FIXED_NOREPLACE but the host linux didn't support it
            // and responded with a different address, so ignore it
        } else {
            if(prot& PROT_EXEC)
                addDBFromAddressRange((uintptr_t)ret, length);
            else
                cleanDBFromAddressRange((uintptr_t)ret, length, prot?0:1);
        }
    }
    #endif
    if(ret!=(void*)-1)
        setProtection((uintptr_t)ret, length, prot);
    return ret;
}
EXPORT void* my_mmap(x64emu_t* emu, void *addr, unsigned long length, int prot, int flags, int fd, int64_t offset) __attribute__((alias("my_mmap64")));

EXPORT void* my_mremap(x64emu_t* emu, void* old_addr, size_t old_size, size_t new_size, int flags, void* new_addr)
{
    (void)emu;
    dynarec_log(LOG_DEBUG, "mremap(%p, %lu, %lu, %d, %p)=>", old_addr, old_size, new_size, flags, new_addr);
    void* ret = mremap(old_addr, old_size, new_size, flags, new_addr);
    dynarec_log(LOG_DEBUG, "%p\n", ret);
    if(ret==(void*)-1)
        return ret; // failed...
    uint32_t prot = getProtection((uintptr_t)old_addr)&~PROT_CUSTOM;
    if(ret==old_addr) {
        if(old_size && old_size<new_size) {
            setProtection((uintptr_t)ret+old_size, new_size-old_size, prot);
            #ifdef DYNAREC
            if(box64_dynarec)
                addDBFromAddressRange((uintptr_t)ret+old_size, new_size-old_size);
            #endif
        } else if(old_size && new_size<old_size) {
            freeProtection((uintptr_t)ret+new_size, old_size-new_size);
            #ifdef DYNAREC
            if(box64_dynarec)
                cleanDBFromAddressRange((uintptr_t)ret+new_size, new_size-old_size, 1);
            #endif
        } else if(!old_size) {
            setProtection((uintptr_t)ret, new_size, prot);
            #ifdef DYNAREC
            if(box64_dynarec)
                addDBFromAddressRange((uintptr_t)ret, new_size);
            #endif
        }
    } else {
        if(old_size
        #ifdef MREMAP_DONTUNMAP
        && ((flags&MREMAP_DONTUNMAP)==0)
        #endif
        ) {
            freeProtection((uintptr_t)old_addr, old_size);
            #ifdef DYNAREC
            if(box64_dynarec)
                cleanDBFromAddressRange((uintptr_t)old_addr, old_size, 1);
            #endif
        }
        setProtection((uintptr_t)ret, new_size, prot); // should copy the protection from old block
        #ifdef DYNAREC
        if(box64_dynarec)
            addDBFromAddressRange((uintptr_t)ret, new_size);
        #endif
    }
    return ret;
}

EXPORT int my_munmap(x64emu_t* emu, void* addr, unsigned long length)
{
    (void)emu;
    dynarec_log(LOG_DEBUG, "munmap(%p, %lu)\n", addr, length);
    #ifdef DYNAREC
    if(box64_dynarec && length) {
        cleanDBFromAddressRange((uintptr_t)addr, length, 1);
    }
    #endif
    int ret = munmap(addr, length);
    if(!ret)
        freeProtection((uintptr_t)addr, length);
    return ret;
}

EXPORT int my_mprotect(x64emu_t* emu, void *addr, unsigned long len, int prot)
{
    (void)emu;
    dynarec_log(LOG_DEBUG, "mprotect(%p, %lu, 0x%x)\n", addr, len, prot);
    if(prot&PROT_WRITE) 
        prot|=PROT_READ;    // PROT_READ is implicit with PROT_WRITE on x86_64
    int ret = mprotect(addr, len, prot);
    #ifdef DYNAREC
    if(box64_dynarec) {
        if(prot& PROT_EXEC)
            addDBFromAddressRange((uintptr_t)addr, len);
        else
            cleanDBFromAddressRange((uintptr_t)addr, len, 0);
    }
    #endif
    if(!ret)
        updateProtection((uintptr_t)addr, len, prot);
    return ret;
}

EXPORT void* my_mallinfo(x64emu_t* emu, void* p)
{
    *(struct mallinfo*)p = mallinfo();
    return p;
}

EXPORT int my_getopt(int argc, char* const argv[], const char *optstring)
{
    my_updateGlobalOpt();
    int ret = getopt(argc, argv, optstring);
    my_checkGlobalOpt();
    return ret;
}

EXPORT int my_getopt_long(int argc, char* const argv[], const char* optstring, const struct option *longopts, int *longindex)
{
    my_updateGlobalOpt();
    int ret = getopt_long(argc, argv, optstring, longopts, longindex);
    my_checkGlobalOpt();
    return ret;
}

EXPORT int my_getopt_long_only(int argc, char* const argv[], const char* optstring, const struct option *longopts, int *longindex)
{
    my_updateGlobalOpt();
    int ret = getopt_long_only(argc, argv, optstring, longopts, longindex);
    my_checkGlobalOpt();
    return ret;
}

#if 0
typedef struct my_cookie_s {
    uintptr_t r, w, s, c;
    void* cookie;
} my_cookie_t;

static ssize_t my_cookie_read(void *p, char *buf, size_t size)
{
    my_cookie_t* cookie = (my_cookie_t*)p;
    return (ssize_t)RunFunction(my_context, cookie->r, 3, cookie->cookie, buf, size);
}
static ssize_t my_cookie_write(void *p, const char *buf, size_t size)
{
    my_cookie_t* cookie = (my_cookie_t*)p;
    return (ssize_t)RunFunction(my_context, cookie->w, 3, cookie->cookie, buf, size);
}
static int my_cookie_seek(void *p, off64_t *offset, int whence)
{
    my_cookie_t* cookie = (my_cookie_t*)p;
    return RunFunction(my_context, cookie->s, 3, cookie->cookie, offset, whence);
}
static int my_cookie_close(void *p)
{
    my_cookie_t* cookie = (my_cookie_t*)p;
    int ret = 0;
    if(cookie->c)
        ret = RunFunction(my_context, cookie->c, 1, cookie->cookie);
    free(cookie);
    return ret;
}
EXPORT void* my_fopencookie(x64emu_t* emu, void* cookie, void* mode, void* read, void* write, void* seek, void* close)
{
    cookie_io_functions_t io_funcs = {read?my_cookie_read:NULL, write?my_cookie_write:NULL, seek?my_cookie_seek:NULL, my_cookie_close};
    my_cookie_t *cb = (my_cookie_t*)calloc(1, sizeof(my_cookie_t));
    cb->r = (uintptr_t)read;
    cb->w = (uintptr_t)write;
    cb->s = (uintptr_t)seek;
    cb->c = (uintptr_t)close;
    cb->cookie = cookie;
    return fopencookie(cb, mode, io_funcs);
}

EXPORT long my_prlimit64(void* pid, uint32_t res, void* new_rlim, void* old_rlim)
{
    return syscall(__NR_prlimit64, pid, res, new_rlim, old_rlim);
}

EXPORT void* my_reallocarray(void* ptr, size_t nmemb, size_t size)
{
    return realloc(ptr, nmemb*size);
}

#ifndef __OPEN_NEEDS_MODE
# define __OPEN_NEEDS_MODE(oflag) \
  (((oflag) & O_CREAT) != 0)
// || ((oflag) & __O_TMPFILE) == __O_TMPFILE)
#endif
EXPORT int my___open_nocancel(x64emu_t* emu, void* file, int oflag, int* b)
{
    int mode = 0;
    if (__OPEN_NEEDS_MODE (oflag))
        mode = b[0];
    return openat(AT_FDCWD, file, oflag, mode);
}

EXPORT int my___libc_alloca_cutoff(x64emu_t* emu, size_t size)
{
    // not always implemented on old linux version...
    library_t* lib = my_lib;
    if(!lib) return 0;
    void* f = dlsym(lib->priv.w.lib, "__libc_alloca_cutoff");
    if(f)
        return ((iFL_t)f)(size);
    // approximate version but it's better than nothing....
    return (size<=(65536*4));
}

// DL functions from wrappedlibdl.c
void* my_dlopen(x64emu_t* emu, void *filename, int flag);
int my_dlclose(x64emu_t* emu, void *handle);
void* my_dlsym(x64emu_t* emu, void *handle, void *symbol);
EXPORT int my___libc_dlclose(x64emu_t* emu, void* handle)
{
    return my_dlclose(emu, handle);
}
EXPORT void* my___libc_dlopen_mode(x64emu_t* emu, void* name, int mode)
{
    return my_dlopen(emu, name, mode);
}
EXPORT void* my___libc_dlsym(x64emu_t* emu, void* handle, void* name)
{
    return my_dlsym(emu, handle, name);
}

// all obstack function defined in obstack.c file
void obstackSetup();

EXPORT int my_nanosleep(const struct timespec *req, struct timespec *rem)
{
    if(!req)
        return 0;   // workaround for some strange calls
    return nanosleep(req, rem);
}
#endif

EXPORT void* my_malloc(unsigned long size)
{
    return calloc(1, size);
}

#if 0
#ifdef PANDORA
#define RENAME_NOREPLACE	(1 << 0)
#define RENAME_EXCHANGE		(1 << 1)
#define RENAME_WHITEOUT		(1 << 2)
EXPORT int my_renameat2(int olddirfd, void* oldpath, int newdirfd, void* newpath, uint32_t flags)
{
    // simulate that function, but
    if(flags&RENAME_NOREPLACE) {
        if(FileExist(newpath, -1)) {
            errno = EEXIST;
            return -1;
        }
        flags &= ~RENAME_NOREPLACE;
    }
    if(!flags) return renameat(olddirfd, oldpath, newdirfd, newpath);
    if(flags&RENAME_WHITEOUT) {
        errno = EINVAL;
        return -1;  // not handling that
    }
    if((flags&RENAME_EXCHANGE) && (olddirfd==-1) && (newdirfd==-1)) {
        // cannot do atomically...
        char* tmp = (char*)malloc(strlen(oldpath)+10); // create a temp intermediary
        tmp = strcat(oldpath, ".tmp");
        int ret = renameat(-1, oldpath, -1, tmp);
        if(ret==-1) return -1;
        ret = renameat(-1, newpath, -1, oldpath);
        if(ret==-1) return -1;
        ret = renameat(-1, tmp, -1, newpath);
        free(tmp);
        return ret;
    }
    return -1; // unknown flags
}
#endif

#ifndef __NR_memfd_create
#define MFD_CLOEXEC		    0x0001U
#define MFD_ALLOW_SEALING	0x0002U
EXPORT int my_memfd_create(x64emu_t* emu, void* name, uint32_t flags)
{
    // try to simulate that function
    uint32_t fl = O_RDWR | O_CREAT;
    if(flags&MFD_CLOEXEC)
        fl |= O_CLOEXEC;
    int tmp = shm_open(name, fl, S_IRWXU);
    if(tmp<0) return -1;
    shm_unlink(name);    // remove the shm file, but it will still exist because it's currently in use
    return tmp;
}
#endif

#ifndef GRND_RANDOM
#define GRND_RANDOM	0x0002
#endif
EXPORT int my_getentropy(x64emu_t* emu, void* buffer, size_t length)
{
    library_t* lib = my_lib;
    if(!lib) return 0;
    void* f = dlsym(lib->priv.w.lib, "getentropy");
    if(f)
        return ((iFpL_t)f)(buffer, length);
    // custom implementation
    if(length>256) {
        errno = EIO;
        return -1;
    }
    int ret = my_getrandom(emu, buffer, length, GRND_RANDOM);
    if(ret!=length) {
        errno = EIO;
        return -1;
    }
    return 0;
}

EXPORT void my_mcount(void* frompc, void* selfpc)
{
    // stub doing nothing...
    return;
}
#endif

union semun {
  int              val;    /* Value for SETVAL */
  struct semid_ds *buf;    /* Buffer for IPC_STAT, IPC_SET */
  unsigned short  *array;  /* Array for GETALL, SETALL */
  struct seminfo  *__buf;  /* Buffer for IPC_INFO
                              (Linux-specific) */
};
#ifndef SEM_STAT_ANY
#define SEM_STAT_ANY 20
#endif

EXPORT int my_semctl(int semid, int semnum, int cmd, union semun b)
{
    struct semid_ds semidds;
    void *backup = NULL;
    if ((cmd == IPC_STAT) || (cmd == IPC_SET) || (cmd == SEM_STAT) || (cmd == SEM_STAT_ANY)) {
        backup = b.buf;
        b.buf = &semidds;
        if (cmd == IPC_SET) {
            AlignSemidDs(&semidds, backup);
        }
    }
    int ret = semctl(semid, semnum, cmd, b);
    if ((cmd == IPC_STAT) || (cmd == IPC_SET) || (cmd == SEM_STAT) || (cmd == SEM_STAT_ANY)) {
        b.buf = backup;
        if (cmd == IPC_STAT) {
            UnalignSemidDs(backup, &semidds);
        }
    }
    return ret;
}

// Backtrace stuff

#include "elfs/elfdwarf_private.h"
EXPORT int my_backtrace(x64emu_t* emu, void** buffer, int size)
{
    if (!size) return 0;
    dwarf_unwind_t *unwind = init_dwarf_unwind_registers(emu);
    int idx = 0;
    char success = 0;
    uintptr_t addr = *(uintptr_t*)R_RSP;
    buffer[0] = (void*)addr;
    while (++idx < size) {
        uintptr_t ret_addr = get_parent_registers(unwind, FindElfAddress(my_context, addr), addr, &success);
        if (ret_addr == (uintptr_t)GetExit()) {
            // TODO: do something to be able to get the function name
            buffer[idx] = (void*)ret_addr;
            success = 2;
            // See elfdwarf_private.c for the register mapping
            unwind->regs[6] = unwind->regs[7]; // mov rsp, rbp
            unwind->regs[7] = *(uint64_t*)unwind->regs[6]; // pop rbp
            unwind->regs[6] += 8;
            ret_addr = *(uint64_t*)unwind->regs[6]; // ret
            unwind->regs[6] += 8;
            if (++idx < size) buffer[idx] = (void*)ret_addr;
        } else if (!success) break;
        else buffer[idx] = (void*)ret_addr;
        addr = ret_addr;
    }
    free_dwarf_unwind_registers(&unwind);
    return idx;
}

EXPORT char** my_backtrace_symbols(x64emu_t* emu, uintptr_t* buffer, int size)
{
    (void)emu;
    char** ret = (char**)calloc(1, size*sizeof(char*) + size*100);  // capping each strings to 100 chars
    char* s = (char*)(ret+size);
    for (int i=0; i<size; ++i) {
        uintptr_t start = 0;
        uint64_t sz = 0;
        elfheader_t *hdr = FindElfAddress(my_context, buffer[i]);
        const char* symbname = FindNearestSymbolName(hdr, (void*)buffer[i], &start, &sz);
        if (symbname && buffer[i]>=start && (buffer[i]<(start+sz) || !sz)) {
            snprintf(s, 100, "%s(%s+%lx) [%p]", ElfName(hdr), symbname, buffer[i] - start, (void*)buffer[i]);
        } else if (hdr) {
            snprintf(s, 100, "%s+%lx [%p]", ElfName(hdr), buffer[i] - (uintptr_t)GetBaseAddress(hdr), (void*)buffer[i]);
        } else {
            snprintf(s, 100, "??? [%p]", (void*)buffer[i]);
        }
        ret[i] = s;
        s += 100;
    }
    return ret;
}

EXPORT void my_backtrace_symbols_fd(x64emu_t* emu, uintptr_t* buffer, int size, int fd)
{
    (void)emu;
    char s[100];
    for (int i=0; i<size; ++i) {
        uintptr_t start = 0;
        uint64_t sz = 0;
        const char* symbname = FindNearestSymbolName(FindElfAddress(my_context, buffer[i]), (void*)buffer[i], &start, &sz);
        if(symbname && buffer[i]>=start && (buffer[i]<(start+sz) || !sz))
            snprintf(s, 100, "%s+%ld [%p]\n", symbname, buffer[i] - start, (void*)buffer[i]);
        else 
            snprintf(s, 100, "??? [%p]\n", (void*)buffer[i]);
        int dummy = write(fd, s, strlen(s));
        (void)dummy;
    }
}

EXPORT int my_iopl(x64emu_t* emu, int level)
{
    // Set I/O permission (so access IN/OUT opcodes) Default is 0. Can Set to 0..3
    // set permission for all 65536 ports addresses
    // note ioperm can set individual permission
    /*static iFi_t real_iopl = NULL;
    static int searched = 0;
    if(!searched) {
        searched = 1;
        real_iopl = (iFi_t)dlsym(my_lib, "iopl");
    }
    if(real_iopl)
        return real_iopl(level);*/
    // For now, lets just return "unsupported"
    errno = ENOSYS;
    return -1;
}

EXPORT int my_stime(x64emu_t* emu, const time_t *t)
{
    // TODO?
    errno = EPERM;
    return -1;
}

typedef struct clone_arg_s {
 x64emu_t* emu;
 uintptr_t fnc;
 void* args;
 int stack_clone_used;
} clone_arg_t;
static int clone_fn(void* p)
{
    clone_arg_t* arg = (clone_arg_t*)p;
    x64emu_t *emu = arg->emu;
    thread_set_emu(emu);
    int ret = RunFunction(my_context, arg->fnc, 1, arg->args);
    FreeX64Emu(&emu);
    if(arg->stack_clone_used)
        my_context->stack_clone_used = 0;
    free(arg);
    return ret;
}

EXPORT int my_clone(x64emu_t* emu, void* fn, void* stack, int flags, void* args, void* parent, void* tls, void* child)
{
    x64emu_t * newemu = NewX64Emu(emu->context, R_RIP, (uintptr_t)stack, 0, 0);
    SetupX64Emu(newemu);
    CloneEmu(newemu, emu);
    void* mystack = NULL;
    clone_arg_t* arg = (clone_arg_t*)calloc(1, sizeof(clone_arg_t));
    if(my_context->stack_clone_used) {
        mystack = malloc(1024*1024);  // stack for own process... memory leak, but no practical way to remove it
    } else {
        if(!my_context->stack_clone)
            my_context->stack_clone = malloc(1024*1024);
        mystack = my_context->stack_clone;
        my_context->stack_clone_used = 1;
        arg->stack_clone_used = 1;
    }
    arg->emu = newemu;
    arg->args = args;
    arg->fnc = (uintptr_t)fn;
    // x86_64 raw clone is long clone(unsigned long flags, void *stack, int *parent_tid, int *child_tid, unsigned long tls);
    int64_t ret = clone(clone_fn, (void*)((uintptr_t)mystack+1024*1024), flags, arg, parent, tls, child);
    return (uintptr_t)ret;
}

EXPORT char** my_environ = NULL;
EXPORT char** my__environ = NULL;
EXPORT char** my___environ = NULL;  // all aliases

EXPORT char* my___progname = NULL;
EXPORT char* my___progname_full = NULL;
EXPORT char* my_program_invocation_name = NULL;
EXPORT char* my_program_invocation_short_name = NULL;

// ignoring this for now
EXPORT char my___libc_single_threaded = 0;

#define PRE_INIT\
    if(box64_tcmalloc_minimal)                                      \
        lib->priv.w.lib = dlopen(NULL, RTLD_LAZY | RTLD_GLOBAL);    \
    else

#define CUSTOM_INIT         \
    box64->libclib = lib;   \
    /*InitCpuModel();*/         \
    ctSetup();              \
    stSetup(box64);         \
    /*obstackSetup();*/         \
    my_environ = my__environ = my___environ = box64->envv;                      \
    my___progname_full = my_program_invocation_name = box64->argv[0];           \
    my___progname = my_program_invocation_short_name =                          \
        strrchr(box64->argv[0], '/') + 1;                                       \
    getMy(lib);                                                                 \
    setNeededLibs(lib, 3,                                              \
        "ld-linux-x86-64.so.2",                                                 \
        "libpthread.so.0",                                                      \
        "librt.so.1");

#define CUSTOM_FINI \
    freeMy();

#include "wrappedlib_init.h"