// Mark stack as non-executable
#if defined(__linux__) && defined(__ELF__)
.section	.note.GNU-stack, "", @progbits
#endif

#include "regs.h"
#define ARG0 RUSTRT_ARG0_S
#define ARG1 RUSTRT_ARG1_S

        .text

/*
According to ABI documentation found at
http://www.x86-64.org/documentation.html
and Microsoft discussion at
http://msdn.microsoft.com/en-US/library/9z1stfyw%28v=VS.80%29.aspx.

BOTH CALLING CONVENTIONS

Callee save registers:
	R12--R15, RDI, RSI, RBX, RBP, RSP
        XMM0--XMM5

Caller save registers:
	RAX, RCX, RDX, R8--R11
        XMM6--XMM15
        Floating point stack

MAC/AMD CALLING CONVENTIONS

Integer arguments go in registers:
        rdi, rsi, rdx, rcx, r8, r9

User flags have no specified role and are not preserved
        across calls, with the exception of DF in %rFLAGS,
        which must be clear (set to "forward" direction)
        on function entry and return.

MICROSOFT CALLING CONVENTIONS

Return value: RAX

First four arguments:
        RCX, RDX, R8, R9
        XMM0, XMM1, XMM2, XMM3
*/

/*
        Stores current registers into arg0/RCX and restores
        registers found in arg1/RDX. This is used by our
	    implementation of getcontext.  Only saves/restores nonvolatile
        registers and the register used for the first argument.
        Volatile registers in general ought to be saved by the caller
        anyhow.
*/

#if defined(__APPLE__)
#define SWAP_REGISTERS _swap_registers
#else
#define SWAP_REGISTERS swap_registers
#endif

// swap_registers(registers_t *oregs, registers_t *regs)
.globl SWAP_REGISTERS
SWAP_REGISTERS:
        // n.b. when we enter, the return address is at the top of
        // the stack (i.e., 0(%RSP)) and the argument is in
        // RUSTRT_ARG0_S.  We
        // simply save all NV registers into oregs.
        // We then restore all NV registers from regs.  This restores
        // the old stack pointer, which should include the proper
        // return address. We can therefore just return normally to
        // jump back into the old code.

        // Save instruction pointer:
        pop %rax
        mov %rax, (RUSTRT_IP*8)(RUSTRT_ARG0_S)

        // Save non-volatile integer registers:
        //   (including RSP)
        mov %rbx, (RUSTRT_RBX*8)(ARG0)
        mov %rsp, (RUSTRT_RSP*8)(ARG0)
        mov %rbp, (RUSTRT_RBP*8)(ARG0)
        mov %r12, (RUSTRT_R12*8)(ARG0)
        mov %r13, (RUSTRT_R13*8)(ARG0)
        mov %r14, (RUSTRT_R14*8)(ARG0)
        mov %r15, (RUSTRT_R15*8)(ARG0)

#if defined(__MINGW32__) || defined(_WINDOWS)
        mov %rdi, (RUSTRT_RDI*8)(ARG0)
        mov %rsi, (RUSTRT_RSI*8)(ARG0)

        // Save stack range
        mov %gs:0x08, %r8
        mov %r8, (RUSTRT_ST1*8)(ARG0)
        mov %gs:0x10, %r9
        mov %r9, (RUSTRT_ST2*8)(ARG0)
#endif

        // Save 0th argument register:
        mov ARG0, (RUSTRT_ARG0*8)(ARG0)

        // Save non-volatile XMM registers:
#if defined(__MINGW32__) || defined(_WINDOWS)
        movapd %xmm6, (RUSTRT_XMM6*8)(ARG0)
        movapd %xmm7, (RUSTRT_XMM7*8)(ARG0)
        movapd %xmm8, (RUSTRT_XMM8*8)(ARG0)
        movapd %xmm9, (RUSTRT_XMM9*8)(ARG0)
        movapd %xmm10, (RUSTRT_XMM10*8)(ARG0)
        movapd %xmm11, (RUSTRT_XMM11*8)(ARG0)
        movapd %xmm12, (RUSTRT_XMM12*8)(ARG0)
        movapd %xmm13, (RUSTRT_XMM13*8)(ARG0)
        movapd %xmm14, (RUSTRT_XMM14*8)(ARG0)
        movapd %xmm15, (RUSTRT_XMM15*8)(ARG0)
#else
        movapd %xmm0, (RUSTRT_XMM0*8)(ARG0)
        movapd %xmm1, (RUSTRT_XMM1*8)(ARG0)
        movapd %xmm2, (RUSTRT_XMM2*8)(ARG0)
        movapd %xmm3, (RUSTRT_XMM3*8)(ARG0)
        movapd %xmm4, (RUSTRT_XMM4*8)(ARG0)
        movapd %xmm5, (RUSTRT_XMM5*8)(ARG0)
#endif

        // Restore non-volatile integer registers:
        //   (including RSP)
        mov (RUSTRT_RBX*8)(ARG1), %rbx
        mov (RUSTRT_RSP*8)(ARG1), %rsp
        mov (RUSTRT_RBP*8)(ARG1), %rbp
        mov (RUSTRT_R12*8)(ARG1), %r12
        mov (RUSTRT_R13*8)(ARG1), %r13
        mov (RUSTRT_R14*8)(ARG1), %r14
        mov (RUSTRT_R15*8)(ARG1), %r15

#if defined(__MINGW32__) || defined(_WINDOWS)
        mov (RUSTRT_RDI*8)(ARG1), %rdi
        mov (RUSTRT_RSI*8)(ARG1), %rsi

        // Restore stack range
        mov (RUSTRT_ST1*8)(ARG1), %r8
        mov %r8, %gs:0x08
        mov (RUSTRT_ST2*8)(ARG1), %r9
        mov %r9, %gs:0x10
#endif

        // Restore 0th argument register:
        mov (RUSTRT_ARG0*8)(ARG1), ARG0

        // Restore non-volatile XMM registers:
#if defined(__MINGW32__) || defined(_WINDOWS)
        movapd (RUSTRT_XMM6*8)(ARG1), %xmm6
        movapd (RUSTRT_XMM7*8)(ARG1), %xmm7
        movapd (RUSTRT_XMM8*8)(ARG1), %xmm8
        movapd (RUSTRT_XMM9*8)(ARG1), %xmm9
        movapd (RUSTRT_XMM10*8)(ARG1), %xmm10
        movapd (RUSTRT_XMM11*8)(ARG1), %xmm11
        movapd (RUSTRT_XMM12*8)(ARG1), %xmm12
        movapd (RUSTRT_XMM13*8)(ARG1), %xmm13
        movapd (RUSTRT_XMM14*8)(ARG1), %xmm14
        movapd (RUSTRT_XMM15*8)(ARG1), %xmm15
#else
        movapd (RUSTRT_XMM0*8)(ARG1), %xmm0
        movapd (RUSTRT_XMM1*8)(ARG1), %xmm1
        movapd (RUSTRT_XMM2*8)(ARG1), %xmm2
        movapd (RUSTRT_XMM3*8)(ARG1), %xmm3
        movapd (RUSTRT_XMM4*8)(ARG1), %xmm4
        movapd (RUSTRT_XMM5*8)(ARG1), %xmm5
#endif

        // Jump to the instruction pointer
        // found in regs:
        jmp *(RUSTRT_IP*8)(ARG1)