diff options
Diffstat (limited to 'src/UnwindRegistersSave.S')
-rw-r--r-- | src/UnwindRegistersSave.S | 457 |
1 files changed, 457 insertions, 0 deletions
diff --git a/src/UnwindRegistersSave.S b/src/UnwindRegistersSave.S new file mode 100644 index 000000000000..1275c2a16175 --- /dev/null +++ b/src/UnwindRegistersSave.S @@ -0,0 +1,457 @@ +//===------------------------ UnwindRegistersSave.S -----------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is dual licensed under the MIT and the University of Illinois Open +// Source Licenses. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "assembly.h" + + .text + +#if defined(__i386__) + +# +# extern int unw_getcontext(unw_context_t* thread_state) +# +# On entry: +# + + +# +-----------------------+ +# + thread_state pointer + +# +-----------------------+ +# + return address + +# +-----------------------+ <-- SP +# + + +# +DEFINE_LIBUNWIND_FUNCTION(unw_getcontext) + push %eax + movl 8(%esp), %eax + movl %ebx, 4(%eax) + movl %ecx, 8(%eax) + movl %edx, 12(%eax) + movl %edi, 16(%eax) + movl %esi, 20(%eax) + movl %ebp, 24(%eax) + movl %esp, %edx + addl $8, %edx + movl %edx, 28(%eax) # store what sp was at call site as esp + # skip ss + # skip eflags + movl 4(%esp), %edx + movl %edx, 40(%eax) # store return address as eip + # skip cs + # skip ds + # skip es + # skip fs + # skip gs + movl (%esp), %edx + movl %edx, (%eax) # store original eax + popl %eax + xorl %eax, %eax # return UNW_ESUCCESS + ret + +#elif defined(__x86_64__) + +# +# extern int unw_getcontext(unw_context_t* thread_state) +# +# On entry: +# thread_state pointer is in rdi +# +DEFINE_LIBUNWIND_FUNCTION(unw_getcontext) + movq %rax, (%rdi) + movq %rbx, 8(%rdi) + movq %rcx, 16(%rdi) + movq %rdx, 24(%rdi) + movq %rdi, 32(%rdi) + movq %rsi, 40(%rdi) + movq %rbp, 48(%rdi) + movq %rsp, 56(%rdi) + addq $8, 56(%rdi) + movq %r8, 64(%rdi) + movq %r9, 72(%rdi) + movq %r10, 80(%rdi) + movq %r11, 88(%rdi) + movq %r12, 96(%rdi) + movq %r13,104(%rdi) + movq %r14,112(%rdi) + movq %r15,120(%rdi) + movq (%rsp),%rsi + movq %rsi,128(%rdi) # store return address as rip + # skip rflags + # skip cs + # skip fs + # skip gs + xorl %eax, %eax # return UNW_ESUCCESS + ret + +#elif defined(__ppc__) + +; +; extern int unw_getcontext(unw_context_t* thread_state) +; +; On entry: +; thread_state pointer is in r3 +; +DEFINE_LIBUNWIND_FUNCTION(unw_getcontext) + stw r0, 8(r3) + mflr r0 + stw r0, 0(r3) ; store lr as ssr0 + stw r1, 12(r3) + stw r2, 16(r3) + stw r3, 20(r3) + stw r4, 24(r3) + stw r5, 28(r3) + stw r6, 32(r3) + stw r7, 36(r3) + stw r8, 40(r3) + stw r9, 44(r3) + stw r10, 48(r3) + stw r11, 52(r3) + stw r12, 56(r3) + stw r13, 60(r3) + stw r14, 64(r3) + stw r15, 68(r3) + stw r16, 72(r3) + stw r17, 76(r3) + stw r18, 80(r3) + stw r19, 84(r3) + stw r20, 88(r3) + stw r21, 92(r3) + stw r22, 96(r3) + stw r23,100(r3) + stw r24,104(r3) + stw r25,108(r3) + stw r26,112(r3) + stw r27,116(r3) + stw r28,120(r3) + stw r29,124(r3) + stw r30,128(r3) + stw r31,132(r3) + + ; save VRSave register + mfspr r0,256 + stw r0,156(r3) + ; save CR registers + mfcr r0 + stw r0,136(r3) + ; save CTR register + mfctr r0 + stw r0,148(r3) + + ; save float registers + stfd f0, 160(r3) + stfd f1, 168(r3) + stfd f2, 176(r3) + stfd f3, 184(r3) + stfd f4, 192(r3) + stfd f5, 200(r3) + stfd f6, 208(r3) + stfd f7, 216(r3) + stfd f8, 224(r3) + stfd f9, 232(r3) + stfd f10,240(r3) + stfd f11,248(r3) + stfd f12,256(r3) + stfd f13,264(r3) + stfd f14,272(r3) + stfd f15,280(r3) + stfd f16,288(r3) + stfd f17,296(r3) + stfd f18,304(r3) + stfd f19,312(r3) + stfd f20,320(r3) + stfd f21,328(r3) + stfd f22,336(r3) + stfd f23,344(r3) + stfd f24,352(r3) + stfd f25,360(r3) + stfd f26,368(r3) + stfd f27,376(r3) + stfd f28,384(r3) + stfd f29,392(r3) + stfd f30,400(r3) + stfd f31,408(r3) + + + ; save vector registers + + subi r4,r1,16 + rlwinm r4,r4,0,0,27 ; mask low 4-bits + ; r4 is now a 16-byte aligned pointer into the red zone + +#define SAVE_VECTOR_UNALIGNED(_vec, _offset) \ + stvx _vec,0,r4 @\ + lwz r5, 0(r4) @\ + stw r5, _offset(r3) @\ + lwz r5, 4(r4) @\ + stw r5, _offset+4(r3) @\ + lwz r5, 8(r4) @\ + stw r5, _offset+8(r3) @\ + lwz r5, 12(r4) @\ + stw r5, _offset+12(r3) + + SAVE_VECTOR_UNALIGNED( v0, 424+0x000) + SAVE_VECTOR_UNALIGNED( v1, 424+0x010) + SAVE_VECTOR_UNALIGNED( v2, 424+0x020) + SAVE_VECTOR_UNALIGNED( v3, 424+0x030) + SAVE_VECTOR_UNALIGNED( v4, 424+0x040) + SAVE_VECTOR_UNALIGNED( v5, 424+0x050) + SAVE_VECTOR_UNALIGNED( v6, 424+0x060) + SAVE_VECTOR_UNALIGNED( v7, 424+0x070) + SAVE_VECTOR_UNALIGNED( v8, 424+0x080) + SAVE_VECTOR_UNALIGNED( v9, 424+0x090) + SAVE_VECTOR_UNALIGNED(v10, 424+0x0A0) + SAVE_VECTOR_UNALIGNED(v11, 424+0x0B0) + SAVE_VECTOR_UNALIGNED(v12, 424+0x0C0) + SAVE_VECTOR_UNALIGNED(v13, 424+0x0D0) + SAVE_VECTOR_UNALIGNED(v14, 424+0x0E0) + SAVE_VECTOR_UNALIGNED(v15, 424+0x0F0) + SAVE_VECTOR_UNALIGNED(v16, 424+0x100) + SAVE_VECTOR_UNALIGNED(v17, 424+0x110) + SAVE_VECTOR_UNALIGNED(v18, 424+0x120) + SAVE_VECTOR_UNALIGNED(v19, 424+0x130) + SAVE_VECTOR_UNALIGNED(v20, 424+0x140) + SAVE_VECTOR_UNALIGNED(v21, 424+0x150) + SAVE_VECTOR_UNALIGNED(v22, 424+0x160) + SAVE_VECTOR_UNALIGNED(v23, 424+0x170) + SAVE_VECTOR_UNALIGNED(v24, 424+0x180) + SAVE_VECTOR_UNALIGNED(v25, 424+0x190) + SAVE_VECTOR_UNALIGNED(v26, 424+0x1A0) + SAVE_VECTOR_UNALIGNED(v27, 424+0x1B0) + SAVE_VECTOR_UNALIGNED(v28, 424+0x1C0) + SAVE_VECTOR_UNALIGNED(v29, 424+0x1D0) + SAVE_VECTOR_UNALIGNED(v30, 424+0x1E0) + SAVE_VECTOR_UNALIGNED(v31, 424+0x1F0) + + li r3, 0 ; return UNW_ESUCCESS + blr + + +#elif defined(__arm64__) || defined(__aarch64__) + +// +// extern int unw_getcontext(unw_context_t* thread_state) +// +// On entry: +// thread_state pointer is in x0 +// + .p2align 2 +DEFINE_LIBUNWIND_FUNCTION(unw_getcontext) + stp x0, x1, [x0, #0x000] + stp x2, x3, [x0, #0x010] + stp x4, x5, [x0, #0x020] + stp x6, x7, [x0, #0x030] + stp x8, x9, [x0, #0x040] + stp x10,x11, [x0, #0x050] + stp x12,x13, [x0, #0x060] + stp x14,x15, [x0, #0x070] + stp x16,x17, [x0, #0x080] + stp x18,x19, [x0, #0x090] + stp x20,x21, [x0, #0x0A0] + stp x22,x23, [x0, #0x0B0] + stp x24,x25, [x0, #0x0C0] + stp x26,x27, [x0, #0x0D0] + stp x28,fp, [x0, #0x0E0] + str lr, [x0, #0x0F0] + mov x1,sp + str x1, [x0, #0x0F8] + str lr, [x0, #0x100] // store return address as pc + // skip cpsr + stp d0, d1, [x0, #0x110] + stp d2, d3, [x0, #0x120] + stp d4, d5, [x0, #0x130] + stp d6, d7, [x0, #0x140] + stp d8, d9, [x0, #0x150] + stp d10,d11, [x0, #0x160] + stp d12,d13, [x0, #0x170] + stp d14,d15, [x0, #0x180] + stp d16,d17, [x0, #0x190] + stp d18,d19, [x0, #0x1A0] + stp d20,d21, [x0, #0x1B0] + stp d22,d23, [x0, #0x1C0] + stp d24,d25, [x0, #0x1D0] + stp d26,d27, [x0, #0x1E0] + stp d28,d29, [x0, #0x1F0] + str d30, [x0, #0x200] + str d31, [x0, #0x208] + mov x0, #0 // return UNW_ESUCCESS + ret + +#elif defined(__arm__) && !defined(__APPLE__) + +#if !defined(__ARM_ARCH_ISA_ARM) + .thumb +#endif + +@ +@ extern int unw_getcontext(unw_context_t* thread_state) +@ +@ On entry: +@ thread_state pointer is in r0 +@ +@ Per EHABI #4.7 this only saves the core integer registers. +@ EHABI #7.4.5 notes that in general all VRS registers should be restored +@ however this is very hard to do for VFP registers because it is unknown +@ to the library how many registers are implemented by the architecture. +@ Instead, VFP registers are demand saved by logic external to unw_getcontext. +@ + .p2align 2 +DEFINE_LIBUNWIND_FUNCTION(unw_getcontext) +#if !defined(__ARM_ARCH_ISA_ARM) + stm r0, {r0-r7} + mov r2, sp + mov r3, lr + str r2, [r0, #52] + str r3, [r0, #56] + str r3, [r0, #60] @ store return address as pc +#else + @ 32bit thumb-2 restrictions for stm: + @ . the sp (r13) cannot be in the list + @ . the pc (r15) cannot be in the list in an STM instruction + stm r0, {r0-r12} + str sp, [r0, #52] + str lr, [r0, #56] + str lr, [r0, #60] @ store return address as pc +#endif +#if __ARM_ARCH_ISA_THUMB == 1 + @ T1 does not have a non-cpsr-clobbering register-zeroing instruction. + @ It is safe to use here though because we are about to return, and cpsr is + @ not expected to be preserved. + movs r0, #0 @ return UNW_ESUCCESS +#else + mov r0, #0 @ return UNW_ESUCCESS +#endif + JMP(lr) + +@ +@ static void libunwind::Registers_arm::saveVFPWithFSTMD(unw_fpreg_t* values) +@ +@ On entry: +@ values pointer is in r0 +@ + .p2align 2 + .fpu vfpv3-d16 +DEFINE_LIBUNWIND_PRIVATE_FUNCTION(_ZN9libunwind13Registers_arm16saveVFPWithFSTMDEPy) + vstmia r0, {d0-d15} + JMP(lr) + +@ +@ static void libunwind::Registers_arm::saveVFPWithFSTMX(unw_fpreg_t* values) +@ +@ On entry: +@ values pointer is in r0 +@ + .p2align 2 + .fpu vfpv3-d16 +DEFINE_LIBUNWIND_PRIVATE_FUNCTION(_ZN9libunwind13Registers_arm16saveVFPWithFSTMXEPy) + vstmia r0, {d0-d15} @ fstmiax is deprecated in ARMv7+ and now behaves like vstmia + JMP(lr) + +@ +@ static void libunwind::Registers_arm::saveVFPv3(unw_fpreg_t* values) +@ +@ On entry: +@ values pointer is in r0 +@ + .p2align 2 + .fpu vfpv3 +DEFINE_LIBUNWIND_PRIVATE_FUNCTION(_ZN9libunwind13Registers_arm9saveVFPv3EPy) + @ VFP and iwMMX instructions are only available when compiling with the flags + @ that enable them. We do not want to do that in the library (because we do not + @ want the compiler to generate instructions that access those) but this is + @ only accessed if the personality routine needs these registers. Use of + @ these registers implies they are, actually, available on the target, so + @ it's ok to execute. + @ So, generate the instructions using the corresponding coprocessor mnemonic. + vstmia r0, {d16-d31} + JMP(lr) + +@ +@ static void libunwind::Registers_arm::saveiWMMX(unw_fpreg_t* values) +@ +@ On entry: +@ values pointer is in r0 +@ + .p2align 2 +DEFINE_LIBUNWIND_PRIVATE_FUNCTION(_ZN9libunwind13Registers_arm9saveiWMMXEPy) +#if (!defined(__ARM_ARCH_6M__) && !defined(__ARM_ARCH_6SM__)) || defined(__ARM_WMMX) + stcl p1, cr0, [r0], #8 @ wstrd wR0, [r0], #8 + stcl p1, cr1, [r0], #8 @ wstrd wR1, [r0], #8 + stcl p1, cr2, [r0], #8 @ wstrd wR2, [r0], #8 + stcl p1, cr3, [r0], #8 @ wstrd wR3, [r0], #8 + stcl p1, cr4, [r0], #8 @ wstrd wR4, [r0], #8 + stcl p1, cr5, [r0], #8 @ wstrd wR5, [r0], #8 + stcl p1, cr6, [r0], #8 @ wstrd wR6, [r0], #8 + stcl p1, cr7, [r0], #8 @ wstrd wR7, [r0], #8 + stcl p1, cr8, [r0], #8 @ wstrd wR8, [r0], #8 + stcl p1, cr9, [r0], #8 @ wstrd wR9, [r0], #8 + stcl p1, cr10, [r0], #8 @ wstrd wR10, [r0], #8 + stcl p1, cr11, [r0], #8 @ wstrd wR11, [r0], #8 + stcl p1, cr12, [r0], #8 @ wstrd wR12, [r0], #8 + stcl p1, cr13, [r0], #8 @ wstrd wR13, [r0], #8 + stcl p1, cr14, [r0], #8 @ wstrd wR14, [r0], #8 + stcl p1, cr15, [r0], #8 @ wstrd wR15, [r0], #8 +#endif + JMP(lr) + +@ +@ static void libunwind::Registers_arm::saveiWMMXControl(unw_uint32_t* values) +@ +@ On entry: +@ values pointer is in r0 +@ + .p2align 2 +DEFINE_LIBUNWIND_PRIVATE_FUNCTION(_ZN9libunwind13Registers_arm16saveiWMMXControlEPj) +#if (!defined(__ARM_ARCH_6M__) && !defined(__ARM_ARCH_6SM__)) || defined(__ARM_WMMX) + stc2 p1, cr8, [r0], #4 @ wstrw wCGR0, [r0], #4 + stc2 p1, cr9, [r0], #4 @ wstrw wCGR1, [r0], #4 + stc2 p1, cr10, [r0], #4 @ wstrw wCGR2, [r0], #4 + stc2 p1, cr11, [r0], #4 @ wstrw wCGR3, [r0], #4 +#endif + JMP(lr) + +#elif defined(__or1k__) + +# +# extern int unw_getcontext(unw_context_t* thread_state) +# +# On entry: +# thread_state pointer is in r3 +# +DEFINE_LIBUNWIND_FUNCTION(unw_getcontext) + l.sw 0(r3), r0 + l.sw 4(r3), r1 + l.sw 8(r3), r2 + l.sw 12(r3), r3 + l.sw 16(r3), r4 + l.sw 20(r3), r5 + l.sw 24(r3), r6 + l.sw 28(r3), r7 + l.sw 32(r3), r8 + l.sw 36(r3), r9 + l.sw 40(r3), r10 + l.sw 44(r3), r11 + l.sw 48(r3), r12 + l.sw 52(r3), r13 + l.sw 56(r3), r14 + l.sw 60(r3), r15 + l.sw 64(r3), r16 + l.sw 68(r3), r17 + l.sw 72(r3), r18 + l.sw 76(r3), r19 + l.sw 80(r3), r20 + l.sw 84(r3), r21 + l.sw 88(r3), r22 + l.sw 92(r3), r23 + l.sw 96(r3), r24 + l.sw 100(r3), r25 + l.sw 104(r3), r26 + l.sw 108(r3), r27 + l.sw 112(r3), r28 + l.sw 116(r3), r29 + l.sw 120(r3), r30 + l.sw 124(r3), r31 +#endif |