| 1 |
|
| 2 | |
| 3 | * Copyright 2016, Cyril Bur, IBM Corp.
|
| 4 |
|
| 5 |
|
| 6 | #ifndef _SELFTESTS_POWERPC_FPU_ASM_H
|
| 7 | #define _SELFTESTS_POWERPC_FPU_ASM_H
|
| 8 | #include "basic_asm.h"
|
| 9 |
|
| 10 | #define PUSH_FPU(stack_size) \
|
| 11 | stfd f31,(stack_size + STACK_FRAME_MIN_SIZE)(%r1); \
|
| 12 | stfd f30,(stack_size + STACK_FRAME_MIN_SIZE - 8)(%r1); \
|
| 13 | stfd f29,(stack_size + STACK_FRAME_MIN_SIZE - 16)(%r1); \
|
| 14 | stfd f28,(stack_size + STACK_FRAME_MIN_SIZE - 24)(%r1); \
|
| 15 | stfd f27,(stack_size + STACK_FRAME_MIN_SIZE - 32)(%r1); \
|
| 16 | stfd f26,(stack_size + STACK_FRAME_MIN_SIZE - 40)(%r1); \
|
| 17 | stfd f25,(stack_size + STACK_FRAME_MIN_SIZE - 48)(%r1); \
|
| 18 | stfd f24,(stack_size + STACK_FRAME_MIN_SIZE - 56)(%r1); \
|
| 19 | stfd f23,(stack_size + STACK_FRAME_MIN_SIZE - 64)(%r1); \
|
| 20 | stfd f22,(stack_size + STACK_FRAME_MIN_SIZE - 72)(%r1); \
|
| 21 | stfd f21,(stack_size + STACK_FRAME_MIN_SIZE - 80)(%r1); \
|
| 22 | stfd f20,(stack_size + STACK_FRAME_MIN_SIZE - 88)(%r1); \
|
| 23 | stfd f19,(stack_size + STACK_FRAME_MIN_SIZE - 96)(%r1); \
|
| 24 | stfd f18,(stack_size + STACK_FRAME_MIN_SIZE - 104)(%r1); \
|
| 25 | stfd f17,(stack_size + STACK_FRAME_MIN_SIZE - 112)(%r1); \
|
| 26 | stfd f16,(stack_size + STACK_FRAME_MIN_SIZE - 120)(%r1); \
|
| 27 | stfd f15,(stack_size + STACK_FRAME_MIN_SIZE - 128)(%r1); \
|
| 28 | stfd f14,(stack_size + STACK_FRAME_MIN_SIZE - 136)(%r1);
|
| 29 |
|
| 30 | #define POP_FPU(stack_size) \
|
| 31 | lfd f31,(stack_size + STACK_FRAME_MIN_SIZE)(%r1); \
|
| 32 | lfd f30,(stack_size + STACK_FRAME_MIN_SIZE - 8)(%r1); \
|
| 33 | lfd f29,(stack_size + STACK_FRAME_MIN_SIZE - 16)(%r1); \
|
| 34 | lfd f28,(stack_size + STACK_FRAME_MIN_SIZE - 24)(%r1); \
|
| 35 | lfd f27,(stack_size + STACK_FRAME_MIN_SIZE - 32)(%r1); \
|
| 36 | lfd f26,(stack_size + STACK_FRAME_MIN_SIZE - 40)(%r1); \
|
| 37 | lfd f25,(stack_size + STACK_FRAME_MIN_SIZE - 48)(%r1); \
|
| 38 | lfd f24,(stack_size + STACK_FRAME_MIN_SIZE - 56)(%r1); \
|
| 39 | lfd f23,(stack_size + STACK_FRAME_MIN_SIZE - 64)(%r1); \
|
| 40 | lfd f22,(stack_size + STACK_FRAME_MIN_SIZE - 72)(%r1); \
|
| 41 | lfd f21,(stack_size + STACK_FRAME_MIN_SIZE - 80)(%r1); \
|
| 42 | lfd f20,(stack_size + STACK_FRAME_MIN_SIZE - 88)(%r1); \
|
| 43 | lfd f19,(stack_size + STACK_FRAME_MIN_SIZE - 96)(%r1); \
|
| 44 | lfd f18,(stack_size + STACK_FRAME_MIN_SIZE - 104)(%r1); \
|
| 45 | lfd f17,(stack_size + STACK_FRAME_MIN_SIZE - 112)(%r1); \
|
| 46 | lfd f16,(stack_size + STACK_FRAME_MIN_SIZE - 120)(%r1); \
|
| 47 | lfd f15,(stack_size + STACK_FRAME_MIN_SIZE - 128)(%r1); \
|
| 48 | lfd f14,(stack_size + STACK_FRAME_MIN_SIZE - 136)(%r1);
|
| 49 |
|
| 50 | |
| 51 | * Careful calling this, it will 'clobber' fpu (by design)
|
| 52 | * Don't call this from C
|
| 53 |
|
| 54 | FUNC_START(load_fpu)
|
| 55 | lfd f14,0(r3)
|
| 56 | lfd f15,8(r3)
|
| 57 | lfd f16,16(r3)
|
| 58 | lfd f17,24(r3)
|
| 59 | lfd f18,32(r3)
|
| 60 | lfd f19,40(r3)
|
| 61 | lfd f20,48(r3)
|
| 62 | lfd f21,56(r3)
|
| 63 | lfd f22,64(r3)
|
| 64 | lfd f23,72(r3)
|
| 65 | lfd f24,80(r3)
|
| 66 | lfd f25,88(r3)
|
| 67 | lfd f26,96(r3)
|
| 68 | lfd f27,104(r3)
|
| 69 | lfd f28,112(r3)
|
| 70 | lfd f29,120(r3)
|
| 71 | lfd f30,128(r3)
|
| 72 | lfd f31,136(r3)
|
| 73 | blr
|
| 74 | FUNC_END(load_fpu)
|
| 75 |
|
| 76 | #endif
|