From b3bd00f159bb5614176d8219a0dfbe0a656defb1 Mon Sep 17 00:00:00 2001 From: gbeauche <> Date: Mon, 1 Nov 2004 17:12:55 +0000 Subject: [PATCH] - affine need_to_preserve[] to get close to linux/x86_64 ABI - optimize NOP fillers on x86-64 (based on GNU as implementation) --- .../src/uae_cpu/compiler/codegen_x86.cpp | 29 ++++++++++++++++++- 1 file changed, 28 insertions(+), 1 deletion(-) diff --git a/BasiliskII/src/uae_cpu/compiler/codegen_x86.cpp b/BasiliskII/src/uae_cpu/compiler/codegen_x86.cpp index 361c8047..e4317d01 100644 --- a/BasiliskII/src/uae_cpu/compiler/codegen_x86.cpp +++ b/BasiliskII/src/uae_cpu/compiler/codegen_x86.cpp @@ -107,7 +107,12 @@ uae_u8 call_saved[]={0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0}; - Special registers (such like the stack pointer) should not be "preserved" by pushing, even though they are "saved" across function calls */ -uae_u8 need_to_preserve[]={1,1,1,1,0,1,1,1,1,1,1,1,1,1,1,1}; +#if defined(__x86_64__) +/* callee-saved registers as defined by Linux/x86_64 ABI: rbx, rbp, rsp, r12 - r15 */ +static const uae_u8 need_to_preserve[]={0,0,0,1,0,1,0,0,0,0,0,0,1,1,1,1}; +#else +static const uae_u8 need_to_preserve[]={1,1,1,1,0,1,1,1}; +#endif /* Whether classes of instructions do or don't clobber the native flags */ #define CLOBBER_MOV @@ -3139,7 +3144,28 @@ static __inline__ void raw_emit_nop_filler(int nbytes) f32_1, f32_2, f32_3, f32_4, f32_5, f32_6, f32_7, f32_8, f32_9, f32_10, f32_11, f32_12, f32_13, f32_14, f32_15 }; + static const uae_u8 prefixes[4] = { 0x66, 0x66, 0x66, 0x66 }; +#if defined(__x86_64__) + /* The recommended way to pad 64bit code is to use NOPs preceded by + maximally four 0x66 prefixes. Balance the size of nops. */ + if (nbytes == 0) + return; + + int i; + int nnops = (nbytes + 3) / 4; + int len = nbytes / nnops; + int remains = nbytes - nnops * len; + + for (i = 0; i < remains; i++) { + emit_block(prefixes, len); + raw_nop(); + } + for (; i < nnops; i++) { + emit_block(prefixes, len - 1); + raw_nop(); + } +#else int nloops = nbytes / 16; while (nloops-- > 0) emit_block(f32_16, sizeof(f32_16)); @@ -3147,6 +3173,7 @@ static __inline__ void raw_emit_nop_filler(int nbytes) nbytes %= 16; if (nbytes) emit_block(f32_patt[nbytes - 1], nbytes); +#endif }