diff options
Diffstat (limited to 'tcg/tci')
| -rw-r--r-- | tcg/tci/README | 120 | ||||
| -rw-r--r-- | tcg/tci/tcg-target-con-set.h | 21 | ||||
| -rw-r--r-- | tcg/tci/tcg-target-con-str.h | 11 | ||||
| -rw-r--r-- | tcg/tci/tcg-target.c.inc | 862 | ||||
| -rw-r--r-- | tcg/tci/tcg-target.h | 180 |
5 files changed, 1194 insertions, 0 deletions
diff --git a/tcg/tci/README b/tcg/tci/README new file mode 100644 index 000000000..f72a40a39 --- /dev/null +++ b/tcg/tci/README @@ -0,0 +1,120 @@ +TCG Interpreter (TCI) - Copyright (c) 2011 Stefan Weil. + +This file is released under the BSD license. + +1) Introduction + +TCG (Tiny Code Generator) is a code generator which translates +code fragments ("basic blocks") from target code (any of the +targets supported by QEMU) to a code representation which +can be run on a host. + +QEMU can create native code for some hosts (arm, i386, ia64, ppc, ppc64, +s390, sparc, x86_64). For others, unofficial host support was written. + +By adding a code generator for a virtual machine and using an +interpreter for the generated bytecode, it is possible to +support (almost) any host. + +This is what TCI (Tiny Code Interpreter) does. + +2) Implementation + +Like each TCG host frontend, TCI implements the code generator in +tcg-target.c.inc, tcg-target.h. Both files are in directory tcg/tci. + +The additional file tcg/tci.c adds the interpreter and disassembler. + +The bytecode consists of opcodes (with only a few exceptions, with +the same same numeric values and semantics as used by TCG), and up +to six arguments packed into a 32-bit integer. See comments in tci.c +for details on the encoding. + +3) Usage + +For hosts without native TCG, the interpreter TCI must be enabled by + + configure --enable-tcg-interpreter + +If configure is called without --enable-tcg-interpreter, it will +suggest using this option. Setting it automatically would need +additional code in configure which must be fixed when new native TCG +implementations are added. + +For hosts with native TCG, the interpreter TCI can be enabled by + + configure --enable-tcg-interpreter + +The only difference from running QEMU with TCI to running without TCI +should be speed. Especially during development of TCI, it was very +useful to compare runs with and without TCI. Create /tmp/qemu.log by + + qemu-system-i386 -d in_asm,op_opt,cpu -D /tmp/qemu.log -singlestep + +once with interpreter and once without interpreter and compare the resulting +qemu.log files. This is also useful to see the effects of additional +registers or additional opcodes (it is easy to modify the virtual machine). +It can also be used to verify native TCGs. + +Hosts with native TCG can also enable TCI by claiming to be unsupported: + + configure --cpu=unknown --enable-tcg-interpreter + +configure then no longer uses the native linker script (*.ld) for +user mode emulation. + + +4) Status + +TCI needs special implementation for 32 and 64 bit host, 32 and 64 bit target, +host and target with same or different endianness. + + | host (le) host (be) + | 32 64 32 64 +------------+------------------------------------------------------------ +target (le) | s0, u0 s1, u1 s?, u? s?, u? +32 bit | + | +target (le) | sc, uc s1, u1 s?, u? s?, u? +64 bit | + | +target (be) | sc, u0 sc, uc s?, u? s?, u? +32 bit | + | +target (be) | sc, uc sc, uc s?, u? s?, u? +64 bit | + | + +System emulation +s? = untested +sc = compiles +s0 = bios works +s1 = grub works +s2 = Linux boots + +Linux user mode emulation +u? = untested +uc = compiles +u0 = static hello works +u1 = linux-user-test works + +5) Todo list + +* TCI is not widely tested. It was written and tested on a x86_64 host + running i386 and x86_64 system emulation and Linux user mode. + A cross compiled QEMU for i386 host also works with the same basic tests. + A cross compiled QEMU for mipsel host works, too. It is terribly slow + because I run it in a mips malta emulation, so it is an interpreted + emulation in an emulation. + A cross compiled QEMU for arm host works (tested with pc bios). + A cross compiled QEMU for ppc host works at least partially: + i386-linux-user/qemu-i386 can run a simple hello-world program + (tested in a ppc emulation). + +* Some TCG opcodes are either missing in the code generator and/or + in the interpreter. These opcodes raise a runtime exception, so it is + possible to see where code must be added. + +* It might be useful to have a runtime option which selects the native TCG + or TCI, so QEMU would have to include two TCGs. Today, selecting TCI + is a configure option, so you need two compilations of QEMU. diff --git a/tcg/tci/tcg-target-con-set.h b/tcg/tci/tcg-target-con-set.h new file mode 100644 index 000000000..ae2dc3b84 --- /dev/null +++ b/tcg/tci/tcg-target-con-set.h @@ -0,0 +1,21 @@ +/* SPDX-License-Identifier: MIT */ +/* + * TCI target-specific constraint sets. + * Copyright (c) 2021 Linaro + */ + +/* + * C_On_Im(...) defines a constraint set with <n> outputs and <m> inputs. + * Each operand should be a sequence of constraint letters as defined by + * tcg-target-con-str.h; the constraint combination is inclusive or. + */ +C_O0_I1(r) +C_O0_I2(r, r) +C_O0_I3(r, r, r) +C_O0_I4(r, r, r, r) +C_O1_I1(r, r) +C_O1_I2(r, r, r) +C_O1_I4(r, r, r, r, r) +C_O2_I1(r, r, r) +C_O2_I2(r, r, r, r) +C_O2_I4(r, r, r, r, r, r) diff --git a/tcg/tci/tcg-target-con-str.h b/tcg/tci/tcg-target-con-str.h new file mode 100644 index 000000000..87c0f19e9 --- /dev/null +++ b/tcg/tci/tcg-target-con-str.h @@ -0,0 +1,11 @@ +/* SPDX-License-Identifier: MIT */ +/* + * Define TCI target-specific operand constraints. + * Copyright (c) 2021 Linaro + */ + +/* + * Define constraint letters for register sets: + * REGS(letter, register_mask) + */ +REGS('r', MAKE_64BIT_MASK(0, TCG_TARGET_NB_REGS)) diff --git a/tcg/tci/tcg-target.c.inc b/tcg/tci/tcg-target.c.inc new file mode 100644 index 000000000..0cb16aaa8 --- /dev/null +++ b/tcg/tci/tcg-target.c.inc @@ -0,0 +1,862 @@ +/* + * Tiny Code Generator for QEMU + * + * Copyright (c) 2009, 2011 Stefan Weil + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "../tcg-pool.c.inc" + +static TCGConstraintSetIndex tcg_target_op_def(TCGOpcode op) +{ + switch (op) { + case INDEX_op_goto_ptr: + return C_O0_I1(r); + + case INDEX_op_ld8u_i32: + case INDEX_op_ld8s_i32: + case INDEX_op_ld16u_i32: + case INDEX_op_ld16s_i32: + case INDEX_op_ld_i32: + case INDEX_op_ld8u_i64: + case INDEX_op_ld8s_i64: + case INDEX_op_ld16u_i64: + case INDEX_op_ld16s_i64: + case INDEX_op_ld32u_i64: + case INDEX_op_ld32s_i64: + case INDEX_op_ld_i64: + case INDEX_op_not_i32: + case INDEX_op_not_i64: + case INDEX_op_neg_i32: + case INDEX_op_neg_i64: + case INDEX_op_ext8s_i32: + case INDEX_op_ext8s_i64: + case INDEX_op_ext16s_i32: + case INDEX_op_ext16s_i64: + case INDEX_op_ext8u_i32: + case INDEX_op_ext8u_i64: + case INDEX_op_ext16u_i32: + case INDEX_op_ext16u_i64: + case INDEX_op_ext32s_i64: + case INDEX_op_ext32u_i64: + case INDEX_op_ext_i32_i64: + case INDEX_op_extu_i32_i64: + case INDEX_op_bswap16_i32: + case INDEX_op_bswap16_i64: + case INDEX_op_bswap32_i32: + case INDEX_op_bswap32_i64: + case INDEX_op_bswap64_i64: + case INDEX_op_extract_i32: + case INDEX_op_extract_i64: + case INDEX_op_sextract_i32: + case INDEX_op_sextract_i64: + case INDEX_op_ctpop_i32: + case INDEX_op_ctpop_i64: + return C_O1_I1(r, r); + + case INDEX_op_st8_i32: + case INDEX_op_st16_i32: + case INDEX_op_st_i32: + case INDEX_op_st8_i64: + case INDEX_op_st16_i64: + case INDEX_op_st32_i64: + case INDEX_op_st_i64: + return C_O0_I2(r, r); + + case INDEX_op_div_i32: + case INDEX_op_div_i64: + case INDEX_op_divu_i32: + case INDEX_op_divu_i64: + case INDEX_op_rem_i32: + case INDEX_op_rem_i64: + case INDEX_op_remu_i32: + case INDEX_op_remu_i64: + case INDEX_op_add_i32: + case INDEX_op_add_i64: + case INDEX_op_sub_i32: + case INDEX_op_sub_i64: + case INDEX_op_mul_i32: + case INDEX_op_mul_i64: + case INDEX_op_and_i32: + case INDEX_op_and_i64: + case INDEX_op_andc_i32: + case INDEX_op_andc_i64: + case INDEX_op_eqv_i32: + case INDEX_op_eqv_i64: + case INDEX_op_nand_i32: + case INDEX_op_nand_i64: + case INDEX_op_nor_i32: + case INDEX_op_nor_i64: + case INDEX_op_or_i32: + case INDEX_op_or_i64: + case INDEX_op_orc_i32: + case INDEX_op_orc_i64: + case INDEX_op_xor_i32: + case INDEX_op_xor_i64: + case INDEX_op_shl_i32: + case INDEX_op_shl_i64: + case INDEX_op_shr_i32: + case INDEX_op_shr_i64: + case INDEX_op_sar_i32: + case INDEX_op_sar_i64: + case INDEX_op_rotl_i32: + case INDEX_op_rotl_i64: + case INDEX_op_rotr_i32: + case INDEX_op_rotr_i64: + case INDEX_op_setcond_i32: + case INDEX_op_setcond_i64: + case INDEX_op_deposit_i32: + case INDEX_op_deposit_i64: + case INDEX_op_clz_i32: + case INDEX_op_clz_i64: + case INDEX_op_ctz_i32: + case INDEX_op_ctz_i64: + return C_O1_I2(r, r, r); + + case INDEX_op_brcond_i32: + case INDEX_op_brcond_i64: + return C_O0_I2(r, r); + + case INDEX_op_add2_i32: + case INDEX_op_add2_i64: + case INDEX_op_sub2_i32: + case INDEX_op_sub2_i64: + return C_O2_I4(r, r, r, r, r, r); + +#if TCG_TARGET_REG_BITS == 32 + case INDEX_op_brcond2_i32: + return C_O0_I4(r, r, r, r); +#endif + + case INDEX_op_mulu2_i32: + case INDEX_op_mulu2_i64: + case INDEX_op_muls2_i32: + case INDEX_op_muls2_i64: + return C_O2_I2(r, r, r, r); + + case INDEX_op_movcond_i32: + case INDEX_op_movcond_i64: + case INDEX_op_setcond2_i32: + return C_O1_I4(r, r, r, r, r); + + case INDEX_op_qemu_ld_i32: + return (TARGET_LONG_BITS <= TCG_TARGET_REG_BITS + ? C_O1_I1(r, r) + : C_O1_I2(r, r, r)); + case INDEX_op_qemu_ld_i64: + return (TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, r) + : TARGET_LONG_BITS <= TCG_TARGET_REG_BITS ? C_O2_I1(r, r, r) + : C_O2_I2(r, r, r, r)); + case INDEX_op_qemu_st_i32: + return (TARGET_LONG_BITS <= TCG_TARGET_REG_BITS + ? C_O0_I2(r, r) + : C_O0_I3(r, r, r)); + case INDEX_op_qemu_st_i64: + return (TCG_TARGET_REG_BITS == 64 ? C_O0_I2(r, r) + : TARGET_LONG_BITS <= TCG_TARGET_REG_BITS ? C_O0_I3(r, r, r) + : C_O0_I4(r, r, r, r)); + + default: + g_assert_not_reached(); + } +} + +static const int tcg_target_reg_alloc_order[] = { + TCG_REG_R2, + TCG_REG_R3, + TCG_REG_R4, + TCG_REG_R5, + TCG_REG_R6, + TCG_REG_R7, + TCG_REG_R8, + TCG_REG_R9, + TCG_REG_R10, + TCG_REG_R11, + TCG_REG_R12, + TCG_REG_R13, + TCG_REG_R14, + TCG_REG_R15, + TCG_REG_R1, + TCG_REG_R0, +}; + +#if MAX_OPC_PARAM_IARGS != 6 +# error Fix needed, number of supported input arguments changed! +#endif + +/* No call arguments via registers. All will be stored on the "stack". */ +static const int tcg_target_call_iarg_regs[] = { }; + +static const int tcg_target_call_oarg_regs[] = { + TCG_REG_R0, +#if TCG_TARGET_REG_BITS == 32 + TCG_REG_R1 +#endif +}; + +#ifdef CONFIG_DEBUG_TCG +static const char *const tcg_target_reg_names[TCG_TARGET_NB_REGS] = { + "r00", + "r01", + "r02", + "r03", + "r04", + "r05", + "r06", + "r07", + "r08", + "r09", + "r10", + "r11", + "r12", + "r13", + "r14", + "r15", +}; +#endif + +static bool patch_reloc(tcg_insn_unit *code_ptr, int type, + intptr_t value, intptr_t addend) +{ + intptr_t diff = value - (intptr_t)(code_ptr + 1); + + tcg_debug_assert(addend == 0); + tcg_debug_assert(type == 20); + + if (diff == sextract32(diff, 0, type)) { + tcg_patch32(code_ptr, deposit32(*code_ptr, 32 - type, type, diff)); + return true; + } + return false; +} + +static void stack_bounds_check(TCGReg base, target_long offset) +{ + if (base == TCG_REG_CALL_STACK) { + tcg_debug_assert(offset >= 0); + tcg_debug_assert(offset < (TCG_STATIC_CALL_ARGS_SIZE + + TCG_STATIC_FRAME_SIZE)); + } +} + +static void tcg_out_op_l(TCGContext *s, TCGOpcode op, TCGLabel *l0) +{ + tcg_insn_unit insn = 0; + + tcg_out_reloc(s, s->code_ptr, 20, l0, 0); + insn = deposit32(insn, 0, 8, op); + tcg_out32(s, insn); +} + +static void tcg_out_op_p(TCGContext *s, TCGOpcode op, void *p0) +{ + tcg_insn_unit insn = 0; + intptr_t diff; + + /* Special case for exit_tb: map null -> 0. */ + if (p0 == NULL) { + diff = 0; + } else { + diff = p0 - (void *)(s->code_ptr + 1); + tcg_debug_assert(diff != 0); + if (diff != sextract32(diff, 0, 20)) { + tcg_raise_tb_overflow(s); + } + } + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 12, 20, diff); + tcg_out32(s, insn); +} + +static void tcg_out_op_r(TCGContext *s, TCGOpcode op, TCGReg r0) +{ + tcg_insn_unit insn = 0; + + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + tcg_out32(s, insn); +} + +static void tcg_out_op_v(TCGContext *s, TCGOpcode op) +{ + tcg_out32(s, (uint8_t)op); +} + +static void tcg_out_op_ri(TCGContext *s, TCGOpcode op, TCGReg r0, int32_t i1) +{ + tcg_insn_unit insn = 0; + + tcg_debug_assert(i1 == sextract32(i1, 0, 20)); + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 20, i1); + tcg_out32(s, insn); +} + +static void tcg_out_op_rl(TCGContext *s, TCGOpcode op, TCGReg r0, TCGLabel *l1) +{ + tcg_insn_unit insn = 0; + + tcg_out_reloc(s, s->code_ptr, 20, l1, 0); + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + tcg_out32(s, insn); +} + +static void tcg_out_op_rr(TCGContext *s, TCGOpcode op, TCGReg r0, TCGReg r1) +{ + tcg_insn_unit insn = 0; + + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 4, r1); + tcg_out32(s, insn); +} + +static void tcg_out_op_rrm(TCGContext *s, TCGOpcode op, + TCGReg r0, TCGReg r1, TCGArg m2) +{ + tcg_insn_unit insn = 0; + + tcg_debug_assert(m2 == extract32(m2, 0, 12)); + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 4, r1); + insn = deposit32(insn, 20, 12, m2); + tcg_out32(s, insn); +} + +static void tcg_out_op_rrr(TCGContext *s, TCGOpcode op, + TCGReg r0, TCGReg r1, TCGReg r2) +{ + tcg_insn_unit insn = 0; + + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 4, r1); + insn = deposit32(insn, 16, 4, r2); + tcg_out32(s, insn); +} + +static void tcg_out_op_rrs(TCGContext *s, TCGOpcode op, + TCGReg r0, TCGReg r1, intptr_t i2) +{ + tcg_insn_unit insn = 0; + + tcg_debug_assert(i2 == sextract32(i2, 0, 16)); + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 4, r1); + insn = deposit32(insn, 16, 16, i2); + tcg_out32(s, insn); +} + +static void tcg_out_op_rrbb(TCGContext *s, TCGOpcode op, TCGReg r0, + TCGReg r1, uint8_t b2, uint8_t b3) +{ + tcg_insn_unit insn = 0; + + tcg_debug_assert(b2 == extract32(b2, 0, 6)); + tcg_debug_assert(b3 == extract32(b3, 0, 6)); + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 4, r1); + insn = deposit32(insn, 16, 6, b2); + insn = deposit32(insn, 22, 6, b3); + tcg_out32(s, insn); +} + +static void tcg_out_op_rrrc(TCGContext *s, TCGOpcode op, + TCGReg r0, TCGReg r1, TCGReg r2, TCGCond c3) +{ + tcg_insn_unit insn = 0; + + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 4, r1); + insn = deposit32(insn, 16, 4, r2); + insn = deposit32(insn, 20, 4, c3); + tcg_out32(s, insn); +} + +static void tcg_out_op_rrrm(TCGContext *s, TCGOpcode op, + TCGReg r0, TCGReg r1, TCGReg r2, TCGArg m3) +{ + tcg_insn_unit insn = 0; + + tcg_debug_assert(m3 == extract32(m3, 0, 12)); + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 4, r1); + insn = deposit32(insn, 16, 4, r2); + insn = deposit32(insn, 20, 12, m3); + tcg_out32(s, insn); +} + +static void tcg_out_op_rrrbb(TCGContext *s, TCGOpcode op, TCGReg r0, + TCGReg r1, TCGReg r2, uint8_t b3, uint8_t b4) +{ + tcg_insn_unit insn = 0; + + tcg_debug_assert(b3 == extract32(b3, 0, 6)); + tcg_debug_assert(b4 == extract32(b4, 0, 6)); + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 4, r1); + insn = deposit32(insn, 16, 4, r2); + insn = deposit32(insn, 20, 6, b3); + insn = deposit32(insn, 26, 6, b4); + tcg_out32(s, insn); +} + +static void tcg_out_op_rrrrr(TCGContext *s, TCGOpcode op, TCGReg r0, + TCGReg r1, TCGReg r2, TCGReg r3, TCGReg r4) +{ + tcg_insn_unit insn = 0; + + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 4, r1); + insn = deposit32(insn, 16, 4, r2); + insn = deposit32(insn, 20, 4, r3); + insn = deposit32(insn, 24, 4, r4); + tcg_out32(s, insn); +} + +static void tcg_out_op_rrrr(TCGContext *s, TCGOpcode op, + TCGReg r0, TCGReg r1, TCGReg r2, TCGReg r3) +{ + tcg_insn_unit insn = 0; + + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 4, r1); + insn = deposit32(insn, 16, 4, r2); + insn = deposit32(insn, 20, 4, r3); + tcg_out32(s, insn); +} + +static void tcg_out_op_rrrrrc(TCGContext *s, TCGOpcode op, + TCGReg r0, TCGReg r1, TCGReg r2, + TCGReg r3, TCGReg r4, TCGCond c5) +{ + tcg_insn_unit insn = 0; + + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 4, r1); + insn = deposit32(insn, 16, 4, r2); + insn = deposit32(insn, 20, 4, r3); + insn = deposit32(insn, 24, 4, r4); + insn = deposit32(insn, 28, 4, c5); + tcg_out32(s, insn); +} + +static void tcg_out_op_rrrrrr(TCGContext *s, TCGOpcode op, + TCGReg r0, TCGReg r1, TCGReg r2, + TCGReg r3, TCGReg r4, TCGReg r5) +{ + tcg_insn_unit insn = 0; + + insn = deposit32(insn, 0, 8, op); + insn = deposit32(insn, 8, 4, r0); + insn = deposit32(insn, 12, 4, r1); + insn = deposit32(insn, 16, 4, r2); + insn = deposit32(insn, 20, 4, r3); + insn = deposit32(insn, 24, 4, r4); + insn = deposit32(insn, 28, 4, r5); + tcg_out32(s, insn); +} + +static void tcg_out_ldst(TCGContext *s, TCGOpcode op, TCGReg val, + TCGReg base, intptr_t offset) +{ + stack_bounds_check(base, offset); + if (offset != sextract32(offset, 0, 16)) { + tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_TMP, offset); + tcg_out_op_rrr(s, (TCG_TARGET_REG_BITS == 32 + ? INDEX_op_add_i32 : INDEX_op_add_i64), + TCG_REG_TMP, TCG_REG_TMP, base); + base = TCG_REG_TMP; + offset = 0; + } + tcg_out_op_rrs(s, op, val, base, offset); +} + +static void tcg_out_ld(TCGContext *s, TCGType type, TCGReg val, TCGReg base, + intptr_t offset) +{ + switch (type) { + case TCG_TYPE_I32: + tcg_out_ldst(s, INDEX_op_ld_i32, val, base, offset); + break; +#if TCG_TARGET_REG_BITS == 64 + case TCG_TYPE_I64: + tcg_out_ldst(s, INDEX_op_ld_i64, val, base, offset); + break; +#endif + default: + g_assert_not_reached(); + } +} + +static bool tcg_out_mov(TCGContext *s, TCGType type, TCGReg ret, TCGReg arg) +{ + switch (type) { + case TCG_TYPE_I32: + tcg_out_op_rr(s, INDEX_op_mov_i32, ret, arg); + break; +#if TCG_TARGET_REG_BITS == 64 + case TCG_TYPE_I64: + tcg_out_op_rr(s, INDEX_op_mov_i64, ret, arg); + break; +#endif + default: + g_assert_not_reached(); + } + return true; +} + +static void tcg_out_movi(TCGContext *s, TCGType type, + TCGReg ret, tcg_target_long arg) +{ + switch (type) { + case TCG_TYPE_I32: +#if TCG_TARGET_REG_BITS == 64 + arg = (int32_t)arg; + /* fall through */ + case TCG_TYPE_I64: +#endif + break; + default: + g_assert_not_reached(); + } + + if (arg == sextract32(arg, 0, 20)) { + tcg_out_op_ri(s, INDEX_op_tci_movi, ret, arg); + } else { + tcg_insn_unit insn = 0; + + new_pool_label(s, arg, 20, s->code_ptr, 0); + insn = deposit32(insn, 0, 8, INDEX_op_tci_movl); + insn = deposit32(insn, 8, 4, ret); + tcg_out32(s, insn); + } +} + +static void tcg_out_call(TCGContext *s, const tcg_insn_unit *func, + ffi_cif *cif) +{ + tcg_insn_unit insn = 0; + uint8_t which; + + if (cif->rtype == &ffi_type_void) { + which = 0; + } else if (cif->rtype->size == 4) { + which = 1; + } else { + tcg_debug_assert(cif->rtype->size == 8); + which = 2; + } + new_pool_l2(s, 20, s->code_ptr, 0, (uintptr_t)func, (uintptr_t)cif); + insn = deposit32(insn, 0, 8, INDEX_op_call); + insn = deposit32(insn, 8, 4, which); + tcg_out32(s, insn); +} + +#if TCG_TARGET_REG_BITS == 64 +# define CASE_32_64(x) \ + case glue(glue(INDEX_op_, x), _i64): \ + case glue(glue(INDEX_op_, x), _i32): +# define CASE_64(x) \ + case glue(glue(INDEX_op_, x), _i64): +#else +# define CASE_32_64(x) \ + case glue(glue(INDEX_op_, x), _i32): +# define CASE_64(x) +#endif + +static void tcg_out_op(TCGContext *s, TCGOpcode opc, + const TCGArg args[TCG_MAX_OP_ARGS], + const int const_args[TCG_MAX_OP_ARGS]) +{ + TCGOpcode exts; + + switch (opc) { + case INDEX_op_exit_tb: + tcg_out_op_p(s, opc, (void *)args[0]); + break; + + case INDEX_op_goto_tb: + tcg_debug_assert(s->tb_jmp_insn_offset == 0); + /* indirect jump method. */ + tcg_out_op_p(s, opc, s->tb_jmp_target_addr + args[0]); + set_jmp_reset_offset(s, args[0]); + break; + + case INDEX_op_goto_ptr: + tcg_out_op_r(s, opc, args[0]); + break; + + case INDEX_op_br: + tcg_out_op_l(s, opc, arg_label(args[0])); + break; + + CASE_32_64(setcond) + tcg_out_op_rrrc(s, opc, args[0], args[1], args[2], args[3]); + break; + + CASE_32_64(movcond) + case INDEX_op_setcond2_i32: + tcg_out_op_rrrrrc(s, opc, args[0], args[1], args[2], + args[3], args[4], args[5]); + break; + + CASE_32_64(ld8u) + CASE_32_64(ld8s) + CASE_32_64(ld16u) + CASE_32_64(ld16s) + case INDEX_op_ld_i32: + CASE_64(ld32u) + CASE_64(ld32s) + CASE_64(ld) + CASE_32_64(st8) + CASE_32_64(st16) + case INDEX_op_st_i32: + CASE_64(st32) + CASE_64(st) + tcg_out_ldst(s, opc, args[0], args[1], args[2]); + break; + + CASE_32_64(add) + CASE_32_64(sub) + CASE_32_64(mul) + CASE_32_64(and) + CASE_32_64(or) + CASE_32_64(xor) + CASE_32_64(andc) /* Optional (TCG_TARGET_HAS_andc_*). */ + CASE_32_64(orc) /* Optional (TCG_TARGET_HAS_orc_*). */ + CASE_32_64(eqv) /* Optional (TCG_TARGET_HAS_eqv_*). */ + CASE_32_64(nand) /* Optional (TCG_TARGET_HAS_nand_*). */ + CASE_32_64(nor) /* Optional (TCG_TARGET_HAS_nor_*). */ + CASE_32_64(shl) + CASE_32_64(shr) + CASE_32_64(sar) + CASE_32_64(rotl) /* Optional (TCG_TARGET_HAS_rot_*). */ + CASE_32_64(rotr) /* Optional (TCG_TARGET_HAS_rot_*). */ + CASE_32_64(div) /* Optional (TCG_TARGET_HAS_div_*). */ + CASE_32_64(divu) /* Optional (TCG_TARGET_HAS_div_*). */ + CASE_32_64(rem) /* Optional (TCG_TARGET_HAS_div_*). */ + CASE_32_64(remu) /* Optional (TCG_TARGET_HAS_div_*). */ + CASE_32_64(clz) /* Optional (TCG_TARGET_HAS_clz_*). */ + CASE_32_64(ctz) /* Optional (TCG_TARGET_HAS_ctz_*). */ + tcg_out_op_rrr(s, opc, args[0], args[1], args[2]); + break; + + CASE_32_64(deposit) /* Optional (TCG_TARGET_HAS_deposit_*). */ + { + TCGArg pos = args[3], len = args[4]; + TCGArg max = opc == INDEX_op_deposit_i32 ? 32 : 64; + + tcg_debug_assert(pos < max); + tcg_debug_assert(pos + len <= max); + + tcg_out_op_rrrbb(s, opc, args[0], args[1], args[2], pos, len); + } + break; + + CASE_32_64(extract) /* Optional (TCG_TARGET_HAS_extract_*). */ + CASE_32_64(sextract) /* Optional (TCG_TARGET_HAS_sextract_*). */ + { + TCGArg pos = args[2], len = args[3]; + TCGArg max = tcg_op_defs[opc].flags & TCG_OPF_64BIT ? 64 : 32; + + tcg_debug_assert(pos < max); + tcg_debug_assert(pos + len <= max); + + tcg_out_op_rrbb(s, opc, args[0], args[1], pos, len); + } + break; + + CASE_32_64(brcond) + tcg_out_op_rrrc(s, (opc == INDEX_op_brcond_i32 + ? INDEX_op_setcond_i32 : INDEX_op_setcond_i64), + TCG_REG_TMP, args[0], args[1], args[2]); + tcg_out_op_rl(s, opc, TCG_REG_TMP, arg_label(args[3])); + break; + + CASE_32_64(neg) /* Optional (TCG_TARGET_HAS_neg_*). */ + CASE_32_64(not) /* Optional (TCG_TARGET_HAS_not_*). */ + CASE_32_64(ext8s) /* Optional (TCG_TARGET_HAS_ext8s_*). */ + CASE_32_64(ext8u) /* Optional (TCG_TARGET_HAS_ext8u_*). */ + CASE_32_64(ext16s) /* Optional (TCG_TARGET_HAS_ext16s_*). */ + CASE_32_64(ext16u) /* Optional (TCG_TARGET_HAS_ext16u_*). */ + CASE_64(ext32s) /* Optional (TCG_TARGET_HAS_ext32s_i64). */ + CASE_64(ext32u) /* Optional (TCG_TARGET_HAS_ext32u_i64). */ + CASE_64(ext_i32) + CASE_64(extu_i32) + CASE_32_64(ctpop) /* Optional (TCG_TARGET_HAS_ctpop_*). */ + case INDEX_op_bswap32_i32: /* Optional (TCG_TARGET_HAS_bswap32_i32). */ + case INDEX_op_bswap64_i64: /* Optional (TCG_TARGET_HAS_bswap64_i64). */ + tcg_out_op_rr(s, opc, args[0], args[1]); + break; + + case INDEX_op_bswap16_i32: /* Optional (TCG_TARGET_HAS_bswap16_i32). */ + exts = INDEX_op_ext16s_i32; + goto do_bswap; + case INDEX_op_bswap16_i64: /* Optional (TCG_TARGET_HAS_bswap16_i64). */ + exts = INDEX_op_ext16s_i64; + goto do_bswap; + case INDEX_op_bswap32_i64: /* Optional (TCG_TARGET_HAS_bswap32_i64). */ + exts = INDEX_op_ext32s_i64; + do_bswap: + /* The base tci bswaps zero-extend, and ignore high bits. */ + tcg_out_op_rr(s, opc, args[0], args[1]); + if (args[2] & TCG_BSWAP_OS) { + tcg_out_op_rr(s, exts, args[0], args[0]); + } + break; + + CASE_32_64(add2) + CASE_32_64(sub2) + tcg_out_op_rrrrrr(s, opc, args[0], args[1], args[2], + args[3], args[4], args[5]); + break; + +#if TCG_TARGET_REG_BITS == 32 + case INDEX_op_brcond2_i32: + tcg_out_op_rrrrrc(s, INDEX_op_setcond2_i32, TCG_REG_TMP, + args[0], args[1], args[2], args[3], args[4]); + tcg_out_op_rl(s, INDEX_op_brcond_i32, TCG_REG_TMP, arg_label(args[5])); + break; +#endif + + CASE_32_64(mulu2) + CASE_32_64(muls2) + tcg_out_op_rrrr(s, opc, args[0], args[1], args[2], args[3]); + break; + + case INDEX_op_qemu_ld_i32: + case INDEX_op_qemu_st_i32: + if (TARGET_LONG_BITS <= TCG_TARGET_REG_BITS) { + tcg_out_op_rrm(s, opc, args[0], args[1], args[2]); + } else { + tcg_out_op_rrrm(s, opc, args[0], args[1], args[2], args[3]); + } + break; + + case INDEX_op_qemu_ld_i64: + case INDEX_op_qemu_st_i64: + if (TCG_TARGET_REG_BITS == 64) { + tcg_out_op_rrm(s, opc, args[0], args[1], args[2]); + } else if (TARGET_LONG_BITS <= TCG_TARGET_REG_BITS) { + tcg_out_op_rrrm(s, opc, args[0], args[1], args[2], args[3]); + } else { + tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_TMP, args[4]); + tcg_out_op_rrrrr(s, opc, args[0], args[1], + args[2], args[3], TCG_REG_TMP); + } + break; + + case INDEX_op_mb: + tcg_out_op_v(s, opc); + break; + + case INDEX_op_mov_i32: /* Always emitted via tcg_out_mov. */ + case INDEX_op_mov_i64: + case INDEX_op_call: /* Always emitted via tcg_out_call. */ + default: + tcg_abort(); + } +} + +static void tcg_out_st(TCGContext *s, TCGType type, TCGReg val, TCGReg base, + intptr_t offset) +{ + stack_bounds_check(base, offset); + switch (type) { + case TCG_TYPE_I32: + tcg_out_op_rrs(s, INDEX_op_st_i32, val, base, offset); + break; +#if TCG_TARGET_REG_BITS == 64 + case TCG_TYPE_I64: + tcg_out_op_rrs(s, INDEX_op_st_i64, val, base, offset); + break; +#endif + default: + g_assert_not_reached(); + } +} + +static inline bool tcg_out_sti(TCGContext *s, TCGType type, TCGArg val, + TCGReg base, intptr_t ofs) +{ + return false; +} + +/* Test if a constant matches the constraint. */ +static bool tcg_target_const_match(int64_t val, TCGType type, int ct) +{ + return ct & TCG_CT_CONST; +} + +static void tcg_out_nop_fill(tcg_insn_unit *p, int count) +{ + memset(p, 0, sizeof(*p) * count); +} + +static void tcg_target_init(TCGContext *s) +{ +#if defined(CONFIG_DEBUG_TCG_INTERPRETER) + const char *envval = getenv("DEBUG_TCG"); + if (envval) { + qemu_set_log(strtol(envval, NULL, 0)); + } +#endif + + /* The current code uses uint8_t for tcg operations. */ + tcg_debug_assert(tcg_op_defs_max <= UINT8_MAX); + + /* Registers available for 32 bit operations. */ + tcg_target_available_regs[TCG_TYPE_I32] = BIT(TCG_TARGET_NB_REGS) - 1; + /* Registers available for 64 bit operations. */ + tcg_target_available_regs[TCG_TYPE_I64] = BIT(TCG_TARGET_NB_REGS) - 1; + /* + * The interpreter "registers" are in the local stack frame and + * cannot be clobbered by the called helper functions. However, + * the interpreter assumes a 64-bit return value and assigns to + * the return value registers. + */ + tcg_target_call_clobber_regs = + MAKE_64BIT_MASK(TCG_REG_R0, 64 / TCG_TARGET_REG_BITS); + + s->reserved_regs = 0; + tcg_regset_set_reg(s->reserved_regs, TCG_REG_TMP); + tcg_regset_set_reg(s->reserved_regs, TCG_REG_CALL_STACK); + + /* The call arguments come first, followed by the temp storage. */ + tcg_set_frame(s, TCG_REG_CALL_STACK, TCG_STATIC_CALL_ARGS_SIZE, + TCG_STATIC_FRAME_SIZE); +} + +/* Generate global QEMU prologue and epilogue code. */ +static inline void tcg_target_qemu_prologue(TCGContext *s) +{ +} diff --git a/tcg/tci/tcg-target.h b/tcg/tci/tcg-target.h new file mode 100644 index 000000000..033e613f2 --- /dev/null +++ b/tcg/tci/tcg-target.h @@ -0,0 +1,180 @@ +/* + * Tiny Code Generator for QEMU + * + * Copyright (c) 2009, 2011 Stefan Weil + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +/* + * This code implements a TCG which does not generate machine code for some + * real target machine but which generates virtual machine code for an + * interpreter. Interpreted pseudo code is slow, but it works on any host. + * + * Some remarks might help in understanding the code: + * + * "target" or "TCG target" is the machine which runs the generated code. + * This is different to the usual meaning in QEMU where "target" is the + * emulated machine. So normally QEMU host is identical to TCG target. + * Here the TCG target is a virtual machine, but this virtual machine must + * use the same word size like the real machine. + * Therefore, we need both 32 and 64 bit virtual machines (interpreter). + */ + +#ifndef TCG_TARGET_H +#define TCG_TARGET_H + +#define TCG_TARGET_INTERPRETER 1 +#define TCG_TARGET_INSN_UNIT_SIZE 4 +#define TCG_TARGET_TLB_DISPLACEMENT_BITS 32 +#define MAX_CODE_GEN_BUFFER_SIZE ((size_t)-1) + +#if UINTPTR_MAX == UINT32_MAX +# define TCG_TARGET_REG_BITS 32 +#elif UINTPTR_MAX == UINT64_MAX +# define TCG_TARGET_REG_BITS 64 +#else +# error Unknown pointer size for tci target +#endif + +#ifdef CONFIG_DEBUG_TCG +/* Enable debug output. */ +#define CONFIG_DEBUG_TCG_INTERPRETER +#endif + +/* Optional instructions. */ + +#define TCG_TARGET_HAS_bswap16_i32 1 +#define TCG_TARGET_HAS_bswap32_i32 1 +#define TCG_TARGET_HAS_div_i32 1 +#define TCG_TARGET_HAS_rem_i32 1 +#define TCG_TARGET_HAS_ext8s_i32 1 +#define TCG_TARGET_HAS_ext16s_i32 1 +#define TCG_TARGET_HAS_ext8u_i32 1 +#define TCG_TARGET_HAS_ext16u_i32 1 +#define TCG_TARGET_HAS_andc_i32 1 +#define TCG_TARGET_HAS_deposit_i32 1 +#define TCG_TARGET_HAS_extract_i32 1 +#define TCG_TARGET_HAS_sextract_i32 1 +#define TCG_TARGET_HAS_extract2_i32 0 +#define TCG_TARGET_HAS_eqv_i32 1 +#define TCG_TARGET_HAS_nand_i32 1 +#define TCG_TARGET_HAS_nor_i32 1 +#define TCG_TARGET_HAS_clz_i32 1 +#define TCG_TARGET_HAS_ctz_i32 1 +#define TCG_TARGET_HAS_ctpop_i32 1 +#define TCG_TARGET_HAS_neg_i32 1 +#define TCG_TARGET_HAS_not_i32 1 +#define TCG_TARGET_HAS_orc_i32 1 +#define TCG_TARGET_HAS_rot_i32 1 +#define TCG_TARGET_HAS_movcond_i32 1 +#define TCG_TARGET_HAS_muls2_i32 1 +#define TCG_TARGET_HAS_muluh_i32 0 +#define TCG_TARGET_HAS_mulsh_i32 0 +#define TCG_TARGET_HAS_direct_jump 0 +#define TCG_TARGET_HAS_qemu_st8_i32 0 + +#if TCG_TARGET_REG_BITS == 64 +#define TCG_TARGET_HAS_extrl_i64_i32 0 +#define TCG_TARGET_HAS_extrh_i64_i32 0 +#define TCG_TARGET_HAS_bswap16_i64 1 +#define TCG_TARGET_HAS_bswap32_i64 1 +#define TCG_TARGET_HAS_bswap64_i64 1 +#define TCG_TARGET_HAS_deposit_i64 1 +#define TCG_TARGET_HAS_extract_i64 1 +#define TCG_TARGET_HAS_sextract_i64 1 +#define TCG_TARGET_HAS_extract2_i64 0 +#define TCG_TARGET_HAS_div_i64 1 +#define TCG_TARGET_HAS_rem_i64 1 +#define TCG_TARGET_HAS_ext8s_i64 1 +#define TCG_TARGET_HAS_ext16s_i64 1 +#define TCG_TARGET_HAS_ext32s_i64 1 +#define TCG_TARGET_HAS_ext8u_i64 1 +#define TCG_TARGET_HAS_ext16u_i64 1 +#define TCG_TARGET_HAS_ext32u_i64 1 +#define TCG_TARGET_HAS_andc_i64 1 +#define TCG_TARGET_HAS_eqv_i64 1 +#define TCG_TARGET_HAS_nand_i64 1 +#define TCG_TARGET_HAS_nor_i64 1 +#define TCG_TARGET_HAS_clz_i64 1 +#define TCG_TARGET_HAS_ctz_i64 1 +#define TCG_TARGET_HAS_ctpop_i64 1 +#define TCG_TARGET_HAS_neg_i64 1 +#define TCG_TARGET_HAS_not_i64 1 +#define TCG_TARGET_HAS_orc_i64 1 +#define TCG_TARGET_HAS_rot_i64 1 +#define TCG_TARGET_HAS_movcond_i64 1 +#define TCG_TARGET_HAS_muls2_i64 1 +#define TCG_TARGET_HAS_add2_i32 1 +#define TCG_TARGET_HAS_sub2_i32 1 +#define TCG_TARGET_HAS_mulu2_i32 1 +#define TCG_TARGET_HAS_add2_i64 1 +#define TCG_TARGET_HAS_sub2_i64 1 +#define TCG_TARGET_HAS_mulu2_i64 1 +#define TCG_TARGET_HAS_muluh_i64 0 +#define TCG_TARGET_HAS_mulsh_i64 0 +#else +#define TCG_TARGET_HAS_mulu2_i32 1 +#endif /* TCG_TARGET_REG_BITS == 64 */ + +/* Number of registers available. */ +#define TCG_TARGET_NB_REGS 16 + +/* List of registers which are used by TCG. */ +typedef enum { + TCG_REG_R0 = 0, + TCG_REG_R1, + TCG_REG_R2, + TCG_REG_R3, + TCG_REG_R4, + TCG_REG_R5, + TCG_REG_R6, + TCG_REG_R7, + TCG_REG_R8, + TCG_REG_R9, + TCG_REG_R10, + TCG_REG_R11, + TCG_REG_R12, + TCG_REG_R13, + TCG_REG_R14, + TCG_REG_R15, + + TCG_REG_TMP = TCG_REG_R13, + TCG_AREG0 = TCG_REG_R14, + TCG_REG_CALL_STACK = TCG_REG_R15, +} TCGReg; + +/* Used for function call generation. */ +#define TCG_TARGET_CALL_STACK_OFFSET 0 +#define TCG_TARGET_STACK_ALIGN 8 + +#define HAVE_TCG_QEMU_TB_EXEC +#define TCG_TARGET_NEED_POOL_LABELS + +/* We could notice __i386__ or __s390x__ and reduce the barriers depending + on the host. But if you want performance, you use the normal backend. + We prefer consistency across hosts on this. */ +#define TCG_TARGET_DEFAULT_MO (0) + +#define TCG_TARGET_HAS_MEMORY_BSWAP 1 + +/* not defined -- call should be eliminated at compile time */ +void tb_target_set_jmp_target(uintptr_t, uintptr_t, uintptr_t, uintptr_t); + +#endif /* TCG_TARGET_H */ |