diff options
Diffstat (limited to 'gas/config/tc-bpf.c')
| -rw-r--r-- | gas/config/tc-bpf.c | 2519 |
1 files changed, 904 insertions, 1615 deletions
diff --git a/gas/config/tc-bpf.c b/gas/config/tc-bpf.c index 3b86f9c..7a54fac 100644 --- a/gas/config/tc-bpf.c +++ b/gas/config/tc-bpf.c @@ -22,42 +22,42 @@ #include "as.h" #include "subsegs.h" #include "symcat.h" -#include "opcodes/bpf-desc.h" -#include "opcodes/bpf-opc.h" -#include "cgen.h" +#include "opcode/bpf.h" #include "elf/common.h" #include "elf/bpf.h" #include "dwarf2dbg.h" +#include "libiberty.h" #include <ctype.h> +/* Data structure representing a parsed BPF instruction. */ + +struct bpf_insn +{ + int size; /* Instruction size in bytes. */ + bpf_insn_word opcode; + uint8_t dst; + uint8_t src; + expressionS offset16; + expressionS imm32; + expressionS imm64; + expressionS disp16; + expressionS disp32; + + unsigned int has_dst : 1; + unsigned int has_src : 1; + unsigned int has_offset16 : 1; + unsigned int has_disp16 : 1; + unsigned int has_disp32 : 1; + unsigned int has_imm32 : 1; + unsigned int has_imm64 : 1; +}; + const char comment_chars[] = ";"; const char line_comment_chars[] = "#"; const char line_separator_chars[] = "`"; const char EXP_CHARS[] = "eE"; const char FLT_CHARS[] = "fFdD"; -static const char *invalid_expression; -static char pseudoc_lex[256]; -static const char symbol_chars[] = -"_ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789"; - -static const char arithm_op[] = "+-/<>%&|^"; - -static void init_pseudoc_lex (void); - -#define LEX_IS_SYMBOL_COMPONENT 1 -#define LEX_IS_WHITESPACE 2 -#define LEX_IS_NEWLINE 3 -#define LEX_IS_ARITHM_OP 4 -#define LEX_IS_STAR 6 -#define LEX_IS_CLSE_BR 7 -#define LEX_IS_OPEN_BR 8 -#define LEX_IS_EQUAL 9 -#define LEX_IS_EXCLA 10 - -#define ST_EOI 100 -#define MAX_TOKEN_SZ 100 - /* Like s_lcomm_internal in gas/read.c but the alignment string is allowed to be optional. */ @@ -110,18 +110,15 @@ const pseudo_typeS md_pseudo_table[] = -/* ISA handling. */ -static CGEN_BITSET *bpf_isa; - - - /* Command-line options processing. */ enum options { OPTION_LITTLE_ENDIAN = OPTION_MD_BASE, OPTION_BIG_ENDIAN, - OPTION_XBPF + OPTION_XBPF, + OPTION_DIALECT, + OPTION_ISA_SPEC, }; struct option md_longopts[] = @@ -129,6 +126,8 @@ struct option md_longopts[] = { "EL", no_argument, NULL, OPTION_LITTLE_ENDIAN }, { "EB", no_argument, NULL, OPTION_BIG_ENDIAN }, { "mxbpf", no_argument, NULL, OPTION_XBPF }, + { "mdialect", required_argument, NULL, OPTION_DIALECT}, + { "misa-spec", required_argument, NULL, OPTION_ISA_SPEC}, { NULL, no_argument, NULL, 0 }, }; @@ -136,18 +135,34 @@ size_t md_longopts_size = sizeof (md_longopts); const char * md_shortopts = ""; -extern int target_big_endian; +/* BPF supports little-endian and big-endian variants. The following + global records what endianness to use. It can be configured using + command-line options. It defaults to the host endianness + initialized in md_begin. */ -/* Whether target_big_endian has been set while parsing command-line - arguments. */ static int set_target_endian = 0; +extern int target_big_endian; + +/* The ISA specification can be one of BPF_V1, BPF_V2, BPF_V3, BPF_V4 + or BPF_XPBF. The ISA spec to use can be configured using + command-line options. It defaults to the latest BPF spec. */ + +static int isa_spec = BPF_V4; -static int target_xbpf = 0; +/* The assembler supports two different dialects: "normal" syntax and + "pseudoc" syntax. The dialect to use can be configured using + command-line options. */ -static int set_xbpf = 0; +enum target_asm_dialect +{ + DIALECT_NORMAL, + DIALECT_PSEUDOC +}; + +static int asm_dialect = DIALECT_NORMAL; int -md_parse_option (int c, const char * arg ATTRIBUTE_UNUSED) +md_parse_option (int c, const char * arg) { switch (c) { @@ -156,12 +171,36 @@ md_parse_option (int c, const char * arg ATTRIBUTE_UNUSED) target_big_endian = 1; break; case OPTION_LITTLE_ENDIAN: - set_target_endian = 1; + set_target_endian = 0; target_big_endian = 0; break; + case OPTION_DIALECT: + if (strcmp (arg, "normal") == 0) + asm_dialect = DIALECT_NORMAL; + else if (strcmp (arg, "pseudoc") == 0) + asm_dialect = DIALECT_PSEUDOC; + else + as_fatal (_("-mdialect=%s is not valid. Expected normal or pseudoc"), + arg); + break; + case OPTION_ISA_SPEC: + if (strcmp (arg, "v1") == 0) + isa_spec = BPF_V1; + else if (strcmp (arg, "v2") == 0) + isa_spec = BPF_V2; + else if (strcmp (arg, "v3") == 0) + isa_spec = BPF_V3; + else if (strcmp (arg, "v4") == 0) + isa_spec = BPF_V4; + else if (strcmp (arg, "xbpf") == 0) + isa_spec = BPF_XBPF; + else + as_fatal (_("-misa-spec=%s is not valid. Expected v1, v2, v3, v4 o xbpf"), + arg); + break; case OPTION_XBPF: - set_xbpf = 1; - target_xbpf = 1; + /* This is an alias for -misa-spec=xbpf. */ + isa_spec = BPF_XBPF; break; default: return 0; @@ -175,43 +214,22 @@ md_show_usage (FILE * stream) { fprintf (stream, _("\nBPF options:\n")); fprintf (stream, _("\ - --EL generate code for a little endian machine\n\ - --EB generate code for a big endian machine\n\ - -mxbpf generate xBPF instructions\n")); +BPF options:\n\ + -EL generate code for a little endian machine\n\ + -EB generate code for a big endian machine\n\ + -mdialect=DIALECT set the assembly dialect (normal, pseudoc)\n\ + -misa-spec set the BPF ISA spec (v1, v2, v3, v4, xbpf)\n\ + -mxbpf alias for -misa-spec=xbpf\n")); } - -static void -init_pseudoc_lex (void) -{ - const char *p; - - for (p = symbol_chars; *p; ++p) - pseudoc_lex[(unsigned char) *p] = LEX_IS_SYMBOL_COMPONENT; - - pseudoc_lex[' '] = LEX_IS_WHITESPACE; - pseudoc_lex['\t'] = LEX_IS_WHITESPACE; - pseudoc_lex['\r'] = LEX_IS_WHITESPACE; - pseudoc_lex['\n'] = LEX_IS_NEWLINE; - pseudoc_lex['*'] = LEX_IS_STAR; - pseudoc_lex[')'] = LEX_IS_CLSE_BR; - pseudoc_lex['('] = LEX_IS_OPEN_BR; - pseudoc_lex[']'] = LEX_IS_CLSE_BR; - pseudoc_lex['['] = LEX_IS_OPEN_BR; - - for (p = arithm_op; *p; ++p) - pseudoc_lex[(unsigned char) *p] = LEX_IS_ARITHM_OP; - - pseudoc_lex['='] = LEX_IS_EQUAL; - pseudoc_lex['!'] = LEX_IS_EXCLA; -} +/* This function is called once, at assembler startup time. This + should set up all the tables, etc that the MD part of the assembler + needs. */ void md_begin (void) { - /* Initialize the `cgen' interface. */ - /* If not specified in the command line, use the host endianness. */ if (!set_target_endian) @@ -223,50 +241,15 @@ md_begin (void) #endif } - /* If not specified in the command line, use eBPF rather - than xBPF. */ - if (!set_xbpf) - target_xbpf = 0; - - /* Set the ISA, which depends on the target endianness. */ - bpf_isa = cgen_bitset_create (ISA_MAX); - if (target_big_endian) - { - if (target_xbpf) - cgen_bitset_set (bpf_isa, ISA_XBPFBE); - else - cgen_bitset_set (bpf_isa, ISA_EBPFBE); - } - else - { - if (target_xbpf) - cgen_bitset_set (bpf_isa, ISA_XBPFLE); - else - cgen_bitset_set (bpf_isa, ISA_EBPFLE); - } - /* Ensure that lines can begin with '*' in BPF store pseudoc instruction. */ lex_type['*'] |= LEX_BEGIN_NAME; - /* Set the machine number and endian. */ - gas_cgen_cpu_desc = bpf_cgen_cpu_open (CGEN_CPU_OPEN_ENDIAN, - target_big_endian ? - CGEN_ENDIAN_BIG : CGEN_ENDIAN_LITTLE, - CGEN_CPU_OPEN_INSN_ENDIAN, - CGEN_ENDIAN_LITTLE, - CGEN_CPU_OPEN_ISAS, - bpf_isa, - CGEN_CPU_OPEN_END); - bpf_cgen_init_asm (gas_cgen_cpu_desc); - - /* This is a callback from cgen to gas to parse operands. */ - cgen_set_parse_operand_fn (gas_cgen_cpu_desc, gas_cgen_parse_operand); - /* Set the machine type. */ bfd_default_set_arch_mach (stdoutput, bfd_arch_bpf, bfd_mach_bpf); - init_pseudoc_lex(); } +/* Round up a section size to the appropriate boundary. */ + valueT md_section_align (segT segment, valueT size) { @@ -310,33 +293,48 @@ md_number_to_chars (char * buf, valueT val, int n) } arelent * -tc_gen_reloc (asection *sec, fixS *fix) +tc_gen_reloc (asection *sec ATTRIBUTE_UNUSED, fixS *fixP) { - return gas_cgen_tc_gen_reloc (sec, fix); -} + bfd_reloc_code_real_type r_type = fixP->fx_r_type; + arelent *reloc; -/* Return the bfd reloc type for OPERAND of INSN at fixup FIXP. This - is called when the operand is an expression that couldn't be fully - resolved. Returns BFD_RELOC_NONE if no reloc type can be found. - *FIXP may be modified if desired. */ + reloc = XNEW (arelent); -bfd_reloc_code_real_type -md_cgen_lookup_reloc (const CGEN_INSN *insn ATTRIBUTE_UNUSED, - const CGEN_OPERAND *operand, - fixS *fixP) -{ - switch (operand->type) + if (fixP->fx_pcrel) + { + r_type = (r_type == BFD_RELOC_8 ? BFD_RELOC_8_PCREL + : r_type == BFD_RELOC_16 ? BFD_RELOC_16_PCREL + : r_type == BFD_RELOC_24 ? BFD_RELOC_24_PCREL + : r_type == BFD_RELOC_32 ? BFD_RELOC_32_PCREL + : r_type == BFD_RELOC_64 ? BFD_RELOC_64_PCREL + : r_type); + } + + reloc->howto = bfd_reloc_type_lookup (stdoutput, r_type); + + if (reloc->howto == (reloc_howto_type *) NULL) { - case BPF_OPERAND_IMM64: - return BFD_RELOC_BPF_64; - case BPF_OPERAND_DISP32: - fixP->fx_pcrel = 1; - return BFD_RELOC_BPF_DISP32; - default: - break; + as_bad_where (fixP->fx_file, fixP->fx_line, + _("relocation is not supported")); + return NULL; } - return BFD_RELOC_NONE; + + //XXX gas_assert (!fixP->fx_pcrel == !reloc->howto->pc_relative); + + reloc->sym_ptr_ptr = XNEW (asymbol *); + *reloc->sym_ptr_ptr = symbol_get_bfdsym (fixP->fx_addsy); + + /* Use fx_offset for these cases. */ + if (fixP->fx_r_type == BFD_RELOC_VTABLE_ENTRY + || fixP->fx_r_type == BFD_RELOC_VTABLE_INHERIT) + reloc->addend = fixP->fx_offset; + else + reloc->addend = fixP->fx_addnumber; + + reloc->address = fixP->fx_frag->fr_address + fixP->fx_where; + return reloc; } + /* *FRAGP has been relaxed to its final size, and now needs to have the bytes inside it modified to conform to the new size. @@ -362,1556 +360,821 @@ md_estimate_size_before_relax (fragS *fragP ATTRIBUTE_UNUSED, } +/* Apply a fixS (fixup of an instruction or data that we didn't have + enough info to complete immediately) to the data in a frag. */ + void -md_apply_fix (fixS *fixP, valueT *valP, segT seg) +md_apply_fix (fixS *fixP, valueT *valP, segT seg ATTRIBUTE_UNUSED) { - /* Some fixups for instructions require special attention. This is - handled in the code block below. */ - if ((int) fixP->fx_r_type >= (int) BFD_RELOC_UNUSED) + char *where = fixP->fx_frag->fr_literal + fixP->fx_where; + + switch (fixP->fx_r_type) { - int opindex = (int) fixP->fx_r_type - (int) BFD_RELOC_UNUSED; - const CGEN_OPERAND *operand = cgen_operand_lookup_by_num (gas_cgen_cpu_desc, - opindex); - char *where; + case BFD_RELOC_BPF_DISP16: + /* Convert from bytes to number of 64-bit words to the target, + minus one. */ + *valP = (((long) (*valP)) - 8) / 8; + break; + case BFD_RELOC_BPF_DISP32: + /* eBPF supports two kind of CALL instructions: the so called + pseudo calls ("bpf to bpf") and external calls ("bpf to + kernel"). + + Both kind of calls use the same instruction (CALL). However, + external calls are constructed by passing a constant argument + to the instruction, whereas pseudo calls result from + expressions involving symbols. In practice, instructions + requiring a fixup are interpreted as pseudo-calls. If we are + executing this code, this is a pseudo call. + + The kernel expects for pseudo-calls to be annotated by having + BPF_PSEUDO_CALL in the SRC field of the instruction. But + beware the infamous nibble-swapping of eBPF and take + endianness into account here. + + Note that the CALL instruction has only one operand, so + this code is executed only once per instruction. */ + md_number_to_chars (where + 1, target_big_endian ? 0x01 : 0x10, 1); + + /* Convert from bytes to number of 64-bit words to the target, + minus one. */ + *valP = (((long) (*valP)) - 8) / 8; + break; + case BFD_RELOC_16_PCREL: + /* Convert from bytes to number of 64-bit words to the target, + minus one. */ + *valP = (((long) (*valP)) - 8) / 8; + break; + default: + break; + } - switch (operand->type) - { - case BPF_OPERAND_DISP32: - /* eBPF supports two kind of CALL instructions: the so - called pseudo calls ("bpf to bpf") and external calls - ("bpf to kernel"). - - Both kind of calls use the same instruction (CALL). - However, external calls are constructed by passing a - constant argument to the instruction, whereas pseudo - calls result from expressions involving symbols. In - practice, instructions requiring a fixup are interpreted - as pseudo-calls. If we are executing this code, this is - a pseudo call. - - The kernel expects for pseudo-calls to be annotated by - having BPF_PSEUDO_CALL in the SRC field of the - instruction. But beware the infamous nibble-swapping of - eBPF and take endianness into account here. - - Note that the CALL instruction has only one operand, so - this code is executed only once per instruction. */ - where = fixP->fx_frag->fr_literal + fixP->fx_where + 1; - where[0] = target_big_endian ? 0x01 : 0x10; - /* Fallthrough. */ - case BPF_OPERAND_DISP16: - /* The PC-relative displacement fields in jump instructions - shouldn't be in bytes. Instead, they hold the number of - 64-bit words to the target, _minus one_. */ - *valP = (((long) (*valP)) - 8) / 8; + if (fixP->fx_addsy == (symbolS *) NULL) + fixP->fx_done = 1; + + if (fixP->fx_done) + { + /* We're finished with this fixup. Install it because + bfd_install_relocation won't be called to do it. */ + switch (fixP->fx_r_type) + { + case BFD_RELOC_8: + md_number_to_chars (where, *valP, 1); + break; + case BFD_RELOC_16: + md_number_to_chars (where, *valP, 2); + break; + case BFD_RELOC_32: + md_number_to_chars (where, *valP, 4); + break; + case BFD_RELOC_64: + md_number_to_chars (where, *valP, 8); + break; + case BFD_RELOC_BPF_DISP16: + md_number_to_chars (where + 2, (uint16_t) *valP, 2); break; - default: + case BFD_RELOC_BPF_DISP32: + md_number_to_chars (where + 4, (uint32_t) *valP, 4); break; - } + case BFD_RELOC_16_PCREL: + md_number_to_chars (where + 2, (uint32_t) *valP, 2); + break; + default: + as_bad_where (fixP->fx_file, fixP->fx_line, + _("internal error: can't install fix for reloc type %d (`%s')"), + fixP->fx_r_type, bfd_get_reloc_code_name (fixP->fx_r_type)); + break; + } } - /* And now invoke CGEN's handler, which will eventually install - *valP into the corresponding operand. */ - gas_cgen_md_apply_fix (fixP, valP, seg); + /* Tuck `value' away for use by tc_gen_reloc. + See the comment describing fx_addnumber in write.h. + This field is misnamed (or misused :-). */ + fixP->fx_addnumber = *valP; } -/* - The BPF pseudo grammar: - - instruction : bpf_alu_insn - | bpf_alu32_insn - | bpf_jump_insn - | bpf_load_store_insn - | bpf_load_store32_insn - | bpf_non_generic_load - | bpf_endianness_conv_insn - | bpf_64_imm_load_insn - | bpf_atomic_insn - ; - - bpf_alu_insn : BPF_REG bpf_alu_operator register_or_imm32 - ; - - bpf_alu32_insn : BPF_REG32 bpf_alu_operator register32_or_imm32 - ; - - bpf_jump_insn : BPF_JA offset - | IF BPF_REG bpf_jump_operator register_or_imm32 BPF_JA offset - | IF BPF_REG32 bpf_jump_operator register_or_imm32 BPF_JA offset - | BPF_CALL offset - | BPF_EXIT - ; - - bpf_load_store_insn : BPF_REG CHR_EQUAL bpf_size_cast BPF_CHR_OPEN_BR \ - register_and_offset BPF_CHR_CLSE_BR - | bpf_size_cast register_and_offset CHR_EQUAL BPF_REG - ; - - bpf_load_store32_insn : BPF_REG CHR_EQUAL bpf_size_cast BPF_CHR_OPEN_BR \ - register32_and_offset BPF_CHR_CLSE_BR - | bpf_size_cast register_and_offset CHR_EQUAL BPF_REG32 - ; - - bpf_non_generic_load : BPF_REG_R0 CHR_EQUAL bpf_size_cast BPF_LD BPF_CHR_OPEN_BR \ - imm32 BPF_CHR_CLSE_BR - ; - - bpf_endianness_conv_insn : BPF_REG_N bpf_endianness_mnem BPF_REG_N - ; - - bpf_64_imm_load_insn : BPF_REG imm64 BPF_LL - ; - - bpf_atomic_insn : BPF_LOCK bpf_size_cast_32_64 register_and_offset BPF_ADD BPF_REG - - register_and_offset : BPF_CHR_OPEN_BR BPF_REG offset BPF_CHR_CLSE_BR - ; - - register32_and_offset : BPF_CHR_OPEN_BR BPF_REG32 offset BPF_CHR_CLSE_BR - ; - - bpf_size_cast : CHR_START BPF_CHR_OPEN_BR bpf_size CHR_START BPF_CHR_CLSE_BR - ; - - bpf_size_cast_32_64 : CHR_START BPF_CHR_OPEN_BR bpf_size_cast_32_64 CHR_STAR BPF_CHR_CLSE_BR - ; - - bpf_size_32_64 : BPF_CAST_U32 - | BPF_CAST_U64 - ; - - bpf_size : BPF_CAST_U8 - | BPF_CAST_U16 - | BPF_CAST_U32 - | BPF_CAST_U64 - ; - - bpf_jump_operator : BPF_JEQ - | BPF_JGT - | BPF_JGE - | BPF_JNE - | BPF_JSGT - | BPF_JSGE - | BPF_JLT - | BPF_JLE - | BPF_JSLT - | BPF_JSLE - ; - - bpf_alu_operator : BPF_ADD - | BPF_SUB - | BPF_MUL - | BPF_DIV - | BPF_OR - | BPF_AND - | BPF_LSH - | BPF_RSH - | BPF_NEG - | BPF_MOD - | BPF_XOR - | BPF_ARSH - | CHR_EQUAL - ; - - bpf_endianness_mnem : BPF_LE16 - | BPF_LE32 - | BPF_LE64 - | BPF_BE16 - | BPF_BE32 - | BPF_BE64 - ; - - offset : BPF_EXPR - | BPF_SYMBOL - ; - - register_or_imm32 : BPF_REG - | expression - ; - - register32_or_imm32 : BPF_REG32 - | expression - ; - - imm32 : BPF_EXPR - | BPF_SYMBOL - ; - - imm64 : BPF_EXPR - | BPF_SYMBOL - ; - - register_or_expression : BPF_EXPR - | BPF_REG - ; - - BPF_EXPR : GAS_EXPR - -*/ - -enum bpf_token_type - { - /* Keep grouped to quickly access. */ - BPF_ADD, - BPF_SUB, - BPF_MUL, - BPF_DIV, - BPF_OR, - BPF_AND, - BPF_LSH, - BPF_RSH, - BPF_MOD, - BPF_XOR, - BPF_MOV, - BPF_ARSH, - BPF_NEG, - - BPF_REG, - - BPF_IF, - BPF_GOTO, - - /* Keep grouped to quickly access. */ - BPF_JEQ, - BPF_JGT, - BPF_JGE, - BPF_JLT, - BPF_JLE, - BPF_JSET, - BPF_JNE, - BPF_JSGT, - BPF_JSGE, - BPF_JSLT, - BPF_JSLE, - - BPF_SYMBOL, - BPF_CHR_CLSE_BR, - BPF_CHR_OPEN_BR, - - /* Keep grouped to quickly access. */ - BPF_CAST_U8, - BPF_CAST_U16, - BPF_CAST_U32, - BPF_CAST_U64, - - /* Keep grouped to quickly access. */ - BPF_LE16, - BPF_LE32, - BPF_LE64, - BPF_BE16, - BPF_BE32, - BPF_BE64, - - BPF_LOCK, - - BPF_IND_CALL, - BPF_LD, - BPF_LL, - BPF_EXPR, - BPF_UNKNOWN, - }; - -static int -valid_expr (const char *e, const char **end_expr) -{ - invalid_expression = NULL; - char *hold = input_line_pointer; - expressionS exp; - - input_line_pointer = (char *) e; - deferred_expression (&exp); - *end_expr = input_line_pointer; - input_line_pointer = hold; +/* Parse an operand expression. Returns the first character that is + not part of the expression, or NULL in case of parse error. - return invalid_expression == NULL; -} + See md_operand below to see how exp_parse_failed is used. */ -static char * -build_bpf_non_generic_load (char *src, enum bpf_token_type cast, - const char *imm32) -{ - char *bpf_insn; - static const char *cast_rw[] = {"b", "h", "w", "dw"}; - - bpf_insn = xasprintf ("%s%s%s %s%s%s%s", - "ld", - src ? "ind" : "abs", - cast_rw[cast - BPF_CAST_U8], - src ? "%" : "", - src ? src : "", - src ? "," : "", - imm32); - return bpf_insn; -} +static int exp_parse_failed = 0; static char * -build_bpf_atomic_insn (char *dst, char *src, - enum bpf_token_type atomic_insn, - enum bpf_token_type cast, - const char *offset) +parse_expression (char *s, expressionS *exp) { - char *bpf_insn; - static const char *cast_rw[] = {"w", "dw"}; - static const char *mnem[] = {"xadd"}; - - bpf_insn = xasprintf ("%s%s [%%%s%s%s],%%%s", mnem[atomic_insn - BPF_ADD], - cast_rw[cast - BPF_CAST_U32], dst, - *offset != '+' ? "+" : "", - offset, src); - return bpf_insn; -} + char *saved_input_line_pointer = input_line_pointer; + char *saved_s = s; -static char * -build_bpf_jmp_insn (char *dst, char *src, - char *imm32, enum bpf_token_type op, - const char *sym, const char *offset) -{ - char *bpf_insn; - static const char *mnem[] = - { - "jeq", "jgt", "jge", "jlt", - "jle", "jset", "jne", "jsgt", - "jsge", "jslt", "jsle" - }; - - const char *in32 = (*dst == 'w' ? "32" : ""); - - *dst = 'r'; - if (src) - *src = 'r'; - - bpf_insn = xasprintf ("%s%s %%%s,%s%s,%s", - mnem[op - BPF_JEQ], in32, dst, - src ? "%" : "", - src ? src : imm32, - offset ? offset : sym); - return bpf_insn; -} + exp_parse_failed = 0; + input_line_pointer = s; + expression (exp); + s = input_line_pointer; + input_line_pointer = saved_input_line_pointer; -static char * -build_bpf_arithm_insn (char *dst, char *src, - int load64, const char *imm32, - enum bpf_token_type type) -{ - char *bpf_insn; - static const char *mnem[] = - { - "add", "sub", "mul", "div", - "or", "and", "lsh", "rsh", - "mod", "xor", "mov", "arsh", - "neg", - }; - const char *in32 = (*dst == 'w' ? "32" : ""); - - *dst = 'r'; - if (src) - *src = 'r'; - - if (type == BPF_NEG) - bpf_insn = xasprintf ("%s%s %%%s", mnem[type - BPF_ADD], in32, dst); - else if (load64) - bpf_insn = xasprintf ("%s %%%s,%s", "lddw", dst, imm32); - else - bpf_insn = xasprintf ("%s%s %%%s,%s%s", mnem[type - BPF_ADD], - in32, dst, - src ? "%" : "", - src ? src: imm32); - return bpf_insn; -} + switch (exp->X_op == O_absent || exp_parse_failed) + return NULL; -static char * -build_bpf_endianness (char *dst, enum bpf_token_type endianness) -{ - char *bpf_insn; - static const char *size[] = {"16", "32", "64"}; - int be = 1; - - if (endianness == BPF_LE16 - || endianness == BPF_LE32 - || endianness == BPF_LE64) - be = 0; - else - gas_assert (endianness == BPF_BE16 || endianness == BPF_BE32 || endianness == BPF_BE64); + /* The expression parser may consume trailing whitespaces. We have + to undo that since the instruction templates may be expecting + these whitespaces. */ + { + char *p; + for (p = s - 1; p >= saved_s && *p == ' '; --p) + --s; + } - bpf_insn = xasprintf ("%s %%%s,%s", be ? "endbe" : "endle", - dst, be ? size[endianness - BPF_BE16] : size[endianness - BPF_LE16]); - return bpf_insn; + return s; } -static char * -build_bpf_load_store_insn (char *dst, char *src, - enum bpf_token_type cast, - const char *offset, int isload) -{ - char *bpf_insn; - static const char *cast_rw[] = {"b", "h", "w", "dw"}; - - *dst = *src = 'r'; - if (isload) - bpf_insn = xasprintf ("%s%s %%%s,[%%%s%s%s]", "ldx", - cast_rw[cast - BPF_CAST_U8], dst, src, - *offset != '+' ? "+" : "", - offset); - else - bpf_insn = xasprintf ("%s%s [%%%s%s%s],%%%s", "stx", - cast_rw[cast - BPF_CAST_U8], dst, - *offset != '+' ? "+" : "", - offset, src); - return bpf_insn; -} +/* Parse a BPF register name and return the corresponding register + number. Return NULL in case of parse error, or a pointer to the + first character in S that is not part of the register name. */ -static int -look_for_reserved_word (const char *token, enum bpf_token_type *type) +static char * +parse_bpf_register (char *s, char rw, uint8_t *regno) { - int i; - static struct - { - const char *name; - enum bpf_token_type type; - } reserved_words[] = + if (asm_dialect == DIALECT_NORMAL) { - { - .name = "if", - .type = BPF_IF - }, - { - .name = "goto", - .type = BPF_GOTO - }, - { - .name = "le16", - .type = BPF_LE16 - }, - { - .name = "le32", - .type = BPF_LE32 - }, - { - .name = "le64", - .type = BPF_LE64 - }, - { - .name = "be16", - .type = BPF_BE16 - }, - { - .name = "be32", - .type = BPF_BE32 - }, - { - .name = "be64", - .type = BPF_BE64 - }, - { - .name = "lock", - .type = BPF_LOCK - }, - { - .name = "callx", - .type = BPF_IND_CALL - }, - { - .name = "skb", - .type = BPF_LD - }, - { - .name = "ll", - .type = BPF_LL - }, - { - .name = NULL, - } - }; + rw = 'r'; + if (*s != '%') + return NULL; + s += 1; - for (i = 0; reserved_words[i].name; ++i) - if (*reserved_words[i].name == *token - && !strcmp (reserved_words[i].name, token)) - { - *type = reserved_words[i].type; - return 1; - } + if (*s == 'f' && *(s + 1) == 'p') + { + *regno = 10; + s += 2; + return s; + } + } - return 0; -} + if (*s != rw) + return NULL; + s += 1; -static int -is_register (const char *token, int len) -{ - if (token[0] == 'r' || token[0] == 'w') - if ((len == 2 && isdigit (token[1])) - || (len == 3 && token[1] == '1' && token[2] == '0')) - return 1; + if (*s == '1') + { + if (*(s + 1) == '0') + { + *regno = 10; + s += 2; + } + else + { + *regno = 1; + s += 1; + } + } + else if (*s >= '0' && *s <= '9') + { + *regno = *s - '0'; + s += 1; + } - return 0; + return s; } -static enum bpf_token_type -is_cast (const char *token) -{ - static const char *cast_rw[] = {"u8", "u16", "u32", "u64"}; - unsigned int i; - - for (i = 0; i < ARRAY_SIZE (cast_rw); ++i) - if (!strcmp (token, cast_rw[i])) - return BPF_CAST_U8 + i; +/* Collect a parse error message. */ - return BPF_UNKNOWN; -} +static int partial_match_length = 0; +static char *errmsg = NULL; -static enum bpf_token_type -get_token (const char **insn, char *token, size_t *tlen) +static void +parse_error (int length, const char *fmt, ...) { -#define GET() \ - (*str == '\0' \ - ? EOF \ - : *(unsigned char *)(str++)) - -#define UNGET() (--str) - -#define START_EXPR() \ - do \ - { \ - if (expr == NULL) \ - expr = str - 1; \ - } while (0) - -#define SCANNER_SKIP_WHITESPACE() \ - do \ - { \ - do \ - ch = GET (); \ - while (ch != EOF \ - && ((ch) == ' ' || (ch) == '\t')); \ - if (ch != EOF) \ - UNGET (); \ - } while (0) - - const char *str = *insn; - int ch, ch2 = 0; - enum bpf_token_type ttype = BPF_UNKNOWN; - size_t len = 0; - const char *expr = NULL; - const char *end_expr = NULL; - int state = 0; - int return_token = 0; - - while (1) + if (length > partial_match_length) { - ch = GET (); - - if (ch == EOF || len > MAX_TOKEN_SZ) - break; - - switch (pseudoc_lex[(unsigned char) ch]) - { - case LEX_IS_WHITESPACE: - SCANNER_SKIP_WHITESPACE (); - return_token = 1; - - switch (state) - { - case 12: /* >' ' */ - ttype = BPF_JGT; - break; - - case 17: /* ==' ' */ - ttype = BPF_JEQ; - break; - - case 18: /* <' ' */ - ttype = BPF_JLT; - break; - - case 20: /* &' ' */ - ttype = BPF_JSET; - break; - - case 22: /* s<' '*/ - ttype = BPF_JSLT; - break; - - case 14: /* s> ' ' */ - ttype = BPF_JSGT; - break; - - case 16: /* =' ' */ - ttype = BPF_MOV; - break; - - default: - return_token = 0; - } - break; - - case LEX_IS_EXCLA: - token[len++] = ch; - state = 21; - break; - - case LEX_IS_ARITHM_OP: - if (state == 16) - { - /* ='-' is handle as '=' */ - UNGET (); - ttype = BPF_MOV; - return_token = 1; - break; - } - - START_EXPR(); - token[len++] = ch; - switch (ch) - { -#define BPF_ARITHM_OP(op, type) \ - case (op): \ - state = 6; \ - ttype = (type); \ - break; - - BPF_ARITHM_OP('+', BPF_ADD); - BPF_ARITHM_OP('-', BPF_SUB); - BPF_ARITHM_OP('*', BPF_MUL); - BPF_ARITHM_OP('/', BPF_DIV); - BPF_ARITHM_OP('|', BPF_OR); - BPF_ARITHM_OP('%', BPF_MOD); - BPF_ARITHM_OP('^', BPF_XOR); - - case '&': - state = 20; /* '&' */ - break; - - case '<': - switch (state) - { - case 0: - state = 18; /* '<' */ - break; - - case 18: - state = 19; /* <'<' */ - break; - - case 8: - state = 22; /* s'<' */ - break; - } - break; - - case '>': - switch (state) - { - case 0: - state = 12; /* '>' */ - break; - - case 12: - state = 13; /* >'>' */ - break; - - case 8: - state = 14; /* s'>' */ - break; - - case 14: - state = 15; /* s>'>' */ - break; - } - break; - } - break; - - case LEX_IS_STAR: - switch (state) - { - case 0: - token[len++] = ch; - START_EXPR (); - state = 2; /* '*', It could be the fist cast char. */ - break; - - case 16: /* ='*' Not valid token. */ - ttype = BPF_MOV; - return_token = 1; - UNGET (); - break; - - case 4: /* *(uXX'*' */ - token[len++] = ch; - state = 5; - break; - } - break; - - case LEX_IS_OPEN_BR: - START_EXPR (); - token[len++] = ch; - return_token = 1; - - switch (state) - { - case 2: - state = 3; /* *'(' second char of a cast or expr. */ - return_token = 0; - break; - - case 6: - if (valid_expr (expr, &end_expr)) - { - len = end_expr - expr; - memcpy (token, expr, len); - ttype = BPF_EXPR; - str = end_expr; - } - else - { - len = 0; - while (*invalid_expression) - token[len++] = *invalid_expression++; - - token[len] = 0; - ttype = BPF_UNKNOWN; - } - break; - - default: - ttype = BPF_CHR_OPEN_BR; - SCANNER_SKIP_WHITESPACE (); - ch2 = GET (); - - if ((isdigit (ch2) || ch2 == '(') - && valid_expr (expr, &end_expr)) - { - len = end_expr - expr; - memcpy (token, expr, len); - ttype = BPF_EXPR; - str = end_expr; - } - else - UNGET (); - } - break; - - case LEX_IS_CLSE_BR: - token[len++] = ch; - - if (state == 0) - { - ttype = BPF_CHR_CLSE_BR; - return_token = 1; - } - else if (state == 5) /* *(uXX*')' */ - return_token = 1; - break; - - case LEX_IS_EQUAL: - token[len++] = ch; - return_token = 1; - - switch (state) - { - case 0: - state = 16; /* '=' */ - return_token = 0; - break; - - case 16: - state = 17; /* ='=' */ - return_token = 0; - break; - - case 2: /* *'=' */ - ttype = BPF_MUL; - break; - - case 10: /* s>>'=' */ - ttype = BPF_ARSH; - break; - - case 12: /* >'=' */ - ttype = BPF_JGE; - break; - - case 13: /* >>'=' */ - ttype = BPF_RSH; - break; - - case 14: /* s>'=' */ - ttype = BPF_JSGE; - break; - - case 15: /* s>>'=' */ - ttype = BPF_ARSH; - break; - - case 18: /* <'=' */ - ttype = BPF_JLE; - break; - - case 19: /* <<'=' */ - ttype = BPF_LSH; - break; - - case 20: /* &'=' */ - ttype = BPF_AND; - break; - - case 21: /* !'=' */ - ttype = BPF_JNE; - break; - - case 22: /* s<'=' */ - ttype = BPF_JSLE; - break; - } - break; - - case LEX_IS_SYMBOL_COMPONENT: - return_token = 1; - - switch (state) - { - case 17: /* =='sym' */ - ttype = BPF_JEQ; - break; - - case 12: /* >'sym' */ - ttype = BPF_JGT; - break; - - case 18: /* <'sym' */ - ttype = BPF_JLT; - break; - - case 20: /* &'sym' */ - ttype = BPF_JSET; - break; - - case 14: /*s>'sym' */ - ttype = BPF_JSGT; - break; - - case 22: /* s<'sym' */ - ttype = BPF_JSLT; - break; - - case 16: /* ='sym' */ - ttype = BPF_MOV; - break; - - default: - return_token = 0; - } - - if (return_token) - { - UNGET (); - break; - } - - START_EXPR (); - token[len++] = ch; - - while ((ch2 = GET ()) != EOF) - { - int type; - - type = pseudoc_lex[(unsigned char) ch2]; - if (type != LEX_IS_SYMBOL_COMPONENT) - break; - token[len++] = ch2; - } - - if (ch2 != EOF) - UNGET (); - - if (state == 0) - { - if (len == 1 && ch == 's') - state = 8; /* signed instructions: 's' */ - else - { - ttype = BPF_SYMBOL; - if (is_register (token, len)) - ttype = BPF_REG; - else if (look_for_reserved_word (token, &ttype)) - ; - else if ((pseudoc_lex[(unsigned char) *token] == LEX_IS_ARITHM_OP - || *token == '(' || isdigit(*token)) - && valid_expr (expr, &end_expr)) - { - len = end_expr - expr; - token[len] = '\0'; - ttype = BPF_EXPR; - str = end_expr; - } - - return_token = 1; - } - } - else if (state == 3) /* *('sym' */ - { - if ((ttype = is_cast (&token[2])) != BPF_UNKNOWN) - state = 4; /* *('uXX' */ - else - { - ttype = BPF_EXPR; - return_token = 1; - } - } - else if (state == 6) - { - if (ttype == BPF_SUB) /* neg */ - { - if (is_register (&token[1], len - 1)) - ttype = BPF_NEG; - else if (valid_expr(expr, &end_expr)) - { - len = end_expr - expr; - memcpy(token, expr, len); - ttype = BPF_EXPR; - str = end_expr; - } - else - { - len = 0; - while (*invalid_expression) - token[len++] = *invalid_expression++; - token[len] = 0; - ttype = BPF_UNKNOWN; - } - } - else if (valid_expr (expr, &end_expr)) - { - len = end_expr - expr; - memcpy(token, expr, len); - ttype = BPF_EXPR; - str = end_expr; - } - else - ttype = BPF_UNKNOWN; - - return_token = 1; - } - break; - } + va_list args; - if (return_token) - { - *tlen = len; - *insn = str; - break; - } + free (errmsg); + va_start (args, fmt); + errmsg = xvasprintf (fmt, args); + va_end (args); + partial_match_length = length; } - - return ttype; - -#undef GET -#undef UNGET -#undef START_EXPR -#undef SCANNER_SKIP_WHITESPACE -#undef BPF_ARITHM_OP } -/* - The parser represent a FSM for the grammar described above. So for example - the following rule: - - ` bpf_alu_insn : BPF_REG bpf_alu_operator register_or_imm32' - - Is parser as follows: - - 1. It starts in state 0. +/* Assemble a machine instruction in STR and emit the frags/bytes it + assembles to. */ - 2. Consumes next token, e.g: `BPF_REG' and set `state' variable to a - particular state to helps to identify, in this case, that a register - token has been read, a comment surrounded by a single quote in the - pseudo-c token is added along with the new `state' value to indicate - what the scanner has read, e.g.: - - state = 6; // dst_reg = str_cast ( 'src_reg' - - So, in `state 6' the scanner has consumed: a destination register - (BPF_REG), an equal character (BPF_MOV), a cast token (BPF_CAST), an - open parenthesis (BPF_CHR_OPEN_BR) and the source register (BPF_REG). - - 3. If the accumulated tokens represent a complete BPF pseudo-c syntax - instruction then, a validation of the terms is made, for example: if - the registers have the same sizes (32/64 bits), if a specific - destination register must be used, etc., after that, a builder: - build_bfp_{non_generic_load,atomic_insn,jmp_insn,arithm_insn,endianness,load_store_insn} - is invoked, internally, it translates the BPF pseudo-c instruction to - a BPF GAS instruction using the previous terms recollected by the - scanner. - - 4. If a successful build of BPF GAS instruction was done, a final - state is set to `ST_EOI' (End Of Instruction) meaning that is not - expecting for more tokens in such instruction. Otherwise if the - conditions to calling builder are not satisfied an error is emitted - and `parse_err' is set. -*/ - -static char * -bpf_pseudoc_to_normal_syntax (const char *str, char **errmsg) +void +md_assemble (char *str ATTRIBUTE_UNUSED) { -#define syntax_err(format, ...) \ - do \ - { \ - if (! parse_err) \ - { \ - parse_err = 1; \ - errbuf = xasprintf (format, ##__VA_ARGS__); \ - } \ - } while (0) - - enum bpf_token_type ttype; - enum bpf_token_type bpf_endianness = BPF_UNKNOWN, - bpf_atomic_insn; - enum bpf_token_type bpf_jmp_op = BPF_JEQ; /* Arbitrary. */ - enum bpf_token_type bpf_cast = BPF_CAST_U8; /* Arbitrary. */ - enum bpf_token_type bpf_arithm_op = BPF_ADD; /* Arbitrary. */ - char *bpf_insn = NULL; - char *errbuf = NULL; - char src_reg[3] = {0}; - char dst_reg[3] = {0}; - char str_imm32[40] = {0}; - char str_offset[40] = {0}; - char str_symbol[MAX_TOKEN_SZ] = {0}; - char token[MAX_TOKEN_SZ] = {0}; - int state = 0; - int parse_err = 0; - size_t tlen; - - while (*str) + /* There are two different syntaxes that can be used to write BPF + instructions. One is very conventional and like any other + assembly language where each instruction is conformed by an + instruction mnemonic followed by its operands. This is what we + call the "normal" syntax. The other syntax tries to look like C + statements. We have to support both syntaxes in this assembler. + + One of the many nuisances introduced by this eccentricity is that + in the pseudo-c syntax it is not possible to hash the opcodes + table by instruction mnemonic, because there is none. So we have + no other choice than to try to parse all instruction opcodes + until one matches. This is slow. + + Another problem is that emitting detailed diagnostics becomes + tricky, since the lack of mnemonic means it is not clear what + instruction was intended by the user, and we cannot emit + diagnostics for every attempted template. So if an instruction + is not parsed, we report the diagnostic corresponding to the + partially parsed instruction that was matched further. */ + + unsigned int idx = 0; + struct bpf_insn insn; + const struct bpf_opcode *opcode; + + /* Initialize the global diagnostic variables. See the parse_error + function above. */ + partial_match_length = 0; + errmsg = NULL; + +#define PARSE_ERROR(...) parse_error (s - str, __VA_ARGS__) + + while ((opcode = bpf_get_opcode (idx++)) != NULL) { - ttype = get_token (&str, token, &tlen); - if (ttype == BPF_UNKNOWN || state == ST_EOI) - { - syntax_err ("unexpected token: '%s'", token); - break; - } - - switch (ttype) - { - case BPF_UNKNOWN: - case BPF_LL: - break; - - case BPF_REG: - switch (state) - { - case 0: - memcpy (dst_reg, token, tlen); - state = 1; /* 'dst_reg' */ - break; - - case 3: - /* dst_reg bpf_op 'src_reg' */ - memcpy (src_reg, token, tlen); - if (*dst_reg == *src_reg) - bpf_insn = build_bpf_arithm_insn (dst_reg, src_reg, 0, - NULL, bpf_arithm_op); - else - { - syntax_err ("different register sizes: '%s', '%s'", - dst_reg, src_reg); - break; - } - state = ST_EOI; - break; - - case 5: - memcpy (src_reg, token, tlen); - state = 6; /* dst_reg = str_cast ( 'src_reg' */ - break; - - case 9: - memcpy (dst_reg, token, tlen); - state = 10; /* str_cast ( 'dst_reg' */ - break; - - case 11: - /* str_cast ( dst_reg offset ) = 'src_reg' */ - memcpy (src_reg, token, tlen); - bpf_insn = build_bpf_load_store_insn (dst_reg, src_reg, - bpf_cast, str_offset, 0); - state = ST_EOI; - break; - - case 14: - memcpy (dst_reg, token, tlen); - state = 15; /* if 'dst_reg' */ - break; - - case 16: - memcpy (src_reg, token, tlen); - state = 17; /* if dst_reg jmp_op 'src_reg' */ - break; - - case 24: - /* dst_reg = endianness src_reg */ - memcpy (src_reg, token, tlen); - if (*dst_reg == 'r' && !strcmp (dst_reg, src_reg)) - bpf_insn = build_bpf_endianness (dst_reg, bpf_endianness); - else - syntax_err ("invalid operand for instruction: '%s'", token); - - state = ST_EOI; - break; - - case 28: - memcpy (dst_reg, token, tlen); - state = 29; /* lock str_cast ( 'dst_reg' */ - break; - - case 32: - { - /* lock str_cast ( dst_reg offset ) atomic_insn 'src_reg' */ - int with_offset = *str_offset != '\0'; - - memcpy (src_reg, token, tlen); - if ((bpf_cast != BPF_CAST_U32 - && bpf_cast != BPF_CAST_U64) - || *dst_reg != 'r' - || *src_reg != 'r') - syntax_err ("invalid wide atomic instruction"); - else - bpf_insn = build_bpf_atomic_insn (dst_reg, src_reg, bpf_atomic_insn, - bpf_cast, with_offset ? str_offset : str_symbol); - } - - state = ST_EOI; - break; - - case 33: - /* callx 'dst_reg' */ - bpf_insn = xasprintf ("%s %%%s", "call", token); - state = ST_EOI; - break; - - case 35: - memcpy (src_reg, token, tlen); - state = 36; /* dst_reg = str_cast skb [ 'src_reg' */ - break; - } - break; - - case BPF_MOV: - case BPF_ADD: - case BPF_SUB: - case BPF_MUL: - case BPF_DIV: - case BPF_OR: - case BPF_AND: - case BPF_LSH: - case BPF_RSH: - case BPF_MOD: - case BPF_XOR: - case BPF_ARSH: - case BPF_NEG: - switch (state) - { - case 1: - state = 3; /* dst_reg 'arith_op' */ - bpf_arithm_op = ttype; - break; - - case 3: - if (ttype == BPF_NEG) - { - /* reg = -reg */ - bpf_arithm_op = ttype; - memcpy (src_reg, token + 1, tlen - 1); - if (strcmp (dst_reg, src_reg)) - { - syntax_err ("found: '%s', expected: -%s", token, dst_reg); - break; - } - - bpf_insn = build_bpf_arithm_insn (dst_reg, src_reg, 0, - NULL, bpf_arithm_op); - state = ST_EOI; - } - break; - - case 23: - memcpy (src_reg, token, tlen); - state = 11; /* str_cast ( dst_reg offset ) '=' */ - break; - - case 12: - if (ttype == BPF_MOV) - state = 13; /* str_cast ( dst_reg offset ) '=' */ - break; - - case 31: - bpf_atomic_insn = ttype; - state = 32; /* lock str_cast ( dst_reg offset ) 'atomic_insn' */ - break; - - default: - syntax_err ("unexpected '%s'", token); - state = ST_EOI; - } - break; - - case BPF_CAST_U8: - case BPF_CAST_U16: - case BPF_CAST_U32: - case BPF_CAST_U64: - bpf_cast = ttype; - switch (state) - { - case 3: - state = 4; /* dst_reg = 'str_cast' */ - break; - - case 0: - state = 8; /* 'str_cast' */ - break; - - case 26: - state = 27; /* lock 'str_cast' */ - break; - } - break; - - case BPF_CHR_OPEN_BR: - switch (state) - { - case 4: - state = 5; /* dst_reg = str_cast '(' */ - break; - - case 8: - state = 9; /* str_cast '(' */ - break; - - case 27: - state = 28; /* lock str_cast '(' */ - break; - - case 34: - state = 35; /* dst_reg = str_cast skb '[' */ - break; - } - break; + const char *p; + char *s; + const char *template + = (asm_dialect == DIALECT_PSEUDOC ? opcode->pseudoc : opcode->normal); + + /* Do not try to match opcodes with a higher version than the + selected ISA spec. */ + if (opcode->version > isa_spec) + continue; + + memset (&insn, 0, sizeof (struct bpf_insn)); + insn.size = 8; + for (s = str, p = template; *p != '\0';) + { + if (*p == ' ') + { + /* Expect zero or more spaces. */ + while (*s != '\0' && (*s == ' ' || *s == '\t')) + s += 1; + p += 1; + } + else if (*p == '%') + { + if (*(p + 1) == '%') + { + if (*s != '%') + { + PARSE_ERROR ("expected '%%'"); + break; + } + p += 2; + s += 1; + } + else if (*(p + 1) == 'w') + { + /* Expect zero or more spaces. */ + while (*s != '\0' && (*s == ' ' || *s == '\t')) + s += 1; + p += 2; + } + else if (*(p + 1) == 'W') + { + /* Expect one or more spaces. */ + if (*s != ' ' && *s != '\t') + { + PARSE_ERROR ("expected white space, got '%s'", + s); + break; + } + while (*s != '\0' && (*s == ' ' || *s == '\t')) + s += 1; + p += 2; + } + else if (strncmp (p, "%dr", 3) == 0) + { + uint8_t regno; + char *news = parse_bpf_register (s, 'r', ®no); + + if (news == NULL || (insn.has_dst && regno != insn.dst)) + { + if (news != NULL) + PARSE_ERROR ("expected register r%d, got r%d", + insn.dst, regno); + else + PARSE_ERROR ("expected register name, got '%s'", s); + break; + } + s = news; + insn.dst = regno; + insn.has_dst = 1; + p += 3; + } + else if (strncmp (p, "%sr", 3) == 0) + { + uint8_t regno; + char *news = parse_bpf_register (s, 'r', ®no); + + if (news == NULL || (insn.has_src && regno != insn.src)) + { + if (news != NULL) + PARSE_ERROR ("expected register r%d, got r%d", + insn.dst, regno); + else + PARSE_ERROR ("expected register name, got '%s'", s); + break; + } + s = news; + insn.src = regno; + insn.has_src = 1; + p += 3; + } + else if (strncmp (p, "%dw", 3) == 0) + { + uint8_t regno; + char *news = parse_bpf_register (s, 'w', ®no); + + if (news == NULL || (insn.has_dst && regno != insn.dst)) + { + if (news != NULL) + PARSE_ERROR ("expected register r%d, got r%d", + insn.dst, regno); + else + PARSE_ERROR ("expected register name, got '%s'", s); + break; + } + s = news; + insn.dst = regno; + insn.has_dst = 1; + p += 3; + } + else if (strncmp (p, "%sw", 3) == 0) + { + uint8_t regno; + char *news = parse_bpf_register (s, 'w', ®no); + + if (news == NULL || (insn.has_src && regno != insn.src)) + { + if (news != NULL) + PARSE_ERROR ("expected register r%d, got r%d", + insn.dst, regno); + else + PARSE_ERROR ("expected register name, got '%s'", s); + break; + } + s = news; + insn.src = regno; + insn.has_src = 1; + p += 3; + } + else if (strncmp (p, "%i32", 4) == 0 + || strncmp (p, "%I32", 4) == 0) + { + if (p[1] == 'I') + { + while (*s == ' ' || *s == '\t') + s += 1; + if (*s != '+' && *s != '-') + { + PARSE_ERROR ("expected `+' or `-', got `%c'", *s); + break; + } + } + + s = parse_expression (s, &insn.imm32); + if (s == NULL) + { + PARSE_ERROR ("expected signed 32-bit immediate"); + break; + } + insn.has_imm32 = 1; + p += 4; + } + else if (strncmp (p, "%o16", 4) == 0) + { + while (*s == ' ' || *s == '\t') + s += 1; + if (*s != '+' && *s != '-') + { + PARSE_ERROR ("expected `+' or `-', got `%c'", *s); + break; + } + + s = parse_expression (s, &insn.offset16); + if (s == NULL) + { + PARSE_ERROR ("expected signed 16-bit offset"); + break; + } + insn.has_offset16 = 1; + p += 4; + } + else if (strncmp (p, "%d16", 4) == 0) + { + s = parse_expression (s, &insn.disp16); + if (s == NULL) + { + PARSE_ERROR ("expected signed 16-bit displacement"); + break; + } + insn.has_disp16 = 1; + p += 4; + } + else if (strncmp (p, "%d32", 4) == 0) + { + s = parse_expression (s, &insn.disp32); + if (s == NULL) + { + PARSE_ERROR ("expected signed 32-bit displacement"); + break; + } + insn.has_disp32 = 1; + p += 4; + } + else if (strncmp (p, "%i64", 4) == 0) + { + s = parse_expression (s, &insn.imm64); + if (s == NULL) + { + PARSE_ERROR ("expected signed 64-bit immediate"); + break; + } + insn.has_imm64 = 1; + insn.size = 16; + p += 4; + } + else + as_fatal (_("invalid %%-tag in BPF opcode '%s'\n"), template); + } + else + { + /* Match a literal character. */ + if (*s != *p) + { + if (*s == '\0') + PARSE_ERROR ("expected '%c'", *p); + else if (*s == '%') + { + /* This is to workaround a bug in as_bad. */ + char tmp[3]; + + tmp[0] = '%'; + tmp[1] = '%'; + tmp[2] = '\0'; + + PARSE_ERROR ("expected '%c', got '%s'", *p, tmp); + } + else + PARSE_ERROR ("expected '%c', got '%c'", *p, *s); + break; + } + p += 1; + s += 1; + } + } - case BPF_CHR_CLSE_BR: - switch (state) - { - case 7: - /* dst_reg = str_cast ( imm32 ')' */ - bpf_insn = build_bpf_load_store_insn (dst_reg, src_reg, - bpf_cast, str_imm32, 1); - state = ST_EOI; - break; - - case 11: - state = 12; /* str_cast ( dst_reg imm32 ')' */ - break; - - case 21: - /* dst_reg = str_cast ( src_reg offset ')' */ - bpf_insn = build_bpf_load_store_insn (dst_reg, src_reg, - bpf_cast, str_offset, 1); - state = ST_EOI; - break; - - case 22: - state = 23; /* str_cast ( dst_reg offset ')' */ - break; - - case 30: - state = 31; /* lock str_cast ( dst_reg offset ')' */ - break; - - case 37: - /* dst_reg = str_cast skb [ src_reg imm32 ']' */ - if (*dst_reg != 'w' && !strcmp ("r0", dst_reg)) - bpf_insn = build_bpf_non_generic_load (*src_reg != '\0' ? src_reg : NULL, - bpf_cast, str_imm32); - else - syntax_err ("invalid register operand: '%s'", dst_reg); - - state = ST_EOI; - break; - } - break; + if (*p == '\0') + { + /* Allow white spaces at the end of the line. */ + while (*s != '\0' && (*s == ' ' || *s == '\t')) + s += 1; + if (*s == '\0') + /* We parsed an instruction successfully. */ + break; + PARSE_ERROR ("extra junk at end of line"); + } + } - case BPF_EXPR: - switch (state) - { - case 3: - { - /* dst_reg bpf_arithm_op 'imm32' */ - int load64 = 0; - - memcpy (str_imm32, token, tlen); - memset (token, 0, tlen); - - if ((ttype = get_token (&str, token, &tlen)) == BPF_LL - && bpf_arithm_op == BPF_MOV) - load64 = 1; - else if (ttype != BPF_UNKNOWN) - syntax_err ("unexpected token: '%s'", token); - - if (load64 && *dst_reg == 'w') - syntax_err ("unexpected register size: '%s'", dst_reg); - - if (! parse_err) - bpf_insn = build_bpf_arithm_insn (dst_reg, NULL, load64, - str_imm32, bpf_arithm_op); - state = ST_EOI; - } - break; - - case 18: - { - /* if dst_reg jmp_op src_reg goto 'offset' */ - int with_src = *src_reg != '\0'; - - memcpy (str_offset, token, tlen); - if (with_src && *dst_reg != *src_reg) - syntax_err ("different register size: '%s', '%s'", - dst_reg, src_reg); - else - bpf_insn = build_bpf_jmp_insn (dst_reg, with_src ? src_reg : NULL, - with_src ? NULL: str_imm32, - bpf_jmp_op, NULL, str_offset); - state = ST_EOI; - } - break; - - case 19: - /* goto 'offset' */ - memcpy (str_offset, token, tlen); - bpf_insn = xasprintf ("%s %s", "ja", str_offset); - state = ST_EOI; - break; - - case 6: - memcpy (str_offset, token, tlen); - state = 21; /* dst_reg = str_cast ( src_reg 'offset' */ - break; - - case 10: - memcpy (str_offset, token, tlen); - state = 22; /* str_cast ( dst_reg 'offset' */ - break; - - case 16: - memcpy (str_imm32, token, tlen); - state = 25; /* if dst_reg jmp_op 'imm32' */ - break; - - case 29: - memcpy (str_offset, token, tlen); - state = 30; /* lock str_cast ( dst_reg 'offset' */ - break; - - case 34: - /* dst_reg = str_cast skb 'imm32' */ - if (*dst_reg != 'w' && !strcmp ("r0", dst_reg)) - { - memcpy (str_imm32, token, tlen); - bpf_insn = build_bpf_non_generic_load (*src_reg != '\0' ? src_reg : NULL, - bpf_cast, str_imm32); - } - else - syntax_err ("invalid register operand: '%s'", dst_reg); - - state = ST_EOI; - break; - - case 36: - memcpy (str_imm32, token, tlen); - state = 37; /* dst_reg = str_cast skb [ src_reg 'imm32' */ - break; - } - break; + if (opcode == NULL) + { + as_bad (_("unrecognized instruction `%s'"), str); + if (errmsg != NULL) + { + as_bad (errmsg); + free (errmsg); + } - case BPF_IF: - if (state == 0) - state = 14; - break; + return; + } + insn.opcode = opcode->opcode; - case BPF_JSGT: - case BPF_JSLT: - case BPF_JSLE: - case BPF_JSGE: - case BPF_JGT: - case BPF_JGE: - case BPF_JLE: - case BPF_JSET: - case BPF_JNE: - case BPF_JLT: - case BPF_JEQ: - if (state == 15) - { - bpf_jmp_op = ttype; - state = 16; /* if dst_reg 'jmp_op' */ - } - break; +#undef PARSE_ERROR - case BPF_GOTO: - switch (state) - { - case 17: - case 25: - state = 18; /* if dst_reg jmp_op src_reg|imm32 'goto' */ - break; - - case 0: - state = 19; - break; - } - break; + /* Generate the frags and fixups for the parsed instruction. */ + { + char *this_frag = frag_more (insn.size); + char bytes[16]; + uint8_t src, dst; + int i; + + /* Zero all the bytes. */ + memset (bytes, 0, 16); + + /* First encode the opcodes. Note that we have to handle the + endianness groups of the BPF instructions: 8 | 4 | 4 | 16 | + 32. */ + if (target_big_endian) + { + /* code */ + bytes[0] = (insn.opcode >> 56) & 0xff; + /* regs */ + bytes[1] = (insn.opcode >> 48) & 0xff; + /* offset16 */ + bytes[2] = (insn.opcode >> 40) & 0xff; + bytes[3] = (insn.opcode >> 32) & 0xff; + /* imm32 */ + bytes[4] = (insn.opcode >> 24) & 0xff; + bytes[5] = (insn.opcode >> 16) & 0xff; + bytes[6] = (insn.opcode >> 8) & 0xff; + bytes[7] = insn.opcode & 0xff; + } + else + { + /* code */ + bytes[0] = (insn.opcode >> 56) & 0xff; + /* regs */ + bytes[1] = (((((insn.opcode >> 48) & 0xff) & 0xf) << 4) + | (((insn.opcode >> 48) & 0xff) & 0xf)); + /* offset16 */ + bytes[3] = (insn.opcode >> 40) & 0xff; + bytes[2] = (insn.opcode >> 32) & 0xff; + /* imm32 */ + bytes[7] = (insn.opcode >> 24) & 0xff; + bytes[6] = (insn.opcode >> 16) & 0xff; + bytes[5] = (insn.opcode >> 8) & 0xff; + bytes[4] = insn.opcode & 0xff; + } - case BPF_SYMBOL: - switch (state) - { - case 18: - { - /* if dst_reg jmp_op src_reg goto 'sym' */ - int with_src = *src_reg != '\0'; - - memcpy (str_symbol, token, tlen); - if (with_src && *dst_reg != *src_reg) - syntax_err ("different register size: '%s', '%s'", - dst_reg, src_reg); - else - bpf_insn = build_bpf_jmp_insn (dst_reg, with_src ? src_reg : NULL, - with_src ? NULL: str_imm32, - bpf_jmp_op, str_symbol, NULL); - state = ST_EOI; - } - break; - - case 19: - /* goto 'sym' */ - memcpy (str_symbol, token, tlen); - bpf_insn = xasprintf ("%s %s", "ja", str_symbol); - state = ST_EOI; - break; - - case 0: - state = ST_EOI; - break; - - case 3: - { - /* dst_reg arithm_op 'sym' */ - int load64 = 0; - - memcpy (str_symbol, token, tlen); - memset (token, 0, tlen); - - if ((ttype = get_token (&str, token, &tlen)) == BPF_LL - && bpf_arithm_op == BPF_MOV) - load64 = 1; - else if (ttype != BPF_UNKNOWN) - syntax_err ("unexpected token: '%s'", token); - - if (load64 && *dst_reg == 'w') - syntax_err ("unexpected register size: '%s'", dst_reg); - - if (! parse_err) - bpf_insn = build_bpf_arithm_insn (dst_reg, NULL, load64, - str_symbol, bpf_arithm_op); - state = ST_EOI; - } - break; - } - break; + /* Now the registers. */ + src = insn.has_src ? insn.src : 0; + dst = insn.has_dst ? insn.dst : 0; - case BPF_LE16: - case BPF_LE32: - case BPF_LE64: - case BPF_BE16: - case BPF_BE32: - case BPF_BE64: - bpf_endianness = ttype; - state = 24; /* dst_reg = 'endianness' */ - break; + if (target_big_endian) + bytes[1] = ((dst & 0xf) << 4) | (src & 0xf); + else + bytes[1] = ((src & 0xf) << 4) | (dst & 0xf); - case BPF_LOCK: - state = 26; - break; + /* Now the immediates. */ + if (insn.has_imm64) + { + switch (insn.imm64.X_op) + { + case O_constant: + { + uint64_t imm64 = insn.imm64.X_add_number; + + if (target_big_endian) + { + bytes[12] = (imm64 >> 56) & 0xff; + bytes[13] = (imm64 >> 48) & 0xff; + bytes[14] = (imm64 >> 40) & 0xff; + bytes[15] = (imm64 >> 32) & 0xff; + bytes[4] = (imm64 >> 24) & 0xff; + bytes[5] = (imm64 >> 16) & 0xff; + bytes[6] = (imm64 >> 8) & 0xff; + bytes[7] = imm64 & 0xff; + } + else + { + bytes[15] = (imm64 >> 56) & 0xff; + bytes[14] = (imm64 >> 48) & 0xff; + bytes[13] = (imm64 >> 40) & 0xff; + bytes[12] = (imm64 >> 32) & 0xff; + bytes[7] = (imm64 >> 24) & 0xff; + bytes[6] = (imm64 >> 16) & 0xff; + bytes[5] = (imm64 >> 8) & 0xff; + bytes[4] = imm64 & 0xff; + } + break; + } + case O_symbol: + case O_subtract: + case O_add: + { + reloc_howto_type *reloc_howto; + int size; + + reloc_howto = bfd_reloc_type_lookup (stdoutput, BFD_RELOC_BPF_64); + if (!reloc_howto) + abort (); + + size = bfd_get_reloc_size (reloc_howto); + + fix_new_exp (frag_now, this_frag - frag_now->fr_literal, + size, &insn.imm64, reloc_howto->pc_relative, + BFD_RELOC_BPF_64); + break; + } + default: + abort (); + } + } - case BPF_IND_CALL: - state = 33; - break; + if (insn.has_imm32) + { + switch (insn.imm32.X_op) + { + case O_constant: + { + uint32_t imm32 = insn.imm32.X_add_number; + + if (target_big_endian) + { + bytes[4] = (imm32 >> 24) & 0xff; + bytes[5] = (imm32 >> 16) & 0xff; + bytes[6] = (imm32 >> 8) & 0xff; + bytes[7] = imm32 & 0xff; + } + else + { + bytes[7] = (imm32 >> 24) & 0xff; + bytes[6] = (imm32 >> 16) & 0xff; + bytes[5] = (imm32 >> 8) & 0xff; + bytes[4] = imm32 & 0xff; + } + break; + } + case O_symbol: + case O_subtract: + case O_add: + case O_uminus: + { + reloc_howto_type *reloc_howto; + int size; + + reloc_howto = bfd_reloc_type_lookup (stdoutput, BFD_RELOC_32); + if (!reloc_howto) + abort (); + + size = bfd_get_reloc_size (reloc_howto); + + fix_new_exp (frag_now, this_frag - frag_now->fr_literal + 4, + size, &insn.imm32, reloc_howto->pc_relative, + BFD_RELOC_32); + break; + } + default: + abort (); + } + } - case BPF_LD: - state = 34; /* dst_reg = str_cast 'skb' */ - break; - } + if (insn.has_disp32) + { + switch (insn.disp32.X_op) + { + case O_constant: + { + uint32_t disp32 = insn.disp32.X_add_number; + + if (target_big_endian) + { + bytes[4] = (disp32 >> 24) & 0xff; + bytes[5] = (disp32 >> 16) & 0xff; + bytes[6] = (disp32 >> 8) & 0xff; + bytes[7] = disp32 & 0xff; + } + else + { + bytes[7] = (disp32 >> 24) & 0xff; + bytes[6] = (disp32 >> 16) & 0xff; + bytes[5] = (disp32 >> 8) & 0xff; + bytes[4] = disp32 & 0xff; + } + break; + } + case O_symbol: + case O_subtract: + case O_add: + { + reloc_howto_type *reloc_howto; + int size; + + reloc_howto = bfd_reloc_type_lookup (stdoutput, BFD_RELOC_BPF_DISP32); + if (!reloc_howto) + abort (); + + size = bfd_get_reloc_size (reloc_howto); + + fix_new_exp (frag_now, this_frag - frag_now->fr_literal, + size, &insn.disp32, reloc_howto->pc_relative, + BFD_RELOC_BPF_DISP32); + break; + } + default: + abort (); + } + } - memset (token, 0, tlen); - } + if (insn.has_offset16) + { + switch (insn.offset16.X_op) + { + case O_constant: + { + uint32_t offset16 = insn.offset16.X_add_number; + + if (target_big_endian) + { + bytes[2] = (offset16 >> 8) & 0xff; + bytes[3] = offset16 & 0xff; + } + else + { + bytes[3] = (offset16 >> 8) & 0xff; + bytes[2] = offset16 & 0xff; + } + break; + } + case O_symbol: + case O_subtract: + case O_add: + { + reloc_howto_type *reloc_howto; + int size; + + reloc_howto = bfd_reloc_type_lookup (stdoutput, BFD_RELOC_BPF_DISP16); + if (!reloc_howto) + abort (); + + size = bfd_get_reloc_size (reloc_howto); + + fix_new_exp (frag_now, this_frag - frag_now->fr_literal, + size, &insn.offset16, reloc_howto->pc_relative, + BFD_RELOC_BPF_DISP16); + break; + } + default: + abort (); + } + } - if (state != ST_EOI) - syntax_err ("incomplete instruction"); + if (insn.has_disp16) + { + switch (insn.disp16.X_op) + { + case O_constant: + { + uint32_t disp16 = insn.disp16.X_add_number; + + if (target_big_endian) + { + bytes[2] = (disp16 >> 8) & 0xff; + bytes[3] = disp16 & 0xff; + } + else + { + bytes[3] = (disp16 >> 8) & 0xff; + bytes[2] = disp16 & 0xff; + } + break; + } + case O_symbol: + case O_subtract: + case O_add: + { + reloc_howto_type *reloc_howto; + int size; + + reloc_howto = bfd_reloc_type_lookup (stdoutput, BFD_RELOC_BPF_DISP16); + if (!reloc_howto) + abort (); + + size = bfd_get_reloc_size (reloc_howto); + + fix_new_exp (frag_now, this_frag - frag_now->fr_literal, + size, &insn.disp16, reloc_howto->pc_relative, + BFD_RELOC_BPF_DISP16); + break; + } + default: + abort (); + } + } - *errmsg = errbuf; - return bpf_insn; + /* Emit bytes. */ + for (i = 0; i < insn.size; ++i) + { + md_number_to_chars (this_frag, (valueT) bytes[i], 1); + this_frag += 1; + } + } -#undef syntax_err + /* Emit DWARF2 debugging information. */ + dwarf2_emit_insn (insn.size); } -void -md_assemble (char *str) -{ - const CGEN_INSN *insn; - char *errmsg; - char *a_errmsg; - CGEN_FIELDS fields; - char *normal; - -#if CGEN_INT_INSN_P - CGEN_INSN_INT buffer[CGEN_MAX_INSN_SIZE / sizeof (CGEN_INT_INSN_P)]; -#else - unsigned char buffer[CGEN_MAX_INSN_SIZE]; -#endif - - gas_cgen_init_parse (); - insn = bpf_cgen_assemble_insn (gas_cgen_cpu_desc, str, &fields, - buffer, &errmsg); - if (insn == NULL) - { - normal = bpf_pseudoc_to_normal_syntax (str, &a_errmsg); - if (normal) - { - insn = bpf_cgen_assemble_insn (gas_cgen_cpu_desc, normal, &fields, - buffer, &a_errmsg); - xfree (normal); - } - - if (insn == NULL) - { - as_bad ("%s", errmsg); - if (a_errmsg) - { - as_bad ("%s", a_errmsg); - xfree (a_errmsg); - } - return; - } - } - - gas_cgen_finish_insn (insn, buffer, CGEN_FIELDS_BITSIZE (&fields), - 0, /* zero to ban relaxable insns. */ - NULL); /* NULL so results not returned here. */ -} +/* Parse an operand that is machine-specific. */ void md_operand (expressionS *expressionP) { - invalid_expression = input_line_pointer - 1; - gas_cgen_md_operand (expressionP); + /* If this hook is invoked it means GAS failed to parse a generic + expression. We should inhibit the as_bad in expr.c, so we can fail + while parsing instruction alternatives. To do that, we change the + expression to not have an O_absent. But then we also need to set + exp_parse_failed to parse_expression above does the right thing. */ + ++input_line_pointer; + expressionP->X_op = O_constant; + expressionP->X_add_number = 0; + exp_parse_failed = 1; } - symbolS * md_undefined_symbol (char *name ATTRIBUTE_UNUSED) { @@ -1929,3 +1192,29 @@ md_atof (int type, char *litP, int *sizeP) { return ieee_md_atof (type, litP, sizeP, false); } + + +/* Determine whether the equal sign in the given string corresponds to + a BPF instruction, i.e. when it is not to be considered a symbol + assignment. */ + +bool +bpf_tc_equal_in_insn (int c ATTRIBUTE_UNUSED, char *str ATTRIBUTE_UNUSED) +{ + uint8_t regno; + + /* Only pseudo-c instructions can have equal signs, and of these, + all that could be confused with a symbol assignment all start + with a register name. */ + if (asm_dialect == DIALECT_PSEUDOC) + { + char *w = parse_bpf_register (str, 'w', ®no); + char *r = parse_bpf_register (str, 'r', ®no); + + if ((w != NULL && *w == '\0') + || (r != NULL && *r == '\0')) + return 1; + } + + return 0; +} |