Age | Commit message (Collapse) | Author | Files | Lines |
|
* ia64-asmtab.c: Revert 2000-12-16 change.
|
|
operands is greater than 127.
2001-02-02 Patrick Macdonald <patrickm@redhat.com>
* cgen.h (CGEN_SYNTAX_CHAR_TYPE): Typedef as unsigned short.
(CGEN_MAX_SYNTAX_ELEMENTS): Rename from CGEN_MAX_SYNTAX_BYTES.
(CGEN_SYNTAX): Define using CGEN_MAX_SYNTAX_ELEMENTS.
* fr30-desc.h: Regenerate with CGEN_MAX_SYNTAX_ELEMENTS.
* m32r-desc.h: Regenerate.
|
|
(putop): Handle 'T', alphabetize order, fix 'I' handling in Intel syntax
|
|
mode, 16 bit forms of ldi, ldo, ldw and stw instructions.
|
|
|
|
* i386.h (i386_optab): Fix pusha and ret templates.
* i386-dis.c (dis386_att, disx86_64_att): Fix ret, lret and iret
templates.
|
|
|
|
|
|
(dis386_twobyt): Add SSE2 instructions.
(twobyte_uses_SSE_prefix: Rename from ... ; add new SSE instructions.
(twobyte_uses_f3_prefix): ... this one.
(grps): Add SSE instructions.
(prefix_user_table): Add two new slots; add SSE2 instructions.
(print_insn_i386): Rename uses_f3_prefix to uses_SSE_prefix;
Handle the REPNZ and Data16 prefixes as well; do proper lookup
to prefix_user_table.
(OP_E): Accept mfence and lfence as well.
(OP_MMX): Data16 prefix turns MMX to SSE; support REX extensions.
(OP_XMM): Support REX extensions.
(OP_EM): Likewise.
(OP_EX): Likewise.
|
|
|
|
* cgen-asm.in (parse_insn_normal): Changed syn to be
CGEN_SYNTAX_CHAR_TYPE. Changed all references to *syn
as character to use CGEN_SYNTAX_CHAR macro and all comparisons
to '\0' to use 0 instead.
* cgen-dis.in (print_insn_normal): Ditto.
* cgen-ibld.in (insert_insn_normal, extract_insn_normal): Ditto.
|
|
(rex): New static variable.
(REX_MODE64, REX_EXTX, REX_EXTY, REX_EXTZ): New constants.
(USED_REX): New macro.
(Ev, Ed, Rm, Iq, Iv64, Cm, Dm, Rm*, Ob64, Ov64): New macros.
(OP_I64, OP_OFF64, OP_IMREG): New functions.
(OP_REG, OP_OFF): Declare.
(get64, get32, get32s): New functions.
(r??_reg): New constants.
(dis386_att): Change templates of instruction implicitly promoted
to 64bit; change e?? to RMe?? for unwind RM byte instructions.
(grps): Likewise.
(dis386_intel): Likewise.
(dixx86_64_att): New table based on dis386_att.
(dixx86_64_intel): New table based on dis386_intel.
(names64, names8rex): New global variable.
(names32, names16): Add extended registers.
(prefix_user_t): Recognize rex prefixes.
(prefix_name): Print REX prefixes nicely.
(op_riprel): New global variable.
(start_pc): Set type to bfd_vma.
(print_insn_i386): Detect the 64bit mode and use proper table;
move ckprefix after initializing the buffer; output unused rex prefixes;
output information about target of RIP relative addresses.
(putop): Support 'O' and 'I'. Update handling of "P', 'Q', 'R' and 'S';
(print_operand_value): New function.
(OP_E, OP_G, OP_REG, OP_I, OP_J, OP_DIR, OP_OFF, OP_D): Add support for
REX prefix and new modes.
(get64, get32s): New.
(get32): Return bfd_signed_vma type.
(set_op): Initialize the op_riprel.
* disassemble.c (disassembler): Recognize the x86-64 disassembly.
|
|
2001-01-03 Richard Sandiford <r.sandiford@redhat.com>
cgen-dis.in (read_insn): Use bfd_get_bits()
|
|
2001-01-02 Richard Sandiford <rsandifo@redhat.com>
* cgen-dis.c (hash_insn_array): Use bfd_put_bits().
(hash_insn_list): Likewise
* cgen-ibld.in (insert_1): Use bfd_put_bits() and bfd_get_bits().
(extract_1): Use bfd_get_bits().
(extract_normal): Apply sign extension to both extraction
methods.
* cgen-opc.c (cgen_get_insn_value): Use bfd_get_bits()
(cgen_put_insn_value): Use bfd_put_bits()
|
|
2000-12-28 Frank Ch. Eigler <fche@redhat.com>
* cgen-asm.in (parse_insn_normal): Print better error message for
instructions with missing operands.
|
|
|
|
|
|
2000-12-12 Frank Ch. Eigler <fche@redhat.com>
* cgen-asm.in (@arch@_cgen_assemble_insn): Prefer printing insert-time
error messages over later parse-time ones.
|
|
* elfxx-ia64.c (get_dyn_sym_info): Cast %p argument to void *.
* config/tc-ia64.h (ia64_init): Add prototype.
* gas/ia64/dv-imply.d, gas/ia64/dv-mutex.d, gas/ia64/dv-safe.d,
gas/ia64/dv-srlz.d, gas/ia64/opc-m.d: Update.
* ia64-dis.c (print_insn_ia64): Cast away const on ia64_free_opcode
argument.
* ia64_gen.c (insert_deplist): Cast sizeof result to int.
(print_dependency_table): Print NULL if semantics field not set.
(insert_opcode_dependencies): Mark cmp parameter as unused.
(print_main_table): Use fprintf_vma to print long long fields.
(main): Mark argv paramter as unused. Convert to old style definition.
* ia64-opc.c (ia64_find_dependency): Cast sizeof result to int.
* ia64-asmtab.c: Regnerate.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
(print_insn_ppi): Make nib1, nib2, nib3 unsigned.
Initialize variable dc to NULL.
(print_insn_shx): Remove unused label d_reg_n.
|
|
|
|
|
|
print_insn_cris_with_register_prefix.
|
|
|
|
|
|
* cgen-dis.in (print_insn): All insns which can fit into insn_value
must be loaded there in their entirety.
|
|
* config/tc-sparc.c (sparc_ip): Fix a bug which caused v9_arg_p
instructions to loose any special insn->architecture mask.
* config/tc-sparc.c (v9a_asr_table): Add v9b ASRs.
(sparc_md_end, sparc_arch_types, sparc_arch,
sparc_elf_final_processing): Handle v8plusb and v9b architectures.
(sparc_ip): Handle siam mode operands. Support v9b ASRs (and
request v9b architecture if they are used).
bfd/
* elf32-sparc.c (elf32_sparc_merge_private_bfd_data,
elf32_sparc_object_p, elf32_sparc_final_write_processing):
Support v8plusb.
* elf64-sparc.c (sparc64_elf_merge_private_bfd_data,
sparc64_elf_object_p): Support v9b.
* archures.c: Declare v8plusb and v9b machines.
* bfd-in2.h: Ditto.
* cpu-sparc.c: Ditto.
include/opcode/
* sparc.h (enum sparc_opcode_arch_val): Add SPARC_OPCODE_ARCH_V9B.
Note that '3' is used for siam operand.
opcodes/
* sparc-dis.c (v9a_asr_reg_names): Add v9b ASRs.
(compute_arch_mask): Add v8plusb and v9b machines.
(print_insn_sparc): siam mode decoding, accept ASRs up to 25.
* opcodes/sparc-opc.c: Support for Cheetah instruction set.
(prefetch_table): Add #invalidate.
|
|
|
|
* fr30-desc.h: Regenerate.
* m32r-desc.h: Regenerate.
* m32r-ibld.c: Regenerate.
|
|
* config/tc-ia64.c (resources_match): Handle IA64_RS_PRr.
* ia64-ic.tbl: Update from Intel.
* ia64-asmtab.c: Regenerate.
|
|
* ia64-gen.c: Convert C++-style comments to C-style comments.
* tic54x-dis.c: Likewise.
|
|
multi-language branch, to work around Solaris' /bin/sh bug. Rebuilt
all affected `configure' scripts.
|
|
|
|
don't have a leading underscore. Fix formatting.
* cris-dis.c (REGISTER_PREFIX_CHAR): New.
(format_reg): Add parameter with_reg_prefix. All callers changed.
(print_with_operands): Ditto.
(print_insn_cris_generic): Renamed from print_insn_cris, add
parameter with_reg_prefix.
(print_insn_cris_with_register_prefix,
print_insn_cris_without_register_prefix, cris_get_disassembler):
New.
* disassemble.c (disassembler) [ARCH_cris]: Call cris_get_disassembler.
|
|
* gas/ia64/opc-f.pl: Add missing fcmp and fpcmp tests.
* gas/ia64/opc-f.s, gas/ia64/opc-f.d: Regenerate.
* ia64-opc-f.c (ia64_opcodes_f): Add fpcmp pseudo-ops for
gt, ge, ngt, and nge.
* ia64-asmtab.c: Regenerate.
|
|
gas/ChangeLog
* config/tc-ia64.c (dv_sem): Add "stop".
(specify_resource, case IA64_RS_PR): Only handles regs 1 to 15 now.
(specify_resource, case IA64_RS_PRr): New for regs 16 to 62.
(specify_resource, case IA64_RS_PR63): Reorder (note == 7) test to
match above.
(mark_resources): Check IA64_RS_PRr.
gas/testsuite/ChangeLog
* gas/ia64/dv-raw-err.s: Add new testcases for PR%, 16 - 62.
* gas/ia64/dv-waw-err.s: Likewise.
* gas/ia64/dv-imply.d: Regenerate.
* gas/ia64/dv-mutex.d, gas/ia64/dv-raw-err.l, gas/ia64/dv-safe.d,
gas/ia64/dv-srlz.d, gas/ia64/dv-war-err.l, gas/ia64/dv-waw-err.l,
gas/ia64/opc-f.d, gas/ia64/opc-i.d, gas/ia64/opc-m.d: Likewise.
include/opcode/ChangeLog
* ia64.h (enum ia64_dependency_semantics): Add IA64_DVS_STOP.
opcodes/ChangeLog
* ia64-dis.c (print_insn_ia64): Revert Aug 7 byte skip count change.
* ia64-gen.c (parse_semantics): Handle IA64_DVS_STOP.
(lookup_specifier): Handle "PR%, 1 to 15" and "PR%, 16 to 62".
* ia64-ic.tbl, ia64-raw.tbl, ia64-war.tbl, ia64-waw.tbl: Update.
* ia64-asmtab.c: Regnerate.
|
|
|
|
* d30v-opc.c (d30v_operand_t): New operand type Rb2.
(d30v_format_tab): Use Rb2 for modinc and moddec.
|