From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 24730 invoked by alias); 11 Apr 2003 05:43:08 -0000 Mailing-List: contact gdb-patches-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sources.redhat.com Received: (qmail 24711 invoked from network); 11 Apr 2003 05:43:05 -0000 Received: from unknown (HELO lacrosse.corp.redhat.com) (66.187.233.200) by sources.redhat.com with SMTP; 11 Apr 2003 05:43:05 -0000 Received: from free.redhat.lsd.ic.unicamp.br (aoliva.cipe.redhat.com [10.0.1.10]) by lacrosse.corp.redhat.com (8.11.6/8.9.3) with ESMTP id h3B5h3u07239; Fri, 11 Apr 2003 01:43:03 -0400 Received: from free.redhat.lsd.ic.unicamp.br (free.redhat.lsd.ic.unicamp.br [127.0.0.1]) by free.redhat.lsd.ic.unicamp.br (8.12.8/8.12.8) with ESMTP id h3B5h18i012301; Fri, 11 Apr 2003 02:43:02 -0300 Received: (from aoliva@localhost) by free.redhat.lsd.ic.unicamp.br (8.12.8/8.12.8/Submit) id h3B5h1uI012297; Fri, 11 Apr 2003 02:43:01 -0300 To: Andrew Cagney Cc: gdb-patches@sources.redhat.com, binutils@sources.redhat.com Subject: Re: use MIPS NewABI register names when disassembling NewABI code References: <3E927BCA.1080702@redhat.com> <3E93391F.2050005@redhat.com> <3E94339A.8030405@redhat.com> From: Alexandre Oliva Organization: GCC Team, Red Hat Date: Fri, 11 Apr 2003 05:43:00 -0000 In-Reply-To: <3E94339A.8030405@redhat.com> Message-ID: User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: multipart/mixed; boundary="=-=-=" X-SW-Source: 2003-04/txt/msg00227.txt.bz2 --=-=-= Content-length: 599 On Apr 9, 2003, Andrew Cagney wrote: > An architecture can't point at anything tied to the lifetime of an > instance of a BFD. tm_print_insn_info doesn't seem to be very tied to a gdbarch, anyway. To wit, note the bug fix in this revised patch. Explicitly setting the ABI to a non-default one, then back, then the non-default one again, would result the wrong tm_print_insn_info settings. This works now, and also disassembles in accordance with gdb's notion of the current ABI. Assuming tm_print_insn_info is actually the right thing to use, of course. Ok to install? --=-=-= Content-Type: text/x-patch Content-Disposition: inline; filename=mips-gdb-abi-disassemble.patch Content-length: 4930 Index: opcodes/ChangeLog from Alexandre Oliva * mips-dis.c (_print_insn_mips): Override reg_names from disassembler_options. Index: opcodes/mips-dis.c =================================================================== RCS file: /cvs/src/src/opcodes/mips-dis.c,v retrieving revision 1.43 diff -u -p -r1.43 mips-dis.c --- opcodes/mips-dis.c 8 Apr 2003 07:14:47 -0000 1.43 +++ opcodes/mips-dis.c 11 Apr 2003 05:39:20 -0000 @@ -1157,6 +1157,14 @@ _print_insn_mips (memaddr, info, endiann return print_insn_mips16 (memaddr, info); #endif + if (info->disassembler_options) + { + if (strcmp (info->disassembler_options, "NewABI") == 0) + reg_names = mips64_reg_names; + else + reg_names = mips32_reg_names; + } + status = (*info->read_memory_func) (memaddr, buffer, INSNLEN, info); if (status == 0) { Index: gdb/ChangeLog from Alexandre Oliva * Makefile.in (libbfd_h): Added missing setting. * mips-tdep.c (mips_gdbarch_init): Set disassembler_options according to the selected ABI. Index: gdb/Makefile.in =================================================================== RCS file: /cvs/src/src/gdb/Makefile.in,v retrieving revision 1.361 diff -u -p -r1.361 Makefile.in --- gdb/Makefile.in 10 Apr 2003 13:15:50 -0000 1.361 +++ gdb/Makefile.in 11 Apr 2003 05:39:25 -0000 @@ -574,6 +574,7 @@ elf_sh_h = $(INCLUDE_DIR)/elf/sh.h elf_arm_h = $(INCLUDE_DIR)/elf/arm.h $(elf_reloc_macros_h) elf_bfd_h = $(BFD_SRC)/elf-bfd.h libaout_h = $(BFD_SRC)/libaout.h +libbfd_h = $(BFD_SRC)/libbfd.h remote_sim_h = $(INCLUDE_DIR)/gdb/remote-sim.h demangle_h = $(INCLUDE_DIR)/demangle.h obstack_h = $(INCLUDE_DIR)/obstack.h Index: gdb/mips-tdep.c =================================================================== RCS file: /cvs/src/src/gdb/mips-tdep.c,v retrieving revision 1.187 diff -u -p -r1.187 mips-tdep.c --- gdb/mips-tdep.c 8 Apr 2003 19:21:15 -0000 1.187 +++ gdb/mips-tdep.c 11 Apr 2003 05:39:29 -0000 @@ -5731,6 +5731,31 @@ mips_gdbarch_init (struct gdbarch_info i if (wanted_abi != MIPS_ABI_UNKNOWN) mips_abi = wanted_abi; + /* We have to set tm_print_insn_info before looking for a + pre-existing architecture, otherwise we may return before we get + a chance to set it up. */ + if (mips_abi == MIPS_ABI_N32 || mips_abi == MIPS_ABI_N64) + { + /* Set up the disassembler info, so that we get the right + register names from libopcodes. */ + tm_print_insn_info.disassembler_options = "NewABI"; + tm_print_insn_info.flavour = bfd_target_elf_flavour; + tm_print_insn_info.arch = bfd_arch_mips; + if (info.bfd_arch_info != NULL + && info.bfd_arch_info->arch == bfd_arch_mips + && info.bfd_arch_info->mach) + tm_print_insn_info.mach = info.bfd_arch_info->mach; + else + tm_print_insn_info.mach = bfd_mach_mips8000; + } + else + /* This string is not recognized explicitly by the disassembler, + but it tells the disassembler to not try to guess the ABI from + the bfd elf headers, such that, if the user overrides the ABI + of a program linked as NewABI, the disassembly will follow the + register naming conventions specified by the user. */ + tm_print_insn_info.disassembler_options = "OldABI"; + if (gdbarch_debug) { fprintf_unfiltered (gdb_stdlog, @@ -5875,18 +5900,6 @@ mips_gdbarch_init (struct gdbarch_info i set_gdbarch_long_bit (gdbarch, 32); set_gdbarch_ptr_bit (gdbarch, 32); set_gdbarch_long_long_bit (gdbarch, 64); - - /* Set up the disassembler info, so that we get the right - register names from libopcodes. */ - tm_print_insn_info.flavour = bfd_target_elf_flavour; - tm_print_insn_info.arch = bfd_arch_mips; - if (info.bfd_arch_info != NULL - && info.bfd_arch_info->arch == bfd_arch_mips - && info.bfd_arch_info->mach) - tm_print_insn_info.mach = info.bfd_arch_info->mach; - else - tm_print_insn_info.mach = bfd_mach_mips8000; - set_gdbarch_use_struct_convention (gdbarch, mips_n32n64_use_struct_convention); set_gdbarch_reg_struct_has_addr (gdbarch, @@ -5906,18 +5919,6 @@ mips_gdbarch_init (struct gdbarch_info i set_gdbarch_long_bit (gdbarch, 64); set_gdbarch_ptr_bit (gdbarch, 64); set_gdbarch_long_long_bit (gdbarch, 64); - - /* Set up the disassembler info, so that we get the right - register names from libopcodes. */ - tm_print_insn_info.flavour = bfd_target_elf_flavour; - tm_print_insn_info.arch = bfd_arch_mips; - if (info.bfd_arch_info != NULL - && info.bfd_arch_info->arch == bfd_arch_mips - && info.bfd_arch_info->mach) - tm_print_insn_info.mach = info.bfd_arch_info->mach; - else - tm_print_insn_info.mach = bfd_mach_mips8000; - set_gdbarch_use_struct_convention (gdbarch, mips_n32n64_use_struct_convention); set_gdbarch_reg_struct_has_addr (gdbarch, --=-=-= Content-length: 289 -- Alexandre Oliva Enjoy Guarana', see http://www.ic.unicamp.br/~oliva/ Red Hat GCC Developer aoliva@{redhat.com, gcc.gnu.org} CS PhD student at IC-Unicamp oliva@{lsd.ic.unicamp.br, gnu.org} Free Software Evangelist Professional serial bug killer --=-=-=--