From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Cagney To: Jiri Smid Cc: gdb-patches@sources.redhat.com Subject: Misc; Was: [RFA]: x86_64 target files Date: Thu, 09 Aug 2001 23:57:00 -0000 Message-id: <3B738589.4050605@cygnus.com> References: X-SW-Source: 2001-08/msg00099.html Jiri, some quick comments on things that caught my eye. Once the *-tdep.c file is multi-arched (as much as possible) I'll go review the *-nat file properly. enjoy, Andrew > + /* This table must line up with REGISTER_NAMES in tm-x86_64.h > + symbols like 'RAX' come from . */ > + static int regmap[] = { > + RAX, RDX, RCX, RBX, > + RSI, RDI, RBP, RSP, > + R8, R9, R10, R11, > + R12, R13, R14, R15, > + RIP, EFLAGS > + }; You are encouraged to create a file like: x86_64-tdep.h so that this sort of thing can be shared amonst the various x86_64 specific files. Both *-nat and *-nat-linux contained that exact table. Just be careful to namespace proof things a little (x86_64 prefixes and the like). -- > + /* This is the variable that is set with "set disassembly-flavor", and > + its legitimate values. */ > + static const char att_flavor[] = "att"; > + static const char intel_flavor[] = "intel"; > + static const char *valid_flavors[] = { > + att_flavor, > + intel_flavor, > + NULL > + }; > + static const char *disassembly_flavor = att_flavor; > Have a look at the ARM which queries opcodes for a list of disassembly variants instead of duplicating those variants here. -- > + static unsigned char > + codestream_fill (int peek_flag) FYI, I don't think that the code stream stuff belongs in a *-tdep.c. Hmm, I suspect you accidently picked it up when clonning i386-tdep.c. Could I recommend going through your new *-tdep.c file and cleaning out anything that isn't being used. Andrew