Re: [PATCH 1/8] Refactor disassembly code

[Yao Qi <qiyaoltc@gmail.com>]

On 17-01-11 15:42:58, Simon Marchi wrote:
> On 2017-01-10 07:26, Yao Qi wrote:
> >@@ -761,15 +768,15 @@ fprintf_disasm (void *stream, const char
> >*format, ...)
> >   return 0;
> > }
> >
> >-struct disassemble_info
> >-gdb_disassemble_info (struct gdbarch *gdbarch, struct ui_file *file)
> >+gdb_disassembler::gdb_disassembler (struct gdbarch *gdbarch,
> >+				    struct ui_file *file,
> >+				    di_read_memory_ftype func)
> 
> Perhaps name this parameter "read_memory_func"?
> 

OK.

> >diff --git a/gdb/disasm.h b/gdb/disasm.h
> >index 4c6fd54..5592cdb 100644
> >--- a/gdb/disasm.h
> >+++ b/gdb/disasm.h
> >@@ -33,6 +33,48 @@ struct gdbarch;
> > struct ui_out;
> > struct ui_file;
> >
> >+class gdb_disassembler
> >+{
> >+  using di_read_memory_ftype = decltype
> >(disassemble_info::read_memory_func);
> >+
> >+public:
> >+  gdb_disassembler (struct gdbarch *gdbarch, struct ui_file *file)
> >+    : gdb_disassembler (gdbarch, file, dis_asm_read_memory)
> >+  {}
> >+
> >+  int print_insn (CORE_ADDR memaddr);
> >+  int print_insn (CORE_ADDR memaddr, int *branch_delay_insns);
> 
> Not very important, but since print_insn(CORE_ADDR) is trivial, you
> could merge those two methods and provide a default parameter value
> of NULL for branch_delay_insns.

OK.  Fixed them locally.

> 
> >+
> >+  /* Prints the instruction INSN into UIOUT and returns the length of
> >+     the printed instruction in bytes.  */
> >+  int pretty_print_insn (struct ui_out *uiout,
> >+			 const struct disasm_insn *insn, int flags);
> 
> It could be a good time to modernize the flags parameter and make it
> an enum flag.
> 

This can be done later.

-- 
Yao (齐尧)