| Message ID | 20180808123934.17450-2-alex.bennee@linaro.org |
|---|---|
| State | New |
| Headers | show |
| Series | add hand-rolled fallback when capstone fails | expand |
On Wed, Aug 08, 2018 at 01:39:31PM +0100, Alex Bennée wrote: > Given our issues with failing disassembly we could try and re-use the > decode tree data to output what instruction is being decoded. This > will be used if registered as a fall-back for when the "proper" > disassembler fails to decode an instruction. > > Signed-off-by: Alex Bennée <alex.bennee@linaro.org> I don't have an opinion on the approach you are taking, but the Python code you are adding is consistent with the existing style of the script. That said, I find the existing code full of output() calls very hard to read. If anybody wants to volunteer to improve the readability of the output generation, it would be welcome. Acked-by: Eduardo Habkost <ehabkost@redhat.com>
Eduardo Habkost <ehabkost@redhat.com> writes: > On Wed, Aug 08, 2018 at 01:39:31PM +0100, Alex Bennée wrote: >> Given our issues with failing disassembly we could try and re-use the >> decode tree data to output what instruction is being decoded. This >> will be used if registered as a fall-back for when the "proper" >> disassembler fails to decode an instruction. >> >> Signed-off-by: Alex Bennée <alex.bennee@linaro.org> > > I don't have an opinion on the approach you are taking, but the > Python code you are adding is consistent with the existing style > of the script. > > That said, I find the existing code full of output() calls very > hard to read. If anybody wants to volunteer to improve the > readability of the output generation, it would be welcome. If we expect to have the script output a number of different forms I guess re-factoring it with some sort of template based approach would be worthwhile. I wonder if there are other examples in the code base? We wouldn't want to have multiple template types. > > Acked-by: Eduardo Habkost <ehabkost@redhat.com> -- Alex Bennée
On Fri, Aug 10, 2018 at 09:55:50AM +0100, Alex Bennée wrote: > > Eduardo Habkost <ehabkost@redhat.com> writes: > > > On Wed, Aug 08, 2018 at 01:39:31PM +0100, Alex Bennée wrote: > >> Given our issues with failing disassembly we could try and re-use the > >> decode tree data to output what instruction is being decoded. This > >> will be used if registered as a fall-back for when the "proper" > >> disassembler fails to decode an instruction. > >> > >> Signed-off-by: Alex Bennée <alex.bennee@linaro.org> > > > > I don't have an opinion on the approach you are taking, but the > > Python code you are adding is consistent with the existing style > > of the script. > > > > That said, I find the existing code full of output() calls very > > hard to read. If anybody wants to volunteer to improve the > > readability of the output generation, it would be welcome. > > If we expect to have the script output a number of different forms I > guess re-factoring it with some sort of template based approach would be > worthwhile. I wonder if there are other examples in the code base? We > wouldn't want to have multiple template types. QAPI scripts also generates C code, and I find them more readable[1]. I think a true template language would be even better than QAPI's approach, but I don't see an obvious candidate that would be worth adding another build dependency to QEMU. [1] Compare: def output_decl(self): global translate_scope global translate_prefix output('typedef ', self.base.base.struct_name(), ' arg_', self.name, ';\n') output(translate_scope, 'bool ', translate_prefix, '_', self.name, '(DisasContext *ctx, arg_', self.name, ' *a, ', insntype, ' insn);\n') And: def gen_visit_members_decl(name): return mcgen(''' void visit_type_%(c_name)s_members(Visitor *v, %(c_name)s *obj, Error **errp); ''', c_name=c_name(name))
diff --git a/scripts/decodetree.py b/scripts/decodetree.py index 277f9a9bba..f4b4318c96 100755 --- a/scripts/decodetree.py +++ b/scripts/decodetree.py @@ -169,6 +169,7 @@ input_file = '' output_file = None output_fd = None insntype = 'uint32_t' +disassemble = False re_ident = '[a-zA-Z][a-zA-Z0-9_]*' @@ -467,6 +468,7 @@ class Pattern(General): def output_code(self, i, extracted, outerbits, outermask): global translate_prefix + global disassemble ind = str_indent(i) arg = self.base.base.name output(ind, '/* line ', str(self.lineno), ' */\n') @@ -474,8 +476,34 @@ class Pattern(General): output(ind, self.base.extract_name(), '(&u.f_', arg, ', insn);\n') for n, f in self.fields.items(): output(ind, 'u.f_', arg, '.', n, ' = ', f.str_extract(), ';\n') - output(ind, 'return ', translate_prefix, '_', self.name, - '(ctx, &u.f_', arg, ', insn);\n') + output(ind, 'return ', translate_prefix, '_', self.name) + if disassemble: + output("(dstr, maxd, ") + else: + output("(ctx, ") + + output('&u.f_', arg) + + if disassemble: + output(");\n") + else: + output(', insn);\n') + + def output_formatter(self): + global translate_prefix + arg = self.base.base.name + output('/* line ', str(self.lineno), ' */\n') + output('typedef ', self.base.base.struct_name(), + ' arg_', self.name, ';\n') + output(translate_scope, 'bool ', translate_prefix, '_', self.name, + '(char *ptr, size_t n, arg_', self.name, ' *a)\n') + output("{\n") + output(str_indent(4), 'snprintf(ptr, n, "', self.name) + # fill in arguments here + output('"); \n') + output(str_indent(4), "return true;\n") + output("}\n") + # end Pattern @@ -973,11 +1001,12 @@ def main(): global insnwidth global insntype global insnmask + global disassemble decode_function = 'decode' decode_scope = 'static ' - long_opts = ['decode=', 'translate=', 'output=', 'insnwidth='] + long_opts = ['decode=', 'translate=', 'output=', 'insnwidth=', 'disassemble'] try: (opts, args) = getopt.getopt(sys.argv[1:], 'o:w:', long_opts) except getopt.GetoptError as err: @@ -998,6 +1027,8 @@ def main(): insnmask = 0xffff elif insnwidth != 32: error(0, 'cannot handle insns of width', insnwidth) + elif o == '--disassemble': + disassemble = True else: assert False, 'unhandled option' @@ -1031,7 +1062,10 @@ def main(): if i.base.base != p.base.base: error(0, i.name, ' has conflicting argument sets') else: - i.output_decl() + if disassemble: + i.output_formatter() + else: + i.output_decl() out_pats[i.name] = i output('\n') @@ -1039,8 +1073,14 @@ def main(): f = formats[n] f.output_extract() - output(decode_scope, 'bool ', decode_function, - '(DisasContext *ctx, ', insntype, ' insn)\n{\n') + output(decode_scope, 'bool ', decode_function) + + if disassemble: + output("(char *dstr, size_t maxd, ") + else: + output('(DisasContext *ctx, ') + + output(insntype, ' insn)\n{\n') i4 = str_indent(4) output(i4, 'union {\n')
Given our issues with failing disassembly we could try and re-use the decode tree data to output what instruction is being decoded. This will be used if registered as a fall-back for when the "proper" disassembler fails to decode an instruction. Signed-off-by: Alex Bennée <alex.bennee@linaro.org> --- scripts/decodetree.py | 52 ++++++++++++++++++++++++++++++++++++++----- 1 file changed, 46 insertions(+), 6 deletions(-)