| Message ID | CAEwic4a4pcYYNVJ2E3A7wmcqy1-LGYoCa+oc6xOEVTvib9S+Xw@mail.gmail.com |
|---|---|
| State | New |
| Headers | show |
On 12/18/2014 10:10 AM, Kai Tietz wrote: > + if (TMPL_ARGS_DEPTH (args) > TMPL_PARMS_DEPTH (tmpl_parms)) > + args = get_innermost_template_args > + (args, TMPL_PARMS_DEPTH (tmpl_parms)); It seems unlikely to be correct to arbitrarily strip off outer template args. Where did the mismatch come from? Jason
2014-12-18 18:26 GMT+01:00 Jason Merrill <jason@redhat.com>: > On 12/18/2014 10:10 AM, Kai Tietz wrote: >> >> + if (TMPL_ARGS_DEPTH (args) > TMPL_PARMS_DEPTH (tmpl_parms)) >> + args = get_innermost_template_args >> + (args, TMPL_PARMS_DEPTH (tmpl_parms)); > > > It seems unlikely to be correct to arbitrarily strip off outer template > args. Where did the mismatch come from? > > Jason > Hmm, we are comming from #2 tsubst_decl (cp/pt.c:11278) #3 instantiate_template_1 (cp/pt.c:15945) #4 instantiate_template (cp/pt.c:15995) #5 instantiate_alias_template (cp/pt.c:11827) #6: lookup_template_class_1 #7: lookup_template_class #8: finish_template_type #9: cp_parser_template_id .... Well, in general I would have assumed to be able to get alias decl of tmpl. Wasn't able to find a simple way to get it. So, by looking into source I found that most cases handling args > tmpl-args by using inner_most_template_args instead. Defaulting to innermost_template_args looks indeed a bit wrong, but seemed to work. Kai
On 12/18/2014 01:16 PM, Kai Tietz wrote: > Well, in general I would have assumed to be able to get alias decl of > tmpl. Wasn't able to find a simple way to get it. So, by looking into > source I found that most cases handling args > tmpl-args by using > inner_most_template_args instead. Defaulting to > innermost_template_args looks indeed a bit wrong, but seemed to work. It's wrong, it just papers over the bug. There shouldn't be a mismatch at this point. The problem seems to be that most_general_template isn't actually returning the most general template, so we're trying to instantiate the partially-instantiated alias template with a full set of arguments: thus the mismatch. Jason
Index: pt.c =================================================================== --- pt.c (Revision 218832) +++ pt.c (Arbeitskopie) @@ -1033,6 +1033,8 @@ optimize_specialization_lookup_p (tree tmpl) static tree retrieve_specialization (tree tmpl, tree args, hashval_t hash) { + tree tmpl_parms; + if (tmpl == NULL_TREE) return NULL_TREE; @@ -1044,10 +1046,20 @@ retrieve_specialization (tree tmpl, tree args, has /* There should be as many levels of arguments as there are levels of parameters. */ - gcc_assert (TMPL_ARGS_DEPTH (args) - == (TREE_CODE (tmpl) == TEMPLATE_DECL - ? TMPL_PARMS_DEPTH (DECL_TEMPLATE_PARMS (tmpl)) - : template_class_depth (DECL_CONTEXT (tmpl)))); + if (TREE_CODE (tmpl) == TEMPLATE_DECL) + { + tmpl_parms = DECL_TEMPLATE_PARMS (tmpl); + if (TMPL_ARGS_DEPTH (args) > TMPL_PARMS_DEPTH (tmpl_parms)) + args = get_innermost_template_args + (args, TMPL_PARMS_DEPTH (tmpl_parms)); + gcc_assert (TMPL_ARGS_DEPTH (args) + == TMPL_PARMS_DEPTH (tmpl_parms)); + } + else + { + tmpl_parms = DECL_CONTEXT (tmpl); + gcc_assert (TMPL_ARGS_DEPTH (args) == template_class_depth (tmpl_parms)); + } if (optimize_specialization_lookup_p (tmpl)) {
Hi, this patch adds handling of aliases within templates and tries to resolve specialization for them. ChangeLog 2014-12-18 Kai Tietz <ktietz@redhat.com> PR c++/61198 * pt.c (retrieve_specialization): Handle using. Tested on x86_64-w64-mingw32. Ok for apply? Regards, Kai ChangeLog testsuite 2014-12-18 Kai Tietz <ktietz@redhat.com> PR c++/61198 * g++.dg/template/using29.C: New file. // { dg-do compile } template<int herp, typename derp_t> struct broken { template<typename target_t> using rebind = broken<herp, target_ }; template<typename derp_t> struct broken<2, derp_t> { template<typename target_t> using rebind = broken<2, target_t>; }; int main(int argc, char **argv) { broken<2, float>::rebind<double> u; return 0; }