Re: [committed] epiphany: Fix up epiphany_setup_incoming_varargs [PR114175]

[Jeff Law]

On 3/20/24 10:05 AM, Jakub Jelinek wrote:

Hi!

Like for x86-64, alpha or rs6000, epiphany seems to be affected too.

Just visually checked differences in c23-stdarg-9.c assembly in a cross
without/with the patch, committed to trunk.

2024-03-20  Jakub Jelinek  

PR target/114175
* config/epiphany/epiphany.cc (epiphany_setup_incoming_varargs): Only
skip function arg advance for TYPE_NO_NAMED_ARGS_STDARG_P functions
if arg.type is NULL.
Built fine.  Platform has been too flakey to run tests with any 
consistency.  So no real test results.


jeff

[committed] epiphany: Fix up epiphany_setup_incoming_varargs [PR114175]

[Jakub Jelinek]

Hi!

Like for x86-64, alpha or rs6000, epiphany seems to be affected too.

Just visually checked differences in c23-stdarg-9.c assembly in a cross
without/with the patch, committed to trunk.

2024-03-20  Jakub Jelinek  

PR target/114175
* config/epiphany/epiphany.cc (epiphany_setup_incoming_varargs): Only
skip function arg advance for TYPE_NO_NAMED_ARGS_STDARG_P functions
if arg.type is NULL.

--- gcc/config/epiphany/epiphany.cc.jj  2024-01-03 11:51:47.401510931 +0100
+++ gcc/config/epiphany/epiphany.cc 2024-03-20 16:10:08.152226880 +0100
@@ -728,7 +728,8 @@ epiphany_setup_incoming_varargs (cumulat
 gcc_assert (arg.mode != BLKmode);
 
   next_cum = *get_cumulative_args (cum);
-  if (!TYPE_NO_NAMED_ARGS_STDARG_P (TREE_TYPE (current_function_decl)))
+  if (!TYPE_NO_NAMED_ARGS_STDARG_P (TREE_TYPE (current_function_decl))
+  || arg.type != NULL_TREE)
 next_cum = (ROUND_ADVANCE_CUM (next_cum, arg.mode, arg.type)
+ ROUND_ADVANCE_ARG (arg.mode, arg.type));
   first_anon_arg = next_cum;

Jakub