On 1/18/22 18:20, Jakub Jelinek wrote:
On Tue, Jan 18, 2022 at 05:56:47PM -0500, Jason Merrill wrote:
Debug information was getting confused because input_location was different
depending on whether we had looked ahead to see if the next tokens look like
a template argument list.
I tried resetting input_location in cp_lexer_rollback_tokens itself, but
that caused regressions, so let's just do it here for now.
Tested x86_64-pc-linux-gnu, applying to trunk.
PR c++/104025
gcc/cp/ChangeLog:
* parser.cc (saved_token_sentinel::rollback): Call
cp_lexer_set_source_position.
(~saved_token_sentinel): Call rollback.
gcc/testsuite/ChangeLog:
* g++.dg/warn/pr104025.C: New test.
The testcase fails for me with this change.
Oops, thanks. Fixed thus:
From bd0ef3534816a1b8ffad544d9ed720690a97d7cc Mon Sep 17 00:00:00 2001
From: Jason Merrill <ja...@redhat.com>
Date: Tue, 18 Jan 2022 22:11:56 -0500
Subject: [PATCH] c++: fix PR104025 change
To: gcc-patches@gcc.gnu.org
Somehow I pushed my earlier patch without it actually fixing the test; we
need input_location to be for the last consumed token, not the next one.
gcc/cp/ChangeLog:
* parser.cc (saved_token_sentinel::rollback): Use
cp_lexer_previous_token.
---
gcc/cp/parser.cc | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/gcc/cp/parser.cc b/gcc/cp/parser.cc
index 60d9f7bb723..00279c43404 100644
--- a/gcc/cp/parser.cc
+++ b/gcc/cp/parser.cc
@@ -1308,7 +1308,8 @@ struct saved_token_sentinel
void rollback ()
{
cp_lexer_rollback_tokens (lexer);
- cp_lexer_set_source_position (lexer);
+ cp_lexer_set_source_position_from_token
+ (cp_lexer_previous_token (lexer));
mode = STS_DONOTHING;
}
~saved_token_sentinel ()
--
2.27.0
