[patch] Slight update to patch for PR c++/27961

["Lee Millward" <lee dot millward at codesourcery dot com>]

Hi,

When looking through my submitted patches to check which ones had been
approved whilst catching up I saw the patch for PR27961 had been
approved so I applied it. Unfortunately, at the time I had totally
forgot that I withdrew the patch as it didn't fully fix the underlying
problem as pointed out by Volker Reichelt.

The attached patch fixes the new testcase by Volker by detecting
error_mark_nodes earlier in cp_parser_parenthesized_expression_list
and cp_parser_init_declarator.

I'm not entirely sure whether this would classify as a regression, the
testcase in PR27961 was listed as a regression from 4.1.1 and the
attached testcase is almost identical to the original.

Bootstrapped and regression tested on i686-pc-linux-gnu, ok for
mainline? Or next stage 1?

Cheers,
Lee.

cp/
2006-10-17  Lee Millward  <lee.millward@codesourcery.com>

       * parser.c (cp_parser_parenthesized_expression_list): Detect
       and propogate invalid expressions as error_mark_node sooner.
       (cp_parser_init_declarator): Likewise for initializations.

testsuite/
2006-10-17  Lee Millward  <lee.millward@codesourcery.com>

       * g++.dg/template/crash60: Add new test for functions
       returning a non-void value.

Attachment: pr27961.txt
Description: Text document