On 2/26/22 19:55, Roger Sayle wrote:
This is a fix for PR c++/39751 which is an ICE-on-invalid regression in
the C++ parser after encountering the end of file. The one line change
is to check that the tokens cached in DECL_PENDING_INLINE_INFO haven't
been purged before processing them in cp_parser_late_parsing_for_member.
Alas in addition to the one line fix (and new test case), I've also
taken the opportunity to silence the -Wmissing-field-initializers
warnings compiling this source file, by replacing the " = { };" with
explicit calls to memset to initialize/reset structures.
This patch has been tested on x86_64-pc-linux-gnu with make bootstrap
and make -k check with no new (unexpected) failures. Ok for mainline?
2022-02-26 Roger Sayle <ro...@nextmovesoftware.com>
gcc/cp/ChangeLog
PR c++/39751
* parser.cc (cp_parser_late_parsing_for_member): Confirm the token
stream hasn't been purged before processing DECL_PENDING_INLINE.
This is OK.
(cp_parser_handle_statement_omp_attributes): Silence compilation
warnings using memset to initialize structure.
(cp_parser_late_parsing_omp_declare_simd): Likewise.
These are not; I don't want to uglify the code to satisfy a broken
warning. The warning is documented to allow this pattern:
Likewise, in C++ this option does not warn about the empty { }
initializer, for example:
struct s { int f, g, h; };
s x = { };
So all the instances you changed indicate a bug in the warning that
should be fixed rather than worked around.
Even if this were OK, it's completely unrelated to PR39751, so should
have been a separate patch.
+++ b/gcc/testsuite/g++.dg/pr39751.C
Should go in g++.dg/parse/
+/* { dg-options "-O2" } */
Unneeded.
Jason
