[Bug c++/86769] g++ destroys condition variable in for statement too early
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86769 Andrew Pinski changed: What|Removed |Added CC||gnu4u at flonatel dot org --- Comment #7 from Andrew Pinski --- *** Bug 36016 has been marked as a duplicate of this bug. ***
[Bug c++/86769] g++ destroys condition variable in for statement too early
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86769 --- Comment #6 from Andrew Pinski --- Created attachment 57684 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=57684=edit Testcase from godbolt
[Bug c++/86769] g++ destroys condition variable in for statement too early
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86769
--- Comment #5 from Marek Polacek ---
Further poking revealed that the patch above mishandles
// PR c++/86769
// { dg-do run }
#define assert(X) do { if (!(X)) __builtin_abort(); } while(0)
int g;
struct X {
X() { g++; }
~X() { g--; }
operator bool() { return g == 0; }
};
void
check_live ()
{
assert (g > 0);
}
void
check_dead ()
{
assert (g == 0);
}
void f(X &) { assert (!g); }
int
main ()
{
for (int i = 0; i < 1; ++i, check_dead ())
{
X x = X();
check_live ();
}
}
So saving it here lest I lose it.
[Bug c++/86769] g++ destroys condition variable in for statement too early
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86769 Marek Polacek changed: What|Removed |Added Status|NEW |ASSIGNED Assignee|unassigned at gcc dot gnu.org |mpolacek at gcc dot gnu.org
[Bug c++/86769] g++ destroys condition variable in for statement too early
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86769
Marek Polacek changed:
What|Removed |Added
CC||mpolacek at gcc dot gnu.org
--- Comment #4 from Marek Polacek ---
Unfinished/untested patch:
commit f873acfa7ed1956a58d02cc383b8d709c446f656 (HEAD -> PR86769)
Author: Marek Polacek
Date: Fri Nov 20 16:40:50 2020 -0500
c++: Condition in for statement destroyed too early [PR86769]
diff --git a/gcc/c-family/c-gimplify.c b/gcc/c-family/c-gimplify.c
index a7c0ec3be0d..f0a19456715 100644
--- a/gcc/c-family/c-gimplify.c
+++ b/gcc/c-family/c-gimplify.c
@@ -205,6 +205,30 @@ expr_loc_or_loc (const_tree expr, location_t or_loc)
return loc;
}
+/* TODO */
+
+static bool
+maybe_inject_incr_to_body (tree body, tree incr)
+{
+ if (TREE_CODE (body) == BIND_EXPR)
+body = BIND_EXPR_BODY (body);
+ if (TREE_CODE (body) == STATEMENT_LIST)
+{
+ tree_stmt_iterator i = tsi_last (body);
+ tree t = tsi_stmt (i);
+ if (TREE_CODE (t) == TRY_FINALLY_EXPR)
+ {
+ if (TREE_CODE (TREE_OPERAND (t, 0)) == STATEMENT_LIST)
+ {
+ i = tsi_last (TREE_OPERAND (t, 0));
+ tsi_link_after (, incr, TSI_CONTINUE_LINKING);
+ return true;
+ }
+ }
+}
+ return false;
+}
+
/* Build a generic representation of one of the C loop forms. COND is the
loop condition or NULL_TREE. BODY is the (possibly compound) statement
controlled by the loop. INCR is the increment expression of a for-loop,
@@ -291,6 +315,9 @@ genericize_c_loop (tree *stmt_p, location_t start_locus,
tree cond, tree body,
append_to_statement_list (top, _list);
}
+ if (maybe_inject_incr_to_body (body, incr))
+incr = NULL_TREE;
+
append_to_statement_list (body, _list);
finish_bc_block (_list, bc_continue, clab);
if (incr)
[Bug c++/86769] g++ destroys condition variable in for statement too early
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86769 Arthur O'Dwyer changed: What|Removed |Added CC||arthur.j.odwyer at gmail dot com --- Comment #3 from Arthur O'Dwyer --- Confirmed. https://godbolt.org/z/MfbrcG
[Bug c++/86769] g++ destroys condition variable in for statement too early
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86769 Marek Polacek changed: What|Removed |Added CC||gcc-bugzilla at contacts dot eelis ||.net --- Comment #2 from Marek Polacek --- *** Bug 68003 has been marked as a duplicate of this bug. ***
[Bug c++/86769] g++ destroys condition variable in for statement too early
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86769 Jonathan Wakely changed: What|Removed |Added Keywords||wrong-code Status|UNCONFIRMED |NEW Last reconfirmed||2018-08-01 Ever confirmed|0 |1
[Bug c++/86769] g++ destroys condition variable in for statement too early
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86769 --- Comment #1 from Richard Smith --- For easy reproduction: https://godbolt.org/g/dfZFxz This does not appear to be a regression: this example appears to be miscompiled by every version of GCC back to at least 4.1.
