On 6/3/22 10:46, Patrick Palka wrote:
On Thu, 2 Jun 2022, Jason Merrill wrote:
On 6/2/22 15:57, Patrick Palka wrote:
On Thu, 2 Jun 2022, Jason Merrill wrote:
On 5/27/22 09:57, Patrick Palka wrote:
On Thu, 26 May 2022, Patrick Palka wrote:
On Thu, 26 May 2022, Jason Merrill wrote:
On 5/26/22 14:57, Patrick Palka wrote:
On Thu, 26 May 2022, Patrick Palka wrote:
Here we expect the calls to BaseClass::baseDevice resolve to the
second,
third and fourth overloads respectively in light of the
cv-qualifiers
of 'this' in each case. But ever since
r12-6075-g2decd2cabe5a4f,
the
calls incorrectly resolve to the first overload at instantiation
time.
This happens because the calls to BaseClass::baseDevice are all
deemed
non-dependent (ever since r7-755-g23cb72663051cd made us ignore
the
dependentness of 'this' when considering the dependence of a
non-static
memfn call), hence we end up checking the call ahead of time,
using
as
the object argument a dummy object of type BaseClass. Since
this
object
argument is cv-unqualified, the calls incoherently resolve to
the
first
overload of baseDevice. Before r12-6075, this incorrect result
would
just get silently discarded and we'd end up redoing OR at
instantiation
time using 'this' as the object argument. But after r12-6075,
we
now
reuse this incorrect result at instantiation time.
This patch fixes this by making finish_call_expr request from
maybe_dummy_object a cv-qualified object consistent with the
cv-quals of
'this'. That way, ahead of time OR using a dummy object will
give
us
the right answer and we could safely reuse it at instantiation
time.
NB: r7-755 is also the cause of the related issue PR105742. Not
sure
if there's a fix that could resolve both PRs at once..
Bootstrapped and regtested on x86_64-pc-linux-gnu, does this
look OK
for trunk/12?
PR c++/105637
gcc/cp/ChangeLog:
* semantics.cc (finish_call_expr): Pass a cv-qualified object
type to maybe_dummy_object that is consistent with the
cv-qualifiers of 'this' if available.
gcc/testsuite/ChangeLog:
* g++.dg/template/non-dependent23.C: New test.
---
gcc/cp/semantics.cc | 15
++++++++---
.../g++.dg/template/non-dependent23.C | 25
+++++++++++++++++++
2 files changed, 37 insertions(+), 3 deletions(-)
create mode 100644
gcc/testsuite/g++.dg/template/non-dependent23.C
diff --git a/gcc/cp/semantics.cc b/gcc/cp/semantics.cc
index cd7a2818feb..1d9348c6cf1 100644
--- a/gcc/cp/semantics.cc
+++ b/gcc/cp/semantics.cc
@@ -2802,16 +2802,25 @@ finish_call_expr (tree fn, vec<tree,
va_gc>
**args, bool disallow_virtual,
[class.access.base] says that we need to convert
'this' to B*
as
part of the access, so we pass 'B' to
maybe_dummy_object. */
+ tree object_type = BINFO_TYPE (BASELINK_ACCESS_BINFO
(fn));
if (DECL_MAYBE_IN_CHARGE_CONSTRUCTOR_P (get_first_fn
(fn)))
{
/* A constructor call always uses a dummy object.
(This
constructor
call which has the form A::A () is actually
invalid and
we are
going to reject it later in
build_new_method_call.) */
- object = build_dummy_object (BINFO_TYPE
(BASELINK_ACCESS_BINFO
(fn)));
+ object = build_dummy_object (object_type);
}
else
- object = maybe_dummy_object (BINFO_TYPE (BASELINK_ACCESS_BINFO
(fn)),
- NULL);
+ {
+ if (current_class_ref)
+ {
+ /* Make sure that if maybe_dummy_object gives us a dummy
object,
+ it'll have the same cv-quals as '*this'. */
+ int quals = cp_type_quals (TREE_TYPE
(current_class_ref));
+ object_type = cp_build_qualified_type (object_type,
quals);
+ }
+ object = maybe_dummy_object (object_type, NULL);
+ }
result = build_new_method_call (object, fn, args,
NULL_TREE,
(disallow_virtual
Drat, this fix doesn't interact well with 'this'-capturing
lambdas:
struct BaseClass {
void baseDevice(); // #1
void baseDevice() const = delete; // #2
};
template<class T>
struct TopClass : T {
void failsToCompile() {
[this] { BaseClass::baseDevice(); }();
}
};
template struct TopClass<BaseClass>;
Here after the fix, we'd incorrectly select the const #2 overload
at
template definition time because current_class_ref is the const
'this'
for the lambda rather than the non-const 'this' for TopClass.. I
suppose
we need something like current_nonlambda_class_type for getting at
the
innermost non-lambda 'this'?
Do you want maybe_resolve_dummy (ob, false)?
That sadly doesn't seem to work -- the object type is BaseClass which
is
not necessarily a base of the dependent TopClass<T>, so
resolvable_dummy_lambda returns NULL_TREE. I guess it would work at
instantiation time though.
Ah, what seems to work well is directly using lambda_expr_this_capture
instead of maybe_resolve_dummy. And we might as well handle this in
maybe_dummy_object for benefit of all callers. How does the following
look? Smoke tested with RUNTESTFLAGS="dg.exp=*.C", full bootstrap and
regtesting in progress.
-- >8 --
Subject: [PATCH] c++: cv-quals of dummy obj for non-dep memfn call
[PR105637]
In non-dependent23.C below we expect the BaseClass::baseDevice calls to
resolve to the second, third and fourth overloads respectively in light
of the cv-qualifiers of 'this' in each case. But ever since
r12-6075-g2decd2cabe5a4f, the calls incorrectly resolve to the first
overload at instantiation time.
This happens because the calls to BaseClass::baseDevice are all deemed
non-dependent (ever since r7-755-g23cb72663051cd made us ignore 'this'
dependence when considering the dependence of a non-static memfn call),
hence we end up checking the call ahead of time, using as the object
argument a dummy object of type BaseClass. Since this object argument
is cv-unqualified, the calls incoherently resolve to the first overload
of baseDevice. Before r12-6075, this incorrect result would just get
silently discarded and we'd end up redoing OR at instantiation time
using 'this' as the object argument. But after r12-6075, we now reuse
this incorrect result at instantiation time.
This patch fixes this by making maybe_dummy_object respect the cv-quals
of (the non-lambda) 'this' when returning a dummy object. Thus, ahead
of time OR using a dummy object will give us the right answer that is
consistent with the instantiation time answer.
An earlier version of this patch didn't handle 'this'-capturing lambdas
correctly, which caused us to mishandle lambda-this22.C below.
PR c++/105637
gcc/cp/ChangeLog:
* tree.cc (maybe_dummy_object): When returning a dummy
object, respect the cv-quals of 'this' if available.
gcc/testsuite/ChangeLog:
* g++.dg/cpp0x/lambda/lambda-this22.C: New test.
* g++.dg/template/non-dependent23.C: New test.
---
gcc/cp/tree.cc | 19 +++++++++++++-
.../g++.dg/cpp0x/lambda/lambda-this22.C | 20 +++++++++++++++
.../g++.dg/template/non-dependent23.C | 25
+++++++++++++++++++
3 files changed, 63 insertions(+), 1 deletion(-)
create mode 100644 gcc/testsuite/g++.dg/cpp0x/lambda/lambda-this22.C
create mode 100644 gcc/testsuite/g++.dg/template/non-dependent23.C
diff --git a/gcc/cp/tree.cc b/gcc/cp/tree.cc
index 09162795801..679bf05b721 100644
--- a/gcc/cp/tree.cc
+++ b/gcc/cp/tree.cc
@@ -4330,7 +4330,24 @@ maybe_dummy_object (tree type, tree* binfop)
(TREE_TYPE (current_class_ref), context)))
decl = current_class_ref;
else
- decl = build_dummy_object (context);
+ {
+ /* Return a dummy object whose cv-quals are consistent with (the
+ non-lambda) 'this' if available. */
+ if (current_class_ref)
+ {
+ int quals = 0;
+ if (current == current_class_type)
+ quals = cp_type_quals (TREE_TYPE (current_class_ref));
+ else if (lambda_function (current_class_type))
+ {
+ tree lambda = CLASSTYPE_LAMBDA_EXPR (current_class_type);
How about
else if (tree lambda = CLASSTYPE_LAMBDA_EXPR (current_class_type))
? OK with that change.
Unfortunately the lambda_function test is necessary to avoid crashing
on lambda-ice11.C; the test mirrors what resolvable_dummy_lambda does
ever since r207999 / r208028 to avoid the crash.
Hmm, how about adjusting lambda_expr_this_capture to avoid the crash?
I'm afraid I'm not sure how to do that :/ In particular for the case
where add_capture_p is nonzero and the given lambda lacks a
lambda_function. I suppose we can relax the assert in the !add_capture_p
case but that seems somewhat hacky.
I noticed that finish_this_expr, another user of lambda_expr_this_capture,
isn't guarded by resolvable_dummy_lambda. I believe it gets away with
this because it checks lambda-ness of TREE_TYPE (current_class_ref) instead
of current_class_type. Perhaps we should do the same in maybe_dummy_object?
This avoids the ICE in lambda-ice11.C without needing to check lambda_function,
and seems like a cleaner approach overall.
-- >8 --
Subject: [PATCH] c++: cv-quals of dummy obj for non-dep memfn call [PR105637]
PR c++/105637
gcc/cp/ChangeLog:
* tree.cc (maybe_dummy_object): When returning a dummy
object, respect the cv-quals of 'this' if available.
gcc/testsuite/ChangeLog:
* g++.dg/cpp0x/lambda/lambda-this22.C: New test.
* g++.dg/template/non-dependent23.C: New test.
---
gcc/cp/tree.cc | 31 ++++++++++++++-----
.../g++.dg/cpp0x/lambda/lambda-this22.C | 20 ++++++++++++
.../g++.dg/template/non-dependent23.C | 25 +++++++++++++++
3 files changed, 69 insertions(+), 7 deletions(-)
create mode 100644 gcc/testsuite/g++.dg/cpp0x/lambda/lambda-this22.C
create mode 100644 gcc/testsuite/g++.dg/template/non-dependent23.C
diff --git a/gcc/cp/tree.cc b/gcc/cp/tree.cc
index 2b9cb7e1c7b..183febffb5d 100644
--- a/gcc/cp/tree.cc
+++ b/gcc/cp/tree.cc
@@ -4319,15 +4319,32 @@ maybe_dummy_object (tree type, tree* binfop)
if (binfop)
*binfop = binfo;
- if (current_class_ref
- /* current_class_ref might not correspond to current_class_type if
- we're in tsubst_default_argument or a lambda-declarator; in either
- case, we want to use current_class_ref if it matches CONTEXT. */
- && (same_type_ignoring_top_level_qualifiers_p
- (TREE_TYPE (current_class_ref), context)))
+ /* current_class_ref might not correspond to current_class_type if
+ we're in tsubst_default_argument or a lambda-declarator; in either
+ case, we want to use current_class_ref if it matches CONTEXT. */
+ tree ctype = current_class_ref ? TREE_TYPE (current_class_ref) : NULL_TREE;
+ if (ctype
+ && same_type_ignoring_top_level_qualifiers_p (ctype, context))
decl = current_class_ref;
else
- decl = build_dummy_object (context);
+ {
+ /* Return a dummy object whose cv-quals are consistent with (the
+ non-lambda) 'this' if available. */
+ if (ctype)
+ {
+ int quals = 0;
+ if (LAMBDA_TYPE_P (ctype))
+ {
+ tree lambda = CLASSTYPE_LAMBDA_EXPR (ctype);
And just checking CLASSTYPE_LAMBDA_EXPR (ctype) still isn't enough?
+ if (tree cap = lambda_expr_this_capture (lambda, false))
+ quals = cp_type_quals (TREE_TYPE (TREE_TYPE (cap)));
+ }
+ else
+ quals = cp_type_quals (ctype);
+ context = cp_build_qualified_type (context, quals);
+ }
+ decl = build_dummy_object (context);
+ }
return decl;
}
diff --git a/gcc/testsuite/g++.dg/cpp0x/lambda/lambda-this22.C
b/gcc/testsuite/g++.dg/cpp0x/lambda/lambda-this22.C
new file mode 100644
index 00000000000..c9e512b1621
--- /dev/null
+++ b/gcc/testsuite/g++.dg/cpp0x/lambda/lambda-this22.C
@@ -0,0 +1,20 @@
+// PR c++/105637
+// { dg-do compile { target c++11 } }
+
+struct BaseClass {
+ void baseDevice(); // #1
+ void baseDevice() const = delete; // #2
+};
+
+template<class T>
+struct TopClass : T {
+ void failsToCompile() {
+ [this] { BaseClass::baseDevice(); }(); // should select #2, not #1
+ }
+
+ void failsToCompile() const {
+ [this] { BaseClass::baseDevice(); }(); // { dg-error "deleted" }
+ }
+};
+
+template struct TopClass<BaseClass>;
diff --git a/gcc/testsuite/g++.dg/template/non-dependent23.C
b/gcc/testsuite/g++.dg/template/non-dependent23.C
new file mode 100644
index 00000000000..ef95c591b75
--- /dev/null
+++ b/gcc/testsuite/g++.dg/template/non-dependent23.C
@@ -0,0 +1,25 @@
+// PR c++/105637
+
+struct BaseClass {
+ void baseDevice(); // #1
+ void baseDevice() const; // #2
+ void baseDevice() volatile; // #3
+ void baseDevice() const volatile; // #4
+};
+
+template<class T>
+struct TopClass : T {
+ void failsToCompile() const {
+ BaseClass::baseDevice(); // should select #2, not #1
+ }
+
+ void failsToCompile() volatile {
+ BaseClass::baseDevice(); // should select #3, not #1
+ }
+
+ void failsToCompile() const volatile {
+ BaseClass::baseDevice(); // should select #4, not #1
+ }
+};
+
+template struct TopClass<BaseClass>;
