[v3] c++/modules: Support lambdas attached to more places in modules [PR111710]

On Tue, Feb 27, 2024 at 11:59:46AM -0500, Patrick Palka wrote:
> On Fri, 16 Feb 2024, Nathaniel Shead wrote:
> 
> > On Tue, Feb 13, 2024 at 07:52:01PM -0500, Jason Merrill wrote:
> > > On 2/10/24 17:57, Nathaniel Shead wrote:
> > > > The fix for PR107398 weakened the restrictions that lambdas must belong
> > > > to namespace scope. However this was not sufficient: we also need to
> > > > allow lambdas keyed to FIELD_DECLs or PARM_DECLs.
> > > 
> > > I wonder about keying such lambdas to the class and function, respectively,
> > > rather than specifically to the field or parameter, but I suppose it doesn't
> > > matter.
> > 
> > I did some more testing and realised my testcase didn't properly
> > exercise whether I'd properly deduplicated or not, and an improved
> > testcase proved that actually keying to the field rather than the class
> > did cause issues. (Parameter vs. function doesn't seem to have mattered
> > however.)
> > 
> > Here's an updated patch that fixes this, and includes the changes for
> > lambdas in base classes that I'd had as a separate patch earlier. I've
> > also added some concepts testcases just in case.
> > 
> > Bootstrapped and regtested on x86_64-pc-linux-gnu, OK for trunk?
> > 
> > -- >8 --
> > 
> > The fix for PR107398 weakened the restrictions that lambdas must belong
> > to namespace scope. However this was not sufficient: we also need to
> > allow lambdas attached to FIELD_DECLs, PARM_DECLs, and TYPE_DECLs.
> > 
> > For field decls we key the lambda to its class rather than the field
> > itself. This avoids some errors with deduplicating fields.
> > 
> > Additionally, by [basic.link] p15.2 a lambda defined anywhere in a
> > class-specifier should not be TU-local, which includes base-class
> > declarations, so ensure that lambdas declared there are keyed
> > appropriately as well.
> > 
> > Because this now requires 'DECL_MODULE_KEYED_DECLS_P' to be checked on a
> > fairly large number of different kinds of DECLs, and that in general
> > it's safe to just get 'false' as a result of a check on an unexpected
> > DECL type, this patch also removes the tree checking from the accessor.
> > 
> > Finally, to handle deduplicating templated lambda fields, we need to
> > ensure that we can determine that two lambdas from different field decls
> > match. The modules code does not attempt to deduplicate expression
> > nodes, which causes issues as the LAMBDA_EXPRs are then considered to be
> > different. However, rather than checking the LAMBDA_EXPR directly we can
> > instead check its type: the generated RECORD_TYPE for a LAMBDA_EXPR must
> > also be unique, and /is/ deduplicated on import, so we can just check
> > for that instead.
> 
> We probably should be deduplicating LAMBDA_EXPR on stream-in, perhaps
> something like
> 
> diff --git a/gcc/cp/module.cc b/gcc/cp/module.cc
> index e8eabb1f6f9..1b2ba2e0fa8 100644
> --- a/gcc/cp/module.cc
> +++ b/gcc/cp/module.cc
> @@ -9183,6 +9183,13 @@ trees_in::tree_value ()
>        return NULL_TREE;
>      }
>  
> +  if (TREE_CODE (t) == LAMBDA_EXPR
> +      && CLASSTYPE_LAMBDA_EXPR (TREE_TYPE (t)))
> +    {
> +      existing = CLASSTYPE_LAMBDA_EXPR (TREE_TYPE (t));
> +      back_refs[~tag] = existing;
> +    }
> +
>    dump (dumper::TREE) && dump ("Read tree:%d %C:%N", tag, TREE_CODE (t), t);
>  
>    if (TREE_CODE (existing) == INTEGER_CST && !TREE_OVERFLOW (existing))
> 
> would suffice?  If not we probably need to take inspiration from the
> TREE_BINFO streaming, and handle LAMBDA_EXPR similarly..
> 

Ah yup, right, that makes more sense. Your suggestion seems to work,
thanks! Here's an updated patch.

Bootstrapped and regtested on x86_64-pc-linux-gnu, OK for trunk?

-- >8 --

The fix for PR107398 weakened the restrictions that lambdas must belong
to namespace scope. However this was not sufficient: we also need to
allow lambdas attached to FIELD_DECLs, PARM_DECLs, and TYPE_DECLs.

For field decls we key the lambda to its class rather than the field
itself. This avoids some errors with deduplicating fields.

Additionally, by [basic.link] p15.2 a lambda defined anywhere in a
class-specifier should not be TU-local, which includes base-class
declarations, so ensure that lambdas declared there are keyed
appropriately as well.

Because this now requires 'DECL_MODULE_KEYED_DECLS_P' to be checked on a
fairly large number of different kinds of DECLs, and that in general
it's safe to just get 'false' as a result of a check on an unexpected
DECL type, this patch also removes the tree checking from the accessor.

Finally, to handle deduplicating templated lambda fields, we need to
ensure that we can determine that two lambdas from different field decls
match, so we ensure that we deduplicate LAMBDA_EXPRs on stream in.

	PR c++/111710

gcc/cp/ChangeLog:

	* cp-tree.h (DECL_MODULE_KEYED_DECLS_P): Remove tree checking.
	(struct lang_decl_base): Update comments and fix whitespace.
	* module.cc (trees_out::lang_decl_bools): Always write
	module_keyed_decls_p flag...
	(trees_in::lang_decl_bools): ...and always read it.
	(trees_out::decl_value): Handle all kinds of keyed decls.
	(trees_in::decl_value): Likewise.
	(trees_in::tree_value): Deduplicate LAMBDA_EXPRs.
	(maybe_key_decl): Also support lambdas attached to fields,
	parameters, and types. Key lambdas attached to fields to their
	class.
	(trees_out::get_merge_kind): Likewise.
	(trees_out::key_mergeable): Likewise.
	(trees_in::key_mergeable): Support keyed decls in a TYPE_DECL
        container.
	* parser.cc (cp_parser_class_head): Start a lambda scope when
	parsing base classes.

gcc/testsuite/ChangeLog:

	* g++.dg/modules/lambda-7.h: New test.
	* g++.dg/modules/lambda-7_a.H: New test.
	* g++.dg/modules/lambda-7_b.C: New test.
	* g++.dg/modules/lambda-7_c.C: New test.

Signed-off-by: Nathaniel Shead <nathanieloshead@gmail.com>
Reviewed-by: Patrick Palka <ppalka@redhat.com>
---
 gcc/cp/cp-tree.h                          |  26 +++---
 gcc/cp/module.cc                          | 101 +++++++++++++---------
 gcc/cp/parser.cc                          |  10 ++-
 gcc/testsuite/g++.dg/modules/lambda-7.h   |  42 +++++++++
 gcc/testsuite/g++.dg/modules/lambda-7_a.H |   4 +
 gcc/testsuite/g++.dg/modules/lambda-7_b.C |   5 ++
 gcc/testsuite/g++.dg/modules/lambda-7_c.C |  41 +++++++++
 7 files changed, 174 insertions(+), 55 deletions(-)
 create mode 100644 gcc/testsuite/g++.dg/modules/lambda-7.h
 create mode 100644 gcc/testsuite/g++.dg/modules/lambda-7_a.H
 create mode 100644 gcc/testsuite/g++.dg/modules/lambda-7_b.C
 create mode 100644 gcc/testsuite/g++.dg/modules/lambda-7_c.C

On 2/27/24 23:12, Nathaniel Shead wrote:
> On Tue, Feb 27, 2024 at 11:59:46AM -0500, Patrick Palka wrote:
>> On Fri, 16 Feb 2024, Nathaniel Shead wrote:
>>
>>> On Tue, Feb 13, 2024 at 07:52:01PM -0500, Jason Merrill wrote:
>>>> On 2/10/24 17:57, Nathaniel Shead wrote:
>>>>> The fix for PR107398 weakened the restrictions that lambdas must belong
>>>>> to namespace scope. However this was not sufficient: we also need to
>>>>> allow lambdas keyed to FIELD_DECLs or PARM_DECLs.
>>>>
>>>> I wonder about keying such lambdas to the class and function, respectively,
>>>> rather than specifically to the field or parameter, but I suppose it doesn't
>>>> matter.
>>>
>>> I did some more testing and realised my testcase didn't properly
>>> exercise whether I'd properly deduplicated or not, and an improved
>>> testcase proved that actually keying to the field rather than the class
>>> did cause issues. (Parameter vs. function doesn't seem to have mattered
>>> however.)
>>>
>>> Here's an updated patch that fixes this, and includes the changes for
>>> lambdas in base classes that I'd had as a separate patch earlier. I've
>>> also added some concepts testcases just in case.
>>>
>>> Bootstrapped and regtested on x86_64-pc-linux-gnu, OK for trunk?
>>>
>>> -- >8 --
>>>
>>> The fix for PR107398 weakened the restrictions that lambdas must belong
>>> to namespace scope. However this was not sufficient: we also need to
>>> allow lambdas attached to FIELD_DECLs, PARM_DECLs, and TYPE_DECLs.
>>>
>>> For field decls we key the lambda to its class rather than the field
>>> itself. This avoids some errors with deduplicating fields.
>>>
>>> Additionally, by [basic.link] p15.2 a lambda defined anywhere in a
>>> class-specifier should not be TU-local, which includes base-class
>>> declarations, so ensure that lambdas declared there are keyed
>>> appropriately as well.
>>>
>>> Because this now requires 'DECL_MODULE_KEYED_DECLS_P' to be checked on a
>>> fairly large number of different kinds of DECLs, and that in general
>>> it's safe to just get 'false' as a result of a check on an unexpected
>>> DECL type, this patch also removes the tree checking from the accessor.
>>>
>>> Finally, to handle deduplicating templated lambda fields, we need to
>>> ensure that we can determine that two lambdas from different field decls
>>> match. The modules code does not attempt to deduplicate expression
>>> nodes, which causes issues as the LAMBDA_EXPRs are then considered to be
>>> different. However, rather than checking the LAMBDA_EXPR directly we can
>>> instead check its type: the generated RECORD_TYPE for a LAMBDA_EXPR must
>>> also be unique, and /is/ deduplicated on import, so we can just check
>>> for that instead.
>>
>> We probably should be deduplicating LAMBDA_EXPR on stream-in, perhaps
>> something like
>>
>> diff --git a/gcc/cp/module.cc b/gcc/cp/module.cc
>> index e8eabb1f6f9..1b2ba2e0fa8 100644
>> --- a/gcc/cp/module.cc
>> +++ b/gcc/cp/module.cc
>> @@ -9183,6 +9183,13 @@ trees_in::tree_value ()
>>         return NULL_TREE;
>>       }
>>   
>> +  if (TREE_CODE (t) == LAMBDA_EXPR
>> +      && CLASSTYPE_LAMBDA_EXPR (TREE_TYPE (t)))
>> +    {
>> +      existing = CLASSTYPE_LAMBDA_EXPR (TREE_TYPE (t));
>> +      back_refs[~tag] = existing;
>> +    }
>> +
>>     dump (dumper::TREE) && dump ("Read tree:%d %C:%N", tag, TREE_CODE (t), t);
>>   
>>     if (TREE_CODE (existing) == INTEGER_CST && !TREE_OVERFLOW (existing))
>>
>> would suffice?  If not we probably need to take inspiration from the
>> TREE_BINFO streaming, and handle LAMBDA_EXPR similarly..
>>
> 
> Ah yup, right, that makes more sense. Your suggestion seems to work,
> thanks! Here's an updated patch.
> 
> Bootstrapped and regtested on x86_64-pc-linux-gnu, OK for trunk?

With that change, do you still need to key to the class instead of the 
field for dedup to work properly?

OK either way.

> -- >8 --
> 
> The fix for PR107398 weakened the restrictions that lambdas must belong
> to namespace scope. However this was not sufficient: we also need to
> allow lambdas attached to FIELD_DECLs, PARM_DECLs, and TYPE_DECLs.
> 
> For field decls we key the lambda to its class rather than the field
> itself. This avoids some errors with deduplicating fields.
> 
> Additionally, by [basic.link] p15.2 a lambda defined anywhere in a
> class-specifier should not be TU-local, which includes base-class
> declarations, so ensure that lambdas declared there are keyed
> appropriately as well.
> 
> Because this now requires 'DECL_MODULE_KEYED_DECLS_P' to be checked on a
> fairly large number of different kinds of DECLs, and that in general
> it's safe to just get 'false' as a result of a check on an unexpected
> DECL type, this patch also removes the tree checking from the accessor.
> 
> Finally, to handle deduplicating templated lambda fields, we need to
> ensure that we can determine that two lambdas from different field decls
> match, so we ensure that we deduplicate LAMBDA_EXPRs on stream in.
> 
> 	PR c++/111710
> 
> gcc/cp/ChangeLog:
> 
> 	* cp-tree.h (DECL_MODULE_KEYED_DECLS_P): Remove tree checking.
> 	(struct lang_decl_base): Update comments and fix whitespace.
> 	* module.cc (trees_out::lang_decl_bools): Always write
> 	module_keyed_decls_p flag...
> 	(trees_in::lang_decl_bools): ...and always read it.
> 	(trees_out::decl_value): Handle all kinds of keyed decls.
> 	(trees_in::decl_value): Likewise.
> 	(trees_in::tree_value): Deduplicate LAMBDA_EXPRs.
> 	(maybe_key_decl): Also support lambdas attached to fields,
> 	parameters, and types. Key lambdas attached to fields to their
> 	class.
> 	(trees_out::get_merge_kind): Likewise.
> 	(trees_out::key_mergeable): Likewise.
> 	(trees_in::key_mergeable): Support keyed decls in a TYPE_DECL
>          container.
> 	* parser.cc (cp_parser_class_head): Start a lambda scope when
> 	parsing base classes.
> 
> gcc/testsuite/ChangeLog:
> 
> 	* g++.dg/modules/lambda-7.h: New test.
> 	* g++.dg/modules/lambda-7_a.H: New test.
> 	* g++.dg/modules/lambda-7_b.C: New test.
> 	* g++.dg/modules/lambda-7_c.C: New test.
> 
> Signed-off-by: Nathaniel Shead <nathanieloshead@gmail.com>
> Reviewed-by: Patrick Palka <ppalka@redhat.com>
> ---
>   gcc/cp/cp-tree.h                          |  26 +++---
>   gcc/cp/module.cc                          | 101 +++++++++++++---------
>   gcc/cp/parser.cc                          |  10 ++-
>   gcc/testsuite/g++.dg/modules/lambda-7.h   |  42 +++++++++
>   gcc/testsuite/g++.dg/modules/lambda-7_a.H |   4 +
>   gcc/testsuite/g++.dg/modules/lambda-7_b.C |   5 ++
>   gcc/testsuite/g++.dg/modules/lambda-7_c.C |  41 +++++++++
>   7 files changed, 174 insertions(+), 55 deletions(-)
>   create mode 100644 gcc/testsuite/g++.dg/modules/lambda-7.h
>   create mode 100644 gcc/testsuite/g++.dg/modules/lambda-7_a.H
>   create mode 100644 gcc/testsuite/g++.dg/modules/lambda-7_b.C
>   create mode 100644 gcc/testsuite/g++.dg/modules/lambda-7_c.C
> 
> diff --git a/gcc/cp/cp-tree.h b/gcc/cp/cp-tree.h
> index 334c11396c2..04c3aa6cd91 100644
> --- a/gcc/cp/cp-tree.h
> +++ b/gcc/cp/cp-tree.h
> @@ -1773,9 +1773,8 @@ check_constraint_info (tree t)
>     (DECL_LANG_SPECIFIC (DECL_MODULE_CHECK (NODE))->u.base.module_entity_p)
>   
>   /* DECL that has attached decls for ODR-relatedness.  */
> -#define DECL_MODULE_KEYED_DECLS_P(NODE)			\
> -  (DECL_LANG_SPECIFIC (TREE_CHECK2(NODE,FUNCTION_DECL,VAR_DECL))\
> -   ->u.base.module_keyed_decls_p)
> +#define DECL_MODULE_KEYED_DECLS_P(NODE) \
> +  (DECL_LANG_SPECIFIC (DECL_MODULE_CHECK (NODE))->u.base.module_keyed_decls_p)
>   
>   /* Whether this is an exported DECL.  Held on any decl that can appear
>      at namespace scope (function, var, type, template, const or
> @@ -2887,21 +2886,20 @@ struct GTY(()) lang_decl_base {
>     unsigned friend_or_tls : 1;		   /* var, fn, type or template */
>     unsigned unknown_bound_p : 1;		   /* var */
>     unsigned odr_used : 1;		   /* var or fn */
> -  unsigned concept_p : 1;                  /* applies to vars and functions */
> +  unsigned concept_p : 1;		   /* applies to vars and functions */
>     unsigned var_declared_inline_p : 1;	   /* var */
>     unsigned dependent_init_p : 1;	   /* var */
>   
> -  /* The following apply to VAR, FUNCTION, TYPE, CONCEPT, & NAMESPACE
> +  /* The following four apply to VAR, FUNCTION, TYPE, CONCEPT, & NAMESPACE
>        decls.  */
> -  unsigned module_purview_p : 1;	   // in named-module purview
> -  unsigned module_attach_p : 1;		   // attached to named module
> -  unsigned module_import_p : 1;     	   /* from an import */
> -  unsigned module_entity_p : 1;		   /* is in the entitity ary &
> -					      hash.  */
> -  /* VAR_DECL or FUNCTION_DECL has keyed decls.     */
> -  unsigned module_keyed_decls_p : 1;
> -
> -  /* 12 spare bits.  */
> +  unsigned module_purview_p : 1;	   /* in named-module purview */
> +  unsigned module_attach_p : 1;		   /* attached to named module */
> +  unsigned module_import_p : 1;		   /* from an import */
> +  unsigned module_entity_p : 1;		   /* is in the entitity ary & hash */
> +
> +  unsigned module_keyed_decls_p : 1;	   /* has keys, applies to all decls */
> +
> +  /* 11 spare bits.  */
>   };
>   
>   /* True for DECL codes which have template info and access.  */
> diff --git a/gcc/cp/module.cc b/gcc/cp/module.cc
> index 106af7bdb3e..1b2ba2e0fa8 100644
> --- a/gcc/cp/module.cc
> +++ b/gcc/cp/module.cc
> @@ -5664,8 +5664,7 @@ trees_out::lang_decl_bools (tree t)
>        want to mark them as in module purview.  */
>     WB (lang->u.base.module_purview_p && !header_module_p ());
>     WB (lang->u.base.module_attach_p);
> -  if (VAR_OR_FUNCTION_DECL_P (t))
> -    WB (lang->u.base.module_keyed_decls_p);
> +  WB (lang->u.base.module_keyed_decls_p);
>     switch (lang->u.base.selector)
>       {
>       default:
> @@ -5738,8 +5737,7 @@ trees_in::lang_decl_bools (tree t)
>     RB (lang->u.base.dependent_init_p);
>     RB (lang->u.base.module_purview_p);
>     RB (lang->u.base.module_attach_p);
> -  if (VAR_OR_FUNCTION_DECL_P (t))
> -    RB (lang->u.base.module_keyed_decls_p);
> +  RB (lang->u.base.module_keyed_decls_p);
>     switch (lang->u.base.selector)
>       {
>       default:
> @@ -7871,8 +7869,7 @@ trees_out::decl_value (tree decl, depset *dep)
>         install_entity (decl, dep);
>       }
>   
> -  if (VAR_OR_FUNCTION_DECL_P (inner)
> -      && DECL_LANG_SPECIFIC (inner)
> +  if (DECL_LANG_SPECIFIC (inner)
>         && DECL_MODULE_KEYED_DECLS_P (inner)
>         && !is_key_order ())
>       {
> @@ -8172,8 +8169,7 @@ trees_in::decl_value ()
>     bool installed = install_entity (existing);
>     bool is_new = existing == decl;
>   
> -  if (VAR_OR_FUNCTION_DECL_P (inner)
> -      && DECL_LANG_SPECIFIC (inner)
> +  if (DECL_LANG_SPECIFIC (inner)
>         && DECL_MODULE_KEYED_DECLS_P (inner))
>       {
>         /* Read and maybe install the attached entities.  */
> @@ -9187,6 +9183,13 @@ trees_in::tree_value ()
>         return NULL_TREE;
>       }
>   
> +  if (TREE_CODE (t) == LAMBDA_EXPR
> +      && CLASSTYPE_LAMBDA_EXPR (TREE_TYPE (t)))
> +    {
> +      existing = CLASSTYPE_LAMBDA_EXPR (TREE_TYPE (t));
> +      back_refs[~tag] = existing;
> +    }
> +
>     dump (dumper::TREE) && dump ("Read tree:%d %C:%N", tag, TREE_CODE (t), t);
>   
>     if (TREE_CODE (existing) == INTEGER_CST && !TREE_OVERFLOW (existing))
> @@ -10484,12 +10487,17 @@ trees_out::get_merge_kind (tree decl, depset *dep)
>   	      if (tree scope
>   		  = LAMBDA_EXPR_EXTRA_SCOPE (CLASSTYPE_LAMBDA_EXPR
>   					     (TREE_TYPE (decl))))
> -		if (TREE_CODE (scope) == VAR_DECL
> -		    && DECL_MODULE_KEYED_DECLS_P (scope))
> -		  {
> -		    mk = MK_keyed;
> -		    break;
> -		  }
> +		{
> +		  /* Lambdas attached to fields are keyed to its class.  */
> +		  if (TREE_CODE (scope) == FIELD_DECL)
> +		    scope = TYPE_NAME (DECL_CONTEXT (scope));
> +		  if (DECL_LANG_SPECIFIC (scope)
> +		      && DECL_MODULE_KEYED_DECLS_P (scope))
> +		    {
> +		      mk = MK_keyed;
> +		      break;
> +		    }
> +		}
>   
>   	    if (RECORD_OR_UNION_TYPE_P (ctx))
>   	      {
> @@ -10789,7 +10797,13 @@ trees_out::key_mergeable (int tag, merge_kind mk, tree decl, tree inner,
>   	    gcc_checking_assert (LAMBDA_TYPE_P (TREE_TYPE (inner)));
>   	    tree scope = LAMBDA_EXPR_EXTRA_SCOPE (CLASSTYPE_LAMBDA_EXPR
>   						  (TREE_TYPE (inner)));
> -	    gcc_checking_assert (TREE_CODE (scope) == VAR_DECL);
> +	    gcc_checking_assert (TREE_CODE (scope) == VAR_DECL
> +				 || TREE_CODE (scope) == FIELD_DECL
> +				 || TREE_CODE (scope) == PARM_DECL
> +				 || TREE_CODE (scope) == TYPE_DECL);
> +	    /* Lambdas attached to fields are keyed to the class.  */
> +	    if (TREE_CODE (scope) == FIELD_DECL)
> +	      scope = TYPE_NAME (DECL_CONTEXT (scope));
>   	    auto *root = keyed_table->get (scope);
>   	    unsigned ix = root->length ();
>   	    /* If we don't find it, we'll write a really big number
> @@ -11067,6 +11081,26 @@ trees_in::key_mergeable (int tag, merge_kind mk, tree decl, tree inner,
>   		}
>   	    }
>   	}
> +      else if (mk == MK_keyed
> +	       && DECL_LANG_SPECIFIC (name)
> +	       && DECL_MODULE_KEYED_DECLS_P (name))
> +	{
> +	  gcc_checking_assert (TREE_CODE (container) == NAMESPACE_DECL
> +			       || TREE_CODE (container) == TYPE_DECL);
> +	  if (auto *set = keyed_table->get (name))
> +	    if (key.index < set->length ())
> +	      {
> +		existing = (*set)[key.index];
> +		if (existing)
> +		  {
> +		    gcc_checking_assert
> +		      (DECL_IMPLICIT_TYPEDEF_P (existing));
> +		    if (inner != decl)
> +		      existing
> +			= CLASSTYPE_TI_TEMPLATE (TREE_TYPE (existing));
> +		  }
> +	      }
> +	}
>         else
>   	switch (TREE_CODE (container))
>   	  {
> @@ -11074,27 +11108,8 @@ trees_in::key_mergeable (int tag, merge_kind mk, tree decl, tree inner,
>   	    gcc_unreachable ();
>   
>   	  case NAMESPACE_DECL:
> -	    if (mk == MK_keyed)
> -	      {
> -		if (DECL_LANG_SPECIFIC (name)
> -		    && VAR_OR_FUNCTION_DECL_P (name)
> -		    && DECL_MODULE_KEYED_DECLS_P (name))
> -		  if (auto *set = keyed_table->get (name))
> -		    if (key.index < set->length ())
> -		      {
> -			existing = (*set)[key.index];
> -			if (existing)
> -			  {
> -			    gcc_checking_assert
> -			      (DECL_IMPLICIT_TYPEDEF_P (existing));
> -			    if (inner != decl)
> -			      existing
> -				= CLASSTYPE_TI_TEMPLATE (TREE_TYPE (existing));
> -			  }
> -		      }
> -	      }
> -	    else if (is_attached
> -		     && !(state->is_module () || state->is_partition ()))
> +	    if (is_attached
> +		&& !(state->is_module () || state->is_partition ()))
>   	      kind = "unique";
>   	    else
>   	      {
> @@ -18984,11 +18999,19 @@ maybe_key_decl (tree ctx, tree decl)
>     if (!modules_p ())
>       return;
>   
> -  // FIXME: For now just deal with lambdas attached to var decls.
> -  // This might be sufficient?
> -  if (TREE_CODE (ctx) != VAR_DECL)
> +  /* We only need to deal with lambdas attached to var, field,
> +     parm, or type decls.  */
> +  if (TREE_CODE (ctx) != VAR_DECL
> +      && TREE_CODE (ctx) != FIELD_DECL
> +      && TREE_CODE (ctx) != PARM_DECL
> +      && TREE_CODE (ctx) != TYPE_DECL)
>       return;
>   
> +  /* For fields, key it to the containing type to handle deduplication
> +     correctly.  */
> +  if (TREE_CODE (ctx) == FIELD_DECL)
> +    ctx = TYPE_NAME (DECL_CONTEXT (ctx));
> +
>     if (!keyed_table)
>       keyed_table = new keyed_map_t (EXPERIMENT (1, 400));
>   
> diff --git a/gcc/cp/parser.cc b/gcc/cp/parser.cc
> index b2ed2baa3a5..3ee9d49fb8e 100644
> --- a/gcc/cp/parser.cc
> +++ b/gcc/cp/parser.cc
> @@ -27678,10 +27678,16 @@ cp_parser_class_head (cp_parser* parser,
>     if (cp_lexer_next_token_is (parser->lexer, CPP_COLON))
>       {
>         if (type)
> -	pushclass (type);
> +	{
> +	  pushclass (type);
> +	  start_lambda_scope (TYPE_NAME (type));
> +	}
>         bases = cp_parser_base_clause (parser);
>         if (type)
> -	popclass ();
> +	{
> +	  finish_lambda_scope ();
> +	  popclass ();
> +	}
>       }
>     else
>       bases = NULL_TREE;
> diff --git a/gcc/testsuite/g++.dg/modules/lambda-7.h b/gcc/testsuite/g++.dg/modules/lambda-7.h
> new file mode 100644
> index 00000000000..6f6080c1324
> --- /dev/null
> +++ b/gcc/testsuite/g++.dg/modules/lambda-7.h
> @@ -0,0 +1,42 @@
> +struct S {
> +  int (*a)(int) = [](int x) { return x * 2; };
> +
> +  int b(int x, int (*f)(int) = [](int x) { return x * 3; }) {
> +    return f(x);
> +  }
> +
> +  static int c(int x, int (*f)(int) = [](int x) { return x * 4; }) {
> +    return f(x);
> +  }
> +};
> +
> +inline int d(int x, int (*f)(int) = [](int x) { return x * 5; }) {
> +  return f(x);
> +}
> +
> +// unevaluated lambdas
> +#if __cplusplus >= 202002L
> +struct E : decltype([](int x) { return x * 6; }) {
> +  decltype([](int x) { return x * 7; }) f;
> +};
> +
> +template <typename T>
> +struct G : decltype([](int x) { return x * 8; }) {
> +  decltype([](int x) { return x * 9; }) h;
> +};
> +
> +template <>
> +struct G<double> : decltype([](int x) { return x * 10; }) {
> +  decltype([](int x) { return x * 11; }) i;
> +};
> +#endif
> +
> +// concepts
> +#if __cpp_concepts >= 201907L
> +template <typename T>
> +concept J = requires { []{ T(); }; };
> +
> +template <typename T>
> +concept K = []{ return sizeof(T) == 1; }();
> +#endif
> +
> diff --git a/gcc/testsuite/g++.dg/modules/lambda-7_a.H b/gcc/testsuite/g++.dg/modules/lambda-7_a.H
> new file mode 100644
> index 00000000000..5197114f76c
> --- /dev/null
> +++ b/gcc/testsuite/g++.dg/modules/lambda-7_a.H
> @@ -0,0 +1,4 @@
> +// { dg-additional-options "-fmodule-header -Wno-subobject-linkage" }
> +// { dg-module-cmi {} }
> +
> +#include "lambda-7.h"
> diff --git a/gcc/testsuite/g++.dg/modules/lambda-7_b.C b/gcc/testsuite/g++.dg/modules/lambda-7_b.C
> new file mode 100644
> index 00000000000..2d781e93067
> --- /dev/null
> +++ b/gcc/testsuite/g++.dg/modules/lambda-7_b.C
> @@ -0,0 +1,5 @@
> +// { dg-additional-options "-fmodules-ts -fno-module-lazy -Wno-subobject-linkage" }
> +// Test for ODR deduplication
> +
> +#include "lambda-7.h"
> +import "lambda-7_a.H";
> diff --git a/gcc/testsuite/g++.dg/modules/lambda-7_c.C b/gcc/testsuite/g++.dg/modules/lambda-7_c.C
> new file mode 100644
> index 00000000000..f283681fa96
> --- /dev/null
> +++ b/gcc/testsuite/g++.dg/modules/lambda-7_c.C
> @@ -0,0 +1,41 @@
> +// { dg-module-do run }
> +// { dg-additional-options "-fmodules-ts -fno-module-lazy -Wno-subobject-linkage" }
> +
> +import "lambda-7_a.H";
> +
> +int main() {
> +  S s;
> +  if (s.a(10) != 20)
> +    __builtin_abort();
> +  if (s.b(10) != 30)
> +    __builtin_abort();
> +  if (s.c(10) != 40)
> +    __builtin_abort();
> +  if (d(10) != 50)
> +    __builtin_abort();
> +
> +#if __cplusplus >= 202002L
> +  E e;
> +  if (e(10) != 60)
> +    __builtin_abort();
> +  if (e.f(10) != 70)
> +    __builtin_abort();
> +
> +  G<int> g1;
> +  if (g1(10) != 80)
> +    __builtin_abort();
> +  if (g1.h(10) != 90)
> +    __builtin_abort();
> +
> +  G<double> g2;
> +  if (g2(10) != 100)
> +    __builtin_abort();
> +  if (g2.i(10) != 110)
> +    __builtin_abort();
> +#endif
> +
> +#if __cpp_concepts >= 201907L
> +  static_assert(J<char>);
> +  static_assert(K<char>);
> +#endif
> +}

On Wed, Feb 28, 2024 at 12:34:51PM -0500, Jason Merrill wrote:
> On 2/27/24 23:12, Nathaniel Shead wrote:
> > On Tue, Feb 27, 2024 at 11:59:46AM -0500, Patrick Palka wrote:
> > > On Fri, 16 Feb 2024, Nathaniel Shead wrote:
> > > 
> > > > On Tue, Feb 13, 2024 at 07:52:01PM -0500, Jason Merrill wrote:
> > > > > On 2/10/24 17:57, Nathaniel Shead wrote:
> > > > > > The fix for PR107398 weakened the restrictions that lambdas must belong
> > > > > > to namespace scope. However this was not sufficient: we also need to
> > > > > > allow lambdas keyed to FIELD_DECLs or PARM_DECLs.
> > > > > 
> > > > > I wonder about keying such lambdas to the class and function, respectively,
> > > > > rather than specifically to the field or parameter, but I suppose it doesn't
> > > > > matter.
> > > > 
> > > > I did some more testing and realised my testcase didn't properly
> > > > exercise whether I'd properly deduplicated or not, and an improved
> > > > testcase proved that actually keying to the field rather than the class
> > > > did cause issues. (Parameter vs. function doesn't seem to have mattered
> > > > however.)
> > > > 
> > > > Here's an updated patch that fixes this, and includes the changes for
> > > > lambdas in base classes that I'd had as a separate patch earlier. I've
> > > > also added some concepts testcases just in case.
> > > > 
> > > > Bootstrapped and regtested on x86_64-pc-linux-gnu, OK for trunk?
> > > > 
> > > > -- >8 --
> > > > 
> > > > The fix for PR107398 weakened the restrictions that lambdas must belong
> > > > to namespace scope. However this was not sufficient: we also need to
> > > > allow lambdas attached to FIELD_DECLs, PARM_DECLs, and TYPE_DECLs.
> > > > 
> > > > For field decls we key the lambda to its class rather than the field
> > > > itself. This avoids some errors with deduplicating fields.
> > > > 
> > > > Additionally, by [basic.link] p15.2 a lambda defined anywhere in a
> > > > class-specifier should not be TU-local, which includes base-class
> > > > declarations, so ensure that lambdas declared there are keyed
> > > > appropriately as well.
> > > > 
> > > > Because this now requires 'DECL_MODULE_KEYED_DECLS_P' to be checked on a
> > > > fairly large number of different kinds of DECLs, and that in general
> > > > it's safe to just get 'false' as a result of a check on an unexpected
> > > > DECL type, this patch also removes the tree checking from the accessor.
> > > > 
> > > > Finally, to handle deduplicating templated lambda fields, we need to
> > > > ensure that we can determine that two lambdas from different field decls
> > > > match. The modules code does not attempt to deduplicate expression
> > > > nodes, which causes issues as the LAMBDA_EXPRs are then considered to be
> > > > different. However, rather than checking the LAMBDA_EXPR directly we can
> > > > instead check its type: the generated RECORD_TYPE for a LAMBDA_EXPR must
> > > > also be unique, and /is/ deduplicated on import, so we can just check
> > > > for that instead.
> > > 
> > > We probably should be deduplicating LAMBDA_EXPR on stream-in, perhaps
> > > something like
> > > 
> > > diff --git a/gcc/cp/module.cc b/gcc/cp/module.cc
> > > index e8eabb1f6f9..1b2ba2e0fa8 100644
> > > --- a/gcc/cp/module.cc
> > > +++ b/gcc/cp/module.cc
> > > @@ -9183,6 +9183,13 @@ trees_in::tree_value ()
> > >         return NULL_TREE;
> > >       }
> > > +  if (TREE_CODE (t) == LAMBDA_EXPR
> > > +      && CLASSTYPE_LAMBDA_EXPR (TREE_TYPE (t)))
> > > +    {
> > > +      existing = CLASSTYPE_LAMBDA_EXPR (TREE_TYPE (t));
> > > +      back_refs[~tag] = existing;
> > > +    }
> > > +
> > >     dump (dumper::TREE) && dump ("Read tree:%d %C:%N", tag, TREE_CODE (t), t);
> > >     if (TREE_CODE (existing) == INTEGER_CST && !TREE_OVERFLOW (existing))
> > > 
> > > would suffice?  If not we probably need to take inspiration from the
> > > TREE_BINFO streaming, and handle LAMBDA_EXPR similarly..
> > > 
> > 
> > Ah yup, right, that makes more sense. Your suggestion seems to work,
> > thanks! Here's an updated patch.
> > 
> > Bootstrapped and regtested on x86_64-pc-linux-gnu, OK for trunk?
> 
> With that change, do you still need to key to the class instead of the field
> for dedup to work properly?
> 
> OK either way.
> 

Thanks. And yes, we still need to key to the class rather than the
field; it looks like otherwise what happens is that when reading the
lambda's TYPE_DECL for a lambda used as an NSDMI, 'key_mergeable' ends
up loading the FIELD_DECL it's attached to (and attempting to install
its keyed declarations) when reading its 'name' before it gets a chance
to decide that the TYPE_DECL has an existing definition and deduplicate,
which causes 'set_overrun' to be called due to apparent mismatch.

By keying to the class instead of the field this isn't an issue because
the keyed declarations of the containing type are handled after the
nested TYPE_DECLs are deduplicated.

I'll added a comment to this effect to the commit message.

> > -- >8 --
> > 
> > The fix for PR107398 weakened the restrictions that lambdas must belong
> > to namespace scope. However this was not sufficient: we also need to
> > allow lambdas attached to FIELD_DECLs, PARM_DECLs, and TYPE_DECLs.
> > 
> > For field decls we key the lambda to its class rather than the field
> > itself. This avoids some errors with deduplicating fields.
> > 
> > Additionally, by [basic.link] p15.2 a lambda defined anywhere in a
> > class-specifier should not be TU-local, which includes base-class
> > declarations, so ensure that lambdas declared there are keyed
> > appropriately as well.
> > 
> > Because this now requires 'DECL_MODULE_KEYED_DECLS_P' to be checked on a
> > fairly large number of different kinds of DECLs, and that in general
> > it's safe to just get 'false' as a result of a check on an unexpected
> > DECL type, this patch also removes the tree checking from the accessor.
> > 
> > Finally, to handle deduplicating templated lambda fields, we need to
> > ensure that we can determine that two lambdas from different field decls
> > match, so we ensure that we deduplicate LAMBDA_EXPRs on stream in.
> > 
> > 	PR c++/111710
> > 
> > gcc/cp/ChangeLog:
> > 
> > 	* cp-tree.h (DECL_MODULE_KEYED_DECLS_P): Remove tree checking.
> > 	(struct lang_decl_base): Update comments and fix whitespace.
> > 	* module.cc (trees_out::lang_decl_bools): Always write
> > 	module_keyed_decls_p flag...
> > 	(trees_in::lang_decl_bools): ...and always read it.
> > 	(trees_out::decl_value): Handle all kinds of keyed decls.
> > 	(trees_in::decl_value): Likewise.
> > 	(trees_in::tree_value): Deduplicate LAMBDA_EXPRs.
> > 	(maybe_key_decl): Also support lambdas attached to fields,
> > 	parameters, and types. Key lambdas attached to fields to their
> > 	class.
> > 	(trees_out::get_merge_kind): Likewise.
> > 	(trees_out::key_mergeable): Likewise.
> > 	(trees_in::key_mergeable): Support keyed decls in a TYPE_DECL
> >          container.
> > 	* parser.cc (cp_parser_class_head): Start a lambda scope when
> > 	parsing base classes.
> > 
> > gcc/testsuite/ChangeLog:
> > 
> > 	* g++.dg/modules/lambda-7.h: New test.
> > 	* g++.dg/modules/lambda-7_a.H: New test.
> > 	* g++.dg/modules/lambda-7_b.C: New test.
> > 	* g++.dg/modules/lambda-7_c.C: New test.
> > 
> > Signed-off-by: Nathaniel Shead <nathanieloshead@gmail.com>
> > Reviewed-by: Patrick Palka <ppalka@redhat.com>
> > ---
> >   gcc/cp/cp-tree.h                          |  26 +++---
> >   gcc/cp/module.cc                          | 101 +++++++++++++---------
> >   gcc/cp/parser.cc                          |  10 ++-
> >   gcc/testsuite/g++.dg/modules/lambda-7.h   |  42 +++++++++
> >   gcc/testsuite/g++.dg/modules/lambda-7_a.H |   4 +
> >   gcc/testsuite/g++.dg/modules/lambda-7_b.C |   5 ++
> >   gcc/testsuite/g++.dg/modules/lambda-7_c.C |  41 +++++++++
> >   7 files changed, 174 insertions(+), 55 deletions(-)
> >   create mode 100644 gcc/testsuite/g++.dg/modules/lambda-7.h
> >   create mode 100644 gcc/testsuite/g++.dg/modules/lambda-7_a.H
> >   create mode 100644 gcc/testsuite/g++.dg/modules/lambda-7_b.C
> >   create mode 100644 gcc/testsuite/g++.dg/modules/lambda-7_c.C
> > 
> > diff --git a/gcc/cp/cp-tree.h b/gcc/cp/cp-tree.h
> > index 334c11396c2..04c3aa6cd91 100644
> > --- a/gcc/cp/cp-tree.h
> > +++ b/gcc/cp/cp-tree.h
> > @@ -1773,9 +1773,8 @@ check_constraint_info (tree t)
> >     (DECL_LANG_SPECIFIC (DECL_MODULE_CHECK (NODE))->u.base.module_entity_p)
> >   /* DECL that has attached decls for ODR-relatedness.  */
> > -#define DECL_MODULE_KEYED_DECLS_P(NODE)			\
> > -  (DECL_LANG_SPECIFIC (TREE_CHECK2(NODE,FUNCTION_DECL,VAR_DECL))\
> > -   ->u.base.module_keyed_decls_p)
> > +#define DECL_MODULE_KEYED_DECLS_P(NODE) \
> > +  (DECL_LANG_SPECIFIC (DECL_MODULE_CHECK (NODE))->u.base.module_keyed_decls_p)
> >   /* Whether this is an exported DECL.  Held on any decl that can appear
> >      at namespace scope (function, var, type, template, const or
> > @@ -2887,21 +2886,20 @@ struct GTY(()) lang_decl_base {
> >     unsigned friend_or_tls : 1;		   /* var, fn, type or template */
> >     unsigned unknown_bound_p : 1;		   /* var */
> >     unsigned odr_used : 1;		   /* var or fn */
> > -  unsigned concept_p : 1;                  /* applies to vars and functions */
> > +  unsigned concept_p : 1;		   /* applies to vars and functions */
> >     unsigned var_declared_inline_p : 1;	   /* var */
> >     unsigned dependent_init_p : 1;	   /* var */
> > -  /* The following apply to VAR, FUNCTION, TYPE, CONCEPT, & NAMESPACE
> > +  /* The following four apply to VAR, FUNCTION, TYPE, CONCEPT, & NAMESPACE
> >        decls.  */
> > -  unsigned module_purview_p : 1;	   // in named-module purview
> > -  unsigned module_attach_p : 1;		   // attached to named module
> > -  unsigned module_import_p : 1;     	   /* from an import */
> > -  unsigned module_entity_p : 1;		   /* is in the entitity ary &
> > -					      hash.  */
> > -  /* VAR_DECL or FUNCTION_DECL has keyed decls.     */
> > -  unsigned module_keyed_decls_p : 1;
> > -
> > -  /* 12 spare bits.  */
> > +  unsigned module_purview_p : 1;	   /* in named-module purview */
> > +  unsigned module_attach_p : 1;		   /* attached to named module */
> > +  unsigned module_import_p : 1;		   /* from an import */
> > +  unsigned module_entity_p : 1;		   /* is in the entitity ary & hash */
> > +
> > +  unsigned module_keyed_decls_p : 1;	   /* has keys, applies to all decls */
> > +
> > +  /* 11 spare bits.  */
> >   };
> >   /* True for DECL codes which have template info and access.  */
> > diff --git a/gcc/cp/module.cc b/gcc/cp/module.cc
> > index 106af7bdb3e..1b2ba2e0fa8 100644
> > --- a/gcc/cp/module.cc
> > +++ b/gcc/cp/module.cc
> > @@ -5664,8 +5664,7 @@ trees_out::lang_decl_bools (tree t)
> >        want to mark them as in module purview.  */
> >     WB (lang->u.base.module_purview_p && !header_module_p ());
> >     WB (lang->u.base.module_attach_p);
> > -  if (VAR_OR_FUNCTION_DECL_P (t))
> > -    WB (lang->u.base.module_keyed_decls_p);
> > +  WB (lang->u.base.module_keyed_decls_p);
> >     switch (lang->u.base.selector)
> >       {
> >       default:
> > @@ -5738,8 +5737,7 @@ trees_in::lang_decl_bools (tree t)
> >     RB (lang->u.base.dependent_init_p);
> >     RB (lang->u.base.module_purview_p);
> >     RB (lang->u.base.module_attach_p);
> > -  if (VAR_OR_FUNCTION_DECL_P (t))
> > -    RB (lang->u.base.module_keyed_decls_p);
> > +  RB (lang->u.base.module_keyed_decls_p);
> >     switch (lang->u.base.selector)
> >       {
> >       default:
> > @@ -7871,8 +7869,7 @@ trees_out::decl_value (tree decl, depset *dep)
> >         install_entity (decl, dep);
> >       }
> > -  if (VAR_OR_FUNCTION_DECL_P (inner)
> > -      && DECL_LANG_SPECIFIC (inner)
> > +  if (DECL_LANG_SPECIFIC (inner)
> >         && DECL_MODULE_KEYED_DECLS_P (inner)
> >         && !is_key_order ())
> >       {
> > @@ -8172,8 +8169,7 @@ trees_in::decl_value ()
> >     bool installed = install_entity (existing);
> >     bool is_new = existing == decl;
> > -  if (VAR_OR_FUNCTION_DECL_P (inner)
> > -      && DECL_LANG_SPECIFIC (inner)
> > +  if (DECL_LANG_SPECIFIC (inner)
> >         && DECL_MODULE_KEYED_DECLS_P (inner))
> >       {
> >         /* Read and maybe install the attached entities.  */
> > @@ -9187,6 +9183,13 @@ trees_in::tree_value ()
> >         return NULL_TREE;
> >       }
> > +  if (TREE_CODE (t) == LAMBDA_EXPR
> > +      && CLASSTYPE_LAMBDA_EXPR (TREE_TYPE (t)))
> > +    {
> > +      existing = CLASSTYPE_LAMBDA_EXPR (TREE_TYPE (t));
> > +      back_refs[~tag] = existing;
> > +    }
> > +
> >     dump (dumper::TREE) && dump ("Read tree:%d %C:%N", tag, TREE_CODE (t), t);
> >     if (TREE_CODE (existing) == INTEGER_CST && !TREE_OVERFLOW (existing))
> > @@ -10484,12 +10487,17 @@ trees_out::get_merge_kind (tree decl, depset *dep)
> >   	      if (tree scope
> >   		  = LAMBDA_EXPR_EXTRA_SCOPE (CLASSTYPE_LAMBDA_EXPR
> >   					     (TREE_TYPE (decl))))
> > -		if (TREE_CODE (scope) == VAR_DECL
> > -		    && DECL_MODULE_KEYED_DECLS_P (scope))
> > -		  {
> > -		    mk = MK_keyed;
> > -		    break;
> > -		  }
> > +		{
> > +		  /* Lambdas attached to fields are keyed to its class.  */
> > +		  if (TREE_CODE (scope) == FIELD_DECL)
> > +		    scope = TYPE_NAME (DECL_CONTEXT (scope));
> > +		  if (DECL_LANG_SPECIFIC (scope)
> > +		      && DECL_MODULE_KEYED_DECLS_P (scope))
> > +		    {
> > +		      mk = MK_keyed;
> > +		      break;
> > +		    }
> > +		}
> >   	    if (RECORD_OR_UNION_TYPE_P (ctx))
> >   	      {
> > @@ -10789,7 +10797,13 @@ trees_out::key_mergeable (int tag, merge_kind mk, tree decl, tree inner,
> >   	    gcc_checking_assert (LAMBDA_TYPE_P (TREE_TYPE (inner)));
> >   	    tree scope = LAMBDA_EXPR_EXTRA_SCOPE (CLASSTYPE_LAMBDA_EXPR
> >   						  (TREE_TYPE (inner)));
> > -	    gcc_checking_assert (TREE_CODE (scope) == VAR_DECL);
> > +	    gcc_checking_assert (TREE_CODE (scope) == VAR_DECL
> > +				 || TREE_CODE (scope) == FIELD_DECL
> > +				 || TREE_CODE (scope) == PARM_DECL
> > +				 || TREE_CODE (scope) == TYPE_DECL);
> > +	    /* Lambdas attached to fields are keyed to the class.  */
> > +	    if (TREE_CODE (scope) == FIELD_DECL)
> > +	      scope = TYPE_NAME (DECL_CONTEXT (scope));
> >   	    auto *root = keyed_table->get (scope);
> >   	    unsigned ix = root->length ();
> >   	    /* If we don't find it, we'll write a really big number
> > @@ -11067,6 +11081,26 @@ trees_in::key_mergeable (int tag, merge_kind mk, tree decl, tree inner,
> >   		}
> >   	    }
> >   	}
> > +      else if (mk == MK_keyed
> > +	       && DECL_LANG_SPECIFIC (name)
> > +	       && DECL_MODULE_KEYED_DECLS_P (name))
> > +	{
> > +	  gcc_checking_assert (TREE_CODE (container) == NAMESPACE_DECL
> > +			       || TREE_CODE (container) == TYPE_DECL);
> > +	  if (auto *set = keyed_table->get (name))
> > +	    if (key.index < set->length ())
> > +	      {
> > +		existing = (*set)[key.index];
> > +		if (existing)
> > +		  {
> > +		    gcc_checking_assert
> > +		      (DECL_IMPLICIT_TYPEDEF_P (existing));
> > +		    if (inner != decl)
> > +		      existing
> > +			= CLASSTYPE_TI_TEMPLATE (TREE_TYPE (existing));
> > +		  }
> > +	      }
> > +	}
> >         else
> >   	switch (TREE_CODE (container))
> >   	  {
> > @@ -11074,27 +11108,8 @@ trees_in::key_mergeable (int tag, merge_kind mk, tree decl, tree inner,
> >   	    gcc_unreachable ();
> >   	  case NAMESPACE_DECL:
> > -	    if (mk == MK_keyed)
> > -	      {
> > -		if (DECL_LANG_SPECIFIC (name)
> > -		    && VAR_OR_FUNCTION_DECL_P (name)
> > -		    && DECL_MODULE_KEYED_DECLS_P (name))
> > -		  if (auto *set = keyed_table->get (name))
> > -		    if (key.index < set->length ())
> > -		      {
> > -			existing = (*set)[key.index];
> > -			if (existing)
> > -			  {
> > -			    gcc_checking_assert
> > -			      (DECL_IMPLICIT_TYPEDEF_P (existing));
> > -			    if (inner != decl)
> > -			      existing
> > -				= CLASSTYPE_TI_TEMPLATE (TREE_TYPE (existing));
> > -			  }
> > -		      }
> > -	      }
> > -	    else if (is_attached
> > -		     && !(state->is_module () || state->is_partition ()))
> > +	    if (is_attached
> > +		&& !(state->is_module () || state->is_partition ()))
> >   	      kind = "unique";
> >   	    else
> >   	      {
> > @@ -18984,11 +18999,19 @@ maybe_key_decl (tree ctx, tree decl)
> >     if (!modules_p ())
> >       return;
> > -  // FIXME: For now just deal with lambdas attached to var decls.
> > -  // This might be sufficient?
> > -  if (TREE_CODE (ctx) != VAR_DECL)
> > +  /* We only need to deal with lambdas attached to var, field,
> > +     parm, or type decls.  */
> > +  if (TREE_CODE (ctx) != VAR_DECL
> > +      && TREE_CODE (ctx) != FIELD_DECL
> > +      && TREE_CODE (ctx) != PARM_DECL
> > +      && TREE_CODE (ctx) != TYPE_DECL)
> >       return;
> > +  /* For fields, key it to the containing type to handle deduplication
> > +     correctly.  */
> > +  if (TREE_CODE (ctx) == FIELD_DECL)
> > +    ctx = TYPE_NAME (DECL_CONTEXT (ctx));
> > +
> >     if (!keyed_table)
> >       keyed_table = new keyed_map_t (EXPERIMENT (1, 400));
> > diff --git a/gcc/cp/parser.cc b/gcc/cp/parser.cc
> > index b2ed2baa3a5..3ee9d49fb8e 100644
> > --- a/gcc/cp/parser.cc
> > +++ b/gcc/cp/parser.cc
> > @@ -27678,10 +27678,16 @@ cp_parser_class_head (cp_parser* parser,
> >     if (cp_lexer_next_token_is (parser->lexer, CPP_COLON))
> >       {
> >         if (type)
> > -	pushclass (type);
> > +	{
> > +	  pushclass (type);
> > +	  start_lambda_scope (TYPE_NAME (type));
> > +	}
> >         bases = cp_parser_base_clause (parser);
> >         if (type)
> > -	popclass ();
> > +	{
> > +	  finish_lambda_scope ();
> > +	  popclass ();
> > +	}
> >       }
> >     else
> >       bases = NULL_TREE;
> > diff --git a/gcc/testsuite/g++.dg/modules/lambda-7.h b/gcc/testsuite/g++.dg/modules/lambda-7.h
> > new file mode 100644
> > index 00000000000..6f6080c1324
> > --- /dev/null
> > +++ b/gcc/testsuite/g++.dg/modules/lambda-7.h
> > @@ -0,0 +1,42 @@
> > +struct S {
> > +  int (*a)(int) = [](int x) { return x * 2; };
> > +
> > +  int b(int x, int (*f)(int) = [](int x) { return x * 3; }) {
> > +    return f(x);
> > +  }
> > +
> > +  static int c(int x, int (*f)(int) = [](int x) { return x * 4; }) {
> > +    return f(x);
> > +  }
> > +};
> > +
> > +inline int d(int x, int (*f)(int) = [](int x) { return x * 5; }) {
> > +  return f(x);
> > +}
> > +
> > +// unevaluated lambdas
> > +#if __cplusplus >= 202002L
> > +struct E : decltype([](int x) { return x * 6; }) {
> > +  decltype([](int x) { return x * 7; }) f;
> > +};
> > +
> > +template <typename T>
> > +struct G : decltype([](int x) { return x * 8; }) {
> > +  decltype([](int x) { return x * 9; }) h;
> > +};
> > +
> > +template <>
> > +struct G<double> : decltype([](int x) { return x * 10; }) {
> > +  decltype([](int x) { return x * 11; }) i;
> > +};
> > +#endif
> > +
> > +// concepts
> > +#if __cpp_concepts >= 201907L
> > +template <typename T>
> > +concept J = requires { []{ T(); }; };
> > +
> > +template <typename T>
> > +concept K = []{ return sizeof(T) == 1; }();
> > +#endif
> > +
> > diff --git a/gcc/testsuite/g++.dg/modules/lambda-7_a.H b/gcc/testsuite/g++.dg/modules/lambda-7_a.H
> > new file mode 100644
> > index 00000000000..5197114f76c
> > --- /dev/null
> > +++ b/gcc/testsuite/g++.dg/modules/lambda-7_a.H
> > @@ -0,0 +1,4 @@
> > +// { dg-additional-options "-fmodule-header -Wno-subobject-linkage" }
> > +// { dg-module-cmi {} }
> > +
> > +#include "lambda-7.h"
> > diff --git a/gcc/testsuite/g++.dg/modules/lambda-7_b.C b/gcc/testsuite/g++.dg/modules/lambda-7_b.C
> > new file mode 100644
> > index 00000000000..2d781e93067
> > --- /dev/null
> > +++ b/gcc/testsuite/g++.dg/modules/lambda-7_b.C
> > @@ -0,0 +1,5 @@
> > +// { dg-additional-options "-fmodules-ts -fno-module-lazy -Wno-subobject-linkage" }
> > +// Test for ODR deduplication
> > +
> > +#include "lambda-7.h"
> > +import "lambda-7_a.H";
> > diff --git a/gcc/testsuite/g++.dg/modules/lambda-7_c.C b/gcc/testsuite/g++.dg/modules/lambda-7_c.C
> > new file mode 100644
> > index 00000000000..f283681fa96
> > --- /dev/null
> > +++ b/gcc/testsuite/g++.dg/modules/lambda-7_c.C
> > @@ -0,0 +1,41 @@
> > +// { dg-module-do run }
> > +// { dg-additional-options "-fmodules-ts -fno-module-lazy -Wno-subobject-linkage" }
> > +
> > +import "lambda-7_a.H";
> > +
> > +int main() {
> > +  S s;
> > +  if (s.a(10) != 20)
> > +    __builtin_abort();
> > +  if (s.b(10) != 30)
> > +    __builtin_abort();
> > +  if (s.c(10) != 40)
> > +    __builtin_abort();
> > +  if (d(10) != 50)
> > +    __builtin_abort();
> > +
> > +#if __cplusplus >= 202002L
> > +  E e;
> > +  if (e(10) != 60)
> > +    __builtin_abort();
> > +  if (e.f(10) != 70)
> > +    __builtin_abort();
> > +
> > +  G<int> g1;
> > +  if (g1(10) != 80)
> > +    __builtin_abort();
> > +  if (g1.h(10) != 90)
> > +    __builtin_abort();
> > +
> > +  G<double> g2;
> > +  if (g2(10) != 100)
> > +    __builtin_abort();
> > +  if (g2.i(10) != 110)
> > +    __builtin_abort();
> > +#endif
> > +
> > +#if __cpp_concepts >= 201907L
> > +  static_assert(J<char>);
> > +  static_assert(K<char>);
> > +#endif
> > +}
>