From: Jason Merrill <jason@redhat.com>
To: Marek Polacek <polacek@redhat.com>
Cc: GCC Patches <gcc-patches@gcc.gnu.org>,
Joseph Myers <joseph@codesourcery.com>
Subject: Re: [PATCH v4] c-family: ICE with [[gnu::nocf_check]] [PR106937]
Date: Fri, 7 Oct 2022 17:56:18 -0400 [thread overview]
Message-ID: <93969753-469e-c5d9-6563-be4bc11ba29f@redhat.com> (raw)
In-Reply-To: <Y0CVRvVh+I5pixLz@redhat.com>
On 10/7/22 17:08, Marek Polacek wrote:
> On Fri, Oct 07, 2022 at 12:17:34PM -0400, Jason Merrill wrote:
>> On 10/6/22 22:12, Marek Polacek wrote:
>>> On Thu, Oct 06, 2022 at 05:42:41PM -0400, Jason Merrill wrote:
>>>> On 10/4/22 19:06, Marek Polacek wrote:
>>>>> On Fri, Sep 30, 2022 at 09:12:24AM -0400, Jason Merrill wrote:
>>>>>> On 9/29/22 18:49, Marek Polacek wrote:
>>>>>>> When getting the name of an attribute, we ought to use
>>>>>>> get_attribute_name, which handles both [[ ]] and __attribute__(())
>>>>>>> forms. Failure to do so may result in an ICE, like here.
>>>>>>>
>>>>>>> Bootstrapped/regtested on x86_64-pc-linux-gnu, ok for trunk?
>>>>>>
>>>>>> How do we print the attributes with this patch? Don't we also want to print
>>>>>> the namespace, and use [[]] in the output?
>>>>>
>>>>> Good point, however: while the testcase indeed has an attribute
>>>>> in the [[]] form in the typedef, here we're printing its "aka":
>>>>>
>>>>> warning: initialization of 'FuncPointerWithNoCfCheck' {aka 'void (__attribute__((nocf_check)) *)(void)'} from incompatible pointer type 'FuncPointer' {aka 'void (*)(void)'}
>>>>>
>>>>> c-pretty-print.cc doesn't seem to know how to print an [[]] attribute.
>>>>> I could do that, but then we'd print
>>>>>
>>>>> aka 'void ([[nocf_check]] *)(void)'
>>>>>
>>>>> in the above, but that's invalid syntax!
>>>>
>>>> Indeed, it should be
>>>>
>>>> void (* [[gnu::nocf_check]])(void)
>>>
>>> Ok, let me fix that too, then. I've updated the description to show
>>> what we print with the patch, and what was printed before.
>>
>> Oops, apparently I was wrongly assuming that the attribute appertained to
>> the pointer.
>
> So was I :/.
>
>> Since it actually appertains to the function type, it should
>> be
>>
>> void (*)(void) [[gnu::nocf_check]].
>>
>> But for GCC attribute syntax, I think we want to leave the __attribute__
>> where it was before, as in the manual's example
>>
>> void (__attribute__((noreturn)) ****f) (void);
>
> Thanks, done here:
>
> Bootstrapped/regtested on x86_64-pc-linux-gnu, ok for trunk?
>
> -- >8 --
> When getting the name of an attribute, we ought to use
> get_attribute_name, which handles both [[]] and __attribute__(())
> forms. Failure to do so may result in an ICE, like here.
>
> pp_c_attributes_display wasn't able to print the [[]] form of
> attributes, so this patch teaches it to.
>
> When printing a pointer to function with a standard attribute, the attribute
> should be printed after the parameter-list. With this patch we print:
>
> aka 'void (*)(int) [[gnu::nocf_check]]'
>
> pp_c_attributes has been unused since its introduction in r56273 so
> this patch removes it.
>
> PR c++/106937
>
> gcc/c-family/ChangeLog:
>
> * c-pretty-print.cc (pp_c_specifier_qualifier_list): Only print
> GNU attributes here.
> (c_pretty_printer::abstract_declarator): For a pointer to function,
> print the standard [[]] attributes here.
> (pp_c_attributes): Remove.
> (pp_c_attributes_display): Print the [[]] form if appropriate. Use
> get_attribute_name. Don't print a trailing space when printing the
> [[]] form.
> * c-pretty-print.h (pp_c_attributes): Remove.
>
> gcc/cp/ChangeLog:
>
> * error.cc: Include "attribs.h".
> (dump_type_prefix): Only print GNU attributes here.
> (dump_type_suffix): Print standard attributes here.
>
> gcc/testsuite/ChangeLog:
>
> * c-c++-common/pointer-to-fn1.c: New test.
> ---
> gcc/c-family/c-pretty-print.cc | 102 +++++++++++---------
> gcc/c-family/c-pretty-print.h | 1 -
> gcc/cp/error.cc | 17 +++-
> gcc/testsuite/c-c++-common/pointer-to-fn1.c | 18 ++++
> 4 files changed, 92 insertions(+), 46 deletions(-)
> create mode 100644 gcc/testsuite/c-c++-common/pointer-to-fn1.c
>
> diff --git a/gcc/c-family/c-pretty-print.cc b/gcc/c-family/c-pretty-print.cc
> index efa1768f4d6..349f0a07d3e 100644
> --- a/gcc/c-family/c-pretty-print.cc
> +++ b/gcc/c-family/c-pretty-print.cc
> @@ -466,7 +466,12 @@ pp_c_specifier_qualifier_list (c_pretty_printer *pp, tree t)
> {
> pp_c_whitespace (pp);
> pp_c_left_paren (pp);
> - pp_c_attributes_display (pp, TYPE_ATTRIBUTES (pointee));
> + /* If we're dealing with the GNU form of attributes, print this:
> + void (__attribute__((noreturn)) *f) ();
> + If it is the standard [[]] attribute, we'll print the attribute
> + in c_pretty_printer::abstract_declarator. */
> + if (!cxx11_attribute_p (TYPE_ATTRIBUTES (pointee)))
> + pp_c_attributes_display (pp, TYPE_ATTRIBUTES (pointee));
> }
> else if (!c_dialect_cxx ())
> pp_c_whitespace (pp);
> @@ -564,15 +569,26 @@ pp_c_parameter_type_list (c_pretty_printer *pp, tree t)
> void
> c_pretty_printer::abstract_declarator (tree t)
> {
> + bool fn_ptr_p = false;
> if (TREE_CODE (t) == POINTER_TYPE)
> {
> if (TREE_CODE (TREE_TYPE (t)) == ARRAY_TYPE
> || TREE_CODE (TREE_TYPE (t)) == FUNCTION_TYPE)
> - pp_c_right_paren (this);
> + {
> + pp_c_right_paren (this);
> + fn_ptr_p = true;
> + }
> t = TREE_TYPE (t);
> }
>
> direct_abstract_declarator (t);
> + /* If this is the standard [[]] attribute, print
> + void (*)() [[noreturn]]; */
> + if (fn_ptr_p && cxx11_attribute_p (TYPE_ATTRIBUTES (t)))
Do you need to check fn_ptr_p? The attributes on a function type go
here whether or not we were dealing with a pointer.
> + {
> + pp_space (this);
> + pp_c_attributes_display (this, TYPE_ATTRIBUTES (t));
> + }
> }
>
> /* direct-abstract-declarator:
> @@ -850,32 +866,7 @@ c_pretty_printer::declaration (tree t)
> pp_c_init_declarator (this, t);
> }
>
> -/* Pretty-print ATTRIBUTES using GNU C extension syntax. */
> -
> -void
> -pp_c_attributes (c_pretty_printer *pp, tree attributes)
> -{
> - if (attributes == NULL_TREE)
> - return;
> -
> - pp_c_ws_string (pp, "__attribute__");
> - pp_c_left_paren (pp);
> - pp_c_left_paren (pp);
> - for (; attributes != NULL_TREE; attributes = TREE_CHAIN (attributes))
> - {
> - pp_tree_identifier (pp, TREE_PURPOSE (attributes));
> - if (TREE_VALUE (attributes))
> - pp_c_call_argument_list (pp, TREE_VALUE (attributes));
> -
> - if (TREE_CHAIN (attributes))
> - pp_separate_with (pp, ',');
> - }
> - pp_c_right_paren (pp);
> - pp_c_right_paren (pp);
> -}
> -
> -/* Pretty-print ATTRIBUTES using GNU C extension syntax for attributes
> - marked to be displayed on disgnostic. */
> +/* Pretty-print ATTRIBUTES marked to be displayed on diagnostic. */
>
> void
> pp_c_attributes_display (c_pretty_printer *pp, tree a)
> @@ -885,10 +876,12 @@ pp_c_attributes_display (c_pretty_printer *pp, tree a)
> if (a == NULL_TREE)
> return;
>
> + const bool std_p = cxx11_attribute_p (a);
> +
> for (; a != NULL_TREE; a = TREE_CHAIN (a))
> {
> - const struct attribute_spec *as;
> - as = lookup_attribute_spec (TREE_PURPOSE (a));
> + const struct attribute_spec *as
> + = lookup_attribute_spec (get_attribute_name (a));
> if (!as || as->affects_type_identity == false)
> continue;
> if (c_dialect_cxx ()
> @@ -896,26 +889,47 @@ pp_c_attributes_display (c_pretty_printer *pp, tree a)
> /* In C++ transaction_safe is printed at the end of the declarator. */
> continue;
> if (is_first)
> - {
> - pp_c_ws_string (pp, "__attribute__");
> - pp_c_left_paren (pp);
> - pp_c_left_paren (pp);
> - is_first = false;
> - }
> + {
> + if (std_p)
> + {
> + pp_c_left_bracket (pp);
> + pp_c_left_bracket (pp);
> + }
> + else
> + {
> + pp_c_ws_string (pp, "__attribute__");
> + pp_c_left_paren (pp);
> + pp_c_left_paren (pp);
> + }
> + is_first = false;
> + }
> else
> - {
> - pp_separate_with (pp, ',');
> - }
> - pp_tree_identifier (pp, TREE_PURPOSE (a));
> + pp_separate_with (pp, ',');
> + tree ns;
> + if (std_p && (ns = get_attribute_namespace (a)))
> + {
> + pp_tree_identifier (pp, ns);
> + pp_colon (pp);
> + pp_colon (pp);
> + }
> + pp_tree_identifier (pp, get_attribute_name (a));
> if (TREE_VALUE (a))
> - pp_c_call_argument_list (pp, TREE_VALUE (a));
> + pp_c_call_argument_list (pp, TREE_VALUE (a));
> }
>
> if (!is_first)
> {
> - pp_c_right_paren (pp);
> - pp_c_right_paren (pp);
> - pp_c_whitespace (pp);
> + if (std_p)
> + {
> + pp_c_right_bracket (pp);
> + pp_c_right_bracket (pp);
> + }
> + else
> + {
> + pp_c_right_paren (pp);
> + pp_c_right_paren (pp);
> + pp_c_whitespace (pp);
> + }
> }
> }
>
> diff --git a/gcc/c-family/c-pretty-print.h b/gcc/c-family/c-pretty-print.h
> index be86bed4fee..92674ab4d06 100644
> --- a/gcc/c-family/c-pretty-print.h
> +++ b/gcc/c-family/c-pretty-print.h
> @@ -119,7 +119,6 @@ void pp_c_space_for_pointer_operator (c_pretty_printer *, tree);
> /* Declarations. */
> void pp_c_tree_decl_identifier (c_pretty_printer *, tree);
> void pp_c_function_definition (c_pretty_printer *, tree);
> -void pp_c_attributes (c_pretty_printer *, tree);
> void pp_c_attributes_display (c_pretty_printer *, tree);
> void pp_c_cv_qualifiers (c_pretty_printer *pp, int qualifiers, bool func_type);
> void pp_c_type_qualifier_list (c_pretty_printer *, tree);
> diff --git a/gcc/cp/error.cc b/gcc/cp/error.cc
> index 4514c8bbb44..8cd011fab0e 100644
> --- a/gcc/cp/error.cc
> +++ b/gcc/cp/error.cc
> @@ -36,6 +36,7 @@ along with GCC; see the file COPYING3. If not see
> #include "internal-fn.h"
> #include "gcc-rich-location.h"
> #include "cp-name-hint.h"
> +#include "attribs.h"
>
> #define pp_separate_with_comma(PP) pp_cxx_separate_with (PP, ',')
> #define pp_separate_with_semicolon(PP) pp_cxx_separate_with (PP, ';')
> @@ -897,7 +898,12 @@ dump_type_prefix (cxx_pretty_printer *pp, tree t, int flags)
> {
> pp_cxx_whitespace (pp);
> pp_cxx_left_paren (pp);
> - pp_c_attributes_display (pp, TYPE_ATTRIBUTES (sub));
> + /* If we're dealing with the GNU form of attributes, print this:
> + void (__attribute__((noreturn)) *f) ();
> + If it is the standard [[]] attribute, we'll print the attribute
> + in dump_type_suffix. */
> + if (!cxx11_attribute_p (TYPE_ATTRIBUTES (sub)))
> + pp_c_attributes_display (pp, TYPE_ATTRIBUTES (sub));
> }
> if (TYPE_PTR_P (t))
> pp_star (pp);
> @@ -1027,6 +1033,15 @@ dump_type_suffix (cxx_pretty_printer *pp, tree t, int flags)
> TREE_CODE (t) == FUNCTION_TYPE
> && (flags & TFF_POINTER));
> dump_ref_qualifier (pp, t, flags);
> + /* If this is the standard [[]] attribute, print
> + void (*)() [[noreturn]]; */
> + if ((flags & TFF_POINTER)
And likewise here.
> + && cxx11_attribute_p (TYPE_ATTRIBUTES (t)))
> + {
> + pp_space (pp);
> + pp_c_attributes_display (pp, TYPE_ATTRIBUTES (t));
> + pp->padding = pp_before;
> + }
> if (tx_safe_fn_type_p (t))
> pp_cxx_ws_string (pp, "transaction_safe");
> dump_exception_spec (pp, TYPE_RAISES_EXCEPTIONS (t), flags);
> diff --git a/gcc/testsuite/c-c++-common/pointer-to-fn1.c b/gcc/testsuite/c-c++-common/pointer-to-fn1.c
> new file mode 100644
> index 00000000000..975885462e9
> --- /dev/null
> +++ b/gcc/testsuite/c-c++-common/pointer-to-fn1.c
> @@ -0,0 +1,18 @@
> +/* PR c++/106937 */
> +/* { dg-options "-fcf-protection" } */
> +/* { dg-additional-options "-std=c++11 -fpermissive" { target c++ } } */
> +/* Test printing a pointer to function with attribute. */
> +
> +__attribute__((nocf_check)) typedef void (*FPA1)();
> +[[gnu::nocf_check]] typedef void (*FPA2)(int);
> +typedef void (*FP1)();
> +typedef void (*FP2)(int);
> +
> +void
> +g (FP1 f1, FP2 f2)
> +{
> + FPA1 p1 = f1; // { dg-warning {aka 'void \(__attribute__\(\(nocf_check\)\) \*\)\(\)'} }
> + FPA2 p2 = f2; // { dg-warning {aka 'void \(\*\)\(int\) \[\[gnu::nocf_check\]\]'} }
> + FP1 p3 = p1; // { dg-warning {aka 'void \(__attribute__\(\(nocf_check\)\) \*\)\(\)'} }
> + FP2 p4 = p2; // { dg-warning {aka 'void \(\*\)\(int\) \[\[gnu::nocf_check\]\]'} }
> +}
>
> base-commit: f30e9fd33e56a5a721346ea6140722e1b193db42
next prev parent reply other threads:[~2022-10-07 21:56 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-29 22:49 [PATCH] " Marek Polacek
2022-09-30 13:12 ` Jason Merrill
2022-10-04 23:06 ` [PATCH v2] " Marek Polacek
2022-10-06 21:42 ` Jason Merrill
2022-10-07 2:12 ` [PATCH v3] " Marek Polacek
2022-10-07 16:17 ` Jason Merrill
2022-10-07 21:08 ` [PATCH v4] " Marek Polacek
2022-10-07 21:56 ` Jason Merrill [this message]
2022-10-07 22:16 ` Marek Polacek
2022-10-10 14:49 ` Jason Merrill
2022-10-10 19:18 ` [PATCH v5] " Marek Polacek
2022-10-10 19:23 ` Jason Merrill
2022-10-11 7:40 ` Andreas Schwab
2022-10-11 20:03 ` Marek Polacek
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=93969753-469e-c5d9-6563-be4bc11ba29f@redhat.com \
--to=jason@redhat.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=joseph@codesourcery.com \
--cc=polacek@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).