* [PATCH] ipa: "naked" attribute implies "noipa" attribute
@ 2021-08-13 9:14 Martin Liška
2021-08-13 10:31 ` Jan Hubicka
2021-08-16 6:15 ` Richard Biener
0 siblings, 2 replies; 5+ messages in thread
From: Martin Liška @ 2021-08-13 9:14 UTC (permalink / raw)
To: gcc-patches; +Cc: Jan Hubicka
Hi.
This is a first part fixing the PR. It makes sense making "naked" functions "noipa".
What's missing is IPA MOD pass support where the pass should not optimize fns
with "noipa" attributes.
@Honza: Can you please implement that?
Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
Ready to be installed?
Thanks,
Martin
PR ipa/101354
gcc/ChangeLog:
* attribs.c (decl_attributes): Make naked functions "noipa"
functions.
---
gcc/attribs.c | 11 +++--------
1 file changed, 3 insertions(+), 8 deletions(-)
diff --git a/gcc/attribs.c b/gcc/attribs.c
index afa485ed37d..e149f5b66dd 100644
--- a/gcc/attribs.c
+++ b/gcc/attribs.c
@@ -517,14 +517,9 @@ decl_attributes (tree *node, tree attributes, int flags,
if (TREE_CODE (*node) == FUNCTION_DECL
&& attributes
&& lookup_attribute ("naked", attributes) != NULL
- && lookup_attribute_spec (get_identifier ("naked")))
- {
- if (lookup_attribute ("noinline", attributes) == NULL)
- attributes = tree_cons (get_identifier ("noinline"), NULL, attributes);
-
- if (lookup_attribute ("noclone", attributes) == NULL)
- attributes = tree_cons (get_identifier ("noclone"), NULL, attributes);
- }
+ && lookup_attribute_spec (get_identifier ("naked"))
+ &&(lookup_attribute ("noipa", attributes) == NULL))
+ attributes = tree_cons (get_identifier ("noipa"), NULL, attributes);
/* A "noipa" function attribute implies "noinline", "noclone" and "no_icf"
for those targets that support it. */
--
2.32.0
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] ipa: "naked" attribute implies "noipa" attribute
2021-08-13 9:14 [PATCH] ipa: "naked" attribute implies "noipa" attribute Martin Liška
@ 2021-08-13 10:31 ` Jan Hubicka
2021-08-13 14:03 ` Martin Liška
2021-08-16 6:15 ` Richard Biener
1 sibling, 1 reply; 5+ messages in thread
From: Jan Hubicka @ 2021-08-13 10:31 UTC (permalink / raw)
To: Martin Liška; +Cc: gcc-patches
> Hi.
>
> This is a first part fixing the PR. It makes sense making "naked" functions "noipa".
> What's missing is IPA MOD pass support where the pass should not optimize fns
> with "noipa" attributes.
>
> @Honza: Can you please implement that?
Hmm, I had patch for that somewhere, will do that.
>
> Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
>
> Ready to be installed?
OK, except for..
> + && lookup_attribute_spec (get_identifier ("naked"))
> + &&(lookup_attribute ("noipa", attributes) == NULL))
missing space or extra braces here. I am not sure how much we play with
NULL_TREE these days.
Honza
> + attributes = tree_cons (get_identifier ("noipa"), NULL, attributes);
> /* A "noipa" function attribute implies "noinline", "noclone" and "no_icf"
> for those targets that support it. */
> --
> 2.32.0
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] ipa: "naked" attribute implies "noipa" attribute
2021-08-13 10:31 ` Jan Hubicka
@ 2021-08-13 14:03 ` Martin Liška
0 siblings, 0 replies; 5+ messages in thread
From: Martin Liška @ 2021-08-13 14:03 UTC (permalink / raw)
To: Jan Hubicka; +Cc: gcc-patches
On 8/13/21 12:31 PM, Jan Hubicka wrote:
>> Hi.
>>
>> This is a first part fixing the PR. It makes sense making "naked" functions "noipa".
>> What's missing is IPA MOD pass support where the pass should not optimize fns
>> with "noipa" attributes.
>>
>> @Honza: Can you please implement that?
>
> Hmm, I had patch for that somewhere, will do that.
Please do so, I'm going to assign you the PR.
>>
>> Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
>>
>> Ready to be installed?
>
> OK, except for..
>> + && lookup_attribute_spec (get_identifier ("naked"))
>> + &&(lookup_attribute ("noipa", attributes) == NULL))
>
> missing space or extra braces here. I am not sure how much we play with
> NULL_TREE these days.
Sure, I'm going to fix it and push to master.
Martin
>
> Honza
>> + attributes = tree_cons (get_identifier ("noipa"), NULL, attributes);
>> /* A "noipa" function attribute implies "noinline", "noclone" and "no_icf"
>> for those targets that support it. */
>> --
>> 2.32.0
>>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] ipa: "naked" attribute implies "noipa" attribute
2021-08-13 9:14 [PATCH] ipa: "naked" attribute implies "noipa" attribute Martin Liška
2021-08-13 10:31 ` Jan Hubicka
@ 2021-08-16 6:15 ` Richard Biener
2021-08-16 6:16 ` Richard Biener
1 sibling, 1 reply; 5+ messages in thread
From: Richard Biener @ 2021-08-16 6:15 UTC (permalink / raw)
To: Martin Liška; +Cc: GCC Patches, Jan Hubicka
On Fri, Aug 13, 2021 at 11:14 AM Martin Liška <mliska@suse.cz> wrote:
>
> Hi.
>
> This is a first part fixing the PR. It makes sense making "naked" functions "noipa".
> What's missing is IPA MOD pass support where the pass should not optimize fns
> with "noipa" attributes.
>
> @Honza: Can you please implement that?
>
> Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
>
> Ready to be installed?
> Thanks,
> Martin
>
> PR ipa/101354
>
> gcc/ChangeLog:
>
> * attribs.c (decl_attributes): Make naked functions "noipa"
> functions.
> ---
> gcc/attribs.c | 11 +++--------
> 1 file changed, 3 insertions(+), 8 deletions(-)
>
> diff --git a/gcc/attribs.c b/gcc/attribs.c
> index afa485ed37d..e149f5b66dd 100644
> --- a/gcc/attribs.c
> +++ b/gcc/attribs.c
> @@ -517,14 +517,9 @@ decl_attributes (tree *node, tree attributes, int flags,
> if (TREE_CODE (*node) == FUNCTION_DECL
> && attributes
> && lookup_attribute ("naked", attributes) != NULL
> - && lookup_attribute_spec (get_identifier ("naked")))
> - {
> - if (lookup_attribute ("noinline", attributes) == NULL)
> - attributes = tree_cons (get_identifier ("noinline"), NULL, attributes);
> -
> - if (lookup_attribute ("noclone", attributes) == NULL)
> - attributes = tree_cons (get_identifier ("noclone"), NULL, attributes);
> - }
> + && lookup_attribute_spec (get_identifier ("naked"))
> + &&(lookup_attribute ("noipa", attributes) == NULL))
> + attributes = tree_cons (get_identifier ("noipa"), NULL, attributes);
that adds noipa only for the purpose of "decomposing" it below.
> /* A "noipa" function attribute implies "noinline", "noclone" and "no_icf"
> for those targets that support it. */
> --
> 2.32.0
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] ipa: "naked" attribute implies "noipa" attribute
2021-08-16 6:15 ` Richard Biener
@ 2021-08-16 6:16 ` Richard Biener
0 siblings, 0 replies; 5+ messages in thread
From: Richard Biener @ 2021-08-16 6:16 UTC (permalink / raw)
To: Martin Liška; +Cc: GCC Patches, Jan Hubicka
On Mon, Aug 16, 2021 at 8:15 AM Richard Biener
<richard.guenther@gmail.com> wrote:
>
> On Fri, Aug 13, 2021 at 11:14 AM Martin Liška <mliska@suse.cz> wrote:
> >
> > Hi.
> >
> > This is a first part fixing the PR. It makes sense making "naked" functions "noipa".
> > What's missing is IPA MOD pass support where the pass should not optimize fns
> > with "noipa" attributes.
> >
> > @Honza: Can you please implement that?
> >
> > Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
> >
> > Ready to be installed?
> > Thanks,
> > Martin
> >
> > PR ipa/101354
> >
> > gcc/ChangeLog:
> >
> > * attribs.c (decl_attributes): Make naked functions "noipa"
> > functions.
> > ---
> > gcc/attribs.c | 11 +++--------
> > 1 file changed, 3 insertions(+), 8 deletions(-)
> >
> > diff --git a/gcc/attribs.c b/gcc/attribs.c
> > index afa485ed37d..e149f5b66dd 100644
> > --- a/gcc/attribs.c
> > +++ b/gcc/attribs.c
> > @@ -517,14 +517,9 @@ decl_attributes (tree *node, tree attributes, int flags,
> > if (TREE_CODE (*node) == FUNCTION_DECL
> > && attributes
> > && lookup_attribute ("naked", attributes) != NULL
> > - && lookup_attribute_spec (get_identifier ("naked")))
> > - {
> > - if (lookup_attribute ("noinline", attributes) == NULL)
> > - attributes = tree_cons (get_identifier ("noinline"), NULL, attributes);
> > -
> > - if (lookup_attribute ("noclone", attributes) == NULL)
> > - attributes = tree_cons (get_identifier ("noclone"), NULL, attributes);
> > - }
> > + && lookup_attribute_spec (get_identifier ("naked"))
> > + &&(lookup_attribute ("noipa", attributes) == NULL))
> > + attributes = tree_cons (get_identifier ("noipa"), NULL, attributes);
>
> that adds noipa only for the purpose of "decomposing" it below.
Hmm, no, we do seem to check for "noipa" in quite some places, so ignore
me.
Richard.
> > /* A "noipa" function attribute implies "noinline", "noclone" and "no_icf"
> > for those targets that support it. */
> > --
> > 2.32.0
> >
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2021-08-16 6:17 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-08-13 9:14 [PATCH] ipa: "naked" attribute implies "noipa" attribute Martin Liška
2021-08-13 10:31 ` Jan Hubicka
2021-08-13 14:03 ` Martin Liška
2021-08-16 6:15 ` Richard Biener
2021-08-16 6:16 ` Richard Biener
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).