[PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[Sam James]

When exploring building bpf_skel with GCC's BPF support, there was a
buid failure because of bpf_core_field_exists vs the mem_hops bitfield:
```
 In file included from util/bpf_skel/sample_filter.bpf.c:6:
util/bpf_skel/sample_filter.bpf.c: In function 'perf_get_sample':
tools/perf/libbpf/include/bpf/bpf_core_read.h:169:42: error: cannot take address of bit-field 'mem_hops'
  169 | #define ___bpf_field_ref1(field)        (&(field))
      |                                          ^
tools/perf/libbpf/include/bpf/bpf_helpers.h:222:29: note: in expansion of macro '___bpf_field_ref1'
  222 | #define ___bpf_concat(a, b) a ## b
      |                             ^
tools/perf/libbpf/include/bpf/bpf_helpers.h:225:29: note: in expansion of macro '___bpf_concat'
  225 | #define ___bpf_apply(fn, n) ___bpf_concat(fn, n)
      |                             ^~~~~~~~~~~~~
tools/perf/libbpf/include/bpf/bpf_core_read.h:173:9: note: in expansion of macro '___bpf_apply'
  173 |         ___bpf_apply(___bpf_field_ref, ___bpf_narg(args))(args)
      |         ^~~~~~~~~~~~
tools/perf/libbpf/include/bpf/bpf_core_read.h:188:39: note: in expansion of macro '___bpf_field_ref'
  188 |         __builtin_preserve_field_info(___bpf_field_ref(field), BPF_FIELD_EXISTS)
      |                                       ^~~~~~~~~~~~~~~~
util/bpf_skel/sample_filter.bpf.c:167:29: note: in expansion of macro 'bpf_core_field_exists'
  167 |                         if (bpf_core_field_exists(data->mem_hops))
      |                             ^~~~~~~~~~~~~~~~~~~~~
cc1: error: argument is not a field access
```

___bpf_field_ref1 was adapted for GCC in 12bbcf8e840f40b82b02981e96e0a5fbb0703ea9
but the trick added for compatibility in 3a8b8fc3174891c4c12f5766d82184a82d4b2e3e
isn't compatible with that as an address is used as an argument.

Workaround this by calling __builtin_preserve_field_info directly as the
bpf_core_field_exists macro does, but without the ___bpf_field_ref use.

Link: https://gcc.gnu.org/PR121420
Co-authored-by: Andrew Pinski <quic_apinski@quicinc.com>
Signed-off-by: Sam James <sam@gentoo.org>
---
 tools/perf/util/bpf_skel/sample_filter.bpf.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/util/bpf_skel/sample_filter.bpf.c b/tools/perf/util/bpf_skel/sample_filter.bpf.c
index b195e6efeb8be..e5666d4c17228 100644
--- a/tools/perf/util/bpf_skel/sample_filter.bpf.c
+++ b/tools/perf/util/bpf_skel/sample_filter.bpf.c
@@ -164,7 +164,7 @@ static inline __u64 perf_get_sample(struct bpf_perf_event_data_kern *kctx,
 		if (entry->part == 8) {
 			union perf_mem_data_src___new *data = (void *)&kctx->data->data_src;
 
-			if (bpf_core_field_exists(data->mem_hops))
+			if (__builtin_preserve_field_info(data->mem_hops, BPF_FIELD_EXISTS))
 				return data->mem_hops;
 
 			return 0;
-- 
2.50.1

Re: [PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[Namhyung Kim]

Hello,

On Wed, Aug 06, 2025 at 01:03:01AM +0100, Sam James wrote:
> When exploring building bpf_skel with GCC's BPF support, there was a
> buid failure because of bpf_core_field_exists vs the mem_hops bitfield:
> ```
>  In file included from util/bpf_skel/sample_filter.bpf.c:6:
> util/bpf_skel/sample_filter.bpf.c: In function 'perf_get_sample':
> tools/perf/libbpf/include/bpf/bpf_core_read.h:169:42: error: cannot take address of bit-field 'mem_hops'
>   169 | #define ___bpf_field_ref1(field)        (&(field))
>       |                                          ^
> tools/perf/libbpf/include/bpf/bpf_helpers.h:222:29: note: in expansion of macro '___bpf_field_ref1'
>   222 | #define ___bpf_concat(a, b) a ## b
>       |                             ^
> tools/perf/libbpf/include/bpf/bpf_helpers.h:225:29: note: in expansion of macro '___bpf_concat'
>   225 | #define ___bpf_apply(fn, n) ___bpf_concat(fn, n)
>       |                             ^~~~~~~~~~~~~
> tools/perf/libbpf/include/bpf/bpf_core_read.h:173:9: note: in expansion of macro '___bpf_apply'
>   173 |         ___bpf_apply(___bpf_field_ref, ___bpf_narg(args))(args)
>       |         ^~~~~~~~~~~~
> tools/perf/libbpf/include/bpf/bpf_core_read.h:188:39: note: in expansion of macro '___bpf_field_ref'
>   188 |         __builtin_preserve_field_info(___bpf_field_ref(field), BPF_FIELD_EXISTS)
>       |                                       ^~~~~~~~~~~~~~~~
> util/bpf_skel/sample_filter.bpf.c:167:29: note: in expansion of macro 'bpf_core_field_exists'
>   167 |                         if (bpf_core_field_exists(data->mem_hops))
>       |                             ^~~~~~~~~~~~~~~~~~~~~
> cc1: error: argument is not a field access
> ```
> 
> ___bpf_field_ref1 was adapted for GCC in 12bbcf8e840f40b82b02981e96e0a5fbb0703ea9
> but the trick added for compatibility in 3a8b8fc3174891c4c12f5766d82184a82d4b2e3e
> isn't compatible with that as an address is used as an argument.
> 
> Workaround this by calling __builtin_preserve_field_info directly as the
> bpf_core_field_exists macro does, but without the ___bpf_field_ref use.

IIUC GCC doesn't support bpf_core_fields_exists() for bitfield members,
right?  Is it gonna change in the future?

> 
> Link: https://gcc.gnu.org/PR121420
> Co-authored-by: Andrew Pinski <quic_apinski@quicinc.com>
> Signed-off-by: Sam James <sam@gentoo.org>
> ---
>  tools/perf/util/bpf_skel/sample_filter.bpf.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/tools/perf/util/bpf_skel/sample_filter.bpf.c b/tools/perf/util/bpf_skel/sample_filter.bpf.c
> index b195e6efeb8be..e5666d4c17228 100644
> --- a/tools/perf/util/bpf_skel/sample_filter.bpf.c
> +++ b/tools/perf/util/bpf_skel/sample_filter.bpf.c
> @@ -164,7 +164,7 @@ static inline __u64 perf_get_sample(struct bpf_perf_event_data_kern *kctx,
>  		if (entry->part == 8) {
>  			union perf_mem_data_src___new *data = (void *)&kctx->data->data_src;
>  
> -			if (bpf_core_field_exists(data->mem_hops))
> +			if (__builtin_preserve_field_info(data->mem_hops, BPF_FIELD_EXISTS))

I believe those two are equivalent (maybe worth a comment?).  But it'd
be great if BPF/clang folks can review if it's ok.

Anyway, I can build it with clang.

Tested-by: Namhyung Kim <namhyung@kernel.org>

Thanks,
Namhyung


>  				return data->mem_hops;
>  
>  			return 0;
> -- 
> 2.50.1
> 

RE: [PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[Andrew Pinski]

> -----Original Message-----
> From: Namhyung Kim <namhyung@kernel.org>
> Sent: Wednesday, August 6, 2025 4:34 PM
> To: Sam James <sam@gentoo.org>
> Cc: Peter Zijlstra <peterz@infradead.org>; Ingo Molnar
> <mingo@redhat.com>; Arnaldo Carvalho de Melo
> <acme@kernel.org>; Mark Rutland
> <mark.rutland@arm.com>; Alexander Shishkin
> <alexander.shishkin@linux.intel.com>; Jiri Olsa
> <jolsa@kernel.org>; Ian Rogers <irogers@google.com>; Adrian
> Hunter <adrian.hunter@intel.com>; Liang, Kan
> <kan.liang@linux.intel.com>; Andrew Pinski
> <quic_apinski@quicinc.com>; linux-perf-
> users@vger.kernel.org; linux-kernel@vger.kernel.org;
> bpf@vger.kernel.org
> Subject: Re: [PATCH] perf: use __builtin_preserve_field_info
> for GCC compatibility
> 
> Hello,
> 
> On Wed, Aug 06, 2025 at 01:03:01AM +0100, Sam James
> wrote:
> > When exploring building bpf_skel with GCC's BPF support,
> there was a
> > buid failure because of bpf_core_field_exists vs the
> mem_hops bitfield:
> > ```
> >  In file included from util/bpf_skel/sample_filter.bpf.c:6:
> > util/bpf_skel/sample_filter.bpf.c: In function
> 'perf_get_sample':
> > tools/perf/libbpf/include/bpf/bpf_core_read.h:169:42:
> error: cannot take address of bit-field 'mem_hops'
> >   169 | #define ___bpf_field_ref1(field)        (&(field))
> >       |                                          ^
> > tools/perf/libbpf/include/bpf/bpf_helpers.h:222:29: note: in
> expansion of macro '___bpf_field_ref1'
> >   222 | #define ___bpf_concat(a, b) a ## b
> >       |                             ^
> > tools/perf/libbpf/include/bpf/bpf_helpers.h:225:29: note: in
> expansion of macro '___bpf_concat'
> >   225 | #define ___bpf_apply(fn, n) ___bpf_concat(fn, n)
> >       |                             ^~~~~~~~~~~~~
> > tools/perf/libbpf/include/bpf/bpf_core_read.h:173:9: note:
> in expansion of macro '___bpf_apply'
> >   173 |         ___bpf_apply(___bpf_field_ref,
> ___bpf_narg(args))(args)
> >       |         ^~~~~~~~~~~~
> > tools/perf/libbpf/include/bpf/bpf_core_read.h:188:39: note:
> in expansion of macro '___bpf_field_ref'
> >   188 |
> __builtin_preserve_field_info(___bpf_field_ref(field),
> BPF_FIELD_EXISTS)
> >       |                                       ^~~~~~~~~~~~~~~~
> > util/bpf_skel/sample_filter.bpf.c:167:29: note: in expansion
> of macro 'bpf_core_field_exists'
> >   167 |                         if (bpf_core_field_exists(data-
> >mem_hops))
> >       |                             ^~~~~~~~~~~~~~~~~~~~~
> > cc1: error: argument is not a field access ```
> >
> > ___bpf_field_ref1 was adapted for GCC in
> > 12bbcf8e840f40b82b02981e96e0a5fbb0703ea9
> > but the trick added for compatibility in
> > 3a8b8fc3174891c4c12f5766d82184a82d4b2e3e
> > isn't compatible with that as an address is used as an
> argument.
> >
> > Workaround this by calling __builtin_preserve_field_info
> directly as
> > the bpf_core_field_exists macro does, but without the
> ___bpf_field_ref use.
> 
> IIUC GCC doesn't support bpf_core_fields_exists() for bitfield
> members, right?  Is it gonna change in the future?

Let's discuss how __builtin_preserve_field_info is handled in the first place for BPF. Right now it seems it is passed some expression as the first argument is never evaluated.
The problem is GCC's implementation of __builtin_preserve_field_info is all in the backend and the front end does not understand of the special rules here.

GCC implements some "special" builtins in the front-end but not by the normal function call rules but parsing them separately; this is how __builtin_offsetof and a few others are implemented in both the C and C++ front-ends (and implemented separately). Now we could have add a hook to allow a backend to something similar and maybe that is the best way forward here.
But it won't be __builtin_preserve_field_info but rather `__builtin_preserve_field_type_info(type,field,kind)` instead. 

__builtin_preserve_enum_type_value would also be added with the following:
__builtin_preserve_enum_type_value(enum_type, enum_value, kind) 

And change all of the rest of the builtins to accept a true type argument rather than having to cast an null pointer to that type.

Will clang implement a similar builtin?

Note this won't be done until at least GCC 16; maybe not until GCC 17 depending on if I or someone else gets time to implement the front-end parts which is acceptable to both the C and C++ front-ends.

Thanks,
Andrew Pinski

> 
> >
> > Link: https://gcc.gnu.org/PR121420
> > Co-authored-by: Andrew Pinski
> <quic_apinski@quicinc.com>
> > Signed-off-by: Sam James <sam@gentoo.org>
> > ---
> >  tools/perf/util/bpf_skel/sample_filter.bpf.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/tools/perf/util/bpf_skel/sample_filter.bpf.c
> > b/tools/perf/util/bpf_skel/sample_filter.bpf.c
> > index b195e6efeb8be..e5666d4c17228 100644
> > --- a/tools/perf/util/bpf_skel/sample_filter.bpf.c
> > +++ b/tools/perf/util/bpf_skel/sample_filter.bpf.c
> > @@ -164,7 +164,7 @@ static inline __u64
> perf_get_sample(struct bpf_perf_event_data_kern *kctx,
> >               if (entry->part == 8) {
> >                       union perf_mem_data_src___new *data = (void
> > *)&kctx->data->data_src;
> >
> > -                     if (bpf_core_field_exists(data->mem_hops))
> > +                     if
> > + (__builtin_preserve_field_info(data->mem_hops,
> BPF_FIELD_EXISTS))
> 
> I believe those two are equivalent (maybe worth a
> comment?).  But it'd be great if BPF/clang folks can review if
> it's ok.
> 
> Anyway, I can build it with clang.
> 
> Tested-by: Namhyung Kim <namhyung@kernel.org>
> 
> Thanks,
> Namhyung
> 
> 
> >                               return data->mem_hops;
> >
> >                       return 0;
> > --
> > 2.50.1
> >

Re: [PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[Yonghong Song]

On 8/6/25 4:33 PM, Namhyung Kim wrote:
> Hello,
>
> On Wed, Aug 06, 2025 at 01:03:01AM +0100, Sam James wrote:
>> When exploring building bpf_skel with GCC's BPF support, there was a
>> buid failure because of bpf_core_field_exists vs the mem_hops bitfield:
>> ```
>>   In file included from util/bpf_skel/sample_filter.bpf.c:6:
>> util/bpf_skel/sample_filter.bpf.c: In function 'perf_get_sample':
>> tools/perf/libbpf/include/bpf/bpf_core_read.h:169:42: error: cannot take address of bit-field 'mem_hops'
>>    169 | #define ___bpf_field_ref1(field)        (&(field))
>>        |                                          ^
>> tools/perf/libbpf/include/bpf/bpf_helpers.h:222:29: note: in expansion of macro '___bpf_field_ref1'
>>    222 | #define ___bpf_concat(a, b) a ## b
>>        |                             ^
>> tools/perf/libbpf/include/bpf/bpf_helpers.h:225:29: note: in expansion of macro '___bpf_concat'
>>    225 | #define ___bpf_apply(fn, n) ___bpf_concat(fn, n)
>>        |                             ^~~~~~~~~~~~~
>> tools/perf/libbpf/include/bpf/bpf_core_read.h:173:9: note: in expansion of macro '___bpf_apply'
>>    173 |         ___bpf_apply(___bpf_field_ref, ___bpf_narg(args))(args)
>>        |         ^~~~~~~~~~~~
>> tools/perf/libbpf/include/bpf/bpf_core_read.h:188:39: note: in expansion of macro '___bpf_field_ref'
>>    188 |         __builtin_preserve_field_info(___bpf_field_ref(field), BPF_FIELD_EXISTS)
>>        |                                       ^~~~~~~~~~~~~~~~
>> util/bpf_skel/sample_filter.bpf.c:167:29: note: in expansion of macro 'bpf_core_field_exists'
>>    167 |                         if (bpf_core_field_exists(data->mem_hops))
>>        |                             ^~~~~~~~~~~~~~~~~~~~~
>> cc1: error: argument is not a field access
>> ```
>>
>> ___bpf_field_ref1 was adapted for GCC in 12bbcf8e840f40b82b02981e96e0a5fbb0703ea9
>> but the trick added for compatibility in 3a8b8fc3174891c4c12f5766d82184a82d4b2e3e
>> isn't compatible with that as an address is used as an argument.
>>
>> Workaround this by calling __builtin_preserve_field_info directly as the
>> bpf_core_field_exists macro does, but without the ___bpf_field_ref use.
> IIUC GCC doesn't support bpf_core_fields_exists() for bitfield members,
> right?  Is it gonna change in the future?
>
>> Link: https://gcc.gnu.org/PR121420
>> Co-authored-by: Andrew Pinski <quic_apinski@quicinc.com>
>> Signed-off-by: Sam James <sam@gentoo.org>
>> ---
>>   tools/perf/util/bpf_skel/sample_filter.bpf.c | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/tools/perf/util/bpf_skel/sample_filter.bpf.c b/tools/perf/util/bpf_skel/sample_filter.bpf.c
>> index b195e6efeb8be..e5666d4c17228 100644
>> --- a/tools/perf/util/bpf_skel/sample_filter.bpf.c
>> +++ b/tools/perf/util/bpf_skel/sample_filter.bpf.c
>> @@ -164,7 +164,7 @@ static inline __u64 perf_get_sample(struct bpf_perf_event_data_kern *kctx,
>>   		if (entry->part == 8) {
>>   			union perf_mem_data_src___new *data = (void *)&kctx->data->data_src;
>>   
>> -			if (bpf_core_field_exists(data->mem_hops))
>> +			if (__builtin_preserve_field_info(data->mem_hops, BPF_FIELD_EXISTS))
> I believe those two are equivalent (maybe worth a comment?).  But it'd
> be great if BPF/clang folks can review if it's ok.

Yes, from clang side, they are almost equnivalent. See tools/lib/bpf/bpf_core_read.h.

#define bpf_core_field_exists(field...)                                     \
         __builtin_preserve_field_info(___bpf_field_ref(field), BPF_FIELD_EXISTS)

bpf_core_field_exists actually relies on clang builtin function
__builtin_preserve_field_info(). This builtin is handled in frontend and
also at early IR stage.

So your above code is okay to me although bpf_core_field_exists() is much
easy to understand the intent.

>
> Anyway, I can build it with clang.
>
> Tested-by: Namhyung Kim <namhyung@kernel.org>
>
> Thanks,
> Namhyung
>
>
>>   				return data->mem_hops;
>>   
>>   			return 0;
>> -- 
>> 2.50.1
>>

Re: [PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[Yonghong Song]

On 8/6/25 4:57 PM, Andrew Pinski wrote:
>
>> -----Original Message-----
>> From: Namhyung Kim <namhyung@kernel.org>
>> Sent: Wednesday, August 6, 2025 4:34 PM
>> To: Sam James <sam@gentoo.org>
>> Cc: Peter Zijlstra <peterz@infradead.org>; Ingo Molnar
>> <mingo@redhat.com>; Arnaldo Carvalho de Melo
>> <acme@kernel.org>; Mark Rutland
>> <mark.rutland@arm.com>; Alexander Shishkin
>> <alexander.shishkin@linux.intel.com>; Jiri Olsa
>> <jolsa@kernel.org>; Ian Rogers <irogers@google.com>; Adrian
>> Hunter <adrian.hunter@intel.com>; Liang, Kan
>> <kan.liang@linux.intel.com>; Andrew Pinski
>> <quic_apinski@quicinc.com>; linux-perf-
>> users@vger.kernel.org; linux-kernel@vger.kernel.org;
>> bpf@vger.kernel.org
>> Subject: Re: [PATCH] perf: use __builtin_preserve_field_info
>> for GCC compatibility
>>
>> Hello,
>>
>> On Wed, Aug 06, 2025 at 01:03:01AM +0100, Sam James
>> wrote:
>>> When exploring building bpf_skel with GCC's BPF support,
>> there was a
>>> buid failure because of bpf_core_field_exists vs the
>> mem_hops bitfield:
>>> ```
>>>   In file included from util/bpf_skel/sample_filter.bpf.c:6:
>>> util/bpf_skel/sample_filter.bpf.c: In function
>> 'perf_get_sample':
>>> tools/perf/libbpf/include/bpf/bpf_core_read.h:169:42:
>> error: cannot take address of bit-field 'mem_hops'
>>>    169 | #define ___bpf_field_ref1(field)        (&(field))
>>>        |                                          ^
>>> tools/perf/libbpf/include/bpf/bpf_helpers.h:222:29: note: in
>> expansion of macro '___bpf_field_ref1'
>>>    222 | #define ___bpf_concat(a, b) a ## b
>>>        |                             ^
>>> tools/perf/libbpf/include/bpf/bpf_helpers.h:225:29: note: in
>> expansion of macro '___bpf_concat'
>>>    225 | #define ___bpf_apply(fn, n) ___bpf_concat(fn, n)
>>>        |                             ^~~~~~~~~~~~~
>>> tools/perf/libbpf/include/bpf/bpf_core_read.h:173:9: note:
>> in expansion of macro '___bpf_apply'
>>>    173 |         ___bpf_apply(___bpf_field_ref,
>> ___bpf_narg(args))(args)
>>>        |         ^~~~~~~~~~~~
>>> tools/perf/libbpf/include/bpf/bpf_core_read.h:188:39: note:
>> in expansion of macro '___bpf_field_ref'
>>>    188 |
>> __builtin_preserve_field_info(___bpf_field_ref(field),
>> BPF_FIELD_EXISTS)
>>>        |                                       ^~~~~~~~~~~~~~~~
>>> util/bpf_skel/sample_filter.bpf.c:167:29: note: in expansion
>> of macro 'bpf_core_field_exists'
>>>    167 |                         if (bpf_core_field_exists(data-
>>> mem_hops))
>>>        |                             ^~~~~~~~~~~~~~~~~~~~~
>>> cc1: error: argument is not a field access ```
>>>
>>> ___bpf_field_ref1 was adapted for GCC in
>>> 12bbcf8e840f40b82b02981e96e0a5fbb0703ea9
>>> but the trick added for compatibility in
>>> 3a8b8fc3174891c4c12f5766d82184a82d4b2e3e
>>> isn't compatible with that as an address is used as an
>> argument.
>>> Workaround this by calling __builtin_preserve_field_info
>> directly as
>>> the bpf_core_field_exists macro does, but without the
>> ___bpf_field_ref use.
>>
>> IIUC GCC doesn't support bpf_core_fields_exists() for bitfield
>> members, right?  Is it gonna change in the future?
> Let's discuss how __builtin_preserve_field_info is handled in the first place for BPF. Right now it seems it is passed some expression as the first argument is never evaluated.
> The problem is GCC's implementation of __builtin_preserve_field_info is all in the backend and the front end does not understand of the special rules here.
>
> GCC implements some "special" builtins in the front-end but not by the normal function call rules but parsing them separately; this is how __builtin_offsetof and a few others are implemented in both the C and C++ front-ends (and implemented separately). Now we could have add a hook to allow a backend to something similar and maybe that is the best way forward here.
> But it won't be __builtin_preserve_field_info but rather `__builtin_preserve_field_type_info(type,field,kind)` instead.
>
> __builtin_preserve_enum_type_value would also be added with the following:
> __builtin_preserve_enum_type_value(enum_type, enum_value, kind)
>
> And change all of the rest of the builtins to accept a true type argument rather than having to cast an null pointer to that type.
>
> Will clang implement a similar builtin?

The clang only has one builtin for some related relocations:
    
    __builtin_preserve_field_info(..., BPF_FIELD_EXISTS)
    __builtin_preserve_field_info(..., BPF_FIELD_BYTE_OFFSET)
    ...
They are all used in bpf_core_read.h.

>
> Note this won't be done until at least GCC 16; maybe not until GCC 17 depending on if I or someone else gets time to implement the front-end parts which is acceptable to both the C and C++ front-ends.
>
> Thanks,
> Andrew Pinski
>
>>> Link: https://gcc.gnu.org/PR121420
>>> Co-authored-by: Andrew Pinski
>> <quic_apinski@quicinc.com>
>>> Signed-off-by: Sam James <sam@gentoo.org>
>>> ---
>>>   tools/perf/util/bpf_skel/sample_filter.bpf.c | 2 +-
>>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/tools/perf/util/bpf_skel/sample_filter.bpf.c
>>> b/tools/perf/util/bpf_skel/sample_filter.bpf.c
>>> index b195e6efeb8be..e5666d4c17228 100644
>>> --- a/tools/perf/util/bpf_skel/sample_filter.bpf.c
>>> +++ b/tools/perf/util/bpf_skel/sample_filter.bpf.c
>>> @@ -164,7 +164,7 @@ static inline __u64
>> perf_get_sample(struct bpf_perf_event_data_kern *kctx,
>>>                if (entry->part == 8) {
>>>                        union perf_mem_data_src___new *data = (void
>>> *)&kctx->data->data_src;
>>>
>>> -                     if (bpf_core_field_exists(data->mem_hops))
>>> +                     if
>>> + (__builtin_preserve_field_info(data->mem_hops,
>> BPF_FIELD_EXISTS))
>>
>> I believe those two are equivalent (maybe worth a
>> comment?).  But it'd be great if BPF/clang folks can review if
>> it's ok.
>>
>> Anyway, I can build it with clang.
>>
>> Tested-by: Namhyung Kim <namhyung@kernel.org>
>>
>> Thanks,
>> Namhyung
>>
>>
>>>                                return data->mem_hops;
>>>
>>>                        return 0;
>>> --
>>> 2.50.1
>>>

Re: [PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[Arnaldo Carvalho de Melo]

On Wed, Aug 06, 2025 at 05:27:02PM -0700, Yonghong Song wrote:
> On 8/6/25 4:57 PM, Andrew Pinski wrote:
> > > -----Original Message-----
> > > From: Namhyung Kim <namhyung@kernel.org>
> > > Sent: Wednesday, August 6, 2025 4:34 PM
> > > To: Sam James <sam@gentoo.org>
> > > Cc: Peter Zijlstra <peterz@infradead.org>; Ingo Molnar
> > > <mingo@redhat.com>; Arnaldo Carvalho de Melo
> > > <acme@kernel.org>; Mark Rutland
> > > <mark.rutland@arm.com>; Alexander Shishkin
> > > <alexander.shishkin@linux.intel.com>; Jiri Olsa
> > > <jolsa@kernel.org>; Ian Rogers <irogers@google.com>; Adrian
> > > Hunter <adrian.hunter@intel.com>; Liang, Kan
> > > <kan.liang@linux.intel.com>; Andrew Pinski
> > > <quic_apinski@quicinc.com>; linux-perf-
> > > users@vger.kernel.org; linux-kernel@vger.kernel.org;
> > > bpf@vger.kernel.org
> > > Subject: Re: [PATCH] perf: use __builtin_preserve_field_info
> > > for GCC compatibility
> > > 
> > > Hello,
> > > 
> > > On Wed, Aug 06, 2025 at 01:03:01AM +0100, Sam James
> > > wrote:
> > > > When exploring building bpf_skel with GCC's BPF support,
> > > there was a
> > > > buid failure because of bpf_core_field_exists vs the
> > > mem_hops bitfield:
> > > > ```
> > > >   In file included from util/bpf_skel/sample_filter.bpf.c:6:
> > > > util/bpf_skel/sample_filter.bpf.c: In function
> > > 'perf_get_sample':
> > > > tools/perf/libbpf/include/bpf/bpf_core_read.h:169:42:
> > > error: cannot take address of bit-field 'mem_hops'
> > > >    169 | #define ___bpf_field_ref1(field)        (&(field))
> > > >        |                                          ^
> > > > tools/perf/libbpf/include/bpf/bpf_helpers.h:222:29: note: in
> > > expansion of macro '___bpf_field_ref1'
> > > >    222 | #define ___bpf_concat(a, b) a ## b
> > > >        |                             ^
> > > > tools/perf/libbpf/include/bpf/bpf_helpers.h:225:29: note: in
> > > expansion of macro '___bpf_concat'
> > > >    225 | #define ___bpf_apply(fn, n) ___bpf_concat(fn, n)
> > > >        |                             ^~~~~~~~~~~~~
> > > > tools/perf/libbpf/include/bpf/bpf_core_read.h:173:9: note:
> > > in expansion of macro '___bpf_apply'
> > > >    173 |         ___bpf_apply(___bpf_field_ref,
> > > ___bpf_narg(args))(args)
> > > >        |         ^~~~~~~~~~~~
> > > > tools/perf/libbpf/include/bpf/bpf_core_read.h:188:39: note:
> > > in expansion of macro '___bpf_field_ref'
> > > >    188 |
> > > __builtin_preserve_field_info(___bpf_field_ref(field),
> > > BPF_FIELD_EXISTS)
> > > >        |                                       ^~~~~~~~~~~~~~~~
> > > > util/bpf_skel/sample_filter.bpf.c:167:29: note: in expansion
> > > of macro 'bpf_core_field_exists'
> > > >    167 |                         if (bpf_core_field_exists(data-
> > > > mem_hops))
> > > >        |                             ^~~~~~~~~~~~~~~~~~~~~
> > > > cc1: error: argument is not a field access ```
> > > > 
> > > > ___bpf_field_ref1 was adapted for GCC in
> > > > 12bbcf8e840f40b82b02981e96e0a5fbb0703ea9
> > > > but the trick added for compatibility in
> > > > 3a8b8fc3174891c4c12f5766d82184a82d4b2e3e
> > > > isn't compatible with that as an address is used as an
> > > argument.
> > > > Workaround this by calling __builtin_preserve_field_info
> > > directly as
> > > > the bpf_core_field_exists macro does, but without the
> > > ___bpf_field_ref use.
> > > 
> > > IIUC GCC doesn't support bpf_core_fields_exists() for bitfield
> > > members, right?  Is it gonna change in the future?
> > Let's discuss how __builtin_preserve_field_info is handled in the first place for BPF. Right now it seems it is passed some expression as the first argument is never evaluated.
> > The problem is GCC's implementation of __builtin_preserve_field_info is all in the backend and the front end does not understand of the special rules here.
> > 
> > GCC implements some "special" builtins in the front-end but not by the normal function call rules but parsing them separately; this is how __builtin_offsetof and a few others are implemented in both the C and C++ front-ends (and implemented separately). Now we could have add a hook to allow a backend to something similar and maybe that is the best way forward here.
> > But it won't be __builtin_preserve_field_info but rather `__builtin_preserve_field_type_info(type,field,kind)` instead.
> > 
> > __builtin_preserve_enum_type_value would also be added with the following:
> > __builtin_preserve_enum_type_value(enum_type, enum_value, kind)
> > 
> > And change all of the rest of the builtins to accept a true type argument rather than having to cast an null pointer to that type.
> > 
> > Will clang implement a similar builtin?
> 
> The clang only has one builtin for some related relocations:
>    __builtin_preserve_field_info(..., BPF_FIELD_EXISTS)
>    __builtin_preserve_field_info(..., BPF_FIELD_BYTE_OFFSET)
>    ...
> They are all used in bpf_core_read.h.
> 
> > 
> > Note this won't be done until at least GCC 16; maybe not until GCC 17 depending on if I or someone else gets time to implement the front-end parts which is acceptable to both the C and C++ front-ends.

So I'm taking the patch as-is, ok?

But first we need the Signed-off-by tag from Andrew Pinski as he is
listed in a Co-authored-by, that I replaced with Co-developed-by as its
the term used for this purpose in:

Yonghong, can I add an Acked-by: you since you participated in this
discussion agreeing with the original patch (If I'm not mistaken)?

RE: [PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[Andrew Pinski]

> -----Original Message-----
> From: Arnaldo Carvalho de Melo <acme@kernel.org>
> Sent: Thursday, October 2, 2025 10:32 AM
> To: Yonghong Song <yonghong.song@linux.dev>
> Cc: Andrew Pinski (QUIC) <quic_apinski@quicinc.com>;
> Namhyung Kim <namhyung@kernel.org>; Sam James
> <sam@gentoo.org>; Peter Zijlstra <peterz@infradead.org>;
> Ingo Molnar <mingo@redhat.com>; Mark Rutland
> <mark.rutland@arm.com>; Alexander Shishkin
> <alexander.shishkin@linux.intel.com>; Jiri Olsa
> <jolsa@kernel.org>; Ian Rogers <irogers@google.com>; Adrian
> Hunter <adrian.hunter@intel.com>; Liang, Kan
> <kan.liang@linux.intel.com>; linux-perf-
> users@vger.kernel.org; linux-kernel@vger.kernel.org;
> bpf@vger.kernel.org
> Subject: Re: [PATCH] perf: use __builtin_preserve_field_info
> for GCC compatibility
> 
> On Wed, Aug 06, 2025 at 05:27:02PM -0700, Yonghong Song
> wrote:
> > On 8/6/25 4:57 PM, Andrew Pinski wrote:
> > > > -----Original Message-----
> > > > From: Namhyung Kim <namhyung@kernel.org>
> > > > Sent: Wednesday, August 6, 2025 4:34 PM
> > > > To: Sam James <sam@gentoo.org>
> > > > Cc: Peter Zijlstra <peterz@infradead.org>; Ingo Molnar
> > > > <mingo@redhat.com>; Arnaldo Carvalho de Melo
> <acme@kernel.org>;
> > > > Mark Rutland <mark.rutland@arm.com>; Alexander
> Shishkin
> > > > <alexander.shishkin@linux.intel.com>; Jiri Olsa
> > > > <jolsa@kernel.org>; Ian Rogers <irogers@google.com>;
> Adrian Hunter
> > > > <adrian.hunter@intel.com>; Liang, Kan
> <kan.liang@linux.intel.com>;
> > > > Andrew Pinski <quic_apinski@quicinc.com>; linux-perf-
> > > > users@vger.kernel.org; linux-kernel@vger.kernel.org;
> > > > bpf@vger.kernel.org
> > > > Subject: Re: [PATCH] perf: use
> __builtin_preserve_field_info for
> > > > GCC compatibility
> > > >
> > > > Hello,
> > > >
> > > > On Wed, Aug 06, 2025 at 01:03:01AM +0100, Sam James
> > > > wrote:
> > > > > When exploring building bpf_skel with GCC's BPF
> support,
> > > > there was a
> > > > > buid failure because of bpf_core_field_exists vs the
> > > > mem_hops bitfield:
> > > > > ```
> > > > >   In file included from
> util/bpf_skel/sample_filter.bpf.c:6:
> > > > > util/bpf_skel/sample_filter.bpf.c: In function
> > > > 'perf_get_sample':
> > > > > tools/perf/libbpf/include/bpf/bpf_core_read.h:169:42:
> > > > error: cannot take address of bit-field 'mem_hops'
> > > > >    169 | #define ___bpf_field_ref1(field)        (&(field))
> > > > >        |                                          ^
> > > > > tools/perf/libbpf/include/bpf/bpf_helpers.h:222:29:
> note: in
> > > > expansion of macro '___bpf_field_ref1'
> > > > >    222 | #define ___bpf_concat(a, b) a ## b
> > > > >        |                             ^
> > > > > tools/perf/libbpf/include/bpf/bpf_helpers.h:225:29:
> note: in
> > > > expansion of macro '___bpf_concat'
> > > > >    225 | #define ___bpf_apply(fn, n) ___bpf_concat(fn,
> n)
> > > > >        |                             ^~~~~~~~~~~~~
> > > > > tools/perf/libbpf/include/bpf/bpf_core_read.h:173:9:
> note:
> > > > in expansion of macro '___bpf_apply'
> > > > >    173 |         ___bpf_apply(___bpf_field_ref,
> > > > ___bpf_narg(args))(args)
> > > > >        |         ^~~~~~~~~~~~
> > > > > tools/perf/libbpf/include/bpf/bpf_core_read.h:188:39:
> note:
> > > > in expansion of macro '___bpf_field_ref'
> > > > >    188 |
> > > > __builtin_preserve_field_info(___bpf_field_ref(field),
> > > > BPF_FIELD_EXISTS)
> > > > >        |                                       ^~~~~~~~~~~~~~~~
> > > > > util/bpf_skel/sample_filter.bpf.c:167:29: note: in
> expansion
> > > > of macro 'bpf_core_field_exists'
> > > > >    167 |                         if (bpf_core_field_exists(data-
> > > > > mem_hops))
> > > > >        |                             ^~~~~~~~~~~~~~~~~~~~~
> > > > > cc1: error: argument is not a field access ```
> > > > >
> > > > > ___bpf_field_ref1 was adapted for GCC in
> > > > > 12bbcf8e840f40b82b02981e96e0a5fbb0703ea9
> > > > > but the trick added for compatibility in
> > > > > 3a8b8fc3174891c4c12f5766d82184a82d4b2e3e
> > > > > isn't compatible with that as an address is used as an
> > > > argument.
> > > > > Workaround this by calling
> __builtin_preserve_field_info
> > > > directly as
> > > > > the bpf_core_field_exists macro does, but without the
> > > > ___bpf_field_ref use.
> > > >
> > > > IIUC GCC doesn't support bpf_core_fields_exists() for
> bitfield
> > > > members, right?  Is it gonna change in the future?
> > > Let's discuss how __builtin_preserve_field_info is handled
> in the first place for BPF. Right now it seems it is passed some
> expression as the first argument is never evaluated.
> > > The problem is GCC's implementation of
> __builtin_preserve_field_info is all in the backend and the
> front end does not understand of the special rules here.
> > >
> > > GCC implements some "special" builtins in the front-end
> but not by the normal function call rules but parsing them
> separately; this is how __builtin_offsetof and a few others are
> implemented in both the C and C++ front-ends (and
> implemented separately). Now we could have add a hook to
> allow a backend to something similar and maybe that is the
> best way forward here.
> > > But it won't be __builtin_preserve_field_info but rather
> `__builtin_preserve_field_type_info(type,field,kind)` instead.
> > >
> > > __builtin_preserve_enum_type_value would also be
> added with the following:
> > > __builtin_preserve_enum_type_value(enum_type,
> enum_value, kind)
> > >
> > > And change all of the rest of the builtins to accept a true
> type argument rather than having to cast an null pointer to
> that type.
> > >
> > > Will clang implement a similar builtin?
> >
> > The clang only has one builtin for some related relocations:
> >    __builtin_preserve_field_info(..., BPF_FIELD_EXISTS)
> >    __builtin_preserve_field_info(...,
> BPF_FIELD_BYTE_OFFSET)
> >    ...
> > They are all used in bpf_core_read.h.
> >
> > >
> > > Note this won't be done until at least GCC 16; maybe not
> until GCC 17 depending on if I or someone else gets time to
> implement the front-end parts which is acceptable to both the
> C and C++ front-ends.
> 
> So I'm taking the patch as-is, ok?
> 
> But first we need the Signed-off-by tag from Andrew Pinski as
> he is listed in a Co-authored-by, that I replaced with Co-
> developed-by as its the term used for this purpose in:
> 
> Yonghong, can I add an Acked-by: you since you participated in
> this discussion agreeing with the original patch (If I'm not
> mistaken)?

Signed-off-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>

Note my email address for doing Linux and GCC development is Andrew.pinski@oss.qualcomm.com . It was apinski_quic@quicinc.com but that changed last month.

Thanks,
Andrew Pinski


> 

Re: [PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[Arnaldo Carvalho de Melo]

On Thu, Oct 02, 2025 at 05:35:09PM +0000, Andrew Pinski wrote:
> > From: Arnaldo Carvalho de Melo <acme@kernel.org>
> > So I'm taking the patch as-is, ok?

> > But first we need the Signed-off-by tag from Andrew Pinski as
> > he is listed in a Co-authored-by, that I replaced with Co-
> > developed-by as its the term used for this purpose in:

> > Yonghong, can I add an Acked-by: you since you participated in
> > this discussion agreeing with the original patch (If I'm not
> > mistaken)?

> Signed-off-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>
 
> Note my email address for doing Linux and GCC development is
> Andrew.pinski@oss.qualcomm.com . It was apinski_quic@quicinc.com but
> that changed last month.

This is what I have in the patch now:

Co-developed-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>
Signed-off-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>

- Arnaldo

Re: [PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[Sam James]

Arnaldo Carvalho de Melo <acme@kernel.org> writes:

> On Thu, Oct 02, 2025 at 05:35:09PM +0000, Andrew Pinski wrote:
>> > From: Arnaldo Carvalho de Melo <acme@kernel.org>
>> > So I'm taking the patch as-is, ok?
>
>> > But first we need the Signed-off-by tag from Andrew Pinski as
>> > he is listed in a Co-authored-by, that I replaced with Co-
>> > developed-by as its the term used for this purpose in:
>
>> > Yonghong, can I add an Acked-by: you since you participated in
>> > this discussion agreeing with the original patch (If I'm not
>> > mistaken)?
>
>> Signed-off-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>
>  
>> Note my email address for doing Linux and GCC development is
>> Andrew.pinski@oss.qualcomm.com . It was apinski_quic@quicinc.com but
>> that changed last month.
>
> This is what I have in the patch now:
>
> Co-developed-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>
> Signed-off-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>

Looks good if Andrew is happy. Thanks!

>
> - Arnaldo

Re: [PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[Andrew Pinski]

On Thu, Oct 2, 2025 at 11:01 AM Sam James <sam@gentoo.org> wrote:
>
> Arnaldo Carvalho de Melo <acme@kernel.org> writes:
>
> > On Thu, Oct 02, 2025 at 05:35:09PM +0000, Andrew Pinski wrote:
> >> > From: Arnaldo Carvalho de Melo <acme@kernel.org>
> >> > So I'm taking the patch as-is, ok?
> >
> >> > But first we need the Signed-off-by tag from Andrew Pinski as
> >> > he is listed in a Co-authored-by, that I replaced with Co-
> >> > developed-by as its the term used for this purpose in:
> >
> >> > Yonghong, can I add an Acked-by: you since you participated in
> >> > this discussion agreeing with the original patch (If I'm not
> >> > mistaken)?
> >
> >> Signed-off-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>
> >
> >> Note my email address for doing Linux and GCC development is
> >> Andrew.pinski@oss.qualcomm.com . It was apinski_quic@quicinc.com but
> >> that changed last month.
> >
> > This is what I have in the patch now:
> >
> > Co-developed-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>
> > Signed-off-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>
>
> Looks good if Andrew is happy. Thanks!

I am.

>
> >
> > - Arnaldo

Re: [PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[Yonghong Song]

On 10/2/25 10:31 AM, Arnaldo Carvalho de Melo wrote:
> On Wed, Aug 06, 2025 at 05:27:02PM -0700, Yonghong Song wrote:
>> On 8/6/25 4:57 PM, Andrew Pinski wrote:
>>>> -----Original Message-----
>>>> From: Namhyung Kim <namhyung@kernel.org>
>>>> Sent: Wednesday, August 6, 2025 4:34 PM
>>>> To: Sam James <sam@gentoo.org>
>>>> Cc: Peter Zijlstra <peterz@infradead.org>; Ingo Molnar
>>>> <mingo@redhat.com>; Arnaldo Carvalho de Melo
>>>> <acme@kernel.org>; Mark Rutland
>>>> <mark.rutland@arm.com>; Alexander Shishkin
>>>> <alexander.shishkin@linux.intel.com>; Jiri Olsa
>>>> <jolsa@kernel.org>; Ian Rogers <irogers@google.com>; Adrian
>>>> Hunter <adrian.hunter@intel.com>; Liang, Kan
>>>> <kan.liang@linux.intel.com>; Andrew Pinski
>>>> <quic_apinski@quicinc.com>; linux-perf-
>>>> users@vger.kernel.org; linux-kernel@vger.kernel.org;
>>>> bpf@vger.kernel.org
>>>> Subject: Re: [PATCH] perf: use __builtin_preserve_field_info
>>>> for GCC compatibility
>>>>
>>>> Hello,
>>>>
>>>> On Wed, Aug 06, 2025 at 01:03:01AM +0100, Sam James
>>>> wrote:
>>>>> When exploring building bpf_skel with GCC's BPF support,
>>>> there was a
>>>>> buid failure because of bpf_core_field_exists vs the
>>>> mem_hops bitfield:
>>>>> ```
>>>>>    In file included from util/bpf_skel/sample_filter.bpf.c:6:
>>>>> util/bpf_skel/sample_filter.bpf.c: In function
>>>> 'perf_get_sample':
>>>>> tools/perf/libbpf/include/bpf/bpf_core_read.h:169:42:
>>>> error: cannot take address of bit-field 'mem_hops'
>>>>>     169 | #define ___bpf_field_ref1(field)        (&(field))
>>>>>         |                                          ^
>>>>> tools/perf/libbpf/include/bpf/bpf_helpers.h:222:29: note: in
>>>> expansion of macro '___bpf_field_ref1'
>>>>>     222 | #define ___bpf_concat(a, b) a ## b
>>>>>         |                             ^
>>>>> tools/perf/libbpf/include/bpf/bpf_helpers.h:225:29: note: in
>>>> expansion of macro '___bpf_concat'
>>>>>     225 | #define ___bpf_apply(fn, n) ___bpf_concat(fn, n)
>>>>>         |                             ^~~~~~~~~~~~~
>>>>> tools/perf/libbpf/include/bpf/bpf_core_read.h:173:9: note:
>>>> in expansion of macro '___bpf_apply'
>>>>>     173 |         ___bpf_apply(___bpf_field_ref,
>>>> ___bpf_narg(args))(args)
>>>>>         |         ^~~~~~~~~~~~
>>>>> tools/perf/libbpf/include/bpf/bpf_core_read.h:188:39: note:
>>>> in expansion of macro '___bpf_field_ref'
>>>>>     188 |
>>>> __builtin_preserve_field_info(___bpf_field_ref(field),
>>>> BPF_FIELD_EXISTS)
>>>>>         |                                       ^~~~~~~~~~~~~~~~
>>>>> util/bpf_skel/sample_filter.bpf.c:167:29: note: in expansion
>>>> of macro 'bpf_core_field_exists'
>>>>>     167 |                         if (bpf_core_field_exists(data-
>>>>> mem_hops))
>>>>>         |                             ^~~~~~~~~~~~~~~~~~~~~
>>>>> cc1: error: argument is not a field access ```
>>>>>
>>>>> ___bpf_field_ref1 was adapted for GCC in
>>>>> 12bbcf8e840f40b82b02981e96e0a5fbb0703ea9
>>>>> but the trick added for compatibility in
>>>>> 3a8b8fc3174891c4c12f5766d82184a82d4b2e3e
>>>>> isn't compatible with that as an address is used as an
>>>> argument.
>>>>> Workaround this by calling __builtin_preserve_field_info
>>>> directly as
>>>>> the bpf_core_field_exists macro does, but without the
>>>> ___bpf_field_ref use.
>>>>
>>>> IIUC GCC doesn't support bpf_core_fields_exists() for bitfield
>>>> members, right?  Is it gonna change in the future?
>>> Let's discuss how __builtin_preserve_field_info is handled in the first place for BPF. Right now it seems it is passed some expression as the first argument is never evaluated.
>>> The problem is GCC's implementation of __builtin_preserve_field_info is all in the backend and the front end does not understand of the special rules here.
>>>
>>> GCC implements some "special" builtins in the front-end but not by the normal function call rules but parsing them separately; this is how __builtin_offsetof and a few others are implemented in both the C and C++ front-ends (and implemented separately). Now we could have add a hook to allow a backend to something similar and maybe that is the best way forward here.
>>> But it won't be __builtin_preserve_field_info but rather `__builtin_preserve_field_type_info(type,field,kind)` instead.
>>>
>>> __builtin_preserve_enum_type_value would also be added with the following:
>>> __builtin_preserve_enum_type_value(enum_type, enum_value, kind)
>>>
>>> And change all of the rest of the builtins to accept a true type argument rather than having to cast an null pointer to that type.
>>>
>>> Will clang implement a similar builtin?
>> The clang only has one builtin for some related relocations:
>>     __builtin_preserve_field_info(..., BPF_FIELD_EXISTS)
>>     __builtin_preserve_field_info(..., BPF_FIELD_BYTE_OFFSET)
>>     ...
>> They are all used in bpf_core_read.h.
>>
>>> Note this won't be done until at least GCC 16; maybe not until GCC 17 depending on if I or someone else gets time to implement the front-end parts which is acceptable to both the C and C++ front-ends.
> So I'm taking the patch as-is, ok?
>
> But first we need the Signed-off-by tag from Andrew Pinski as he is
> listed in a Co-authored-by, that I replaced with Co-developed-by as its
> the term used for this purpose in:
>
> Yonghong, can I add an Acked-by: you since you participated in this
> discussion agreeing with the original patch (If I'm not mistaken)?

LGTM.

Acked-by: Yonghong Song <yonghong.song@linux.dev>

Re: [PATCH] perf: use __builtin_preserve_field_info for GCC compatibility

[Arnaldo Carvalho de Melo]

On Sat, Oct 04, 2025 at 04:19:48PM -0700, Yonghong Song wrote:
> On 10/2/25 10:31 AM, Arnaldo Carvalho de Melo wrote:
> > But first we need the Signed-off-by tag from Andrew Pinski as he is
> > listed in a Co-authored-by, that I replaced with Co-developed-by as its
> > the term used for this purpose in:
 
> > Yonghong, can I add an Acked-by: you since you participated in this
> > discussion agreeing with the original patch (If I'm not mistaken)?
 
> LGTM.
 
> Acked-by: Yonghong Song <yonghong.song@linux.dev>
 

Thanks, added to the cset,

- Arnaldo