[PATCH] MAINTAINERS: add RELAY entry

[PATCH] MAINTAINERS: add RELAY entry

[Jason Xing]

From: Jason Xing <kernelxing@tencent.com>

RELAYFS was originally developed by Tom Zanussi and Karim
Yaghmour in 2005[1]. Jens Axboe converted it from filesystem
into a generic API in 2006[2] and made it widely known through
the notable I/O tracing tool blktrace. In the decade, there remain
a few users scatterred across different subsystems, like recently
added wifi commit[3] that is an example to show how to communicate
between users and kernel. Last year I've already done some
maintenance and added/corrected some diagnostic counters.

At Tencent, we internally maintain RELAY as one of most crucial
components of network observibility platform which was shared a
bit at LPC 2025[4][5] and hopefully will be published in the paper
this year. RELAY has proven highly efficient due to its inherent
design essence. This design becomes the indispensable way to build
a 7x24 platform monitoring various hot paths even without any
selectively sampling (yes, sampling is commonly used to avoid the
overall performance degradation). One of the recommended usages is
use its zerocopy function relay_reserve() to transfer data in a
raw format that can be recognized and parsed by the corresponding
application to userspace without introducing heavy locks and
complicated logic that appears in other types of approaches, like
printk. More details can be discovered by reading through the
Documentation :)

Credits are given to the all the contributors and reviewers for
RELAY/RELAYFS in the past and future! Many thanks!

[1]: commit e82894f84dbb ("[PATCH] relayfs")
[2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to
a generic relay API")
[3]: commit c1bf6959dd81 ("wifi: ath11k: Register relayfs entries
for CFR dump")
[4]: https://lpc.events/event/19/contributions/2055/
[5]: https://lpc.events/event/19/contributions/2010/

Acked-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jason Xing <kernelxing@tencent.com>
---
 MAINTAINERS | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/MAINTAINERS b/MAINTAINERS
index b8d8a5c41597..5d751760136c 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -22283,6 +22283,16 @@ L:	linux-wireless@vger.kernel.org
 S:	Orphan
 F:	drivers/net/wireless/rsi/
 
+RELAY
+M:	Andrew Morton <akpm@linux-foundation.org>
+M:	Jens Axboe <axboe@kernel.dk>
+M:	Jason Xing <kernelxing@tencent.com>
+L:	linux-kernel@vger.kernel.org
+S:	Maintained
+F:	Documentation/filesystems/relay.rst
+F:	include/linux/relay.h
+F:	kernel/relay.c
+
 REGISTER MAP ABSTRACTION
 M:	Mark Brown <broonie@kernel.org>
 L:	linux-kernel@vger.kernel.org
-- 
2.43.5

Re: [PATCH] MAINTAINERS: add RELAY entry

[Jens Axboe]

On 2/28/26 7:09 PM, Jason Xing wrote:
> From: Jason Xing <kernelxing@tencent.com>
> 
> RELAYFS was originally developed by Tom Zanussi and Karim
> Yaghmour in 2005[1]. Jens Axboe converted it from filesystem
> into a generic API in 2006[2] and made it widely known through
> the notable I/O tracing tool blktrace. In the decade, there remain
> a few users scatterred across different subsystems, like recently
> added wifi commit[3] that is an example to show how to communicate
> between users and kernel. Last year I've already done some
> maintenance and added/corrected some diagnostic counters.
> 
> At Tencent, we internally maintain RELAY as one of most crucial
> components of network observibility platform which was shared a
> bit at LPC 2025[4][5] and hopefully will be published in the paper
> this year. RELAY has proven highly efficient due to its inherent
> design essence. This design becomes the indispensable way to build
> a 7x24 platform monitoring various hot paths even without any
> selectively sampling (yes, sampling is commonly used to avoid the
> overall performance degradation). One of the recommended usages is
> use its zerocopy function relay_reserve() to transfer data in a
> raw format that can be recognized and parsed by the corresponding
> application to userspace without introducing heavy locks and
> complicated logic that appears in other types of approaches, like
> printk. More details can be discovered by reading through the
> Documentation :)
> 
> Credits are given to the all the contributors and reviewers for
> RELAY/RELAYFS in the past and future! Many thanks!
> 
> [1]: commit e82894f84dbb ("[PATCH] relayfs")
> [2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to
> a generic relay API")
> [3]: commit c1bf6959dd81 ("wifi: ath11k: Register relayfs entries
> for CFR dump")
> [4]: https://lpc.events/event/19/contributions/2055/
> [5]: https://lpc.events/event/19/contributions/2010/
> 
> Acked-by: Andrew Morton <akpm@linux-foundation.org>
> Signed-off-by: Jason Xing <kernelxing@tencent.com>
> ---
>  MAINTAINERS | 10 ++++++++++
>  1 file changed, 10 insertions(+)
> 
> diff --git a/MAINTAINERS b/MAINTAINERS
> index b8d8a5c41597..5d751760136c 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -22283,6 +22283,16 @@ L:	linux-wireless@vger.kernel.org
>  S:	Orphan
>  F:	drivers/net/wireless/rsi/
>  
> +RELAY
> +M:	Andrew Morton <akpm@linux-foundation.org>
> +M:	Jens Axboe <axboe@kernel.dk>
> +M:	Jason Xing <kernelxing@tencent.com>
> +L:	linux-kernel@vger.kernel.org
> +S:	Maintained
> +F:	Documentation/filesystems/relay.rst
> +F:	include/linux/relay.h
> +F:	kernel/relay.c
> +
>  REGISTER MAP ABSTRACTION
>  M:	Mark Brown <broonie@kernel.org>
>  L:	linux-kernel@vger.kernel.org

Acked-by: Jens Axboe <axboe@kernel.dk>

-- 
Jens Axboe

Re: [PATCH] MAINTAINERS: add RELAY entry

[Andy Shevchenko]

On Sun, Mar 01, 2026 at 10:09:02AM +0800, Jason Xing wrote:

> RELAYFS was originally developed by Tom Zanussi and Karim
> Yaghmour in 2005[1]. Jens Axboe converted it from filesystem
> into a generic API in 2006[2] and made it widely known through
> the notable I/O tracing tool blktrace. In the decade, there remain
> a few users scatterred across different subsystems, like recently
> added wifi commit[3] that is an example to show how to communicate
> between users and kernel. Last year I've already done some
> maintenance and added/corrected some diagnostic counters.
> 
> At Tencent, we internally maintain RELAY as one of most crucial
> components of network observibility platform which was shared a
> bit at LPC 2025[4][5] and hopefully will be published in the paper
> this year. RELAY has proven highly efficient due to its inherent
> design essence. This design becomes the indispensable way to build
> a 7x24 platform monitoring various hot paths even without any
> selectively sampling (yes, sampling is commonly used to avoid the
> overall performance degradation). One of the recommended usages is
> use its zerocopy function relay_reserve() to transfer data in a
> raw format that can be recognized and parsed by the corresponding
> application to userspace without introducing heavy locks and
> complicated logic that appears in other types of approaches, like
> printk. More details can be discovered by reading through the
> Documentation :)
> 
> Credits are given to the all the contributors and reviewers for
> RELAY/RELAYFS in the past and future! Many thanks!
> 
> [1]: commit e82894f84dbb ("[PATCH] relayfs")
> [2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to
> a generic relay API")

It's better to read as a single line here

[2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to > a generic relay API")

OR you can even have it as a proper Link tag:

Link: https://git.kernel.org/torvalds/c/b86ff981a825 [2]

Same for the rest.

> [3]: commit c1bf6959dd81 ("wifi: ath11k: Register relayfs entries
> for CFR dump")
> [4]: https://lpc.events/event/19/contributions/2055/
> [5]: https://lpc.events/event/19/contributions/2010/

> 

Note, when converting to Link tags the above, this blank line should be
removed.

> Acked-by: Andrew Morton <akpm@linux-foundation.org>
> Signed-off-by: Jason Xing <kernelxing@tencent.com>

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH] MAINTAINERS: add RELAY entry

[Jason Xing]

Hi Andy,

On Mon, Mar 2, 2026 at 4:35 PM Andy Shevchenko
<andriy.shevchenko@linux.intel.com> wrote:
>
> On Sun, Mar 01, 2026 at 10:09:02AM +0800, Jason Xing wrote:
>
> > RELAYFS was originally developed by Tom Zanussi and Karim
> > Yaghmour in 2005[1]. Jens Axboe converted it from filesystem
> > into a generic API in 2006[2] and made it widely known through
> > the notable I/O tracing tool blktrace. In the decade, there remain
> > a few users scatterred across different subsystems, like recently
> > added wifi commit[3] that is an example to show how to communicate
> > between users and kernel. Last year I've already done some
> > maintenance and added/corrected some diagnostic counters.
> >
> > At Tencent, we internally maintain RELAY as one of most crucial
> > components of network observibility platform which was shared a
> > bit at LPC 2025[4][5] and hopefully will be published in the paper
> > this year. RELAY has proven highly efficient due to its inherent
> > design essence. This design becomes the indispensable way to build
> > a 7x24 platform monitoring various hot paths even without any
> > selectively sampling (yes, sampling is commonly used to avoid the
> > overall performance degradation). One of the recommended usages is
> > use its zerocopy function relay_reserve() to transfer data in a
> > raw format that can be recognized and parsed by the corresponding
> > application to userspace without introducing heavy locks and
> > complicated logic that appears in other types of approaches, like
> > printk. More details can be discovered by reading through the
> > Documentation :)
> >
> > Credits are given to the all the contributors and reviewers for
> > RELAY/RELAYFS in the past and future! Many thanks!
> >
> > [1]: commit e82894f84dbb ("[PATCH] relayfs")
> > [2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to
> > a generic relay API")
>
> It's better to read as a single line here
>
> [2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to > a generic relay API")
>
> OR you can even have it as a proper Link tag:
>
> Link: https://git.kernel.org/torvalds/c/b86ff981a825 [2]
>
> Same for the rest.

Thanks for the review.

The fact is that originally I used a single line but checkpatch warned
me because of exceeding 75 chars. The current format is just to split
it into two lines as script requested.

Thanks,
Jason

>
> > [3]: commit c1bf6959dd81 ("wifi: ath11k: Register relayfs entries
> > for CFR dump")
> > [4]: https://lpc.events/event/19/contributions/2055/
> > [5]: https://lpc.events/event/19/contributions/2010/
>
> >
>
> Note, when converting to Link tags the above, this blank line should be
> removed.
>
> > Acked-by: Andrew Morton <akpm@linux-foundation.org>
> > Signed-off-by: Jason Xing <kernelxing@tencent.com>
>
> --
> With Best Regards,
> Andy Shevchenko
>
>

Re: [PATCH] MAINTAINERS: add RELAY entry

[Jason Xing]

On Mon, Mar 2, 2026 at 4:40 PM Jason Xing <kerneljasonxing@gmail.com> wrote:
>
> Hi Andy,
>
> On Mon, Mar 2, 2026 at 4:35 PM Andy Shevchenko
> <andriy.shevchenko@linux.intel.com> wrote:
> >
> > On Sun, Mar 01, 2026 at 10:09:02AM +0800, Jason Xing wrote:
> >
> > > RELAYFS was originally developed by Tom Zanussi and Karim
> > > Yaghmour in 2005[1]. Jens Axboe converted it from filesystem
> > > into a generic API in 2006[2] and made it widely known through
> > > the notable I/O tracing tool blktrace. In the decade, there remain
> > > a few users scatterred across different subsystems, like recently
> > > added wifi commit[3] that is an example to show how to communicate
> > > between users and kernel. Last year I've already done some
> > > maintenance and added/corrected some diagnostic counters.
> > >
> > > At Tencent, we internally maintain RELAY as one of most crucial
> > > components of network observibility platform which was shared a
> > > bit at LPC 2025[4][5] and hopefully will be published in the paper
> > > this year. RELAY has proven highly efficient due to its inherent
> > > design essence. This design becomes the indispensable way to build
> > > a 7x24 platform monitoring various hot paths even without any
> > > selectively sampling (yes, sampling is commonly used to avoid the
> > > overall performance degradation). One of the recommended usages is
> > > use its zerocopy function relay_reserve() to transfer data in a
> > > raw format that can be recognized and parsed by the corresponding
> > > application to userspace without introducing heavy locks and
> > > complicated logic that appears in other types of approaches, like
> > > printk. More details can be discovered by reading through the
> > > Documentation :)
> > >
> > > Credits are given to the all the contributors and reviewers for
> > > RELAY/RELAYFS in the past and future! Many thanks!
> > >
> > > [1]: commit e82894f84dbb ("[PATCH] relayfs")
> > > [2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to
> > > a generic relay API")
> >
> > It's better to read as a single line here
> >
> > [2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to > a generic relay API")
> >
> > OR you can even have it as a proper Link tag:
> >
> > Link: https://git.kernel.org/torvalds/c/b86ff981a825 [2]
> >
> > Same for the rest.
>
> Thanks for the review.
>
> The fact is that originally I used a single line but checkpatch warned
> me because of exceeding 75 chars. The current format is just to split
> it into two lines as script requested.

Oh interesting thing is Andrew already kindly revised the patch just
like you said:
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm.git/commit/?h=mm-hotfixes-unstable&id=46947075c75b9b213ba31c1af8c45e0a9c85bf74

So I think there is no need to repost it :)

Thank you!

Re: [PATCH] MAINTAINERS: add RELAY entry

[Andy Shevchenko]

On Mon, Mar 02, 2026 at 04:40:40PM +0800, Jason Xing wrote:
> Hi Andy,
> 
> On Mon, Mar 2, 2026 at 4:35 PM Andy Shevchenko
> <andriy.shevchenko@linux.intel.com> wrote:
> > On Sun, Mar 01, 2026 at 10:09:02AM +0800, Jason Xing wrote:

...

> > > [1]: commit e82894f84dbb ("[PATCH] relayfs")
> > > [2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to
> > > a generic relay API")
> >
> > It's better to read as a single line here
> >
> > [2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to > a generic relay API")
> >
> > OR you can even have it as a proper Link tag:
> >
> > Link: https://git.kernel.org/torvalds/c/b86ff981a825 [2]
> >
> > Same for the rest.
> 
> Thanks for the review.
> 
> The fact is that originally I used a single line but checkpatch warned
> me because of exceeding 75 chars. The current format is just to split
> it into two lines as script requested.

Checkpatch doesn't parse the tags correctly (or lines that can be recognized as
similar to tags). TL;DR: ignore checkpatch for this certain case.

> > > [3]: commit c1bf6959dd81 ("wifi: ath11k: Register relayfs entries
> > > for CFR dump")
> > > [4]: https://lpc.events/event/19/contributions/2055/
> > > [5]: https://lpc.events/event/19/contributions/2010/
> >
> > Note, when converting to Link tags the above, this blank line should be
> > removed.

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH] MAINTAINERS: add RELAY entry

[Andy Shevchenko]

On Mon, Mar 02, 2026 at 04:47:49PM +0800, Jason Xing wrote:
> On Mon, Mar 2, 2026 at 4:40 PM Jason Xing <kerneljasonxing@gmail.com> wrote:
> > On Mon, Mar 2, 2026 at 4:35 PM Andy Shevchenko
> > <andriy.shevchenko@linux.intel.com> wrote:
> > > On Sun, Mar 01, 2026 at 10:09:02AM +0800, Jason Xing wrote:

...

> > > > [1]: commit e82894f84dbb ("[PATCH] relayfs")
> > > > [2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to
> > > > a generic relay API")
> > >
> > > It's better to read as a single line here
> > >
> > > [2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to a generic relay API")
> > >
> > > OR you can even have it as a proper Link tag:
> > >
> > > Link: https://git.kernel.org/torvalds/c/b86ff981a825 [2]
> > >
> > > Same for the rest.
> >
> > Thanks for the review.
> >
> > The fact is that originally I used a single line but checkpatch warned
> > me because of exceeding 75 chars. The current format is just to split
> > it into two lines as script requested.
> 
> Oh interesting thing is Andrew already kindly revised the patch just
> like you said:
> https://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm.git/commit/?h=mm-hotfixes-unstable&id=46947075c75b9b213ba31c1af8c45e0a9c85bf74
> 
> So I think there is no need to repost it :)

LGTM, but if people wants real Link tags... Up to you.
(I think we are fine now with the current code.)

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH] MAINTAINERS: add RELAY entry

[Jason Xing]

On Mon, Mar 2, 2026 at 4:54 PM Andy Shevchenko
<andriy.shevchenko@linux.intel.com> wrote:
>
> On Mon, Mar 02, 2026 at 04:47:49PM +0800, Jason Xing wrote:
> > On Mon, Mar 2, 2026 at 4:40 PM Jason Xing <kerneljasonxing@gmail.com> wrote:
> > > On Mon, Mar 2, 2026 at 4:35 PM Andy Shevchenko
> > > <andriy.shevchenko@linux.intel.com> wrote:
> > > > On Sun, Mar 01, 2026 at 10:09:02AM +0800, Jason Xing wrote:
>
> ...
>
> > > > > [1]: commit e82894f84dbb ("[PATCH] relayfs")
> > > > > [2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to
> > > > > a generic relay API")
> > > >
> > > > It's better to read as a single line here
> > > >
> > > > [2]: commit b86ff981a825 ("[PATCH] relay: migrate from relayfs to a generic relay API")
> > > >
> > > > OR you can even have it as a proper Link tag:
> > > >
> > > > Link: https://git.kernel.org/torvalds/c/b86ff981a825 [2]
> > > >
> > > > Same for the rest.
> > >
> > > Thanks for the review.
> > >
> > > The fact is that originally I used a single line but checkpatch warned
> > > me because of exceeding 75 chars. The current format is just to split
> > > it into two lines as script requested.
> >
> > Oh interesting thing is Andrew already kindly revised the patch just
> > like you said:
> > https://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm.git/commit/?h=mm-hotfixes-unstable&id=46947075c75b9b213ba31c1af8c45e0a9c85bf74
> >
> > So I think there is no need to repost it :)
>
> LGTM, but if people wants real Link tags... Up to you.
> (I think we are fine now with the current code.)

Great! Thanks!

Thanks,
Jason