Re: [PATCH 6/6] async_tx: Mark as Deprecated in Kconfig

[Frank Li]

On Thu, Oct 15, 2026 at 08:51:46PM +0000, Eric Biggers wrote:
> On Wed, Oct 14, 2026 at 03:43:50PM -0400, Frank.Li@xxxxxxxxxxx wrote:
> > From: Frank Li <Frank.Li@xxxxxxx>
> >
> > The async_tx framework was introduced to support asynchronous RAID
> > memory operations and DMA offload engines that were common at the time.
> > Today, adoption is limited, most clients use DMAEngine interfaces
> > directly, and the framework's dependency-chaining and software fallback
> > features provide little value for new users.
> >
> > Maintaining async_tx adds an extra abstraction layer and ongoing
> > maintenance burden while serving a shrinking set of users. Deprecate the
> > framework and discourage new users in favor of direct DMAEngine APIs.
> >
> > Signed-off-by: Frank Li <Frank.Li@xxxxxxx>
> > ---
> >  crypto/async_tx/Kconfig | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/crypto/async_tx/Kconfig b/crypto/async_tx/Kconfig
> > index 89bafa2efc300..02d472fc7a476 100644
> > --- a/crypto/async_tx/Kconfig
> > +++ b/crypto/async_tx/Kconfig
> > @@ -1,6 +1,6 @@
> >  # SPDX-License-Identifier: GPL-2.0
> >  config ASYNC_CORE
> > -	tristate
> > +	tristate "Async tx core (DEPRECATED)"
>
> If possible we should just remove async_tx entirely
> (https://lore.kernel.org/linux-crypto/20261015-asycn_tx_remove-v1-0-056a7085e0a1@xxxxxxx/)
> as your other series does.

Yes, I am not sure if it can go smooth, some xor dmaengine added at 2017.

> Adding the prompt here actually does the
> opposite of deprecating it: it starts asking people to unnecessarily
> enable it, while doing nothing to make it easier to disable.

what's good ways if remove meet problem.

Frank

>
> - Eric