[PATCH 0/2] mmc: core: Avoid export of mmc_alloc|free_host()

[PATCH 0/2] mmc: core: Avoid export of mmc_alloc|free_host()

[Ulf Hansson]

From: Ulf Hansson <ulfh@kernel.org>

The only remaining user of mmc_alloc|free_host() is the greybus sdio driver.
Let's convert it into using the resource managed variant so we can drop the
export of mmc_alloc|free_host().

Ulf Hansson (2):
  staging: greybus: sdio: Convert to devm_mmc_alloc_host()
  mmc: core: Turn mmc_alloc|free_host() into static functions

 drivers/mmc/core/host.c        | 35 +++++++++-------------------------
 drivers/staging/greybus/sdio.c | 11 +++--------
 include/linux/mmc/host.h       |  2 --
 3 files changed, 12 insertions(+), 36 deletions(-)

-- 
2.43.0

[PATCH 1/2] staging: greybus: sdio: Convert to devm_mmc_alloc_host()

[Ulf Hansson]

From: Ulf Hansson <ulfh@kernel.org>

Simplify the code by converting from mmc_alloc_host() to the resource
managed devm_mmc_alloc_host().

Signed-off-by: Ulf Hansson <ulfh@kernel.org>
---
 drivers/staging/greybus/sdio.c | 11 +++--------
 1 file changed, 3 insertions(+), 8 deletions(-)

diff --git a/drivers/staging/greybus/sdio.c b/drivers/staging/greybus/sdio.c
index 3952f3d225db..642d785fe918 100644
--- a/drivers/staging/greybus/sdio.c
+++ b/drivers/staging/greybus/sdio.c
@@ -767,17 +767,15 @@ static int gb_sdio_probe(struct gbphy_device *gbphy_dev,
 	struct gb_sdio_host *host;
 	int ret = 0;
 
-	mmc = mmc_alloc_host(sizeof(*host), &gbphy_dev->dev);
+	mmc = devm_mmc_alloc_host(&gbphy_dev->dev, sizeof(*host));
 	if (!mmc)
 		return -ENOMEM;
 
 	connection = gb_connection_create(gbphy_dev->bundle,
 					  le16_to_cpu(gbphy_dev->cport_desc->id),
 					  gb_sdio_request_handler);
-	if (IS_ERR(connection)) {
-		ret = PTR_ERR(connection);
-		goto exit_mmc_free;
-	}
+	if (IS_ERR(connection))
+		return PTR_ERR(connection);
 
 	host = mmc_priv(mmc);
 	host->mmc = mmc;
@@ -835,8 +833,6 @@ static int gb_sdio_probe(struct gbphy_device *gbphy_dev,
 	gb_connection_disable(connection);
 exit_connection_destroy:
 	gb_connection_destroy(connection);
-exit_mmc_free:
-	mmc_free_host(mmc);
 
 	return ret;
 }
@@ -863,7 +859,6 @@ static void gb_sdio_remove(struct gbphy_device *gbphy_dev)
 	mmc_remove_host(mmc);
 	gb_connection_disable(connection);
 	gb_connection_destroy(connection);
-	mmc_free_host(mmc);
 }
 
 static const struct gbphy_device_id gb_sdio_id_table[] = {
-- 
2.43.0

[PATCH 2/2] mmc: core: Turn mmc_alloc|free_host() into static functions

[Ulf Hansson]

From: Ulf Hansson <ulfh@kernel.org>

As there are no longer any users of these functions, let's make them
internal to the mmc core. While at it, let's also flip the order of the
in-parameters to mmc_alloc_host() to be consistent with devm_alloc_host().

Signed-off-by: Ulf Hansson <ulfh@kernel.org>
---
 drivers/mmc/core/host.c  | 35 +++++++++--------------------------
 include/linux/mmc/host.h |  2 --
 2 files changed, 9 insertions(+), 28 deletions(-)

diff --git a/drivers/mmc/core/host.c b/drivers/mmc/core/host.c
index 1bcf0e917b59..8e9e61839198 100644
--- a/drivers/mmc/core/host.c
+++ b/drivers/mmc/core/host.c
@@ -505,14 +505,7 @@ static int mmc_first_nonreserved_index(void)
 	return max + 1;
 }
 
-/**
- *	mmc_alloc_host - initialise the per-host structure.
- *	@extra: sizeof private data structure
- *	@dev: pointer to host device model structure
- *
- *	Initialise the per-host structure.
- */
-struct mmc_host *mmc_alloc_host(int extra, struct device *dev)
+static struct mmc_host *mmc_alloc_host(struct device *dev, int extra)
 {
 	int index;
 	struct mmc_host *host;
@@ -583,7 +576,13 @@ struct mmc_host *mmc_alloc_host(int extra, struct device *dev)
 	return host;
 }
 
-EXPORT_SYMBOL(mmc_alloc_host);
+static void mmc_free_host(struct mmc_host *host)
+{
+	cancel_delayed_work_sync(&host->detect);
+	cancel_work_sync(&host->sdio_irq_work);
+	mmc_pwrseq_free(host);
+	put_device(&host->class_dev);
+}
 
 static void devm_mmc_host_release(struct device *dev, void *res)
 {
@@ -598,7 +597,7 @@ struct mmc_host *devm_mmc_alloc_host(struct device *dev, int extra)
 	if (!dr)
 		return NULL;
 
-	host = mmc_alloc_host(extra, dev);
+	host = mmc_alloc_host(dev, extra);
 	if (!host) {
 		devres_free(dr);
 		return NULL;
@@ -692,19 +691,3 @@ void mmc_remove_host(struct mmc_host *host)
 }
 
 EXPORT_SYMBOL(mmc_remove_host);
-
-/**
- *	mmc_free_host - free the host structure
- *	@host: mmc host
- *
- *	Free the host once all references to it have been dropped.
- */
-void mmc_free_host(struct mmc_host *host)
-{
-	cancel_delayed_work_sync(&host->detect);
-	cancel_work_sync(&host->sdio_irq_work);
-	mmc_pwrseq_free(host);
-	put_device(&host->class_dev);
-}
-
-EXPORT_SYMBOL(mmc_free_host);
diff --git a/include/linux/mmc/host.h b/include/linux/mmc/host.h
index ddb32bc2946f..198554e48346 100644
--- a/include/linux/mmc/host.h
+++ b/include/linux/mmc/host.h
@@ -585,11 +585,9 @@ struct mmc_host {
 
 struct device_node;
 
-struct mmc_host *mmc_alloc_host(int extra, struct device *);
 struct mmc_host *devm_mmc_alloc_host(struct device *dev, int extra);
 int mmc_add_host(struct mmc_host *);
 void mmc_remove_host(struct mmc_host *);
-void mmc_free_host(struct mmc_host *);
 void mmc_of_parse_clk_phase(struct device *dev,
 			    struct mmc_clk_phase_map *map);
 int mmc_of_parse(struct mmc_host *host);
-- 
2.43.0

Re: [PATCH 1/2] staging: greybus: sdio: Convert to devm_mmc_alloc_host()

[Rui Miguel Silva]

Hey Ulf,
Thanks for the patch.

On Tue Sep 15, 2026 at 2:20 PM WEST, Ulf Hansson wrote:

> From: Ulf Hansson <ulfh@kernel.org>
>
> Simplify the code by converting from mmc_alloc_host() to the resource
> managed devm_mmc_alloc_host().
>
> Signed-off-by: Ulf Hansson <ulfh@kernel.org>

LGTM,

Acked-by: Rui Miguel Silva <rui.silva@linaro.org>

Cheers,
    Rui
> ---
>  drivers/staging/greybus/sdio.c | 11 +++--------
>  1 file changed, 3 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/staging/greybus/sdio.c b/drivers/staging/greybus/sdio.c
> index 3952f3d225db..642d785fe918 100644
> --- a/drivers/staging/greybus/sdio.c
> +++ b/drivers/staging/greybus/sdio.c
> @@ -767,17 +767,15 @@ static int gb_sdio_probe(struct gbphy_device *gbphy_dev,
>  	struct gb_sdio_host *host;
>  	int ret = 0;
>  
> -	mmc = mmc_alloc_host(sizeof(*host), &gbphy_dev->dev);
> +	mmc = devm_mmc_alloc_host(&gbphy_dev->dev, sizeof(*host));
>  	if (!mmc)
>  		return -ENOMEM;
>  
>  	connection = gb_connection_create(gbphy_dev->bundle,
>  					  le16_to_cpu(gbphy_dev->cport_desc->id),
>  					  gb_sdio_request_handler);
> -	if (IS_ERR(connection)) {
> -		ret = PTR_ERR(connection);
> -		goto exit_mmc_free;
> -	}
> +	if (IS_ERR(connection))
> +		return PTR_ERR(connection);
>  
>  	host = mmc_priv(mmc);
>  	host->mmc = mmc;
> @@ -835,8 +833,6 @@ static int gb_sdio_probe(struct gbphy_device *gbphy_dev,
>  	gb_connection_disable(connection);
>  exit_connection_destroy:
>  	gb_connection_destroy(connection);
> -exit_mmc_free:
> -	mmc_free_host(mmc);
>  
>  	return ret;
>  }
> @@ -863,7 +859,6 @@ static void gb_sdio_remove(struct gbphy_device *gbphy_dev)
>  	mmc_remove_host(mmc);
>  	gb_connection_disable(connection);
>  	gb_connection_destroy(connection);
> -	mmc_free_host(mmc);
>  }
>  
>  static const struct gbphy_device_id gb_sdio_id_table[] = {
> -- 
> 2.43.0

Re: [PATCH 2/2] mmc: core: Turn mmc_alloc|free_host() into static functions

[Shawn Lin]

在 2026/09/15 星期二 21:20, Ulf Hansson 写道:
> From: Ulf Hansson <ulfh@kernel.org>
> 
> As there are no longer any users of these functions, let's make them
> internal to the mmc core. While at it, let's also flip the order of the
> in-parameters to mmc_alloc_host() to be consistent with devm_alloc_host().
> 

Nice move when vub300 was gone,

Reviewed-by: Shawn Lin <shawn.lin@linux.dev>

> Signed-off-by: Ulf Hansson <ulfh@kernel.org>
> ---
>   drivers/mmc/core/host.c  | 35 +++++++++--------------------------
>   include/linux/mmc/host.h |  2 --
>   2 files changed, 9 insertions(+), 28 deletions(-)
> 
> diff --git a/drivers/mmc/core/host.c b/drivers/mmc/core/host.c
> index 1bcf0e917b59..8e9e61839198 100644
> --- a/drivers/mmc/core/host.c
> +++ b/drivers/mmc/core/host.c
> @@ -505,14 +505,7 @@ static int mmc_first_nonreserved_index(void)
>   	return max + 1;
>   }
>   
> -/**
> - *	mmc_alloc_host - initialise the per-host structure.
> - *	@extra: sizeof private data structure
> - *	@dev: pointer to host device model structure
> - *
> - *	Initialise the per-host structure.
> - */
> -struct mmc_host *mmc_alloc_host(int extra, struct device *dev)
> +static struct mmc_host *mmc_alloc_host(struct device *dev, int extra)
>   {
>   	int index;
>   	struct mmc_host *host;
> @@ -583,7 +576,13 @@ struct mmc_host *mmc_alloc_host(int extra, struct device *dev)
>   	return host;
>   }
>   
> -EXPORT_SYMBOL(mmc_alloc_host);
> +static void mmc_free_host(struct mmc_host *host)
> +{
> +	cancel_delayed_work_sync(&host->detect);
> +	cancel_work_sync(&host->sdio_irq_work);
> +	mmc_pwrseq_free(host);
> +	put_device(&host->class_dev);
> +}
>   
>   static void devm_mmc_host_release(struct device *dev, void *res)
>   {
> @@ -598,7 +597,7 @@ struct mmc_host *devm_mmc_alloc_host(struct device *dev, int extra)
>   	if (!dr)
>   		return NULL;
>   
> -	host = mmc_alloc_host(extra, dev);
> +	host = mmc_alloc_host(dev, extra);
>   	if (!host) {
>   		devres_free(dr);
>   		return NULL;
> @@ -692,19 +691,3 @@ void mmc_remove_host(struct mmc_host *host)
>   }
>   
>   EXPORT_SYMBOL(mmc_remove_host);
> -
> -/**
> - *	mmc_free_host - free the host structure
> - *	@host: mmc host
> - *
> - *	Free the host once all references to it have been dropped.
> - */
> -void mmc_free_host(struct mmc_host *host)
> -{
> -	cancel_delayed_work_sync(&host->detect);
> -	cancel_work_sync(&host->sdio_irq_work);
> -	mmc_pwrseq_free(host);
> -	put_device(&host->class_dev);
> -}
> -
> -EXPORT_SYMBOL(mmc_free_host);
> diff --git a/include/linux/mmc/host.h b/include/linux/mmc/host.h
> index ddb32bc2946f..198554e48346 100644
> --- a/include/linux/mmc/host.h
> +++ b/include/linux/mmc/host.h
> @@ -585,11 +585,9 @@ struct mmc_host {
>   
>   struct device_node;
>   
> -struct mmc_host *mmc_alloc_host(int extra, struct device *);
>   struct mmc_host *devm_mmc_alloc_host(struct device *dev, int extra);
>   int mmc_add_host(struct mmc_host *);
>   void mmc_remove_host(struct mmc_host *);
> -void mmc_free_host(struct mmc_host *);
>   void mmc_of_parse_clk_phase(struct device *dev,
>   			    struct mmc_clk_phase_map *map);
>   int mmc_of_parse(struct mmc_host *host);

Re: [PATCH 1/2] staging: greybus: sdio: Convert to devm_mmc_alloc_host()

[Greg Kroah-Hartman]

On Tue, Sep 15, 2026 at 03:20:45PM +0200, Ulf Hansson wrote:
> From: Ulf Hansson <ulfh@kernel.org>
> 
> Simplify the code by converting from mmc_alloc_host() to the resource
> managed devm_mmc_alloc_host().
> 
> Signed-off-by: Ulf Hansson <ulfh@kernel.org>
> ---
>  drivers/staging/greybus/sdio.c | 11 +++--------
>  1 file changed, 3 insertions(+), 8 deletions(-)

Acked-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Re: [PATCH 2/2] mmc: core: Turn mmc_alloc|free_host() into static functions

[Johan Hovold]

On Tue, Sep 15, 2026 at 03:20:46PM +0200, Ulf Hansson wrote:
> From: Ulf Hansson <ulfh@kernel.org>
> 
> As there are no longer any users of these functions, let's make them
> internal to the mmc core. 

Why would you want to do that? The devres helpers should just be simple
wrappers around these and sometimes devres just isn't a good fit.

Especially with the work cancellations present in mmc_free_host() (which
I have pointed out elsewhere should not be there), a driver may need to
free the host before tearing down other non-devres managed resources
during unbind.

This may even be needed for greybus which currently destroys the
connection before freeing the host.

I suggest you just keep the non-devres interface around (after updating
the prototype).

> While at it, let's also flip the order of the
> in-parameters to mmc_alloc_host() to be consistent with devm_alloc_host().

Johan

Re: [PATCH 2/2] mmc: core: Turn mmc_alloc|free_host() into static functions

[Ulf Hansson]

On Wed, Sep 16, 2026 at 9:43 AM Johan Hovold <johan@kernel.org> wrote:
>
> On Tue, Sep 15, 2026 at 03:20:46PM +0200, Ulf Hansson wrote:
> > From: Ulf Hansson <ulfh@kernel.org>
> >
> > As there are no longer any users of these functions, let's make them
> > internal to the mmc core.
>
> Why would you want to do that? The devres helpers should just be simple
> wrappers around these and sometimes devres just isn't a good fit.

At the moment there seems to be no need for them. I would rather keep
API/interfaces as simple as possible, so I prefer to remove them at
this point.

If we see a need for them, we can always add them back.

>
> Especially with the work cancellations present in mmc_free_host() (which
> I have pointed out elsewhere should not be there), a driver may need to
> free the host before tearing down other non-devres managed resources
> during unbind.

Can you please point me to such an example so I can try to understand better?

>
> This may even be needed for greybus which currently destroys the
> connection before freeing the host.

I looked closer at gb_sdio_remove() (the greybus sdio driver's
->remove() callback) and I think the problem isn't about freeing the
host.

Instead it seems like the call to mmc_remove_host() is done too late.
To me it looks like when the mmc core tries to power off the card
gracefully, through mmc_remove_host() the driver has already moved
into a state where it no longer accepts any requests.

>
> I suggest you just keep the non-devres interface around (after updating
> the prototype).
>
> > While at it, let's also flip the order of the
> > in-parameters to mmc_alloc_host() to be consistent with devm_alloc_host().
>
> Johan

Kind regards
Uffe

Re: [PATCH 2/2] mmc: core: Turn mmc_alloc|free_host() into static functions

[Johan Hovold]

On Wed, Sep 16, 2026 at 11:36:25AM +0200, Ulf Hansson wrote:
> On Wed, Sep 16, 2026 at 9:43 AM Johan Hovold <johan@kernel.org> wrote:
> >
> > On Tue, Sep 15, 2026 at 03:20:46PM +0200, Ulf Hansson wrote:
> > > From: Ulf Hansson <ulfh@kernel.org>
> > >
> > > As there are no longer any users of these functions, let's make them
> > > internal to the mmc core.
> >
> > Why would you want to do that? The devres helpers should just be simple
> > wrappers around these and sometimes devres just isn't a good fit.
> 
> At the moment there seems to be no need for them. I would rather keep
> API/interfaces as simple as possible, so I prefer to remove them at
> this point.
> 
> If we see a need for them, we can always add them back.

Devres generally only works when all resources are device managed.
Therefore you should always provide the underlying non-devres manages
interface as well so that you don't force devres on drivers where it
could cause trouble.
 
> > Especially with the work cancellations present in mmc_free_host() (which
> > I have pointed out elsewhere should not be there), a driver may need to
> > free the host before tearing down other non-devres managed resources
> > during unbind.
> 
> Can you please point me to such an example so I can try to understand better?

We just discussed the renesas driver which can schedule rescan work
before registering the host controller. [1]

If such a driver also has non-devres managed resources that are freed
before the work is cancelled you have a use-after-free.

> > This may even be needed for greybus which currently destroys the
> > connection before freeing the host.
> 
> I looked closer at gb_sdio_remove() (the greybus sdio driver's
> ->remove() callback) and I think the problem isn't about freeing the
> host.
> 
> Instead it seems like the call to mmc_remove_host() is done too late.
> To me it looks like when the mmc core tries to power off the card
> gracefully, through mmc_remove_host() the driver has already moved
> into a state where it no longer accepts any requests.

Yes, that looks wrong, but that's a separate issue.

I only pointed at greybus as an example of a driver which has non-devres
managed resources. If there is (rescan) work still scheduled after
probe() or remove() returns, there's a potential use-after-free.

Johan


[1] https://lore.kernel.org/lkml/ap7CaVj82BJZgjf6@hovoldconsulting.com/

Re: [PATCH 2/2] mmc: core: Turn mmc_alloc|free_host() into static functions

[Ulf Hansson]

On Wed, Sep 16, 2026 at 12:00 PM Johan Hovold <johan@kernel.org> wrote:
>
> On Wed, Sep 16, 2026 at 11:36:25AM +0200, Ulf Hansson wrote:
> > On Wed, Sep 16, 2026 at 9:43 AM Johan Hovold <johan@kernel.org> wrote:
> > >
> > > On Tue, Sep 15, 2026 at 03:20:46PM +0200, Ulf Hansson wrote:
> > > > From: Ulf Hansson <ulfh@kernel.org>
> > > >
> > > > As there are no longer any users of these functions, let's make them
> > > > internal to the mmc core.
> > >
> > > Why would you want to do that? The devres helpers should just be simple
> > > wrappers around these and sometimes devres just isn't a good fit.
> >
> > At the moment there seems to be no need for them. I would rather keep
> > API/interfaces as simple as possible, so I prefer to remove them at
> > this point.
> >
> > If we see a need for them, we can always add them back.
>
> Devres generally only works when all resources are device managed.
> Therefore you should always provide the underlying non-devres manages
> interface as well so that you don't force devres on drivers where it
> could cause trouble.
>
> > > Especially with the work cancellations present in mmc_free_host() (which
> > > I have pointed out elsewhere should not be there), a driver may need to
> > > free the host before tearing down other non-devres managed resources
> > > during unbind.
> >
> > Can you please point me to such an example so I can try to understand better?
>
> We just discussed the renesas driver which can schedule rescan work
> before registering the host controller. [1]

That isn't a problem, but it's not the right thing to do as
host->rescan_disable is set.

>
> If such a driver also has non-devres managed resources that are freed
> before the work is cancelled you have a use-after-free.

If you are referring to the mmc rescan work; mmc_rescan() will just
bail out as host->rescan_disable has been set as soon as
mmc_remove_host() is called (see mmc_stop_host()). In other words, I
don't see how a use-after-free would be possible in this regard.

If you are referring to an internal work for the host, yes that needs
to be managed correctly. Although, that doesn't matter whether the mmc
host has been allocated with the resource managed variant or not.

>
> > > This may even be needed for greybus which currently destroys the
> > > connection before freeing the host.
> >
> > I looked closer at gb_sdio_remove() (the greybus sdio driver's
> > ->remove() callback) and I think the problem isn't about freeing the
> > host.
> >
> > Instead it seems like the call to mmc_remove_host() is done too late.
> > To me it looks like when the mmc core tries to power off the card
> > gracefully, through mmc_remove_host() the driver has already moved
> > into a state where it no longer accepts any requests.
>
> Yes, that looks wrong, but that's a separate issue.
>
> I only pointed at greybus as an example of a driver which has non-devres
> managed resources. If there is (rescan) work still scheduled after
> probe() or remove() returns, there's a potential use-after-free.

See above. This should not be an issue, at least it has nothing to do
whether we are using managed resources or not.

For the mmc_rescan() work to access the host internal data (for
potential causing a use-after-free), the host must have been
"started". See mmc_start|stop_host().

>
> Johan
>
>
> [1] https://lore.kernel.org/lkml/ap7CaVj82BJZgjf6@hovoldconsulting.com/

Kind regards
Uffe