[PATCH v2 1/2] soc: imx: gpcv2: Do not pass static memory as platform data

[PATCH v2 1/2] soc: imx: gpcv2: Do not pass static memory as platform data

[Andrey Smirnov]

Platform device core assumes the ownership of dev.platform_data as
well as that it is dynamically allocated and it will try to kfree it
as a part of platform_device_release(). Change the code to pass
kzalloc'ed chunk of memory instead of a pointer to a static memory to
avoid causing a BUG() when calling platform_device_put().

The problem can be reproduced by artificially enabling the error path
of platform_device_add() call (around line 357).

Note that this change also allows us to constify imx7_pgc_domains,
since we no longer need to be able to modify it.

Cc: Shawn Guo <shawnguo@kernel.org>
Cc: Stefan Agner <stefan@agner.ch>
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: linux-arm-kernel at lists.infradead.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
---

Changes since [v1]:

    - Replaced devm_kzalloc() with devm_kmalloc() as per suggestion from Stefan

[1] lkml.kernel.org/r/20180110161608.13015-1-andrew.smirnov at gmail.com

 drivers/soc/imx/gpcv2.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)

diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
index afc7ecc3c187..9947598ec4b5 100644
--- a/drivers/soc/imx/gpcv2.c
+++ b/drivers/soc/imx/gpcv2.c
@@ -18,6 +18,7 @@
 #include <linux/pm_domain.h>
 #include <linux/regmap.h>
 #include <linux/regulator/consumer.h>
+#include <linux/slab.h>
 #include <dt-bindings/power/imx7-power.h>
 
 #define GPC_LPCR_A7_BSC			0x000
@@ -155,7 +156,7 @@ static int imx7_gpc_pu_pgc_sw_pdn_req(struct generic_pm_domain *genpd)
 	return imx7_gpc_pu_pgc_sw_pxx_req(genpd, false);
 }
 
-static struct imx7_pgc_domain imx7_pgc_domains[] = {
+static const struct imx7_pgc_domain imx7_pgc_domains[] = {
 	[IMX7_POWER_DOMAIN_MIPI_PHY] = {
 		.genpd = {
 			.name      = "mipi-phy",
@@ -321,7 +322,14 @@ static int imx_gpcv2_probe(struct platform_device *pdev)
 			continue;
 		}
 
-		domain = &imx7_pgc_domains[domain_index];
+		domain = kmalloc(sizeof(*domain), GFP_KERNEL);
+		if (!domain) {
+			of_node_put(np);
+			return -ENOMEM;
+		}
+		memcpy(domain, &imx7_pgc_domains[domain_index],
+		       sizeof(*domain));
+
 		domain->regmap = regmap;
 		domain->genpd.power_on  = imx7_gpc_pu_pgc_sw_pup_req;
 		domain->genpd.power_off = imx7_gpc_pu_pgc_sw_pdn_req;
@@ -330,6 +338,7 @@ static int imx_gpcv2_probe(struct platform_device *pdev)
 						domain_index);
 		if (!pd_pdev) {
 			dev_err(dev, "Failed to allocate platform device\n");
+			kfree(domain);
 			of_node_put(np);
 			return -ENOMEM;
 		}
-- 
2.14.3

[PATCH v2 2/2] soc: imx: gpc: Do not pass static memory as platform data

[Andrey Smirnov]

Platform device core assumes the ownership of dev.platform_data as
well as that it is dynamically allocated and it will try to kfree it
as a part of platform_device_release(). Change the code to pass
kzalloc'ed chunk of memory instead of a pointer to a static memory to
avoid causing a BUG() when calling platform_device_put().

The problem can be reproduced by artificially enabling the error path
of platform_device_add() call (around line 452).

Cc: Shawn Guo <shawnguo@kernel.org>
Cc: Stefan Agner <stefan@agner.ch>
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: linux-arm-kernel at lists.infradead.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
---

Changes since [v1]:

    - Replaced devm_kzalloc() with devm_kmalloc() as per suggestion from Stefan

[1] lkml.kernel.org/r/20180110161608.13015-2-andrew.smirnov at gmail.com


 drivers/soc/imx/gpc.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/drivers/soc/imx/gpc.c b/drivers/soc/imx/gpc.c
index 47e7aa963dbb..769aae647fed 100644
--- a/drivers/soc/imx/gpc.c
+++ b/drivers/soc/imx/gpc.c
@@ -18,6 +18,7 @@
 #include <linux/pm_domain.h>
 #include <linux/regmap.h>
 #include <linux/regulator/consumer.h>
+#include <linux/slab.h>
 
 #define GPC_CNTR		0x000
 
@@ -428,13 +429,19 @@ static int imx_gpc_probe(struct platform_device *pdev)
 			if (domain_index >= of_id_data->num_domains)
 				continue;
 
-			domain = &imx_gpc_domains[domain_index];
+			domain = kmalloc(sizeof(*domain), GFP_KERNEL);
+			if (!domain) {
+				of_node_put(np);
+				return -ENOMEM;
+			}
+			memcpy(domain, &imx_gpc_domains[domain_index], sizeof(*domain));
 			domain->regmap = regmap;
 			domain->ipg_rate_mhz = ipg_rate_mhz;
 
 			pd_pdev = platform_device_alloc("imx-pgc-power-domain",
 							domain_index);
 			if (!pd_pdev) {
+				kfree(domain);
 				of_node_put(np);
 				return -ENOMEM;
 			}
-- 
2.14.3

[PATCH v2 1/2] soc: imx: gpcv2: Do not pass static memory as platform data

[Shawn Guo]

On Mon, Jan 22, 2018 at 07:07:47AM -0800, Andrey Smirnov wrote:
> Platform device core assumes the ownership of dev.platform_data as
> well as that it is dynamically allocated and it will try to kfree it
> as a part of platform_device_release(). Change the code to pass
> kzalloc'ed chunk of memory instead of a pointer to a static memory to
> avoid causing a BUG() when calling platform_device_put().
> 
> The problem can be reproduced by artificially enabling the error path
> of platform_device_add() call (around line 357).
> 
> Note that this change also allows us to constify imx7_pgc_domains,
> since we no longer need to be able to modify it.
> 
> Cc: Shawn Guo <shawnguo@kernel.org>
> Cc: Stefan Agner <stefan@agner.ch>
> Cc: Lucas Stach <l.stach@pengutronix.de>
> Cc: linux-arm-kernel at lists.infradead.org
> Cc: linux-kernel at vger.kernel.org
> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>

Applied both, thanks.

[PATCH v2 1/2] soc: imx: gpcv2: Do not pass static memory as platform data

[Shawn Guo]

On Thu, Feb 08, 2018 at 02:47:10PM +0800, Shawn Guo wrote:
> On Mon, Jan 22, 2018 at 07:07:47AM -0800, Andrey Smirnov wrote:
> > Platform device core assumes the ownership of dev.platform_data as
> > well as that it is dynamically allocated and it will try to kfree it
> > as a part of platform_device_release(). Change the code to pass
> > kzalloc'ed chunk of memory instead of a pointer to a static memory to
> > avoid causing a BUG() when calling platform_device_put().
> > 
> > The problem can be reproduced by artificially enabling the error path
> > of platform_device_add() call (around line 357).
> > 
> > Note that this change also allows us to constify imx7_pgc_domains,
> > since we no longer need to be able to modify it.
> > 
> > Cc: Shawn Guo <shawnguo@kernel.org>
> > Cc: Stefan Agner <stefan@agner.ch>
> > Cc: Lucas Stach <l.stach@pengutronix.de>
> > Cc: linux-arm-kernel at lists.infradead.org
> > Cc: linux-kernel at vger.kernel.org
> > Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
> 
> Applied both, thanks.

I decided to drop these two patches.  Triggered by Fabio's patch on top
of you changes, I just took another look at them, and found that
platform_device_add_data() can help us do the same thing in a nicer way.

Shawn

[1] https://patchwork.kernel.org/patch/10274969/

[PATCH v2 1/2] soc: imx: gpcv2: Do not pass static memory as platform data

[Andrey Smirnov]

On Wed, Mar 14, 2018 at 8:08 PM, Shawn Guo <shawnguo@kernel.org> wrote:
> On Thu, Feb 08, 2018 at 02:47:10PM +0800, Shawn Guo wrote:
>> On Mon, Jan 22, 2018 at 07:07:47AM -0800, Andrey Smirnov wrote:
>> > Platform device core assumes the ownership of dev.platform_data as
>> > well as that it is dynamically allocated and it will try to kfree it
>> > as a part of platform_device_release(). Change the code to pass
>> > kzalloc'ed chunk of memory instead of a pointer to a static memory to
>> > avoid causing a BUG() when calling platform_device_put().
>> >
>> > The problem can be reproduced by artificially enabling the error path
>> > of platform_device_add() call (around line 357).
>> >
>> > Note that this change also allows us to constify imx7_pgc_domains,
>> > since we no longer need to be able to modify it.
>> >
>> > Cc: Shawn Guo <shawnguo@kernel.org>
>> > Cc: Stefan Agner <stefan@agner.ch>
>> > Cc: Lucas Stach <l.stach@pengutronix.de>
>> > Cc: linux-arm-kernel at lists.infradead.org
>> > Cc: linux-kernel at vger.kernel.org
>> > Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
>>
>> Applied both, thanks.
>
> I decided to drop these two patches.  Triggered by Fabio's patch on top
> of you changes, I just took another look at them, and found that
> platform_device_add_data() can help us do the same thing in a nicer way.
>

Shawn:

I am slightly confused by your phrasing. Are you:

a) saying that you dropped my patches in favor of your own that use
platform_device_add_data()

or

b) asking me to convert my patches to use platform_device_add_data()?

Just trying to figure out if any action on my part is necessary.

Thanks,
Andrey Smirnov

[PATCH v2 1/2] soc: imx: gpcv2: Do not pass static memory as platform data

[Shawn Guo]

On Thu, Mar 15, 2018 at 06:00:06AM -0700, Andrey Smirnov wrote:
> I am slightly confused by your phrasing. Are you:
> 
> a) saying that you dropped my patches in favor of your own that use
> platform_device_add_data()
> 
> or
> 
> b) asking me to convert my patches to use platform_device_add_data()?
> 
> Just trying to figure out if any action on my part is necessary.

Sorry for the blur.  I'm asking you to repost your patches with using
platform_device_add_data().  In other word, that's b).

Shawn

[PATCH v2 1/2] soc: imx: gpcv2: Do not pass static memory as platform data

[Andrey Smirnov]

On Thu, Mar 15, 2018 at 6:52 AM, Shawn Guo <shawnguo@kernel.org> wrote:
> On Thu, Mar 15, 2018 at 06:00:06AM -0700, Andrey Smirnov wrote:
>> I am slightly confused by your phrasing. Are you:
>>
>> a) saying that you dropped my patches in favor of your own that use
>> platform_device_add_data()
>>
>> or
>>
>> b) asking me to convert my patches to use platform_device_add_data()?
>>
>> Just trying to figure out if any action on my part is necessary.
>
> Sorry for the blur.  I'm asking you to repost your patches with using
> platform_device_add_data().  In other word, that's b).

OK, cool, will do.

Thanks,
Andrey Smirnov