The code allocates sizeof(regulator_dev) for a pointer. Make it less
generous. Let kcalloc() calculate the size, while at it.
Cc: [email protected]
Fixes: d8ca7d184b33 ("regulator: core: Introduce API for regulators coupling customization")
Signed-off-by: Michał Mirosław <[email protected]>
---
drivers/regulator/core.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
index 75ff7c563c5d..9e18997777d3 100644
--- a/drivers/regulator/core.c
+++ b/drivers/regulator/core.c
@@ -5011,20 +5011,20 @@ static void regulator_remove_coupling(struct regulator_dev *rdev)
static int regulator_init_coupling(struct regulator_dev *rdev)
{
+ struct regulator_dev **coupled;
int err, n_phandles;
- size_t alloc_size;
if (!IS_ENABLED(CONFIG_OF))
n_phandles = 0;
else
n_phandles = of_get_n_coupled(rdev);
- alloc_size = sizeof(*rdev) * (n_phandles + 1);
-
- rdev->coupling_desc.coupled_rdevs = kzalloc(alloc_size, GFP_KERNEL);
- if (!rdev->coupling_desc.coupled_rdevs)
+ coupled = kcalloc(n_phandles + 1, sizeof(*coupled), GFP_KERNEL);
+ if (!coupled)
return -ENOMEM;
+ rdev->coupling_desc.coupled_rdevs = coupled;
+
/*
* Every regulator should always have coupling descriptor filled with
* at least pointer to itself.
--
2.20.1
09.08.2020 22:21, Michał Mirosław пишет:
> The code allocates sizeof(regulator_dev) for a pointer. Make it less
> generous. Let kcalloc() calculate the size, while at it.
>
> Cc: [email protected]
> Fixes: d8ca7d184b33 ("regulator: core: Introduce API for regulators coupling customization")
> Signed-off-by: Michał Mirosław <[email protected]>
> ---
> drivers/regulator/core.c | 10 +++++-----
> 1 file changed, 5 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
> index 75ff7c563c5d..9e18997777d3 100644
> --- a/drivers/regulator/core.c
> +++ b/drivers/regulator/core.c
> @@ -5011,20 +5011,20 @@ static void regulator_remove_coupling(struct regulator_dev *rdev)
>
> static int regulator_init_coupling(struct regulator_dev *rdev)
> {
> + struct regulator_dev **coupled;
> int err, n_phandles;
> - size_t alloc_size;
>
> if (!IS_ENABLED(CONFIG_OF))
> n_phandles = 0;
> else
> n_phandles = of_get_n_coupled(rdev);
>
> - alloc_size = sizeof(*rdev) * (n_phandles + 1);
> -
> - rdev->coupling_desc.coupled_rdevs = kzalloc(alloc_size, GFP_KERNEL);
> - if (!rdev->coupling_desc.coupled_rdevs)
> + coupled = kcalloc(n_phandles + 1, sizeof(*coupled), GFP_KERNEL);
> + if (!coupled)
> return -ENOMEM;
>
> + rdev->coupling_desc.coupled_rdevs = coupled;
> +
> /*
> * Every regulator should always have coupling descriptor filled with
> * at least pointer to itself.
>
Hello, Michał! Thank you for the patch! Not sure whether it's worthwhile
to backport this change since it's an improvement, I'll leave it to Mark
to decide, otherwise looks good to me.
Reviewed-by: Dmitry Osipenko <[email protected]>
On Sun, Aug 09, 2020 at 10:44:25PM +0300, Dmitry Osipenko wrote:
> 09.08.2020 22:21, Michał Mirosław пишет:
> > The code allocates sizeof(regulator_dev) for a pointer. Make it less
> > generous. Let kcalloc() calculate the size, while at it.
> >
> > Cc: [email protected]
> > Fixes: d8ca7d184b33 ("regulator: core: Introduce API for regulators coupling customization")
> Hello, Michał! Thank you for the patch! Not sure whether it's worthwhile
> to backport this change since it's an improvement, I'll leave it to Mark
> to decide, otherwise looks good to me.
Yeah, this is more a performance improvement than a fix.
On Mon, Aug 10, 2020 at 01:37:47PM +0100, Mark Brown wrote:
> On Sun, Aug 09, 2020 at 10:44:25PM +0300, Dmitry Osipenko wrote:
> > 09.08.2020 22:21, Michał Mirosław пишет:
> > > The code allocates sizeof(regulator_dev) for a pointer. Make it less
> > > generous. Let kcalloc() calculate the size, while at it.
> > >
> > > Cc: [email protected]
> > > Fixes: d8ca7d184b33 ("regulator: core: Introduce API for regulators coupling customization")
>
> > Hello, Michał! Thank you for the patch! Not sure whether it's worthwhile
> > to backport this change since it's an improvement, I'll leave it to Mark
> > to decide, otherwise looks good to me.
>
> Yeah, this is more a performance improvement than a fix.
Should I resend without Cc: stable?
Best Regards,
Michał Mirosław
On Mon, Aug 10, 2020 at 06:25:56PM +0200, Michał Mirosław wrote:
> On Mon, Aug 10, 2020 at 01:37:47PM +0100, Mark Brown wrote:
> > Yeah, this is more a performance improvement than a fix.
> Should I resend without Cc: stable?
I've already queued it to be applied to -next with that removed and an
edited subject line.
On Mon, Aug 10, 2020 at 06:33:37PM +0100, Mark Brown wrote:
> On Mon, Aug 10, 2020 at 06:25:56PM +0200, Micha? Miros?aw wrote:
> > On Mon, Aug 10, 2020 at 01:37:47PM +0100, Mark Brown wrote:
> > > Yeah, this is more a performance improvement than a fix.
> > Should I resend without Cc: stable?
> I've already queued it to be applied to -next with that removed and an
> edited subject line.
Ok, thanks!