rtc: sun6i: Use struct_size() in kzalloc()

One of the more common cases of allocation size calculations is finding the size of a structure that has a zero-sized array at the end, along with memory for some number of elements for that array. For example:

struct foo {
        int stuff;
        void *entry[];
};

instance = kzalloc(sizeof(struct foo) + sizeof(void *) * count, GFP_KERNEL);

Instead of leaving these open-coded and prone to type mistakes, we can now use the new struct_size() helper:

instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL);

Signed-off-by Gustavo A. R. Silva <[email protected]>

---
 drivers/rtc/rtc-sun6i.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

drivers/rtc/rtc-sun6i.c +0 -0

index 2cd5a7b..fe07310 100644

Forwarding a question about coccinelle and isomorphisms from Kees Cook

---------- Forwarded message ----------
From: Kees Cook <[email protected]>
To: "Gustavo A. R. Silva" <[email protected]>
Cc: Alessandro Zummo <[email protected]>, Alexandre Belloni <[email protected]>, Maxime Ripard <[email protected]>, Chen-Yu Tsai <[email protected]>, linux-rtc at vger.kernel.org, linux-arm-kernel <[email protected]>, LKML <[email protected]> Bcc:
Date: Thu, 23 Aug 2018 13:56:35 -0700
Subject: Re: [PATCH] rtc: sun6i: Use struct_size() in kzalloc()

Quoted 29 lines · Gustavo A. R. Silva

One of the more common cases of allocation size calculations is finding the size of a structure that has a zero-sized array at the end, along with memory for some number of elements for that array. For example:

struct foo {
        int stuff;
        void *entry[];
};

instance = kzalloc(sizeof(struct foo) + sizeof(void *) * count, GFP_KERNEL);

Instead of leaving these open-coded and prone to type mistakes, we can now use the new struct_size() helper:

instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL);

Signed-off-by Gustavo A. R. Silva <[email protected]>

---
 drivers/rtc/rtc-sun6i.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

drivers/rtc/rtc-sun6i.c +0 -0

index 2cd5a7b..fe07310 100644
@@ -199,8 +199,7 @@ static void __init sun6i_rtc_clk_init(struct device_node *node)
        if (!rtc)
                return;
 

drivers/rtc/rtc-sun6i.c static void __init sun6i_rtc_clk_init(struct device_node *node) · line 205

Quoted 6 lines

drivers/rtc/rtc-sun6i.c +1 -2

-       clk_data = kzalloc(sizeof(*clk_data) + (sizeof(*clk_data->hws) * 2),
-                          GFP_KERNEL);
+       clk_data = kzalloc(struct_size(clk_data, hws, 2), GFP_KERNEL);
        if (!clk_data) {
                kfree(rtc);
                return;

This looks like entirely correct to me, but I'm surprised the Coccinelle script didn't discover this. I guess the isomorphisms don't cover the parenthesis?

One of the more common cases of allocation size calculations is finding the size of a structure that has a zero-sized array at the end, along with memory for some number of elements for that array. For example:

struct foo {
        int stuff;
        void *entry[];
};

instance = kzalloc(sizeof(struct foo) + sizeof(void *) * count, GFP_KERNEL);

Instead of leaving these open-coded and prone to type mistakes, we can now use the new struct_size() helper:

instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL);

Signed-off-by Gustavo A. R. Silva <[email protected]>

---
 drivers/rtc/rtc-sun6i.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

drivers/rtc/rtc-sun6i.c +0 -0

index 2cd5a7b..fe07310 100644

Quoted 40 lines · Joe Perches

Forwarding a question about coccinelle and isomorphisms from Kees Cook

---------- Forwarded message ----------
From: Kees Cook <[email protected]>
To: "Gustavo A. R. Silva" <[email protected]>
Cc: Alessandro Zummo <[email protected]>, Alexandre Belloni <[email protected]>, Maxime Ripard <[email protected]>, Chen-Yu Tsai <[email protected]>, linux-rtc at vger.kernel.org, linux-arm-kernel <[email protected]>, LKML <[email protected]> Bcc:
Date: Thu, 23 Aug 2018 13:56:35 -0700
Subject: Re: [PATCH] rtc: sun6i: Use struct_size() in kzalloc()

Quoted 29 lines · Gustavo A. R. Silva

One of the more common cases of allocation size calculations is finding the size of a structure that has a zero-sized array at the end, along with memory for some number of elements for that array. For example:

struct foo {
        int stuff;
        void *entry[];
};

instance = kzalloc(sizeof(struct foo) + sizeof(void *) * count, GFP_KERNEL);

Instead of leaving these open-coded and prone to type mistakes, we can now use the new struct_size() helper:

instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL);

Signed-off-by Gustavo A. R. Silva <[email protected]>

---
 drivers/rtc/rtc-sun6i.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

drivers/rtc/rtc-sun6i.c +0 -0

index 2cd5a7b..fe07310 100644
@@ -199,8 +199,7 @@ static void __init sun6i_rtc_clk_init(struct device_node *node)
        if (!rtc)
                return;
 

Quoted 10 lines

drivers/rtc/rtc-sun6i.c static void __init sun6i_rtc_clk_init(struct device_node *node) · line 205

Quoted 6 lines

drivers/rtc/rtc-sun6i.c +1 -2

-       clk_data = kzalloc(sizeof(*clk_data) + (sizeof(*clk_data->hws) * 2),
-                          GFP_KERNEL);
+       clk_data = kzalloc(struct_size(clk_data, hws, 2), GFP_KERNEL);
        if (!clk_data) {
                kfree(rtc);
                return;

This looks like entirely correct to me, but I'm surprised the Coccinelle script didn't discover this. I guess the isomorphisms don't cover the parenthesis?

I had this:

@@
identifier alloc =~ "kmalloc|kzalloc|kmalloc_node|kzalloc_node|vmalloc|vzalloc|kvmalloc|kvzalloc"; identifier VAR, ELEMENT;
expression COUNT;
@@

- alloc(sizeof(*VAR) + COUNT * sizeof(*VAR->ELEMENT)
+ alloc(struct_size(VAR, ELEMENT, COUNT)
, ...)

But I needed to explicitly change the rule to:

(
- alloc(sizeof(*VAR) + COUNT * sizeof(*VAR->ELEMENT)
+ alloc(struct_size(VAR, ELEMENT, COUNT)
, ...)
|
- alloc(sizeof(*VAR) + (COUNT * sizeof(*VAR->ELEMENT))
+ alloc(struct_size(VAR, ELEMENT, COUNT)
, ...)
)

to add the ()s. I would expect arithmetic commutative expressions to match? But I had to add parens?

drivers/rtc/rtc-sun6i.c static void __init sun6i_rtc_clk_init(struct device_node *node) · line 205

Quoted 7 lines

drivers/rtc/rtc-sun6i.c +1 -2

-       clk_data = kzalloc(sizeof(*clk_data) + (sizeof(*clk_data->hws) * 2),
-                          GFP_KERNEL);
+       clk_data = kzalloc(struct_size(clk_data, hws, 2), GFP_KERNEL);
        if (!clk_data) {
                kfree(rtc);
                return;

This looks like entirely correct to me, but I'm surprised the Coccinelle script didn't discover this. I guess the isomorphisms don't cover the parenthesis?

Quoted 77 lines · Kees Cook

Quoted 40 lines · Joe Perches

Forwarding a question about coccinelle and isomorphisms from Kees Cook

---------- Forwarded message ----------
From: Kees Cook <[email protected]>
To: "Gustavo A. R. Silva" <[email protected]>
Cc: Alessandro Zummo <[email protected]>, Alexandre Belloni <[email protected]>, Maxime Ripard <[email protected]>, Chen-Yu Tsai <[email protected]>, linux-rtc at vger.kernel.org, linux-arm-kernel <[email protected]>, LKML <[email protected]> Bcc:
Date: Thu, 23 Aug 2018 13:56:35 -0700
Subject: Re: [PATCH] rtc: sun6i: Use struct_size() in kzalloc()

Quoted 29 lines · Gustavo A. R. Silva

One of the more common cases of allocation size calculations is finding the size of a structure that has a zero-sized array at the end, along with memory for some number of elements for that array. For example:

struct foo {
        int stuff;
        void *entry[];
};

instance = kzalloc(sizeof(struct foo) + sizeof(void *) * count, GFP_KERNEL);

Instead of leaving these open-coded and prone to type mistakes, we can now use the new struct_size() helper:

instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL);

Signed-off-by Gustavo A. R. Silva <[email protected]>

---
 drivers/rtc/rtc-sun6i.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

drivers/rtc/rtc-sun6i.c +0 -0

index 2cd5a7b..fe07310 100644
@@ -199,8 +199,7 @@ static void __init sun6i_rtc_clk_init(struct device_node *node)
        if (!rtc)
                return;
 

Quoted 10 lines

drivers/rtc/rtc-sun6i.c static void __init sun6i_rtc_clk_init(struct device_node *node) · line 205

Quoted 6 lines

drivers/rtc/rtc-sun6i.c +1 -2

-       clk_data = kzalloc(sizeof(*clk_data) + (sizeof(*clk_data->hws) * 2),
-                          GFP_KERNEL);
+       clk_data = kzalloc(struct_size(clk_data, hws, 2), GFP_KERNEL);
        if (!clk_data) {
                kfree(rtc);
                return;

This looks like entirely correct to me, but I'm surprised the Coccinelle script didn't discover this. I guess the isomorphisms don't cover the parenthesis?

I had this:

@@
identifier alloc =~ "kmalloc|kzalloc|kmalloc_node|kzalloc_node|vmalloc|vzalloc|kvmalloc|kvzalloc"; identifier VAR, ELEMENT;
expression COUNT;
@@

- alloc(sizeof(*VAR) + COUNT * sizeof(*VAR->ELEMENT)
+ alloc(struct_size(VAR, ELEMENT, COUNT)
, ...)

But I needed to explicitly change the rule to:

(
- alloc(sizeof(*VAR) + COUNT * sizeof(*VAR->ELEMENT)
+ alloc(struct_size(VAR, ELEMENT, COUNT)
, ...)
|
- alloc(sizeof(*VAR) + (COUNT * sizeof(*VAR->ELEMENT))
+ alloc(struct_size(VAR, ELEMENT, COUNT)
, ...)
)

Quoted 2 lines

to add the ()s. I would expect arithmetic commutative expressions to match? But I had to add parens?

Isomorphisms don't add parens. They only remove them. If they added them, you would end up with the possibility of having them everywhere, in all permutations, which would be slow and useless.

Quoted 81 lines · Julia Lawall

Quoted 77 lines · Kees Cook

Quoted 40 lines · Joe Perches

Forwarding a question about coccinelle and isomorphisms from Kees Cook

---------- Forwarded message ----------
From: Kees Cook <[email protected]>
To: "Gustavo A. R. Silva" <[email protected]>
Cc: Alessandro Zummo <[email protected]>, Alexandre Belloni <[email protected]>, Maxime Ripard <[email protected]>, Chen-Yu Tsai <[email protected]>, linux-rtc at vger.kernel.org, linux-arm-kernel <[email protected]>, LKML <[email protected]> Bcc:
Date: Thu, 23 Aug 2018 13:56:35 -0700
Subject: Re: [PATCH] rtc: sun6i: Use struct_size() in kzalloc()

Quoted 29 lines · Gustavo A. R. Silva

One of the more common cases of allocation size calculations is finding the size of a structure that has a zero-sized array at the end, along with memory for some number of elements for that array. For example:

struct foo {
        int stuff;
        void *entry[];
};

instance = kzalloc(sizeof(struct foo) + sizeof(void *) * count, GFP_KERNEL);

Instead of leaving these open-coded and prone to type mistakes, we can now use the new struct_size() helper:

instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL);

Signed-off-by Gustavo A. R. Silva <[email protected]>

---
 drivers/rtc/rtc-sun6i.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

drivers/rtc/rtc-sun6i.c +0 -0

index 2cd5a7b..fe07310 100644
@@ -199,8 +199,7 @@ static void __init sun6i_rtc_clk_init(struct device_node *node)
        if (!rtc)
                return;
 

Quoted 10 lines

drivers/rtc/rtc-sun6i.c static void __init sun6i_rtc_clk_init(struct device_node *node) · line 205

Quoted 6 lines

drivers/rtc/rtc-sun6i.c +1 -2

-       clk_data = kzalloc(sizeof(*clk_data) + (sizeof(*clk_data->hws) * 2),
-                          GFP_KERNEL);
+       clk_data = kzalloc(struct_size(clk_data, hws, 2), GFP_KERNEL);
        if (!clk_data) {
                kfree(rtc);
                return;

This looks like entirely correct to me, but I'm surprised the Coccinelle script didn't discover this. I guess the isomorphisms don't cover the parenthesis?

I had this:

@@
identifier alloc =~ "kmalloc|kzalloc|kmalloc_node|kzalloc_node|vmalloc|vzalloc|kvmalloc|kvzalloc"; identifier VAR, ELEMENT;
expression COUNT;
@@

- alloc(sizeof(*VAR) + COUNT * sizeof(*VAR->ELEMENT)
+ alloc(struct_size(VAR, ELEMENT, COUNT)
, ...)

But I needed to explicitly change the rule to:

(
- alloc(sizeof(*VAR) + COUNT * sizeof(*VAR->ELEMENT)
+ alloc(struct_size(VAR, ELEMENT, COUNT)
, ...)
|
- alloc(sizeof(*VAR) + (COUNT * sizeof(*VAR->ELEMENT))
+ alloc(struct_size(VAR, ELEMENT, COUNT)
, ...)
)

Quoted 6 lines

Quoted 2 lines

to add the ()s. I would expect arithmetic commutative expressions to match? But I had to add parens?

Isomorphisms don't add parens. They only remove them. If they added them, you would end up with the possibility of having them everywhere, in all permutations, which would be slow and useless.

Would a rule for:

a + (b * c)

match:

a + b * c

Quoted 90 lines · Kees Cook

Quoted 82 lines · Julia Lawall

Quoted 78 lines · Kees Cook

Quoted 40 lines · Joe Perches

Forwarding a question about coccinelle and isomorphisms from Kees Cook

---------- Forwarded message ----------
From: Kees Cook <[email protected]>
To: "Gustavo A. R. Silva" <[email protected]>
Cc: Alessandro Zummo <[email protected]>, Alexandre Belloni <[email protected]>, Maxime Ripard <[email protected]>, Chen-Yu Tsai <[email protected]>, linux-rtc at vger.kernel.org, linux-arm-kernel <[email protected]>, LKML <[email protected]> Bcc:
Date: Thu, 23 Aug 2018 13:56:35 -0700
Subject: Re: [PATCH] rtc: sun6i: Use struct_size() in kzalloc()

Quoted 29 lines · Gustavo A. R. Silva

One of the more common cases of allocation size calculations is finding the size of a structure that has a zero-sized array at the end, along with memory for some number of elements for that array. For example:

struct foo {
        int stuff;
        void *entry[];
};

instance = kzalloc(sizeof(struct foo) + sizeof(void *) * count, GFP_KERNEL);

Instead of leaving these open-coded and prone to type mistakes, we can now use the new struct_size() helper:

instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL);

Signed-off-by Gustavo A. R. Silva <[email protected]>

---
 drivers/rtc/rtc-sun6i.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

drivers/rtc/rtc-sun6i.c +0 -0

index 2cd5a7b..fe07310 100644
@@ -199,8 +199,7 @@ static void __init sun6i_rtc_clk_init(struct device_node *node)
        if (!rtc)
                return;
 

Quoted 10 lines

drivers/rtc/rtc-sun6i.c static void __init sun6i_rtc_clk_init(struct device_node *node) · line 205

Quoted 6 lines

drivers/rtc/rtc-sun6i.c +1 -2

-       clk_data = kzalloc(sizeof(*clk_data) + (sizeof(*clk_data->hws) * 2),
-                          GFP_KERNEL);
+       clk_data = kzalloc(struct_size(clk_data, hws, 2), GFP_KERNEL);
        if (!clk_data) {
                kfree(rtc);
                return;

This looks like entirely correct to me, but I'm surprised the Coccinelle script didn't discover this. I guess the isomorphisms don't cover the parenthesis?

I had this:

@@
identifier alloc =~ "kmalloc|kzalloc|kmalloc_node|kzalloc_node|vmalloc|vzalloc|kvmalloc|kvzalloc"; identifier VAR, ELEMENT;
expression COUNT;
@@

- alloc(sizeof(*VAR) + COUNT * sizeof(*VAR->ELEMENT)
+ alloc(struct_size(VAR, ELEMENT, COUNT)
, ...)

But I needed to explicitly change the rule to:

(
- alloc(sizeof(*VAR) + COUNT * sizeof(*VAR->ELEMENT)
+ alloc(struct_size(VAR, ELEMENT, COUNT)
, ...)

Quoted 1 line

- alloc(sizeof(*VAR) + (COUNT * sizeof(*VAR->ELEMENT))
+ alloc(struct_size(VAR, ELEMENT, COUNT)
, ...)
)

Quoted 6 lines

Quoted 2 lines

to add the ()s. I would expect arithmetic commutative expressions to match? But I had to add parens?

Isomorphisms don't add parens. They only remove them. If they added them, you would end up with the possibility of having them everywhere, in all permutations, which would be slow and useless.

Quoted 7 lines

Would a rule for:

a + (b * c)

match:

a + b * c

I would say yes. Basically it removes the parentheses but doesn't reparse the code, so it doesn't redo the associativity. Since * has higher precedence than +, then both will be matched. On the other hand, if you put:

(a + b) * c

It will consider a pattern with the parentheses removed, but that pattern won't match anything, because real trees that consist of a + b * c are parsed in a different way.

Quoted 8 lines · Julia Lawall

On Thu, 23 Aug 2018, Kees Cook wrote:

(a + b) * c

It will consider a pattern with the parentheses removed, but that pattern won't match anything, because real trees that consist of a + b * c are parsed in a different way.

spatch 1.0.4 doesn't seem to:

$ spatch --version
spatch version 1.0.4 with Python support and with PCRE support
$ cat match_mul.cocci
@@
expression a, b, c;
int d;
@@

* d = a * b + c;

Quoted 20 lines · Joe Perches

Quoted 8 lines · Julia Lawall

On Thu, 23 Aug 2018, Kees Cook wrote:

(a + b) * c

It will consider a pattern with the parentheses removed, but that pattern won't match anything, because real trees that consist of a + b * c are parsed in a different way.

spatch 1.0.4 doesn't seem to:

$ spatch --version
spatch version 1.0.4 with Python support and with PCRE support
$ cat match_mul.cocci
@@
expression a, b, c;
int d;
@@

Quoted 1 line

* d = a * b + c;

But "(a * b) + c" for the rule does:

$ cat isomorphisms.c
#include <stdio.h>
#include <stdlib.h>

int main(int argc, char *argv[])

{
        int a, b, c;

        a = atoi(argv[1]);
        b = atoi(argv[2]);
        c = atoi(argv[3]);

        printf("%d\n", a + b * c);
        printf("%d\n", (a + b) * c);
        printf("%d\n", a + (b * c));
        printf("%d\n", b * c + a);
        printf("%d\n", b * (c + a));
        printf("%d\n", (b * c) + a);

        return 0;
}
$ cat match.cocci
@@
identifier A, B, C;
@@

* (A * B) + C

$ spatch -sp-file match.cocci isomorphisms.c
init_defs_builtins: /usr/lib/coccinelle/standard.h
HANDLING: isomorphisms.c
diff =

tmp/cocci-output-8402-870fd6-isomorphisms.c renamed from isomorphisms.c +0 -4

@@ -9,12 +9,8 @@ int main(int argc, char *argv[])
        b = atoi(argv[2]);
        c = atoi(argv[3]);
 
-       printf("%d\n", a + b * c);
        printf("%d\n", (a + b) * c);
-       printf("%d\n", a + (b * c));
-       printf("%d\n", b * c + a);
        printf("%d\n", b * (c + a));
-       printf("%d\n", (b * c) + a);
 
        return 0;
 }

So it sounds like I should put parens around all kinds of things in my rules...

Quoted 20 lines · Joe Perches

Quoted 8 lines · Julia Lawall

On Thu, 23 Aug 2018, Kees Cook wrote:

(a + b) * c

It will consider a pattern with the parentheses removed, but that pattern won't match anything, because real trees that consist of a + b * c are parsed in a different way.

spatch 1.0.4 doesn't seem to:

$ spatch --version
spatch version 1.0.4 with Python support and with PCRE support
$ cat match_mul.cocci
@@
expression a, b, c;
int d;
@@

Quoted 1 line

* d = a * b + c;

But "(a * b) + c" for the rule does:

Quoted 21 lines · Kees Cook

Quoted 20 lines · Joe Perches

Quoted 8 lines · Julia Lawall

On Thu, 23 Aug 2018, Kees Cook wrote:

(a + b) * c

It will consider a pattern with the parentheses removed, but that pattern won't match anything, because real trees that consist of a + b * c are parsed in a different way.

spatch 1.0.4 doesn't seem to:

$ spatch --version
spatch version 1.0.4 with Python support and with PCRE support
$ cat match_mul.cocci
@@
expression a, b, c;
int d;
@@

Quoted 3 lines

Quoted 1 line

* d = a * b + c;

But "(a * b) + c" for the rule does:

Yes, but not for other valid uses like below:

$ cat match_mul.cocci
@@
expression a, b, c;
int d;
@@

* d = (a * b) + c;

$ cat test_mul.c
int a, b, c, d;

void foo(void)

{
	d = ((a * b) + c);
	d = ((a * b)) + c;
	d = (a * b) + c;
	d = a * b + c;
}
$ spatch -sp-file match_mul.cocci test_mul.c
init_defs_builtins: /usr/lib/coccinelle/standard.h
HANDLING: test_mul.c
diff =

Quoted 52 lines · Joe Perches

Quoted 21 lines · Kees Cook

Quoted 20 lines · Joe Perches

Quoted 8 lines · Julia Lawall

On Thu, 23 Aug 2018, Kees Cook wrote:

(a + b) * c

It will consider a pattern with the parentheses removed, but that pattern won't match anything, because real trees that consist of a + b * c are parsed in a different way.

spatch 1.0.4 doesn't seem to:

$ spatch --version
spatch version 1.0.4 with Python support and with PCRE support
$ cat match_mul.cocci
@@
expression a, b, c;
int d;
@@

Quoted 3 lines

Quoted 1 line

* d = a * b + c;

But "(a * b) + c" for the rule does:

Yes, but not for other valid uses like below:

$ cat match_mul.cocci
@@
expression a, b, c;
int d;
@@

* d = (a * b) + c;

$ cat test_mul.c
int a, b, c, d;

void foo(void)

{
	d = ((a * b) + c);
	d = ((a * b)) + c;
	d = (a * b) + c;
	d = a * b + c;
}
$ spatch -sp-file match_mul.cocci test_mul.c
init_defs_builtins: /usr/lib/coccinelle/standard.h
HANDLING: test_mul.c
diff =

tmp/cocci-output-23856-7e83f7-test_mul.c void foo(void) · line 7

Quoted 6 lines

tmp/cocci-output-23856-7e83f7-test_mul.c renamed from test_mul.c +0 -2

@@ -4,6 +4,4 @@ void foo(void)
 {
 	d = ((a * b) + c);
 	d = ((a * b)) + c;
-	d = (a * b) + c;
-	d = a * b + c;
 }

I'm not sure to get the point.

If you think that people may fully parenthesize the code, then ((a * b) +
c) would be a reasonable option for the pattern. I'm not sure why double parentheses, ((a * b)) + c, should exist in the first place.

One of the more common cases of allocation size calculations is finding the size of a structure that has a zero-sized array at the end, along with memory for some number of elements for that array. For example:

struct foo { 
	int stuff;
        void *entry[];
};

instance = kzalloc(sizeof(struct foo) + sizeof(void *) * count, GFP_KERNEL);

Instead of leaving these open-coded and prone to type mistakes, we can now use the new struct_size() helper:

instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL);

Signed-off-by Gustavo A. R. Silva <[email protected]>

---
 drivers/rtc/rtc-sun6i.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)