Skip to contentSkip to conversation

doc: ioctl: Add some clarifications to botching-up-ioctls

PATCH Applied Laura Abbott 3 participants 3 messages Sep 2, 2016 – Sep 6, 2016 Documentation

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 doc: ioctl: Add some clarifications to botching-up-ioctls

- The guide currently says to pad the structure to a multiple of 64-bits. This is not necessary in cases where the structure contains no 64-bit types. Clarify this concept to avoid unnecessary padding.
- When using __u64 to hold user pointers, blindly trying to do a cast to a void __user * may generate a warning on 32-bit systems about a cast from an integer to a pointer of different size. There is a macro to deal with this which hides an ugly double cast. Add a reference to this macro.

---
I can split these into two separate patches if necessary but it seemed simpler just to colapse them into a single patch.

---
 Documentation/ioctl/botching-up-ioctls.txt | 13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)
Documentation/ioctl/botching-up-ioctls.txt +8 -5
index cc30b14..36138c6 100644
@@ -34,15 +34,18 @@ will need to add a a 32-bit compat layer:
    64-bit platforms do. So we always need padding to the natural size to get
    this right.
 
- * Pad the entire struct to a multiple of 64-bits - the structure size will
-   otherwise differ on 32-bit versus 64-bit. Having a different structure size
-   hurts when passing arrays of structures to the kernel, or if the kernel
-   checks the structure size, which e.g. the drm core does.
+ * Pad the entire struct to a multiple of 64-bits if the structure contains
+   64-bit types - the structure size will otherwise differ on 32-bit versus
+   64-bit. Having a different structure size hurts when passing arrays of
+   structures to the kernel, or if the kernel checks the structure size, which
+   e.g. the drm core does.
 
  * Pointers are __u64, cast from/to a uintprt_t on the userspace side and
    from/to a void __user * in the kernel. Try really hard not to delay this
    conversion or worse, fiddle the raw __u64 through your code since that
-   diminishes the checking tools like sparse can provide.
+   diminishes the checking tools like sparse can provide. The macro
+   u64_to_user_ptr can be used in the kernel to avoid warnings about integers
+   and pointres of different sizes.
 
 
 Basics
Signature · 1 line
2.7.4
Acked-by on v1 1/1
#1 lore raw
Quoted 11 lines · CEST Laura Abbott

- The guide currently says to pad the structure to a multiple of 64-bits. This is not necessary in cases where the structure contains no 64-bit types. Clarify this concept to avoid unnecessary padding.
- When using __u64 to hold user pointers, blindly trying to do a cast to a void __user * may generate a warning on 32-bit systems about a cast from an integer to a pointer of different size. There is a macro to deal with this which hides an ugly double cast. Add a reference to this macro.

Looks good to me,

maintainer Applied
#2 lore raw
Quoted 10 lines · On Sat, 03 Sep 2016 00:44:02 +0200 Arnd Bergmann <[email protected]> wrote:
Quoted 9 lines · CEST Laura Abbott

- The guide currently says to pad the structure to a multiple of 64-bits. This is not necessary in cases where the structure contains no 64-bit types. Clarify this concept to avoid unnecessary padding.
- When using __u64 to hold user pointers, blindly trying to do a cast to a void __user * may generate a warning on 32-bit systems about a cast from an integer to a pointer of different size. There is a macro to deal with this which hides an ugly double cast. Add a reference to this macro.

Quoted 6 lines
Quoted 2 lines

Looks good to me,

Applied to the docs tree, thanks.

jon