[PATCH] doc: botching-up-ioctls: Make it clearer why structs must be padded

[PATCH] doc: botching-up-ioctls: Make it clearer why structs must be padded

[Daniel Vetter]

This came up in discussions when reviewing drm patches.

Cc: Eric Anholt <eric@anholt.net>
Cc: linux-doc@vger.kernel.org
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>

--

Aside: I wonder whether we shouldn't move this to some other place and
rst-ify it? Any good suggestions?
-Daniel
---
 Documentation/ioctl/botching-up-ioctls.txt | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/Documentation/ioctl/botching-up-ioctls.txt b/Documentation/ioctl/botching-up-ioctls.txt
index d02cfb48901c..883fb034bd04 100644
--- a/Documentation/ioctl/botching-up-ioctls.txt
+++ b/Documentation/ioctl/botching-up-ioctls.txt
@@ -73,7 +73,9 @@ will have a second iteration or at least an extension for any given interface.
    future extensions is going right down the gutters since someone will submit
    an ioctl struct with random stack garbage in the yet unused parts. Which
    then bakes in the ABI that those fields can never be used for anything else
-   but garbage.
+   but garbage. This is also the reason why you must explicitly pad all
+   structures, even if you never use them in an array - the padding the compiler
+   might insert could contain garbage.
 
  * Have simple testcases for all of the above.
 
-- 
2.17.0

Re: [PATCH] doc: botching-up-ioctls: Make it clearer why structs must be padded

[Eric Anholt]

[-- Attachment #1: Type: text/plain, Size: 1538 bytes --]

Daniel Vetter <daniel.vetter@ffwll.ch> writes:

> This came up in discussions when reviewing drm patches.
>
> Cc: Eric Anholt <eric@anholt.net>
> Cc: linux-doc@vger.kernel.org
> Cc: Jonathan Corbet <corbet@lwn.net>
> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
>
> --
>
> Aside: I wonder whether we shouldn't move this to some other place and
> rst-ify it? Any good suggestions?
> -Daniel
> ---
>  Documentation/ioctl/botching-up-ioctls.txt | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/Documentation/ioctl/botching-up-ioctls.txt b/Documentation/ioctl/botching-up-ioctls.txt
> index d02cfb48901c..883fb034bd04 100644
> --- a/Documentation/ioctl/botching-up-ioctls.txt
> +++ b/Documentation/ioctl/botching-up-ioctls.txt
> @@ -73,7 +73,9 @@ will have a second iteration or at least an extension for any given interface.
>     future extensions is going right down the gutters since someone will submit
>     an ioctl struct with random stack garbage in the yet unused parts. Which
>     then bakes in the ABI that those fields can never be used for anything else
> -   but garbage.
> +   but garbage. This is also the reason why you must explicitly pad all
> +   structures, even if you never use them in an array - the padding the compiler
> +   might insert could contain garbage.

I hadn't realized that we had this document in git, or I probably would
have written this patch.  I think this makes it clear enough how I got
vc4 and v3d wrong.  Thanks!

Reviewed-by: Eric Anholt <eric@anholt.net>

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]

Re: [PATCH] doc: botching-up-ioctls: Make it clearer why structs must be padded

[Jonathan Corbet]

On Wed,  2 May 2018 09:51:06 +0200
Daniel Vetter <daniel.vetter@ffwll.ch> wrote:

> This came up in discussions when reviewing drm patches.

Applied, thanks.

> Aside: I wonder whether we shouldn't move this to some other place and
> rst-ify it? Any good suggestions?

For the moment, probably in Documentation/process, next to
volatile-considered-harmful.rst and such.  Even better, of course, would
be to have some nice document on designing user-space APIs in general...
one can always dream ... :)

Thanks,

jon