Most fbdev drivers depend on framebuffer_alloc() to initialize the
allocated memory to 0. Document this guarantee.
Suggested-by: Miguel Ojeda <[email protected]>
Signed-off-by: Thomas Zimmermann <[email protected]>
Cc: Helge Deller <[email protected]>
---
drivers/video/fbdev/core/fb_info.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/video/fbdev/core/fb_info.c b/drivers/video/fbdev/core/fb_info.c
index 8bdbefdd4b70..1b7615fc393d 100644
--- a/drivers/video/fbdev/core/fb_info.c
+++ b/drivers/video/fbdev/core/fb_info.c
@@ -13,7 +13,8 @@
*
* Creates a new frame buffer info structure. Also reserves @size bytes
* for driver private data (info->par). info->par (if any) will be
- * aligned to sizeof(long).
+ * aligned to sizeof(long). Both, the instance of struct fb_info and
+ * the driver private data, are cleared to zero.
*
* Returns the new structure, or NULL if an error occurred.
*
--
2.41.0
On Thu, Jul 13, 2023 at 3:03 PM Thomas Zimmermann <[email protected]> wrote:
>
> Most fbdev drivers depend on framebuffer_alloc() to initialize the
> allocated memory to 0. Document this guarantee.
>
> Suggested-by: Miguel Ojeda <[email protected]>
> Signed-off-by: Thomas Zimmermann <[email protected]>
> Cc: Helge Deller <[email protected]>
Thanks for sending this! Maybe this would be best earlier in the
series, so that later patches make more sense (since they use the
guarantee), but it is not a big deal.
> + * aligned to sizeof(long). Both, the instance of struct fb_info and
> + * the driver private data, are cleared to zero.
I think both commas may be best omitted (but I am not a native speaker).
Reviewed-by: Miguel Ojeda <[email protected]>
Cheers,
Miguel
On 7/13/23 06:21, Miguel Ojeda wrote:
> On Thu, Jul 13, 2023 at 3:03 PM Thomas Zimmermann <[email protected]> wrote:
>>
>> Most fbdev drivers depend on framebuffer_alloc() to initialize the
>> allocated memory to 0. Document this guarantee.
>>
>> Suggested-by: Miguel Ojeda <[email protected]>
>> Signed-off-by: Thomas Zimmermann <[email protected]>
>> Cc: Helge Deller <[email protected]>
>
> Thanks for sending this! Maybe this would be best earlier in the
> series, so that later patches make more sense (since they use the
> guarantee), but it is not a big deal.
>
>> + * aligned to sizeof(long). Both, the instance of struct fb_info and
>> + * the driver private data, are cleared to zero.
>
> I think both commas may be best omitted (but I am not a native speaker).
Yes, it would be better to omit them.
> Reviewed-by: Miguel Ojeda <[email protected]>
>
> Cheers,
> Miguel
--
~Randy