Re: [PATCH] drm/msm/dpu: avoid open-coded 64-bit division

[Georgi Djakov]

On 27.05.20 16:35, Arnd Bergmann wrote:
> 64-bit integer division is normally not allowed in the kernel
> because of the large overhead on 32-bit machines:
> 
> drivers/gpu/drm/msm/disp/dpu1/dpu_core_perf.o: in function `_dpu_core_perf_crtc_update_bus':
> dpu_core_perf.c:(.text+0x810): undefined reference to `__aeabi_uldivmod'
> 
> The function already contains a call to do_div(), so I assume this
> is never called in a performance critical context, and we can
> use div_u64 for the second one as well.
> 
> Fixes: 04d9044f6c57 ("drm/msm/dpu: add support for clk and bw scaling for display")
> Signed-off-by: Arnd Bergmann <arnd@xxxxxxxx>

Reviewed-by: Georgi Djakov <georgi.djakov@xxxxxxxxxx>

Thanks for the patch Arnd. I just noticed this in linux-next.

Rob, could you please take it into msm-next, as the patch which
triggers the issue is also there?

Thanks,
Georgi

> ---
>  include/linux/interconnect.h | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/include/linux/interconnect.h b/include/linux/interconnect.h
> index 3a63d98613fc..8279fe9b2082 100644
> --- a/include/linux/interconnect.h
> +++ b/include/linux/interconnect.h
> @@ -11,7 +11,7 @@
>  #include <linux/types.h>
>  
>  /* macros for converting to icc units */
> -#define Bps_to_icc(x)	((x) / 1000)
> +#define Bps_to_icc(x)	div_u64((x), 1000)
>  #define kBps_to_icc(x)	(x)
>  #define MBps_to_icc(x)	((x) * 1000)
>  #define GBps_to_icc(x)	((x) * 1000 * 1000)
>