Re: [PATCH v2] drm/mediatek: separate color module to fixup error memory reallocation

2017-06-19 Thread CK Hu
On Fri, 2017-06-16 at 22:02 +0800, YT Shen wrote: > Previous patch (c5f228ef6c drm/mediatek: add *driver_data for different > hardware settings) calls devm_kfree() and then devm_kzalloc() to > reallocate color module data structure. But this reallocation cannnot > guarantee the new address is

Re: [PATCH v2] drm/mediatek: separate color module to fixup error memory reallocation

2017-06-19 Thread CK Hu
On Fri, 2017-06-16 at 22:02 +0800, YT Shen wrote: > Previous patch (c5f228ef6c drm/mediatek: add *driver_data for different > hardware settings) calls devm_kfree() and then devm_kzalloc() to > reallocate color module data structure. But this reallocation cannnot > guarantee the new address is

[PATCH v2] drm/mediatek: separate color module to fixup error memory reallocation

2017-06-16 Thread YT Shen
Previous patch (c5f228ef6c drm/mediatek: add *driver_data for different hardware settings) calls devm_kfree() and then devm_kzalloc() to reallocate color module data structure. But this reallocation cannnot guarantee the new address is unchanged, but the caller will use the old address, which is

[PATCH v2] drm/mediatek: separate color module to fixup error memory reallocation

2017-06-16 Thread YT Shen
Previous patch (c5f228ef6c drm/mediatek: add *driver_data for different hardware settings) calls devm_kfree() and then devm_kzalloc() to reallocate color module data structure. But this reallocation cannnot guarantee the new address is unchanged, but the caller will use the old address, which is