At Tue, 1 Jul 2014 17:33:25 +0530, Sachin Kamat wrote:
On Tue, Jul 1, 2014 at 5:26 PM, Takashi Iwai tiwai@suse.de wrote:
At Tue, 1 Jul 2014 17:02:54 +0530, Sachin Kamat wrote:
'status' is not used in the function. Remove it.
Signed-off-by: Sachin Kamat sachin.kamat@samsung.com Cc: Stephen Warren swarren@wwwdotorg.org Cc: Thierry Reding thierry.reding@gmail.com
All patches in this series only compile tested.
Removing such a line is unsafe, thus it must be tested. It changes the hardware access pattern and sometimes the read is intentionally there although the value isn't used.
That said, without testing, I won't apply such patches.
Right. I understand. That is the reason I explicitly mentioned that patches are untested. However, in the above cases how about removing the local variable (return value) and just keeping the read/write calls?
If the call is needed, yes, at best with some comments. But until someone really tests that it's unnecessary, there is no big reason to touch the code.
Takashi