[PATCH] staging: media: davinci_vpfe: Remove variable vpfe_dev

[PATCH] staging: media: davinci_vpfe: Remove variable vpfe_dev

[Nishka Dasgupta]

Remove variable vpfe_dev and replace it with its value (since the
function otherwise uses values directly instead of local variables).
Issue found with Coccinelle.

Signed-off-by: Nishka Dasgupta <nishkadg.linux@gmail.com>
---
 drivers/staging/media/davinci_vpfe/vpfe_video.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/staging/media/davinci_vpfe/vpfe_video.c b/drivers/staging/media/davinci_vpfe/vpfe_video.c
index 510202a3b091..8927b744b13e 100644
--- a/drivers/staging/media/davinci_vpfe/vpfe_video.c
+++ b/drivers/staging/media/davinci_vpfe/vpfe_video.c
@@ -448,7 +448,6 @@ vpfe_video_get_next_buffer(struct vpfe_video_device *video)
 /* schedule the next buffer which is available on dma queue */
 void vpfe_video_schedule_next_buffer(struct vpfe_video_device *video)
 {
-	struct vpfe_device *vpfe_dev = video->vpfe_dev;
 	unsigned long addr;
 
 	if (list_empty(&video->dma_queue))
@@ -463,19 +462,18 @@ void vpfe_video_schedule_next_buffer(struct vpfe_video_device *video)
 	list_del(&video->next_frm->list);
 	video->next_frm->vb.vb2_buf.state = VB2_BUF_STATE_ACTIVE;
 	addr = vb2_dma_contig_plane_dma_addr(&video->next_frm->vb.vb2_buf, 0);
-	video->ops->queue(vpfe_dev, addr);
+	video->ops->queue(video->vpfe_dev, addr);
 	video->state = VPFE_VIDEO_BUFFER_QUEUED;
 }
 
 /* schedule the buffer for capturing bottom field */
 void vpfe_video_schedule_bottom_field(struct vpfe_video_device *video)
 {
-	struct vpfe_device *vpfe_dev = video->vpfe_dev;
 	unsigned long addr;
 
 	addr = vb2_dma_contig_plane_dma_addr(&video->cur_frm->vb.vb2_buf, 0);
 	addr += video->field_off;
-	video->ops->queue(vpfe_dev, addr);
+	video->ops->queue(video->vpfe_dev, addr);
 }
 
 /* make buffer available for dequeue */
-- 
2.19.1

Re: [PATCH] staging: media: davinci_vpfe: Remove variable vpfe_dev

[Greg KH]

On Fri, May 31, 2019 at 02:17:18AM +0530, Nishka Dasgupta wrote:
> Remove variable vpfe_dev and replace it with its value (since the
> function otherwise uses values directly instead of local variables).

This says _what_ you do.  But we can see that in the patch itself.

You need to say _why_ you are doing this.

There's no need for this change at all.  Again, as I have said before,
we write code for developers to read first, the compiler second.  By
making these types of changes you are making it harder to
read/understand by a developer, and providing absolutely no benifit to
the compiler at all.

So it's actually making the code worse!

not good at all.

Please reconsider this type of change, as I keep asking you to.

greg k-h

Re: [PATCH] staging: media: davinci_vpfe: Remove variable vpfe_dev

[Nishka Dasgupta]

On 31/05/19 2:25 AM, Greg KH wrote:
> On Fri, May 31, 2019 at 02:17:18AM +0530, Nishka Dasgupta wrote:
>> Remove variable vpfe_dev and replace it with its value (since the
>> function otherwise uses values directly instead of local variables).
> 
> This says _what_ you do.  But we can see that in the patch itself.
> 
> You need to say _why_ you are doing this.
> 
> There's no need for this change at all.  Again, as I have said before,
> we write code for developers to read first, the compiler second.  By
> making these types of changes you are making it harder to
> read/understand by a developer, and providing absolutely no benifit to
> the compiler at all.
> 
> So it's actually making the code worse!
> 
> not good at all.
> 
> Please reconsider this type of change, as I keep asking you to.

Okay. In this case I thought it wouldn't make it worse since the 
function is low on local variables anyway? Clearly I was wrong, so I 
won't count this case as an exception in future.

Thanking you,
Nishka

> greg k-h
> 

Re: [PATCH] staging: media: davinci_vpfe: Remove variable vpfe_dev

[Greg KH]

On Fri, May 31, 2019 at 02:39:23AM +0530, Nishka Dasgupta wrote:
> On 31/05/19 2:25 AM, Greg KH wrote:
> > On Fri, May 31, 2019 at 02:17:18AM +0530, Nishka Dasgupta wrote:
> > > Remove variable vpfe_dev and replace it with its value (since the
> > > function otherwise uses values directly instead of local variables).
> > 
> > This says _what_ you do.  But we can see that in the patch itself.
> > 
> > You need to say _why_ you are doing this.
> > 
> > There's no need for this change at all.  Again, as I have said before,
> > we write code for developers to read first, the compiler second.  By
> > making these types of changes you are making it harder to
> > read/understand by a developer, and providing absolutely no benifit to
> > the compiler at all.
> > 
> > So it's actually making the code worse!
> > 
> > not good at all.
> > 
> > Please reconsider this type of change, as I keep asking you to.
> 
> Okay. In this case I thought it wouldn't make it worse since the function is
> low on local variables anyway? Clearly I was wrong, so I won't count this
> case as an exception in future.

Did you check the object file output to verify this?  Try it and see :)