Re: [PATCH] counter: ti-eqep: Use container_of instead of struct counter_device::priv

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



hello William,

On Thu, Dec 16, 2021 at 02:25:03PM +0900, William Breathitt Gray wrote:
> On Mon, Dec 13, 2021 at 12:43:12PM +0100, Uwe Kleine-König wrote:
> > Using counter->priv is a memory read and so more expensive than
> > container_of which is only an addition. (In this case even a noop
> > because the offset is 0.)
> > 
> > So container_of is expected to be a tad faster, it's type-safe, and
> > produces smaller code (ARCH=arm allmodconfig):
> > 
> > 	$ source/scripts/bloat-o-meter drivers/counter/ti-eqep.o-pre drivers/counter/ti-eqep.o
> > 	add/remove: 0/0 grow/shrink: 0/9 up/down: 0/-108 (-108)
> > 	Function                                     old     new   delta
> > 	ti_eqep_position_enable_write                132     120     -12
> > 	ti_eqep_position_enable_read                 260     248     -12
> > 	ti_eqep_position_ceiling_write               132     120     -12
> > 	ti_eqep_position_ceiling_read                236     224     -12
> > 	ti_eqep_function_write                       220     208     -12
> > 	ti_eqep_function_read                        372     360     -12
> > 	ti_eqep_count_write                          312     300     -12
> > 	ti_eqep_count_read                           236     224     -12
> > 	ti_eqep_action_read                          664     652     -12
> > 	Total: Before=4598, After=4490, chg -2.35%
> > 
> > Signed-off-by: Uwe Kleine-König <u.kleine-koenig@xxxxxxxxxxxxxx>
> 
> I'll pick this up and submit it with the rest of the Counter changes for
> this cycle.

This is great. The same transformation could be done for the other
drivers using the priv pointer. Then priv could be removed from struct
counter_device. Good idea?

A quick prototype patch yields:

drivers/counter/104-quad-8.o
  add/remove: 0/0 grow/shrink: 5/17 up/down: 76/-172 (-96)
  Total: Before=11802, After=11706, chg -0.81%
drivers/counter/ftm-quaddec.o
  add/remove: 0/0 grow/shrink: 0/5 up/down: 0/-60 (-60)
  Total: Before=5096, After=5036, chg -1.18%
drivers/counter/intel-qep.o
  add/remove: 0/0 grow/shrink: 0/10 up/down: 0/-116 (-116)
  Total: Before=4867, After=4751, chg -2.38%
drivers/counter/interrupt-cnt.o
  add/remove: 0/0 grow/shrink: 0/6 up/down: 0/-60 (-60)
  Total: Before=2841, After=2781, chg -2.11%
drivers/counter/microchip-tcb-capture.o
  add/remove: 0/0 grow/shrink: 1/6 up/down: 12/-68 (-56)
  Total: Before=5920, After=5864, chg -0.95%
drivers/counter/stm32-lptimer-cnt.o
  add/remove: 0/0 grow/shrink: 0/10 up/down: 0/-140 (-140)
  Total: Before=6458, After=6318, chg -2.17%
drivers/counter/stm32-timer-cnt.o
  add/remove: 0/0 grow/shrink: 0/11 up/down: 0/-132 (-132)
  Total: Before=5504, After=5372, chg -2.40%
drivers/counter/ti-eqep.o
  add/remove: 0/0 grow/shrink: 0/10 up/down: 0/-120 (-120)
  Total: Before=4598, After=4478, chg -2.61%

(the ti-eqep object file got a bit smaller, probably because I removed
the priv member from struct counter_device.)

Best regards
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | https://www.pengutronix.de/ |

Attachment: signature.asc
Description: PGP signature


[Index of Archives]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Input]     [Linux Kernel]     [Linux SCSI]     [X.org]

  Powered by Linux