Re: [RFC PATCH 2/2] leds: trigger: implement block trigger

[Enzo Matsumiya]

On 04/30, Marek Behun wrote:
> On Fri, 30 Apr 2021 15:32:11 -0300
> Enzo Matsumiya <ematsumiya@xxxxxxx> wrote:
>
> > Activity is then represented in an accumulated manner (part_read_stat_accum()),
> > with a fixed blinking interval of 50ms.
>
> part_stat_read_accum, not part_read_stat_accum

Good catch, will fix in v2. Thanks.

> Why only accum? With the netdev trigger, you can choose whether rx, tx,
> or both are blinking the LED.

The original patch from Akinobu Mita could distinct between READ,
WRITE, and DISCARD. My reasoning to not follow that was I've seen
NICs with a TX and RX LED (i.e. netdev follows that), but I've never
seen any disk activity indicator with separated LEDs for read and write,
for example. So accum made sense to me.

If this is really desired, I can come up with this, but I'd like to hear
from others.

> Also I think the trigger should be called "blockdev" instead of
> "block". This is consistent with "netdev", and avoids misinterpretation
> with the verb "to block".

Thanks. I'll change this in v2.

> Marek

Cheers,

Enzo