Thanks! Given that the driver is fully serialized, there really
shouldn't be any subtleties there.
I'll update the patch after some more testing.
On 04/24/2015 04:00 AM, Daniel Kurtz wrote:
Hi guys,
On Fri, Apr 24, 2015 at 6:08 PM, Jean Delvare <jdelvare@xxxxxxx> wrote:
Hi Ellen,
On Mon, 20 Apr 2015 15:52:47 -0700, Ellen Wang wrote:
This leads to a related question: If the driver is serialized, then the
(status = priv->status) inside wait_event_timeout() isn't strictly
necessary, correct? It can just be priv->status. I'm just want to
double check that there's no race and I can add a priv->auxsts and not
have to stick something like this inside the wait_event_timeout():
(auxsts = priv->auxsts, status = priv->status)
To be honest I'm not 100% certain. I think this was only a minor
optimization, but I may be wrong. Adding Daniel to Cc, he added that
code to the i2c-i801 driver. Daniel, any comment on this?
Equally mysterious to me is:
priv->status |= status;
in i801_isr() where it would seem that a simple "=" would suffice.
Sorry, it has been a long time since i looked at i801 code,
I think you guys are correct. You can probably just do:
wait_event(priv->waitq, priv->status);
status = priv->status;
priv->status = 0;
return i801_check_post(priv, status);
And, I agree with Jean in i801_isr(),
priv->status |= status;
Could probably just be:
priv->status = status;
-Dan
--
Jean Delvare
SUSE L3 Support
--
To unsubscribe from this list: send the line "unsubscribe linux-i2c" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
