On Sun, Sep 14, 2014 at 10:27:43AM +0800, Jianqun wrote:
> ? 09/14/2014 12:37 AM, Mark Brown ??:
> >> + ret = clk_prepare_enable(i2s->hclk);
> >> + if (ret) {
> >> + dev_err(i2s->dev, "hclock enable failed %d\n", ret);
> >> + return ret;
> >> + }
> > BTW: you're also missing a clk_disable_unprepare() in the remove path
> > here, please send a followup fixing that.
> remove function has done the clk_disable_unprepare for "hclk".
> One more thing, since "i2s_clk" is enabled at runtime_resume, and is disabled in runtime_suspend,
> hasn't enable in probe, so do the "i2s_clk" to be disabled in remove ?
> The current driver has disable in remove.
Again, please try to write clear changelogs which describe what the goal
of the patch is and cover obvious questions like this which a reviewer
might ask.
This is all extremely unclear - you're adding an enable here with no
matching disable. It seems that what you saying that there was
previously a bug where the clock was disabled without being enabled?
I had to look at the code to work that out...
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 473 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-rockchip/attachments/20140915/4f25d2fa/attachment.sig>
