| Message ID | 1431670969-1996-1-git-send-email-jcliang@chromium.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On 05/15, Ricky Liang wrote: > diff --git a/drivers/clk/mediatek/clk-pll.c b/drivers/clk/mediatek/clk-pll.c > index 66154ca..72fe307 100644 > --- a/drivers/clk/mediatek/clk-pll.c > +++ b/drivers/clk/mediatek/clk-pll.c > @@ -289,6 +289,7 @@ static struct clk *mtk_clk_register_pll(const struct mtk_pll_data *data, > init.ops = &mtk_pll_ops; > init.parent_names = &parent_name; > init.num_parents = 1; > + init.flags = data->flags; Wrong? It looks like plls[] in drivers/clk/mediatek/clk-mt8135.c is putting mtk specific clock flags in this structure member. How about initializing the structure to { } instead? That way if we ever add more structure members we don't have to detect the landmine laid here.
diff --git a/drivers/clk/mediatek/clk-pll.c b/drivers/clk/mediatek/clk-pll.c index 66154ca..72fe307 100644 --- a/drivers/clk/mediatek/clk-pll.c +++ b/drivers/clk/mediatek/clk-pll.c @@ -289,6 +289,7 @@ static struct clk *mtk_clk_register_pll(const struct mtk_pll_data *data, init.ops = &mtk_pll_ops; init.parent_names = &parent_name; init.num_parents = 1; + init.flags = data->flags; clk = clk_register(NULL, &pll->hw);
The variable init (struct clk_init_data) is allocated on the stack. We weren't initializing the .flags field, so it contains random junk, which can cause all kinds of interesting issues when the flags are parsed by clk_register. The best solution seems to just pass on the flags passed in to our clk_register wrappers. Signed-off-by: Ricky Liang <jcliang@chromium.org> --- drivers/clk/mediatek/clk-pll.c | 1 + 1 file changed, 1 insertion(+)