On 09/01/2023 16:11, Andy Shevchenko wrote:
On Mon, Jan 09, 2023 at 04:01:43PM +0200, Tomi Valkeinen wrote:
On 09/01/2023 15:58, Andy Shevchenko wrote:
On Mon, Jan 09, 2023 at 03:55:39PM +0200, Andy Shevchenko wrote:
On Mon, Jan 09, 2023 at 02:59:01PM +0200, Tomi Valkeinen wrote:
On 09/01/2023 13:07, Andy Shevchenko wrote:
On Sun, Jan 08, 2023 at 06:06:34AM +0200, Laurent Pinchart wrote:
On Thu, Jan 05, 2023 at 04:03:06PM +0200, Tomi Valkeinen wrote:
...
+ gc->of_node = priv->client->dev.of_node;
We don't have of_node anymore in gc. And if the parent device is set, you can
drop this line (it will work with older and newer kernels. Otherwise, use
fwnode.
What do you mean "we don't have of_node anymore"?
Just to add that you missed "gc" in the quote which is crucial to get
what I meant.
Well, the original text was just above...
Yes, but my point is that that part is crucial and since you left it out makes
me think that you are talking about general of_node, which is of course won't
gone any time soon :-)
There is no of_node member of struct gpio_chip anymore. This will fail
to compile.
It's in Linus' tree, but yes, now that I fetched linux-next I see it's gone.
Right. I expressed to Bart to include that change into one of the rcX, but he
decided to go for next. Dunno how many cases like yours we may experience in
this cycle.
Alright. Well, good that you caught this already. In the worst case we
would have had a compilation break in the next merge window.
Tomi
