On 8/12/24 06:13, Krzysztof Kozlowski wrote:
On 12/08/2024 14:21, Wolfram Sang wrote:
Yep, but to be fair the patchset did not say anything about
dependencies. There is absolutely nothing in cover letter, nothing in
the patches, so I do not wonder that this mishap happened.
Still, one shouldn't take DT patches (which are even the last ones in
this series) until all other patches are at least in -next, or? Yes,
mistakes happen, so no big deal, but i2c is not to blame IMHO.
No, it's not. It was just a ping. The issue is here not describing
dependency, allowing Guenter to take the patch and not even telling him
Oh, I knew that the i2c patches were not yet in the tree. I just didn't
know that I must not apply patches in this situation (where the actual
patches are perfectly fine but assume that something else completely elsewhere
is applied). After all, the amc6821 patches do not actually trigger anything
in i2c mux, they just trigger instantiation of nested devices.
We live and learn. Patches now dropped from linux-next.
I do wonder though if the rules for applying a sequence of patches with
non-technical dependencies is documented somewhere.
Thanks,
Guenter
