On 2016/4/8 18:46, Marc Zyngier wrote: > On 08/04/16 11:33, Kefeng Wang wrote: >> >> >> On 2016/4/8 16:53, Marc Zyngier wrote: >>> On Fri, 8 Apr 2016 16:26:21 +0800 >>> Kefeng Wang <[email protected]> wrote: >> [...] >> But there is still one issue in the for_each_child_of_node loop of >> mbigen_device_probe. Assume that there are 3+ child node(mbigen_gmac, >> mbigen_i2c, mbigen_xxx, etc) in mbigen_chip_dsa, if one of them with >> incorrect configuration, then the loop will end, but we still need >> parse the left child node. we should consider this situation, right? > > That's up to whoever designed the driver to decide, really. I don't know > if it is worth continuing in that case (and you are in a better position > than me to find out).
Ok, will talk with Ma Jun directly, and if the change is needed, I will send a new patch after reviewed by Ma Jun. Thanks, Kefeng
