On Tue, 02 Jul 2019, Lothar Waßmann wrote: > Hi, > > On Tue, 2 Jul 2019 19:47:16 +0800 Fuqian Huang wrote: > > Andy Shevchenko <andy.shevchenko@xxxxxxxxx> 於 2019年7月2日週二 下午5:51寫道： > > > > > > On Tue, Jul 2, 2019 at 11:20 AM Fuqian Huang <huangfq.daxian@xxxxxxxxx> wrote: > > > > > > > > I am not an expert on this. I just write a coccinelle script to search > > > > this kind of misuse and fix it in a naive way. > > > > Could you tell me about how to use the proper bus accessors? Then I > > > > will fix it up and resend a v2 patch set. > > > > > > First, don't top post. > > > And answering to this, simple drop the patch. > > > Proper bus accessors is exactly what it's used in the current code. > > > > But why not use dev_get_drvdata directly. > > It simplifies getting the 'driver_data' from 'struct device' directly. > > And the platform_device here is not required. > > Replace it can remove the unnecessary step back and forth. (dev -> pdev -> dev). > > > Did you check whether the compiler generates different (better) code > with and without your patch? My guess is it won't. I can see Fuqian's point. If bus APIs are preferred, maybe it would be nicer if the function was adapted to accept a platform_device instead? Caveat: I haven't taken the time to look into the call-site details. This comment is based on just the patch alone. -- Lee Jones [李琼斯] Linaro Services Technical Lead Linaro.org │ Open source software for ARM SoCs Follow Linaro: Facebook | Twitter | Blog
