On Tue, 15 Jan 2019 at 09:48, Arend Van Spriel <arend.vanspriel@xxxxxxxxxxxx> wrote: > On 1/15/2019 7:19 AM, Rafał Miłecki wrote: > > From: Rafał Miłecki <rafal@xxxxxxxxxx> > > > > So far __brcmf_err() was using pr_err() which didn't allow identifying > > device that was affected by an error. It's crucial for systems with more > > than 1 device supported by brcmfmac (a common case for home routers). > > > > This change allows passing struct brcmf_bus to the __brcmf_err(). That > > struct has been agreed to be the most common one. It allows accessing > > struct device easily & using dev_err() printing helper. > > Acked-by: Arend van Spriel <arend.vanspriel@xxxxxxxxxxxx> > > Signed-off-by: Rafał Miłecki <rafal@xxxxxxxxxx> > > --- > > This is my another try on improving brcmf_err after the failure from 2 > > years ago: > > [PATCH V3 4/9] brcmfmac: add struct brcmf_pub parameter to the __brcmf_err > > https://patchwork.kernel.org/patch/9553255/ > > > > Back then my change has been rejected due to miscommunication and late > > realisation that struct brcmf_pub (a previous choice instead of struct > > brcmf_bus) was a bad idea. Back then Arend wrote: > >> So I would think using struct brcmf_bus in brcmf_err() would be best > >> fit. > > > > So this patch follows that suggestion & updates __brcmf_err() > > accordingly. > > Thanks, Rafał > > Little less than two years ago I played with your idea and using GCC > builtin __builtin_types_compatible_p(t1,t2). Anyway, it looks good. So > you want to limit it to brcmf_err() or brcmf_dbg() as well? I believe all messages printed by brcmfmac should specify a device. brcmf_err, brcmf_info & brcmf_dbg. I can work on brcmf_info & brcmf_dbg once I get done with brcmf_err :) -- Rafał
