Messages in this thread | | | Date | Thu, 12 Sep 2013 15:56:51 +0100 | From | Lee Jones <> | Subject | Re: [PATCH 21/33] clk: ux500: Add Device Tree support for the PRCC Kernel clock |
| |
On Thu, 12 Sep 2013, Linus Walleij wrote:
> On Tue, Aug 27, 2013 at 10:23 AM, Lee Jones <lee.jones@linaro.org> wrote: > > On Fri, 23 Aug 2013, Linus Walleij wrote: > >> On Thu, Aug 22, 2013 at 11:21 AM, Lee Jones <lee.jones@linaro.org> wrote: > > >> > 1. Duplicate each of the; clk_reg_prcmu_*(), clk_reg_prcc_pclk(), > >> > clk_reg_prcc_kclk() calls into your proposed u8500_clk_init_dt(), > >> > which, while keeping everything separate would be unrealistic. > >> > >> I think this is perfectly realistic. > >> > >> You're not going to duplicate each clk_register_clkdev(), > >> which makes it way smaller than the original function, > >> and since one of the function will be inside a > >> > >> #ifdef CONFIG_OF > >> #endif > >> > >> After we switch the entire platform to DT-only it will be pretty > >> obvious which big chunk of code that needs to go away, it's > >> a clean cut. > >> > >> (Note: I know the #ifdef CONFIG_OF is not necessary anymore > >> since we switched to multiplatform, but I intend that marker for > >> humans, not machines.) > > > > This sounds gross. To duplicate; u8500_clk_init(), u8540_clk_init() > > and u9540_clk_init() just for the sake of loading a few pointers into > > an array for a small part of the development cycle sounds obscene. > > > > I genuinely think keeping the current patch in this series and then > > removing the clk_register_clkdev() in the remove ATAG support series > > is the best way to go. > > So what I am worrying about is not only the looks of the code > and what is beautiful or not may be something of an opinion > anway. > > What I worry about is leaving all the calls to clk_register_clkdev() > in the DT boot path. Because that has the potential to hide a lot > of bugs, as clk_get() from drivers that should've got named and > probed randomly now will still find their clocks from their old > device names, instead of using the <&ersand> clocks from > the device tree. > > But if you still don't like this, let me cook a counter-patch so > I can realized on my own how terribly wrong I am...
I'm going to yank all of the clk_register_clkdev() calls out imminently anyway, so all those these hiding bugs will soon become apparent in any case. Why don't I place my 'remove ATAGs' patch-set on top of this one and then finally remove the clk_register_clkdev()s and watch the carnage ensue?
-- Lee Jones Linaro STMicroelectronics Landing Team Lead Linaro.org │ Open source software for ARM SoCs Follow Linaro: Facebook | Twitter | Blog -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |