On Mon, 2012-10-22 at 17:08 +0200, Thierry Reding wrote: > On Mon, Oct 22, 2012 at 01:52:08PM +0000, Arnd Bergmann wrote: > > On Monday 22 October 2012, Thierry Reding wrote: > > > > As long as we get build warnings for leaving out the __devinit/__devexit > > > > annotations, I would generally recommend putting them in. If we do a > > > > patch to remove all of them, a couple extra instances will not cause > > > > any more troubles than we already have. > > > > > > I've never seen any build warnings for leaving __devinit/__devexit out. > > > Where does that happen? > > > > Section mismatches usually result into warnings from modpost, like > > > > WARNING: modpost: Found 1 section mismatch(es). > > To see full details build your kernel with: > > 'make CONFIG_DEBUG_SECTION_MISMATCH=y' > > > > Actually doing that gives you an output like this (currently on > > exynos_defconfig): > > > > $ make CONFIG_DEBUG_SECTION_MISMATCH=y > > WARNING: drivers/pinctrl/built-in.o(.devinit.text+0x124): Section mismatch > > in reference from the function samsung_pinctrl_probe() to the function > > .init.text:samsung_gpiolib_register() > > The function __devinit samsung_pinctrl_probe() references > > a function __init samsung_gpiolib_register(). > > If samsung_gpiolib_register is only used by samsung_pinctrl_probe then > > annotate samsung_gpiolib_register with a matching annotation. > > > > or like this (now fixed in socfpga_defconfig): > > > > WARNING: drivers/net/ethernet/stmicro/stmmac/stmmac.o(.text+0x5d4c): > > Section mismatch in reference from the function stmmac_pltfr_probe() to the > > function .devinit.text:stmmac_probe_config_dt() > > The function stmmac_pltfr_probe() references > > the function __devinit stmmac_probe_config_dt(). > > This is often because stmmac_pltfr_probe lacks a __devinit > > annotation or the annotation of stmmac_probe_config_dt is wrong. > > > > I believe you normally don't get warnings for functions that could be > > marked __devinit and only call regular functions, but there are > > a couple of __devinit infrastructure functions that you can't call > > from a function that isn't __init or __devinit. > > Right. If you get those warnings you shouldn't be dropping the > annotations. But I don't think that is the case for this driver. Tony, > can you confirm that the driver still builds properly without warnings > if you drop the __devinit/__devexit? > > Thierry > _______________________________________________ > linux-arm-kernel mailing list > linux-arm-ker...@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Correct - it builds without warnings without __devinit/__devexit. If it had introduced warnings when I tested it, I would have mentioned it :) Regards Tony P -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
