Prev: x86 platform driver updates for 2.6.35
Next: [PATCH] firewire: core: check for 1394a compliant IRM, fix inaccessibility of Sony camcorder
From: Linus Torvalds on 30 May 2010 12:40 On Fri, 28 May 2010, Samuel Ortiz wrote: > > This is the MFD pull request for the 2.6.35 merge window. Samuel - this adds the CONFIG_MFD thing, with absolutely _no_ explanation. So now people see this totally incomprehensible question: Enable Multifunction device drivers? (Y/n) with no sane way to decide whether they should actually enable it or not. I'm ok with it being 'y' default, since it's a whole new config option that enables pre-existing config options, but _dammit_ people, you need to think way more about the Kconfig options you add! "MFD" isn't even a well-defined term. A lot of people think about printers when they hear it. So you really _really_ need to explain what the h*ll you are asking people! Linus -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
From: George Spelvin on 30 May 2010 18:40
Linus wrote: > Samuel - this adds the CONFIG_MFD thing, with absolutely _no_ explanation. > So now people see this totally incomprehensible question: > > Enable Multifunction device drivers? (Y/n) > > with no sane way to decide whether they should actually enable it or not. Seconded. This actually happens a lot; I don't mean to pick on the MFD drivers especially. Sometimes the best way to answer a new config question is to RTFS, which is a bit tedious. For example, the "ACPI 4.0 Power Meter" (CONFIG_ACPI_POWER_METER). To me, a power meter is a large glass-encased object containing a spinning aluminum disc and mounted on the wall of my house which determines my monthly power bill. How this relates to ACPI 4.0 firmware is unclear. Indeed, how do I figure out of I have ACPI 4.0 firmware? Then there's CONFIG_SFI. Nowhere in the Kconfig text does it explain that it's intended as a lightweight alternative to ACPI for hand-held devices and not for desktop machines. Then there are the PCI_GO* options... I still have no idea. (I just pick "any", figuring it's the most likely to work.) I think the CONFIG_PCI help text could use some updating. The CONFIG_PCIEPORTBUS is unnecessarily confusing. It doesn't clearly explain, if I have PCI express slots on my computer: - Do I actually need this to boot? - Does it do anything besides enable a config submenu? CONFIG_MCA (admittedly a critically endangered interest group, but still listed as "Maintained" in MAINTAINERS) stresses the importance of the web site address given in in Documentation/mca.txt, which is http://www.dgmicro.com/mca/. That domain no longer exists (expired March '07) and has been grabbed by a squatter. It can be found on the wayback machine at http://web.archive.org/web/20070225223240/http://www.dgmicro.com/mca/ and gives a last update date of 1999. It would be easy to go on... -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/ |