Message ID | 56D6C623.8030908@gmail.com |
---|---|
State | New |
Headers | show |
On Wednesday 02 March 2016 11:53:23 Maxime Coquelin wrote: > Highlights: > ----------- > - Enable GPIO led driver in stm32_defconfig > - Introduce a config fragment to override RAM base address > Our first config fragment ;-) Merged into next/defconfig, thanks! Arnd
2016-03-02 23:21 GMT+01:00 Arnd Bergmann <arnd@arndb.de>: > On Wednesday 02 March 2016 11:53:23 Maxime Coquelin wrote: >> Highlights: >> ----------- >> - Enable GPIO led driver in stm32_defconfig >> - Introduce a config fragment to override RAM base address >> > > Our first config fragment ;-) Yes! Out of curiosity, did you see the question I asked you about fragments? See "Re: [PATCH v3 4/4] ARM: STM32: Enable Ethernet in stm32_defconfig". (https://lkml.org/lkml/2016/3/1/474) This is not urgent, it can wait for after the merge window, but I will be happy to have your view on this. Thanks! Maxime
On Thursday 03 March 2016 09:59:35 Maxime Coquelin wrote: > 2016-03-02 23:21 GMT+01:00 Arnd Bergmann <arnd@arndb.de>: > > On Wednesday 02 March 2016 11:53:23 Maxime Coquelin wrote: > >> Highlights: > >> ----------- > >> - Enable GPIO led driver in stm32_defconfig > >> - Introduce a config fragment to override RAM base address > >> > > > > Our first config fragment > > Yes! Out of curiosity, did you see the question I asked you about fragments? > See "Re: [PATCH v3 4/4] ARM: STM32: Enable Ethernet in stm32_defconfig". > (https://lkml.org/lkml/2016/3/1/474) > > This is not urgent, it can wait for after the merge window, but I will > be happy to have your view on this. > No, I missed that. Generally, I'd say we should limit the use of Kconfig fragments to more fundamental settings, it should not replace the manual configuration that users apply to pick the set of drivers they want. In the end there is not much difference between someone locally turning on CONFIG_INET when they need it, or using a fragment that does the same. Please just pick one default here (ethernet and ipv4 both enabled or both disabled) that makes sense for most users, and have the others change it manually. Arnd
2016-03-03 11:33 GMT+01:00 Arnd Bergmann <arnd@arndb.de>: > On Thursday 03 March 2016 09:59:35 Maxime Coquelin wrote: >> 2016-03-02 23:21 GMT+01:00 Arnd Bergmann <arnd@arndb.de>: >> > On Wednesday 02 March 2016 11:53:23 Maxime Coquelin wrote: >> >> Highlights: >> >> ----------- >> >> - Enable GPIO led driver in stm32_defconfig >> >> - Introduce a config fragment to override RAM base address >> >> >> > >> > Our first config fragment >> >> Yes! Out of curiosity, did you see the question I asked you about fragments? >> See "Re: [PATCH v3 4/4] ARM: STM32: Enable Ethernet in stm32_defconfig". >> (https://lkml.org/lkml/2016/3/1/474) >> >> This is not urgent, it can wait for after the merge window, but I will >> be happy to have your view on this. >> > > No, I missed that. Generally, I'd say we should limit the use of Kconfig > fragments to more fundamental settings, it should not replace the > manual configuration that users apply to pick the set of drivers > they want. In the end there is not much difference between someone > locally turning on CONFIG_INET when they need it, or using a fragment > that does the same. > > Please just pick one default here (ethernet and ipv4 both enabled or > both disabled) that makes sense for most users, and have the others > change it manually. Ok, I understand your point. In this case, I think having a fragment for Ethernet might not be a good idea because of the quantity of possible combinations. Thanks, Maxime