-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Somebody in the thread at some point said:
|> These guys need to uplevel to more recent kernels anyway, over time the |> pressure for them to do so will only increase. After some time we can |> go back to Qi doing the right thing and the whole boot sequence will be |> fine for this issue without workarounds in the rootfs. | Well... speaking for SHR... I strongly prefer Qi doing the right thing | and fix our kernel/rootfs instead (which we will do with the next kernel | build). It's good to hear, obviously since I found and fixed this problem I share the desire to not have to rip the fix out again. But --> | Maybe a strong notice about this on openmoko-devel ML or on planet would | be enough? At least people cannot complain about silently breaking | things then... I think practically we have to be compatible with old kernels for a while. The cost of the compatibility is quite limited in time since the fault current goes away when GSM is turned on in initscripts; whereas the cost of incompatibility is a sticky breakage of GSM function. U-Boot has always "done the wrong thing" with regards to this as well, so every U-Boot user is experiencing the workaround proposed for Qi already. I have an idea though, what do you guys do about the U-Boot mkudfu content? Is there anything in there we can use to make a decision about in Qi's GTA02 code as to how to leave the UARTs? Eg, kernel revision? - -Andy -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iEYEARECAAYFAkmr06oACgkQOjLpvpq7dMooWACfXY24oYARDauLy1gtpxBLZeEU 9z8Aniv5ZYzLEPbzy/YDm1cDLsrYtd6a =j6h9 -----END PGP SIGNATURE-----
