Heitzso wrote:
now, with framebuffer and console turned off, ... boot went much, much further, but not to completion
2.4 for me seems to be a game of Russian Roulette. Best grab your self a bag of chips and get used to flipping the power on the SS20. Eventually it will boot. The question is "when". My ss20 has a dual Ross 125MHz module (512kb cache I think), running 2.4.22-rc3. I had to add in the asm clobber patch, and the context table alignment patch Keith mentioned. Built with GCC-3.3.1, and most config options as modules and such, it took about 21 reboots before I got the kernel to load completely and boot into userland.
That was a month ago. I've refused to cut the thing off since until 2.4.23 (and that's a maybe) just to see how long it'll last. The boot process threw Level 15 interrupts, watchdog resets, and init hangs at me. Out of 21 reboots, 3 loaded the kernel to Init. Out of those 3, only one (the last) booted to userland. So all in all, it is really a guessing game.
The machine runes headless too. Serial console works flawlessly, sshd runs fine. Aside from getting it to boot, it is rather a functional little machine, if a tad on the slow side.
--Kumba --"Such is oft the course of deeds that move the wheels of the world: small hands do them because they must, while the eyes of the great are elsewhere." --Elrond