Re: ia64 qualification for Wheezy
Aníbal Monsalve Salazar wrote:
> I'm happy to test a possible fix for #638068 on my ia64 machines.
> Does someone have a hint about fixing it?
Émeric seems to have had luck working around the problem by using
busybox instead of klibc.
Thorsten mentioned that some low-level voodoo will probably be
needed before klibc is properly fixed:
| I’ve just put some 「write(2, "here1\n", 6)」 style debugging
| into klibc’s libc_main.c and found out that shared binaries
| seem to crash before reaching __libc_main there. And I cannot
| debug an ELF interpreter with gdb, or at least I don’t know
| how. Would an IA64 expert please look at this? (Note the way
| how klibc does shared binaries is… special.)
For the short term, perhaps someone interested could take
inspiration from Émeric's instructions and write a patch for
initramfs-tools to disable use of klibc on ia64?
Hope that helps,