On Fri, Sep 02, 2005 at 10:33:21AM +0100, Nick Clifton wrote: > I suspect however that this might actually be a compiler problem, or > even a user problem. What I think might be happening is that the > compiler is creating code that assumes that __divqu function will be > within the 16-bit offset available to the R_ALPHA_LITERAL and telling > the assembler to go ahead and generate the reloc. Presumably this used > to be OK when the libafs library was smaller, but now that it has grown > this assumption is no longer true.
This turns out to be correct. The assembly output from the compiler confirms that the compiler explicitly specifies the use of the literal relocation type. As it turns out though, the whole issue is a red herring. I checked the size of the .got section of the loaded module, and it was 75K, beyond the range that the literal relocation type can specify an offset (64K). Thus, all R_ALPHA_LITERAL relocated symbols fail, it just happens that __divqu is the first symbol of that type. Why is .got so large? There's a kernel bug in Linux's alpha module loader that inflates the got. After applying the patch, the .got is only 8K and everything is good: http://www.ussg.iu.edu/hypermail/linux/kernel/0509.0/1342.html http://www.ussg.iu.edu/hypermail/linux/kernel/0509.0/1362.html _______________________________________________ bug-binutils mailing list bug-binutils@gnu.org http://lists.gnu.org/mailman/listinfo/bug-binutils
