From: Peter M. <pm...@fr...> - 2016-08-25 12:34:16
|
On Thu, Aug 25 2016, Christophe Rhodes wrote: > then disassemble the code around that address Thanks for your help with gdb. Here the result: --8<---------------cut here---------------start------------->8--- 0x31b91708: bnez a3,0x31b928f4 0x31b9170c: nop 0x31b91710: lw t0,48(s6) 0x31b91714: nop => 0x31b91718: lwc1 $f0,-3(t0) 0x31b9171c: nop 0x31b91720: lw t0,49(s8) --8<---------------cut here---------------end--------------->8--- > On the other hand, I have no idea about the variety of MIPS instruction > sets. I would have expected that instructions would be added rather > than removed as time went on (given that the MIPS build of sbcl worked a > decade ago or so), I expect that too. But MIPS III has been already introduced in 1992 (from https://en.wikipedia.org/wiki/MIPS_instruction_set). > but that is probably a naïve expectation. I don't know about your > platform; it may be that there are just one or two "missing" > instructions, or it may be that things have changed fundamentally. But > the first step is to find out what's causing the first fault, I think. I guess, that the lwc1 and perhaps some other instructions have to be replaced for my target? -- Peter |