Ronald G Minnich <[EMAIL PROTECTED]> writes: > I've got the 1.12 mkelfImage. I build an initrd and kernel and load into > compact flash. With a brand-new linuxbios I get this: > > Found ELF candiate at offset 0 > header_offset is 0 > Try to load at offset 0x0 > Loading Linux version: 2.4.17-lanl.10customslave (root@localhost) #8 Tue > Apr 30 03:42:48 M2 > Dropping non PT_LOAD segment > New segment addr 0x80000 size 0x1c83ae offset 0x1000 filesize 0x1c40f6 > (cleaned up) New segment addr 0x80000 size 0x1c83ae offset 0x1000 filesize > 0x1c40f6 > Loading Segment: addr: 0x0000000000080000 memsz: 0x00000000001c83ae > filesz: 0x00000000001c6 > Clearing Segment: addr: 0x00000000002440f6 memsz: 0x00000000000042b8 > Jumping to boot code > > It gets the linux version right. > > The segment loading looks good, extending from 0x80000 to the end of my > ramdisk at 0x2483ae (0x8000+0x1c83ae above). > > So that seems ok. > > But jumping to boot code is the last think I see :-=( > > do I need to patch my kernel to use the new elf loader?
No but if you have ld newer that 1.10 you may have problems. Loading data a segment at 512K for 1824K bytes looks very suspcious. If it works the new mkelfImage should be more loader agnostic. In particular it should work just fine under etherboot as well. What does readelf -a say about your image? Eric
