On 24.06.2009 18:29, Myles Watson wrote:
at the default 8KB). This might solve booting, but I'll still have APs
and
BSP talking all at once, which I'm also seeing on K10.
I tried with a 32KB stack and a 64KB stack, no change, the machine still resets itself:
http://ward.vandewege.net/coreboot/h8dme/minicom-20090624e.cap
Maybe this is a silly suggestion, but how hard would it be to write a script that unscrambled the output?
I did it by hand. Not pleasant, but see below. [...] Copying data from cache to RAM -- switching to use RAM as stack... Done testx = 5a5a5a5a Disabling cache as ram now Clearing initial memory region: Done Jumping to image. Jumping to image. Check CBFS header at fffe0fd0 Check CBFS header at fffe0fd0 magic is 4f524243 Found CBFS header at fffe0fd0 magic is 4f524243 Found CBFS header at fffe0fd0 Check normal/payload CBFS: follow chain: Check normal/payload ffCfB0F0S0:0 0f o+l l2o8w +c h1a0i0n7:6 f+f fa0l0i0g0n0 -+> 2f8f f+1 0100a0076 +Ch eaclki gnno r-m>a lf/fcfo1r0e0bao0ot _Crhaemck
It's obvious that the CBFS code is called on each core. Note that even if CBFS can handle this, concurrent lzma decompression to the same area is guaranteed to kill the machine: lzma uses the destination memory range as scratch space during decompression.
Regards, Carl-Daniel