OOPS on AMD CFI chips

[mark.langsdorf at amd.com]

	I generated an oops by copying a file.  Here's the oops
itself, followed by the assembly code immediately proceeding
the crash (from cfi_amdstd_write_bytes_32 in cfi_cmdset_0002.c).
	It looks to me like the crash is caused by edi being
0xC1000000.  Am I misinterpreting or should I follow up on this
more?

CPU:    0
EIP:    0010:[<c01a27fb>]
EFLAGS: 00010217
eax: 3fffffff   ebx: fffffffe   ecx: 3ffa2774   edx: 00000003
esi: c03445ec   edi: c1000000   ebp: 005cbd14   esp: c0e89dc8
ds: 0018   es: 0018   ss: 0018
Process cp (pid: 231, process nr: 35, stackpage=c0e89000)
Stack: 00000001 c003dad0 c0e89dd4 00ffffff 00000000 00000000 00000000
00000000
       00000000 00000000 00000000 00000000 c011b680 c014d760 00000000
00000000
       c014dbd0 c014ddd0 00000000 00000000 00000000 c014de00 00000000
00000000
Call Trace: [<c011b680>] [<c014d760>] [<c014dbd0>] [<c014ddd0>] [<c014de00>]
[<c014c1f0>] [<c014c490>]
       [<c014bff0>] [<c014d4f0>] [<c014c1f0>] [<c014c960>] [<c014cf90>]
[<c014c650>] [<c014c8f0>] [<c014ccd0>]
       [<c014bca0>] [<c014d2a0>] [<c014d3d0>] [<c014de10>] [<c014dff0>]
[<c014b790>] [<c014b710>] [<c014e050>]
       [<c014bbe0>] [<c01b4cea>] [<c014b590>] [<c014e090>]
Code: f3 a5 f6 c3 02 74 02 66 a5 f6 c3 01 74 01 a4 8b 74 24 2c e9
Unable to handle kernel paging request at virtual address 83d0ff28
current->tss.cr3 = 008e7000, %cr3 = 008e7000
*pde = 00000000


7ee:	movl		0x10(%esp, 1), %edi
7f2:  movl		0x28(%esp, 1), %esi
7f6:  movl		0x2c(%esp, 1), %ebx
7fa:  cld
7fb:  repz movsl	%ds:(%esi), %es(%edi)

Mark Langsdorf
Advanced Micro Devices, Inc             Tel: 512.602.3756
5204 E. Ben White Blvd. M/S 590         Fax: 512.602.5051
Austin, TX 78741                        mark.langsdorf at amd.com


To unsubscribe, send "unsubscribe mtd" to majordomo at infradead.org