I noticed one further thing about these hpmcs. There seems to be
a relationship between the exception address recorded in cr21 and
the address used in the faulting instruction. I have seen this in
at least four hpmcs, including the lba_pat_out8 one posted previously
to the list. Here is one from yesterday:
----------------- Processor 1 HPMC Information - PDC Version: 46.34 ------
Timestamp = Sun May 15 18:52:36 GMT 2011 (20:11:05:15:18:52:36)
HPMC Chassis Codes
Chassis Code Extension
------------ ---------
0xe800035c00e00000 0x00000000403140c0
0x57000f7300e00000 0x8040004000000000
0x5600100b00e00000 0x0000000000000194
0x5600106400e00000 0xfffffff0f0436fc0
General Registers 0 - 31
00-03 0000000000000000 0000000040622080 00000000403189b0 00000000405c64b0
04-07 0000000040614080 0000000000000000 00000000404e8100 000000007f1048d0
08-11 0000000000000001 0000000000000100 000000007f0b1020 0000000000200200
12-15 000000004062e880 0000000040622080 000000007f0b1420 000000007f0b1820
16-19 000000007f0b1c20 0000000000000001 0000000040622080 000000007f1047f8
20-23 ffe0000000000000 ffffffffffffffff 000000004058d110 8000000000000000
24-27 0000000100028109 0000000000000001 00000000405c64b0 0000000040614080
28-31 0000000004082001 000000007f104970 000000007f1049a0 0000000004082000
Control Registers 0 - 31
00-03 0000000000000000 0000000000000000 0000000000000000 0000000000000000
04-07 0000000000000000 0000000000000000 0000000000000000 0000000000000000
08-11 0000000000000da6 0000000000000000 00000000000000c0 000000000000003f
12-15 0000000000000000 0000000000000000 0000000000103000 ffe0000000000000
16-19 000000d398385ff0 0000000000000000 00000000403140c0 000000000f80001c
20-23 00000000a607ffd0 0000000014082001 000000ff0804ff0f 0000000000000000
24-27 0000000000573000 000000003c11e000 ffffffffffffffff 0000000040002f80
28-31 0000000040002140 ffffffffffffffff 000000007f104000 ffffffffffffffff
Space Registers 0 - 7
00-03 0000000000369800 0000000000000000 0000000000000000 0000000000369800
04-07 0000000000000000 0000000000000000 0000000000000000 0000000000000000
IIA Space (back entry) = 0x0000000000000000
IIA Offset (back entry) = 0x00000000403140c4
Processor 1 was executing ldb in mem_serial_in:
00000000403140a8 <mem_serial_in>:
403140a8: 43 5c 00 a2 ldb 51(r26),ret0
403140ac: 53 5f 00 20 ldd 10(r26),r31
403140b0: 01 7c 18 c0 mtsarcm ret0
403140b4: d7 99 00 00 depw,z r25,sar,32,ret0
403140b8: db 9c 0f e0 extrd,s ret0,63,32,ret0
403140bc: 0b 9f 0a 1c add,l r31,ret0,ret0
=> 403140c0: 0f 80 00 1c ldb r0(ret0),ret0
403140c4: e8 40 d0 00 bve (rp)
403140c8: db 9c 0b f8 extrd,u ret0,63,8,ret0
403140cc: 00 00 00 00 break 0,0
Note the address in $ret0 and the value in $cr21. I'm thinking
some form of cache aliasing may be the cause. The fault appears
related to an io space access.
Dave
--
J. David Anglin dave.anglin@xxxxxxxxxxxxxx
National Research Council of Canada (613) 990-0752 (FAX: 952-6602)
--
To unsubscribe from this list: send the line "unsubscribe linux-parisc" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
