There are mainboards with the MPV3 chipset which run perfectly
with both options enabled, the "CPU to PCI write buffer" and the
"Peer Concurrency". So do (in four cases) the DFI-PB5V3+ mainboards,
and it seems to be a problem related to the implementation of the chipset
on the mainboards, and NOT a generic chipset-problem of the MVP3.
Disabling these options generally would mean a big performance loss
for those who have such mainboards it works (on the new DFI mainboard
it works also, I heard a few days ago).
I agree. I have a TMC TI5VG+ / Via MVP3 chipset board with both the
options set and it's been quite stable.
lspci -vvx of 00.0 and 00.1:
00:00.0 Host bridge: VIA Technologies, Inc. VT82C597 [Apollo VP3]
(rev 04)
Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop-
ParErr- Stepping- SERR- FastB2B-
Status: Cap+ 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium
>TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 16 set
Region 0: Memory at e0000000 (32-bit, prefetchable)
Capabilities: [a0] AGP version 1.0
Status: RQ=7 SBA+ 64bit- FW- Rate=1
Command: RQ=7 SBA+ AGP- 64bit- FW- Rate=1
00: 06 11 97 05 06 00 90 02 04 00 00 06 00 10 00 00
10: 08 00 00 e0 00 00 00 00 00 00 00 00 00 00 00 00
20: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
30: 00 00 00 00 a0 00 00 00 00 00 00 00 00 00 00 00
00:01.0 PCI bridge: VIA Technologies, Inc. VT82C598 [Apollo MVP3 AGP]
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop-
ParErr- Stepping- SERR- FastB2B-
Status: Cap- 66Mhz+ UDF- FastB2B- ParErr- DEVSEL=medium
>TAbort- <TAbort- <MAbort+ >SERR- <PERR-
Latency: 0 set
Bus: primary=00, secondary=01, subordinate=01, sec-latency=0
I/O behind bridge: 0000d000-0000dfff
Memory behind bridge: e4000000-e7ffffff
Prefetchable memory behind bridge: e8000000-e8ffffff
BridgeCtl: Parity- SERR- NoISA+ VGA+ MAbort- >Reset- FastB2B-
00: 06 11 98 85 07 00 20 22 00 00 04 06 00 00 01 00
10: 00 00 00 00 00 00 00 00 00 01 01 00 d0 d0 00 00
20: 00 e4 f0 e7 00 e8 f0 e8 00 00 00 00 00 00 00 00
30: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0c 00
regards,
-- Paul Jakma paul@clubi.ie http://hibernia.clubi.ie PGP5 key: http://www.clubi.ie/jakma/publickey.txt ------------------------------------------- Fortune: Of course there's no reason for it, it's just our policy.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/