masafumi and Ken, thank you for your fast, very helpful replies.
Though I have double checked the OpenSolaris HCL, I still could not find any relevant references, neither to the 890FX chipset nor to the 1090T CPU. Ken, the most relevant/recent AMD hardware reference I found in your HCL sumbissions was an ASUS M2N-E nForce 570 Chipset based board with an Athlon 64 X2 CPU which is IMHO way too distinct for extrapolating 1090T/890FX compatibility on. Is that what you have been referring to in your reply? Nevertheless, encouraged by the positive feedback from both of you, I have decided to test it out and contribute the findings to the HCL ASAP. (I bought the motherboard and CPU already 2 weeks ago, but due to the difficulty of finding a reasonable distributor of 4GB unregistered ECC DDR3 modules in central Europe I was able to test drive the system for the first time just 2 days ago.) IMPRESSIONS and OBSERVATIONS: Stress testing the system with 2 Kingston 4GB unregistered ECC DDR3 modules (KVR1333D3E9SK2/8G) with MemTest86+ for many hours was flawless. After 2 more days of extensive OpenSolaris usage I have to say, I am very pleased with the results. The system is rock stable and very fluid, the strengths of the Solaris kernel, the Phenom architecture and 8GB ECC RAM combine synergistically to form a powerful, unique and quite impressive experience (both as server and workstation). With cca. 25% CPU load and a CoolIT Vantage liquid cooler operating in Quiet mode the CPU runs at respectable 35,5 degrees Celsius already and is expected to drop further 2-5 degrees after the thermal compound (Arctic Silver 5) breaks in completely (expected after cca. 200 hours and several thermal cycles). I have submitted a detailed report to the OpenSolaris HCL few hours ago. It should be available soon (after being reviewed I suppose). REMAINING DRIVER ISSUES: There are just 3 driver issues with this board I am trying to resolve now: 1) Missing driver: Marvell 8059 Gigabit LAN controller $ pfexec scanpci | grep 0x4381 pci bus 0x0003 cardnum 0x00 function 0x00: vendor 0x11ab device 0x4381 I have tried the latest Marvel Yukon 2 Gigabit drivers from IPS and Marvell, both don't seem to support 8059: $ pfexec update_drv -a -i '"pci11ab,4381"' yge devfsadm: driver failed to attach: yge Warning: Driver (yge) successfully added to system but failed to attach $ pfexec update_drv -a -i '"pci11ab,4381"' yukonx devfsadm: driver failed to attach: yukonx Warning: Driver (yukonx) successfully added to system but failed to attach 2) Missing driver: NEC USB 3.0 controller $ pfexec scanpci | grep 0x0194 pci bus 0x0002 cardnum 0x00 function 0x00: vendor 0x1033 device 0x0194 3) audiohd driver problems: SupremeFX X-FI 3.1) The mixer does not provide a control to turn on spread (i.e. mirror the signal for the 2 front speakers to the rear speakers). 3.2) When attaching a 4.1 speaker system to the standard lime green and black connectors, the lime connector works as expected, but the black connector does not output any signal to the rear speakers at all. Tested with mplayer multichannel audio and audiotest. This does not seem to be a hardware problem. I had a similar issue with the audiols driver on MSI K8N Diamond MB (SB0438) before, the only way to make the rear speakers play was to turn on spread. I was hoping the overall experience with the audiohd driver would be an improvement, but right now, due to missing spread it turns on to be quite the opposite. Can someone with more experience and deeper insight provide suggestions on how to approach these issues? Thanks. -dusan -- This message posted from opensolaris.org _______________________________________________ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
