Quiet Sun Blade 2000; Fujitsu PRIMEPOWER 250; Books
Recently, on a whim, I thought that it would be nice to see the fan speed on my Sun Blade 2000. Looking through the tda driver source and searching didn't show up any way of finding this out, but I could display the values that the driver had set. As noted in the mailing list posting, I noticed that the system fan was always running at full speed. With that bug fixed, my SB2000 is now much quieter and I'll use it as my desktop. It has 8GB memory, two 300GB disks, plus a Creator 3D and an XVR-100 to use with my Samsung SyncMaster 245B monitors. These monitors run at 1920x1200, and also handle a wide range of inputs, so are great for machines with older graphics cards. While here, I added a basic manual page, as there was nothing previously. Note, that I haven't yet tried to make the fans on the E280R run at variable speed, although it looks like this should be possible (the E280R uses the same mainboard as the Blade 1000/2000, but it has different firmware and three fans instead of two).
Whilst looking at the devices on the I²C bus, I noticed that we already had a seeprom driver for the AT24C64 EEPROM's that show up like:
although the DIMM's (where I expected a similar message starting with dimm-fru) were showing up as:
instead. It turns out that the Philips PCF8584 can have more than one I²C bus attached. We already supported that in our PCF8584 driver, so with a simple fix to the Open Firmware code, direct configuration support for the seeprom driver, and passing the correct size to the seeprom driver, it's now possible to access the EEPROM's on devices that have the compatible property set to i2c-at24c64. On the E280R, these now show up like:
and I can read their contents:
# dd if=seeprom0 bs=1 skip=6144 count=2048 | hexdump -C 00000320 00 00 00 00 0c 00 00 00 00 f8 00 10 b7 00 00 00 |................| 00000330 31 30 32 34 20 4d 42 20 4e 47 20 53 44 52 41 4d |1024 MB NG SDRAM| 00000340 20 44 49 4d 4d 00 00 00 00 00 00 00 00 00 00 00 | DIMM...........| 00000350 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................| * 00000380 00 44 41 54 41 52 41 4d 2c 55 53 41 20 20 00 00 |.DATARAM,USA ..| 00000390 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................| * 000003c0 00 35 30 31 36 32 36 36 31 2e 20 20 20 20 20 20 |.50162661. | 000003d0 30 36 30 31 44 41 54 41 52 41 4d 20 44 49 4d 4d |0601DATARAM DIMM| 000003e0 00 00 00 00 12 0b 01 f0 01 04 00 4b 00 80 0c 00 |...........K....| 000003f0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
# dd if=seeprom16 bs=1 skip=6144 count=2048 | hexdump -C 00000710 00 00 00 00 c1 78 01 00 07 02 00 03 04 04 17 f2 |.....x..........| 00000720 01 f4 14 8c 01 4a 02 26 04 b0 00 32 14 b0 0c b7 |.....J.&...2....| 00000730 31 f2 9b f8 00 10 b7 3e 38 42 b2 50 57 52 20 53 |1......>8B.PWR S| 00000740 55 50 50 4c 59 2c 20 4d 55 4c 54 49 20 4f 55 54 |UPPLY, MULTI OUT| 00000750 50 55 54 2c 20 54 59 50 45 20 41 31 35 31 00 00 |PUT, TYPE A151..| 00000760 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................| * 00000780 00 00 00 00 00 00 00 00 00 00 00 4d 61 74 61 6d |...........Matam| 00000790 6f 72 6f 73 2c 20 54 61 6d 70 73 2c 20 4d 65 78 |oros, Tamps, Mex| 000007a0 69 63 6f 00 00 00 00 00 00 00 00 00 00 00 00 00 |ico.............| 000007b0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................| 000007c0 00 00 00 00 00 00 00 00 00 00 00 33 30 30 31 34 |...........30014| 000007d0 35 37 4c 38 33 34 35 37 03 1a 30 33 35 31 50 53 |57L83457..0351PS| 000007e0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 c0 2b |...............+| 000007f0 38 30 36 2d 37 38 39 31 2d 30 32 0c f0 ee d8 1c |806-7891-02.....|
(non-text lines omitted for brevity). Now, I can check part numbers without opening the lid ;-) On the SB2000, where the I²C controller has the second bus, I see:
One remaining bug is that the I²C driver busy-waits for the read or write to complete. This means that the machine becomes unresponsive while reading the seeprom. This requires a re-write of the driver (probably to use a kernel thread) to fix.
Recent purchase: a Fujitsu
PRIMEPOWER 250. This is a Fujitsu SPARC64 V machine, and this one
came with two 1.98 GHz processors, 8GB RAM, an extra quad-port gigabit
ethernet card and single-port gigabit ethernet card, and two 136GB disks,
via
1st Technologies Limited on Ebay. It was very well packaged and they
delivered it on the day that I asked.
It's currently sitting on the floor, but I hope to add it to my 19" rack above the Sun T2000 and help with the NetBSD port to the SPARC64 V machines which Matthew Green has started. Thanks to Mike Owens, I've been able to reset the XSCF password, so that I can control it remotely. It also provided a bit of initial puzzlement, as the XSCF RJ45 serial port is wired in reverse when compared to Sun and Cisco machines, but an old Cisco roll over cable came to the rescue.
Recent book purchases:
Having grown up at Prestatyn, a recent purchase of Larry Goddard's photographs brought back memories of travelling on the North Wales main line. A very nice set of photographs, and well reproduced too. The book starts at the Holyhead end of the line, and covers a number of locations before finishing at Chester on page 88. The dates range from the late 1970's to the end of the century, and mainly cover loco-hauled freight and passenger trains, although there are a few DMU photographs too.
A Railwayman's Odyssey is an acount of the first 40 years of railway life of the author of a number of excellent books on railway safety. It starts before the author's railway career with his interest in railways at a young age, and continues through his various railway operating appointments from junior clerk to assistant movements superindentent. The book covers a wide range of locations, and the period from 1930 to 1970. An extensive selection of photographs compliments the text, and is representative of the dates and locations. A good read, and I'm looking forward to the sequel.
-^- More notes -^-