I'm looking to buy a new desktop machine today or tomorrow. There were several posts last November bout the MSI K9N Neo-F:
http://www.openbios.org/pipermail/linuxbios/2006-November/thread.html
Where are we with that board? It's cheap and on the face of it should be fairly easy to port - MCP55 chipset, etc. There was some concern about the ehternet card?
I think I'll get this board or perhaps one of the Abit KN9 ones that are very similar to the Gigabyte M57SLI-S4. All have a PLCC socket which is a big plus.
If anyone has been working on these boards give me a shout - I'd like to put some work into adding support for one of these.
Thanks, Ward.
On Sun, Aug 12, 2007 at 06:25:26PM -0400, Ward Vandewege wrote:
Where are we with that board?
I don't think anyone got around to it.
It's cheap and on the face of it should be fairly easy to port - MCP55 chipset, etc. There was some concern about the ehternet card?
For LB purposes the NIC is not likely to cause problems, I think the uncertainty was about a Linux driver for it. drivers/net/forcedeth.c does support the relevant PCI IDs (PCI_DEVICE_ID_NVIDIA_NVENET_14/15, 0x0372/0x0373) according to the FreeBSD nfe driver.
I think I'll get this board or perhaps one of the Abit KN9 ones that are very similar to the Gigabyte M57SLI-S4. All have a PLCC socket which is a big plus.
If anyone has been working on these boards give me a shout - I'd like to put some work into adding support for one of these.
I don't know if Carl-Daniel got a response from the desktop BIOS people at MSI, I got the impression we'd be able to get some help from them if we needed it, but I suppose it would be easier for them to answer specific questions.
//Peter
On 8/12/07, Peter Stuge peter@stuge.se wrote:
On Sun, Aug 12, 2007 at 06:25:26PM -0400, Ward Vandewege wrote:
Where are we with that board?
I don't think anyone got around to it.
I never got around to it; stalled at flashrom support. I still have two of the boards sitting here, waiting for me to do my worst.
AFIAK the only one in this "ultra-low-end" class that never got tried was the
BIOSTAR Tforce 550
-dhbarr.
On Sun, Aug 12, 2007 at 07:25:17PM -0500, David H. Barr wrote:
Where are we with that board?
I don't think anyone got around to it.
I never got around to it; stalled at flashrom support.
Didn't flashrom work at least with the flash chip on the Gigabyte board? (I remember there were problems with BIOS savior, but seem to recall that flashing the chip alone worked OK.)
PLCC in socket? Which flash part?
Is the problem the flash chip or the southbridge?
I still have two of the boards sitting here, waiting for me to do my worst.
If you have another LPC board you could use that as a flasher with pushpinflash or a BIOS savior. (1Mbyte RD1-LPC8 is easy to get.)
AFIAK the only one in this "ultra-low-end" class that never got tried was the BIOSTAR Tforce 550
The K9N NEO-F is even nicer on the budget over here. Wow. 2GHz dual core X2 + mobo + 2x512MB RAM = 140EUR! (ex 25% VAT)
//Peter
On Mon, Aug 13, 2007 at 02:46:53AM +0200, Peter Stuge wrote:
Didn't flashrom work at least with the flash chip on the Gigabyte board? (I remember there were problems with BIOS savior, but seem to recall that flashing the chip alone worked OK.)
Yeah, it works fine. I was having some problems with flashrom while running LB (but *not* while running the proprietary bios) a few months ago, but I need to revisit that. I'll make a trac bug if the problem persists with the latest and greatest flashrom.
AFIAK the only one in this "ultra-low-end" class that never got tried was the BIOSTAR Tforce 550
Yeah - it's out of stock here it seems.
The K9N NEO-F is even nicer on the budget over here. Wow. 2GHz dual core X2 + mobo + 2x512MB RAM = 140EUR! (ex 25% VAT)
Actually, the Neo-F also seems to be out of stock in most places. It seems that MCP55-based boards are disappearing from the market...
Thanks, Ward.
On Sun, Aug 12, 2007 at 10:26:20PM -0400, Ward Vandewege wrote:
I was having some problems with flashrom while running LB (but *not* while running the proprietary bios) a few months ago, but I need to revisit that. I'll make a trac bug if the problem persists with the latest and greatest flashrom.
Great.
AFIAK the only one in this "ultra-low-end" class that never got tried was the BIOSTAR Tforce 550
Yeah - it's out of stock here it seems.
The K9N NEO-F is even nicer on the budget over here. Wow. 2GHz dual core X2 + mobo + 2x512MB RAM = 140EUR! (ex 25% VAT)
Actually, the Neo-F also seems to be out of stock in most places.
I can still get both locally. Neo-F at 50EUR, Tforce at 65EUR.
It seems that MCP55-based boards are disappearing from the market...
About time I guess, they're already what, 9-10 months old.. Crazy.
If the hardware doesn't live longer than that, the vendors definately can't afford having software stuck in some legal department for two full months.
//Peter
On Mon, Aug 13, 2007 at 02:46:53AM +0200, Peter Stuge wrote:
PLCC in socket? Which flash part?
For the MSI K9N Neo (MS-7260), which I happen to own: PLCC, in a socket.
I'll test flashrom later, but IIRC correctly it didn't work out of the box when I first tried it.
Is the problem the flash chip or the southbridge?
Not sure, but I'll find out.
AFIAK the only one in this "ultra-low-end" class that never got tried was the BIOSTAR Tforce 550
The K9N NEO-F is even nicer on the budget over here. Wow. 2GHz dual core X2 + mobo + 2x512MB RAM = 140EUR! (ex 25% VAT)
Yep, the board alone is available for 50,- in Germany, for example.
Uwe.
On Tue, Aug 14, 2007 at 02:21:22PM +0200, Uwe Hermann wrote:
Is the problem the flash chip or the southbridge?
Not sure, but I'll find out.
OK, seems to be a mainboard issue:
$ flashrom -wv foo.bin Calibrating delay loop... ok No LinuxBIOS table found. Found chipset "NVIDIA MCP55": Enabling flash write... OK. Pm49FL004 found at physical address: 0xfff80000 Flash part is Pm49FL004 (512 KB) Flash image seems to be a legacy BIOS. Disabling checks. Programming Page: 0007 at address: 0x00070000 Verifying flash - FAILED
Uwe.
On Tue, Aug 14, 2007 at 03:28:10PM +0200, Uwe Hermann wrote:
On Tue, Aug 14, 2007 at 02:21:22PM +0200, Uwe Hermann wrote:
Is the problem the flash chip or the southbridge?
Not sure, but I'll find out.
OK, seems to be a mainboard issue:
Found chipset "NVIDIA MCP55": Enabling flash write... OK. Pm49FL004 found at physical address: 0xfff80000
All other problems have also been with the PMC parts IIRC.
Do you have any other LPC flash chips?
//Peter
On Tue, Aug 14, 2007 at 06:42:27PM +0200, Peter Stuge wrote:
On Tue, Aug 14, 2007 at 03:28:10PM +0200, Uwe Hermann wrote:
On Tue, Aug 14, 2007 at 02:21:22PM +0200, Uwe Hermann wrote:
Is the problem the flash chip or the southbridge?
Not sure, but I'll find out.
OK, seems to be a mainboard issue:
Found chipset "NVIDIA MCP55": Enabling flash write... OK. Pm49FL004 found at physical address: 0xfff80000
All other problems have also been with the PMC parts IIRC.
Do you have any other LPC flash chips?
I don't think the chip is the problem; I just tested another one which I know works fine in the ASUS A8NE-FM/S (however, it's a Pm49FL004 too). The writing fails with that chip.
Uwe.
On Tue, Aug 14, 2007 at 08:52:44PM +0200, Uwe Hermann wrote:
All other problems have also been with the PMC parts IIRC.
Do you have any other LPC flash chips?
I don't think the chip is the problem;
Not the particular chip, but I would like to rule out the chip type.
I just tested another one which I know works fine in the ASUS A8NE-FM/S (however, it's a Pm49FL004 too).
That board has a CK804 southbridge though, right?
The writing fails with that chip.
AFAIK all problems involve MCP55 and PMC flash, even if it's just the inactive chip in a BIOS savior and flashrom should write the other, active chip in a BIOS savior.
Ward's problems are with MCP55+SST-in-BIOS-savior. As I recall Ward got SST chips working if they were put directly in the socket on the board.
The PMC chips can do both LPC and FWH so they have their own state machine which is a bit different from those in a single-protocol chip.
Not a lot, but maybe enough to cause trouble under some circumstances, which happens to collide with what flashrom does.
Yes, it's just a suspicion.
//Peter
On 13.08.2007 01:29, Peter Stuge wrote:
On Sun, Aug 12, 2007 at 06:25:26PM -0400, Ward Vandewege wrote:
It's cheap and on the face of it should be fairly easy to port - MCP55 chipset, etc. There was some concern about the ehternet card?
For LB purposes the NIC is not likely to cause problems, I think the uncertainty was about a Linux driver for it. drivers/net/forcedeth.c does support the relevant PCI IDs (PCI_DEVICE_ID_NVIDIA_NVENET_14/15, 0x0372/0x0373) according to the FreeBSD nfe driver.
If there are problems with forcedeth, I'd like to know about them as I'm one of the authors (although most of my work was in reverse engineering prior to Nvidia supporting the driver).
I think I'll get this board or perhaps one of the Abit KN9 ones that are very similar to the Gigabyte M57SLI-S4. All have a PLCC socket which is a big plus.
If anyone has been working on these boards give me a shout - I'd like to put some work into adding support for one of these.
I don't know if Carl-Daniel got a response from the desktop BIOS people at MSI, I got the impression we'd be able to get some help from them if we needed it, but I suppose it would be easier for them to answer specific questions.
As you say: They were willing to answer specific questions, but I got sidetracked by university papers and unfortunately had no time for porting in the last months.
Regards, Carl-Daniel
-
Carl-Daniel Hailfinger -
David H. Barr -
Peter Stuge -
Uwe Hermann -
Ward Vandewege