tel me what's printed on the chip, open the socket, don't remove the chip or make a picture to remember the orientation.Where can I get ASRock BMC flash chips? I want to try putting one in from some 3647 Board and test if the BMC works.
tel me what's printed on the chip, open the socket, don't remove the chip or make a picture to remember the orientation.Where can I get ASRock BMC flash chips? I want to try putting one in from some 3647 Board and test if the BMC works.
I want to try a ROM chip from some other motherboard. Maybe it will work.tel me what's printed on the chip, open the socket, don't remove the chip or make a picture to remember the orientation.
Highly unlikely unless the PCH/CPU and all eSPI/LPC devices are wired up and addressed exactly the same. One bit of luck might be that the design is based off of something very standard because spinning a completely new design would be extremely expensive. That said, it has an FPGA wired in so perhaps not as standard as we'd like.I want to try a ROM chip from some other motherboard. Maybe it will work.
But the chip that contains the BMC rom is a Winbond 25q256jvfq
To blue right slotWhere can I get ASRock BMC flash chips? I want to try putting one in from some 3647 Board and test if the BMC works.
Yes, but as far as I know the BMC discovers all devices. The FPGA is not really important for me. But many boards from ASRock have exactly the same firmware, let alone the ASMB9-IKVM module. It is supported by more than two motherboards with different chipsets.Highly unlikely unless the PCH/CPU and all eSPI/LPC devices are wired up and addressed exactly the same. One bit of luck might be that the design is based off of something very standard because spinning a completely new design would be extremely expensive. That said, it has an FPGA wired in so perhaps not as standard as we'd like.
I haven't been able to trace this off of the pictures, but it's possible that the 2nd 10G port is wired up to the FPGA and not the PCIe bus for the host CPU.
poweron with keyboard ?I've been shorting PS_ON when appling power to the PSU to get it to turn on. I don't see any other settings in bios or jumper on board.
Neither of those are options in the bios.poweron with keyboard ?
there should be a setting in the BIOS what todo if AC loss
stay OFF
turn ON
last state
The BMC is supposed to be powered by 5vSB, PMBus be damned. If it isn't it probably means that either it was never going to work OOB/LOM in S5. The PMBus can't really be accessed by anything that doesn't have any brains, so unless there is some EC that is powered by 5vSB which would then control the PMBus, SMBus and eSPI/LPC, the BMC should simply always be in.Neither of those are options in the bios.
It seems that with only 5VSB the BMC isn't running. I'm not sure what the deal with this board is, or if it needs the PMBus PSU hooked up.
So even if we get BMC up on the board you wouldn't be able to turn the board on (at least that the way its looking right now). Prolly going to just need to probe out how PS_ON is triggered by this board.
The BMC is supposed to be powered by 5vSB, PMBus be damned. If it isn't it probably means that either it was never going to work OOB/LOM in S5. The PMBus can't really be accessed by anything that doesn't have any brains, so unless there is some EC that is powered by 5vSB which would then control the PMBus, SMBus and eSPI/LPC, the BMC should simply always be in.
Looking at the extra headers, there do seem to be BMC-specific I2C headers and serial headers; heck, that USB-B port might even be wired up to the BMC instead of the SIO and used for primary OOB management if they went super custom. That might be testable; if you plug in a USB-A to USB-B cable in there and in a working machine it should pop up as a generic serial port.
this is BMC I2C(PMBus/SMBus) Debug port, not BMC UART (sure I2C is serial too)There is a BMC_DEBUG with some pins on the board, next to the missing dedicated lan port for the BMC. Afaik this is the serial port for the BMC
There is an unpopulated JTAG header, and I wouldn't be surprised if there was a serial interface around there as well. The FPGA has its own header more close to itself, and Intel's debug header is on the back. But none of that helps if nobody with a board also has a debuggerthis is BMC I2C(PMBus/SMBus) Debug port, not BMC UART (sure I2C is serial too)
Does it not show up at all, like not even in some lspci -vv with a weird ID? Does dmesg say anything about a bad PCIe thing?3A 5Vsb supply fixed the booting problem. Now I just need to figure out why this OCP NIC isn't showing in LSPCI... Tried installing drivers but if it's not detected I'm not sure what to do.
Tried a second OCP NIC (CX3) and the result is the same, the board doesn't detect the card at all, tried changing bifurication and ASPM settings with no luck.Does it not show up at all, like not even in some lspci -vv with a weird ID? Does dmesg say anything about a bad PCIe thing?
As far as I know the mezzanine slot doesn't really need any special sauce to be picked up by the OS if something is plugged in normally.