Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
9 years ago

fpga load gives error code -2

Hello,

We have created a custom board with a Cyclone V on it. The boot process works: We can go from preloader to u-boot to linux.

But now I want to load something in the FPGA, and am trying to do from u-boot. However, I get stuck at error -2:

SOCFPGA_CYCLONE5#  mmc rescan
SOCFPGA_CYCLONE5#  fatload mmc 0:2 0x2000000 switchbox_master.rbf
reading switchbox_master.rbf
7007204 bytes read in 393 ms (17 MiB/s)
SOCFPGA_CYCLONE5#  fpga load 0 0x2000000 7007204
MSEL = 0
put FPGA in reset phase
FPGA is in reset phase
FPGA is out of reset phase
altera_load: Failed with error code -2

(I added some debug printout)

According to https://rocketboards.org/foswiki/view/documentation/gsrd131programmingfpga Error code -2 means: The FPGA is not able to enter configuration phase after FPGA reset release request being made at FPGA Manager

Any idea what can be wrong?

Thanks,

Michiel

3 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor

    The data transfer yet happen which indicate the error not coming from the RBF.

    A quick check on the FPGA config signal tie off against dev kit would be useful
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor

    We based our hardware design in the EBV SoCrates board. I looked at the nCONFIG pin, but could not see any difference. Is there some other pin? Or maybe a clock? I can't probe the pins at the moment, because I need our hardware guy for that, and he is away on holiday.

    I now started the SoCrates board, and with the same rbf file, I get error -4. I think that is a seperate error, maybe related to the RBF format.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor

    WE solved it now. It turned out that there was a short between CONF_DONE and nSTATUS. With that fixed, programming the FPGA works. Thanks for the suggestion to look there!