FPGA: Added a workaround for the GW1N-4D IDCODE(Flaws of Apicula).
This commit is contained in:
1 parent
e69b791612
commit
4f01016608
4 files changed
+118
-3
No files matched your search
@@ -59,6 +59,15 @@ jobs:
|
||||
test "$fs_size" -gt $((bin_size * 6))
|
||||
test "$fs_size" -lt $((bin_size * 10))
|
||||
|
||||
# The device only accepts a bitstream naming its own IDCODE and
|
||||
# otherwise stays unconfigured, with nothing driving the antennas.
|
||||
# apycula's GW1N-4 database writes the plain GW1N-4, so the build
|
||||
# stamps the part's in (fpga/common/gowin_devid.py) - keep it that way.
|
||||
# The IDCODE is the four bytes at 0x1C of the header.
|
||||
idcode=$(od -An -tx1 -j 28 -N 4 fpga/fpga_pm5.bin | tr -d ' \n')
|
||||
echo "device idcode in the bitstream: 0x${idcode}"
|
||||
test "$idcode" = "1100381b"
|
||||
|
||||
# A rebuild from scratch has to produce the same bytes; the tools carry no
|
||||
# timestamp, so a difference means something started leaking one in, and
|
||||
# a released bitstream could no longer be checked against the sources.
|
||||
|
||||
Reference in new issue
Block a user