When icebram is called to replace RAM contents but no RAM contents are replaced, it currently simply exits. This indicates a build failure. (Why else would anyone call icebram?)
This becomes more important due to this issue, where icebram fails after switching to a new memory_libmap. Existing projects that used to build just fine are currently creating broken bitstreams, and unless the -v flag is used, no warnings are generated. (Not that anyone actually looks at warnings...)
Ultimately, icebram should be fixed so that it works with the new memory_libmap, but until that's done, it should at least exit with any error instead, which is what this PR does.
When
icebram
is called to replace RAM contents but no RAM contents are replaced, it currently simply exits. This indicates a build failure. (Why else would anyone callicebram
?)This becomes more important due to this issue, where
icebram
fails after switching to a newmemory_libmap
. Existing projects that used to build just fine are currently creating broken bitstreams, and unless the-v
flag is used, no warnings are generated. (Not that anyone actually looks at warnings...)Ultimately,
icebram
should be fixed so that it works with the newmemory_libmap
, but until that's done, it should at least exit with any error instead, which is what this PR does.