Attention is currently required from: Namyoon Woo, Thomas Heijligen, Edward O'Callaghan, Anastasia Klimchuk.
2 comments:
Patchset:
I have a question. I created https://ticket.coreboot. […]
We discussed this in a meeting today. I think the best is to
temporarily move the code as is into a separate file that re-
flects that there is more to do, e.g. `fixme.c`. Then this
patch can go in and Thomas doesn't have to wait for some-
body to fix the code.
File spi.c:
int probe_variable_size(struct flashctx *flash)
{
unsigned int i;
const struct emu_data *emu_data = flash->mst->spi.data;
Also it relies on emu_data, what happens if there is some other data in the context? another struct type?
You are spot-on, this is most likely undefined behavior. Probably what I had
in mind when I added this feature to the list of release-blocking issues. I
guess I commented about that on Gerrit a while ago (CB:44879). It's a perfect
example how merging without proper review and keeping broken code(!) affects
the project.
To view, visit change 63831. To unsubscribe, or for help writing mail filters, visit settings.