diff options
author | Peter Marheine <pmarheine@chromium.org> | 2021-09-01 13:14:55 +1000 |
---|---|---|
committer | Felix Held <felix-coreboot@felixheld.de> | 2021-09-02 14:53:59 +0000 |
commit | 5d91bdd3126f507d5c338bcd6f252b2cc814fe4a (patch) | |
tree | 33260c76796a024e2591262dc22668d5e8447508 /src/mainboard/google/jecht/bootblock.c | |
parent | 4a359401370d179b6b37dd6a6ff0c87ca8e1d241 (diff) |
mb/google/zork: correct MST probes
It turns out that putting a device ref in an overridetree at a different
point in the tree will generate a duplicate device definition, such that
the change introducing this support was ignoring the device presence
specified by overridetree.cb and only using the baseboard configuration.
I believe testing of that change was not redone after the baseboard was
changed to disable the MST, so that conflicting behavior was not
noticed.
The incorrect behavior generated a disabled device for the MST at the
location specified by the baseboard, and one with the probe as a child
of the soc. At runtime this did a fw_config probe of the "I2C 00:4a"
device, and later probed a different "I2C 00:4a" which was already
disabled. As the disabled one came later, it seems to have completely
disabled the MST, discarding the results of the variant-specific probe.
BUG=b:185862297
TEST=10EC2141 device is now present on a Dali berknip
BRANCH=zork
Change-Id: I2a8feb544f3fc198fe6313b226ad8995aad31c3e
Signed-off-by: Peter Marheine <pmarheine@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/57298
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Diffstat (limited to 'src/mainboard/google/jecht/bootblock.c')
0 files changed, 0 insertions, 0 deletions