https://codereview.chromium.org/1183563003 made it so that slave checkouts ignore master state and use slave state instead.

It looks like this isn't working fully reliably.

https://build.chromium.org/p/tryserver.chromium.win/builders/win_clang?numbuilds=200 has occasional blocks of "Exception steps exception setup_properties" exceptions, an example is here:

https://build.chromium.org/p/tryserver.chromium.win/builders/win_clang/builds/147

@@@STEP_CURSOR setup_properties@@@
@@@STEP_STARTED@@@
@@@STEP_CURSOR setup_properties@@@
@@@STEP_LOG_LINE@exception@Traceback (most recent call last):@@@
@@@STEP_LOG_LINE@exception@  File "../../../scripts/slave/annotated_run.py", line 599, in get_recipe_properties@@@
@@@STEP_LOG_LINE@exception@    workdir, mastername, buildername)@@@
@@@STEP_LOG_LINE@exception@  File "../../../scripts/slave/annotated_run.py", line 681, in get_factory_properties_from_disk@@@
@@@STEP_LOG_LINE@exception@    (buildername, mastername))@@@
@@@STEP_LOG_LINE@exception@LookupError: builder "win_clang" not found on in master "tryserver.chromium.win"@@@
@@@STEP_LOG_END@exception@@@
@@@STEP_EXCEPTION@@@

I think I configured the master correctly and restarted it yesterday night, and based on most builds succeeding, I think that's true.

It looks like these mostly happen on few slaves:
https://build.chromium.org/p/tryserver.chromium.win/buildslaves/vm612-m4
https://build.chromium.org/p/tryserver.chromium.win/buildslaves/vm624-m4

...oh, it looks like update_scripts fails with "fatal: Unable to create 'E:/b/build/.git/refs/remotes/origin/master.lock': File exists."

Maybe the code from https://codereview.chromium.org/1183563003 should print a better error message if the slave checkout is out of date. I'll file a different ticket for getting these two slaves cleaned up, so I suppose this bug isn't super important. Maybe it helps someone googling for this error in the future.