Owners check treats imported files as OWNERS of the directory containing them |
|||||
Issue descriptionIf foo/OWNERS contains "file://bar/NOT_OWNERS" then if a CL contains changes in both foo/ and bar/, bar/NOT_OWNERS will be treated as an OWNERS file for bar/ (either instead of or in addition to the actual bar/OWNERS file, depending on iteration order). This can cause ipc/SECURITY_OWNERS to be used instead of or in addition to ipc/OWNERS.
,
Sep 1 2016
,
Sep 4 2017
This issue has been Available for over a year. If it's no longer important or seems unlikely to be fixed, please consider closing it out. If it is important, please re-triage the issue. Sorry for the inconvenience if the bug really should have been left as Available. If you change it back, also remove the "Hotlist-Recharge-Cold" label. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Sep 5 2017
Oh man, this is definitely a problem, and one I'd totally forgotten about until now. We need a full overhaul of how OWNERS works now that we're using Gerrit.
,
Sep 5 2017
I could've sworn we fixed this, or something like it, at some point, so we should re-test to verify. @agable - why do you think we need a full overhaul of how OWNERS work?
,
Sep 6 2017
We don't need an overhaul of how OWNERS files work, but we do need an overhaul of how other systems (presubmit, cq, git-cl, gerrit, etc) use them. For example, checking OWNERS in presubmit on a trybot is horridly broken. The patch has been applied before the owners check runs, so I can add myself to an OWNERS file and then approve my own change; the code to prevent that is messy and involves lots of special casing. For example, when TBRing something, nothing checks to make sure that the person being TBRed is an OWNER of the relevant code. For example, git-cl is unintelligent about how it suggests owners, and Gerrit can't make suggestions at all yet. I think this is the fix you're thinking of: https://chromium-review.googlesource.com/447962 . I think that does fully resolve this bug; please verify.
,
Sep 6 2017
> For example, checking OWNERS in presubmit on a trybot is horridly broken. > The patch has been applied before the owners check runs, so I can add myself > to an OWNERS file and then approve my own change; the code to prevent that > is messy and involves lots of special casing. I don't think it's *that* messy, but I take your point. > For example, when TBRing something, nothing checks to make sure that the > person being TBRed is an OWNER of the relevant code. True. I don't remember why we did things this way, but it wouldn't be hard to change this. > For example, git-cl is unintelligent about how it suggests owners, > and Gerrit can't make suggestions at all yet. I'm not sure what your definition of unintelligent here. There is *an* algorithm to do this, though it isn't as smart as Google3's :). |
|||||
►
Sign in to add a comment |
|||||
Comment 1 by martiniss@chromium.org
, Aug 10 2016Components: -Infra Infra>SDK