On Thu, Mar 22, 2018 at 4:20 PM, Amir Goldstein <amir73il@xxxxxxxxx> wrote: > On Thu, Mar 22, 2018 at 1:55 PM, Amir Goldstein <amir73il@xxxxxxxxx> wrote: > Yes, it is a mistake as you observed. > I was trying to understand why it didn't affect the tests. > > Apparently, the case where underlying dentry is disconnected > and overlay inode is in cache is harder to reproduce and is not covered by > the current tests. I think it requires: > - encoding non-upper non-dir overlay file handle > - drop caches > - open by non-upper file handle (disconnected underlying and not in > overlay icache) > - open by non-upper file handle again (disconnected underlying and in > overlay icache) > > The current tests only cover: > - encoding non-upper non-dir overlay file handle > - drop caches > - open by non-upper file handle (disconnected underlying and not in > overlay icache) > AND: > - encoding non-upper non-dir overlay file handle AND keep the file open > - drop caches > - open by non-upper file handle (connected underlying and in overlay icache) > > I will add the missing tests to my TODO. > I suppose you will be fixing this bug on commit? Yes. Thanks, Miklos -- To unsubscribe from this list: send the line "unsubscribe linux-unionfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
