On 16 June 2011 17:15, J. R. Okajima <hooanon05@xxxxxxxxxxx> wrote: > > Michal Suchanek: >> Probably swap the two above, you can't make a whiteout in presence of >> the directory, right? >> Anyway, you could just mark dirA as whiteout and remove any whiteouts >> contained in it asynchronously, and only jump through these hoops when >> trying to create a new entry in place of non-empty whiteout, or sync >> on emptying the old whiteout before making a new entry. > > Unfortunately I cannot understand what you wrote. > > First, the order of >> - create whiteout for dirA >> - rename dirA to .wh..wh.XXXXXXXX > is correct and I think it should be, in order to make a little help for Yes, it's correct for aufs which uses reserved file names for whiteouts. Filesystems that don't reserve filenames cannot make whiteout for an existing entry but aufs can. > fsck/auchk. > And what is "non-empty whiteout" and "emptying the old whiteout"? > The whiteout is a size zero-ed and hardlinked regular file in aufs. Is there any reason why a directory cannot be whiteout? > > >> Yes, it can only cause pollution with whiteouts unrelated to any files >> that ever existed which is not too much of an issue unless people want >> to add random stuff to the lower layer and see it in the union when >> they reconstruct it again. > > ?? > Do you think that the .wh..wh.XXXXXXXX hides something on the lower > layer? If so, it is wrong. Such doubly whiteout hides nothing except > itself. It may possibly hide a XXXXXXXX file if it is later added to the lower layer. But if .wh.XXXXXXXX is in itself a reserved filename that is never brought up from the lower layer then this is a non-issue, it works consistently regardless of existence of the superfluous whiteout. > > >> It is only valid when in the upper layer of a union. However, so is >> whiteout, and so are files that were visible in the union but are not >> visible in the top layer if examined separately, outside of the union. > > Do you mean that your special symlink has totally different file-type > from a symlink? Just as whiteout has totally different file-type from a file. It's specific to the union. > Anyway what I want to say is, what such symlink refers may differ > from what users originally expect. But I may misunderstand what you call > "fallthru symlink". How is this different from other files that are taken from the lower layer and not copied into the upper layer? If you are concerned about that you want a full copy, not a union. Thanks Michal -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
