--- On Tue, 13/11/12, akpm@xxxxxxxxxxxxxxxxxxxx <akpm@xxxxxxxxxxxxxxxxxxxx> wrote: > From: akpm@xxxxxxxxxxxxxxxxxxxx <akpm@xxxxxxxxxxxxxxxxxxxx> > Subject: + hfsplus-avoid-crash-on-failed-block-map-free.patch added to -mm tree > To: mm-commits@xxxxxxxxxxxxxxx > Cc: alan@xxxxxxxxxxxxxxx, hch@xxxxxx, htl10@xxxxxxxxxxxxxxxxxxxxx, jack@xxxxxxx, slava@xxxxxxxxxxx, viro@xxxxxxxxxxxxxxxxxx > Date: Tuesday, 13 November, 2012, 22:35 > > The patch titled > Subject: hfsplus: avoid crash on > failed block map free > has been added to the -mm tree. Its filename is > > hfsplus-avoid-crash-on-failed-block-map-free.patch > > Before you just go and hit "reply", please: > a) Consider who else should be cc'ed > b) Prefer to cc a suitable mailing list as > well > c) Ideally: find the original patch on the > mailing list and do a > reply-to-all to that, adding suitable > additional cc's > > *** Remember to use Documentation/SubmitChecklist when > testing your code *** > > The -mm tree is included into linux-next and is updated > there every 3-4 working days > > ------------------------------------------------------ > From: Alan Cox <alan@xxxxxxxxxxxxxxx> > Subject: hfsplus: avoid crash on failed block map free > > If the read fails we kmap an error code. This doesn't > end well. Instead > print a critical error and pray. This mirrors the rest > of the fs > behaviour with critical error cases. > > Acked-by: Vyacheslav Dubeyko <slava@xxxxxxxxxxx> > Signed-off-by: Alan Cox <alan@xxxxxxxxxxxxxxx> > Signed-off-by: Vyacheslav Dubeyko <slava@xxxxxxxxxxx> > Cc: Hin-Tak Leung <htl10@xxxxxxxxxxxxxxxxxxxxx> > Cc: Al Viro <viro@xxxxxxxxxxxxxxxxxx> > Cc: Christoph Hellwig <hch@xxxxxx> > Cc: Jan Kara <jack@xxxxxxx> > Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> Acked-by: Hin-Tak Leung <htl10@xxxxxxxxxxxxxxxxxxxxx> For some reason I don't seem to have received this the first time it was posted (despite being in CC). I'll have this a try to see if it helps my kmap problem (where running 'du' on a large dir on hfs+ eventually fails). BTW, there is a PATCH v2 for this; I'll give that a go as well. Since I last posted about my kmap problem, I tried switching all the dprint() to dynamic debugging (as suggested by Vyacheslav, private e-mail exchange). This seems to have achieved a similar effect as my voodoo kmap mutex - slowing down the hfs+ driver enough for it to be more difficult for du to fail - I need to try for ~10 times before it fails. Also a bit earlier than du finally triggers a kernel trace, the hfs+ driver reports cat b-tree inconsistent. (and fsck afterwards again says nothing is wrong). So Alan might be on to something there - it seems it is just the kernel itself getting confused when pages are mapped and unmapped. I also found some of the dprint() not detailed enough (when dyn-enabled). So I have a few patches, some just more informative, sitting in my private repo. I hope to get to the bottom of the kmap problem eventually. Hin-Tak -- 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
