On Thu, 2009-04-16 at 13:12 -0400, Chris Mason wrote: > On Thu, 2009-04-16 at 10:56 -0400, Chris Mason wrote: > > On Thu, 2009-04-16 at 13:39 +0200, Mike Galbraith wrote: > > > On Wed, 2009-04-15 at 13:22 -0400, Chris Mason wrote: > > > > Hello everyone, > > > > > > Greetings, > > > > > > > This passed a longer stress test and generally seems to be working. I > > > > don't think anyone would recommend it as a default for 2.6.30, but it > > > > may be a good idea to have a review party and decide if it is safe enough > > > > to include so people can experiment with it. > > > > > > I know you didn't say RFT, but I did some anyway, and found a 100% > > > repeatable corruption scenario wrt git+umount. > > > > > > > Well, that's a surprise. I can trigger it here too, trying to figure > > out how this is different from the fsx and fsstress hammering. It looks > > like git is just going good old fashioned writes, so I must be losing > > one or two of them. > > Ah ok, it is just a missed i_size update. Basically because file_write > doesn't wait for page writeback to finish, someone can be updating > i_size at the same time the end_io handler for the last page is running. > > Git triggers this when it does the sha1flush just before closing the > file. A brief note (preferably with 8x10 color glossies with circles and arrows) on how you figured that out so quick would be a good LKML archive investment :) -Mike -- To unsubscribe from this list: send the line "unsubscribe linux-ext4" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
