2012-10-16 (화), 16:14 +0000, Arnd Bergmann: > On Tuesday 16 October 2012, Jaegeuk Kim wrote: > > Thank you for a lot of points to be addressed. :) > > Maybe it's time to summarize them. > > Please let me know what I misunderstood. > > > > [In v2] > > - Extension list > > : Mkfs supports configuring extensions by user, and that information > > will be stored in the superblock. In order to reduce the cleaning > > overhead, > > f2fs supports an additional interface, ioctl, likewise ext4. > > That is what I suggested but actually Dave Chinner is the person that you > need to listen to rather than me in this regard. Using an extended attribute > in the root node would be more appropriate to configure this than an ioctl. > > > - The number of active logs > > : No change will be done in on-disk layout (i.e., max 6 logs). > > Instead, f2fs supports changing the number with a mount option. > > Currently, I think 4, 5, and 6 would be enough. > > Right, that would be the minimum that I would ask for. If it is relatively > easy to support more than six logs in the file format without actually > implementing them in the code, you might want to support up to 16, just > to be future-proof.
Ok, got it. > > For the lower bound, being able to support as little as 2 logs for > cheap hardware would be nice, but 4 logs is the important one. > > 5 logs is probably not all that important, as long as you have the > choice between 4 and 6. If you implement three different ways, I > would prefer have the choice of 2/4/6 over 4/5/6 logs. Ok, I'll try, but in the case of 2 logs, it may need to change recovery routines. > > > - Section size > > : Mkfs supports multiples of segments for a section, not power-of-two. > > Right. > > > [Future optimization] > > - Data separation > > : file access pattern, and else? > > : Investigate the option to make large files erase block indirect rather than > part of the normal logs > > There is one more more point that I have not mentioned before, which is the > alignment of write requests. As far as I can tell, you try to group writes > as much as possible, but the alignment and the minimum size is still just > 4 KB. Yes. > I fear that this might not be good enough for a lot of cases when > the page sizes grow and there is no sufficient amount of nonvolatile > write cache in the device. I wonder whether there is something that can > be done to ensure we always write with a minimum alignment, and pad > out the data with zeroes if necessary in order to avoid getting into > garbage collection on devices that can't handle sub-page writes. You're very familiar with flash. :) Yes, as the page size grows, the sub-page write issue is one of the most critical problems. I also thought this before, but I have not made a conclusion until now. Because, I don't know how to deal with this in other companies, but, I've seen that so many firmware developers in samsung have tried to reduce this overhead by adapting many schemes. I guess very cautiously that other companies also handle this well. Therefore, I keep a question whether file system should care about this perfectly or not. Thanks, > > Arnd -- Jaegeuk Kim Samsung -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/