On Fri, Oct 10, 2008 at 06:02:49PM -0400, Valerie Aurora Henson wrote: > On Fri, Oct 10, 2008 at 04:43:43PM -0400, Theodore Ts'o wrote: > > > > There was discussion about how to do 64-bit ABI conversation, and an > > assertion that it might be hard to do ext2fs_block_iterate3. So I > > decided give an example of how I believe the most efficient way to do > > this sort of thing. > > I'm afraid there's been a misunderstanding; I wasn't saying it was > hard, just pointing out that different versions of interfaces are > profilerating as a result of continuing to support the old ABI at the > same time as the new 64-bit ABI - perhaps more versions of more > functions than was originally anticipated. (I wrote a similar patch > and I agree it's not hard, just more cruft to maintain.) But it's not that much more cruft to maintain. Take a look at my patch; the original interface gets implemented on top of the old interface, so the amount of extra cruft to maintain is minimal. The implementation of ext2fs_block_iterate2 simply becomes: struct xlate64 xl; xl.real_private = priv_data; xl.func = func; return ext2fs_block_iterate3(fs, ino, flags, block_buf, xlate64_func, &xl); That wasn't that bad, was it? > Eagerly awaiting ext2fs_block_iterate4, If there are other changes that we need to implement at the same time, if we haven't done a release yet, we can fold those changes into the ext2fs_block_iterate3. - Ted -- 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
