| 1 |
On 10:43 Thu 20 Dec , Jan Kundrát wrote: |
| 2 |
> Donnie Berkholz wrote: |
| 3 |
> > Looking at my kernel config, ext3 and reiser explicitly support |
| 4 |
> > xattrs, and I see jfs and xfs have acls and security labels, which |
| 5 |
> > might be usable. |
| 6 |
> |
| 7 |
> Extended attributes can be turned off during compile time for each |
| 8 |
> filesystem you mentioned. |
| 9 |
|
| 10 |
And? |
| 11 |
|
| 12 |
If you turn off features you need, things break. There's nothing new |
| 13 |
about that. If you disable ext3 support in your kernel, you can't mount |
| 14 |
an ext3 partition and you'll get an error during boot about not finding |
| 15 |
the root. |
| 16 |
|
| 17 |
> NFSv3 doesn't support them (yes, I do share $PORTDIR). |
| 18 |
|
| 19 |
While doing a search about this, I came across a page on the Beagle site |
| 20 |
describing how Beagle deals with a similar issue that says: |
| 21 |
|
| 22 |
"Extended attributes are used by Beagle to keep track of which files |
| 23 |
have been indexed and which need to be re-indexed. There is a |
| 24 |
sqlite-based fallback which is a bit slower, so it is recommended that |
| 25 |
you do use extended attributes. ... |
| 26 |
|
| 27 |
"Your kernel will need support for extended attributes on the filesystem |
| 28 |
you are indexing. If you are using XFS or JFS, extended attributes are |
| 29 |
always enabled and you can skip this section and move on to Installing |
| 30 |
prerequisites. For Reiser3, ext2 and ext3 filesystems, the default |
| 31 |
kernel config does not enable the attributes. ... |
| 32 |
|
| 33 |
"Also note that neither Reiser4 nor NFS support extended attributes, so |
| 34 |
the sqlite-based fallback will be used by default." |
| 35 |
|
| 36 |
The idea of the sqlite-based fallback is what's interesting here. |
| 37 |
|
| 38 |
> Also note that in some circumstances like when running in a |
| 39 |
> virtualized environment, imposing additional requirements on the kernel |
| 40 |
> might be problematic. |
| 41 |
|
| 42 |
Why's that? Here's some performance tests from 3 years ago on a 2.6.10 |
| 43 |
kernel that hopefully have improved in the meanwhile: |
| 44 |
|
| 45 |
http://lwn.net/Articles/112566/ |
| 46 |
|
| 47 |
> It wouldn't be great to require extended attributes for each and every |
| 48 |
> Gentoo box... |
| 49 |
|
| 50 |
Why not? |
| 51 |
|
| 52 |
Thanks, |
| 53 |
Donnie |
| 54 |
-- |
| 55 |
gentoo-dev@g.o mailing list |
Archives