On Fri, 2006-10-20 at 09:44 +0200, [EMAIL PROTECTED] wrote:

> I now have a semi-working program that can extract files with filenames
> and paths. I am test-driving it on a healthy jfs-image. It does not work
> that relaible yet. So, heres my questions:
> 
> Does an inode-extent _always_ consist of 32 consecutive inodes, with
> consecutive di_number?

Yes

>  The reason I ask is that I found inodes, that seems
> to be valid, but which are parts of a much smaller extent (5-25inos or
> so). Is this possible? 

I'm not sure how.  If a freed inode extent is partially reused, I would
expect entire 4K pages to be overwritten, leaving you with 8 or 16
inodes together, probably all with di_nlink == 0.

> It might also be due to that neighbouring inodes,
> that should have been a part of the extent, failed the sanity-check
> (mentioned elseware in this thread).

The extent will contain both allocated and unallocated inodes.  For both
of these, di_inostamp should be the same, di_number should be valid, and
di_fileset should be 16.  di_nlink being non-zero indicates that the
inode is allocated.

> I am getting some dtree.header.nextindex==-1 when I am parsing the dtrees.
> Is this really a valid value or does it indicate that there are errors in
> the dtree?

I don't think that's valid, so I'm not sure why you are seeing that.
nextindex should always be <= dtree.header.maxslot, which should never
be greater than 128.  nextindex is unsigned, so -1 = 255.
> 
> Best regards,
> Simon
> 
-- 
David Kleikamp
IBM Linux Technology Center


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Jfs-discussion mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jfs-discussion

Reply via email to