rfsd 0.25.1 released

rfsd 0.25.1 released

[Bo Brantén]

rfsd 0.25.1 has been released, download it here: 
http://www.acc.umu.se/~bosse/rfsd-0.25.1.zip

Changed since last release:
Fixed the directory listing bug. (This bug apeared when the caller
supplied buffer didn't fit all the information for a directory so that
more than one query directory call was needed, when processing the first
directory entry in the second (and consecutive) calls the filename length
was misscalculated since information on the directory entry before wasn't
saved between the calls)

Known problems:
- Some files misses the last block when reading data from them. At a test
on the linux kernel source code that contains aproximatly 24000 files,
10% or aproximatly 2400 files showed this problem, however if the file
system was mounted with the notail option while untaring the files on
Linux less than 1% or 108 files showed this problem.
- No write support yet.

Bo Branten

Question on how file tails are stored

[Bo Brantén]

Hello,

I would like to ask you how file tails are stored on ReiserFs 3.6, am I 
right that they are stored as either one or two items but not more than 
two?

Also I found this statement on the Internet:

"If a tail is larger than the maximum size of a file that can fit into a 
formatted node but is smaller than the unformatted node size (4K), then it 
is stored in an unformatted node, and a pointer to it plus a count of the 
space used is stored in an indirect item."

Does this represent a third case not covered of what I said above, and if 
so how does one detect this case and find the tail data?

Bo Branten

Re: Question on how file tails are stored

[Edward Shishkin]

Bo Brantén wrote:
>
> Hello,

Hello.

>
> I would like to ask you how file tails are stored on ReiserFs 3.6, am
> I right that they are stored as either one or two items but not more
> than two?

Yes, if you mean (default) "small" tail suppression policy.

Indeed, a data flow of size < blocksize can not be
split into more then 2 chunks, if you partition it by a lattice
which has a step == blocksize. 

>
> Also I found this statement on the Internet:
>
> "If a tail is larger than the maximum size of a file that can fit into
> a formatted node but is smaller than the unformatted node size (4K),
> then it is stored in an unformatted node, 

Right.
It fits perfectly the "small" policy .


> and a pointer to it plus a count of the space used is stored in an
> indirect item."

This looks like some not related calculation..

>
> Does this represent a third case not covered of what I said above, and
> if so how does one detect this case and find the tail data?

There is no special interfaces to access tail data.
You can try to parse debugreiserfs -d output..

Thanks,
Edward.
--
To unsubscribe from this list: send the line "unsubscribe reiserfs-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html