hdparm 9.43 -> 9.45+ update request

[A. Eibach]

This is, in fact, NOT a real bug report.

However, as Manfred asked me to report it here anyway (https://answers.launchpad.net/ubuntu/+source/hdparm/+question/270310 ) I will do so.
(Let alone because the genuine Debian OS is also still at 9.43...)

Yes, (*)Ubuntu is VERY much behind with hdparm, quasi-stuck at 9.43 for more than 2 years. This is a bit much.

If you'd like to know why it may be sensible to update, here's an example:

https://sourceforge.net/p/hdparm/bugs/51/

--read-sector is byte-swapped in every version <= 9.43. Mark fixed this in 9.45+. unfortunately you guys were asleep with both eyes open and still haven't considered a tool update for 2 + years :)

[racitup (racitup)]

I have found a bug in hdparm v9.43 with the --fibmap option.
For certain filesystems it returns incorrect values when compared with filefrag, example below for vfat. Notice the factor of 8 difference between number of extent sectors: 125092 vs 1000736.
I have reported this to Mark Lord (author) and am awaiting feedback whether this has been fixed in a later version.

hdparm --fibmap ./edisk5/mnt.erox9t4T/emptyspace.zeros, out=./edisk5/mnt.erox9t4T/emptyspace.zeros:
filesystem blocksize 512, begins at LBA 4102144; assuming 512 byte sectors.
byte_offset begin_LBA end_LBA sectors
0 4125406 4250497 125092, err=, ret=0

filefrag -b512 -e ./edisk5/mnt.erox9t4T/emptyspace.zeros, out=Filesystem type is: 4d44
File size of ./edisk5/mnt.erox9t4T/emptyspace.zeros is 512376832 (1000736 blocks of 512 bytes)
ext: logical_offset: physical_offset: length: expected: flags:
0: 0.. 1000735: 186096.. 1186831: 1000736: 1186824: merged,eof
./edisk5/mnt.erox9t4T/emptyspace.zeros: 1 extent found, err=, ret=0

[racitup (racitup)]

Further to my previous post, it appears this was fixed in v9.45 according to this bug report and patch:

https://sourceforge.net/p/hdparm/bugs/58/