Chris Elston wrote:
This problem also does not occur on XFS data partitions because they make
use of allocation groups, across which extents records may not span.
Allocation groups have a maximum size of just under 4Gb.
FWIW, this is not the case any longer; AGs can now be > 4Gb.
However, I still haven't shown this problem on the data subvol...
I made a fs with 2 AGs, first is 5gb. Then I pre-allocated 5G into a file, and
got -almost- 1 extent ;-) but first extent is nearly 5g. I wrote "1" to 0
bytes and "2" to 2^32+4097 bytes, and read it back. Seems ok, oddly enough.
penguin3:~ # ./allocsp /mnt/test/bigfile
pwrite at offset 0
pwrite at offset 4294972296
penguin3:~ # xfs_bmap -v /mnt/test/bigfile
/mnt/test/bigfile:
EXT: FILE-OFFSET BLOCK-RANGE AG AG-OFFSET TOTAL
FLAGS
0: [0..10485663]: 96..10485759 0 (96..10485759) 10485664
10000
1: [10485664..10485759]: 10506304..10506399 1 (20544..20639) 96
10000
penguin3:~ # hexdump /mnt/test/bigfile
0000000 0031 0000 0000 0000 0000 0000 0000 0000
0000010 0000 0000 0000 0000 0000 0000 0000 0000
*
100001380 0000 0000 0000 0000 0032
100001389
penguin3:~ #
|