It's the usual problem, need to do all partitioning with OS/2 aware tools, and do some math to have the 4k sector partitioning, basically JFS 4k sectors need to align with the SSD's 4k sectors.
My SSD works fine, Linux is happy with it and can trim it.
You might have to start over with blanking the beginning of the drive and then creating the partitions.
I have a 1TB SSD
(I put 4 partition on it)
I just could have all aligned seen under windows using msinfo32 (or linux $ sudo fdisk -l to get starting offset - To have aligned partitions, first should start at 1MiB ofset)
I could have all aligned with the starting offest been a multiple of 4096. The easiest was to create partition under my win7 which set first ofset at 1MiB by defaut, compatible with HDD and SDD.
All was looking good but...
(I noted that L-Geo and S-geo CHS are now 976761-64-32

I checked using testdisk and geo seems ok
I do xcopy from an other JFS disk having saved on it my boot partitions to restore one of the new aligned partition.
I tried to boot on it but it failed... Normal, MBR boot partition information was lost and JFS bootloader removed.
I put back using DFSEE the partition as boot partition and reloaded the jfs loader.
Then, I could boot ArcaOS from the restored partition...
But above a GB value, none where operationnal.
A new look under Dfsee and this time:
L-Geo 976761-64-32
S-Geo 65535-64-32 = 65535MiB ! (a big amount of GiB missing to go up to 1TiB (1024GB)
Before, I had H255 and S63 (of course, sector size always left at default 512...)
But this resulted to unaligned partition while Dfsee found it aligned (ok for HDD) but it isn't for SSD.
For Dfsee display, it is possible to adjust CHS to get acurate information but it did not correct the disk, only virtually
Note: I checked my older SSD and found that all partitions where unaligned but it worked (badly with more write operations) and the fstrim worked under linux but this time, it will not due the partition can't be mounted.
All errors results from Dfsee after a disk clone operation and dfsee doesn't check nor does it warn when the destination
disk is a SSD with aligment requirements.... Disk clone rewrites MBR but it clones the disk uniqueid too which gives then a disk signature collision problem...
Well,
testdisk allow me to set CHS but its maximum S value is 63 (56 is not a nice value)
Now, my question is what will be maximum cylinder value for OS/2 LVM ? Which CHS value should I use ?
Now, trying again 124519-255-63 under testdisk not modifying my aligned partitions
aftet this I now see under dfsee:
L-Geo 976761-64-32
S-Geo 124519-255-63
under msinfo32, starting offset are left aligned...
It always show me partitions names and volume label.... (will now check how ArcaOS will react..)