Oh, it's so great how its broken under Linux; my 840 EVO was bought with the explicit intention of trickling down to be the Fedora disk next upgrade time, replacing the Kingston V100. Sigh.
Anyway,
accordingto Martin K Petersen at Oracle,
Yes, this confirms that the old firmware did not support queued
TRIM. Whereas the new one does.
I'll try to reach out to my contacts at Samsung again so we can get this
resolved.
Until then, I have sent out a patch that blacklists queued TRIM for all
Samsung 800-series drives. It appears to be generic to their
implementation and not tied to a particular drive model.
Patch can be seen
hereThe issue was seemingly introduced in the
850 PRO f/w update a few months ago. The one they pulled from distribution and have not superceded. Despite which, they've introduced the same TRIM issue into the new 840 EVO f/w now; wonder if they used the same code?
There does seem to be an available
test to verify disk trim functionality on Win 8 "if the device includes ATA non-NCQ Trim support" (which is how Linux needs to run the 840 EVO current f/w and 850 PRO pulled f/w), but I don't have Win 8 on a box with a testable drive.
There's also a
test which "verifies that a SATA device that supports Trim performs according to the certification requirements" but
This test may trim or overwrite any or all of the sectors on the test device. As a result, the data on the drive will be lost. This test must be run against a secondary SATA device running in AHCI mode
so nope, not this week.