[Info-vax] VMS and the (lack of the) TRIM facility.

[JF Mezei]

On 15-06-21 11:27, Stephen Hoffman wrote:

> In years past, OpenVMS device drivers maintain (very messy) lists of 
> quirks as in-line code, usually with comments, and OpenVMS never got 
> around to having any particular list of or database of those quirks.   
> For rather fewer devices than Linux and Windows support, too.

When VMS boots, the driver interrogates each disk device to see what
they can cand can't do, right ? (for instance what level of SCSI they
support).  If that is correct, then the driver can simply ask if the
device supports TRIM and keep a flag for that device of whether to
ingnore TRIM  IO requests and pass them onto the device.

The actual logic of issuing TRIM requests will lie within the code that
deletes files. The disk driver doesn't know about dick blocks becoming free.


> 
> Apple has gotten around to adding TRIM with their upcoming release,

Nop. TRIM has been around on OS-X for a number of years. But Apple
limited it to only Apple installed SSDs, did not enable it for others,
and one needed a "hack" to tell the driver to accept the non-Apple SSD.
(which is what I have on my laptop).

What is happening with latest release is that Apple provides a tool to
enable TRIM on non-Apple devices, with the warning that Apple hasn't
tested that drive and can't garantee anything.

> Do I think that TRIM should be available?  Yes.  Do I think it'll be 
> simple or quick or possibly even ever available on Itanium — other than 
> the existing HP SSDs — at all?  No.

If VMS doesn't support TRIM today, then those HP SSDs are being treated
as rust coated spinning platters, with write amplification and all other
warts.