TRIM on third party SSD

[Freyqq]

Ah but if we limited progress to only what we had before, we wouldn't progress.

I've been running 2 SSDs since Yosemite came out without Trim and no noticeable or measurable speed dropoff on either so far...

You won't notice much of a dropoff unless the SSD has a lot of writes on it since a TRIM was run

Anyway, I'd personally rather have TRIM than a security feature that didn't exist until 10.10. It's a shame I have to choose at all...Windows doesn't have this problem.

 

[PeterHolbrook]

It would be nice if someone came up with an application capable of writing a fake ID to an SSD drive, so that Yosemite would recognise it as trim-worthy, but, failing that, we don't really need to wait.

Naturally, there'll be people who say trim is unimportant, or that TrimEnabler, despite its pitfalls in Yosemite, is more than enough for their needs, etc. For those who want trim RIGHT NOW with no hacks, there's the choice of purchasing an Angelbird SSD, which is recognised as an APPLE SSD by Yosemite and trim is enabled by the OS itself without any hacks.

Naturally, there'll be those who say they don't want to purchase that make, because it isn't as well known as Samsung, Toshiba, et cetera and, in addition, is more expensive than some of those makes. But for those not so ready to pass judgment on what they haven't seen, the choice of optimal performance is really there.

 

[chrfr]

It would be nice if someone came up with an application capable of writing a fake ID to an SSD drive, so that Yosemite would recognise it as trim-worthy, but, failing that, we don't really need to wait.

I just discovered this morning that the latest 10.10.4 beta (14e26a) apparently enables trim for 3rd party disks. The Mac Mini I'm typing on has a Fusion drive I created with a Samsung 840 pro, and running fsck reported that it was trimming the disk, without any 3rd party trim enabler installed. The same happened with my 2012 Air which has an OWC SSD installed.

 

[mikeboss]

first we hear that Apple is focussing on stability with OS X 10.11 and now this hell froze over if true. installing 14E26a right now...

EDIT:
hell didn't freeze over. TRIM still not supported.

EDIT II:
started in single-user mode and ran "fsck -ffy" -> same outcome, no TRIM.

 

[PeterHolbrook]

I just discovered this morning that the latest 10.10.4 beta (14e26a) apparently enables trim for 3rd party disks. The Mac Mini I'm typing on has a Fusion drive I created with a Samsung 840 pro, and running fsck reported that it was trimming the disk, without any 3rd party trim enabler installed. The same happened with my 2012 Air which has an OWC SSD installed.

I haven't checked myself (I don't run operating system betas), but it seems to me you are confusing your wish with reality. It is likely that fsck will always display that "trim message", whether trim actually works or doesn't. As it doesn't work (because Yosemite kext signing disabled it), the message is misleading. Better go the Angelbird route.

 

[chrfr]

first we hear that Apple is focussing on stability with OS X 10.11 and now this hell froze over if true. installing 14E26a right now...

EDIT:
hell didn't freeze over. TRIM still not supported.

It's reporting as "not supported" for me in System Report, but did display "trimming unused blocks" which is new behavior in fsck.

----------

I haven't checked myself (I don't run operating system betas), but it seems to me you are confusing your wish with reality. It is likely that fsck will always display that "trim message", whether trim actually works or doesn't. As it doesn't work (because Yosemite kext signing disabled it), the message is misleading. Better go the Angelbird route.

Extension signing has nothing to do with why trim support doesn't work natively. I don't personally care whether or not trim works in 10.10.x as I do not use any 3rd party tools to enable it and have no performance issues with these SSDs, I just found it interesting that it reports that it's trimming unused blocks in 10.10.4. There's no way to tell what it's actually doing either way.

 

[PeterHolbrook]

Extension signing has nothing to do with why trim support doesn't work natively.

Sorry, but it has EVERYTHING to do with it. All TrimEnabler does is get rid of kext signing and patch a controller so that Yosemite will think that any SSD is manufactured by Apple, so that trim will work. The patching wouldn't be possible with kext signing on.

 

[jbarley]

It's reporting as "not supported" for me in System Report, but did display "trimming unused blocks" which is new behavior in fsck.

I have the latest Beta installed, and after reading your post booted into 'single user mode' and ran the 'fsck' command. I see no mention of trim in the generated report whasoever. Too bad because like others here I also got all excited.

 

[chrfr]

Sorry, but it has EVERYTHING to do with it. All TrimEnabler does is get rid of kext signing and patch a controller so that Yosemite will think that any SSD is manufactured by Apple, so that trim will work. The patching wouldn't be possible with kext signing on.

You're misunderstanding what happens. Apple could have written the built-in extension to allow trim on 3rd party SSDs, and signed that extension, but didn't. Extension signing is why the 3rd party tool like Trim Enabler doesn't work automatically but is not why Apple's native extension doesn't trim aftermarket SSDs.

As for why my computers are reporting that they're trimming blocks but others aren't... I'm mystified.

 

[PeterHolbrook]

You're misunderstanding what happens.

Not at all, sir. Period.

 

[DeltaMac]

Maybe it's only a limited list of SSDs that are now (maybe) supported.
I have the beta 10.10.4, with a Crucial MX100, latest firmware MU02.
I have not attempted in any way to enable trim on this system.
Profiler shows trim not supported, as I would expect.
Booted to single-user, ran fsck, and NO text showed any trim operation. Just a usual listing for a non-trim fsck, which doesn't list trim in any way. No change in this beta, at least for the Crucial SSD that I have.

 

[jbarley]

Maybe it's only a limited list of SSDs that are now (maybe) supported.
I have the beta 10.10.4, with a Crucial MX100, latest firmware MU02.
I have not attempted in any way to enable trim on this system.
Profiler shows trim not supported, as I would expect.
Booted to single-user, ran fsck, and NO text showed any trim operation. Just a usual listing for a non-trim fsck, which doesn't list trim in any way. No change in this beta, at least for the Crucial SSD that I have.

If this is the case then you can rule out the Samsung 840 evo's also.

Edit: my SSD is standalone, maybe this has something to do with "fusion" drives?

 

[chrfr]

This is the output from fsck on my 2012 Air with the OWC SSD. The SSD is still listed as " Trim Support: No" and I don't have a 3rd party trim enabler installed.

 

[sim667]

So I've been considering this today, and I like the idea of editing the firmware, so I'm going to be looking into seeing whether thats possible on the samsung evo 850

One thing I was wondering is whether something like using chameleon to enable trim, booting into single user and doing and fsck -fy command would trim the drive, and then rebooting into the ui and disabling chameleon would be a solution..... One thing for me is that I dj and sometimes have to restart my computer and very occasionally reset my ram on the fly (and after quite a lot of booze).....

i.e. could i use this as an occasional thing, like running a defrag or something? Or is trim purely a realtime process?

 

[Ethosik]

TRIM is not that important and it is definitely not worth the hassle. I never understand what the big fuss is about. I have been using SSDs for years on my Mac Pro and never enabled TRIM and I have never had any issues.

You won't notice a severe difference if you leave TRIM disabled.

 

[PeterHolbrook]

TRIM is not that important and it is definitely not worth the hassle. I never understand what the big fuss is about. I have been using SSDs for years on my Mac Pro and never enabled TRIM and I have never had any issues.

You won't notice a severe difference if you leave TRIM disabled.

Naturally, there'll be people who say trim is unimportant, or that TrimEnabler, despite its pitfalls in Yosemite, is more than enough for their needs, etc.

Quod erat demonstrandum.

 

[rbart]

So, to summarize, are there any news regarding TRIM in 10.10.4 ?
When you boot in single mode an run fsck -fy is it different than in 10.10.3 ?
Because if now we can manually trim from time to time, maybe trim enabler becomes less useful ?

 

[PeterHolbrook]

So, to summarize, are there any news regarding TRIM in 10.10.4 ?

Unless you consider as news-worthy the fact that a probably misleading line appears in the output of fsck -fy, run in single mode, on fusion disk created with an otherwise non-trim-compliant SSD in Yosemite, I don't think there is any news at all.

Even if the trim command were to work in such a situation (which is very doubtful), rebooting a Mac into single mode from time to time in order to run fsck -fy is impractical. No production machine can be used in such a way.

 

[SlCKB0Y]

I think it was optional in mav, default in yosumfinky, nonexistent before, and probably mandatory on the next release or the one after.

Mandatory? As in can't be disabled? Won't happen...ever.

----------

No production machine can be used in such a way.

The same could be said of modifying storage drivers and breaking their signing.

 

[rbart]

For my usage, it's acceptable to boot every month in single mode and run the command if it avoids to hack the system ...

 

[PeterHolbrook]

The same could be said of modifying storage drivers and breaking their signing.

I didn't say otherwise. I'm not advocating the use of TrimEnabler AT ALL.

 

[definitive]

To the people who are saying that TRIM works great on their system ... Could you please elaborate? It's like saying snake oil make everything better... There's no actual proof being posted, so I hope you understand how it makes it sound when you post something like that.

To those thinking editing firmware to trick the system into thinking that it's a genuine Apple SSD: be careful. If you don't know what you're doing, you can brick the drive.

 

[PeterHolbrook]

To the people who are saying that TRIM works great on their system ... Could you please elaborate? It's like saying snake oil make everything better... There's no actual proof being posted, so I hope you understand how it makes it sound when you post something like that.

I don't have an SSD, so I'm not enabling trim or anything like that, but your claim that no proof has been posted is far from accurate. There's plenty of material about this issue. Have a look, for instance, at https://www.youtube.com/watch?v=sK3iTz5C46c. Notice the progressive loss of write speed when trim is disabled. Of course, you could as well say that you can live with that, et cetera. Summing it up, do as you wish.

I've made up my mind. Unless Apple change their trim policy for 3rd party SSDs, I'll buy an Angelbird SSD when the time comes. For the time being, I'm perfectly happy with a Seagate SSHD running Yosemite 10.10.3 on my Mac Pro 1,1.

Edit: In any case, the degradation effect of "trirmless" SSDs on Macs, possibly due to the nature of HFS, is less dramatic than what would happen in Windows without trim.

 

[b0fh666]

Mandatory? As in can't be disabled? Won't happen...ever.

----------

never say never

it's a matter of time until we need to jailbreak our macs

 

[definitive]

I don't have an SSD, so I'm not enabling trim or anything like that, but your claim that no proof has been posted is far from accurate. There's plenty of material about this issue. Have a look, for instance, at https://www.youtube.com/watch?v=sK3iTz5C46c. Notice the progressive loss of write speed when trim is disabled. Of course, you could as well say that you can live with that, et cetera. Summing it up, do as you wish.

I've made up my mind. Unless Apple change their trim policy for 3rd party SSDs, I'll buy an Angelbird SSD when the time comes. For the time being, I'm perfectly happy with a Seagate SSHD running Yosemite 10.10.3 on my Mac Pro 1,1.

That video did not prove anything. It was a benchmark with TRIM on and off. It didn't show two drives (one with TRIM on, and one with TRIM off) side by side, and how their performance fairs after a year of use... The very minor differences in benchmarks are just that - minor differences. It was a single benchmark, and not multiples which were averaged and compared. Even the description states that the difference is minor, if any.