Mac OS X 10.10.4 Supports TRIM for Third-Party SSD Hard Drives

[Taz Mangus]

OK, I am a bit clueless on all this, so do forgive if this is a naive question. I have a crucial SSD installed in the CD bay of my 2011 MacBoom Pro (m550). I have been using Trim Enabler 3.3 to this point. Is there any improvement to be gained from switching now over to the Apple Trimforce function? Note: I just updated to 10.10.4 and the Terminal command listed here does not seem to work. It returns "function not found"... Is it possible some versions of these SSDs are not recognized by Trimforce?

Did you enter: sudo trimforce enable? I would disable trim enabler before enabling trimforce. The advantage of trimforce vs trim enabler is that trim enabler modifies a system kext file and trimforce is now a native OS X command which can be left on without causing an issue when an OS X update is done.

 

[b_scott]

how would this work differently than Trim Enabler?

 

[Reason077]

how would this work differently than Trim Enabler?

The end result is exactly the same. It tells the OS the it's ok to send TRIM commands to your SSD.

The difference is that where Trim Enabler modified a system kernel extension to do so (which in recent OS X versions, means disabling kext signing, making your system slightly less secure), the "trimforce" command is built in and uses a new API that Apple have added.

 

[Reason077]

if you also have a Crucial SSD, it's worth checking your firmware and seeing if there is a more recent version that fixes trim handling before enabling trim.

The bug is with the NCQ variant of the TRIM command, which is different to normal TRIM.

On a Mac, there's no reason to worry about enabling TRIM on drives that may have bugs with NCQ TRIM. Linux is the only OS that supports the NCQ TRIM command, which is a new feature added very recently in the ATA spec. Mac OS X and Windows only use the standard, non-queued TRIM command.

The reason so many SSD firmwares have faulty NCQ TRIM implementations is precisely because most OSes don't support it and thus their firmware updates to enable it haven't been properly tested!

 

[Reason077]

Not gonna do it.

Any good SSD has enough overprovisioning that TRIM shouldn't make much of a speed difference.

Wrong. If you think that overprovisioning has anything much to do with TRIM then you don't understand how SSDs work.

Plus, my root partition is so close to full, it REALLY wouldn't make any difference (TRIM's performance impact increases with the amount of free space in your file system). There MIGHT be a slight impact on longevity, but how much longer am I going to have this 17" MacBook Pro as my primary system anyhow?

Performance improvements from TRIM have little to do with how much free space you have. If you have, say, 1GB free on your SSD and you write a 1GB file to it, then that write will be always be much faster if the space has been previously TRIMed.

TRIM also has little (or nothing) to do with SSD longevity. You're subjecting the NAND to the same number of erase-write cycles regardless of whether you use TRIM or not. The difference is that the slowest-by-far part of the cycle (ERASE) can be performed in advance if you use TRIM.

(overprovisioning, on the other hand, does help with longevity, by ensuring the same physical NAND cells aren't constantly being used for a file that is overwritten very frequently, for example)

I could cross-reference the exact model of my Crucial SSD against white lists and black lists, but I don't trust it. New devices get added to the blacklist all the time. It's not worth the risk for the extremely minor benefit.

The blacklists in Linux are for NCQ TRIM. Normal non-NCQ TRIM works fine in any modern drive. Mac OS X doesn't use NCQ TRIM.

Oh, and on my Linux box, I tried enabling TRIM, and it actually SLOWED DOWN. That's because it doesn't support queued TRIM, so the TRIM commands really hurt efficiency.

Actually Linux is the only OS that supports queued TRIM. But many drives (e.g. Samsung, Sandisk, Crucial, and others) have bugs in their NCQ TRIM implementations, hence the blacklist.

 

[olletsocmit]

"sudo trimforce enable" turns TRIM on, what is the command to turn it off? Apparently today TRIM Enabler app is coming out with an update to enable TRIM on 10.10.4 but will not disable kext signing so i thought it may be better to enable trim via the app.

 

[Reason077]

"sudo trimforce enable" turns TRIM on, what is the command to turn it off?

sudo trimforce disable

 

[Superhai]

so i thought it may be better to enable trim via the app.

Why would you think so? Why keep a third-party app that does exact same thing as the built-in solution?

 

[msh]

The fsck -fy command after enabling trim didn't change performance but found 5GB additional space on my 120GB drive. I am impressed.

 

[chrfr]

The fsck -fy command after enabling trim didn't change performance but found 5GB additional space on my 120GB drive. I am impressed.

That just means you had filesystem corruption. It's not related to trim.

 

[msh]

That just means you had filesystem corruption. It's not related to trim.

Ok. So not impressed. I thought diskutility would take care of corrupted files but apparently you have to run fsck. I will run that periodically now.

 

[scrmtrey]

I enabled trim via apple trimforce on 850EVO and it is fine.

 

[chrfr]

Ok. So not impressed. I thought diskutility would take care of corrupted files but apparently you have to run fsck. I will run that periodically now.

Disk Utility will, but it may also be that you freed up some swap files with the reboot. There's lots of possibilities why you got more space, but none of them are due to trim.

 

[Cindori]

Disk Sensei now offers safe Trim enabling:

https://www.cindori.org/safely-enable-trim-on-yosemite-and-el-capitan/

 

[JD575]

Is there a way to determine if its working or enabled? I ran the command and it said it was successful, but I use a PCIe card in a RAID 0 config. My Mac only show the RAID status, not individual drives.

 

[yukyuklee]

running fine on my mac mini with crucial mx100

 

[allanfries]

After doing a bit of research, etc, I decided to enable this feature. So last night I shut off my current Trim Enabler by Oskar Groth (Cindori.org), and enabled the built in. Haven't really noticed any difference with exception of apps opening a touch slower. Seems better this morning. So far no major issues and here's fingers crossed.

 

[chrfr]

Haven't really noticed any difference with exception of apps opening a touch slower.

This isn't something that switching from Trim Enabler to trimforce could possibly cause.

 

[0x0x0x0]

The number of people that post here who don't understand how SSDs work or what TRIM does is absolutely startling. There's no reason why TRIM would work worse with non-Apple-branded SSD than with an Apple-branded SSD (note the -branded part). TRIM is nothing "fancy" it is just an ATA command that tells SSD that a particular LBA no longer stores "useful" data, and depending on what brand you have, different algorithms deal with such markings. Off the top of my head, I can think of at least two possibilities: erase the LBA now, or, more efficiently, leave the erasing till the next round of GC…

My suspicion is that this trim enabler has nothing to do with "support" but more with some lawyer at Apple suddenly realising that Apple was walking into an anti-trust suit or an abuse of monopoly investigation…

 

[yukyuklee]

The number of people that post here who don't understand how SSDs work or what TRIM does is absolutely startling. There's no reason why TRIM would work worse with non-Apple-branded SSD than with an Apple-branded SSD (note the -branded part). TRIM is nothing "fancy" it is just an ATA command that tells SSD that a particular LBA no longer stores "useful" data, and depending on what brand you have, different algorithms deal with such markings. Off the top of my head, I can think of at least two possibilities: erase the LBA now, or, more efficiently, leave the erasing till the next round of GC…

My suspicion is that this trim enabler has nothing to do with "support" but more with some lawyer at Apple suddenly realising that Apple was walking into an anti-trust suit or an abuse of monopoly investigation…

well said! just like "Fusion Drive"

 

[avkdm]

Is there a way to determine if its working or enabled? I ran the command and it said it was successful, but I use a PCIe card in a RAID 0 config. My Mac only show the RAID status, not individual drives.

You could try this. Bit clunky but may help.
Install Trim enabler but do not activate.
Trim Enabler will give you this if its activated....
"The patch is not active, but Trim is working. You probably have an Apple SSD, which supports Trim by default."

 

[allanfries]

You could try this. Bit clunky but may help.
Install Trim enabler but do not activate.
Trim Enabler will give you this if its activated....
"The patch is not active, but Trim is working. You probably have an Apple SSD, which supports Trim by default."[/QUO

Yes I've noticed this.

 

[Superhai]

Is there a way to determine if its working or enabled?

TRIM is only for SATA SSD drives, if you are using non-ATA protocols they use other means. You can check by using this in Terminal:

system_profiler SPSerialATADataType SPParallelATADataType

If you find "TRIM Support: Yes" under one of your drives and the drive is not named "APPLE SSD" it is enabled and working. If it is saying "TRIM Support: No" then it is not.

 

[Ledgem]

The bug is with the NCQ variant of the TRIM command, which is different to normal TRIM.

On a Mac, there's no reason to worry about enabling TRIM on drives that may have bugs with NCQ TRIM. Linux is the only OS that supports the NCQ TRIM command, which is a new feature added very recently in the ATA spec. Mac OS X and Windows only use the standard, non-queued TRIM command.

According to the article I linked in my initial post, the problem was with standard TRIM, not queued TRIM.

To those who are saying "I haven't had any problems," the concern is that the data corruption would be silent. The operating system would still have files marked as being in the location, but the SSD would incorrectly use the TRIM command to zero out blocks in a location occupying legitimate data. You wouldn't find out that the data was gone until you tried to access it and discovered that the files were corrupt because the data was missing.

Proceed at your own risk.

 

[0x0x0x0]

According to the article I linked in my initial post, the problem was with standard TRIM, not queued TRIM.

…

Proceed at your own risk.

That seems to be the first and only time I ever heard of a claim that the SSD itself is marking wrong LBAs as free. Go figure, bearing in mind that non-Apple-branded SSD get "tested" by way more people than the Apple-branded ones…