Apple Macbook Pro SSD Controller
Brennan Shacklett
SM1024F, but which is apparently manufactured by Samsung. The drive
works great in OSX; however, my primary OS is linux, and as shown in
this kernel bug: https://bugzilla.kernel.org/show_bug.cgi?id=60731
there are numerous issues with the device. The _GTF for the drive
appears to be a non standard length, NCQ fails, the drive stop command
fails, and the drive suffers from terrible performance in some
workloads.
This all points to Apple and Samsung using some sort of non standard
AHCI interface, but there appears to be zero publicly available
documentation on the matter.
Could anyone on this list point me to some documentation or shed some
light on how OSX initializes and uses this drive properly?
Thanks,
Brennan
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Ata-scsi-dev mailing list (Ata-sc...@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
https://lists.apple.com/mailman/options/ata-scsi-dev/ata-scsi-dev-garchive-72467%40googlegroups.com
This email sent to ata-scsi-dev-...@googlegroups.com
Chris Sarcone
My MBP 11,3 has a SSD which System Profiler identifies as an APPLE SSD
SM1024F, but which is apparently manufactured by Samsung. The drive
works great in OSX; however, my primary OS is linux, and as shown in
this kernel bug: https://bugzilla.kernel.org/show_bug.cgi?id=60731
there are numerous issues with the device. The _GTF for the drive
appears to be a non standard length, NCQ fails, the drive stop command
fails, and the drive suffers from terrible performance in some
workloads.
This all points to Apple and Samsung using some sort of non standard
AHCI interface, but there appears to be zero publicly available
documentation on the matter.
Could anyone on this list point me to some documentation or shed some
light on how OSX initializes and uses this drive properly?
------------------
6 Infinite Loop
M/S 306-2MS
Cupertino CA 95014
phone: (408) 974-4033
fax: (408) 862-7577
email: sar...@apple.com
Brennan Shacklett
someone from Apple took the time to respond to a Linux issue.
> The official way to report bugs is through the http://bugreport.apple.com
> website. I encourage you to file bugs with as much relevant information as
> possible so Apple engineers can investigate the issue.
I didn't report it there because it is an issue with using the device
under Linux which I figured Apple engineers wouldn't be interested in.
Should I report the fact that the device doesn't work properly under
Linux and link the kernel bug, or do I need to to find out more
specific information?
I'm sort of in a time crunch because I need this computer for
development in a week, so the prospect of the issue disappearing into
the possible black hole of a bugreporter is a tad frightening.
> It initializes the drive the same way it does any other drive. Typical
> commands include IDENTIFY DEVICE and SET FEATURES.
So Linux calls SET FEATURES on the drive using the commands from the
acpi _GTF (which linux believes is a length of 8 which doesn't exactly
fit with the standard 7 registers) Regardless if the length 8 issue is
bypassed the drive winds up rejecting the SET FEATURES command which
isn't surprising because one of the features it sets is SATA DISABLE
(0x90) which seems odd (this is why I thought something non standard
was going on).
Is there any way to get the commands which Apple sends to the drive to
compare them with what linux is doing? Or is it possible that
something specialized needs to happen with initializing the PCI device
itself?
Thank you again for your time,
--Brennan