Difference between revisions of "Ultrabay Slim SATA HDD Adapter"

From ThinkWiki
Jump to: navigation, search
(Add warning regarding poor performance with some 3rd-party adapters)
(Poor performance with 3rd-party adapters: Actually, you *can* get UDMA5 out of a 3rd-party adapter.)
Line 37: Line 37:
 
*ThinkPad {{T40}}, {{T40p}}, {{T41}}, {{T41p}}, {{T42}}, {{T42p}}, {{T43}}, {{T43p}}
 
*ThinkPad {{T40}}, {{T40p}}, {{T41}}, {{T41p}}, {{T42}}, {{T42p}}, {{T43}}, {{T43p}}
  
=== Poor performance with 3rd-party adapters ===
+
=== Kernel Issues with 3rd-party adapters ===
Many (most?) of the SATA adapters listed on eBay are 3rd-party clones, whose poor build quality limits them to UDMA/33 speeds. They are therefore slower than the common IDE adapters, and especially unsuitable for SSD use.
+
Many (most?) of the SATA adapters listed on eBay are 3rd-party clones, and the kernel will often limit them to UDMA/33 speeds, with a maximum of 33.3 MB/s.  They are thereby rendered slower than the common IDE adapters.
  
 
A kernel log showing such an adapter being limited to UDMA/33:
 
A kernel log showing such an adapter being limited to UDMA/33:
Line 48: Line 48:
 
[   55.650676] ata2: EH complete   
 
[   55.650676] ata2: EH complete   
 
</pre>
 
</pre>
 +
 +
However, this problem can possibly be overcome by adding the boot parameter {{bootparm|libata.force|X:80c}} (where X is the IDE line which is being affected by this UDMA problem).
 +
 +
A kernel log after using the {{bootparm|libata.force|X:80c}} bootparm:
 +
<pre>
 +
[ 1863.865187] ata2: ACPI event
 +
[ 1865.191117] ata2: soft resetting link
 +
[ 1865.345444] ata2: FORCE: cable set to 80c
 +
[ 1865.345729] ata2.00: ATA-7: TOSHIBA MK6034GSX, AH101A, max UDMA/100
 +
[ 1865.345741] ata2.00: 117210240 sectors, multi 16: LBA48 NCQ (depth 0/32)
 +
[ 1865.349602] ata2.00: configured for UDMA/100
 +
[ 1865.349629] ata2: EH complete
 +
</pre>
 +
 +
'''NOTE:''' Using "libata.force=X:80c" is oft referred to as a "work-around"--the better solution would be to patch the kernel, e.g., [http://bugs.launchpad.net/ubuntu/+bug/195221 Ubuntu Bug #195221].  Then again, this problem has been around at least since [http://www.mail-archive.com/linux-ide@vger.kernel.org/msg14622.html January of 2008]....
 +
 +
  
 
[[Category:Components]]
 
[[Category:Components]]

Revision as of 16:22, 16 January 2012

Ultrabay Slim SATA HDD Adapter

This is an adapter to connect an additional hard disk drive to the Ultrabay Slim.

Features

  • Takes any ThinkPad SATA 9.5mm thick, 2.5" HDD

Ultrabay Slim SATA HDD adapter

NOTE!
This SATA HDD adapter only works with select ThinkPads as listed below, for other machines with UltraBay Slim, use the PATA base UltraBay Slim HDD Adapter instead.

Part numbers

Marketing PN: 40Y8725 FRU PN: 26R9246

Linux support

The adapter contains a PATA to SATA bridge chip, which means that although the disk is SATA it connects as a PATA (IDE) device.

If connected at boot time, the drive is normally located at /dev/sdb

Supported with

Unsupported use with

The adapter has a black plastic grommet at the back that restricts its use to the 60 series ThinkPads. Some people have successfully removed this grommet with a pair of channel lock pliers, or sliced it off with a Dremel after which it will fit into older ThinkPads like the T40 series. This works because the adapter actually bridges the SATA HDD to PATA as those older machines have no SATA support.

This has been tested with the following ThinkPads, although this is obviously not supported!

Kernel Issues with 3rd-party adapters

Many (most?) of the SATA adapters listed on eBay are 3rd-party clones, and the kernel will often limit them to UDMA/33 speeds, with a maximum of 33.3 MB/s. They are thereby rendered slower than the common IDE adapters.

A kernel log showing such an adapter being limited to UDMA/33:

[   55.610787] ata2.00: ATA-7: FUJITSU MHV2020BH, 0093002C, max UDMA/100                          
[   55.610793] ata2.00: 39070080 sectors, multi 16: LBA48 NCQ (depth 0/32)                        
[   55.610801] ata2.00: limited to UDMA/33 due to 40-wire cable                                   
[   55.650666] ata2.00: configured for UDMA/33                                                    
[   55.650676] ata2: EH complete  

However, this problem can possibly be overcome by adding the boot parameter libata.force=X:80c (where X is the IDE line which is being affected by this UDMA problem).

A kernel log after using the libata.force=X:80c bootparm:

[ 1863.865187] ata2: ACPI event
[ 1865.191117] ata2: soft resetting link
[ 1865.345444] ata2: FORCE: cable set to 80c
[ 1865.345729] ata2.00: ATA-7: TOSHIBA MK6034GSX, AH101A, max UDMA/100
[ 1865.345741] ata2.00: 117210240 sectors, multi 16: LBA48 NCQ (depth 0/32)
[ 1865.349602] ata2.00: configured for UDMA/100
[ 1865.349629] ata2: EH complete

NOTE: Using "libata.force=X:80c" is oft referred to as a "work-around"--the better solution would be to patch the kernel, e.g., Ubuntu Bug #195221. Then again, this problem has been around at least since January of 2008....