Skip to contentSkip to conversation

Serial ATA (SATA) status report

DISCUSSION Jeff Garzik 8 participants 13 messages Feb 26, 2004 – Mar 1, 2004

lore
Share & tools
lore
#0 lore raw

attached.

Serial ATA (SATA) for Linux
status report
Feb 25, 2004

Hardware support
================

Intel ICH5 and ICH5-R
---------------------
Summary: No TCQ. Looks like a PATA controller, but with a few added, non-standard SATA port controls. Hardware does not support hotplug. "Coldplug" support is potentially feasible.

libata driver status: Production, but see issue #2, #3, #4.

drivers/ide driver status: Production, but see issue #1, #2.

Issue #1: Depending on BIOS settings, IDE driver may lock up computer when probing drives.

Issue #2: Excessive interrupts are seen in some configurations.

Issue #3: "Enhanced mode" or "SATA-only mode" may need to be set in BIOS.

Issue #4: Driver disables SATA port on unload, but does not re-enable SATA port on load.

Intel ICH6 ("AHCI")
-------------------
Summary: Per-device queues, full SATA control including hotplug and PM.

libata driver status: "looks like ICH5" support available in ata_piix very soon. Full support available in a week or two.

Promise TX2/TX4/SX4
-------------------
Summary: Per-host queues on all controllers. Full SATA control including hotplug and PM on all but one controller.

libata driver status: Beta, but see issue #5.

Issue #5: Some boards appear to have PATA as well as SATA ports. PATA is not currently supported, and no plans have yet been made to rectify this. Ideally drivers/ide would drive PATA, but if they are the same PCI device, that would not be feasible.

Silicon Image 3112/3114
-----------------------
Summary: No TCQ. Looks like a PATA controller, but with full SATA control including hotplug and PM.

libata driver status: Alpha.

drivers/ide driver status: Production, but see issue #4.

Issue #4: Need to have the most recent fixes posted to lkml, for stable operation and full performance (where possible).

Silicon Image 3124
------------------
Soon, hopefully. User-written.

Broadcom/ServerWorks/Apple
--------------------------
Summary: Huge per-device queues, full SATA control including hotplug and PM.

libata driver status: Beta.

VIA
---
Summary: No TCQ. Looks like a PATA controller, but with full SATA control including hotplug and PM.

libata driver status: Beta.

SiS
---

libata driver status: in progress (user-written)

Vitesse
-------

libata driver status: in progress

Marvell
-------
libata driver status: in progress

Software support
================

Basic Serial ATA support
------------------------
The "ATA host state machine", the core of the entire driver, is considered production-stable.

The error handling is _very_ simple, but at this stage that is an advantage. Error handling code anywhere is inevitably both complex and sorely under-tested. libata error handling is intentionally simple.

Positives:  Easy to review and verify correctness.  Never data
corruption.  Negatives:  if an error occurs, libata will simply send

the error back the block layer. There are limited retries by the block layer, depending on the type of error, but there is never a bus reset.

Or in other words: "it's better to stop talking to the disk than compound existing problems with further problems."

As Serial ATA matures, and host- and device-side errata become apparent, the error handling will be slowly refined. I am planning to work with a few (kind!) disk vendors, to obtain special drives/firmwares that allow me to inject faults, and otherwise exercise error handling code.

Queueing support
----------------
Even though some SATA host controllers on the market already support command queueing (a.k.a. "TCQ"), libata does not yet support it.

However, libata was designed from the ground-up to support queueing, so I need only change a few lines of code, and write two functions, to enable this behavior.

Queueing will be enabled in libata soon, but to do so requires a long stretch of testing on a large variety of controllers and drives. This is very time-intensive, and is the largest part of this task.

Tangent: Host-based queueing and Native Command Queueing

Queueing is the process of sending multiple commands to a single device, without waiting for prior commands to finish. This increases performance and reduces latency. There are three types of queueing in the ATA world:

1) "legacy TCQ" -- some PATA devices support this. Just ignore it, it's going away.

2) "host-based TCQ" -- the host controller supports a queue of drive commands, whether or not the drive supports it.

3) "Native Command Queueing" -- both host and drive cooperate in the queueing and execution of drive commands. This should provide the highest performance and lowest latency of all three options.

#1 is support by drivers/ide _only_. libata will not support this. #2 will soon be supported by libata.
#3 will be supported by libata when hardware is available from drive manufacturers.

Hotplug support
---------------
All SATA is hotplug.

libata does not support hotplug... yet.

Power Management support
------------------------
Over and above the power management specified in the ATA/ATAPI specification, one can aggressively control the power consumption of SATA hosts, the SATA bus, and the SATA device.

SMART support
-------------
Soon. Requires the capability to directly submit ATA commands from userspace to the low-level device, which must be added with care.

#1 lore raw
Quoted 6 lines · Jeff Garzik

VIA
---
Summary: No TCQ. Looks like a PATA controller, but with full SATA control including hotplug and PM.

libata driver status: Beta.

Does drivers/ide support this also?

#2 lore raw
Quoted 11 lines · Mike Fedyk
Quoted 6 lines · Jeff Garzik

VIA
---
Summary: No TCQ. Looks like a PATA controller, but with full SATA control including hotplug and PM.

libata driver status: Beta.

Does drivers/ide support this also?

Yes. Glad you mentioned that...

Jeff

#3 lore raw
In reply to Jeff Garzik

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 1 line

attached.

Which one of these chips are the 3Ware cards based on?

Mark.

- --
Mark Watts
Senior Systems Engineer
QinetiQ TIM
St Andrews Road, Malvern
GPG Public Key ID: 455420ED

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)

iD8DBQFAPbirBn4EFUVUIO0RAj2FAKCxOV+tWN99yqtHiCB8lCpV4cdvhQCgknbq fJ8nwJTyCX1MiuhhNsu2bW0=
=vEVb
-----END PGP SIGNATURE-----

#4 lore raw
Quoted 1 line · Mark Watts

Which one of these chips are the 3Ware cards based on?

None of them. 3ware has its own chip, supported by the 3w-xxxx driver in mainline 2.4 and 2.6. It's basicly exports the logical arrays as SCSI devices.

#5 lore raw

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 6 lines
Quoted 1 line · Mark Watts

Which one of these chips are the 3Ware cards based on?

None of them. 3ware has its own chip, supported by the 3w-xxxx driver in mainline 2.4 and 2.6. It's basicly exports the logical arrays as SCSI devices.

Neat. Are there any known issues with these cards? (Do they work with AMD-64?)

- --
Mark Watts
Senior Systems Engineer
QinetiQ TIM
St Andrews Road, Malvern
GPG Public Key ID: 455420ED

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)

iD8DBQFAPgF8Bn4EFUVUIO0RAg/JAJ9L5vNQVaoW37ElAc+OIPOJUusvCgCg4YrA b2+lYzV5MVt820JH+PLD3Vs=
=ycsm
-----END PGP SIGNATURE-----

#6 lore raw
Quoted 8 lines · Mark Watts
Quoted 3 lines · Andre Tomt

None of them. 3ware has its own chip, supported by the 3w-xxxx driver in mainline 2.4 and 2.6. It's basicly exports the logical arrays as SCSI devices.

Neat. Are there any known issues with these cards? (Do they work with AMD-64?)

the 3w-xxxx-module works well enough in 32bit mode on AMD64. sadly enough, we have had some other issues with 64bit mode, but the _driver_ seems to load there as well.

Signature · 1 line
Terje
#10 lore raw
In reply to Mark Watts
Quoted 9 lines · Mark Watts

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 1 line

attached.

Which one of these chips are the 3Ware cards based on?

None of the above. All of these chips are non-RAID. I suppose I should list stuff like 3ware as well...

Jeff

#7 lore raw
In reply to Terje Kvernes

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 3 lines

the 3w-xxxx-module works well enough in 32bit mode on AMD64. sadly enough, we have had some other issues with 64bit mode, but the _driver_ seems to load there as well.

Do these 'issues' prevent the cards from being used at all in 64bit mode on AMD-64?

We'd really like to use the 4-port SATA 3Ware card on a Tyan Thunder K8W (S2885) but it'd be a bit of a waste if we can only use it in 32bit mode. (I assume 32bit mode means you compile for i686 instead of AMD-64 ?)

Mark.

- --
Mark Watts
Senior Systems Engineer
QinetiQ TIM
St Andrews Road, Malvern
GPG Public Key ID: 455420ED

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)

iD8DBQFAPhcCBn4EFUVUIO0RAgYlAKCIdWX36mkFko8kdlyhCtzTcpAgsQCeOCc4 hZYOYnuFKuZLa5v4KJ7XGjI=
=M/Ej
-----END PGP SIGNATURE-----

#8 lore raw
Quoted 6 lines · Mark Watts
Quoted 3 lines

the 3w-xxxx-module works well enough in 32bit mode on AMD64. sadly enough, we have had some other issues with 64bit mode, but the _driver_ seems to load there as well.

Do these 'issues' prevent the cards from being used at all in 64bit mode on AMD-64?

our issues are mostly related to getting X up and running. we've been looking at keeping most of the userland as 32bit, but still running a 64bit kernel. this works fine, except for the nVidia kernel module and its attempt to talk to nVidias openGL libraries. it's a very long story. but, as far as I remember, the 3ware card worked fine.

Quoted 4 lines

We'd really like to use the 4-port SATA 3Ware card on a Tyan Thunder K8W (S2885) but it'd be a bit of a waste if we can only use it in 32bit mode. (I assume 32bit mode means you compile for i686 instead of AMD-64 ?)

correct, the kernels on our AMD64-boxes are currently built for i686. I _might_ be able to use one of our FX-51-boxes to test 64bit usage of a 3ware PATA controller if you feel you can't take the slight gamble. I'm not quite sure _when_ I can get this done though, maybe within a week or two? but, since you really need to test SATA as well, I'd advice you to get a card and test it yourself. I'd be very surprised if it didn't work.

Signature · 1 line
Terje
#9 lore raw
In reply to Mark Watts
Quoted 9 lines · Mark Watts
Quoted 6 lines
Quoted 1 line · Mark Watts

Which one of these chips are the 3Ware cards based on?

None of them. 3ware has its own chip, supported by the 3w-xxxx driver in mainline 2.4 and 2.6. It's basicly exports the logical arrays as SCSI devices.

Neat. Are there any known issues with these cards? (Do they work with AMD-64?)

We have build Raid boxes with 3ware on x86_64. We have not had to run a 32bit OS to get it working and it works fine. There were some issues with more than 4GB of memory but I believe 3ware has resolved them now.

Signature · 5 lines
Mark Lane, CET mailto:[email protected] 
Hard Data Ltd. http://www.harddata.com 
T: 01-780-456-9771   F: 01-780-456-9772
11060 - 166 Avenue Edmonton, AB, Canada, T5X 1Y3
--> Ask me about our Excellent 1U Systems! <--
#11 lore raw
In reply to Jeff Garzik
Quoted 11 lines · Jeff Garzik

Silicon Image 3112/3114
-----------------------
Summary: No TCQ. Looks like a PATA controller, but with full SATA control including hotplug and PM.

libata driver status: Alpha.

drivers/ide driver status: Production, but see issue #4.

Issue #4: Need to have the most recent fixes posted to lkml, for stable operation and full performance (where possible).

Having had problems with SiI 3112A under the standard IDE drivers
http://marc.theaimsgroup.com/?l=linux-kernel&m=107539618032329&w=2
I thought I would test my set of disks with some bad sectors that oops or hang the standard driver. I also tested them under high load (which also causes problems with the legacy drivers which can drop to PIO mode or just hang).

libata (2.4.25-libata3) is much more stable, managing 22 hours without problems, however has just hung with
ata3: DMA timeout, stat 0x1
as the last log message (nothing before), box completely unresponsive. Will recompile with magic sysrq and with DPRINTK and VPRINTK defined. It wasnt actually accessing a part of the disk with bad sectors at the point, just under high load and accessing 3 disks. What is the best way of debugging this?

SCSI subsystem driver Revision: 1.00
libata version 1.01 loaded.
sata_sil version 0.53
ata1: SATA max UDMA/133 cmd 0xF8816080 ctl 0xF881608A bmdma 0xF8816000 irq 26
ata2: SATA max UDMA/133 cmd 0xF88160C0 ctl 0xF88160CA bmdma 0xF8816008 irq 26
ata1: dev 0 cfg 49:2f00 82:7c6b 83:7b09 84:4003 85:7c49 86:3a01 87:4003 88:207f
ata1: dev 0 ATA, max UDMA/133, 240121728 sectors
ata1: dev 0 configured for UDMA/133
ata2: no device found (phy stat 00000000)
ata2: thread exiting
ata3: SATA max UDMA/133 cmd 0xF8818480 ctl 0xF881848A bmdma 0xF8818400 irq 27
ata4: SATA max UDMA/133 cmd 0xF88184C0 ctl 0xF88184CA bmdma 0xF8818408 irq 27
ata3: dev 0 cfg 49:2f00 82:346b 83:7f21 84:4003 85:3469 86:3c01 87:4003 88:203f
ata3: dev 0 ATA, max UDMA/100, 488397168 sectors (lba48)
ata3: dev 0 configured for UDMA/100
ata4: dev 0 cfg 49:2f00 82:346b 83:7f21 84:4003 85:3469 86:3c01 87:4003 88:203f
ata4: dev 0 ATA, max UDMA/100, 488397168 sectors (lba48)
ata4: dev 0 configured for UDMA/100
scsi0 : sata_sil
scsi1 : sata_sil
scsi2 : sata_sil
scsi3 : sata_sil

  Vendor: ATA       Model: Maxtor 6Y120M0    Rev: 1.01
  Type:   Direct-Access                      ANSI SCSI revision: 05
  Vendor: ATA       Model: WDC WD2500JD-00F  Rev: 1.01
  Type:   Direct-Access                      ANSI SCSI revision: 05
  Vendor: ATA       Model: WDC WD2500JD-00F  Rev: 1.01
  Type:   Direct-Access                      ANSI SCSI revision: 05

Attached scsi disk sda at scsi0, channel 0, id 0, lun 0 Attached scsi disk sdb at scsi2, channel 0, id 0, lun 0 Attached scsi disk sdc at scsi3, channel 0, id 0, lun 0 SCSI device sda: 240121728 512-byte hdwr sectors (122942 MB)
sda: unknown partition table
SCSI device sdb: 488397168 512-byte hdwr sectors (250059 MB)
sdb: unknown partition table
SCSI device sdc: 488397168 512-byte hdwr sectors (250059 MB)
sdc: unknown partition table

#12 lore raw

All,

Quoted 1 line
Quoted 1 line

drivers/ide driver status: Production, but see issue #4.

is there somewhere a web-site where one can look for the status of the current drivers?

I have an ICH5-Controllor on my board and I need Hot-Plug support - and if I followed the thread correctly, this combination isn't supported right now...

Bye
Michael