Port Multiplier with Sil3132 and Sil4726

[heavyLIGHT]

I am currently trying to get a 5 drive "Venus T5" enclosure, which
uses an Sil4726 Storage Processor, connected to my system which uses a Sil3132 SATA Host Controller. Everything I've read
indicates that this should work, but it never connects, none of
the drives get initialized and mapped in.

I'm running Fedora 10 with kernel-2.6.27.21-170.2.56.fc10.


When the external controller is powered up I get...

ata8: exception Emask 0x10 SAct 0x0 SErr 0x0 action 0xe frozen
ata8: irq_stat 0x00b40090, PHY RDY changed
ata8: hard resetting link
ata8: SATA link up 3.0 Gbps (SStatus 123 SControl 0)
ata8.15: Port Multiplier 1.1, 0x1095:0x4726 r31, 7 ports, feat 0x1/0x9
ata8.00: hard resetting link
ata8.00: SATA link down (SStatus 0 SControl 300)
ata8.01: hard resetting link
ata8.01: SATA link down (SStatus 0 SControl 300)
ata8.02: hard resetting link
ata8.02: SATA link down (SStatus 0 SControl 300)
ata8.03: hard resetting link
ata8.03: SATA link down (SStatus 0 SControl 320)
ata8.04: hard resetting link
ata8.04: SATA link down (SStatus 0 SControl 320)
ata8.05: hard resetting link
ata8.05: SATA link up 3.0 Gbps (SStatus 123 SControl 0)
ata8.06: hard resetting link
ata8.06: SATA link down (SStatus 0 SControl 320)
ata8.05: unsupported device, disabling
ata8.05: disabled
ata8: EH complete


And if I hit the reconnect button on the external box I get...


ata8.00: failed to read SCR 1 (Emask=0x40)
ata8.01: failed to read SCR 1 (Emask=0x40)
ata8.02: failed to read SCR 1 (Emask=0x40)
ata8.03: failed to read SCR 1 (Emask=0x40)
ata8.04: failed to read SCR 1 (Emask=0x40)
ata8.05: failed to read SCR 1 (Emask=0x40)
ata8.06: failed to read SCR 1 (Emask=0x40)
ata8.15: exception Emask 0x10 SAct 0x0 SErr 0x80000 action 0xe frozen
ata8.15: irq_stat 0x01140010, PHY RDY changed
ata8.15: SError: { 10B8B }
ata8.00: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata8.01: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata8.02: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata8.03: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata8.04: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata8.05: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata8.06: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata8.15: hard resetting link
ata8.15: SATA link up 1.5 Gbps (SStatus 113 SControl 10)
ata8.00: hard resetting link
ata8.00: SATA link down (SStatus 0 SControl 300)
ata8.01: hard resetting link
ata8.01: SATA link down (SStatus 0 SControl 300)
ata8.02: hard resetting link
ata8.02: SATA link down (SStatus 0 SControl 300)
ata8.03: hard resetting link
ata8.03: SATA link down (SStatus 0 SControl 300)
ata8.04: hard resetting link
ata8.04: SATA link down (SStatus 0 SControl 300)
ata8.05: hard resetting link
ata8.05: SATA link up 1.5 Gbps (SStatus 113 SControl 0)
ata8.06: hard resetting link
ata8.06: SATA link down (SStatus 0 SControl 310)
ata8.05: unsupported device, disabling
ata8.05: disabled
ata8: EH complete


Any help would be greatly appreciated.
Thank You.

[theraidbox]

Consider using a hardware port multiplier solution instead of the software raid implementation of SATA controllers and port multipliers using Silicon Image chips. A 4x1 eSATA/USB Hardware Port Multiplier (HPM) that is OS independent and will work with any eSATA port is an alternative solution. The 4-port SATA integrated HPM with an Oxford OXUFS936QSE chipset uses hardware raid technology.

RAID mode selection is done using a rotary switch and setup is
completed by pressing a raid setting button. Once configured all is needed is to connect the HPM to your system via USB 2.0 or eSATA and detect, partition & mount the RAID volume.

RAID mode supported are FAST2 (2 drive RAID 0 Striping),, SAFE2 (RAID1 Mirroring), SAFE FAST (RAID 1+0 Mirrored Stripped), BIG 2 (2 drives Concatenation), FAST4 (4 drive RAID 0 Striping), BIG 4(4 drives Concatenation), RAID 1+0, RAID 5 over 4 drives, or RAID 5+S.

Connecting the HPM to onboard SATA ports on the motherboard is possible by using 2 eSATA ports PCI slot mounting bracket with two eSATA connectors to convert internal Serial ATA to
eSATA.

2TB, 3TB, 4TB, 6TB & 8TB Turnkey RAID solutions utilizing the 4x1 eSATA/USB Hardware Port Multiplier (HPM) are available at
https://www.theraidbox.com

The way software RAID works using SI controllers and port multipliers (PM) is that, you use the SATARAID5 array manager installed in Windows to see and configure as a raid volume all 5 drives connected to the PM. If the PM is connected to non-PM aware controller card (legacy) it will only detect one drive out of the 5.

But using a Hardware Port Multiplier which uses an integrated hardware controller, the raid setup is done on the HPM itself. So once you connect it to a SATA port, it will be seen as a single RAID volume.

[drees]

I've got basically the same thing here except that it's a Storbox

On my F10 install running 2.6.27.25-170.2.72.fc10.x86_64, it's been working except that under heavy access things lock up for a bit before reseting and I get messages the same as yours, except my drives are detected. I made sure mine was on before booting up and haven't tried the button on the front.

ata5.00: failed to read SCR 1 (Emask=0x40)
ata5.01: failed to read SCR 1 (Emask=0x40)
ata5.02: failed to read SCR 1 (Emask=0x40)
ata5.03: failed to read SCR 1 (Emask=0x40)
ata5.04: failed to read SCR 1 (Emask=0x40)
ata5.05: failed to read SCR 1 (Emask=0x40)
ata5.06: failed to read SCR 1 (Emask=0x40)
ata5.15: exception Emask 0x4 SAct 0x0 SErr 0x0 action 0x6 frozen
ata5.00: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata5.01: exception Emask 0x100 SAct 0x14 SErr 0x0 action 0x6 frozen
ata5.01: cmd 60/00:10:c0:f2:08/01:00:00:00:00/40 tag 2 ncq 131072 in
         res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
ata5.01: status: { DRDY }
ata5.01: cmd 60/70:20:50:f2:08/00:00:00:00:00/40 tag 4 ncq 57344 in
         res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
ata5.01: status: { DRDY }
ata5.02: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata5.03: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata5.04: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata5.05: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata5.06: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen
ata5.15: hard resetting link
ata5.15: SATA link up 3.0 Gbps (SStatus 123 SControl 0)
ata5.00: hard resetting link
ata5.00: SATA link up 3.0 Gbps (SStatus 123 SControl 320)
ata5.01: hard resetting link
ata5.01: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata5.02: hard resetting link
ata5.02: SATA link up 1.5 Gbps (SStatus 113 SControl 320)
ata5.03: hard resetting link
ata5.03: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata5.04: hard resetting link
ata5.04: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata5.05: hard resetting link
ata5.05: SATA link up 3.0 Gbps (SStatus 123 SControl 0)
ata5.06: hard resetting link
ata5.06: SATA link up 1.5 Gbps (SStatus 113 SControl 320)
ata5.00: configured for UDMA/100
ata5.01: configured for UDMA/100
ata5.02: configured for UDMA/100
ata5.03: configured for UDMA/100
ata5.04: configured for UDMA/100
ata5.05: unsupported device, disabling
ata5.05: disabled
ata5: EH complete
sd 5:0:0:0: [sdd] 156250000 512-byte hardware sectors (80000 MB)
sd 5:0:0:0: [sdd] Write Protect is off
sd 5:0:0:0: [sdd] Mode Sense: 00 3a 00 00
sd 5:0:0:0: [sdd] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
sd 5:1:0:0: [sde] 156250000 512-byte hardware sectors (80000 MB)
sd 5:1:0:0: [sde] Write Protect is off
sd 5:1:0:0: [sde] Mode Sense: 00 3a 00 00
sd 5:1:0:0: [sde] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
sd 5:2:0:0: [sdf] 156250000 512-byte hardware sectors (80000 MB)
sd 5:2:0:0: [sdf] Write Protect is off
sd 5:2:0:0: [sdf] Mode Sense: 00 3a 00 00
sd 5:2:0:0: [sdf] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
sd 5:3:0:0: [sdg] 156301488 512-byte hardware sectors (80026 MB)
sd 5:3:0:0: [sdg] Write Protect is off
sd 5:3:0:0: [sdg] Mode Sense: 00 3a 00 00
sd 5:3:0:0: [sdg] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
sd 5:4:0:0: [sdh] 156250000 512-byte hardware sectors (80000 MB)
sd 5:4:0:0: [sdh] Write Protect is off
sd 5:4:0:0: [sdh] Mode Sense: 00 3a 00 00
sd 5:4:0:0: [sdh] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
sd 5:0:0:0: [sdd] 156250000 512-byte hardware sectors (80000 MB)
sd 5:0:0:0: [sdd] Write Protect is off
sd 5:0:0:0: [sdd] Mode Sense: 00 3a 00 00
sd 5:0:0:0: [sdd] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
sd 5:1:0:0: [sde] 156250000 512-byte hardware sectors (80000 MB)
sd 5:1:0:0: [sde] Write Protect is off
sd 5:1:0:0: [sde] Mode Sense: 00 3a 00 00
sd 5:1:0:0: [sde] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
sd 5:2:0:0: [sdf] 156250000 512-byte hardware sectors (80000 MB)
sd 5:2:0:0: [sdf] Write Protect is off
sd 5:2:0:0: [sdf] Mode Sense: 00 3a 00 00
sd 5:2:0:0: [sdf] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
sd 5:3:0:0: [sdg] 156301488 512-byte hardware sectors (80026 MB)
sd 5:3:0:0: [sdg] Write Protect is off
sd 5:3:0:0: [sdg] Mode Sense: 00 3a 00 00
sd 5:3:0:0: [sdg] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
sd 5:4:0:0: [sdh] 156250000 512-byte hardware sectors (80000 MB)
sd 5:4:0:0: [sdh] Write Protect is off
sd 5:4:0:0: [sdh] Mode Sense: 00 3a 00 00
sd 5:4:0:0: [sdh] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA

My heavy load test is to run

dd if=/dev/<drive> bs=16k of=/dev/null

on 4-5 drives. Seems pretty reliable to trigger so far, reads from one of the drives will not start.

I'm going to try the latest updates-testing kernel and see how that goes before opening some bug reports...

[drees]

Looks like I have worked around the hang/resets on my system by disabling NCQ on the drives that supported it. I suspect that the port multiplier starts getting flaky under high load when NCQ is going. When I tried building an array using drives on the device it would hang up basically instantly.

I may try setting the NCQ values to something higher than 1 but lower than 31 and see what happens...