Smart Array Controller and Kernel 2.2.18/19

[Ayman Altounji]

Hi, all.

I have a Compaq Proliant 1600R with 6 internal scsi drives and 13 external scsi drives, all hooked to a SmartArray 4200 controller, dual 500MHz cpus.

For a long time, it has been running RedHat w/ kernel 2.2.13. I keep it relatively up to date from the updates site, but recently decided to move to kernel 2.2.19 - RedHat current "official" kernel for 6.2.

The machine was built with minimal RAID - most drives were individual drives, thus the logical devices were /dev/ida/c0d0-d14 or something similar. We re-arranged the drives into two RAID-1 and two RAID-5 arrays, with a couple of individual drives left over. This changed the logical device naming to /dev/ida/c0d0-/dev/ida/c0d4, /dev/ida/c0d10 and /dev/ida/c0d11.

When booting from kernel 2.2.13 or 2.2.14 I get the following output:
cpqarray: Device 1011 has been found at 0 78
Compaq SMART2 Driver (v 1.0.6)
Found 1 controller(s)
cpqarray: Finding drives on ida0 (Smart Array 4200)
cpqarray ida/c0d0: blksz=512 nr_blks=35561280
cpqarray ida/c0d1: blksz=512 nr_blks=35561280
cpqarray ida/c0d2: blksz=512 nr_blks=35561280
cpqarray ida/c0d3: blksz=512 nr_blks=71122560
cpqarray ida/c0d4: blksz=512 nr_blks=106692000
cpqarray ida/c0d10: blksz=512 nr_blks=142261440
cpqarray ida/c0d11: blksz=512 nr_blks=284522880
ida/c0d0: ida/c0d0p1 ida/c0d0p2 ida/c0d0p3 ida/c0d0p4 < ida/c0d0p5 ida/c0d0p6 ida/c0d0p7 ida/c0d0p8 ida/c0d0p9 >
ida/c0d1: ida/c0d1p1
ida/c0d2: ida/c0d2p1 ida/c0d2p2 ida/c0d2p3
ida/c0d3: ida/c0d3p1
ida/c0d4: ida/c0d4p1
ida/c0d10: ida/c0d10p1
ida/c0d11: ida/c0d11p1

No problem. But when I attempt to boot from a 2.2.18 or 2.2.19 kernel I get the following output:
cpqarray: Device 1011 has been found at 0 78
Compaq SMART2 Driver (v 1.0.6)
Found 1 controller(s)
cpqarray: Finding drives on ida0 (Smart Array 4200)
cpqarray ida/c0d0: blksz=512 nr_blks=35561280
cpqarray ida/c0d1: blksz=512 nr_blks=35561280
cpqarray ida/c0d2: blksz=512 nr_blks=35561280
cpqarray ida/c0d3: blksz=512 nr_blks=71122560
cpqarray ida/c0d4: blksz=512 nr_blks=106692000
cpqarray ida/c0d10: blksz=512 nr_blks=142261440
cpqarray ida/c0d11: blksz=512 nr_blks=284522880
Partition check:
ida/c0d0: ida/c0d0p1 ida/c0d0p2 ida/c0d0p3 ida/c0d0p4 < ida/c0d0p5 ida/c0d0p6 ida/c0d0p7 ida/c0d0p8 ida/c0d0p9 >
ida/c0d1: ida/c0d1p1
ida/c0d2: ida/c0d2p1 ida/c0d2p2 ida/c0d2p3
ida/c0d3: ida/c0d3p1
ida/c0d4: ida/c0d4p1

Note drives d10 and d11 are no longer listed in the partition check. The driver does appear to see the disks (pre-Partition check) but not the partitions (post-Partition check)

Attempts to mount d10 or d11 after the system boots results inan error message saying that /dev/ida/c0d[10/11]p1 is not a valid block device. It seems as if the kernel is scanning the devices, finds no valid device on d5, and stops looking.

Anyone have any ideas on this behavior ?