pata_hpt drivers are experimental and should be disabled
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
linux (Ubuntu) |
Expired
|
Undecided
|
Unassigned |
Bug Description
The libata drivers for hpt366 are buggy/experimental: http://
I tried the 2.6.28 kernel from jaunty on my Intrepid system and this driver detects the hpt366 but it fails with the same DMA problems as the kernel bugid. My software raid mirror failed within minutes of booting. I don't think this driver should be in a release version until it's solid. Also, the intrepid 2.6.27 kernels do not detect the libata hpt366 properly, and load the 37x driver which does not work on the hpt302/302n.
The old IDE hpt366 driver was stable back in feisty - ran it for a year at least without issues.
I think Intrepid/Jaunty should re-enable the old IDE modules for hpt366
I have the same problem. I have a rocketraid 302n-card (i'm not using the hardware-raid). With the hpt366-driver, I had no problems. Since ubuntu 8.10, I have been forced to use the pata_hpt3x2n and it is very buggy. It sometimes does not detect some or all of the ide-drives connected to the controller card. When the pata_hpt3x2n do detect my ide-disks, throughput are constandly at 39K/sec, which in turn makes my computer extremley slow to respond (my OS is on these disks).
I agree totaly with the above poster. Please re-enable old IDE modules for hpt366!