Re: libata: ide-drive mounted as sda : how to deal with 32-bit/multisector?



Shadow_7 wrote:
I just got a new ide-controllercard, cause my old one was broken.

if you say so.


it was broken - and it was really broken, cause I stepped on it and broke it :)

the new card is a pci-card which is a raid card, but also works as "normal" ide card in socalled "pass-through"-mode. The attached drive is definitely an IDE-drive.

With kernel 2.6.17 or 2.6.15 the drive attached to the card is recognized as hde (which is what I expect when adding an extra ide-controller) and the module in question ist it821x.
With the new kernel 2.6.20 this module does not exist anymore and seems to be replaced with pata_it821x.

I use Ubuntu 6.10 (2.6.17) and ubuntu 7.04 (2.6.20) and did not compile the kernel on my own. I tried but could not boot the new kernel yet, cause it could not mount the root-drive and I havent figured out yet why.

lspci reports:
00:0f.0 RAID bus controller: <pci_lookup_name: buffer too small> (rev 13)

lspci -n
00:0f.0 0104: 1283:8212 (rev 13)


Which doesn't seem right since pata is supposedly the early generation of SATA support. I thought you said this was an IDE card?

<skip>

This only happens with me when I use the ide-scsi module(s). Which is somewhat deprecated in 2.6.x kernels. It was intended(AFAIK) for cd burning way back when, since scsi supposedly had better support of the command structure needed for burning media through cdrecord. In theory you can burn using the ATAPI support in cdrecord (wodim) with the 2.6.x kernels and newer versions of cdrecord.


as I see, libata is under highly development in recent 2.6-kernel and somehow trying to unify the different protocols.


What are you trying to accomplish? Better drive performance? A different mount device for your drive? hdparm to give you a warm fuzzy, even if it doesn't actually affect the drive? Just venting? I just don't get your purpose/intent from your original post.


I'm trying to understand my system. I'm trying to understand whats going on. I stepped on the problem when trying to activate multisec-support on the drive which I experienced as performace-improvment on other drives on other servers.
And never turn away a warm fuzzy ;)

pata_it821x is listed in 2.6.20-kernel as (very experimental) module for SATA-cards while it821x is listed as module for ATA/IDE-cards.
Nevertheless on my actual system the pata-module is able to control my IDE-card as well.

I have several servers with external IDE-cards. Some servers have more then 4 IDE-drives and therefor need an extra card. Some other servers have a JMicron-controller onboard which I could not get running on 2.6.17-kernels, so I added an extra card. When trying to compile new kernels on my own and boot this kernel I run into kernel-panic because the root-drive cannot be found. I think this somehow is related to this issue and I need to understand how the kernel sees and adresses the drives to configure the boot-loader proper, put the correct modules in ramdisk or compile into the kernel.

I always had the feeling that knowing your systems gives you a warm fuzzy. Although I've to admit that my english is not good enough to get the exact meaning of this expression. But I liked it in your posting cause it made me feel grown up and wise :)

peter










.



Relevant Pages

  • athlon-xp + fakeraid regression
    ... The build completes fine, the kernel boots fine, the machine will seem to be fine as long as it remains quiescent. ... At the beginning, just after hitting enter on the make command, one of the ad4 disk light goes on solid for several seconds. ... There is a well known thing where these cheap pata fakeraid cards will try to do ata133 if the drive says it can, when really, even if he drives are new ata133 drives and the cables are new and short and shielded, you still shouldn't try to do ata133 since the spec is too tight and you'll just get bit errors or other failures. ... The fix is use ata100 somehow, either by disabling dma entirely in loader.conf (since you have no more selective option there, and the raid card bios never has an option for controlling pio/dma mode like motherboard bios's have) and then use atacontrol in rc.early to set udma5, or by using disks that can only do ata100 and only advertise ata100 to the controller. ...
    (freebsd-current)
  • Re: pata_sil680 regression 2.6.22->2.6.24
    ... Having recently upgaded my Ubuntu install from Gutsy to Hardy, ... Seagate disk connected via a SiI680 PCI card is no longer detected. ... I'm not sure if the kernel config has changed as I ... Note, my bios does not support 750GB drives, so the drive is not ...
    (Linux-Kernel)
  • How to turn off ide-scsi-emulation (not quite straightforward..)
    ... I have two linux servers, both running Fedora Core 1 which is kept ... No parameters requiring ide-scsi emulation are given to kernel at ... the cd-rom drives are very seldom needed and I can't trace when the change ...
    (comp.os.linux.setup)
  • Re: Cant boot if SCSI (AIC7xxx) not connected to any drives
    ... > would let me indicate that there are no drives on the card. ... > The boot process does seem to find my boot drive on the 3-ware raid and even ... Seems like your kernel has not the correct modules for that card. ...
    (comp.os.linux.setup)
  • Re: Migrating to a RAID 1 setup (hardware RAID)
    ... I don't have an identical drive handy unfortunately, but drives are ... need to do is make sure the linux kernel supports them. ... - Configure your boot loader to add the new kernel so you can boot it ... Install card. ...
    (comp.os.linux.hardware)