LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - Hardware (https://www.linuxquestions.org/questions/linux-hardware-18/)
-   -   RocketRAID 1520 driver compiling problems on Kernel 2.6.11-8 (https://www.linuxquestions.org/questions/linux-hardware-18/rocketraid-1520-driver-compiling-problems-on-kernel-2-6-11-8-a-323823/)

Icedman 05-15-2005 06:55 PM

RocketRAID 1520 driver compiling problems on Kernel 2.6.11-8
 
Hi all!

I'm having trouble compiling the Open Source drivers from the Highpoint Technologies website
I figured out what was wrong with the makefile (which made it *not* compile the module),
but now I've got three errors:

icedman:/home/icedman/Software/hpt2 # make NON_RAID=1 RR1520=1 KERNELDIR=/usr/src/linux-2.6.11.8
make -C /usr/src/linux-2.6.11.8 SUBDIRS=`pwd` modules
make[1]: Entering directory `/usr/src/linux-2.6.11.8'
CC [M] /home/icedman/Software/hpt2/hpt.o
In file included from /home/icedman/Software/hpt2/hpt.c:49:
drivers/scsi/hosts.h:1:2: warning: #warning "This file is obsolete, please use <scsi/scsi_host.h> instead"
/home/icedman/Software/hpt2/hpt.c:91: warning: `sd_inuse' defined but not used
/home/icedman/Software/hpt2/ioctl.c:88: warning: `nothing' used but never defined
LD [M] /home/icedman/Software/hpt2/hpt37x2.o
Building modules, stage 2.
MODPOST
*** Warning: "nothing" [/home/icedman/Software/hpt2/hpt37x2.ko] undefined!
CC /home/icedman/Software/hpt2/hpt37x2.mod.o
LD [M] /home/icedman/Software/hpt2/hpt37x2.ko
make[1]: Leaving directory `/usr/src/linux-2.6.11.8'

Anybody got an idea how to fix those. Since the driver hangs on insmod.

Thanks for your answers

Electro 05-16-2005 04:02 PM

Those are just warnings. It will install. Make sure you configure the kernel to not include highpoint. Make sure you update the firmware or highpoint BIOS.

In the readme, you need to load sd_mod first or use modprobe hpt37x2.ko. The utility modprobe will load the modules that hpt37x2.ko depends on.

Icedman 05-16-2005 04:20 PM

I know those. I'm running the same software on my comp now with kernel 2.6.8-24.14.

The problem is, that the module hangs when compiled to 2.6.11-8.
So I'd really want to get rid of those warnings. What changes in the
SCSI modules did they do to kernel 2.6.11-8 (if any)?

Edit:
Ok. I'll give you the errors the module gives under 2.6.11-8

RocketRAID 1520 SATA Controller May 18 2005 03:08:57

ACPI:PCI Interrupt 0000:00:0b.0[A] -> GSI 19 (level, low) -> IRQ 193

scsi2: hpt37x2
Vendor: WDC WD20 Model: 00JD-00GBB0 rev 2.0
Type: Direct-Access Ansi SCSI-revision 00

SCSI Device SDC: 390721967 512-byte hdwr sectors 200050MB

sdc: asking for cache data failed
sdc: assuming drive cache: write-through

abort called PID 3423 target:0 lun:0 flags:1 reason:0
reset called PID 3423 target:0 lun:0 flags:1

(PIDs 3483, 3486, 3487 and 3488 are same, then this comes:)

SCSI error : <2 0 0 0> return code = 0x6000000

Anybody got any Idea what the heck is going on? Since this works perfectly on kernel 2.6.8-24.14

Electro 05-17-2005 05:24 PM

At the bootloader try adding apci=off.

Add -DDEBUG -DDEBUG_LEVEL=3 to the make command. It will give you more information. I suggest relaying the information to Highpoint so they can fix the problem.

I'm sorry, I can not help you on what line of code to change because I have not yet write software to control hardware.

For now, try using the Highpoint module that comes with the kernel until Highpoint fixes the problem.

Icedman 05-17-2005 05:55 PM

No offense, but I ain't gonna use that Crap of a "module" that's integrated in the kernel.
If I want stability I'll go with doing some modding to the driver itself. I think I'll go and grab
those 2.6.8-24.14 SCSI kernel modules and throw them on the 2.6.11.8.

No offense, but something really bad has been done to the SCSI modules after 2.6.8-24.14,
since the Highpoint module works just fine on that. I can compile the module for 2.6.11.8, but..
The Makefile's really funny. It has a nice layout etc, but when compiling with the option NON_RAID=1,
the automake doesn't compile. It seems the Makefile is trying to copy raid.o to nothing.

Oh well, I think I'll just boot few hundred times this POS to get it working.

j_col 09-06-2005 07:32 AM

I am having the same issues with the RocketRaid 1520 from Highpoint, anybody ever managed to get this sucker installed on an distro, or to be more specific Redhat FC4?

I am going to install the kernel source code and see if I can compile the open source driver from highpoint, but I am not feeling too optimistic after reading this post. Icedman you were saying you had some trouble with the make file distributed with the driver? Can you show me how you revised it?

Electro 09-07-2005 03:14 AM

It should work with the kernel Highpoint software (hpt366) because it is PATA controller with PATA to SATA bridge. It does not use SCSI at all. The code that you are trying to compile uses SCSI.

It is not the Makefile is a few lines of code. Ask the kernel developers that has knowledge of Highpoint controllers.

I have two Highpoint controllers. They use the HPT370 IDE controller chip. I use the hpt366 module that comes with the kernel without any stablity or reliability problems. Right now I am using kernel version 2.6.13 and no problems yet with the hpt366 module.

j_col 09-08-2005 07:52 AM

Please excuse my ignorance Electro, while I am not a Linux newbie I have never played around with the Kernal before. Can you tell me what exactly the hpt366 module is, and how I can use this to run my RocketRAID 1520 controller?

Electro 09-08-2005 09:27 PM

You have to re-configure and compile your kernel if you can not find it in /lib/modules/`uname`. Some distributions have it built-in, as a module, or did not include it at all. The HPT3xx option is under the IDE branch. If it is a module, typing 'modprobe hpt366' will control your controller and look up any IDE drives. If there are, the controller will automatically set DMA. The hpt366 module have to be included in the ramdisk file. The kernel have to have ramdisk and initrd support builtin, but not as a module.


All times are GMT -5. The time now is 05:26 AM.