Purchased new server (HP proliant ML330) and am having problems locating a driver for the SATA raid controller. I did locate a driver I think was intended for RH 2.4.20 kernel and was able to force load in an 8.2 install (would prefer to install 9.0 but can't force load the driver) but it will not load during a normal boot (force load). I have extracted the following from the driver using strings: LD XY IDERAID LSILOGIC kernel_version=2.4.19-4GB-SMP using_checksums=1 parm_debug_level=i author=LSI Logic description=LSI Logic CSB6 driver license=GPL The part I find interesting is the last line. If the source is available I could recompile the driver and use it for a normal boot. Does anyone have knowledge of the sources for this driver and if so where I might be able do download them? The closest I have been able to find is for the megaraid controller -not- the ideraid controller. I have emailed a request to LSI but have not had a reply, same for HP. Ken
"Ken Schneider"
Does anyone have knowledge of the sources for this driver and if so where I might be able do download them? The closest I have been able to find is for the megaraid controller -not- the ideraid controller. I have emailed a request to LSI but have not had a reply, same for HP.
This driver is closed source, i.e. it's *only* available in binary form. Philipp
-----Original Message-----
From: Philipp Thomas
"Ken Schneider"
[Sat, 07 Feb 2004 17:02:26 -0500]: Does anyone have knowledge of the sources for this driver and if so where I might be able do download them? The closest I have been able to find is for the megaraid controller -not- the ideraid controller. I have emailed a request to LSI but have not had a reply, same for HP.
This driver is closed source, i.e. it's *only* available in binary form.
Philipp
After contacting LSI Logic tech support I was supplied with the source code which I have compiled, with the help of Anders Johansson, for the 2.4 kernels, default and smp. I further contacted tech support for inclusion of the sources into the kernel tree and was told that they are open and can be included. The correct controller is a Ultra ATA 100 IDE raid controller, (as used in the HP proliant ML330), by LSI Logic Corp. as can be seen by the beginning of the source files: /*megaide_ata.c*/ /* * Linux MegaRAID IDEal Software RAID Stack * Copyright (c) 2003 LSI Logic Corporation *----------------------------------------------------------------------------- * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. * *----------------------------------------------------------------------------- */ Incorporating the sources in the kernel tree would make it much easier to install linux on this machine asa well as any machine that uses this controller. Ken Schneider
On Sat, 2004-02-14 at 21:18 -0500, Ken Schneider wrote:
I further contacted tech support for inclusion of the sources into the kernel tree and was told that they are open and can be included.
Something's odd about the source package you sent to me. The module has MODULE_LICENSE("GPL"), yet there is a binary only megaraid_lib.o in there. I wonder if that's a mistake on the part of the person creating the tgz or if they just don't understand the GPL fully.
-----Original Message-----
From: Anders Johansson
On Sat, 2004-02-14 at 21:18 -0500, Ken Schneider wrote:
I further contacted tech support for inclusion of the sources into the kernel tree and was told that they are open and can be included.
Something's odd about the source package you sent to me. The module has MODULE_LICENSE("GPL"), yet there is a binary only megaraid_lib.o in there. I wonder if that's a mistake on the part of the person creating the tgz or if they just don't understand the GPL fully.
That's a good question and why I am going to contact the tech that I have been in contact with on monday. I do thank you again for your help, 2.4 is working so far on the machine but because of a couple of minor problems with USB I would like to try installing the 2.6 kernel to see if it would help, (a different way of compiling the kernel and modules). I have an HP LaserJet 3300 with scanning and I can only scan if I don't use the smp kernel. USB sound is the other item giving me fits. 2.4.21-166-default works fine, 2.4.21-166-smp4G always locks up. I also have someone from kernel.org interested in finding out if there are sources available for the lib file and if they are available it could lead to incorporation into the kernel tree. Once I have things sorted out I plan on dropping an email explaining what I went through to get 2.4 installed and possibly the 2.6 upgrade. You and Phillip seem to be burning a lot of midnight oil. 11:27pm in southwest Florida. Ken Schneider
Anders Johansson
I wonder if that's a mistake on the part of the person creating the tgz or if they just don't understand the GPL fully.
I'd guess whoever made that package doesn't have a clue as to what GPL means. Philipp
-----Original Message-----
From: Philipp Thomas
Anders Johansson
[Sun, 15 Feb 2004 04:46:05 +0100]: I wonder if that's a mistake on the part of the person creating the tgz or if they just don't understand the GPL fully.
I'd guess whoever made that package doesn't have a clue as to what GPL means.
Philipp
I'm going to email the tech that gave the tar package and see if the sources for the lib file are available and if they are not they are needed in order to add the package to the kernel tree. Ken Schneider
On Sun, 2004-02-15 at 16:23 +0100, Philipp Thomas wrote:
Anders Johansson
[Sun, 15 Feb 2004 04:46:05 +0100]: I wonder if that's a mistake on the part of the person creating the tgz or if they just don't understand the GPL fully.
I'd guess whoever made that package doesn't have a clue as to what GPL means.
I'd guess you're right. From megaide_pci.c * Closed Part of the * Source contains LSI's RAID intelligence and it will be provided as binary * image Something like what nvidia does with their driver, except nvidia doesn't claim it's GPLed.
-----Original Message-----
From: Anders Johansson
On Sun, 2004-02-15 at 16:23 +0100, Philipp Thomas wrote:
Anders Johansson
[Sun, 15 Feb 2004 04:46:05 +0100]: I wonder if that's a mistake on the part of the person creating the tgz or if they just don't understand the GPL fully.
I'd guess whoever made that package doesn't have a clue as to what GPL means.
I'd guess you're right. From megaide_pci.c
* Closed Part of the * Source contains LSI's RAID intelligence and it will be provided as binary * image
Something like what nvidia does with their driver, except nvidia doesn't claim it's GPLed.
Barring the ability to obtain the sources for megaide_lib.o do you think it would be possible for me to build a driver for the 2.6 kernel with these sources? The build process is quite different then 2.4. Ken Schneider
"Ken Schneider"
Barring the ability to obtain the sources for megaide_lib.o do you think it would be possible for me to build a driver for the 2.6 kernel with these sources? The build process is quite different then 2.4.
Without knowledge of the sources for megaide_lib it's impossible to tell if the driver will work under 2.6. You'd have to ask LSI tech support if that would be possible. If LSI encapsulated their code well enough and all necessary interface code is in the source files, chances would be rather good to get the driver working under 2.6. Philipp
-----Original Message-----
From: Philipp Thomas
Without knowledge of the sources for megaide_lib it's impossible to tell if the driver will work under 2.6. You'd have to ask LSI tech support if that would be possible.
If LSI encapsulated their code well enough and all necessary interface code is in the source files, chances would be rather good to get the driver working under 2.6.
Philipp
I'm willing to give it a try (I have the hardware to test with), I just need some assistance in how to incorporate the sources provided into the tree and compile them. I am able to compile in 2.4 but 2.6 is quite different, no make dep to set environment etc. Ken Schneider
On Sun, 2004-02-15 at 15:37 -0500, Ken Schneider wrote:
I'm willing to give it a try (I have the hardware to test with), I just need some assistance in how to incorporate the sources provided into the tree and compile them.
I am able to compile in 2.4 but 2.6 is quite different, no make dep to set environment etc.
No, there is no need for "make dep" in 2.6 But this is bigger than just putting the sources in the right place. Someone who knows something about kernel development will have to take the 'glue' code and port it to 2.6, you can't just compile it as is.
participants (3)
-
Anders Johansson
-
Ken Schneider
-
Philipp Thomas