S-ATA 3114, SuSE 9.0 and more than 2 disks
Hi, actually I'm testing S-ATA support on a CELSIUS V810 dual opteron workstation with several distributions. Now it came out, that SuSE 9.1 works very nice, but SuSE 9.0 dies when more than 2 S-ATA disks are attached to the Silicon Image 3114 controller. The differences I found in the kernel are: SuSE 9.0: Kernel 2.4.21-215-smp with sata-sil module version 0.52 SuSE 9.1: Kernel 2.6.5-52-smp with sata-sil module version 0.54 So I guess the problem is caused by "features" in the older version of the sata_sil module. I tried to just copy the newer to the place of the older and recompile the 2.4.21-215 kernel with that, but its not that simple. Are there any intentions from SuSE to fix that bug and provide an updated 2.4.21 kernel for the SuSE 9.0 for AMD64 or do I have to practise kernel hacking on my own? :-) Regards Rainer -- Dipl.-Inf. (FH) Rainer Koenig Project Manager Linux Fujitsu-Siemens VP BC E SW OS Phone: +49-821-804-3321 Fax: +49-821-804-2131
Does the 9.0 dist. Have driver for 3112a and 9.1 for 3114 ? 3112a is only a 2 device PCI controller, whereas 3114 a 4 device; I know that we can here under MS Windows force a 3112 driver to use 2 ports of 3114 (as we did 12 months ago whilst developing board using 3114 before working driver existed). Andrew Cotterill Product Manager - Mainboards & VGA TEL: +44 (0)1908 370288 FAX: +44 (0)1908 370686 CELL: +44 (0)7813 802808 http://www.epox.org Please provide original Text if replying. ________________________________ CONFIDENTIALITY NOTICE: This Email is confidential and may also be privileged. If you are not the intended recipient, please notify the sender IMMEDIATELY; you should not copy the email or use it for any purpose or disclose its contents to any other person. GENERAL STATEMENT: Any statements made, or intentions expressed in this communication may not necessarily reflect the view of Actron Electronics. Be advised that no content herein may be held binding upon Actron Electronics or any associated company unless confirmed by the issuance of a formal contractual document or Purchase Order.E&OE. All specifications subject to change without prior notice. ________________________________ -----Original Message----- From: Rainer Koenig [mailto:Rainer.Koenig@fujitsu-siemens.com] Sent: 07 May 2004 07:23 To: suse-amd64@suse.com Subject: [suse-amd64] S-ATA 3114, SuSE 9.0 and more than 2 disks Hi, actually I'm testing S-ATA support on a CELSIUS V810 dual opteron workstation with several distributions. Now it came out, that SuSE 9.1 works very nice, but SuSE 9.0 dies when more than 2 S-ATA disks are attached to the Silicon Image 3114 controller. The differences I found in the kernel are: SuSE 9.0: Kernel 2.4.21-215-smp with sata-sil module version 0.52 SuSE 9.1: Kernel 2.6.5-52-smp with sata-sil module version 0.54 So I guess the problem is caused by "features" in the older version of the sata_sil module. I tried to just copy the newer to the place of the older and recompile the 2.4.21-215 kernel with that, but its not that simple. Are there any intentions from SuSE to fix that bug and provide an updated 2.4.21 kernel for the SuSE 9.0 for AMD64 or do I have to practise kernel hacking on my own? :-) Regards Rainer -- Dipl.-Inf. (FH) Rainer Koenig Project Manager Linux Fujitsu-Siemens VP BC E SW OS Phone: +49-821-804-3321 Fax: +49-821-804-2131 -- Check the List-Unsubscribe header to unsubscribe For additional commands, email: suse-amd64-help@suse.com
"Andy C (EPoX UK)" <andy@epox.org> writes:
Does the 9.0 dist. Have driver for 3112a and 9.1 for 3114 ?
Yes. The SuSE 9.0 for AMD64 has a backported module from the 2.6-kernel series. On the 9.1 it works perfectly fine with 4 disks attached, on the 9.0 SuSE tries to do a partition check during boot and it stops responding here.
3112a is only a 2 device PCI controller, whereas 3114 a 4 device; I know that we can here under MS Windows force a 3112 driver to use 2 ports of 3114 (as we did 12 months ago whilst developing board using 3114 before working driver existed).
I know. I'm watching the sources for the S-ATA modules since several months and I remember that the first attempts were based on a 3112 module and the wish of the module developer to find out the registers he has to access to make use of the other 2 channels. Well, since the module is just about 400 LoC I guess it shouldn't be a big deal to find out the things that need to be changed. Only that I would prefer if someone that already "backported" that module would do it instead of me. :-) Ok, I'm not afraid to dig into the kernel sources and it will be a nice practise but I wonder if my boss is happy if I start kernel hacking now. :-) Regards Rainer -- Dipl.-Inf. (FH) Rainer Koenig Project Manager Linux Fujitsu-Siemens VP BC E SW OS Phone: +49-821-804-3321 Fax: +49-821-804-2131
I know. I'm watching the sources for the S-ATA modules since several months and I remember that the first attempts were based on a 3112 module and the wish of the module developer to find out the registers he has to access to make use of the other 2 channels.
Well, since the module is just about 400 LoC I guess it shouldn't be a big deal to find out the things that need to be changed. Only that I would prefer if someone that already "backported" that module would do it instead of me. :-) Ok, I'm not afraid to dig into the kernel sources and it will be a nice practise but I wonder if my boss is happy if I start kernel hacking now. :-)
Most likely libata.c needs to be updated too. Our sata backport is based on the patches in ftp://ftp.kernel.org/pub/linux/kernel/people/jgarzik/libata/2.4* There are newer versions there. -Andi
On Fri, May 07, 2004 at 07:37:24AM +0100, Andy C (EPoX UK) wrote:
Does the 9.0 dist. Have driver for 3112a and 9.1 for 3114 ?
Yes. In fact 9.1 has two drivers; the old one and a new libata based one. The new one should be used by default. The 9.0 SATA update CD also had an older version of this driver without 3114 support.
3112a is only a 2 device PCI controller, whereas 3114 a 4 device; I know that we can here under MS Windows force a 3112 driver to use 2 ports of 3114 (as we did 12 months ago whilst developing board using 3114 before working driver existed).
At least the libata (sata_sil driver) will always probe all available ports; there is no number hardcoded in the driver. -Andi
Rainer Koenig <Rainer.Koenig@fujitsu-siemens.com> writes:
Hi,
actually I'm testing S-ATA support on a CELSIUS V810 dual opteron workstation with several distributions. Now it came out, that SuSE 9.1 works very nice, but SuSE 9.0 dies when more than 2 S-ATA disks are attached to the Silicon Image 3114 controller. The differences I found in the kernel are:
SuSE 9.0: Kernel 2.4.21-215-smp with sata-sil module version 0.52 SuSE 9.1: Kernel 2.6.5-52-smp with sata-sil module version 0.54
So I guess the problem is caused by "features" in the older version of the sata_sil module. I tried to just copy the newer to the place of the older and recompile the 2.4.21-215 kernel with that, but its not that simple.
Are there any intentions from SuSE to fix that bug and provide an updated 2.4.21 kernel for the SuSE 9.0 for AMD64 or do I have to practise kernel hacking on my own? :-)
No - but you should be able to run the 9.1 2.6 kernel on 9.0, Andreas -- Andreas Jaeger, aj@suse.de, http://www.suse.de/~aj SUSE Linux AG, Maxfeldstr. 5, 90409 Nürnberg, Germany GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
Andreas Jaeger <aj@suse.de> writes:
Are there any intentions from SuSE to fix that bug and provide an updated 2.4.21 kernel for the SuSE 9.0 for AMD64 or do I have to practise kernel hacking on my own? :-)
No - but you should be able to run the 9.1 2.6 kernel on 9.0,
Yes, surely *I* will be able to do that. The question is if my customers are able or willing to upgrade to the 2.6 kernel. Of course I suggest an upgrade to SuSE 9.1, but if for whatever reasons they decide to stick to SuSE 9.0 I need a solution for them. Regards Rainer -- Dipl.-Inf. (FH) Rainer Koenig Project Manager Linux Fujitsu-Siemens VP BC E SW OS Phone: +49-821-804-3321 Fax: +49-821-804-2131
Hello, two P-ATA Disks attached to an on board Promise SATA378 (MSI K8T Neo) are not found by Suse 9.1 64bit. Any idea what I can do? Regards, Alexander
"Zaaap" <Zaaap@gmx.de> writes:
Hello,
two P-ATA Disks attached to an on board Promise SATA378 (MSI K8T Neo) are not found by Suse 9.1 64bit.
Any idea what I can do?
Please send the hwinfo output, Andreas -- Andreas Jaeger, aj@suse.de, http://www.suse.de/~aj SUSE Linux AG, Maxfeldstr. 5, 90409 Nürnberg, Germany GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
måndag 10 maj 2004 07:18 skrev Zaaap:
Hello,
two P-ATA Disks attached to an on board Promise SATA378 (MSI K8T Neo) are not found by Suse 9.1 64bit.
Any idea what I can do?
Are you sure they aren't found as hde, hdf ... that's what kernel 2.6 does on my K8T Neo board.
On Mon, May 10, 2004 at 12:21:10PM +0200, ?rn Hansen wrote:
m?ndag 10 maj 2004 07:18 skrev Zaaap:
Hello,
two P-ATA Disks attached to an on board Promise SATA378 (MSI K8T Neo) are not found by Suse 9.1 64bit.
Any idea what I can do?
Are you sure they aren't found as hde, hdf ... that's what kernel 2.6 does on my K8T Neo board.
ide=reverse may help with that. -Andi
On Mon, 2004-05-10 at 07:18, Zaaap wrote:
Hello,
two P-ATA Disks attached to an on board Promise SATA378 (MSI K8T Neo) are not found by Suse 9.1 64bit.
Any idea what I can do?
Regards, Alexander
Alexander... If you are refering to the "odd ball" (2 sata, and one pata) pata connector on the TX2 chip, then I fear it is not yet support with the GPL 2.6 drivers. At least that is what my research on the topic for mine came up with... Somebody please tell me I'm wrong! Jerry
participants (7)
-
Andi Kleen -
Andreas Jaeger -
Andy C (EPoX UK) -
Jerome R. Westrick -
Rainer Koenig -
Zaaap -
Örn Hansen