Hi,
I have an Adaptec eSATA card, model 1225SA: http://www.adaptec.com/en-US/products/Controllers/Hardware/sata/entry/AAR-12...
Although the chipset is sata_sil24, it is not recognized by default by the kernel because its PCI ID is different than the ones recognized by this driver on the CentOS kernel. The driver recognizes PCI IDs such as 0x3132, 0x0242, 0x3131, but not 0x0244, which is the one this card has.
I found out that it works with the latest kernel, and that the patch to make it work only adds the 0x0244 PCI ID to the list of IDs recognized by the sata_sil24 driver. Here is the patch: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdif...
Following the instructions on the Wiki, I built a custom kernel based on the CentOS Plus kernel, but adding this patch. The adapter worked as expected with that kernel.
So I ask: Would it be possible to incorporate this patch to the default set of patches applied on the CentOS Plus kernel? I would really appreciate it since that would make it easier for me to keep updated without worrying about the card.
Thanks a lot! Filipe
Filipe Brandenburger wrote:
Hi,
I have an Adaptec eSATA card, model 1225SA: http://www.adaptec.com/en-US/products/Controllers/Hardware/sata/entry/AAR-12...
Although the chipset is sata_sil24, it is not recognized by default by the kernel because its PCI ID is different than the ones recognized by this driver on the CentOS kernel. The driver recognizes PCI IDs such as 0x3132, 0x0242, 0x3131, but not 0x0244, which is the one this card has.
I found out that it works with the latest kernel, and that the patch to make it work only adds the 0x0244 PCI ID to the list of IDs recognized by the sata_sil24 driver. Here is the patch: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdif...
Following the instructions on the Wiki, I built a custom kernel based on the CentOS Plus kernel, but adding this patch. The adapter worked as expected with that kernel.
So I ask: Would it be possible to incorporate this patch to the default set of patches applied on the CentOS Plus kernel? I would really appreciate it since that would make it easier for me to keep updated without worrying about the card.
You picked a good day to ask, I am getting ready to build the new plus kernel for c5 right now :D
I will put that patch in there too if it applies cleanly for me.
Thanks, Johnny Hughes
On Wed, Oct 1, 2008 at 12:57, Johnny Hughes johnny@centos.org wrote:
You picked a good day to ask, I am getting ready to build the new plus kernel for c5 right now :D
Great to know! :-D
I will put that patch in there too if it applies cleanly for me.
Let me know if it doesn't. I included it in the specfile as Patch400000 to go after the Patch99999 (last one in CentOS Plus), but I don't think that's needed.
In any case, if you look at the patch you will see it's one line only, so it's really trivial.
Let me know if I can help with anything else to get this patch included!
Thanks, Filipe
Hi,
On Wed, Oct 8, 2008 at 11:46, Matthew Kent matt@bravenet.com wrote:
On Wed, 2008-10-01 at 11:57 -0500, Johnny Hughes wrote:
You picked a good day to ask, I am getting ready to build the new plus kernel for c5 right now :D
Any luck with this? :)
I contacted Johnny last Friday, he was still working on the kmods. He is making the kmods use the RHEL "weak updates" method and not the fedora method. This will improve things a lot, because they won't have to be rebuilt each time a new kernel is released. Last Friday, 2 of the 3 kmods were working correctly, he was working on the last one at the time.
I'm sure it will be ready when it's ready! :-)
Filipe
Matthew Kent wrote:
Any luck with this? :)
Not wanting to sound like a spoil sport but if its critical to your server(s) can't you build your own kernel, pretty sure I've seen this discussed on the docs list. I suppose the CentOS plus one has to go through QA and regression testing no?
Spike.
Spike Turner wrote:
Matthew Kent wrote:
Any luck with this? :)
Not wanting to sound like a spoil sport but if its critical to your server(s) can't you build your own kernel, pretty sure I've seen this discussed on the docs list. I suppose the CentOS plus one has to go through QA and regression testing no?
Spike.
Hey, if he has a critical need, he could simply trek over to http://people.centos.org/hughesjr/kernel/5/plus/2.6.18-92.1.13.el5.centos.pl... and help Johnny test a build that stands a *much* better chance of working right that a home-made one.
On Wed, 2008-10-08 at 12:56 -0500, Robert wrote:
Spike Turner wrote:
Matthew Kent wrote:
Any luck with this? :)
Not wanting to sound like a spoil sport but if its critical to your server(s) can't you build your own kernel, pretty sure I've seen this discussed on the docs list. I suppose the CentOS plus one has to go through QA and regression testing no?
Not sure where I said its critical? It's merely a friendly question I thought might be interesting to the other centosplus users on the list.
Spike.
Hey, if he has a critical need, he could simply trek over to http://people.centos.org/hughesjr/kernel/5/plus/2.6.18-92.1.13.el5.centos.pl... and help Johnny test a build that stands a *much* better chance of working right that a home-made one.
Wasn't aware that existed, thanks!
-
Filipe Brandenburger -
Johnny Hughes -
Matthew Kent -
Robert -
Spike Turner