linux-kernel - Re: Problem with ata layer in 2.6.24

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <479E9BD0.3040900@rtr.ca>
Date:	Mon, 28 Jan 2008 22:21:52 -0500
From:	Mark Lord <liml@....ca>
To:	Gene Heskett <gene.heskett@...il.com>
Cc:	Robert Hancock <hancockr@...w.ca>,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: Problem with ata layer in 2.6.24

Gene Heskett wrote:
> On Monday 28 January 2008, Gene Heskett wrote:
>> On Monday 28 January 2008, Robert Hancock wrote:
>> [...]
>>
>>> Check the /etc/modprobe.conf file, a lot of distributions use this to
>>> generate the initrd. If there's references to pata_amd it'll try and
>>> include it.
>> Bingo!  Thanks Robert, I'll try it again with that line commented.  I wasn't
>> aware of that connection at all.  Yup, it worked, I feel a reboot coming
>> on. :)
> 
> But it didn't work, apparently commenting that line out needs to be balanced 
> by adding another line telling it amd74xx is the 'hostadapter', not 
> necessarily scsi.
> 
> Can this be made more universal so I don't have to edit /etc/modprobe.conf?
>..

You could really do it like Linus (and me), and not bother with modules
for critical services like hard disks.

Just build them *into* the core kernel (select "y" or "checkmark" rather
than "m" or "dot" for modules).  This eliminates a ton of crap that can fail,
and may also make your kernel a micro-MIP faster (core memory is often mapped
without page table entries, whereas loaded modules use page tables.. slower, slightly).

Linus just edits the /boot/grub/menu.lst, and clones an existing boot entry
for the new kernel, editing the "kernel" line to match the name of the file
that got installed in /boot by "make install" (from the kernel directory).
He just leaves the ramdisk/initrd line as-was --> wrong version, but that's okay.

I totally get rid of them here, but that requires hardcoding the root=/dev/xxxx
part on the "kernel" line.  No big deal, it works just fine that way.

Cheers
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/