Hello,
I wanted to bring attention changes in kernels <= 2.6.13 and >=
2.6.14
Used to for atapi ata support with sata optical drives libata.h had
to have ATA_ENABLE_ATAPI defined.
In 2.6.14 and later the support is within the kernel now and a kernel
option of atapi_enabled=1.
drivers/scsi/libataa-core.c has atapi_enabled by default as integer 0
(meaning off). I have put it as one and
compiled a kernel and put it to use with the current 6.1.1-2 x86
livecd and the system recognized the
PX-716SA drive, however, I couldn't get past that, probably because I
was trying a 2.6.14 kernel.
This is just a heads up for future releases of the lfs livecd. I am
sure more sata optical devices will be used
on systems in the future. The hack was used for systems that were
using systems in enhanced mode
where only SATA was being utilized. There is a compatible mode in the
BIOS which lets the sata drives
work with older operating systems just like they are legacy PATA
devices.
A future release which utilizes a 2.6.14 or greater kernel will not
require editing a header and thus, the option
to enable the atapi feature should be noted somewhere when pressing
F1 along with all the vga information.
I would definitely like to do this and so I am asking, what is the
easiest way to introduce the 2.6.14 kernel tree
to the livecd with modules so that it will work? Right now it can't
load the unionfs.
Right now as it is, the current livecd has to be downloaded, a person
has to build a new kernel same as the version
on the livecd, edit libata.h and rebuild then the SATA optical device
can be used and the LFS livecd will find the
CD as the drive is initialized and attached to a block device. The
reason libata.h had the option to define
ATA_ENABLE_ATAPI is because there were still people having system
instabilities. I do not know the direction
the lfs livecd developers want to go with this, and I think it is a
critical step in supporting future hardware.
Sincerely,
William
--
http://linuxfromscratch.org/mailman/listinfo/livecd
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page