[CLUE-Tech] Kernel compiling questions

[Mike Staver]

Yeah, I actually like Suse as well - it's just that this is a server at
work, and well, I have to use Red Hat for several reasons.  The main one
being that I know it 10x better than any other distro.  The second
reason being that I know Macromedia's ColdFusion works with Red Hat. 
They say it runs on other distros, but they don't really provide support
for those.  After paying thousands of dollars for the enterprise
versions of CF 5 and 6, I'd hate to lose their support because I picked
the wrong distro :)  

Back to the patch - here is exactly what they say to get it running:

----------------------------------------------
OK, if you want to do it: Get the driver distribution dc395-1??.tar.gz.
Unpack it:
> tar xvzf dc395-1??.tar.gz
Copy the driver files to the linux source tree:
> cp -p dc395/dc395x_trm.? /usr/src/linux/drivers/scsi/
> cp -p dc395/README.dc395x /usr/src/linux/drivers/scsi/
and apply the appropriate patch to your kernel source tree. Depending on
your kernel version, this is dc395-integ20.diff, dc395-integ22.diff or
dc395-integ24.diff. (Replace XX by 20, 22, 23 or 24)
> patch -p1 -d /usr/src/linux <dc395/dc395-integXX.diff
The patches have been built against a 2.0.38, a 2.2.13, a 2.3.48
and a 2.4.6 kernel.
Note that you might get rejections, if you use different kernels, but
those
should be easy to fix manually.
Look out for the .rej files (find . -name "*.rej")
Note: If you use a SuSE kernel source tree from SuSE 6.2 or later, you
don't
need to apply the kernel patch, as those come with some version of the
dc395x driver. SuSE 6.2/6.3 kernels lack the init/main.c part of the
patch,
though, so you can't use boot time params for a DC315, then. Of course,
you 
can isolate the diff part to init/main.c and apply it to make this work
...   

Now you have to reconfigure your kernel (make oldconfig or make
menuconfig),
enable the DC395x driver (compiled into the kernel or as a module; note
that
you need to enable CONFIG_EXPERIMENTAL to be able to select the DC395
driver; of course you need it in the kernel, if your root fs resides on
a
partition connected to the DC395 and you don't do advanced INITRD
tricks)
and recompile your kernel and your modules:
> cd /usr/src/linux
> make oldconfig  (enable CONFIG_SCSI_DC395x_TRMS1040) 
> make bzlilo
> make modules
> make modules_install
> depmod -ae

NEW: You may be able to skip compilation of the kernel by just using the
Makefile. For 2.4 kernels, if you compile outside the kernel source
tree,
you don't even need to apply a patch to your kernel any more.

I will provide a complete patch which makes kernel integration easier
and
will ask to apply it to the main tree, as soon as I got enough reports
about
the stability of the driver.
Currently, there are still too many problems left :-(
----------------------------------------------------------------

So, it simply sounds like it's not in the main tree because you can lose
data.  OH well, this is simply for one thing on my machine - mounting a
SCSI cdrom device, actually many of them.  So, losing data isn't a
concern I have.  I was going to mount some hard drives and stuff, but
for the time being, I'm happy just reading cdrom files.  The problem for
me with menuconfig is that I do enable that experimental part, and I
still don't see it anywhere.  I followed these directions to the letter,
except for the part where I'd configure the kernel via command line. 
So, I guess I was just curious if all experimental options were in a
specific part of that menu, or thrown in where ever they are relevant.  


Lynn Danielson wrote:
> 
> Dave Price wrote:
> > ... Lynn Danielson wrote:
> >>In the SuSE kernel "Tekram DC395/U/UW and DC315/U SCSI support" shows up
> >>just after "Symbios 53c416 SCSI support" and just before "Tekram DC390(T)
> >>and Am53/79C974 SCSI support".  I hate to rub it in, but no patching was
> >>required.
> >
> > debian's 2.4.18 only has: Tekram DC390(T) and Am53/79C974 SCSI support
> 
> SuSE modifies their kernel a fair bit.  Not everybody likes this and in
> one instance I found that a stock Linux kernel worked better for me. (I
> was compiling in support for a TV card if I recall correctly.)  What I
> typically like about the SuSE kernel is that they tend to include drivers
> that haven't made it into the lastest stable kernel version.  I'm not
> sure whether this is because the kernel developers don't think the code
> is ready for prime time yet or if haven't had time to test the code.
> SuSE works on developing their own drivers, especiall X servers, which
> they release community after they've included it in their product for a
> period of time.
> 
> But back to the Mike's original question.  I wasn't trying to be flippant
> in my response.  I would expect the driver to show up in the SCSI low
> level driver section.  In my .config file the parameter is
> CONFIG_SCSI_DC395x_TRMS1040=m for compiling in modular support.
> 
> Looking in the ./drivers/scsi directory I find a README file which
> describes the dc395x driver as beta, "I would currently not suggest to
> use this driver, if you are dealing with important data, which you don't
> have a backup of! YOU RISK DATA LOSS!"  Mike probably got the same
> readme with his patch, so I'll assume he's OK with that.
> 
> Looking in my ./drivers/scsi/Makefile I see the following line:
> obj-$(CONFIG_SCSI_DC395x_TRMS1040)      += dc395x_trm.o
> I would expect Mike to have a similar line in his scsi/Makefile.
> 
> Hope this helps,
> 
> Lynn
> 
> _______________________________________________
> CLUE-Tech mailing list
> CLUE-Tech at clue.denver.co.us
> http://clue.denver.co.us/mailman/listinfo/clue-tech

-- 
                                 -Mike Staver
                                  mstaver at globaltaxnetwork.com
                                  http://www.fimble.com/staver
                                  Home: 303-306-2303
                                  Work: 303-894-3818
                                  ICQ: 45682755