1 |
After several nights of hardware configuration issues and Gentoo/doc |
2 |
issues, she's up and running. Woo-hoo! My system is: |
3 |
|
4 |
Miata 500au 500Mhz |
5 |
1GB Mem (1x512, 2x256) |
6 |
DEC/Symbios Logic 53c875 dual-channel FW SCSI card |
7 |
2x4.3GB RZ29B (Seagate ST15150W) drives |
8 |
LG CRD-8322B 32x IDE CDROM (Orig 32x Toshiba XM-6102B swapped out in |
9 |
testing bad LiveCDs) |
10 |
DLT7000 bare tape drive (not installed yet). |
11 |
|
12 |
Here's some of the installation notes I've made (pasted from my local Wiki): |
13 |
|
14 |
* The 1.4-test4 LiveCD does not work on a Miata. |
15 |
|
16 |
* The 1.4-test3 LiveCD continually causes seg faults and other errors |
17 |
when copying files from the CD to RAMdisk during the boot. In order to |
18 |
possibly get the CD to boot after several tries: |
19 |
1. Alpha must be at the SRM prompt (">>>"). |
20 |
2. Make sure no commands have been entered since the last |
21 |
init/power on. If there are, init or warmstart the Alpha. |
22 |
3. show dev |
23 |
4. b dka0 -flags 4 |
24 |
5. At the aboot prompt, l |
25 |
6. Enter "0" at the aboot prompt to boot off of the first entry |
26 |
listed. |
27 |
7. The file copy to RAM will still cause errors, but it should |
28 |
ignore them. |
29 |
8. If it still fails, start over making sure to init/warmstart |
30 |
first. |
31 |
9. Not sure why this sometimes works (serendipity is a wonderful |
32 |
thing), but it did for me... |
33 |
|
34 |
* Save the headaches and grab the Gentoo2004 LiveCD from the |
35 |
experimental directory on the Gentoo mirrors. It worked without a |
36 |
hitch. |
37 |
|
38 |
* The stage1 ISO in the "mirrors" directory will not work. The |
39 |
portage package is hosed and I could not make the reported patches |
40 |
work. Instead grab |
41 |
http://gentoo.chem.wisc.edu/gentoo/experimental/alpha/stages/stage1-alpha-20040730.tar.bz2 |
42 |
(substitute your local mirror in the URL) |
43 |
|
44 |
* CFLAGS="-mieee -O3 -pipe -mcpu=ev56" |
45 |
|
46 |
* Kernel |
47 |
o gentoo-dev-sources (2.6.8-gentoo-r1) |
48 |
o ACCEPT_KEYWORDS="~alpha" emerge -v gentoo-dev-sources |
49 |
|
50 |
* aboot |
51 |
o Package "aboot" (v0.9) will error out trying to compile. |
52 |
o Grab 1.0 instead |
53 |
o ACCEPT_KEYWORDS="~alpha" emerge -v aboot |
54 |
|
55 |
* DO NOT USE QLOGIC ISP1020 SCSI CARD! There are many folks on various |
56 |
lists who have complained that the driver causes FS corruption under |
57 |
2.6 because of a lack of proper porting. |
58 |
|
59 |
I've also noticed several (all?) aboot docs saying that "dqa0" denotes the |
60 |
IDE device 0, but it's not true for all Alphas, including the Miata. OK, |
61 |
at least *my* Miata has the IDE CDROM as device "DKA0". My two SCSI |
62 |
drives are "DKA0" and "DKA100" (SCSI ID 0 and 1, respectively). |
63 |
|
64 |
That's all. I've had a helluva (in a wierd way, fun) time getting this to |
65 |
work. If I hadn't done several Intel/AMD Gentoo installs and if I hadn't |
66 |
worked with DEC stuff for 10 years in a previous life, I don't know that I |
67 |
could have pulled this off. There's several little gotchas whose answers |
68 |
could not be found in Google, but took a little experimentation, |
69 |
guesswork, and experience. I guess I'm hoping that someone else doing a |
70 |
Google search on the same stuff I ran into would have one-stop shopping. |
71 |
|
72 |
And now back to my 80MB emerge, already in progress. :) |
73 |
|
74 |
Rich J. |
75 |
|
76 |
|
77 |
-- |
78 |
gentoo-alpha@g.o mailing list |