RM7900 datasheet boot-mode settings

Post new topic   Reply to topic    CPU-World.com forums Forum Index -> Information Requests (Collectible Chips only)
View previous topic :: View next topic  
Author Message
ruckusman



Joined: 02 Jan 2025
Posts: 6

PostPosted: Fri Jan 03, 2025 9:00 pm    Post subject: RM7900 datasheet boot-mode settings Reply with quote

Hi guys,

Long term project here to upgrade the SGI O2 with an RM7900 CPU - serious OCD...
For years, literally I couldn't find any datasheet on the RM7900, but recently lucked out and found the long form of the RM7965 datasheet here

https://www.datasheets360.com/pdf/-6475857418499999415
(part [subset] of the RM7900 family instruction manual)

My primary interest is in the boot-mode setting bit stream which begins on page 48 of the above document.

Now I believe that the boot-mode bit stream is also applicable to the RM7900 because, on page 48 - field (bit) 10 enables/disables the external cache.
The RM7965 doesn't have an L3 external cache controller, so to my mind that boot-mode bit stream setting is applicable to the RM7900.

Ideally I'd get my hands on this document, however it's nowhere to be found

For additional Detail Information regarding the operation of the PMC-Sierra see the latest PMC-Sierra datasheet for the RM79xx Family Microprocessors Data Sheet (doc. # PMC-2030581), Issue No. 11: September, 2006


Anyone have any thoughts on the matter?

I don't even have the CPU as yet, but need to sort this element out before attempting anything anyway...
Back to top
View user's profile Send private message  
mattst88



Joined: 09 Sep 2010
Posts: 6

PostPosted: Thu May 22, 2025 8:09 am    Post subject: Reply with quote

I have the same plans!

It's been a long time since I looked at any of this, but in 2018 I contacted Microsemi (which acquired PMC-Sierra) and asked for documentation no the RM7900, and they were unwilling to help. Their reply:

Quote:
Thank you for your interest in Microchip's products and services. While Microchip greatly values the teaching and research initiatives of the academic sector, we have very limited ability to support product information requests from universities and other academic institutions. Your access request has been forwarded to our sales group for review, however, if you do not receive further reply from us within 14 days, you will not be granted access to the requested information. Thank you for your understanding and good luck with your project.


I suspect you're exactly right that the info in the RM7965 documentation is applicable for the RM7900. I guess we won't know until we actually try!
Back to top
View user's profile Send private message   Visit poster's website
ruckusman



Joined: 02 Jan 2025
Posts: 6

PostPosted: Sat May 24, 2025 5:39 am    Post subject: Reply with quote

Hiya Matts88

I have fairly copious notes on various potential sources for the info - there were sources that sold dev kits Wind River systems is one such Org.

There are RM7900 spec sheets out there somewhere, but it's true internet quest territory finding one.

Are you on irixnet.org or forums.sgi.sh or their discord channel?
I haven't been on for a few months as I dosed the older laptop with a cup of coffee - the data (passwords) are intact just awaiting transfer - I am lazy -there was some interesting info re: RM7900 and the needed PROM mods AND apparently one of the guys on there knew an ex-SGI engineer that had the know-how to do the mods.
Back to top
View user's profile Send private message  
mattst88



Joined: 09 Sep 2010
Posts: 6

PostPosted: Tue May 27, 2025 5:59 pm    Post subject: Reply with quote

I'm on irixnet.org, but not very active there. I don't believe I've ever created an account on forums.sgi.sh.

In terms of the PROM, I've disassembled it and found the code that checks the CPU type. Of course it doesn't know about RM7900 CPUs, so presumably it would fail there.
Back to top
View user's profile Send private message   Visit poster's website
ruckusman



Joined: 02 Jan 2025
Posts: 6

PostPosted: Thu May 29, 2025 1:30 am    Post subject: Reply with quote

Hi Matts

I knew I recognised your handle from somewhere, I've read your blog about the chat with Jesse Barnes about this very possibility.

Have you acquired the 6.5.7 source with the PROM code?

Given that you've successfully decompiled the PROM, you're further along than I am - if you've seen any of the posts from Chicago-Joe about this, he had a RM7900-835 CPU on a board, it wouldn't POST, BUT before he read mention of using the terminal to monitor the boot progress he'd had it desoldered again.

I think the pieces of the puzzle all exist, they just need to be collected and assembled to make this work.
Back to top
View user's profile Send private message  
mattst88



Joined: 09 Sep 2010
Posts: 6

PostPosted: Thu May 29, 2025 9:27 am    Post subject: Reply with quote

ruckusman wrote:
Have you acquired the 6.5.7 source with the PROM code?


Is that part of one of the IRIX leaks?

In any case, I can't imagine it would be possible to rebuild into a working image.

ruckusman wrote:
Given that you've successfully decompiled the PROM, you're further along than I am - if you've seen any of the posts from Chicago-Joe about this, he had a RM7900-835 CPU on a board, it wouldn't POST, BUT before he read mention of using the terminal to monitor the boot progress he'd had it desoldered again.


I've still got a ton of work to do just to understand the necessary bits of the PROM, and I haven't looked at it in years.

In any case, here's just one example of some code in it checking for CPU PrIDs:

Code:
    19d0:       23000924        li      t1,35                                                                               
    19d4:       00000000        nop
    19d8:       00780840        mfc0    t0,$15                  Read PrID Register                                           
    19dc:       00ff0831        andi    t0,t0,0xff00                                                                         
    19e0:       02420800        srl     t0,t0,0x8               Extract Processor ID                                         
    19e4:       3f000911        beq     t0,t1,0x1ae4            Branch to 0x1ae4 if R5000 (35, 0x23)                         
    19e8:       00000000        nop                                                                                         
    19ec:       28000924        li      t1,40                   Branch to 0x1ae4 if RM52xx (40, 0x28)                       
    19f0:       3c000911        beq     t0,t1,0x1ae4                                                                         
    19f4:       00000000        nop                                                                                         
    19f8:       27000924        li      t1,39                                                                               
    19fc:       6f000915        bne     t0,t1,0x1bbc            Branch to 0x1bbc if not RM7000 (39, 0x27)                   
    1a00:       00000000        nop


I'll see if I can clean up some of the tooling I wrote and post it.

ruckusman wrote:
I think the pieces of the puzzle all exist, they just need to be collected and assembled to make this work.


Yeah, I agree!
Back to top
View user's profile Send private message   Visit poster's website
ruckusman



Joined: 02 Jan 2025
Posts: 6

PostPosted: Fri May 30, 2025 10:37 pm    Post subject: Reply with quote

Hiya Matt,

I've sent you a PM on irixnet - a thread over there to aggregate the knowledge is probably useful.
Back to top
View user's profile Send private message  
Display posts from previous:   
Post new topic   Reply to topic    CPU-World.com forums Forum Index -> Information Requests (Collectible Chips only) All times are GMT - 5 Hours
Page 1 of 1
Jump to:  
You can post new topics in this forum
You can reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You cannot download files in this forum

Powered by phpBB © 2001 phpBB Group