[GTALUG] "Golden Age for Computer Architecture"

[D. Hugh Redelmeier via talk]

This Turing Award Winners talk is quite insigtful.

Full title: "A New Golden Age for Computer Architecture:
Domain-Specific Hardware/Software Co-Design, Enhanced Security, Open
Instruction Sets, and Agile Chip Development"

The speakers are key figures in the RISC revolution.
John L. Hennessy and David A. Patterson

<https://www.acm.org/hennessy-patterson-turing-lecture>

Note: they mean that it is a golden age for researchers. Not so good
for users.
---
Talk Mailing List
***@gtalug.org
https://gtalug.org

[Russell Reiter via talk]

Post by D. Hugh Redelmeier via talk
This Turing Award Winners talk is quite insigtful.
Domain-Specific Hardware/Software Co-Design, Enhanced Security, Open
Instruction Sets, and Agile Chip Development"
The speakers are key figures in the RISC revolution.
John L. Hennessy and David A. Patterson
<https://www.acm.org/hennessy-patterson-turing-lecture>
Note: they mean that it is a golden age for researchers. Not so good
for users.

Interesting stuff thanks.

I just came across this info on the hidden RISC in VIA chipsets now aka the
rosenbridge backdoor.

Apparently it was discovered by reading through some linked patent records.
He has fuzzy tools so you can check your own hardware.

Some wag deemed it a "ring 4" exploit.

https://github.com/xoreaxeaxeax/rosenbridge

The rosenbridge backdoor is a small, non-x86 core embedded alongside the
main x86 core in the CPU. It is enabled by a model-specific-register
control bit, and then toggled with a launch-instruction. The embedded core
is then fed commands, wrapped in a specially formatted x86 instruction. The
core executes these commands (which we call the 'deeply embedded
instruction set'), bypassing all memory protections and privilege checks.

---

Post by D. Hugh Redelmeier via talk
Talk Mailing List
https://gtalug.org/mailman/listinfo/talk

[James Knott via talk]

Post by Russell Reiter via talk
Some wag deemed it a "ring 4" exploit.

Many years ago, I was a computer tech, working on a variety of systems,
including the VAX 11/780 and Data General Eclipse.  Both of those had a
writable control store and the VAX even loaded it's microcode from a
floppy at boot.  One ot the things I had to do, when repairing Eclipse
computers was work my way through the microcode, using a very large
listing and a logic probe, along with the microstep switch on the front
panel.  The Eclipse used 4 AMD 4 bit slice processors and, IIRC, over
100 bits wide ROM to contain the instruction set.  With the WCS, I
expect it would be fairly easy to insert malware.

---
Talk Mailing List
talk

[Russell Reiter via talk]

Post by James Knott via talk

Post by Russell Reiter via talk
Some wag deemed it a "ring 4" exploit.

Many years ago, I was a computer tech, working on a variety of systems,
including the VAX 11/780 and Data General Eclipse. Both of those had a
writable control store and the VAX even loaded it's microcode from a
floppy at boot. One ot the things I had to do, when repairing Eclipse
computers was work my way through the microcode, using a very large
listing and a logic probe, along with the microstep switch on the front
panel. The Eclipse used 4 AMD 4 bit slice processors and, IIRC, over
100 bits wide ROM to contain the instruction set. With the WCS, I
expect it would be fairly easy to insert malware.

Thats an interesting paradigm shift.

Searching through the physical setup using logic tools old school, vs new
school; searching through patent records just to be able to find the
hardware.

Post by James Knott via talk
---
Talk Mailing List
https://gtalug.org/mailman/listinfo/talk