Intel Developing Universal Scalable Firmware As Next-Gen Firmware Platform

[phoronix]

Phoronix: Intel Developing Universal Scalable Firmware As Next-Gen Firmware Platform

Intel passed along news today of their development efforts around Universal Scalable Firmware, a new initiative they are pursuing to simplify and scale firmware development for hardware from edge computing devices to the cloud...

https://www.phoronix.com/scan.php?page=news_item&px=Intel-USF-Firmware

[avph]

This is joke in very poor taste by Intel. It's basically calling what SlimBootloader, an Intel project, does a 'universal spec' and then expecting other firmware solutions to implement the same interface. This is just Intel forcibly wanting control over firmware solutions. Renaming existing and well established open source firmware solutions like u-boot and coreboot as Platform Orchestration Layers is simply demeaning.

Besides that the solutions proposed by the spec are also just bad from a technical and security point of view. Scalable FSP == how to the role of open source code in firmware to a minimum.

• It proposed that even the first instruction run by the CPU is located inside a blob??
• SMM inside FSP == closed source code that at runtime can do whatever it wants at the highest possible level of privilege? Your OS won't even know that Intel has bugs inside its SMM code wreaking havoc (and if you ever had a look at Intel UEFI code, you can be sure there will be plenty!!)
• ...

[milkylainen]

Besides that the solutions proposed by the spec are also just bad from a technical and security point of view.

Well. Not only that. Beside the horrendous security, it's usually designed by a committee and over-complicated to no end.
No Intel 4-letter acronyms named here, none forgotten either...

[Slartifartblast]

Remember this ?

From: Bill Gates
Sent: Sunday, January 24, 1999 8:41 AM
To: Jeff Westorinen; Ben Fathi
Cc: Carl Stork (Exchange); Nathan Myhrvold; Eric Rudder
Subject: ACPI extensions

One thing I find myself wondering about is whether we shouldn’t try and make the “ACPI” extensions somehow Windows specific.

It seems unfortunate if we do this work and get our partners to do the work and the results is that Linux works great without having to do the work.

Maybe there is no way to avoid this problem but it does bother me.

Maybe we could define the APIs so that they work well with NT and not
the others even if they are open.

Or maybe we could patent something related to this.

[Developer12]

Looks to me like intel simply trying to push everyone up the software stack and out of their business.

They already have lots of guarded domains to themselves, like the ME and the FSP's early silicon init, or arguably SMM. This just implements an API that acts as a territorial boundary and seals off lots of other low-level details in addition to the early boot process. Everyone else then just gets a shim role in the middle of the stack, adapting the intel interface to legacy BIOS/UEFI/coreboot payload/linux boot/whatever. A "library" as intel put it.