No, AMD Will Not Be Opening Up Its Firmware/Microcode

[bridgman]

Bridgman does not believe me wen i tell him that in the end we will have full opensource hardware down to the firmware and if it is not AMD hardware then we will have a solution WITHOUT AMD...

he believes that we the FLOSS community can not do this but in fact he will be wrong.

You have an uncanny ability to be consistently wrong about what I believe. If you were guessing you would be right more often, so you must have some good insight but just a little problem with the interpretation.

Open source hardware right down to the firmware has existed for a long time (at least 30 years)... it just tends to fall behind after a while because nobody is pouring R&D money into it on an ongoing basis. The question is where the big R&D money is going to come from if all the hardware is open source and prices get decoupled from R&D costs.

I do think the industry will reach a "good enough" point where a lot of users can live without that big ongoing R&D investment, but I don't think the GPU market is there yet. The CPU market is closer, obviously, and as others have pointed out we are already involved in some of those initiatives.

[bibaheu]

Sorry for changing the topic, but is still AMD publishing the internal documentation for the Radeon GPUs? I remember reading once a very long PDF about the internal registers and atomBIOS stuff.

[libv]

Sorry for changing the topic, but is still AMD publishing the internal documentation for the Radeon GPUs? I remember reading once a very long PDF about the internal registers and atomBIOS stuff.

They did so because we (egbert eich and I) wrote that into our proposal for an open source strategy (which was posted on phoronix many years ago). Even then, ATI provided the bare minimum of what was written in the proposal, and what AMD commanded ATI to do, and hoped that we would fail (good thing that the one guy with a clue of display engines at the time was around to do the work). ATI of course stopped doing this at the earliest possible time, which was when AMD ran out of cash in the credit crunch, sold off its fabs, and killed, amongst others, the RadeonHD project (which at that point needed to die, and i state that about my own baby). Due to the graphics cards having more of a market share and higher margins than the CPUs/chipsets, ATI slowly gained more and more political control inside the AMD/ATI conglomorate during the RadeonHD project lifetime, and this also played into the whole political and technological mess. The credit crunch was just the final straw that broke AMDs back.

Afaik, ATI has not resumed this level of documentation, they actually have only further increased their dependence on extra layers and encapsulation with atomBIOS and DAL.

Intel started doing documentation because of this (it can well be argued that intel started doing BIOS free display drivers because of me as well) and kept it up for a long time, i do not currently know whether this is still true today, i know for a fact that intel was still producing display engine register level documentation in 2012.

[dungeon]

Sorry for changing the topic, but is still AMD publishing the internal documentation for the Radeon GPUs? I remember reading once a very long PDF about the internal registers and atomBIOS stuff.

They do it al the time, look here GCN3 ISA updated now in august

http://developer.amd.com/resources/developer-guides-manuals/

[libv]

They do it al the time, look here GCN3 ISA updated now in august

http://developer.amd.com/resources/d...uides-manuals/

ISA is the instruction set architecture of the shaders. It is a far cry from register level display, media, CP, etc, documentation.

Also, before the RadeonHD project had started, AMD had formed a GPGPU team with handpicked people from AMD and ATI. This team was not as ATI centric politically. They needed the ISA to become public, and at some point, in an attempt to go around Mr Bridgman we were brought into contact with them. Sadly, in this call, in January 2008, Mr Bridgman did manage to be present, and the whole thing was a major waste of time, again. The one hope i still had vanished then. This was the one chance to get a proper communication channel into AMD and ATI, one that would actually be helpful and would not play a double game, come up with excuses all the time, and would not filibuster each phone conference tpo death. That call is when RadeonHD died for me.

But at least the ISA releases kept on coming over the years. AMDs tiger team approach did work here.

[dungeon]

Well that is OK, even when docs are there something is not supported.

I readed once BKDG from there, i just wanted sensors to allow somehow to reading watts on Kabini... but seems fam15h_power driver is dumb simple to support reading that there

[libv]

Well that is OK, even when docs are there something is not supported.

I readed once BKDG from there, i just wanted sensors to allow somehow to reading watts on Kabini... but seems fam15h_power driver is dumb simple to support reading that there

Yes, you get that, or used to get that, for most of the AMD hw. You only got some of that sort of information for just a small time window for ATI hardware, and only so because we, those evil corporate microsoft drones (if you believe the people who forked radeonhd), requested this at the right time, in the right way, from the right people (being AMD, the party of the AMD/ATI conglomerate which wanted open source drivers).

[coder111]

Is there bugreport about this? For example mine 6650M is fine.

Hi,

Yes, there is: https://bugzilla.kernel.org/show_bug.cgi?id=51381

And Alex added code to the kernel to blacklist this specific card by PCI ID and disable dynamic power management for it. (Thanks to Alex for quick response by the way). Which made my system usable, but still I wish dynamic PM worked. From time to time I recompile the kernel and remove my card from the quirks blacklist just to see if the error is still there. It is still there...

My card is in an Asus laptop. What make is yours?

[bibaheu]

They do it al the time, look here GCN3 ISA updated now in august

http://developer.amd.com/resources/d...uides-manuals/

Thanks, there is actually quite interesting stuff there!

Afaik, ATI has not resumed this level of documentation, they actually have only further increased their dependence on extra layers and encapsulation with atomBIOS and DAL.

It seems that documents for Volcanic Islands and Polaris are missing, and for the others, I didn't find any mention of atomBIOS and outputs, just 3D engine and the GCN ISA.

I hope they don't move too much output control into the firmware; I remember from your blog (I think) the story about the old Intel GPUs that wouldn't work correctly until the Windows driver had set up the outputs correctly. Probably that i915something package that was required to get proper resolutions on my laptop was related to the firmware thing. I don't want to go back to that, I think that program destroyed the laptop's internal panel... happily it was under warranty

[libv]

coder111 and bibaheu: it's too late, in both cases. That ship sailed a long time again, and feel free to thank Mr Bridgman and the people who forked RadeonHD for it.