This Is What Started AMD's Open-Source Strategy

[AnonymousCoward]

i think its hopeless to arguing to every one in the way that everyone can not understand why there is no working dynpm as default.

Why is there no working dynpm by default? Because we are all waiting for you to write it.

Dynamic power management is complicated stuff so this code has to be written by a highly skilled professional who knows what he's talking about. And since you've clearly demonstrated that you know better than everyone else (bridgman included), it's only logical that you write the dynpm code.

You're the best man for this job Q, do not disappoint us. I'll be bugging you from time to time for status updates.

[Qaridarium]

Why is there no working dynpm by default? Because we are all waiting for you to write it.

Dynamic power management is complicated stuff so this code has to be written by a highly skilled professional who knows what he's talking about. And since you've clearly demonstrated that you know better than everyone else (bridgman included), it's only logical that you write the dynpm code.

You're the best man for this job Q, do not disappoint us. I'll be bugging you from time to time for status updates.

LOL ;-) sure but i never write programs with more than 3 lines of code (copied out of a handbook)

but yes everyone should try his best to clear this "ugly" situation.

[Qaridarium]

Q, you either believe him or you don't. If you think he's lying about the problems, then even if he gave you a link you'd just think he faked that discussion or told agd5f to make up that log. If you don't think he's lying, then why not just take him at his word and if you really want to look into the problems yourself then do the search for it rather than making it sound like you think he's lying?

i do not write such a stuff and i don't think he is a liar.

i just follow rules of rhetoric style.

and i will not search in the hole universe only he claim that there is the source of his claims.

[del_diablo]

smitty3268: Overheating is the largest inconvinience a card can have. Why is that not a valid complaint?

[bridgman]

Is overheating a larger inconvenience than crashing or having an unreadable display ?

If the driver defaults to "default" without power savings then the user can set a lower power mode. If the driver defaults to a power management mode which makes a system unuseable then the user's options are more limited. It's really that simple.

When the driver reaches the point that power savings can be turned on by default without causing problems for a non-trivial number of users then I think you can safely assume that the default will be changed.

[del_diablo]

If the card needs to be in "overheating mode" in order to not choke itself to death, or not randomly crash, then its useless ain't it?

[bridgman]

For the relatively small set of systems where both of the following apply ...

(a) crash or don't display with *any* power saving mode enabled
(b) don't have a cooling solution capable of running with default settings

... then the open source drivers would not be a good solution and the proprietary driver would have to be used for now. It doesn't mean the card is useless.

The same would apply in any case where none of the power savings modes actually reduced power *and* the cooling solution wasn't capable of running at default settings... don't think I have seen any of those yet but I have seen one recently where the chip ran at a reasonable temperature but the fan noise was high enough to annoy the user.

Note that the problem is not so much "for a single system no power savings mode works" as "there isn't a single power savings mode which works on *all* systems". The former relates to useability of a single system, while the latter relates to whether one power savings mode can be picked as default (rather than defaulting to "default" and having the user pick a mode which works for their system).

BTW it's really important to distinguish between discussion about "should the current driver turn power savings on by default ?" with discussion about "will power management improve over time to the point where it can be turned on by default ?".

I suspect that people are confusing the two questions, and thus misinterpreting my statement that the driver should probably not enable the current PM code by default as a statement that PM is not going to improve over time. That's only a guess, but I don't remember ever having to answer the same question so many times in a single thread before.

[Figueiredo]

Fellows,

Brigdman, I understand perfectly your explanation. IMO this comes down to one of the same problems Linux faces again and again. There is no graphical interface that allows a newbie user to swap between these PM modes. If there was such a tool, i believe there wouldn't be so much fighting between the better default. Just my 2 cents.

Regards

[RussianNeuroMancer]

Brigdman, but how all such problems solved in properietary driver? Isn't you can just look at already existing solution and then use it in open source driver?

[bridgman]

The power management code in the proprietary driver is huge - maybe 4x the size of the entire open source graphics stack. It's not written so that you can just take pieces out of it, unfortunately.

We realized a couple of years into the program that power management was going to be a big task, so we set an initial goal of making sure that drivers allowed users to trade off power draw vs performance, using hardware that was relatively common from one generation to the next. That's what the current PM code does, along with an initial attempt at dynamic power switching. There are a number of improvements which could be made using information that is already publicly available, but there hasn't been a lot of community interest in doing that work.

We probably need 10x the current effort to get to the next level. We have started some work internally, but it will take time.

In the meantime, previous posts suggest that Qaridarium has been tasked with improving the current power management code