On 22.02.23 20:50, Limonciello, Mario wrote:
Richard and I both sent out relatively similar patches in the past, but they never went anywhere.
The problem here is that the default will never satisfy a great majority. just a majority. So if we admit that there is no optimal solution, why not pick the tested, that is, current solution, as long as it does least?
We did confirm that Windows does set a similar policy as well though, which is what prompted us to attempt this. As there is more interest again maybe we can revive that discussion and merge together some ideas from the sets of patches.
Well, the most obvious question, as this can also be done with a udev rule, is, what are the risks of a change?
Previous submissions: v4: https://lore.kernel.org/linux-usb/20220825045517.16791-1-mario.limonciello@xxxxxxx/
This one at least restricts it to some cases. And as much as it pains me, we need to ask whether there are risks in not emulating Windows. And on the other hand, what happens if we do this. Are you sure, for example, that you do not break use cases for S4 with this change? If a laptop goes into S4 because power is too low, we really do not want it to wake up just because somebody accidentally hits a mouse button. Regards Oliver
