PCIe, power management, and problematic BIOSes

Posted Jul 2, 2011 18:54 UTC (Sat) by mjg59 (subscriber, #23239)
In reply to: PCIe, power management, and problematic BIOSes by Lennie
Parent article: PCIe, power management, and problematic BIOSes

That's usually how things are done in the embedded world. In PC-land, the BIOS abstracts away a lot of the very low-level hardware. Slimming down the BIOS implies the kernel having to learn a lot more about the underlying hardware than it currently knows.

PCIe, power management, and problematic BIOSes

Posted Jul 3, 2011 20:15 UTC (Sun) by elanthis (guest, #6227) [Link] (1 responses)

I think some of the arguments made by Linus and co. was that the hardware interfaces should just be more standardized. A pipe dream for sure, but it would certainly solve a lot of the problems.

A further problem is a lack of conformance tests. This is the same flaw OpenGL suffers from, and is the primary reason why nobody in their right mind uses OpenGL unless they're forced to (e.g. because they're targetting non-Windows OSes, none of which have bothered to replace OpenGL with something not designed by amateur C hackers in the 1980's).

PCIe, power management, and problematic BIOSes

Posted Sep 28, 2016 14:10 UTC (Wed) by Hi-Angel (guest, #110915) [Link]

In part this was a Khronos problem, and mind you, DirectX a long time have been a much worse API than OpenGL was. There's a really great article on the matter http://programmers.stackexchange.com/a/88055/148803