1 |
On 2017-05-15 14:33, Corbin Bird wrote: |
2 |
|
3 |
> Gigabyte has long been known to ship 'broken for Linux' e820 firmware. |
4 |
> 'e820' is basically the APG Aperature setup in the firmware. |
5 |
> Gigabyte kills OR overrides the IOMMU support in the firmware, to setup |
6 |
> the AGP Aperature. |
7 |
> |
8 |
> Closed source / binary video drivers WANT an AGP Aperature. |
9 |
|
10 |
So is _that_ what it's all about? Is the default assumption that I run |
11 |
non-free video drivers? Why? I _never_ use these, and if you bet on me |
12 |
using them in the future and win the bet, you can probably make a cool |
13 |
million or two. |
14 |
|
15 |
> The kernel wants an IOMMU on x86_64, as an IOMMU. |
16 |
|
17 |
But it has one! The GART_IOMMU. It was even added specifically for |
18 |
Linux, according to kernel.org discussions I remember seeing but can't |
19 |
locate now. |
20 |
|
21 |
I also found [1] , which I think is the clearest and most human-oriented |
22 |
explanation of this issue, yet. And ... the recommendation is pretty |
23 |
much the opposite of yours. The ball is in your court, sir. |
24 |
|
25 |
> > CONFIG_X86_X2APIC=y |
26 |
|
27 |
I did this long ago, and it has no effect - still only one is found, |
28 |
according to dmesg. As I wrote in the earlier thread. |
29 |
|
30 |
Also, the Kconfig help for this option says it does something quite |
31 |
different. |
32 |
|
33 |
> > CONFIG_X86_MPPARSE=y |
34 |
|
35 |
Ok, this one I have not set so far. Will read up on it and try, |
36 |
|
37 |
[1] |
38 |
http://bogdan.org.ua/2009/09/30/iommu-this-costs-you-64-mb-of-ram.html |
39 |
|
40 |
-- |
41 |
Please *no* private Cc: on mailing lists and newsgroups |
42 |
Personal signed mail: please _encrypt_ and sign |
43 |
Don't clear-text sign: |
44 |
http://primate.net/~itz/blog/the-problem-with-gpg-signatures.html |