Re: [PATCH] x86_32: trim memory by updating e820 v3

[Gabriel C <nix.or.die@googlemail.com>]

Andrew Morton wrote:
> On Tue, 22 Jan 2008 16:23:20 -0800 Yinghai Lu <Yinghai.Lu@Sun.COM> wrote:
> 
>> [PATCH] x86_32: trim memory by updating e820 v3
>>
>> when mtrr is not covering all e820 table, need to trim the ram, need to update e820
>>
>> reuse some code for x86_64
>>
>> here need to add early_get_cap and use it in early_cpu_detect, and move mtrr_bp_init early
>>
>> need Justine to test with his special system with bug bios.
>>
>> Signed-off-by: Yinghai Lu <yinghai.lu@sun.com>
> 
> Speaking of mtrr and e820....
> 
> Could someone please take a peek at
> http://bugzilla.kernel.org/show_bug.cgi?id=10508 ?
> 
> For some reason we seem to have turned this:
> 
> [    0.000000]  BIOS-e820: 0000000000000000 - 000000000009f800 (usable)
> [    0.000000]  BIOS-e820: 000000000009f800 - 00000000000a0000 (reserved)
> [    0.000000]  BIOS-e820: 00000000000dc000 - 0000000000100000 (reserved)
> [    0.000000]  BIOS-e820: 0000000000100000 - 00000000bf6d0000 (usable)
> [    0.000000]  BIOS-e820: 00000000bf6d0000 - 00000000bf6e3000 (ACPI NVS)
> [    0.000000]  BIOS-e820: 00000000bf6e3000 - 00000000c0000000 (reserved)
> [    0.000000]  BIOS-e820: 00000000e0000000 - 00000000f0000000 (reserved)
> [    0.000000]  BIOS-e820: 00000000fec00000 - 00000000fec10000 (reserved)
> [    0.000000]  BIOS-e820: 00000000fed00000 - 00000000fed00400 (reserved)
> [    0.000000]  BIOS-e820: 00000000fed14000 - 00000000fed1a000 (reserved)
> [    0.000000]  BIOS-e820: 00000000fed1c000 - 00000000fed90000 (reserved)
> [    0.000000]  BIOS-e820: 00000000fee00000 - 00000000fee01000 (reserved)
> [    0.000000]  BIOS-e820: 00000000ff000000 - 0000000100000000 (reserved)
> [    0.000000]  BIOS-e820: 0000000100000000 - 0000000140000000 (usable)
> 
> into this:
> 
> reg00: base=0xc0000000 (3072MB), size=1024MB: uncachable, count=1
> reg01: base=0x00000000 ( 0MB), size=4096MB: write-back, count=1
> reg02: base=0x100000000 (4096MB), size=1024MB: write-back, count=1
> reg03: base=0xbf700000 (3063MB), size= 1MB: uncachable, count=1
> reg04: base=0xbf800000 (3064MB), size= 8MB: uncachable, count=1
> 
> which screws up the X server's attempt to map the video memory at
> 0xd0000000.

I see that on my box with ASUS P5E-VM DO and 4G RAM.
It has an Q35 intel chipset and the intel card is set to use 256MB in BIOS.

I tested 2.6.{24*,25,25-next,25-latest-git} 32/64 bit.

Also I get the mtrr type mismatch message on 32 and 64 bit kernels.

...

[    0.000000] BIOS-provided physical RAM map:
[    0.000000]  BIOS-e820: 0000000000000000 - 000000000009ec00 (usable)
[    0.000000]  BIOS-e820: 000000000009ec00 - 00000000000a0000 (reserved)
[    0.000000]  BIOS-e820: 00000000000e4000 - 0000000000100000 (reserved)
[    0.000000]  BIOS-e820: 0000000000100000 - 00000000cf550000 (usable)
[    0.000000]  BIOS-e820: 00000000cf550000 - 00000000cf55e000 (ACPI data)
[    0.000000]  BIOS-e820: 00000000cf55e000 - 00000000cf5e0000 (ACPI NVS)
[    0.000000]  BIOS-e820: 00000000cf5e0000 - 00000000cf600000 (reserved)
[    0.000000]  BIOS-e820: 00000000fee00000 - 00000000fee01000 (reserved)
[    0.000000]  BIOS-e820: 00000000ffc00000 - 0000000100000000 (reserved)
[    0.000000]  BIOS-e820: 0000000100000000 - 000000012c000000 (usable)
...

cat /proc/mtrr :

reg00: base=0xd0000000 (3328MB), size= 256MB: uncachable, count=1
reg01: base=0xe0000000 (3584MB), size= 512MB: uncachable, count=1
reg02: base=0x00000000 (   0MB), size=4096MB: write-back, count=1
reg03: base=0x100000000 (4096MB), size= 512MB: write-back, count=1
reg04: base=0x120000000 (4608MB), size= 128MB: write-back, count=1
reg05: base=0x128000000 (4736MB), size=  64MB: write-back, count=1
reg06: base=0xcf600000 (3318MB), size=   2MB: uncachable, count=1


dmesg is saying :

..

[   24.012694] [drm] Initialized i915 1.6.0 20060119 on minor 0
[   24.175500] mtrr: type mismatch for d0000000,10000000 old: write-back new: write-combining
[   24.423740] set status page addr 0x00033000

..

from Xorg log :

...

(WW) intel(0): Failed to set up write-combining range (0xd0000000,0x10000000)
(II) intel(0): vgaHWGetIOBase: hwp->IOBase is 0x03d0, hwp->PIOOffset is 0x0000

...


Regards,

Gabriel