Dave Jones <davej@xxxxxxxxxx> writes:
> On Fri, Jun 24, 2005 at 12:43:29PM -0700, Wolfgang S. Rupprecht wrote:
> > My motherboard, a tyan k8e (s2865) has entries in the bios for
> > something called:
> >
> > S/W memory hole Remapping disable / enable
> > H/W memory hole Remapping disable / enable
> > MTRR mapping mode continuous / discrete
> >
> Hmm, usually such an option does the trick.
Ok, I decided to see what booting in single user would show me.
turning on the "H/W remapping" and the "mtrr => discrete" did allow
the last 512MB to show up in /proc/mtrr.
The problem is that my PCI ethernet card, a linksys eg1032 (syskonnect
chipset) refuses to work. I wonder if the kernel is handing it memory
above 4GB and it can't deal with this.
What is disconcerting is that I used to have 5 mtrr regions. The last
one was deleted to make room for the last dram hunk.
(Normal mode when everything is working)
hw off, sw off:
reg00: base=0x00000000 ( 0MB), size=2048MB: write-back, count=1
reg01: base=0x80000000 (2048MB), size=1024MB: write-back, count=1
reg02: base=0xc0000000 (3072MB), size= 512MB: write-back, count=1
reg03: base=0xdff00000 (3583MB), size= 1MB: uncachable, count=1
reg04: base=0xfc000000 (4032MB), size= 8MB: write-combining, count=1
with various remapping switches on:
hw on, sw off:
reg00: base=0x00000000 ( 0MB), size=2048MB: write-back, count=1
reg01: base=0x80000000 (2048MB), size=1024MB: write-back, count=1
reg02: base=0xc0000000 (3072MB), size= 512MB: write-back, count=1
reg03: base=0xdff00000 (3583MB), size= 1MB: uncachable, count=1
reg04: base=0x100000000 (4096MB), size= 512MB: write-back, count=1
hw on, sw on:
reg00: base=0x00000000 ( 0MB), size=2048MB: write-back, count=1
reg01: base=0x80000000 (2048MB), size=1024MB: write-back, count=1
reg02: base=0xc0000000 (3072MB), size= 512MB: write-back, count=1
reg03: base=0xdff00000 (3583MB), size= 1MB: uncachable, count=1
reg04: base=0x100000000 (4096MB), size= 512MB: write-back, count=1
hw off, sw on
reg00: base=0x00000000 ( 0MB), size=2048MB: write-back, count=1
reg01: base=0x80000000 (2048MB), size=1024MB: write-back, count=1
reg02: base=0xbff00000 (3071MB), size= 1MB: uncachable, count=1
reg03: base=0x100000000 (4096MB), size=1024MB: write-back, count=1
With 3.5G active, hw off, sw off (normal mode when everything is working)
eth1 Link encap:Ethernet HWaddr 00:0C:41:EB:C4:20
inet addr:192.83.197.8 Bcast:192.83.197.127 Mask:255.255.255.128
inet6 addr: 2001:5a8:4:7d1:20c:41ff:feeb:c420/64 Scope:Global
inet6 addr: fe80::20c:41ff:feeb:c420/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:166 errors:0 dropped:0 overruns:0 frame:0
TX packets:149 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:23681 (23.1 KiB) TX bytes:20873 (20.3 KiB)
Interrupt:50 Memory:fdff8000-0
With 4G active, hw on or sw on or both:
eth1 Link encap:Ethernet HWaddr 00:0C:41:EB:C4:20
UP BROADCAST RUNNING MULTICAST DYNAMIC MTU:1500 Metric:1
RX packets:11 errors:0 dropped:0 overruns:0 frame:0
TX packets:7 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:660 (660.0 b) TX bytes:1858 (1.8 KiB)
Interrupt:50 Memory:fdff8000-0
I wonder if the reg04 mtrr at 0xfc000000 => 0xfc800000 is used for
anything that would effect my network card or if the network card is
being handed memory above 4G which it can't use over the PCI bus.
> Is there a newer BIOS you can try perhaps ? It's not really a
> problem that we can do anything about sadly.
Not yet. I'll keep an eye out though.
-wolfgang
--
Wolfgang S. Rupprecht http://www.wsrcc.com/wolfgang/
