Re: Linux-2.4.0-test2

[werner darkwing uoregon edu]

Ivan Kokshaysky wrote:

> On Sat, Jun 24, 2000 at 02:56:18PM -0700, Richard Henderson wrote:
> > "Fixed" thus.  Upgrade to aboot 0.7 for srm.  I don't know if
> > any version of MILO honors the load address properly.
> >
> Milo-2.2 does. Though it puts command line in the wrong place
> and passes incorrect memory map. I've almost finished a fix.
>
> But there is another problem using the new start address - currently
> it's just a waste of 5Mb of RAM on anything except Wildfire.
> Is it possible to utilize a memory not used by console?
>
> Ivan.
>
> --
> To unsubscribe: send e-mail to axp-list-request@redhat.com with
> 'unsubscribe' as the subject.  Do not send it to axp-list@redhat.com

aboot-0.6 with 2.2.17pre5
---------------------
Jun 24 17:37:09 d23-155 kernel: Linux version 2.2.17pre5  (gcc version
2.96 20000621 (experimental)) #3 Wed Jun 21 21:07:11 PDT 2000
Jun 24 17:37:09 d23-155 kernel: Booting on Tsunami variation Webbrick
using machine vector Webbrick from SRM
Jun 24 17:37:09 d23-155 kernel: HWRPB cycle frequency (462962962) seems
inaccurate - using the measured value of 462384180 Hz
Jun 24 17:37:09 d23-155 kernel: Console: colour VGA+ 80x25
Jun 24 17:37:09 d23-155 kernel: Calibrating delay loop... 920.65
BogoMIPS
Jun 24 17:37:09 d23-155 kernel: Memory: 255072k available
Jun 24 17:37:09 d23-155 kernel: Dentry hash table entries: 32768 (order
6, 512k)
Jun 24 17:37:09 d23-155 kernel: Buffer cache hash table entries: 262144
(order 8, 2048k)
Jun 24 17:37:09 d23-155 kernel: Page cache hash table entries: 32768
(order 5, 256k)
Jun 24 17:37:09 d23-155 kernel: POSIX conformance testing by UNIFIX
Jun 24 17:37:09 d23-155 kernel: Alpha PCI BIOS32 revision 0.04
Jun 24 17:37:09 d23-155 kernel: PCI: Probing PCI hardware
Jun 24 17:37:09 d23-155 kernel: Linux NET4.0 for Linux 2.2
Jun 24 17:37:09 d23-155 kernel: Based upon Swansea University Computer
Society NET3.039
Jun 24 17:37:09 d23-155 kernel: NET4: Unix domain sockets 1.0 for Linux
NET4.0. Jun 24 17:37:09 d23-155 kernel: NET4: Linux TCP/IP 1.0 for
NET4.0
Jun 24 17:37:09 d23-155 kernel: IP Protocols: ICMP, UDP, TCP, IL
Jun 24 17:37:09 d23-155 kernel: TCP: Hash tables configured (ehash
262144 bhash 65536)

aboot-0.7 with 2.4.0-test2
---------------------
Jun 24 18:18:56 d23-155 kernel: Linux version 2.4.0-test2 (gcc version
2.96 20000623 (experimental)) #2 Sat Jun 24 18:14:50 PDT 2000
Jun 24 18:18:56 d23-155 kernel: Booting on Tsunami variation Webbrick
using machine vector Webbrick from SRM
Jun 24 18:18:56 d23-155 kernel: Command line: root=/dev/hdb3
Jun 24 18:18:56 d23-155 kernel: memcluster 0, usage 1, start        0,
end      256
Jun 24 18:18:56 d23-155 kernel: memcluster 1, usage 0, start      256,
end    32675
Jun 24 18:18:56 d23-155 kernel: memcluster 2, usage 1, start    32675,
end    32768
Jun 24 18:18:56 d23-155 kernel: freeing pages 1296:32675
Jun 24 18:18:56 d23-155 kernel: On node 0 totalpages: 32768
Jun 24 18:18:56 d23-155 kernel: zone(0): 32768 pages.
Jun 24 18:18:56 d23-155 kernel: zone(1): 0 pages.
Jun 24 18:18:56 d23-155 atd: atd startup succeeded
Jun 24 18:18:56 d23-155 kernel: zone(2): 0 pages.
Jun 24 18:18:56 d23-155 kernel: Kernel command line: root=/dev/hdb3
Jun 24 18:18:56 d23-155 kernel: Console: colour VGA+ 80x25
Jun 24 18:18:56 d23-155 kernel: Calibrating delay loop... 920.65
BogoMIPS
Jun 24 18:18:56 d23-155 kernel: Memory: 246248k/261400k available (1260k
kernel code, 13104k reserved, 322k data, 224k init)
Jun 24 18:18:56 d23-155 kernel: Dentry-cache hash table entries: 32768
(order: 6, 524288 bytes)
Jun 24 18:18:56 d23-155 kernel: Buffer-cache hash table entries: 8192
(order: 3, 65536 bytes)
Jun 24 18:18:56 d23-155 kernel: Page-cache hash table entries: 32768
(order: 5, 262144 bytes)
Jun 24 18:18:56 d23-155 kernel: kmem_create: Poisoning requested, but
con given - bdev_cache
Jun 24 18:18:56 d23-155 kernel: Inode-cache hash table entries: 16384
(order: 5, 262144 bytes)
Jun 24 18:18:56 d23-155 kernel: kmem_create: Poisoning requested, but
con given - inode_cache
Jun 24 18:18:56 d23-155 kernel: Inode-cache hash table entries: 16384
(order: 5, 262144 bytes)
Jun 24 18:18:56 d23-155 kernel: kmem_create: Poisoning requested, but
con given - inode_cache


The difference is quite significant 255072k versus 246248k. I'm glad I
bought two
extra  DIMMS recently.