Ronald G. Minnich
8eb557da83
Remove code that AMD feels is not releasable.
2001-09-07 18:17:23 +00:00
Ronald G. Minnich
bfa9200c03
Fixed the raminit.inc so they don't need a working %esp.
...
Removed set of %esp from entry16.inc
2001-08-24 16:47:35 +00:00
Ronald G. Minnich
68221584cc
One-line change to set the ESP to a sane value before starting crt0.S.
...
Some of the .inc files require a sane stack value. This is not a full stack init
as found in crt0.S but just a quick set to make sure call/return works.
2001-08-23 19:58:31 +00:00
Eric W. Biederman
62edf8e35e
Minor glitch fix for reset16 from last checkin.
2001-08-21 04:50:37 +00:00
Eric W. Biederman
e1b09246c2
Compile fixes from linuxBIOS hopefully it works for everyone again.
...
crt0.base
- SMP compile fixes
- No longer unconditionally include cpu/p5/start32.inc
(it's been renamed cpu/i386/entry16.inc)
ldscript.base
- Simplfied, and moved some content into reset16.lds and entry16.lds
stddef.h
- Use C style comments for #endif /* I386_STDDEF_H */
pirq_routing.h
- The define is HAVE_PIRQ_TABLE not HAVE_PIRQ_ROUTING_TABLE
mtrr.c
- Remove the need for defining INTEL_PPRO_MTRR
mainboard/xxx/Config
- start32.inc is no longer automatically included include
entry16.inc and reset16.inc where appropriate
In particular if we are using a docipl we don't want
reset16.inc or reset16.lds
tyan/guiness/mptable.c
- Removed spurious define USE_ALL_CPUS
northbridge/intel/440gx/
- Updated to compile with the factored generic memory code.
nortsouthbridge/sis/xxx/
- Remove includes of northsouthbridge/sis/630/param.h did I add them?
The file is gone now so including it is just bad...
NLBConfig.py
- Allow relative filename arguments
- Add directive ldscript to add an include file into
our linker script
- Fixed docipl to use strings when calling set_option
- Allow target the target directive to use relative paths
- Add extra dependencies to ldscript.lds
2001-08-21 02:37:35 +00:00
Eric W. Biederman
ed8c9d7e0d
Code changes from my work and the AMD760MP chipset.
...
The primary changes is the way in which printk works.
But there are many other assorted code cleanups.
Just look and see :)
2001-08-08 02:45:10 +00:00
Eric W. Biederman
228148aa23
Start of merge from work on the AMD760MP platform.
...
This is the safe part just additions to files, and comment changes
2001-08-07 19:46:37 +00:00
Ronald G. Minnich
bde25c214d
Fixed so all IORRs get 0'ed
2001-06-09 18:18:59 +00:00
Ronald G. Minnich
80f9c91e66
Works now.
...
pcchips is now pretty fast!
2001-06-08 23:22:43 +00:00
Ronald G. Minnich
8df16b00d8
Added support for a cpufixup function.
...
See src/cpu/k7/Config for details.
2001-06-08 17:18:43 +00:00
Li-Ta Lo
1115c4d2af
use #ifdef i686 to protect Pentium class processors from hanging
2001-04-25 02:34:00 +00:00
Eric W. Biederman
39b133810a
Minor fixes
...
- Add IMB to the palcode in start.inc
- Reduce the heap to only 128K in alpha/ldscript.base
- In elfboot add checks to make certain we don't overwrite linuxBIOS...
2001-04-13 22:04:26 +00:00
Ronald G. Minnich
3a8ce19f87
we need cache_enable and disable visible.
...
failed attempts to get the acer northbridge to size ram
CV:S ----------------------------------------------------------------------
2001-04-12 18:36:53 +00:00
Eric W. Biederman
0785f9f233
Final updates for my code cleanup and alpha code merge.
2001-03-23 22:56:05 +00:00
Ronald G. Minnich
faf0c47b79
testing I HATE CVS
2001-03-22 21:26:32 +00:00
Ronald G. Minnich
227ca052f7
Changes from Eric for Alpha and other support
2001-03-13 04:22:19 +00:00
Ronald G. Minnich
af5eaf5b29
Changes to support 440BX
2001-02-08 16:17:38 +00:00
Li-Ta Lo
70d361e5f8
fix some rounding problem
2001-01-29 02:12:35 +00:00
Ronald G. Minnich
b56912ec3a
Fixes for CMD_LINE, and other fixes to set up ga-6bxc.
...
mtrr.c, remove redundant define
string.h bug with 0 in strlen
2001-01-18 23:11:26 +00:00
Li-Ta Lo
719f89882d
found an very very old bug on setting MTRR range
2001-01-04 07:50:56 +00:00
Li-Ta Lo
27a812455f
fix some typo
2000-12-07 01:43:28 +00:00
Li-Ta Lo
54e665686e
more clean up for printk messages
2000-12-06 11:02:36 +00:00
Li-Ta Lo
ba4403a177
finally, a fullproof implementation for MTRR setting
2000-12-06 03:58:57 +00:00
Li-Ta Lo
7f6e048fc0
a smarter Late MTRR implementation
2000-12-05 07:25:32 +00:00
Li-Ta Lo
9b604b8ca7
use the safer movzwl instead of movl
2000-12-05 02:07:37 +00:00
Li-Ta Lo
fd4ec49f7d
clear and set MTRRs correctly, I hope
2000-12-04 09:54:04 +00:00
Li-Ta Lo
d3e4c27442
use TSC to minotoring the performance of LinuxBIOS
2000-12-02 05:48:29 +00:00
Li-Ta Lo
69f93c224c
add more
...
#ifndef lint
static char rcsid[] = "$Id:$";
#endif
to *.c source
2000-12-02 03:51:28 +00:00
Li-Ta Lo
7d651021b2
l2_cache.c works fine, put DBG back to printk.h
2000-12-02 03:33:36 +00:00
Ronald G. Minnich
d96429d1e2
l2 cache bug
2000-12-01 23:35:17 +00:00
Li-Ta Lo
f0ea4a2bc0
Debug message is disabled by default now. If you want to show the message,
...
please add "option DEBUG" in the config file or in the C source
#define DEBUG
#include <printk.h>
2000-12-01 08:59:47 +00:00
Li-Ta Lo
075e8642af
add cvs identification string
...
#ifndef lint
static char rcsid[] = "$Id:$";
#endif
2000-12-01 01:43:51 +00:00
Ronald G. Minnich
4cf942c02f
added l2_cache.o object
2000-11-29 21:22:50 +00:00
Ronald G. Minnich
48607dd0b7
L2 cache code is solid. Experimenting with VIA SPD code -- no good.
2000-11-29 19:46:39 +00:00
Li-Ta Lo
bb441faa1c
minor clean up in mtrr.c
2000-11-29 09:05:58 +00:00
Ronald G. Minnich
cd664e0cd3
support for variable size memory on sis
...
added debugging to fill_inbuf
2000-11-28 05:23:02 +00:00
Ronald G. Minnich
a2edf7acdd
Missing stuff.
2000-11-28 04:55:17 +00:00
Ronald G. Minnich
a627e0bf31
per denis, add cache_disable call.
2000-11-25 04:13:05 +00:00
Li-Ta Lo
ab0636c99d
reformt with indent -kr -i8
2000-11-25 01:07:15 +00:00
Li-Ta Lo
4aec1cf42a
clean up on Makefiles and MTRR/L2 stuff
2000-11-24 04:16:50 +00:00
Ronald G. Minnich
8fa4c17322
We now support l2 cache!
...
first cut at acer m1631 support.
2000-11-21 17:30:21 +00:00
Ronald G. Minnich
ab80324461
Tweaks for VIA SPD support.
...
Added L2 cache support.
2000-11-21 01:17:56 +00:00
Ronald G. Minnich
21322ba3b7
mtrr fixes for 630, they are a hack now.
...
ipl.S: check for FRAMEBUFFER, don't set 0x63 if HAVE_FRAMEBUFFER not
defined.
Starting support for L440GX
2000-11-01 03:08:35 +00:00
Ronald G. Minnich
787deb4ae4
Fixes so we don't use any of the standard include paths
2000-10-17 22:58:51 +00:00
Ronald G. Minnich
934e5e0403
reorg
2000-10-17 13:18:43 +00:00
Ronald G. Minnich
69e023b385
Code reorg
2000-10-16 23:47:28 +00:00
Ronald G. Minnich
96fa389618
Building the new src tree
2000-10-16 03:03:03 +00:00