| Age | Commit message (Collapse) | Author |
|---|
|
This reverts commit eb7bb49eb5b48c39baf7a256b7c74e23e3da5660.
Stepan pointed out that "s" means string, which makes the following statement
in this commit message invalid: "Since we either have reserved space (which
we shouldn't do anything with in these two functions), an enum or a
hexadecimal value, unsigned int seemed like the way to go."
Signed-off-by: Luc Verhaegen <libv@skynet.be>
Acked-by: Luc Verhaegen <libv@skynet.be>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4335 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
|
|
To ease some of my debugging pain on the unichrome, i decided i needed to
move FB size selection into cmos, so i could test a size and then reset it
to the default after loading this value so that the next reboot uses the
(working) default again. This meant implementing set_option in parallel to
get_option.
get_option was then found to have inversed argument ordering (like outb) and
passing char * and then depending on the cmos layout length, which made me
feel quite uncomfortable. Since we either have reserved space (which we
shouldn't do anything with in these two functions), an enum or a
hexadecimal value, unsigned int seemed like the way to go. So all users of
get_option now have their arguments inversed and switched from using ints
to unsigned ints now.
The way get_cmos_value was implemented forced us to not overlap byte and to
have multibyte values be byte aligned. This logic is now adapted to do a
full uint32_t read (when needed) at any offset and any length up to 32, and
the shifting all happens inside an uint32_t as well. set_cmos_value was
implemented similarly. Both routines have been extensively tested in a
quick separate little program as it is not easy to get this stuff right.
build_opt_tbl.c was altered to function correctly within these new
parameters. The enum value retrieval has been changed strol(..., NULL, 10)
to stroul(..., NULL, 0), so that we not only are able to use unsigned ints
now but so that we also interprete hex values correctly. The 32bit limit
gets imposed on all entries not marked reserved, an unused "user_data" field
that appeared in a lot of cmos.layouts has been changed to reserved as well.
Signed-off-by: Luc Verhaegen <libv@skynet.be>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4332 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
|
|
- Replace $(PWD) with $(CURDIR) in Makefiles. I don't know why
the Solaris version behaves differently, but CURDIR is a safe
choice on gnu make (and we require gnu make already)
- Use tail -1 instead of tail -n1 in a file that already relies on
tail -1 support in another place
- Use tail -1 as alternative to tail -n1 in another place
- Use #define for ulong_t in romcc, as that name is used on Solaris
- Avoid fprinting a null pointer. The standard doesn't mandate that
this is a special case, and Solaris doesn't implement it that way.
Signed-off-by: Patrick Georgi <patrick.georgi@coresystems.de>
Acked-by: Ronald G. Minnich <rminnich@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4305 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
|
|
3 (with one of them way too much assembler code).
On the way, I had to make some changes to the way the code is built,
which is an effort I want to expand over time.
Right now, large portions of the in-ROM part of coreboot is compiled as
a single file, with lots of .c files including other .c files.
That has its justification for pre-raminit code, but it also affects
lots of post-raminit code (memcpy doesn't really make sense before
raminit, or at least CAR)
The coreboot_apc code (AMD boards) gained some .c includes because I
don't know that part of the code enough to really rework it and only
have limited possibilities to test it. The includes should give an
identical situation for this part of the code.
This change was posted as set of 6 patches to the list, but they
were mostly split for review purposes, hence commit them all at once.
They can still be backed up using the patch files, if necessary.
Signed-off-by: Patrick Georgi <patrick.georgi@coresystems.de>
Acked-by: Ronald G. Minnich <rminnich@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4233 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
|
|
(scan-build chokes on this)
Signed-off-by: Stefan Reinauer <stepan@coresystems.de>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4181 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
|
|
Signed-off-by: Stefan Reinauer <stepan@coresystems.de>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4157 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
|
|
refactoring Config.lb became possible.
Factor out ROM size calculation from Config.lb.
This patch converts 87 boards (with and without USE_FAILOVER_IMAGE),
but it has to work around a parser bug.
89 files changed, 209 insertions(+), 2415 deletions(-)
A total of 2206 removed lines.
Abuild works for all changed boards on khepri.
Myles writes:
I've tested serengeti for the failover portion and s2892 for the
nofailover portion. ldoptions are exactly the same and they both boot
the same.
Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
Acked-by: Myles Watson <mylesgw@gmail.com>
Acked-by: Peter Stuge <peter@stuge.se>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4147 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
|
|
This also has the config tool changes in v2/util.
Rename romfs.[ch]->cbfs.[ch] and sed romfs->cbfs romtool->cbfstool ROMFS->CBFS
Signed-off-by: Peter Stuge <peter@stuge.se>
Acked-by: Ronald G. Minnich <rminnich@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4113 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
|
|
Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
It should fix the build break introduced in r4101
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4105 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
|
|
now all have the same parameter order.
action "$(CC) $(DISTRO_CFLAGS) $(CFLAGS) $(CPPFLAGS) $(DEBUG_CFLAGS)
-I$(TOP)/src -I. -nostdinc -nostdlib -fno-builtin -Wall -Os -c -S
$(MAINBOARD)/$(CACHE_AS_RAM_AUTO_C) -o $@"
The idea behind this parameter order is:
- *FLAGS at the beginning.
- Use a common set of *FLAGS.
- Include files and directories listed afterwards.
- nostdinc, nostdlib, no-builtin tell the compiler this is standalone
code.
- Warnings. They do not influence source or compilation.
- Compilation strategy (small) and output mode (asm or binary).
- File to be compiled.
- Output name.
- $(DEBUG_CFLAGS) and -S are only used for asm output.
Other changes in this patch:
- src/supermicro/h8dme/Config.lb now uses $DEBUG_CFLAGS instead of
hardcoding the respective flags.
- $DEBUG_CFLAGS was added to asm outputting $CC calls:
supermicro/h8dme/Config.lb
lippert/roadrunner-lx/Config.lb
- $DISTRO_CFLAGS was added to some $CC calls in:
iwill/dk8_htx/Config.lb (CAR AP code)
supermicro/h8dmr/Config.lb (CAR AP code)
supermicro/h8dme/Config.lb (CAR AP code)
gigabyte/m57sli/Config.lb (CAR AP code)
gigabyte/ga_2761gxdk/Config.lb (CAR AP code)
amd/serengeti_cheetah_fam10/Config.lb (everywhere)
msi/ms7135/Config.lb (everywhere)
nvidia/l1_2pvv/Config.lb (CAR AP code)
-$CFLAGS was added to all $CC calls in:
amd/db800/Config.lb
amd/dbm690t/Config.lb
amd/norwich/Config.lb
amd/pistachio/Config.lb
amd/serengeti_cheetah/Config.lb
amd/serengeti_cheetah_fam10/Config.lb
arima/hdama/Config.lb
artecgroup/dbe61/Config.lb
asus/a8n_e/Config.lb
asus/a8v-e_se/Config.lb
asus/m2v-mx_se/Config.lb
broadcom/blast/Config.lb
digitallogic/msm800sev/Config.lb
gigabyte/ga_2761gxdk/Config.lb
gigabyte/m57sli/Config.lb
ibm/e325/Config.lb
ibm/e326/Config.lb
iei/pcisa-lx-800-r10/Config.lb
iwill/dk8_htx/Config.lb
iwill/dk8s2/Config.lb
iwill/dk8x/Config.lb
kontron/986lcd-m/Config.lb
lippert/roadrunner-lx/Config.lb
lippert/spacerunner-lx/Config.lb
msi/ms7135/Config.lb
msi/ms7260/Config.lb
msi/ms9185/Config.lb
msi/ms9282/Config.lb
newisys/khepri/Config.lb
nvidia/l1_2pvv/Config.lb
pcengines/alix1c/Config.lb
sunw/ultra40/Config.lb
supermicro/h8dme/Config.lb
supermicro/h8dmr/Config.lb
technexion/tim8690/Config.lb
tyan/s2735/Config.lb
tyan/s2850/Config.lb
tyan/s2875/Config.lb
tyan/s2880/Config.lb
tyan/s2881/Config.lb
tyan/s2882/Config.lb
tyan/s2885/Config.lb
tyan/s2891/Config.lb
tyan/s2892/Config.lb
tyan/s2895/Config.lb
tyan/s2912/Config.lb
tyan/s2912_fam10/Config.lb
tyan/s4880/Config.lb
tyan/s4882/Config.lb
- Use $@ wherever appropriate.
- Kill that evil CACHE_AS_RAM_AUTO_C variable.
- Trailing whitespace fixups on lines which were touched anyway.
We now only have 6 remaining different calls to $CC whereas before there
were 20.
If I am allowed to rename src/mainboard/kontron/986lcd-m/auto.c to
src/mainboard/kontron/986lcd-m/cache_as_ram_auto.c, we're down to 4
different calls.
If we can decide on the use of $CPU_OPT, we are down to 3 different
calls.
One additional point I'd like to clear up:
if ASSEMBLER_DEBUG
makedefine DEBUG_CFLAGS := -g -dA -fverbose-asm
end
"-dA -fverbose-asm" is only useful for asm output. For these flags,
DEBUG_CFLAGS is a total misnomer. What about calling them
DEBUG_ASMCFLAGS or somesuch?
"-g" should be controllable by a separate switch. It is useful even for
object code.
The following targets are broken by this patch because they contain
implicit declarations, but the error did not trigger due to missing
CFLAGS:
amd/serengeti_cheetah
asus/a8v-e_se
asus/m2v-mx_se
digitallogic/msm800sev
pcengines/alix1c
supermicro/h8dme
supermicro/h8dmr
Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
Acked-by: Ronald G. Minnich <rminnich@gmail.com>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4097 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
|
|
Lots of Config.lb files still have "if USE_DCACHE_RAM" sections although
USE_DCACHE_RAM is always set for them. Such checks are not only
pointless, they actively make the files hard to read.
A full abuild run confirmed that compilation did not change with this
patch applied.
The patch does not change whitespace of the remaining code to ease
review and svn blame.
With this change, it should be possible to have two or three Config.lb
variants in total (except the actual hardware config). Right now, some
Config.lb have comments, some don't, some have empty lines for better
readability, some don't, some have leading whitespace, some don't. This
is an utter mess and unifying these files would certainly reduce the
headaches I have when looking at them.
Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4093 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
|
|
As I mentioned a few weeks ago, I am in the process of porting this board:
http://www.technexion.com/products/embedded_boards/tim-8690-mt.html
This board has a dual BIOS , choosable with a jumper - much like the BIOS
savier from before - so it is a pleasure to work with as a linuxbios developer.
It is still a work in progress, however , I already submit the patch.
All on board devices and slots work as expected, only need some more stress
testing with the RAM, acpi, etc..
Signed-off-by: Daniel Toussaint <daniel@dmhome.net>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4075 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
|
