coreboot.git
13 years agoAfter this has been brought up many times before, rename src/arch/i386 to
Stefan Reinauer [Sat, 11 Dec 2010 20:33:41 +0000 (20:33 +0000)]
After this has been brought up many times before, rename src/arch/i386 to
src/arch/x86.

Signed-off-by: Stefan Reinauer <stepan@coreboot.org>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6161 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoTwo hda_verb.h files: Add more comments.
Uwe Hermann [Fri, 10 Dec 2010 12:52:50 +0000 (12:52 +0000)]
Two hda_verb.h files: Add more comments.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6160 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAdd TINY_BOOTBLOCK support for AMD SB700.
Uwe Hermann [Fri, 10 Dec 2010 09:02:50 +0000 (09:02 +0000)]
Add TINY_BOOTBLOCK support for AMD SB700.

Factor out the ROM decode enable functionality into bootblock.c and
handle it via the usual TINY_BOOTBLOCK mechanism.

Use "select TINY_BOOTBLOCK" in the southbridge, not individual boards.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6159 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMerge enable_rom.c files into bootblock.c files.
Uwe Hermann [Thu, 9 Dec 2010 18:09:14 +0000 (18:09 +0000)]
Merge enable_rom.c files into bootblock.c files.

All southbridges using TINY_BOOTBLOCK have a bootblock.c files which
simply includes an enable_rom.c files. As discussed on the mailing
list, drop the enable_rom.c file by merging it into bootblock.c.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6158 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoBuild fix, forgot to run abuild on the latest tree.
Uwe Hermann [Thu, 9 Dec 2010 13:10:57 +0000 (13:10 +0000)]
Build fix, forgot to run abuild on the latest tree.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6157 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoDeduplicate various ACPI .asl files.
Uwe Hermann [Thu, 9 Dec 2010 12:39:48 +0000 (12:39 +0000)]
Deduplicate various ACPI .asl files.

The files debug.asl, globutil.asl, and statdef.asl are duplicated in
many K8/Fam10h boards. However, they're neither board-specific nor
K8/Fam10h-specific nor AMD-specific, so move them to src/arch/i386/acpi.

debug.asl contains generic chunks for I/O port 0x80 handling, and debug
output over serial port (init COM port, send byte, send string, etc).

globutil.asl contains utility methods for string comparison, string length
and similar stuff.

statdef.asl contains generic ACPI bit definitions / status codes from
the ACPI spec (not board- or chipset-specific).

This patch was mostly generated by:

mkdir src/arch/i386/acpi
svn add src/arch/i386/acpi
svn cp src/mainboard/amd/dbm690t/acpi/debug.asl src/arch/i386/acpi/
svn cp src/mainboard/amd/dbm690t/acpi/globutil.asl src/arch/i386/acpi/
svn cp src/mainboard/amd/dbm690t/acpi/statdef.asl src/arch/i386/acpi/
cd src/mainboard
find . -name debug.asl -exec svn rm {} \;
find . -name globutil.asl -exec svn rm {} \;
find . -name statdef.asl -exec svn rm {} \;

Abuild-tested.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Stefan Reinauer <stepan@coreboot.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6156 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAdd missing instruction break.
Zheng Bao [Thu, 9 Dec 2010 06:18:29 +0000 (06:18 +0000)]
Add missing instruction break.

Signed-off-by: Zheng Bao <zheng.bao@amd.com>
Acked-by: Zheng Bao <zheng.bao@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6155 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoThese empty files sneaked in from another patch and shouldn't have been included...
Tobias Diedrich [Wed, 8 Dec 2010 21:45:57 +0000 (21:45 +0000)]
These empty files sneaked in from another patch and shouldn't have been included in r6153, remove them.

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6154 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoTobias Diedrich wrote:
Tobias Diedrich [Wed, 8 Dec 2010 21:40:12 +0000 (21:40 +0000)]
Tobias Diedrich wrote:
> Definitively a iasl problem, it can't even disassemble it's own
> output back to something equivalent to the input file.
> It seems to be generating Bytecode for the Add where it shouldn't.

Here is a solution using the SSDT.

Unfortunately iasl does not resolve simple arithmetic at compile
time, so we can not use Add(DEFAULT_PMBASE, PCNTRL) in the
Processor statement.
This patch instead dynamically generates the processor statement.
I can't use the speedstep generate_cpu_entries() directly since the
cpu doesn't support speedstep.
For now the code is in the southbridge directory, but maybe it
should go into cpu/intel/ somewhere.
IIRC notebook cpus of the era can already have speedstep, so it
would probably be possible to pair the i82371eb with a
speedstep-capable cpu...
Also, I don't know if multiprocessor boards (abit bp6?) would need
to be handled differently.

Abuild-tested.

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6153 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAllow user to define location for Kconfig config via
Patrick Georgi [Wed, 8 Dec 2010 19:58:30 +0000 (19:58 +0000)]
Allow user to define location for Kconfig config via
DOTCONFIG make variable (defaults to .config).
Let abuild use that.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6152 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMove "select CACHE_AS_RAM" lines from boards into CPU socket.
Uwe Hermann [Wed, 8 Dec 2010 08:22:04 +0000 (08:22 +0000)]
Move "select CACHE_AS_RAM" lines from boards into CPU socket.

All K8/Fam10h boards use CAR, so move the "select CACHE_AS_RAM"
into the socket directories, and remove it from the individual boards.

Do the same for Intel CPUs/sockets where all boards use CAR.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Stefan Reinauer <stepan@coreboot.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6151 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agosecond round name simplification. drop the <component>_ prefix.
stepan [Wed, 8 Dec 2010 07:07:33 +0000 (07:07 +0000)]
second round name simplification. drop the <component>_ prefix.

the prefix was introduced in the early v2 tree many years ago
because our old build system "newconfig" could not handle two files with
the same name in different paths like /path/to/usb.c and
/another/path/to/usb.c correctly. Only one of the files would end up
being compiled into the final image.

Since Kconfig (actually since shortly before we switched to Kconfig) we
don't suffer from that problem anymore. So we could drop the sb700_
prefix from all those filenames (or, the <componentname>_ prefix in general)

- makes it easier to fork off a new chipset
- makes it easier to diff against other chipsets
- storing redundant information in filenames seems wrong

Signed-off-by: <stepan@coresystems.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6150 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agofirst round name simplification. drop the <component>_ prefix.
stepan [Wed, 8 Dec 2010 05:42:47 +0000 (05:42 +0000)]
first round name simplification. drop the <component>_ prefix.

the prefix was introduced in the early v2 tree many years ago
because our old build system "newconfig" could not handle two files with
the same name in different paths like /path/to/usb.c and
/another/path/to/usb.c correctly. Only one of the files would end up
being compiled into the final image.

Since Kconfig (actually since shortly before we switched to Kconfig) we
don't suffer from that problem anymore. So we could drop the sb700_
prefix from all those filenames (or, the <componentname>_ prefix in general)

- makes it easier to fork off a new chipset
- makes it easier to diff against other chipsets
- storing redundant information in filenames seems wrong

Signed-off-by: <stepan@coresystems.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6149 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMove MMCONF resource into the domain for fam10 for the resource allocator.
Myles Watson [Tue, 7 Dec 2010 19:34:01 +0000 (19:34 +0000)]
Move MMCONF resource into the domain for fam10 for the resource allocator.

Signed-off-by: Myles Watson <mylesgw@gmail.com>
Acked-by: Scott Duplichan <scott@notabs.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6148 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoGet rid of some unneeded function prototypes in romstage.c files.
Uwe Hermann [Tue, 7 Dec 2010 19:16:07 +0000 (19:16 +0000)]
Get rid of some unneeded function prototypes in romstage.c files.

Abuild-tested.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6147 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoTrivial. Fix typo.
Zheng Bao [Tue, 7 Dec 2010 06:27:44 +0000 (06:27 +0000)]
Trivial. Fix typo.
sh> find -name "acpi_tables.c" | xargs sed -i "s/FDAT/FADT/g"

Signed-off-by: Zheng Bao <zheng.bao@amd.com>
Acked-by: Zheng Bao <zheng.bao@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6146 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoDrop unused/obsolete CONFIG_COMPRESS from a few board Kconfigs.
Uwe Hermann [Mon, 6 Dec 2010 20:27:12 +0000 (20:27 +0000)]
Drop unused/obsolete CONFIG_COMPRESS from a few board Kconfigs.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6145 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoGet rid of some useless/empty *_fixups.c files.
Uwe Hermann [Mon, 6 Dec 2010 18:20:48 +0000 (18:20 +0000)]
Get rid of some useless/empty *_fixups.c files.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6144 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoWinbond W83627HF: Use existing functions instead of open-coding.
Uwe Hermann [Mon, 6 Dec 2010 18:17:01 +0000 (18:17 +0000)]
Winbond W83627HF: Use existing functions instead of open-coding.

Use w83627hf_set_clksel_48() where needed instead or open-coding the same
functionality, and also use w83627hf_enable_serial() instead of
w83627hf_enable_dev() (which does exactly the same, but isn't wrapped in the
enter/exit config mode functions).

Abuild-tested.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6143 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoBefore lane reversal,
Zheng Bao [Mon, 6 Dec 2010 08:19:38 +0000 (08:19 +0000)]
Before lane reversal,
De-asserts STRAP_BIF_all_valid for
PCIE-GFX core.
After lane reversal,
Asserts STRAP_BIF_all_valid for
PCIE-GFX core.

Signed-off-by: Zheng Bao <zheng.bao@amd.com>
Acked-by: QingPei Wang <wangqingpei@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6142 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAdd initial support for the ASUS M4A78-EM.
Juhana Helovuo [Mon, 6 Dec 2010 01:11:12 +0000 (01:11 +0000)]
Add initial support for the ASUS M4A78-EM.

Signed-off-by: Juhana Helovuo <juhe@iki.fi>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6141 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoW83627DHG/W83627EHG fixups for virtual LDNs.
Uwe Hermann [Sun, 5 Dec 2010 22:36:14 +0000 (22:36 +0000)]
W83627DHG/W83627EHG fixups for virtual LDNs.

W83627DHG:

 - Add proper "virtual LDN" handling for the LDNs that need it (i.e., those
   that don't have their "enable" bit in bit 0 of the 0x30 register).

 - Fix various I/O masks in the pnp_dev_info[] array as per
   datasheet. Add missing PNP_IRQ0 to the W83627DHG_ACPI LDN.

W83627EHG:

 - Similar to W83627DHG, improve the "virtual LDN" setup a bit (it was
   mostly implemented already, though).

 - Add missing PNP_IRQ0 to the W83627EHG_ACPI LDN.

Also: Fix up devicetree.cb of all boards using W83627DHG/W83627EHG to adapt
for the virtual LDNs.

include/device/pnp.h: Add comment that 'function' (which refers to the
LDN and should probably be renamed later) has to be at least 16 bits
wide. In theory LDNs could use u8, but due to the virtual LDN info being
encoded in the "high byte" of 'function' it must be at least u16.

asrock/939a785gmh/romstage.c: Drop unused GPIO6_DEV.

ibase/mb899/romstage.c: Use DUMMY_DEV instead of a specific LDN (serial
port 1 in this case) to avoid confusion. The global registers
manipulated there are accessible from any LDN.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6140 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoACPI table dumping wrapper script
Stefan Reinauer [Sat, 4 Dec 2010 20:50:39 +0000 (20:50 +0000)]
ACPI table dumping wrapper script

Signed-off-by: Stefan Reinauer <stepan@coreboot.org>
Acked-by: Stefan Reinauer <stepan@coreboot.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6139 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoFollowing patch removes the cut-and-paste stuff from Mahagony and fixes the _CRS...
Rudolf Marek [Sat, 4 Dec 2010 10:08:55 +0000 (10:08 +0000)]
Following patch removes the cut-and-paste stuff from Mahagony and fixes the _CRS object to make it work (same code as on M2V-MX SE)

Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Acked-by: Marc Jones <marcj303@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6138 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoUpdate coreboot crossgcc toolchain, GCC 4.5.1, MPFR 3.0.0, GDB 7.2.
Marc Jones [Fri, 3 Dec 2010 00:45:56 +0000 (00:45 +0000)]
Update coreboot crossgcc toolchain, GCC 4.5.1, MPFR 3.0.0, GDB 7.2.
Add libelf_cv_elf_h_works=no to produce a libelf.h for Cygwin.
Add GDB patch to handle #pragma pack in the i386-elf gcc target.

Signed-off-by: Marc Jones <marcj303@gmail.com>
Acked-by: Stefan Reinauer <stepan@coreboot.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6137 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMore explicite and straight way to set seed.
Zheng Bao [Thu, 2 Dec 2010 01:50:38 +0000 (01:50 +0000)]
More explicite and straight way to set seed.
The read-modify-write wasn't needed. This is easier to understand.

Signed-off-by: Zheng Bao <zheng.bao@amd.com>
Acked-by: Marc Jones <marcj303@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6136 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoThe patch just make the power LED on.
Rudolf Marek [Tue, 30 Nov 2010 21:21:33 +0000 (21:21 +0000)]
The patch just make the power LED on.

Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6135 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoFix the SPD to channel mapping. Please note that there is something wrong with UMA.
Rudolf Marek [Tue, 30 Nov 2010 20:18:53 +0000 (20:18 +0000)]
Fix the SPD to channel mapping. Please note that there is something wrong with UMA.

Single channel (in slot DDR1 and DDR3) produces strange artefacts on screen (and hang)
Dual Channel (in DDR1 and DDR2 aka blue slot) - works nice
All slots populated - same case as Single channel - must be something wrong with UMA.

Tested with 2x 512MB CAS 2.5 DDR400

Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6134 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoTrivial. Reindent and dos2unix.
Zheng Bao [Tue, 30 Nov 2010 02:05:17 +0000 (02:05 +0000)]
Trivial. Reindent and dos2unix.

Signed-off-by: Zheng Bao <zheng.bao@amd.com>
Acked-by: Zheng Bao <zheng.bao@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6133 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoTobias Diedrich wrote:
Tobias Diedrich [Mon, 29 Nov 2010 20:40:33 +0000 (20:40 +0000)]
Tobias Diedrich wrote:
> Stefan Reinauer wrote:
> > The specified IO port is most likely wrong. As the comment mentions, the
> > SSDT is a good place for that. A preprocessor define used both in the
> > CPU init code and in the asl would solve the problem without an SSDT.
> > For some info on CPU SSDT creation on intel check out
> > src/cpu/intel/speedstep/acpi.c
>
> The IO port is ok (and I wrote the comment myself ;)):
> DEFAULT_PMBASE is 0xe400
> PCNTRL reg offset is 0x10
>
> Using the preprocessor will probably work too if iasl can do simple
> arithmetic (likely yes), I'll look into that.

BTW, my first idea was to use an acpi method that looks up pmbase in
the pci cfg space, but when I define a method like this:

        Method(TEST, 2)
        {
                Return (Add(Arg0, Arg1))
        }

I get:
|build/mainboard/asus/p2b/dsdt.ramstage.asl     9:   Processor (CPU0,
|0x01, TEST(0xe400, 0x10), 0x06) {}
|Error    4096 -       syntax error, unexpected PARSEOP_NAMESEG,
|expecting ')' ^

While using the builtin Add() directly works.

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6132 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAdd Fintek F71889 detection and dump support.
David Hendricks [Mon, 29 Nov 2010 11:56:39 +0000 (11:56 +0000)]
Add Fintek F71889 detection and dump support.

The patch was tested by a user on IRC who had the F71889FG. I
wrote it using documentation from Fintek's website available here:
http://www.fintek.com.tw/files/productfiles/F71889_V0.28P.pdf

This patch also seems to work for the F71889ED, which uses 0x09 and 0x09 for
chip ID bytes 1 & 2. However, I have not been able to find documentation to
verify that the two chips are identical from superiotool's perspective.

Signed-off-by: David Hendricks <dhendrix@google.com>
Signed-off-by: Alec Ari <neotheuser@ymail.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6131 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agofix typo
Stefan Reinauer [Mon, 29 Nov 2010 00:20:20 +0000 (00:20 +0000)]
fix typo

Signed-off-by: Stefan Reinauer <stepan@coreboot.org>
Acked-by: Stefan Reinauer <stepan@coreboot.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6130 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agodevicetree.cb: Only add as many entries as there are DIMM slots.
Uwe Hermann [Sun, 28 Nov 2010 14:24:07 +0000 (14:24 +0000)]
devicetree.cb: Only add as many entries as there are DIMM slots.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6129 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years ago- Add support for Intel Pentium III MSRs
Tobias Diedrich [Sat, 27 Nov 2010 14:44:19 +0000 (14:44 +0000)]
- Add support for Intel Pentium III MSRs
- pmbase is on southbridge function 3 on I82371XX

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6128 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAfter finding the missing bit poweroff works now.
Tobias Diedrich [Sat, 27 Nov 2010 09:40:16 +0000 (09:40 +0000)]
After finding the missing bit poweroff works now.
I cleaned up the patch and moved most of the dsdt.dsl and
acpi_tables.c into the southbrige/northbridge directory.
Updated patch should fix abuild error and incorporates suggestions
on irc by uwe (thanks for the comments).
Thanks to Idwer Vollering <vidwer@gmail.com> for the original patch.

Tested:
  Linux (poweroff, powerbutton event)
  XP (poweroff, powerbutton event)

Abuild-tested

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Stefan Reinauer <stepan@coreboot.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6127 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoBroadcom BCM5785: Add TINY_BOOTBLOCK support.
Uwe Hermann [Fri, 26 Nov 2010 22:42:41 +0000 (22:42 +0000)]
Broadcom BCM5785: Add TINY_BOOTBLOCK support.

In bcm5785_enable_rom(): Use PCI IDs from pci_ids.h instead of hardcoding,
and use 'dev' instead of 'addr' as device_t variable name.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6126 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAMD SB600: Add TINY_BOOTBLOCK support.
Uwe Hermann [Fri, 26 Nov 2010 22:39:40 +0000 (22:39 +0000)]
AMD SB600: Add TINY_BOOTBLOCK support.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6125 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAMD-8111: Add TINY_BOOTBLOCK support.
Uwe Hermann [Fri, 26 Nov 2010 22:35:11 +0000 (22:35 +0000)]
AMD-8111: Add TINY_BOOTBLOCK support.

Also, add missing license header to amd8111_enable_rom.c, add some more code
comments and use PCI IDs from pci_ids.h instead of hardcoding.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6124 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMCP55: Add TINY_BOOTBLOCK support.
Uwe Hermann [Thu, 25 Nov 2010 09:03:55 +0000 (09:03 +0000)]
MCP55: Add TINY_BOOTBLOCK support.

Also, move CONFIG_HT_CHAIN_END_UNITID_BASE #ifdef block to mcp55.h to make
the build work (but this is a good idea anyway, as it's used in
multiple files).

Abuild-tested.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Stefan Reinauer <stepan@coreboot.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6123 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoS3 support for ASUS M2V
Tobias Diedrich [Wed, 24 Nov 2010 20:03:09 +0000 (20:03 +0000)]
S3 support for ASUS M2V

This adds the board-specific parts for S3 support on the M2V board.

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Stefan Reinauer <stepan@coreboot.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6122 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoWith low serial console loglevels a pcie graphics card is not
Tobias Diedrich [Wed, 24 Nov 2010 19:57:08 +0000 (19:57 +0000)]
With low serial console loglevels a pcie graphics card is not
initialized properly because the pcie link takes some time to come
up.

I set the timeout rather arbitrary to 100ms, this is what a BIOS_ERR
and higher only boot looks like on my system (with pcie printks set
to BIOS_ERR so they show up):

|Device error
|Device error
|PCI: 00:02.0 PCIe link up after 35800 us
|PCI: 00:03.0 PCIe link up after 12900 us
|PCI: 00:03.1 PCIe link timeout
|PCI: 00:03.2 PCIe link up after 32000 us
|APIC: 00 missing read_resources
|I2C: 01:50 missing read_resources
|I2C: 01:51 missing read_resources
|I2C: 01:52 missing read_resources
|I2C: 01:53 missing read_resources
|Start bios (version pre-0.6.2-20101025_023503-nukunuku)

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Stefan Reinauer <stepan@coreboot.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6121 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoThis patch solves crashes and BSODs that occur when booting Win7 with
Scott Duplichan [Wed, 24 Nov 2010 00:39:44 +0000 (00:39 +0000)]
This patch solves crashes and BSODs that occur when booting Win7 with
AMD RS780 uma graphics. Tested with frame buffer sizes 64m through 1GB
by running dxdiag and Windows media player at 1600x1200 true color.
Additional changes needed to boot Win7 on Mahogany_fam10 will follow.

-- Enable and program the debug bar as required by the ATI graphics driver.
   First, make the debug bar writable and allow resource allocation code
   to program it. Once programmed, enable its operation.
-- Disable the family 10h processor mmconf while the RS780 mmconf is in use.
-- Make strap programming more closely follow the reference BIOS.
-- Disable PCIe bar 3 after using it.
-- UMA size is no longer hardcoded.
-- Disable write combining for all steppings to eliminate stability problem.
-- Correct task file data.
-- Improve the accuracy of the Atom table that passes information to the driver.

Signed-off-by: Scott Duplichan <scott@notabs.org>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6120 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoUSBDEBUG by default in abuild was committed by mistake and
Patrick Georgi [Tue, 23 Nov 2010 07:30:50 +0000 (07:30 +0000)]
USBDEBUG by default in abuild was committed by mistake and
then left in because USBDEBUG was actively worked on.
This isn't true anymore, so drop it

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6119 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMake smp_write_bus static (local scope), to prevent new boards from
Patrick Georgi patrick [Tue, 23 Nov 2010 07:19:54 +0000 (07:19 +0000)]
Make smp_write_bus static (local scope), to prevent new boards from
using it directly again.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6118 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years ago1) wraps the s3 parts of chipset code/memory init code with if CONFIG_HAVE_ACPI_RESUM...
Rudolf Marek [Mon, 22 Nov 2010 22:00:52 +0000 (22:00 +0000)]
1) wraps the s3 parts of chipset code/memory init code with if CONFIG_HAVE_ACPI_RESUME == 1 getting rid of ugly define in romstage.c

2) the patch implements get_cbmem_toc in chipset specific way if defined.
On Intel targets it should be unchanged. On K8T890 the the cbmem_toc is read from NVRAM. Why you ask? Because we cannot do it as on intel, because the framebuffer might be there making it hard to look for it in memory (and remember we need it so early that everying is uncached)

3) The patch removes hardcoded limits for suspend/resume save area (it was 1MB) on intel. Now it computes right numbers itself.

4) it impelements saving the memory during CAR to reserved range in sane way. First the sysinfo area (CAR data) is copied, then the rest after car is disabled (cached copy is used). I changed bit also the the copy of CAR area is now done uncached for target which I feel is more right.

I think I did not change the Intel suspend/resume behaviour but best would be if someone can test it. Please note this patch was unfinished on my drive since ages and it would be very nice to get it in to prevent bit rotten it again.
Now I feel it is done good way and should not break anything. I did a test with abuild and it seems fine.

Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Acked-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Stefan Reinauer <stepan@coreboot.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6117 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoDrop unused ACPI_WRITE_MADT_IOAPIC #define.
Uwe Hermann [Mon, 22 Nov 2010 16:23:54 +0000 (16:23 +0000)]
Drop unused ACPI_WRITE_MADT_IOAPIC #define.

This should probably be C code in some .c file anyway.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6116 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoDrop per-board ram_check() calls for now.
Uwe Hermann [Mon, 22 Nov 2010 15:57:57 +0000 (15:57 +0000)]
Drop per-board ram_check() calls for now.

Every board had a slightly different invokation, very often commented out
anyway. We could either decide that this is only to be used by developers
during bringup (and thus added manually to romstage.c and removed before
the board gets committed). This method seems to be preferred from what I
have heard on IRC / mailing list in the past.

Or, we add the ram_check() somewhere globally and allow the user to enable
it via menuconfig (possibly only if EXPERT is selected).

Either way, the current method of spreading the calls all over the place is
not really the way to go.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6115 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoFinal set of smp_write_bus -> mptable_write_buses changes.
Patrick Georgi [Mon, 22 Nov 2010 14:14:56 +0000 (14:14 +0000)]
Final set of smp_write_bus -> mptable_write_buses changes.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6114 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoWorkaround to get die.c to work with romcc.
Patrick Georgi [Mon, 22 Nov 2010 13:07:10 +0000 (13:07 +0000)]
Workaround to get die.c to work with romcc.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6113 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoi855: Remove useless memctrl indirection.
Uwe Hermann [Mon, 22 Nov 2010 12:59:36 +0000 (12:59 +0000)]
i855: Remove useless memctrl indirection.

This needlessly complicates the code and increases register pressure on romcc
chipsets. We did the same conversion on i440BX, i830, and others.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Peter Stuge <peter@stuge.se>
Acked-by: Stefan Reinauer <stepan@coreboot.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6112 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoPrinting coreboot debug messages on VGA console is pretty much useless, since
Stefan Reinauer [Mon, 22 Nov 2010 08:09:50 +0000 (08:09 +0000)]
Printing coreboot debug messages on VGA console is pretty much useless, since
initializing VGA happens pretty much as the last thing before starting the
payload. Hence, drop VGA console support, as we did in coreboot v3.

- Drop VGA and BTEXT console support.
  Console is meant to be debugging only, and by the time graphics comes up
  99% of the risky stuff has already happened. Note: This patch does not remove
  hardware init but only the actual output functionality.

  The ragexl driver needs some extra love, but that's for another day
- factor out die() and post()
- drop some leftover RAMBASE < 0x100000 checks.

Signed-off-by: Stefan Reinauer <stepan@coreboot.org>
Acked-by: QingPei Wang<wangqingpei@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6111 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoacpi.h: Small fixes and adding comments.
Uwe Hermann [Mon, 22 Nov 2010 00:42:42 +0000 (00:42 +0000)]
acpi.h: Small fixes and adding comments.

 - Mention full name of all the tables (SSDT, FADT, etc).

 - Drop obsolete / incorrect "LXBIOS" reference.

 - Add missing ACPI address space type #defines specified in newer versions:
   ACPI_ADDRESS_SPACE_EC, ACPI_ADDRESS_SPACE_SMBUS.

 - Add missing "enum acpi_apic_types" entries: Localx2Apic, Localx2ApicNMI.

 - Add ACPI_FACS_64BIT_WAKE_F #define.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6110 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMove CK804_PCI_E_X and CK804B_PCI_E_X defines (which have been 4 by
Jonathan Kollasch [Sun, 21 Nov 2010 22:55:46 +0000 (22:55 +0000)]
Move CK804_PCI_E_X and CK804B_PCI_E_X defines (which have been 4 by
default on all boards) into Kconfig.

Signed-off-by: Jonathan Kollasch <jakllsch@kollasch.net>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6109 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoSimplify a few code chunks, fix whitespace and indentation.
Uwe Hermann [Sun, 21 Nov 2010 22:47:22 +0000 (22:47 +0000)]
Simplify a few code chunks, fix whitespace and indentation.

Also, remove some less useful comments, some dead code / unused functions.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6108 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoDrop excessive whitespace randomly sprinkled in romstage.c files.
Uwe Hermann [Sun, 21 Nov 2010 17:29:59 +0000 (17:29 +0000)]
Drop excessive whitespace randomly sprinkled in romstage.c files.

Also drop some dead or useless code snippets.

Abuild-tested.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6107 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoConvert more boards to use mptable_write_buses.
Patrick Georgi [Sun, 21 Nov 2010 14:41:07 +0000 (14:41 +0000)]
Convert more boards to use mptable_write_buses.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6106 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoConvert boards to use mptable_write_buses.
Patrick Georgi [Sun, 21 Nov 2010 14:40:09 +0000 (14:40 +0000)]
Convert boards to use mptable_write_buses.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6105 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMove MCP55_PCI_E_X_* to Kconfig. Any useless values in romstage.cs were
Patrick Georgi [Sun, 21 Nov 2010 14:38:24 +0000 (14:38 +0000)]
Move MCP55_PCI_E_X_* to Kconfig. Any useless values in romstage.cs were
not brought over to Kconfig (this applies to all #defines to 4, as
that's the default anyway)

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6104 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoUse DIMM0 et al in lots more places instead of hardocding values.
Uwe Hermann [Sun, 21 Nov 2010 11:36:03 +0000 (11:36 +0000)]
Use DIMM0 et al in lots more places instead of hardocding values.

The (0xa << 3) expression equals 0x50, i.e. DIMM0.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6103 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoBuild fix.
Uwe Hermann [Sun, 21 Nov 2010 10:26:04 +0000 (10:26 +0000)]
Build fix.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6102 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMerge all spd_addr.h into the resp. romstage.c files.
Uwe Hermann [Sat, 20 Nov 2010 20:36:40 +0000 (20:36 +0000)]
Merge all spd_addr.h into the resp. romstage.c files.

Except for one instance the spd_addr.h were now very tiny, there's not
much point in keeping that stuff in an extra file. The only user of those
files is the romstage.c file anyway.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6101 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoSome more DIMM0 related cleanups and deduplication.
Uwe Hermann [Sat, 20 Nov 2010 20:23:08 +0000 (20:23 +0000)]
Some more DIMM0 related cleanups and deduplication.

 - VIA VT8235: Do the shift in smbus_read_byte() as all other chipsets do.

 - spd.h: Move RC00-RC63 #defines here, they were duplicated in lots of
   romstage.c files and lots of spd_addr.h files. Don't even bother for
   those spd_addr.h which aren't even actually used, drop them right away.

 - Replace various 0x50 hardcoded numbers with DIMM0, 0x51 with DIMM1,
   and 0xa0 with (DIMM0 << 1) where appropriate.

 - Various debug.c files: Replace SMBUS_MEM_DEVICE_START with DIMM0,
   SMBUS_MEM_DEVICE_END with DIMM7, and drop useless SMBUS_MEM_DEVICE_INC.

 - VIA VX800: Drop unused SMBUS_ADDR_CH* #defines.

 - VIA VT8623: Do the shift in smbus_read_byte() as all other chipsets do.
   Then, replace 0xa0 (which now becomes 0x50) with DIMM0.

 - alix1c/romstage.c, alix2d/romstage.c: Adapt to recent bit shift changes.

 - Various files: Drop DIMM_SPD_BASE and/or replace it with DIMM0.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6100 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoUnify DIMM SPD addressing. For Geode, change the
Patrick Georgi [Sat, 20 Nov 2010 10:31:00 +0000 (10:31 +0000)]
Unify DIMM SPD addressing. For Geode, change the
addressing scheme to match the rest of the tree
(0x50 instead of 0xa0).

abuild tested.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6099 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoCosmetic fixes and comment additions in acpi.c.
Uwe Hermann [Fri, 19 Nov 2010 15:14:42 +0000 (15:14 +0000)]
Cosmetic fixes and comment additions in acpi.c.

 - Fix whitespace, coding style, and indentation in some places.

 - Add comments for less obvious entries and hardcoded numbers (e.g. 'type').

 - Add comments for all/most 'revision' fields, mention in which version
   of the ACPI spec which revision number is to be used.

 - Add URLs to a few external documents which describe tables that are
   not mentioned in the ACPI spec (or where the external document may
   provide further info), e.g. SRAT, SLIT, HPET, MCFG, etc.

 - Use the ASLC #define instead of hardcoding "CORE" in one instance
   (ASLC is already used everywhere else).

 - Add some TODOs for additional stuff which is in the spec but not yet
   handled by our code / #defines.

Abuild-tested.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6098 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAdd test to check for up-to-date GPL license headers to lint.
Patrick Georgi [Fri, 19 Nov 2010 10:16:43 +0000 (10:16 +0000)]
Add test to check for up-to-date GPL license headers to lint.
"make lint" should not stop after first failed test.
Improve "make lint" output.

Signed-off-by: Patrick Georgi <patrick.georgi@coresystems.de>
Acked-by: Patrick Georgi <patrick.georgi@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6097 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agodrop temp file from coreboot tree
Stefan Reinauer [Fri, 19 Nov 2010 00:29:32 +0000 (00:29 +0000)]
drop temp file from coreboot tree
Signed-off-by: Stefan Reinauer <stepan@coresystems.de>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6096 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoFix/drop some obsolete comments,
Uwe Hermann [Thu, 18 Nov 2010 20:12:13 +0000 (20:12 +0000)]
Fix/drop some obsolete comments,

 - s/Options.lb/devicetree.cb/

 - s/Config.lb/devicetree.cb/

 - s/cache_as_ram_auto.c/romstage.c/

 - h8dmr_fam10/README: Drop obsolete comment, we have mc_patch_01000086.h in
   the tree now.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6095 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoDrop unused and incorrect RTC_DEV for Winbond W83627EHG.
Uwe Hermann [Thu, 18 Nov 2010 19:40:33 +0000 (19:40 +0000)]
Drop unused and incorrect RTC_DEV for Winbond W83627EHG.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6094 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMake lint script executable, otherwise invocation fails.
Uwe Hermann [Thu, 18 Nov 2010 18:12:09 +0000 (18:12 +0000)]
Make lint script executable, otherwise invocation fails.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6093 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agolint tests can now describe what they do (for the benefit of
Patrick Georgi [Thu, 18 Nov 2010 15:07:06 +0000 (15:07 +0000)]
lint tests can now describe what they do (for the benefit of
make lint users)

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6092 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAdd "make lint" target that calls all util/lint/lint-* scripts
Patrick Georgi [Thu, 18 Nov 2010 15:05:06 +0000 (15:05 +0000)]
Add "make lint" target that calls all util/lint/lint-* scripts
and fails if any of these output text to stdout.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6091 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoSet locale to POSIX to avoid problems with invalid 8bit character
Patrick Georgi [Thu, 18 Nov 2010 14:33:02 +0000 (14:33 +0000)]
Set locale to POSIX to avoid problems with invalid 8bit character
sequences.
Increase scanning speed.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6090 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMove DIMM_MAP_LOGICAL to Kconfig.
Patrick Georgi [Thu, 18 Nov 2010 11:36:16 +0000 (11:36 +0000)]
Move DIMM_MAP_LOGICAL to Kconfig.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6089 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMove register block definitions out of board code into
Patrick Georgi [Thu, 18 Nov 2010 10:48:15 +0000 (10:48 +0000)]
Move register block definitions out of board code into
chipset code (where it belongs)

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6088 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoFor completeness sake: License header.
Patrick Georgi [Thu, 18 Nov 2010 00:46:53 +0000 (00:46 +0000)]
For completeness sake: License header.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6087 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoEliminate SET_NB_CFG_54 option. There was no board that
Patrick Georgi [Thu, 18 Nov 2010 00:11:32 +0000 (00:11 +0000)]
Eliminate SET_NB_CFG_54 option. There was no board that
deselected it, and very likely there won't ever be any
hardware that requires it deselected.

Keep the "selected" code path around, leading to no
functional change.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Scott Duplichan <scott@notabs.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6086 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMove Intel power management related defines to some central location.
Patrick Georgi [Wed, 17 Nov 2010 21:52:15 +0000 (21:52 +0000)]
Move Intel power management related defines to some central location.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6085 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoDynamically generate PNP0C02 mainboard resources in SSDT
Tobias Diedrich [Wed, 17 Nov 2010 16:27:06 +0000 (16:27 +0000)]
Dynamically generate PNP0C02 mainboard resources in SSDT

Updated patch with improved comments and small bugfix (use same
value for min and max on io resource).

While adding the area between TOM1 and 4GB to \SB.PCI0._CRS seems to be the
easiest way to get both Linux and Windows happy, it is not quite correct
because reserved areas like APIC, MMCONF etc. ranges need to be excluded.

This is a proof of concept patch for the M2V board that dynamically creates a
ResourceTemplate() containing these in the SSDT and adds a corresponding
PNP0C02 device to the DSDT.

All resources that have IORESOURCE_RESERVE and (IORESOURCE_MEM or IORESOURCE_IO) set
are added.

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Added M2V-MX SE too.
Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6084 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoLinux also needs the MMCONF area to be reserved either in E820 or
Tobias Diedrich [Wed, 17 Nov 2010 11:30:50 +0000 (11:30 +0000)]
Linux also needs the MMCONF area to be reserved either in E820 or
as an ACPI motherboard resource or it will not enable MMCONFIG
and the extended pcie configuration area will be unaccessible:

This patch adds the IORESOURCE_RESERVE flag to the APIC and MMCONF
resource flags to do this.
I also added a new resource for the mapped bios rom area just below 4GB.
I'm not sure if the choice for the index parameter of new_resource()
is correct though.
Note that the bios rom decode is enabled in
src/southbridge/via/vt8237r/vt8237r_early_smbus.c
for the whole 4MB area (even though the comment says 1MB).

Ruik: I extended the flash range to 16MB (This is what VT8237S can decode)
Remove the MMCONFIG region reserve in the mainboard file (this patch makes it obsolete)

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6083 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoThis problem was introduced with
Tobias Diedrich [Wed, 17 Nov 2010 11:02:05 +0000 (11:02 +0000)]
This problem was introduced with
http://tracker.coreboot.org/trac/coreboot/changeset/3953

Note that all corresponding DSDTs only ever check TOM2 against 0.

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6082 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoThe only southbridge having a pirq_assign_irqs function (needed for
Tobias Diedrich [Wed, 17 Nov 2010 10:58:13 +0000 (10:58 +0000)]
The only southbridge having a pirq_assign_irqs function (needed for
CONFIG_PIRQ_ROUTE) so far is the amd cs5530.
Add one for vt8237 too.
Setting up the pci routing is important in case you want to boot DOS,
OSes that don't support ACPI or MP tables and ROMs for add-in storage
controllers may depend on this too.
TODO: Fix the 4 routing links limitation in
      src/arch/i386/boot/pirq_routing.c

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6081 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoDrop W83627THF, it's the same device as W83627THG.
Uwe Hermann [Tue, 16 Nov 2010 23:15:37 +0000 (23:15 +0000)]
Drop W83627THF, it's the same device as W83627THG.

The only difference is that the "G" version is in a Pb-free package, which
is not relevant from a programmer's view.

We keep W83627THG (and drop W83627THF) because:

 - The W83627THF had a CIR device / LDN which doesn't actually exist.

 - The W83627THF had no GPIO2, GPIO3 LDNs (were commented out).

 - The W83627THF didn't use the PNP_MSC0/1 which is needed/used by boards.

This also fixes an issue on MSI MS7135's devicetree.cb:

  device pnp 4e.6 off end           # XXX keep allocator happy

The line above can be (and is) removed, as it was only needed due to the
incorrect CIR LDN in the W83627THF.

In the iwill/dk8x target: Drop incorrect LDNs 4 and 6, add 0xb.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6080 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoDrop commented out debug defines
Patrick Georgi [Tue, 16 Nov 2010 22:15:09 +0000 (22:15 +0000)]
Drop commented out debug defines

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6079 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoForgot to remove one set of SET_FIDVID defines
Patrick Georgi [Tue, 16 Nov 2010 22:10:55 +0000 (22:10 +0000)]
Forgot to remove one set of SET_FIDVID defines

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6078 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMove the SET_FIDVID* family of configuration options to Kconfig and
Patrick Georgi [Tue, 16 Nov 2010 21:25:29 +0000 (21:25 +0000)]
Move the SET_FIDVID* family of configuration options to Kconfig and
make their defaults more obvious.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Stefan Reinauer <stepan@coreboot.org>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6077 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoback out parts of #6073
Stefan Reinauer [Tue, 16 Nov 2010 00:41:17 +0000 (00:41 +0000)]
back out parts of #6073

Signed-off-by: Stefan Reinauer <stepan@coresystems.de>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6076 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agofix random breakage
Stefan Reinauer [Mon, 15 Nov 2010 21:09:57 +0000 (21:09 +0000)]
fix random breakage

Signed-off-by: Stefan Reinauer <stepan@coresystems.de>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6075 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMove RCBA defines to northbridge (instead of mainboard)
Patrick Georgi [Mon, 15 Nov 2010 19:44:42 +0000 (19:44 +0000)]
Move RCBA defines to northbridge (instead of mainboard)

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6074 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoC and other Super I/O cosmetic fixes.
Uwe Hermann [Mon, 15 Nov 2010 19:35:14 +0000 (19:35 +0000)]
C and other Super I/O cosmetic fixes.

 - Random coding style, whitespace and cosmetic fixes.

 - Consistently use the same spacing and 4-hexdigit port number format
   in the pnp_dev_info[] arrays.

 - Drop dead/unused code and less useful comments.

 - Add missing "(C)" characters and copyright years.

 - Shorten and simplify some code snippets.

 - Use u8/u16/etc. everywhere.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6073 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAdd a target for the ASUS A8V-E Deluxe (trivial).
Uwe Hermann [Sun, 14 Nov 2010 21:48:14 +0000 (21:48 +0000)]
Add a target for the ASUS A8V-E Deluxe (trivial).

For now this is a plain copy of the ASUS A8V-E SE target, I reported
that most of the code also works (sort of) for the ASUS A8V-E Deluxe
a long while ago, see

  http://www.coreboot.org/pipermail/coreboot/2008-March/031866.html
  http://www.coreboot.org/ASUS_A8V-E_Deluxe

There will be a bunch of changes necessary though (devicetree.cb, mptable.c,
ACPI, etc) which do not apply to the A8V-E SE, so we need an extra target.

Also: Increase ID_SECTION_OFFSET on the VIA K8T890/K8M890 southbridge, as
otherwise there will be build errors if the MAINBOARD_PART_NUMBER string
gets too long (as is the case for "A8V-E Deluxe"). The error is:

  ld: section .id loaded at [00000000ffffffd2,00000000ffffffef] overlaps
  section .romstrap loaded at [00000000ffffff80,00000000ffffffd3]

(both with stock Debian gcc and with xgcc)

Increase ID_SECTION_OFFSET (default 0x10) to 0x80 as other southbridges do.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6072 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoCK804/MCP55 devicetree.cb cosmetic and indentation fixes.
Uwe Hermann [Sun, 14 Nov 2010 20:10:11 +0000 (20:10 +0000)]
CK804/MCP55 devicetree.cb cosmetic and indentation fixes.

Add a few more comments for the entries, and also change the devicetree.cb
files to the more compact and better readable variant with indentation level
of 2 spaces (instead of random mix of tabs and spaces).

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6071 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoReturn 0, (as for 40pin cable if SB not found)
Rudolf Marek [Sun, 14 Nov 2010 14:39:29 +0000 (14:39 +0000)]
Return 0, (as for 40pin cable if SB not found)
Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6070 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMove cable detect logic to a weak function in vt8237r_ide.c and add
Tobias Diedrich [Sun, 14 Nov 2010 14:17:29 +0000 (14:17 +0000)]
Move cable detect logic to a weak function in vt8237r_ide.c and add
an override function in m2v/mainboard.c

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6069 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoCurrently the
Tobias Diedrich [Sun, 14 Nov 2010 14:12:14 +0000 (14:12 +0000)]
Currently the
        cablesel |= (sb->ide0_80pin_cable << 28) |
                    (sb->ide0_80pin_cable << 20) |
                    (sb->ide1_80pin_cable << 12) |
                    (sb->ide1_80pin_cable << 4);
in vt8237r_ide.c ends up doing
cablesel |= 0xfffffff0;
(with both bits set to 1) which is probably not the intended result. ;)

After a short discussion on irc the consensus was to change the
bitfields to u8 as it's probably not worth it using bitfields here.

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6068 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoMTRR related improvements for AMD family 10h and family 0Fh systems
Scott Duplichan [Sat, 13 Nov 2010 19:07:59 +0000 (19:07 +0000)]
MTRR related improvements for AMD family 10h and family 0Fh systems

-- When building for UMA, reduce the limit for DRAM below 4GB
   from E0000000 to C0000000. This is needed to accomodate the
   UMA frame buffer.
-- Correct problem where msr C0010010 bits 21 and 22 (MtrrTom2En
   and Tom2ForceMemTypeWB) are not set consistently across cores.
-- Enable TOM2 only if DRAM is present above 4GB.
-- Use AMD Tom2ForceMemTypeWB feature to avoid the need for
   variable MTRR ranges above 4GB.
-- Add above4gb flag argument to function x86_setup_var_mtrrs. Clearing
   this flag causes x86_setup_var_mtrrs() to omit MTRR ranges for
   DRAM above 4GB. AMD systems use this option to conserve MTRRs.
-- Northbridge.c change to deduct UMA memory from DRAM size reported
   by ram_resource. This corrects a problem where mtrr.c generates an
   unexpected variable MTRR range.
-- Correct problem causing build failure when CONFIG_GFXUMA=1 and
   CONFIG_VAR_MTRR_HOLE=0.
-- Reserve the UMA DRAM range for AMD K8 as is already done for AMD
   family 10h.
Tested with mahogany on ECS A780G-GM with 2GB and 4GB.
Tested with mahogany_fam10 on ECS A780G-GM with 2GB and 4GB.

Signed-off-by: Scott Duplichan <scott@notabs.org>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6067 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agomainboard/asus/m2v: Set DDR2 voltage to 1.8V
Tobias Diedrich [Fri, 12 Nov 2010 20:46:02 +0000 (20:46 +0000)]
mainboard/asus/m2v: Set DDR2 voltage to 1.8V

The power-on default is 1.95V, set the DDR2 voltage to
standards-conforming 1.8V.

I also measured with a multimeter to confirm this.

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6066 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoConsensus seems that this is wanted, integrated into the tree somehow.
Patrick Georgi [Fri, 12 Nov 2010 09:46:30 +0000 (09:46 +0000)]
Consensus seems that this is wanted, integrated into the tree somehow.
This isn't hooked up anywhere, so won't affect anything, except for
developers trying to remove configuration #defines.

Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6065 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAdd support for Fintek F71872 superio.
Jonathan Kollasch [Thu, 11 Nov 2010 22:25:55 +0000 (22:25 +0000)]
Add support for Fintek F71872 superio.

Signed-off-by: Jonathan Kollasch <jakllsch@kollasch.net>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6064 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoRemove superfluous Super I/O res0/res1 lines.
Uwe Hermann [Thu, 11 Nov 2010 13:14:55 +0000 (13:14 +0000)]
Remove superfluous Super I/O res0/res1 lines.

The pc_keyboard_init() function no longer takes any base addresses
since r5152 (passed in via res0/res1 variables previously), so drop them.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Patrick Georgi <patrick@georgi-clan.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6063 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

13 years agoAdd VT8237A id to src/southbridge/via/vt8237r/bootblock.c
Tobias Diedrich [Thu, 11 Nov 2010 05:12:01 +0000 (05:12 +0000)]
Add VT8237A id to src/southbridge/via/vt8237r/bootblock.c

I missed this one since it was working anyway, since
"The LPC BIOS ROM is always accessed when ISA addresses
 FFF80000-FFFFFFFF and 000F0000-000FFFFF are decoded" (VT8237R datasheet)
And the rom I use for testing is smaller than this 512KB default range.

Signed-off-by: Tobias Diedrich <ranma+coreboot@tdiedrich.de>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@6062 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1