coreboot.git
16 years agocs5536 IDE PWB flag was not getting set since it is 1<<14 and it was only doing a
Marc Jones [Tue, 6 May 2008 16:56:47 +0000 (16:56 +0000)]
cs5536 IDE PWB flag was not getting set since it is 1<<14 and it was only doing a
pci_write_config8.

Signed-off-by: Marc Jones <marc.jones@amd.com>
Acked-by: Ronald G. Minnich <rminnich@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3282 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoThis patch changes Config-lab.lb for qemu to use lzma like the other targets.
Myles Watson [Tue, 6 May 2008 15:17:43 +0000 (15:17 +0000)]
This patch changes Config-lab.lb for qemu to use lzma like the other targets.

Signed-off-by: Myles Watson <mylesgw@gmail.com>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3281 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoThis patch adds pc keyboard init function call for qemu in v2 since some payloads...
Aaron Lwe [Tue, 6 May 2008 15:02:22 +0000 (15:02 +0000)]
This patch adds pc keyboard init function call for qemu in v2 since some payloads assume
Coreboot initializes it.  Coreboot v3 already does it.

Signed-off-by: Aaron Lwe <aaron.lwe@gmail.com>
Acked-by: Myles Watson <mylesgw@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3280 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoFix various issues on MSI MS-7135 board.
Jonathan A. Kollasch [Tue, 6 May 2008 13:26:32 +0000 (13:26 +0000)]
Fix various issues on MSI MS-7135 board.

 - W83627THF is strapped to 0x4e, not 0x2e
 - there's no device 9 on PCI-E x1 bus, it should be device 0
 - add mptable entries for AGR slot, based on info in user manual
 - enable floppy drive controller so that some legacy VGA ROMs will work

Signed-off-by: Jonathan A. Kollasch <jakllsch@kollasch.net>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3279 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoThis patch changes the payload path for Config.lb; this board is supported by
Ward Vandewege [Mon, 5 May 2008 20:50:58 +0000 (20:50 +0000)]
This patch changes the payload path for Config.lb; this board is supported by
buildrom and this bit was forgotten during r3092.

This is a trivial patch.

Signed-off-by: Ward Vandewege <ward@gnu.org>
Acked-by: Ward Vandewege <ward@gnu.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3278 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoflashrom: Add a tested bitmap field to the flash chip table.
Peter Stuge [Sat, 3 May 2008 04:34:37 +0000 (04:34 +0000)]
flashrom: Add a tested bitmap field to the flash chip table.

Two bits indicate OK and BAD for each operation PROBE READ ERASE WRITE.
8 bits out of 32 are in use now. No bits set means nothing has been tested.
For chips with at least one operation that is not tested or not working, the
user is asked to email a report to a special email adress so that the table
can be updated.

All chips are TEST_UNTESTED for now.

Signed-off-by: Peter Stuge <peter@stuge.se>
Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3277 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoBy default, the Intel 3100 LPC interface enables only I/O range 0x3f8
Ed Swierk [Wed, 30 Apr 2008 18:29:35 +0000 (18:29 +0000)]
By default, the Intel 3100 LPC interface enables only I/O range 0x3f8
for both serial ports, making it challenging to use COM2 for the early
console.

Enable the traditional I/O ranges 0x3f8 for COM1 and 0x2f8 for COM2.

Signed-off-by: Ed Swierk <eswierk@arastra.com>
Acked-by: Joseph Smith <joe@settoplinux.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3276 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoflashrom: Enable ROM decode range to 1MB for vt8237r
Bari Ari [Tue, 29 Apr 2008 13:46:38 +0000 (13:46 +0000)]
flashrom: Enable ROM decode range to 1MB for vt8237r

Signed-off-by: Bari Ari <bari@onelabs.com>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3275 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoThe generic jedec.c does not work for the ST M50FLW flash
Claus Gindhart [Mon, 28 Apr 2008 17:51:09 +0000 (17:51 +0000)]
The generic jedec.c does not work for the ST M50FLW flash
devices, because they need an unlock command first.
For this reason, ST M50FLW support is moved to a
new HW support module, because any change in jedec.c
would bear the risk to cause problems with the already
supported devices.

It's already tested with ST M50FLW080A; the other
chips of this family i dont have available, so i couldnt
test it.

Signed-off-by: Claus Gindhart <claus.gindhart@kontron.com>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3274 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoflashrom: Handle NULL probe, erase and write function pointers in the
Peter Stuge [Mon, 28 Apr 2008 14:47:30 +0000 (14:47 +0000)]
flashrom: Handle NULL probe, erase and write function pointers in the
flashchips table. The read pointer was already checked properly.

Signed-off-by: Peter Stuge <peter@stuge.se>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3273 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: Add gettimeofday() and friends
Jordan Crouse [Fri, 25 Apr 2008 23:11:02 +0000 (23:11 +0000)]
libpayload:  Add gettimeofday() and friends

Add a gettimeofday() implementation - it works pretty well, but it
drifts a little bit so its not very suitable for keeping time.  It
works best to track changes in time over small periods of time.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3272 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: Fix a small but aggressive bug in printf()
Jordan Crouse [Fri, 25 Apr 2008 23:10:23 +0000 (23:10 +0000)]
libpayload:  Fix a small but aggressive bug in printf()

This was causing the returned counter value to be one more then it
should be when printing a single character.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Ward Vandewege <ward@gnu.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3271 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: Enable keyboard translation so that we can use scancode set 1
Jordan Crouse [Fri, 25 Apr 2008 23:09:39 +0000 (23:09 +0000)]
libpayload:  Enable keyboard translation so that we can use scancode set 1

The qemu keyboard controller defaults to using scancode set 2, we use set 1.
Turn on the translate mode in the keyboard controller to force the issue.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3270 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: Fix malloc allocation
Jordan Crouse [Fri, 25 Apr 2008 23:08:47 +0000 (23:08 +0000)]
libpayload:  Fix malloc allocation

Apparently the previous version worked on luck.  Fix the allocation
and add parens to better guide the compiler.  Also, halt() if
the heap is poisoned (like by an overrun).  Finally, fix calloc()
so that it actually works.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3269 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: Add the null terminator to the end of the duplicated string
Jordan Crouse [Fri, 25 Apr 2008 23:07:39 +0000 (23:07 +0000)]
libpayload:  Add the null terminator to the end of the duplicated string

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Myles Watson <mylesgw@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3268 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoChange abuild ROM_IMAGE_SIZE to match the standard s_c_fam10 Config.lb.
Marc Jones [Fri, 25 Apr 2008 22:56:57 +0000 (22:56 +0000)]
Change abuild ROM_IMAGE_SIZE to match the standard s_c_fam10 Config.lb.
The FAM10 code takes up more space in the uncompressed "ROMCC" portion
of coreboot. Also, It is still growing as features are added.

Signed-off-by: Marc Jones <marc.jones@amd.com>
Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3267 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoRemove inline from FAM10 CPU initialization functions.
Marc Jones [Fri, 25 Apr 2008 21:34:25 +0000 (21:34 +0000)]
Remove inline from FAM10 CPU initialization functions.
This doesn't save any space for me but it is the right thing to allow GCC to
optimize.

Signed-off-by: Marc Jones <marc.jones@amd.com>
Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3266 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoFix so pci device memory allocation does not use memory base address at 0xfec00000...
Aaron Lwe [Fri, 25 Apr 2008 02:02:33 +0000 (02:02 +0000)]
Fix so pci device memory allocation does not use memory base address at 0xfec00000, this is reserved for APIC.

Signed-off-by: Aaron Lwe <aaron.lwe@gmail.com>
Acked-by: Joseph Smith <joe@settoplinux.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3265 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoPayload location fix for buildrom (trivial).
Uwe Hermann [Fri, 25 Apr 2008 00:38:41 +0000 (00:38 +0000)]
Payload location fix for buildrom (trivial).

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@3264 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd CPUID processor name string support for Fam10 CPUs.
Marc Jones [Thu, 24 Apr 2008 20:03:13 +0000 (20:03 +0000)]
Add CPUID processor name string support for Fam10 CPUs.
Peter did a nice job cleaning up my initial patch. Thanks!

Signed-off-by: Marc Jones <marc.jones@amd.com>
Signed-off-by: Peter Stuge <peter@stuge.se>
Acked-by: Marc Jones <marc.jones@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3263 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoOn APs the ClLinesToNbDis was being left enabled from CAR setup.
Marc Jones [Thu, 24 Apr 2008 19:49:59 +0000 (19:49 +0000)]
On APs the ClLinesToNbDis was being left enabled from CAR setup.
Disabling it should help performance.

Signed-off-by: Marc Jones <marc.jones@amd.com>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3262 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoThis board (http://www.ieiworld.com/en/product_IPC.asp?model=PCISA-LX)
Nikolay Petukhov [Thu, 24 Apr 2008 13:37:01 +0000 (13:37 +0000)]
This board (ieiworld.com/en/product_IPC.asp?model=PCISA-LX)
is based on amd-lx800/cs5536.

Tutorial: http://www.coreboot.org/IEI_LX_800_Build_Tutorial

Signed-off-by: Nikolay Petukhov <nikolay.petukhov@gmail.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3261 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoFlash pages, which where excluded from updating using the exclude or the
Claus Gindhart [Thu, 24 Apr 2008 09:07:57 +0000 (09:07 +0000)]
Flash pages, which where excluded from updating using the exclude or the
layout option, as well as areas, whose flash contents already contain the
desired data, will be skipped.
These ensures absolute data security of critical areas (BIOS boot block),
e.g. against a sudden power off or a CPU hangup during flashing. As a
nice side effect, it speeds up the flash process, if the BIOS to be flashed
is very similar to the version in flash.

Signed-off-by: Claus Gindhart <claus.gindhart@kontron.com>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3260 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoSame old story: Fam10 needs more space again. My calculations say it
Carl-Daniel Hailfinger [Wed, 23 Apr 2008 22:54:40 +0000 (22:54 +0000)]
Same old story: Fam10 needs more space again. My calculations say it
needs 172 more bytes, give it 512 and hope that's enough for a while.
Trivial.

Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3259 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoTrivial payload location changes for buildrom.
Myles Watson [Wed, 23 Apr 2008 22:01:55 +0000 (22:01 +0000)]
Trivial payload location changes for buildrom.

Signed-off-by: Myles Watson <mylesgw@gmail.com>
Acked-by: Myles Watson <mylesgw@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3258 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoThese config files are so that buildrom can use these two boards.
Myles Watson [Wed, 23 Apr 2008 21:06:08 +0000 (21:06 +0000)]
These config files are so that buildrom can use these two boards.

Myles

Signed-off-by: Myles Watson <mylesgw@gmail.com>
Acked-by: Ward Vandewege <ward@gnu.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3257 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoThis is the sata irq patch for s2895 and ultra40. It also changes some broken
Myles Watson [Wed, 23 Apr 2008 20:40:55 +0000 (20:40 +0000)]
This is the sata irq patch for s2895 and ultra40. It also changes some broken
white space in the s2892 and s2891 mptable.c files.

Signed-off-by: Myles Watson <mylesgw@gmail.com>
Acked-by: Ward Vandewege <ward@gnu.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3256 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoFix irqs for secondary ports on both sata controllers.
Myles Watson [Wed, 23 Apr 2008 17:55:25 +0000 (17:55 +0000)]
Fix irqs for secondary ports on both sata controllers.

Signed-off-by: Myles Watson <mylesgw@gmail.com>
Acked-by: Ward Vandewege <ward@gnu.org>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3255 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoDetect SMSC SCH5027 (trivial).
Uwe Hermann [Wed, 23 Apr 2008 09:27:18 +0000 (09:27 +0000)]
Detect SMSC SCH5027 (trivial).

This chip seems to be very similar to the SMSC DME1737, for coreboot
purposes it might even work without any code changes.

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@3254 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoThis patch fixes the 3 broken sata ports on the Tyan s2891 (primary port on
Ward Vandewege [Wed, 23 Apr 2008 00:40:39 +0000 (00:40 +0000)]
This patch fixes the 3 broken sata ports on the Tyan s2891 (primary port on
secondary controller was ok). There were two problems: the master sata
controller was not being initialized, and the irqs for the secondary ports on
both controllers were not being set in the mptable.

Thanks for Jonathan Kollasch for all the help figuring out the IRQ problem.

While all ports work reliably under a recent kernel (2.6.24), sata is about
half as fast as under the proprietary bios, according to bonnie++. That still
needs fixing...

Signed-off-by: Ward Vandewege <ward@gnu.org>
Acked-by: Myles Watson <mylesgw@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3253 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoClean up and remove late initialization code that is no longer needed.
Marc Jones [Tue, 22 Apr 2008 23:32:56 +0000 (23:32 +0000)]
Clean up and remove late initialization code that is no longer needed.

Pstate intialization has moved to early init because it requires a warm reset.
Add CPUID setup and disable SMM access to late initialization.
Much of this code is leftover from porting from K8.

Signed-off-by: Marc Jones <marc.jones@amd.com>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3252 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoFind matching settings for each CPUs FID, VID, and P-state registers and initialize...
Marc Jones [Tue, 22 Apr 2008 23:27:53 +0000 (23:27 +0000)]
Find matching settings for each CPUs FID, VID, and P-state registers and initialize them.

Supports single and split plane systems. Set P0 on all cores for best performance.
All APs will be in hlt(C1).

The platform warm rest logic has been updated to alway reset for HT and FID/VID setup. It is not optional anymore.

Signed-off-by: Marc Jones <marc.jones@amd.com>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3251 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoUpdate the FAM10 microcode to current versions.
Marc Jones [Tue, 22 Apr 2008 23:20:07 +0000 (23:20 +0000)]
Update the FAM10 microcode to current versions.
In addition, AP microcode is now updated in early initialization to support errata settings that require it.

Signed-off-by: Marc Jones <marc.jones@amd.com>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3250 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoMissed this file in the previous check-in, r3248.
Marc Jones [Tue, 22 Apr 2008 23:09:34 +0000 (23:09 +0000)]
Missed this file in the previous check-in, r3248.

Add early MSR and PCI register initialization.
This fixes many default setting as well as erratas.
Some CPU core functions were moved from the HT init and platform specific code
to the generic Fam10 CPU code.

Signed-off-by: Marc Jones <marc.jones@amd.com>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3249 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd early MSR and PCI register initialization.
Marc Jones [Tue, 22 Apr 2008 22:11:31 +0000 (22:11 +0000)]
Add early MSR and PCI register initialization.
This fixes many default setting as well as erratas.
Some CPU core functions were moved from the HT init and platform specific code to the generic Fam10 CPU code.

Signed-off-by: Marc Jones <marc.jones@amd.com>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3248 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd support for a 'bootlog' module to coreinfo.
Uwe Hermann [Tue, 22 Apr 2008 20:19:53 +0000 (20:19 +0000)]
Add support for a 'bootlog' module to coreinfo.

It displays the coreboot printk buffer in RAM and let's you scroll through it.
This feature is only available for coreboot v3 though, as v2 doesn't have a
printk-buffer feature, yet.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3247 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoShow index numbers in the NVRAM dump, similar to the PCI config space dump.
Uwe Hermann [Tue, 22 Apr 2008 16:56:21 +0000 (16:56 +0000)]
Show index numbers in the NVRAM dump, similar to the PCI config space dump.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3246 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: Fix keyboard buglet
Jordan Crouse [Mon, 21 Apr 2008 22:33:58 +0000 (22:33 +0000)]
libpayload:  Fix keyboard buglet

This solves the multiple keystroke issue that popped up recently.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3245 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoThis trivial patch adds the SMSC SCH3112 Super I/O chip ID to the
Christopher Kilgour [Sat, 19 Apr 2008 13:32:19 +0000 (13:32 +0000)]
This trivial patch adds the SMSC SCH3112 Super I/O chip ID to the
generic SMSC support, and corrects a small typo.

With this patch, coreboot v2 on a mainboard with SCH3112 has been
demonstrated to correctly use the serial port.  No other chip
functions were tested.

Signed-off-by: Christopher Kilgour <techie@whiterocker.com>
Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3244 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoReplace buildtarget's check for --build-id with something
Ed Swierk [Fri, 18 Apr 2008 20:48:22 +0000 (20:48 +0000)]
Replace buildtarget's check for --build-id with something
a bit less awkward (pun intended).

Signed-off-by: Ed Swierk <eswierk@arastra.com>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3243 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAlter buildtarget to invoke the cross-compiler when
Ed Swierk [Fri, 18 Apr 2008 20:47:11 +0000 (20:47 +0000)]
Alter buildtarget to invoke the cross-compiler when
checking for --build-id, if the user has specified one by setting CC
in the environment; there's no point in checking the native linker in
this case.

Signed-off-by: Ed Swierk <eswierk@arastra.com>
Acked-by: Ronald G. Minnich <rminnich@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3242 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoChange default payload location for easier buildrom support (trivial).
Uwe Hermann [Wed, 16 Apr 2008 00:46:08 +0000 (00:46 +0000)]
Change default payload location for easier buildrom support (trivial).

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@3241 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoMove curses/speaker.c to drivers/ as it's not curses-specific (trivial).
Uwe Hermann [Tue, 15 Apr 2008 17:24:08 +0000 (17:24 +0000)]
Move curses/speaker.c to drivers/ as it's not curses-specific (trivial).

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@3240 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd the patch for building tint as payload, as well as a small README,
Uwe Hermann [Tue, 15 Apr 2008 16:47:20 +0000 (16:47 +0000)]
Add the patch for building tint as payload, as well as a small README,
into the payloads/external/tint directory.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3239 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoCorrect upper boundary for isxdigit.
Ulf Jordan [Sat, 12 Apr 2008 20:09:46 +0000 (20:09 +0000)]
Correct upper boundary for isxdigit.

Signed-off-by: Ulf Jordan <jordan@chalmers.se>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3238 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoDrop unneeded #includes, add EXIT_SUCCESS/EXIT_FAILURE (trivial).
Uwe Hermann [Fri, 11 Apr 2008 20:16:24 +0000 (20:16 +0000)]
Drop unneeded #includes, add EXIT_SUCCESS/EXIT_FAILURE (trivial).

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@3237 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoRewrite and filling of libc/ctype.c (thus adjusting copyright line).
Uwe Hermann [Fri, 11 Apr 2008 19:43:55 +0000 (19:43 +0000)]
Rewrite and filling of libc/ctype.c (thus adjusting copyright line).

Use a simple one-liner for each of the functions. You can surely optimize
the code some more, but I chose not to do that in order to keep it readable.
When compiling with -Os the size differences are minimal.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3236 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoConvert BIN2HEX/HEX2BIN to functions and add the abs() family
Uwe Hermann [Fri, 11 Apr 2008 18:38:04 +0000 (18:38 +0000)]
Convert BIN2HEX/HEX2BIN to functions and add the abs() family
of functions while we're at it.

hex2bin() now also supports upper-case input characters.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3235 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoVarious small consistency fixes (trivial):
Uwe Hermann [Fri, 11 Apr 2008 18:01:50 +0000 (18:01 +0000)]
Various small consistency fixes (trivial):

 - Use _FOO_H include guard format everywhere.

 - Add missing speaker.c prototypes to libpayload.h.

 - Consistently use short form u8/u16/u32 instead of uint8_t et. al.

 - kcofig: Use 'depends on' instead of 'depends', which seems deprecated.

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@3234 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: Add a Geode video driver
Jordan Crouse [Fri, 11 Apr 2008 15:48:21 +0000 (15:48 +0000)]
libpayload:  Add a Geode video driver

Add a Geode video driver in lieu of VGA on Geode LX devices

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3233 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoBring Fam10 memory controller init up to date with the latest AMD BKDG
Marc Jones (marc.jones [Fri, 11 Apr 2008 03:20:28 +0000 (03:20 +0000)]
Bring Fam10 memory controller init up to date with the latest AMD BKDG
recomendations.
Changes include the following:
fix > 4GB dqs tests
fix channel interleaving
ecc memory scrub updates
MC tristating updates
debug print changes
fix memory hoisting across nodes -
    The DRAM Hole Address Register is set via devx in each node, but the Node
    number <-> DRAM Base mapping and the Node number <-> DstNode mapping is
    set in Node 0. The memmap is setup on node0 and copied to the other nodes
    later. so dev, not devx. The bug was the hole was always being set on the
    first node.

Signed-off-by: Marc Jones (marc.jones@amd.com)
Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3232 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: Support functions for Geode
Jordan Crouse [Thu, 10 Apr 2008 22:50:44 +0000 (22:50 +0000)]
libpayload: Support functions for Geode

The Geode video driver will require a number of support functions,
including udelay(), PCI bus walking and MSRs.  This adds those functions
in preparation for the actual code.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3231 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: Add video console framework
Jordan Crouse [Thu, 10 Apr 2008 22:49:02 +0000 (22:49 +0000)]
libpayload:  Add video console framework

Add a framework for multiple video console drivers.  This is to prepare
for the Geode driver.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3230 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: Avoid .svn files in the header install
Jordan Crouse [Thu, 10 Apr 2008 17:57:42 +0000 (17:57 +0000)]
libpayload: Avoid .svn files in the header install

Slight tweak to to the install target to avoid copying .svn
files.  Trivial self ack.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3229 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agocoreinfo: Make coreinfo use the gcc-wrappers from libpayload
Jordan Crouse [Thu, 10 Apr 2008 00:05:41 +0000 (00:05 +0000)]
coreinfo:  Make coreinfo use the gcc-wrappers from libpayload

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3228 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoMake the sample Makefile a bit more generic, so it can be adapted more
Uwe Hermann [Wed, 9 Apr 2008 23:48:48 +0000 (23:48 +0000)]
Make the sample Makefile a bit more generic, so it can be adapted more
easily for other payloads. Also, add a 'distclean' target (trivial).

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@3227 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: Implement gcc wrappers for libpayload
Jordan Crouse [Wed, 9 Apr 2008 23:05:59 +0000 (23:05 +0000)]
libpayload: Implement gcc wrappers for libpayload

libpayload uses a ton of flags and other scary gcc and ld options.  These
wrappers hide most of that from the user, so that using libpayload is as
easy as lpgcc -o hello hello.c

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3226 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd missing prototypes for libc/rand.c functions (trivial).
Uwe Hermann [Tue, 8 Apr 2008 23:38:15 +0000 (23:38 +0000)]
Add missing prototypes for libc/rand.c functions (trivial).

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@3225 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoSmall curses fixes/additions for libpayload (trivial).
Uwe Hermann [Tue, 8 Apr 2008 23:30:22 +0000 (23:30 +0000)]
Small curses fixes/additions for libpayload (trivial).

 - Properly set LINES and COLS, needed for a real curses application.

 - Implement notimeout() and wtimeout(), which are trivial.

 - Implement a dummy flushinp() for now as it's needed by a curses
   application I'm porting, will be replaced by something useful later.

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@3224 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: Add a timeout function for getchar and getch
Jordan Crouse [Tue, 8 Apr 2008 23:21:33 +0000 (23:21 +0000)]
libpayload:  Add a timeout function for getchar and getch

Implement a timeout option for getchar() to return after so many
milliseconds.  Also implement the same thing for curses using
the halfdelay() function.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3223 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd rand/rand_r/srand functions for generating pseudo-random bytes.
Uwe Hermann [Mon, 7 Apr 2008 23:33:50 +0000 (23:33 +0000)]
Add rand/rand_r/srand functions for generating pseudo-random bytes.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3222 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoST M50FW016 and ST M50FW040 support the 82802ab command set, not jedec.
Ed Swierk [Mon, 7 Apr 2008 22:33:33 +0000 (22:33 +0000)]
ST M50FW016 and ST M50FW040 support the 82802ab command set, not jedec.

Signed-off-by: Ed Swierk <eswierk@arastra.com>
Acked-by: Joseph Smith <joe@smittys.pointclark.net>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3221 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoRe-add files I deleted by mistake in r3219. They are meant for a different
Marc Jones (marc.jones [Mon, 7 Apr 2008 18:11:03 +0000 (18:11 +0000)]
Re-add files I deleted by mistake in r3219. They are meant for a different
patch.

Signed-off-by: Marc Jones (marc.jones@amd.com)
Acked-by: Marc Jones (marc.jones@amd.com)
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3220 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoDon't check exclusive IRQ fieldin the PIR table.
Marc Jones(marc.jones [Mon, 7 Apr 2008 17:49:57 +0000 (17:49 +0000)]
Don't check exclusive IRQ fieldin the PIR table.
This field is rarely used (and not used in the LX tables).
There is not a good reason to mask off non-exclusive IRQs.

Signed-off-by: Marc Jones(marc.jones@amd.com)
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3219 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoThis patch halts the tco timer early in the boot process on all ICH series southbridges.
Joseph Smith [Sun, 6 Apr 2008 04:26:19 +0000 (04:26 +0000)]
This patch halts the tco timer early in the boot process on all ICH series southbridges.
It also keeps the boot processes from rebooting through out the coreboot process.

Signed-off-by: Joseph Smith <joe@smittys.pointclark.net>
Acked-by: Corey Osgood <corey.osgood@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3218 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agolibpayload: remove unneeded stack stuff
Jordan Crouse [Sat, 5 Apr 2008 01:07:27 +0000 (01:07 +0000)]
libpayload: remove unneeded stack stuff

Following on the previous code to streamline the libpayload init code,
this removes the now unneeded stack structures.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3217 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoFix the case where the user selects no modules in Kconfig at all.
Uwe Hermann [Fri, 4 Apr 2008 16:49:09 +0000 (16:49 +0000)]
Fix the case where the user selects no modules in Kconfig at all.
Until now, the build would break, and even if it didn't the ELF would
triple-fault in QEMU.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3216 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd a workaround for a bug in some binutils version which strictly
Carl-Daniel Hailfinger [Fri, 4 Apr 2008 15:02:45 +0000 (15:02 +0000)]
Add a workaround for a bug in some binutils version which strictly
interpret whitespace as macro argument delimiter. Since the code is
preprocessed by gcc and the tokenizer may insert whitespace, that can
fail. http://sourceware.org/bugzilla/show_bug.cgi?id=669

The same change was committed in r3044 to the AMD CAR code.

Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3215 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoDocument a rough estimate of how much space in the ELF file each of the
Uwe Hermann [Fri, 4 Apr 2008 13:28:10 +0000 (13:28 +0000)]
Document a rough estimate of how much space in the ELF file each of the
coreinfo features / modules will consume (trivial).

The measurements were done with libpayload r3213 (but compiled with -Os),
and coreinfo r3211 (also compiled with -Os).

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@3214 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd BIN2HEX and HEX2BIN macros (trivial).
Uwe Hermann [Fri, 4 Apr 2008 13:16:33 +0000 (13:16 +0000)]
Add BIN2HEX and HEX2BIN macros (trivial).

They're generally useful for lots of stuff, but especially for converting
to/from the compact 160 bit (20 byte) representation of SHA-1 hashes to
the "hex" representation (same as 'sha1sum' output), which is 40 bytes long.

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@3213 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd a SHA-1 implementation to libpayload.
Uwe Hermann [Thu, 3 Apr 2008 23:01:23 +0000 (23:01 +0000)]
Add a SHA-1 implementation to libpayload.

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@3212 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoDoing another 'make' after a 'make clean' was broken until now. Fix it
Uwe Hermann [Thu, 3 Apr 2008 22:20:35 +0000 (22:20 +0000)]
Doing another 'make' after a 'make clean' was broken until now. Fix it
by not deleting build/config.h during 'make clean' (only in 'make distclean').

Also, change the default behaviour of 'make' from asking the user to
run 'make config' (or similar) to actually _run_ 'make config' without
asking questions. It's always possible to explicitly invoke
'make menuconfig' or 'make xconfig' and so on, of course.

Finally, make _all_ targets (allyesconfig, randconfig, and so on)
generate a build/config.h file, as we always #include it.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3211 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd missing snprintf() to libc/printf.c (trivial).
Uwe Hermann [Wed, 2 Apr 2008 12:35:45 +0000 (12:35 +0000)]
Add missing snprintf() to libc/printf.c (trivial).
This is also taken from the HelenOS project.

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@3210 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoSetting an integrated southbridge device (like SATA or USB2.0) to
Ed Swierk [Tue, 1 Apr 2008 17:14:57 +0000 (17:14 +0000)]
Setting an integrated southbridge device (like SATA or USB2.0) to
"off" in Config.lb should cause the PCI device not to respond to
configuration requests.

Replace the existing code that I naively copied from esb6300 with
something that actually works on the 3100.

Signed-off-by: Ed Swierk <eswierk@arastra.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3209 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoRemove i82801DB files that I meant to delete in r3206.
Joseph Smith [Tue, 1 Apr 2008 17:05:22 +0000 (17:05 +0000)]
Remove i82801DB files that I meant to delete in r3206.

Signed-off-by: Joseph Smith <joe@smittys.pointclark.net>
Acked-by: Corey Osgood <corey.osgood@gmail.com>
Acked-by: Ed Swierk <eswierk@arastra.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3208 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoTiny style fix for consistency (trivial).
Ed Swierk [Tue, 1 Apr 2008 02:48:12 +0000 (02:48 +0000)]
Tiny style fix for consistency (trivial).

Signed-off-by: Ed Swierk <eswierk@arastra.com>
Acked-by: Ed Swierk <eswierk@arastra.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3207 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoRemoval of i82801DB (ICH4)
Joseph Smith [Tue, 1 Apr 2008 02:42:52 +0000 (02:42 +0000)]
Removal of i82801DB (ICH4)

There are no boards that use the i82801DB (ICH4). The code does NOT work.

Signed-off-by: Joseph Smith <joe@smittys.pointclark.net>
Acked-by: Corey Osgood <corey.osgood@gmail.com>
Acked-by: Ed Swierk <eswierk@arastra.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3206 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoThe early init code of several Intel southbridge chipsets calls
Ed Swierk [Tue, 1 Apr 2008 02:36:59 +0000 (02:36 +0000)]
The early init code of several Intel southbridge chipsets calls
pci_locate_device() to locate the SMBus controller and LPC bridge
devices on the PCI bus. Since these devices are always located at a
fixed PCI bus:device:function, the code can be simplified by
hardcoding the devices.

Signed-off-by: Ed Swierk <eswierk@arastra.com>
Acked-by: Corey Osgood <corey.osgood@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3205 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoLibpayload fixes to prevent triple-faults when running in QEMU.
Klaus Schnass [Mon, 31 Mar 2008 21:02:29 +0000 (21:02 +0000)]
Libpayload fixes to prevent triple-faults when running in QEMU.

Let the linker figure out the correct address and just CALL the
start_main entry point.

Signed-off-by: Klaus Schnass <dev@stuffit.at>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3204 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd support for an "NVRAM Dump" screen in coreinfo (optional), as well as for
Uwe Hermann [Mon, 31 Mar 2008 20:30:18 +0000 (20:30 +0000)]
Add support for an "NVRAM Dump" screen in coreinfo (optional), as well as for
displaying the current date/time in the lower-right corner (optional).

Also, only build/use coreinfo modules which were selected in kconfig. This
makes coreinfo truly modular, and you can save quite a bit of ROM space
by disabling unwanted parts of coreinfo.

Finally, simplify the Makefile a bit by getting rid of MODULES (and only
using OBJECTS).

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3203 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoFix the NVRAM access functions to work correctly for the
Uwe Hermann [Mon, 31 Mar 2008 20:21:49 +0000 (20:21 +0000)]
Fix the NVRAM access functions to work correctly for the
upper 128 bytes of NVRAM (if enabled).

For most chipsets this means using I/O ports 0x72/0x73, but at least
on some VIA chipsets (I tested the VIA VT8237R on actual hardware)
these ports won't work and you have to use 0x74/0x75. Thus, make this
a Kconfig option for now.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3202 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoWhitespace fixes (trivial).
Uwe Hermann [Mon, 31 Mar 2008 15:21:24 +0000 (15:21 +0000)]
Whitespace fixes (trivial).

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@3201 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoRename drivers/cmos.c to drivers/nvram.c (trivial).
Uwe Hermann [Mon, 31 Mar 2008 15:18:56 +0000 (15:18 +0000)]
Rename drivers/cmos.c to drivers/nvram.c (trivial).

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@3200 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoDue to popular demand, rename "CMOS" to "NVRAM" (trivial).
Uwe Hermann [Mon, 31 Mar 2008 15:17:39 +0000 (15:17 +0000)]
Due to popular demand, rename "CMOS" to "NVRAM" (trivial).

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@3199 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoLike other Intel chipsets, the Intel 3100 has a TCO timer that reboots
Ed Swierk [Sun, 30 Mar 2008 11:31:15 +0000 (11:31 +0000)]
Like other Intel chipsets, the Intel 3100 has a TCO timer that reboots
the system automatically unless software resets the timer
periodically. The extra reboot extends boot time by several seconds.

The attached patch adds a function to the Intel 3100 southbridge code
that halts the TCO timer, thus preventing this extra reboot, and calls
the function early in the boot process on the Mt. Arvon board.

It also fixes a bug in the LPC device initialization -- the ACPI BAR
enable flag is bit 7, not bit 4.

Signed-off-by: Ed Swierk <eswierk@arastra.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3198 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd support for the TeleVideo TC7020.
Kenji Noguchi [Sat, 29 Mar 2008 17:24:58 +0000 (17:24 +0000)]
Add support for the TeleVideo TC7020.

Signed-off-by: Kenji Noguchi <tokyo246@gmail.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3197 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoNow coreboot performs IRQ routing for some boards.
Nikolay Petukhov [Sat, 29 Mar 2008 16:59:27 +0000 (16:59 +0000)]
Now coreboot performs IRQ routing for some boards.
You can see this by executing commands like this:
grep -r pci_assign_irqs coreboot/src/*

This basically AMD/LX based boards: pcengines/alix1c,
digitallogic/msm800sev, artecgroup/dbe61, amd/norwich, amd/db800.

Also for AMD/GX1 based boards need a patch
[http://www.pengutronix.de/software/ptxdist/temporary-src/references/geode-5530.patch]
for the right IRQ setup.
AMD/GX1 based boards is: advantech/pcm-5820, asi/mb_5blmp, axus/tc320,
bcom/winnet100, eaglelion/5bcm, iei/nova4899r, iei/juki-511p.

I have two ideas.
1. Delete duplicate code from AMD/LX based boards.
2. Add IRQ routing for AMD/GX1 boards in coreboot.

The pirq.patch for IRQ routing logically consist from of two parts:

First part of pirq.patch independent from type chipsets and assign IRQ for
ever PCI device. It part based on AMD/LX write_pirq_routing_table() function.

Second part of pirq.patch depends of type chipset and set PIRQx lines
in interrupt router. This part supports only CS5530/5536 interrupt routers.

IRQ routing functionality is included through PIRQ_ROUTE in Config.lb.

Tested on iei/juki-511p(cs5530a), iei/pcisa-lx(cs5536) and also on
TeleVideo TC7020, see
http://www.coreboot.org/pipermail/coreboot/2007-December/027973.html.

Signed-off-by: Nikolay Petukhov <nikolay.petukhov@gmail.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3196 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years ago[libpayload] Work around sign-extending issue
Jordan Crouse [Sat, 29 Mar 2008 16:13:22 +0000 (16:13 +0000)]
[libpayload] Work around sign-extending issue

Somewhere characters are getting sign-extended, meaning that the
attributes of the drawing chars (>= 128) are wrong. Cast the value
before sending it to VGA.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3195 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoMake a few array entries only as big as they absolutely need to be (trivial).
Uwe Hermann [Sat, 29 Mar 2008 01:35:21 +0000 (01:35 +0000)]
Make a few array entries only as big as they absolutely need to be (trivial).

This decreases the size of the superiotool binary from ca. 1.1 MB to 600 KB.

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@3194 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoFix for irq routing issues.
Joseph Smith [Fri, 28 Mar 2008 03:35:11 +0000 (03:35 +0000)]
Fix for irq routing issues.
Signed-off-by: Joseph Smith <joe@smittys.pointclark.net>
Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3193 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd initial support for some basic CMOS read/write functions and the
Uwe Hermann [Thu, 27 Mar 2008 23:26:40 +0000 (23:26 +0000)]
Add initial support for some basic CMOS read/write functions and the
bcd2dec()/dec2bcd() functions we'll need for (among other things)
converting some date/time parameters in CMOS.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3192 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoVarious tiny fixes (trivial):
Uwe Hermann [Thu, 27 Mar 2008 20:46:49 +0000 (20:46 +0000)]
Various tiny fixes (trivial):

 - Show PCI IDs as 4-digit numbers always.

 - Cosmetic changes to make UI look more consistent.

 - Drop MODULE_COUNT #define and use ARRAY_SIZE() where needed.

 - Small fix to improve build system (create build/ when not there).

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@3191 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoDrop -Os in libpayload for now, it causes run-time problems for some
Uwe Hermann [Thu, 27 Mar 2008 19:11:44 +0000 (19:11 +0000)]
Drop -Os in libpayload for now, it causes run-time problems for some
strange reason (broken curses/VGA display, maybe others).

It'll be re-enabled when we fixed that.

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@3190 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoMake the getpir output compile (Closes #70).
Jon Dufresne [Tue, 25 Mar 2008 19:43:01 +0000 (19:43 +0000)]
Make the getpir output compile (Closes #70).

Signed-off-by: Jon Dufresne <jon.dufresne@gmail.com>
Acked-by: Uwe Hermann <uwe@hermann-uwe.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3189 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoAdd a recent kconfig version to coreinfo, in order to make the
Uwe Hermann [Mon, 24 Mar 2008 15:47:49 +0000 (15:47 +0000)]
Add a recent kconfig version to coreinfo, in order to make the
supported features configurable later (currently unused). Store all
build files and results (coreinfo.elf) in build/ now.

I'm self-acking this as (though non-"trivial") it doesn't affect coreinfo
in its functionality, this is more or less a "cosmetic" change to the
build interface ("make" -> "make menuconfig && make").

This is a kconfig checkout from the Linux kernel (where kconfig is being
actively maintained) from 03/2008. The hash identifying the last commit
to kconfig is 587c90616a5b44e6ccfac38e64d4fecee51d588c.

The amount of changes to kconfig itself has been kept as small as possible
to keep the diff small and to ease updating/porting to newer kconfig versions.
The following changes were performed on the upstream Linux kconfig:

 - s/kernel/coreinfo/, and s/Linux/coreinfo/ in various strings.

 - Consistently use the env. variable KERNELVERSION in all kconfig
   interfaces -- e.g. config/menuconfig/gconfig/xconfig -- as version number.

 - Hardcode our paths/filenames in some places (could be improved upstream).

 - Always write .config and build/config.h, no matter which kconfig
   interface is used (config/menuconfig/gconfig/xconfig). We want to
   include build/config.h in our code.

 - Adapt the kconfig Makefile for our purposes (build/ directory, rules, etc).

In addition, a few items in the coreinfo Makefile are needed for this to work.

This kconfig setup is successfully tested with all targets from 'make help':

  config          - Update current config utilising a line-oriented program
  menuconfig      - Update current config utilising a menu based program
  xconfig         - Update current config utilising a QT based front-end
  gconfig         - Update current config utilising a GTK based front-end
  oldconfig       - Update current config utilising a provided .config as base
  silentoldconfig - Same as oldconfig, but quietly
  randconfig      - New config with random answer to all options
  defconfig       - New config with default answer to all options
  allmodconfig    - New config selecting modules when possible
  allyesconfig    - New config where all options are accepted with yes
  allnoconfig     - New config where all options are answered with no

For 'make defconfig' to work you have to do (which we don't need in coreinfo):

  $ mkdir configs; touch configs/defconfig

You can also use 'make foo_defconfig' in which case kconfig will use a
file called 'configs/foo_defconfig' as basis.

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@3188 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoMake functions static (where possible) to reduce code size (trivial).
Uwe Hermann [Sun, 23 Mar 2008 15:34:04 +0000 (15:34 +0000)]
Make functions static (where possible) to reduce code size (trivial).
Also, disable header() for now, as it's not being used.

Here are some stats on size differences:

 - ls

23820 coreinfo.old.elf
23564 coreinfo.new.elf

 - size *elf

   text    data     bss     dec     hex filename
  15199    2468  181904  199571   30b93 coreinfo.old.elf
  14934    2468  181912  199314   30a92 coreinfo.new.elf

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@3187 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoMake cursor positioning work by using both halves of the
Jonathan A. Kollasch [Sat, 22 Mar 2008 15:27:26 +0000 (15:27 +0000)]
Make cursor positioning work by using both halves of the
VGA cursor position register.

Have vga_scroll_up() and vga_clear_line() present row/column
arguments to the VIDEO() macro in the right order.

Signed-off-by: Jonathan A. Kollasch <jakllsch@kollasch.net>
Acked-by: Jordan Crouse <jordan.crouse@amd.com>
-This line, and those below, will be ignored--

M    libpayload/drivers/vga.c

git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3186 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoCosmetics, fix typos (trivial).
Uwe Hermann [Fri, 21 Mar 2008 18:37:23 +0000 (18:37 +0000)]
Cosmetics, fix typos (trivial).

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@3185 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoQuickfix for libpayload's strcpy() to properly NUL-terminate strings (trivial).
Uwe Hermann [Fri, 21 Mar 2008 15:47:38 +0000 (15:47 +0000)]
Quickfix for libpayload's strcpy() to properly NUL-terminate strings (trivial).

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@3184 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

16 years agoFollowing patch adds K8M890 support. It initializes the AGP and graphics UMA.
Rudolf Marek [Thu, 20 Mar 2008 21:19:50 +0000 (21:19 +0000)]
Following patch adds K8M890 support. It initializes the AGP and graphics UMA.
The V-link setup and HT bridge is redone, because VT8237A has it in another
device. So far following combination of chipsets should now work:

K8T890CE + VT8237R
K8M890(CE) + VT8237R

VIA PC1 brige moved to NB code (vt8237r_bridge.c -> k8t890_bridge.c) and
notes about K8M890 support were added.

Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@3183 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1