Kyösti Mälkki (kyosti.malkki(a)gmail.com) just uploaded a new patch set to gerrit, which you can find at http://review.coreboot.org/435
-gerrit
commit 1eefdb3bbbd582042a2b332f55df0a54c3d8cae9
Author: Kyösti Mälkki <kyosti.malkki(a)gmail.com>
Date: Sun Nov 20 20:08:15 2011 +0200
Via Epia-N and C3: Set ioapic delivery type in Kconfig
The original comment says it's a Via C3 and not Epia requirement
to deliver IOAPIC interrupts on APIC serial bus. If someone can
confirm this, set delivery type under cpu/via/model_c3 instead.
Change-Id: I73c55755e0ec1ac5756b4ee7ccdfc8eb93184e4f
Signed-off-by: Kyösti Mälkki <kyosti.malkki(a)gmail.com>
---
src/Kconfig | 4 ++++
src/arch/x86/lib/ioapic.c | 9 ---------
src/mainboard/via/epia-n/Kconfig | 5 +++++
3 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/src/Kconfig b/src/Kconfig
index 525d452..1df339d 100644
--- a/src/Kconfig
+++ b/src/Kconfig
@@ -239,6 +239,10 @@ config IOAPIC
bool
default n
+config IOAPIC_DELIVERY_TYPE
+ int
+ default 1
+
# TODO: Can probably be removed once all chipsets have kconfig options for it.
config VIDEO_MB
int
diff --git a/src/arch/x86/lib/ioapic.c b/src/arch/x86/lib/ioapic.c
index 91e43a6..fee2ca1 100644
--- a/src/arch/x86/lib/ioapic.c
+++ b/src/arch/x86/lib/ioapic.c
@@ -180,15 +180,6 @@ static void load_vectors(u32 ioapic_base, u8 bsp_lapicid)
}
}
-// XXX this decision should probably be made elsewhere, and
-// it's the C3, not the EPIA this depends on.
-
-#if CONFIG_EPIA_VT8237R_INIT
-#define CONFIG_IOAPIC_DELIVERY_TYPE 0
-#else
-#define CONFIG_IOAPIC_DELIVERY_TYPE 1
-#endif
-
/**
* Assign IOAPIC with an ID
*
diff --git a/src/mainboard/via/epia-n/Kconfig b/src/mainboard/via/epia-n/Kconfig
index 4806753..5a88f6a 100644
--- a/src/mainboard/via/epia-n/Kconfig
+++ b/src/mainboard/via/epia-n/Kconfig
@@ -28,4 +28,9 @@ config IRQ_SLOT_COUNT
int
default 7
+# IRQ delivery on APIC bus may be a Via C3 CPU requirement.
+config IOAPIC_DELIVERY_TYPE
+ int
+ default 0
+
endif # BOARD_VIA_EPIA_N
Having put off to reply, this may be a reason to do so once more :)
I have 5 or 6 Wyse S10 (which are very very similar to the S50's)
thinclients with 1gb 'disk-on-chip' IDE disks and 256MB Ram (over the
stock 128mb ram and no d-o-c, e.g. using bios only) of which I'm willing
to donate 1 or even 2 if demand is high enough. I'll cover shipping
within europe (and other places we can talk) and throw in the 100-240V
adapter with EU power plug. I'm not skilled enough to do the porting
myself, unless i'd have loads of help (i know C but no clue where to
state and little time).
Let me know if your interested, or if there's some other procedure i'd
have to adhere to when donating 'stuff'.
oliver
On 18-11-11 21:00, Marc Jones wrote:
> On Wed, Nov 16, 2011 at 9:06 AM, JM<fijam(a)archlinux.us> wrote:
>> Hello,
>>
>> I am interested in replacing the BIOS on WYSE S50 with coreboot. I
>> have browsed the mailing list archives, however, some of the reports
>> are a bit unclear. I'd appreciate if someone could shed some light on
>> the following concerns of mine:
>>
>> 1. Does the board work after RAMBASE rework (i.e. is ticket #163 really fixed)?
>>
>> 2. Is the current support of this board sufficient to boot Linux? More
>> specifically, I have come across some reports of broken vsa and acpi,
>> is this still an issue? If so, how does the lack of those features
>> affect normal Linux operation?
>>
>> 3. When booting Linux from USB using default manufacturer's BIOS the
>> ATA port is being disabled. Is this a software or hardware limitation?
>> In other words, if I boot from USB with coreboot/SeaBIOS would I have
>> ATA enabled?
>>
>> 4. When I boot the Linux kernel with the default BIOS it occasionally
>> hangs on initializing the PCI subsystem. Is this a result of a bug in
>> vsa code (virtualized pci) or BIOS? Does the problem persist with
>> coreboot/SeaBIOS and AMD's vsa? I noticed something called OpenVSA,
>> does it support Geode GX?
>>
>> 5. Can I help in any way with the development not being a programmer?
>>
>> Thanks,
>> JM
> Hi JM,
>
> It has been some time since anyone has worked on Geode. You may have
> to try it to find out. Before you do, make sure you have a way to
> recover the BIOS image.
>
> It is dated and the build process has changed, but you should read the
> geode porting page. It talks about VSA and graphics and stuff. The VSA
> was openned and used on the OLPC project. It requires the Microsoft
> tools to build and no one has ported it to gcc yet. There is a link to
> the binary on the geode page. You should stop by the IRC #coreboot if
> you need a hand building coreboot.
>
> http://www.coreboot.org/AMD_Geode_Porting_Guide
>
> Marc
>
> --
> http://se-eng.com
>
Hi! I'm new to this list so I'll start by introducing myself a bit: I'm
a french student involved in free software: I use GNU/Linux distros on
all my PCs, mostly with the fully free Trisquel distro and I'm also a
developer (I know C quite well) mostly involved in the Replicant project
(fully free Android derivate) as a hacker (lower-level stuff).
The lead dev on the project (GNUtoo) decided to look at how doable it is
to port his board to coreboot, so I decided to follow his lead and look
at which of the motherboards I own would be the best one for coreboot. I
have 2 boards that might be good candidates, ASRock ConRoeXFire-eSATA2
(which has a socketed PLCC32 chip) and Asus P5LP-LE (Lithium) with a
soldered PLCC32 chip. Both boards should have supported
North/Southbridge and SuperIO already supported in coreboot.
I decided to start the work on the ASRock ConRoeXFire-eSATA2 one and
thanks to the help I got on IRC, it has been possible to add support for
the board on flashrom in a few hours. I also found another compatible
flash (same size, even same model apparently) so I can use it to flash
coreboot images.
So here are the technical details now:
* northbridge is Intel Corporation 82945G/GZ/P/PL
* southbridge is Intel Corporation N10/ICH 7 Family
* superio is Winbond W83627EHG
(see lspci out for more details)
So the board coming closer to mine is Kontron 986LCD-M/mITX. I forked
the device files for this one and I did this:
* replaced W83627THG references to W83627EHG
* used superio init function from another board (ASUS A8V-E SE)
* modified this function to have clcksel to 48MHz (as told on IRC)
* called w83627ehg_enable_serial(SERIAL_DEV, CONFIG_TTYS0_BASE); between
superio init and console init
So far I got serial working, but the northbridge I use isn't the exact
model on the board and the model on the board isn't supported by
coreboot. Though: some stuff seems to work with the other northbridge
code and I found wide documentation about the intel 945G/GZ/P/PL chip
used.
So I'm wondering what the next steps are… Adding support for the correct
northbridge?
--
Paul Kocialkowski
* Website : http://www.paulk.fr/
* Blog : http://blog.paulk.fr/
* Microblogging : http://status.paulk.fr/
> It has been some time since anyone has worked on Geode. You may have
> to try it to find out. Before you do, make sure you have a way to
> recover the BIOS image.
Thanks for your reply. The chip in question (SST49LF020A in PLCC) only
supports LPC and parallel programming. Do you think it would be
feasible to purchase another device (they come at about 10 euros) and
hot-swap the ROM for reflashing if something went wrong?
Alternatively, any hints for 'pulling out' the LPC interface from
another device? Sorry if my questions are a bit elementary.
> It is dated and the build process has changed, but you should read the
> geode porting page. It talks about VSA and graphics and stuff. The VSA
> was openned and used on the OLPC project. It requires the Microsoft
> tools to build and no one has ported it to gcc yet. There is a link to
> the binary on the geode page. You should stop by the IRC #coreboot if
> you need a hand building coreboot.
>
> http://www.coreboot.org/AMD_Geode_Porting_Guide
I'll be sure to take a look at that.
Regards,
Jan
the following patch was just integrated into master:
commit ac9431d26ba477811c2e27c8ee1a70109ca671fb
Author: Kerry Sheh <shekairui(a)gmail.com>
Date: Tue Nov 15 21:27:57 2011 +0800
mainboard: Add AMD unionstation RDK support
AMD unionstation Reference Design Kit is Designed for hd settop box application.
This platform using family14 APU, SB800 southbridge.
Vgabios is required, can download vgabios from AMD NDA website.
Verified Feature:
HDMI, LAN, mini-pcie slots, sata, usb, analog audio and
optical fiber digital audio output.
Change-Id: Ib1d1d8c889d6fb29f4298b57dfe5c5c1cea1431c
Signed-off-by: Kerry She <kerry.she(a)amd.com>
Signed-off-by: Kerry She <shekairui(a)gmail.com>
Build-Tested: build bot (Jenkins) at Tue Nov 15 14:43:53 2011, giving +1
Reviewed-By: Marc Jones <marcj303(a)gmail.com> at Fri Nov 18 21:43:19 2011, giving +2
See http://review.coreboot.org/434 for details.
-gerrit
the following patch was just integrated into master:
commit 8558795ef1369052cf634e2d6c5f33a2a11ad718
Author: Kerry Sheh <shekairui(a)gmail.com>
Date: Tue Nov 15 21:27:07 2011 +0800
mainboard: Add AMD southstation RDK support
AMD southstation Reference Design Kit is designed for NAS application.
This platform using family14 RevC0 processor, SB850 southbridge.
Vgabios and Promise RAID Option ROM is required for hardware RAID support,
can retrieve from the AMD NDA website.
Verified feature:
HDMI, LAN, usb and mini-pcie slot.
RAID0, RAID1 RAID10 and RAID5 upto 6 sata hard drive with ubuntu server 10.10.
Change-Id: I16e6f5dab8b0d634e186068c81436db77fb4475a
Signed-off-by: Kerry She <kerry.she(a)amd.com>
Signed-off-by: Kerry She <shekairui(a)gmail.com>
Build-Tested: build bot (Jenkins) at Tue Nov 15 14:33:48 2011, giving +1
Reviewed-By: Marc Jones <marcj303(a)gmail.com> at Fri Nov 18 21:38:58 2011, giving +2
See http://review.coreboot.org/433 for details.
-gerrit
On Wed, Nov 16, 2011 at 9:06 AM, JM <fijam(a)archlinux.us> wrote:
> Hello,
>
> I am interested in replacing the BIOS on WYSE S50 with coreboot. I
> have browsed the mailing list archives, however, some of the reports
> are a bit unclear. I'd appreciate if someone could shed some light on
> the following concerns of mine:
>
> 1. Does the board work after RAMBASE rework (i.e. is ticket #163 really fixed)?
>
> 2. Is the current support of this board sufficient to boot Linux? More
> specifically, I have come across some reports of broken vsa and acpi,
> is this still an issue? If so, how does the lack of those features
> affect normal Linux operation?
>
> 3. When booting Linux from USB using default manufacturer's BIOS the
> ATA port is being disabled. Is this a software or hardware limitation?
> In other words, if I boot from USB with coreboot/SeaBIOS would I have
> ATA enabled?
>
> 4. When I boot the Linux kernel with the default BIOS it occasionally
> hangs on initializing the PCI subsystem. Is this a result of a bug in
> vsa code (virtualized pci) or BIOS? Does the problem persist with
> coreboot/SeaBIOS and AMD's vsa? I noticed something called OpenVSA,
> does it support Geode GX?
>
> 5. Can I help in any way with the development not being a programmer?
>
> Thanks,
> JM
Hi JM,
It has been some time since anyone has worked on Geode. You may have
to try it to find out. Before you do, make sure you have a way to
recover the BIOS image.
It is dated and the build process has changed, but you should read the
geode porting page. It talks about VSA and graphics and stuff. The VSA
was openned and used on the OLPC project. It requires the Microsoft
tools to build and no one has ported it to gcc yet. There is a link to
the binary on the geode page. You should stop by the IRC #coreboot if
you need a hand building coreboot.
http://www.coreboot.org/AMD_Geode_Porting_Guide
Marc
--
http://se-eng.com