Asus F2A85-M progress

[Rudolf Marek]

      Hi all,

I have working serial now! As the result I created the SerialICE support for 
this board. Next struggle is AGESA. It seems we dont have any FM2 board and even 
if I tweaked the thatcher board config for FM2 and SPD addresses too it simply 
does not do any memory init.

I would appreciate any help in this direction. I'm attaching my WIP patch for a 
reference.


Thanks
Rudolf

[Rudolf Marek]

      Hi

Anyone knows what happen to my attachment?

Trying second time, first time it was also included...

Thanks
Rudolf

[Stefan Reinauer]

      * Rudolf Marek <r.marek(a)assembler.cz> [121105 23:02]:
  Hi
 
 Anyone knows what happen to my attachment?
 
 Trying second time, first time it was also included...
 
 Thanks
 Rudolf
Weird... there does not seem to be an attachment, either time.
Can you upload the patch to gerrit instead? 

Stefan

[Rudolf Marek]

      Hi all,

The attachment was:

  Content-Type: text/x-diff;
  name="dump.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
  filename="dump.patch"

Lets try again ;)

Thanks
Rudolf

[Rudolf Marek]

      Hi all,

I'm able to boot into linux with one 2GB module in "A2" slot!

Major issue: no VGA PCI device enabled - no clue what do I need to do, most 
likely PlatformGnbPcie.c is wrong but I dont know how to fix it.

I think I succeed to enable the GPP on FCH and now I see soldered realtek chip. 
This was done via #define DFLT_FCH_GPP_LINK_CONFIG.

XHCI is disabled, I need to enable firmware load.

Is anyone buying this board too?

Thanks
Rudolf

[Rudolf Marek]

      Hi all,

Good news is that the new Asus F2A85X seems to work pretty fine here except for 
VGA (but PCIe/PCI slot do work) I tried only one AHCI port and I need to figure 
out how to enable USB30 ports,but linux boots and does not complain about 
anything. I even use dual channel and 4GB RAM and fixed ACPI IRQ routing too.

Maybe it is slowly time to prepare a patch, I think I will need to write a 
porting guide + specific AGESA bits.

1) is it OK to in one patch just copy amd/tchatcher to my new board folder (git 
add style)
2) second patchset will do the changes...

Is this equivalent of svn copy?

Thanks
Rudolf

[Rudolf Marek]

      Hi all,

Here comes first FM2 based board to coreboot!

http://review.coreboot.org/#/c/2038/

The patch is based on Thatcher board. So far it boots Linux (3.2/3.7), no VGA,
internal network adapter works, AHCI works. External PCI/PCIe slots
works too. Power management/ACPI seems to work. This is a good start.

USB and XHCI untested but visible.

The external VGA does work but it is not POSTed in SeaBIOS for some reason. The 
post by DRI driver works... Reason so far unknown.

Thanks
Rudolf

Asus F2A85-M progress: patch send to Gerrit for review

[Paul Menzel]

      Dear Rudolf,


Am Sonntag, den 16.12.2012, 20:39 +0100 schrieb Rudolf Marek:

  Here comes first FM2 based board to coreboot!
 
 http://review.coreboot.org/#/c/2038/
awesome! If this gets merged, let’s not forget to blog about it at
<http://blogs.coreboot.org/>.

  The patch is based on Thatcher board. So far it boots
Linux (3.2/3.7), no VGA,
 internal network adapter works, AHCI works. External PCI/PCIe slots
 works too. Power management/ACPI seems to work. This is a good start.
 
 USB and XHCI untested but visible.
 
 The external VGA does work but it is not POSTed in SeaBIOS for some reason. The 
 post by DRI driver works... Reason so far unknown.
Please also add that to the commit message (`git commit --amend`).

Could you please attach all the log files to your response or to the
Wiki? coreboot serial output, `dmesg`, `/var/log/syslog`,
`/var/log/Xorg.0.log`. That would be great!


Thanks,

Paul

[Rudolf Marek]

      Hi all,

Good news, I managed to get external and internal VGA to work!

A) for internal VGA:
Boot the legacy BIOS, and use http://www.coreboot.org/VGA_support chapter 
extracting from your system:

dd if=/dev/mem of=vgabios.bin bs=1k count=64 skip=768

You will need following patch to seabios:

--- a/src/optionroms.c
+++ b/src/optionroms.c
@@ -215,7 +215,10 @@ is_pci_vga(struct pci_device *pci)
  {
      if (pci->class != PCI_CLASS_DISPLAY_VGA)
          return 0;
-    u16 cmd = pci_config_readw(pci->bdf, PCI_COMMAND);
+    u16 cmd = pci_config_readw(pci->bdf, PCI_COMMAND) | PCI_COMMAND_IO | 
PCI_COMMAND_MEMORY;
+
+    pci_config_writew(pci->bdf, PCI_COMMAND, cmd);
+
      if (!(cmd & PCI_COMMAND_IO && cmd & PCI_COMMAND_MEMORY))
          return 0;
      while (pci->parent) {

Reason is unknown, I see coreboot is writing 7 to cmd, but there is actually 
6... Maybe there is some magic about IO decode bit...

B) for external VGA... this is working now out of the box, I had to enable the 
bridge devices in devictree.cb otherwise the PCI setup is real mess.

C) Known issues:

So far I observe that Linux complains about pref regions:

[    0.648901] pci 0000:00:02.0: no compatible bridge window for [mem 
0xb0000000-0xbfffffff 64bit pref]
[    0.658001] pci 0000:00:15.1: no compatible bridge window for [mem 
0xc0000000-0xc00fffff 64bit pref]
[    0.667124] pci 0000:01:00.0: no compatible bridge window for [mem 
0xb0000000-0xbfffffff 64bit pref]
[    0.676229] pci 0000:04:00.0: no compatible bridge window for [mem 
0xc0004000-0xc0004fff 64bit pref]
[    0.685332] pci 0000:04:00.0: no compatible bridge window for [mem 
0xc0000000-0xc0003fff 64bit pref]

But looking to the setup from coreboot, I can't see where is the problem. Full 
boot log is attached, it was done without a internal GFX with external GFX (the 
internal GFX gets disabled if VGA ROM BIOS is not found)

Kernel 3.2 gets always a oops in HD audio, not sure what this can be, maybe I 
have something wrong with VERB configs, which were blindly copied from Thatcher.

I'm attaching the bootlog.

Thanks
Rudolf

      Rudolf Marek writes:

  Maybe it is slowly time to prepare a patch, I think I
will need to write a  
 porting guide + specific AGESA bits.
I would be really helpful if you would write a guide. I've been trying to
port the asus m5a88 board to agesa but I haven't managed to get it working
correctly yet.
So a guide explaining how to port + AGESA would be great.

[Bernhard Urban]

      On Mon, Dec 17, 2012 at 2:33 PM,  <darkdefende(a)gmail.com> wrote:
  Rudolf Marek writes:
  Maybe it is slowly time to prepare a patch, I
think I will need to write a
 porting guide + specific AGESA bits.
 I would be really helpful if you would write a guide. I've been trying to
 port the asus m5a88 board to agesa but I haven't managed to get it working
 correctly yet.
 So a guide explaining how to port + AGESA would be great.
+1
similar situation here with the asus m5a99x (yeah, progress is slow... :))


btw, great work ruik :-)

[Andrey Gursky]

      Hi Rudolf!

2012/12/15, Rudolf Marek <r.marek(a)assembler.cz>cz>:
  I think I will need to write a porting guide +
specific AGESA bits.
It would be really great to have something more specifically than
http://www.coreboot.org/AMD_Family_10h_Porting_Guide!

It turned out that I've made not the best choice of my gigabyte dual
bios AM3+-motherboard, because they have changed the procedure of BIOS
restoring from manual user controlled to automatic. And I couldn't
find yet any information where this automatic is located. Maybe it
would be possible to workaround it. Otherwise I must wait until I can
get a new one.

Andrey