PATCH: lx loads payload again, with new vsa, proper PCI bus scan etc.

[ron minnich]

With this set of changes, the lx target loads VSA and configures the PCI bus,
with VSA operating correctly. This is tested with AMD's recently released
new-model VSA code. Note that v3 build is correctly compressing the uncompressed
vsa payload. I am really happy with v3 so far.

Signed-off-by: Ronald G. Minnich

Sadly, the first jump to payload crashes and burns. Marc, can you
check this with fs2 for me? I still have not found mine ...

PS: PLEASE, if you are going to the coreboot summit April 3-5,
register at hpcsw.org. You can get good rates at the hotel through
april 6 even. I am staying there through sunday.

thanks

ron

PATCH: lx loads payload again, with new vsa, proper PCI bus scan etc.

[Carl-Daniel Hailfinger]

On 01.02.2008 20:28, ron minnich wrote:
> With this set of changes, the lx target loads VSA and configures the PCI bus,
> with VSA operating correctly. This is tested with AMD's recently released
> new-model VSA code. Note that v3 build is correctly compressing the uncompressed
> vsa payload. I am really happy with v3 so far.
>
> Signed-off-by: Ronald G. Minnich

That's great!

Acked-by: Carl-Daniel Hailfinger

This passed my compile tests for all targets except the artec dbe61,
which has been broken for a long time anyway.
A few minor comments. It would be great if you could address them before
committing.

northbridge/amd/geodelx/domain is a copy of northbridge/amd/geodelx/dts.
You probably want to use "svn mv" for that because it preserves history
and the old file was probably intended to have been moved, not copied.

northbridge/amd/geodelx/vsmsetup.c:247: warning: ‘biosint’ defined but
not used
Since the new VSA does not use BIOSINT services anymore, deleting
biosint and related functions from vsmsetup.c would shrink vsmsetup.c by
one fourth. Patch follows (could you merge it into your patch?):

Signed-off-by: Carl-Daniel Hailfinger

Index: northbridge/amd/geodelx/vsmsetup.c
===================================================================
--- northbridge/amd/geodelx/vsmsetup.c	(Revision 570)
+++ northbridge/amd/geodelx/vsmsetup.c	(Arbeitskopie)
@@ -227,83 +227,3 @@
    MEMSIZE = 0x12
 };

-int pcibios(unsigned long *pedi, unsigned long *pesi, unsigned long *pebp,
-	    unsigned long *pesp, unsigned long *pebx, unsigned long *pedx,
-	    unsigned long *pecx, unsigned long *peax, unsigned long *pflags);
-
-int handleint21(unsigned long *pedi, unsigned long *pesi, unsigned long *pebp,
-		unsigned long *pesp, unsigned long *pebx, unsigned long *pedx,
-		unsigned long *pecx, unsigned long *peax,
-		unsigned long *pflags);
-
-/* see the vga_exit() call. until we clean this up this function will remain here */
-static int biosint(unsigned long intnumber,
-	    unsigned long gsfs, unsigned long dses,
-	    unsigned long edi, unsigned long esi,
-	    unsigned long ebp, unsigned long esp,
-	    unsigned long ebx, unsigned long edx,
-	    unsigned long ecx, unsigned long eax,
-	    unsigned long cs_ip, unsigned short stackflags)
-{
-	unsigned long ip;
-	unsigned long cs;
-	unsigned long flags;
-	int ret = -1;
-
-	ip = cs_ip & 0xffff;
-	cs = cs_ip >> 16;
-	flags = stackflags;
-
-	printk(BIOS_DEBUG, "biosint: INT# 0x%lx\n", intnumber);
-	printk(BIOS_DEBUG, "biosint: eax 0x%lx ebx 0x%lx ecx 0x%lx edx 0x%lx\n",
-		     eax, ebx, ecx, edx);
-	printk(BIOS_DEBUG, "biosint: ebp 0x%lx esp 0x%lx edi 0x%lx esi 0x%lx\n",
-		     ebp, esp, edi, esi);
-	printk(BIOS_DEBUG, "biosint:  ip 0x%lx   cs 0x%lx  flags 0x%lx\n",
-		     ip, cs, flags);
-	printk(BIOS_DEBUG, "biosint: gs 0x%lx fs 0x%lx ds 0x%lx es 0x%lx\n",
-		     gsfs >> 16, gsfs & 0xffff, dses >> 16, dses & 0xffff);
-
-	// cases in a good compiler are just as good as your own tables.
-	switch (intnumber) {
-	case 0 ... 15:
-		// These are not BIOS service, but the CPU-generated exceptions
-		printk(BIOS_INFO, "biosint: Oops, exception %lu\n", intnumber);
-		if (esp < 0x1000) {
-			printk(BIOS_DEBUG, "Stack contents: ");
-			while (esp < 0x1000) {
-				printk(BIOS_DEBUG, "0x%04x ", *(unsigned short *)esp);
-				esp += 2;
-			}
-			printk(BIOS_DEBUG, "\n");
-		}
-		printk(BIOS_DEBUG, "biosint: Bailing out ... not now\n");
-		// "longjmp"
-		//vga_exit();
-		break;
-
-	case PCIBIOS:
-		ret = pcibios(&edi, &esi, &ebp, &esp,
-			      &ebx, &edx, &ecx, &eax, &flags);
-		break;
-	case MEMSIZE:
-		// who cares.
-		eax = 128 * 1024;
-		ret = 0;
-		break;
-	case 0x15:
-		ret = handleint21(&edi, &esi, &ebp, &esp,
-				  &ebx, &edx, &ecx, &eax, &flags);
-		break;
-	default:
-	  printk(BIOS_INFO, "BIOSINT: Unsupport int #0x%lx\n", intnumber);
-		break;
-	}
-	if (ret)
-		flags |= 1;	// carry flags
-	else
-		flags &= ~1;
-	stackflags = flags;
-	return ret;
-}
-

[ron minnich]

Committed revision 571.

I included your svn mv suggestion, and have also removed the pci
biosint support in vsmsetup.c

Thanks

ron

PATCH: lx loads payload again, with new vsa, proper PCI bus scan etc.

[Carl-Daniel Hailfinger]

On 01.02.2008 21:36, ron minnich wrote:
> Committed revision 571.
>
> I included your svn mv suggestion, and have also removed the pci
> biosint support in vsmsetup.c

Thanks, great!

Regards,
Carl-Daniel

PATCH: lx loads payload again, with new vsa, proper PCI bus scan etc.

[Marc Jones]

ron minnich wrote:
> Committed revision 571.
>
> I included your svn mv suggestion, and have also removed the pci
> biosint support in vsmsetup.c
>
> Thanks
>
> ron

Ron,

How is VSA being compressed and built in? I seem to have missed something.

Marc

[ron minnich]

On Feb 1, 2008 1:44 PM, Marc Jones  wrote:
> How is VSA being compressed and built in? I seem to have missed something.

No, I missed something: it's not being compressed. Oops. That's for buildrom.

./build/util/lar/lar -a build/bios.bin lxvsa:blob/vsa

I could nrv2b encode it by hand, then it would also work.

ron

PATCH: lx loads payload again, with new vsa, proper PCI bus scan etc.

[Marc Jones]

ron minnich wrote:
> On Feb 1, 2008 1:44 PM, Marc Jones  wrote:
> > How is VSA being compressed and built in? I seem to have missed something.
>
> No, I missed something: it's not being compressed. Oops. That's for buildrom.
>
> ./build/util/lar/lar -a build/bios.bin lxvsa:blob/vsa
>
> I could nrv2b encode it by hand, then it would also work.
>
> ron

OK, I agree it is for buildrom to compress.
Marc

PATCH: lx loads payload again, with new vsa, proper PCI bus scan etc.

[Carl-Daniel Hailfinger]

On 01.02.2008 22:50, ron minnich wrote:
> On Feb 1, 2008 1:44 PM, Marc Jones  wrote:
> > How is VSA being compressed and built in? I seem to have missed something.
>
> No, I missed something: it's not being compressed. Oops. That's for buildrom.
>
> ./build/util/lar/lar -a build/bios.bin lxvsa:blob/vsa
>
> I could nrv2b encode it by hand, then it would also work.

Why nrv2b? Default compression does work out just fine.

Regards,
Carl-Daniel

[ron minnich]

On Feb 1, 2008 2:08 PM, Carl-Daniel Hailfinger
 wrote:
> Why nrv2b? Default compression does work out just fine.

no real reason.

ron