Re: [9fans] Standalone cpu/auth..

Re: [9fans] Standalone cpu/auth..

[David Gordon Hogan]

> > I just tried the cpu/auth standalone on a dual PIII with Intel
> > server mother board. At boot time, i see the message..
> > .....
> > pcirouting: South bridge FFFF, FFFF not found
> > .....
> > pcirouting: Can't find south bridge PCI 255.37.7

I've just realized that the first message is coming out
of 9load, which has an older version of pcirouting()!

It's generally safe to ignore these messages (assuming
that everything else is working ok).  We'll add some
more sanity checking in a future version, which should
get rid of messages like the above.

Re: [9fans] Standalone cpu/auth..

[Ish Rattan]

On Thu, 16 May 2002, David Gordon Hogan wrote:

> > > I just tried the cpu/auth standalone on a dual PIII with Intel
> > > server mother board. At boot time, i see the message..
> > > .....
> > > pcirouting: South bridge FFFF, FFFF not found
> > > .....
> > > pcirouting: Can't find south bridge PCI 255.37.7
>
> I've just realized that the first message is coming out
> of 9load, which has an older version of pcirouting()!
Looks like it.

I have not tested it much but seems to be working ok.
-ishwar

[9fans] Standalone cpu/auth..

[Ish Rattan]

I am seeing wren i/o errors and pahse errors..

Is there a max size limitation on /dev/sdC0/fs, I installed it
with a size close to 6Gb?

-ishwar

Re: [9fans] Standalone cpu/auth..

[David Gordon Hogan]

> For got the program output :-|
>
> ---
> PCI interrupt routing table version 1.0 at fdf10

Thanks.  What does the "pci" command report?

Re: [9fans] Standalone cpu/auth..

[Ish Rattan]

Ouput generated by pci command.

-ishwar
----
0.0.0:	06.00.00 1166/0009   0
0.0.1:	06.00.00 1166/0009   0
0.15.0:	06.01.00 1166/0201   0
0.15.1:	01.01.8a 1166/0212   0 0:000001f1 16 1:000003f5 16 2:00000171 16 3:00000375 16 4:00002441 16
0.15.2:	0c.03.10 1166/0220   9 0:fb002000 4096
0.15.3:	06.00.00 1166/0225   0
0.2.0:	03.00.00 1002/4752 255 0:fa000000 16777216 1:00002001 256 2:fb000000 4096
0.3.0:	02.00.00 8086/1229  11 0:fb001000 4096 1:00002401 64 2:fb100000 1048576
---

On Thu, 16 May 2002, David Gordon Hogan wrote:

> > For got the program output :-|
> >
> > ---
> > PCI interrupt routing table version 1.0 at fdf10
>
> Thanks.  What does the "pci" command report?
>

Re: [9fans] Standalone cpu/auth..

[rsc]

the drawterm sessions should be slower,
since they are going through srvold9p rather
than plugging the network connection
into the kernel directly.  it shouldn't be very
much slower, but perhaps a little.  if you're
comparing drawterm-linux->3e vs ->4e,
i don't think you should see a big difference.

if you're comparing drawterm-linux->3e with
drawterm-windows->4e, you will see a big difference.
repeating yesterday's theme, the 4e drawterm sources
from windows now disable nagle's algorithm,
which makes an enormous difference.

replace /sys/src/cmd/unix/drawterm/devip-unix.c:/^so_socket
with this one and recompile.  you should see a big
difference.

int
so_socket(int type)
{
	int fd, one;

	switch(type) {
	default:
		error("bad protocol type");
	case S_TCP:
		type = SOCK_STREAM;
		break;
	case S_UDP:
		type = SOCK_DGRAM;
		break;
	}

	fd = socket(AF_INET, type, 0);
	if(fd < 0)
		error(strerror(errno));

	one = 1;
	if(setsockopt(fd, IPPROTO_TCP, TCP_NODELAY, (char*)&one, sizeof(one)) > 0)
		print("setsockopt: %s", sys_errlist[errno]);

	return fd;
}

russ

Re: [9fans] Standalone cpu/auth..

[David Gordon Hogan]

> I just tried the cpu/auth standalone on a dual PIII with Intel
> server mother board. At boot time, i see the message..
> .....
> pcirouting: South bridge FFFF, FFFF not found
> .....
> pcirouting: Can't find south bridge PCI 255.37.7

The first message is ``impossible''.  I'd love to know
why it's happening.  The second one is just improbable.
What does the "pci" command report?  Also, here is
a program for dumping the contents of the BIOS PCI
routing table, please save it as "pcir.c", compile and
run, then send me the output:

#include <u.h>
#include <libc.h>

typedef struct Router Router;
typedef struct Slot Slot;

struct Slot
{
	uchar	bus;			// Pci bus number
	uchar	dev;			// Pci device number
	uchar	maps[12];		// Avoid structs!  Link and mask.
	uchar	slot;			// Add-in/built-in slot
	uchar	reserved;
};

struct Router
{
	uchar	signature[4];	// Routing table signature
	uchar	version[2];	// Version number
	uchar	size[2];		// Total table size
	uchar	bus;			// Interrupt router bus number
	uchar	devfn;		// Router's devfunc
	uchar	pciirqs[2];		// Exclusive PCI irqs
	uchar	compat[4];	// Compatible PCI interrupt router
	uchar	miniport[4];	// Miniport data
	uchar	reserved[11];
	uchar	checksum;
};

#define	GET2(p)		((p)[0]|((p)[1]<<8))
#define	DEV(devfn)	(devfn>>3)
#define	FN(devfn)		(devfn&7)

void
main(int argc, char *argv[])
{
	Slot *e;
	Router *r;
	int fd, size, i;
	char file[64];
	uchar *buf, *p, *ep, *m;

	USED(argc, argv);

	snprint(file, sizeof file, "#p/%d/mem", getpid());
	fd = open(file, OREAD);
	if(fd < 0)
		sysfatal("open proc mem");
	if(seek(fd, 0x80000000|0xf0000, 0) < 0)
		sysfatal("seek proc mem");

	buf = malloc(0x10000);
	if(read(fd, buf, 0x10000) != 0x10000)
		sysfatal("read proc mem");

	ep = buf+0x10000;
	for(p = buf; p < ep; p += 16)
		if(p[0] == '$' && p[1] == 'P' && p[2] == 'I' && p[3] == 'R')
			break;
	if(p == ep) {
		fprint(2, "pcir: PCI routing table not found\n");
		exits("not found");
	}

	r = (Router *)p;

	print("PCI interrupt routing table version %d.%d at %.5ux\n",
		r->version[1], r->version[0], p-buf+0xf0000);
	print("South Bridge %d.%d.%d, irqs %.4uX compat %.4uX/%.4uX\n",
		r->bus, DEV(r->devfn), FN(r->devfn), GET2(r->pciirqs),
		GET2(&r->compat[0]), GET2(&r->compat[2]));
	print("miniport data: %.2uX %.2uX %.2uX %.2uX\n",
		r->miniport[0], r->miniport[1], r->miniport[2], r->miniport[3]);
	size = GET2(r->size);
	for(e = (Slot *)&r[1]; (uchar *)e < p + size; e++) {
		print("%d.%d.%d %.2uX:\t", e->bus, DEV(e->dev), FN(e->dev), e->slot);
		for(i = 0; i != 4; i++) {
			m = &e->maps[i * 3];
			print("[%d] %.2uX %.4uX ", i, m[0], (m[2] << 8)|m[1]);
		}
		print("\n");
	}

	exits(nil);
}

Re: [9fans] Standalone cpu/auth..

[Ish Rattan]

Here is the output of the program. BTW the motherboard in question is
Intel Server board SAI2 with 32-bit and 64-bit PCI slots with on board
NIC and video.

-ishwar

On Wed, 15 May 2002, David Gordon Hogan wrote:

> > I just tried the cpu/auth standalone on a dual PIII with Intel
> > server mother board. At boot time, i see the message..
> > .....
> > pcirouting: South bridge FFFF, FFFF not found
> > .....
> > pcirouting: Can't find south bridge PCI 255.37.7
>
> The first message is ``impossible''.  I'd love to know
> why it's happening.  The second one is just improbable.
> What does the "pci" command report?  Also, here is
> a program for dumping the contents of the BIOS PCI
> routing table, please save it as "pcir.c", compile and
> run, then send me the output:
>
> #include <u.h>
> #include <libc.h>
>
> typedef struct Router Router;
> typedef struct Slot Slot;
>
> struct Slot
> {
> 	uchar	bus;			// Pci bus number
> 	uchar	dev;			// Pci device number
> 	uchar	maps[12];		// Avoid structs!  Link and mask.
> 	uchar	slot;			// Add-in/built-in slot
> 	uchar	reserved;
> };
>
> struct Router
> {
> 	uchar	signature[4];	// Routing table signature
> 	uchar	version[2];	// Version number
> 	uchar	size[2];		// Total table size
> 	uchar	bus;			// Interrupt router bus number
> 	uchar	devfn;		// Router's devfunc
> 	uchar	pciirqs[2];		// Exclusive PCI irqs
> 	uchar	compat[4];	// Compatible PCI interrupt router
> 	uchar	miniport[4];	// Miniport data
> 	uchar	reserved[11];
> 	uchar	checksum;
> };
>
> #define	GET2(p)		((p)[0]|((p)[1]<<8))
> #define	DEV(devfn)	(devfn>>3)
> #define	FN(devfn)		(devfn&7)
>
> void
> main(int argc, char *argv[])
> {
> 	Slot *e;
> 	Router *r;
> 	int fd, size, i;
> 	char file[64];
> 	uchar *buf, *p, *ep, *m;
>
> 	USED(argc, argv);
>
> 	snprint(file, sizeof file, "#p/%d/mem", getpid());
> 	fd = open(file, OREAD);
> 	if(fd < 0)
> 		sysfatal("open proc mem");
> 	if(seek(fd, 0x80000000|0xf0000, 0) < 0)
> 		sysfatal("seek proc mem");
>
> 	buf = malloc(0x10000);
> 	if(read(fd, buf, 0x10000) != 0x10000)
> 		sysfatal("read proc mem");
>
> 	ep = buf+0x10000;
> 	for(p = buf; p < ep; p += 16)
> 		if(p[0] == '$' && p[1] == 'P' && p[2] == 'I' && p[3] == 'R')
> 			break;
> 	if(p == ep) {
> 		fprint(2, "pcir: PCI routing table not found\n");
> 		exits("not found");
> 	}
>
> 	r = (Router *)p;
>
> 	print("PCI interrupt routing table version %d.%d at %.5ux\n",
> 		r->version[1], r->version[0], p-buf+0xf0000);
> 	print("South Bridge %d.%d.%d, irqs %.4uX compat %.4uX/%.4uX\n",
> 		r->bus, DEV(r->devfn), FN(r->devfn), GET2(r->pciirqs),
> 		GET2(&r->compat[0]), GET2(&r->compat[2]));
> 	print("miniport data: %.2uX %.2uX %.2uX %.2uX\n",
> 		r->miniport[0], r->miniport[1], r->miniport[2], r->miniport[3]);
> 	size = GET2(r->size);
> 	for(e = (Slot *)&r[1]; (uchar *)e < p + size; e++) {
> 		print("%d.%d.%d %.2uX:\t", e->bus, DEV(e->dev), FN(e->dev), e->slot);
> 		for(i = 0; i != 4; i++) {
> 			m = &e->maps[i * 3];
> 			print("[%d] %.2uX %.4uX ", i, m[0], (m[2] << 8)|m[1]);
> 		}
> 		print("\n");
> 	}
>
> 	exits(nil);
> }
>

Re: [9fans] Standalone cpu/auth..

[Ish Rattan]

For got the program output :-|

---
PCI interrupt routing table version 1.0 at fdf10
South Bridge 255.31.7, irqs 0000 compat 0000/0000
miniport data: 00 00 00 00
0.0.0 00:	[0] 00 DEF8 [1] 00 DEF8 [2] 00 DEF8 [3] 00 DEF8
0.6.0 01:	[0] 1B 1EF8 [1] 1C 1EF8 [2] 1D 1EF8 [3] 1E 1EF8
0.7.0 02:	[0] 18 1EF8 [1] 1C 1EF8 [2] 1D 1EF8 [3] 0A 1EF8
0.8.0 05:	[0] 19 1EF8 [1] 1E 1EF8 [2] 1C 1EF8 [3] 1D 1EF8
0.9.0 06:	[0] 1A 1EF8 [1] 1D 1EF8 [2] 1E 1EF8 [3] 1C 1EF8
0.3.0 00:	[0] 13 1EF8 [1] 00 1EF8 [2] 00 1EF8 [3] 00 1EF8
0.2.0 00:	[0] 14 1EF8 [1] 00 1EF8 [2] 00 1EF8 [3] 00 1EF8
1.0.0 00:	[0] 00 DEF8 [1] 00 DEF8 [2] 00 DEF8 [3] 00 DEF8
1.10.0 03:	[0] 15 1EF8 [1] 1C 1EF8 [2] 1D 1EF8 [3] 1E 1EF8
1.11.0 04:	[0] 16 1EF8 [1] 1E 1EF8 [2] 1C 1EF8 [3] 1D 1EF8
0.15.0 00:	[0] 02 DEF8 [1] 02 DEF8 [2] 02 DEF8 [3] 02 DEF8

---
On Wed, 15 May 2002, David Gordon Hogan wrote:

> > I just tried the cpu/auth standalone on a dual PIII with Intel
> > server mother board. At boot time, i see the message..
> > .....
> > pcirouting: South bridge FFFF, FFFF not found
> > .....
> > pcirouting: Can't find south bridge PCI 255.37.7
>
> The first message is ``impossible''.  I'd love to know
> why it's happening.  The second one is just improbable.
> What does the "pci" command report?  Also, here is
> a program for dumping the contents of the BIOS PCI
> routing table, please save it as "pcir.c", compile and
> run, then send me the output:
>
> #include <u.h>
> #include <libc.h>
>
> typedef struct Router Router;
> typedef struct Slot Slot;
>
> struct Slot
> {
> 	uchar	bus;			// Pci bus number
> 	uchar	dev;			// Pci device number
> 	uchar	maps[12];		// Avoid structs!  Link and mask.
> 	uchar	slot;			// Add-in/built-in slot
> 	uchar	reserved;
> };
>
> struct Router
> {
> 	uchar	signature[4];	// Routing table signature
> 	uchar	version[2];	// Version number
> 	uchar	size[2];		// Total table size
> 	uchar	bus;			// Interrupt router bus number
> 	uchar	devfn;		// Router's devfunc
> 	uchar	pciirqs[2];		// Exclusive PCI irqs
> 	uchar	compat[4];	// Compatible PCI interrupt router
> 	uchar	miniport[4];	// Miniport data
> 	uchar	reserved[11];
> 	uchar	checksum;
> };
>
> #define	GET2(p)		((p)[0]|((p)[1]<<8))
> #define	DEV(devfn)	(devfn>>3)
> #define	FN(devfn)		(devfn&7)
>
> void
> main(int argc, char *argv[])
> {
> 	Slot *e;
> 	Router *r;
> 	int fd, size, i;
> 	char file[64];
> 	uchar *buf, *p, *ep, *m;
>
> 	USED(argc, argv);
>
> 	snprint(file, sizeof file, "#p/%d/mem", getpid());
> 	fd = open(file, OREAD);
> 	if(fd < 0)
> 		sysfatal("open proc mem");
> 	if(seek(fd, 0x80000000|0xf0000, 0) < 0)
> 		sysfatal("seek proc mem");
>
> 	buf = malloc(0x10000);
> 	if(read(fd, buf, 0x10000) != 0x10000)
> 		sysfatal("read proc mem");
>
> 	ep = buf+0x10000;
> 	for(p = buf; p < ep; p += 16)
> 		if(p[0] == '$' && p[1] == 'P' && p[2] == 'I' && p[3] == 'R')
> 			break;
> 	if(p == ep) {
> 		fprint(2, "pcir: PCI routing table not found\n");
> 		exits("not found");
> 	}
>
> 	r = (Router *)p;
>
> 	print("PCI interrupt routing table version %d.%d at %.5ux\n",
> 		r->version[1], r->version[0], p-buf+0xf0000);
> 	print("South Bridge %d.%d.%d, irqs %.4uX compat %.4uX/%.4uX\n",
> 		r->bus, DEV(r->devfn), FN(r->devfn), GET2(r->pciirqs),
> 		GET2(&r->compat[0]), GET2(&r->compat[2]));
> 	print("miniport data: %.2uX %.2uX %.2uX %.2uX\n",
> 		r->miniport[0], r->miniport[1], r->miniport[2], r->miniport[3]);
> 	size = GET2(r->size);
> 	for(e = (Slot *)&r[1]; (uchar *)e < p + size; e++) {
> 		print("%d.%d.%d %.2uX:\t", e->bus, DEV(e->dev), FN(e->dev), e->slot);
> 		for(i = 0; i != 4; i++) {
> 			m = &e->maps[i * 3];
> 			print("[%d] %.2uX %.4uX ", i, m[0], (m[2] << 8)|m[1]);
> 		}
> 		print("\n");
> 	}
>
> 	exits(nil);
> }
>

[9fans] Standalone cpu/auth..

[Ish Rattan]

I just tried the cpu/auth standalone on a dual PIII with Intel
server mother board. At boot time, i see the message..
.....
pcrouting: South bridge FFFF, FFFF not found
.....
pcirouting: Can't find south bridge PCI 255.37.7
.....

but does boot correctly.

Also, can Plan 9 handle 100Mb IOMEGA Zip drive (IDE/ATPI)?

Connecting from Linux with drwaterm, the session is slow (CPU
power does not seem to help) as comapred to 3rd version case.

-ishwar