TG3 doesn't work in kernel 2.4.27

TG3 doesn't work in kernel 2.4.27

[Tetsuo Handa]

Hello,

I'm using tg3.o with DHCP and PXE boot environment
and I updated from 2.4.26 to 2.4.27,
but tg3.o became not working with IBM BladeCenter.

I think tg3.o in 2.4.27 is generating something broken arp.
When I run 'arp' in the DHCP server (who doesn't use tg3.o),
the entry with <incomplete> status appears.
The IP address which has the <incomplete> status is
the DHCP client's (who is using tg3.o in 2.4.27).

The workaround I took is to replace tg3.h and tg3.c
in 2.4.27 with the files in 2.4.26, and it seems working fine.

Thanks.

--
Tetsuo Handa

Re: TG3 doesn't work in kernel 2.4.27

[Marcelo Tosatti]

On Sun, Aug 15, 2004 at 01:53:49AM +0900, Tetsuo Handa wrote:
> Hello,
> 
> I'm using tg3.o with DHCP and PXE boot environment
> and I updated from 2.4.26 to 2.4.27,
> but tg3.o became not working with IBM BladeCenter.
> 
> I think tg3.o in 2.4.27 is generating something broken arp.
> When I run 'arp' in the DHCP server (who doesn't use tg3.o),
> the entry with <incomplete> status appears.
> The IP address which has the <incomplete> status is
> the DHCP client's (who is using tg3.o in 2.4.27).
> 
> The workaround I took is to replace tg3.h and tg3.c
> in 2.4.27 with the files in 2.4.26, and it seems working fine.

David Miller is the tg3 maintainer, he will help you.

Re: TG3 doesn't work in kernel 2.4.27

[David S. Miller]

On Sun, 15 Aug 2004 16:57:58 -0300
Marcelo Tosatti <marcelo.tosatti@cyclades.com> wrote:

> On Sun, Aug 15, 2004 at 01:53:49AM +0900, Tetsuo Handa wrote:
> > Hello,
> > 
> > I'm using tg3.o with DHCP and PXE boot environment
> > and I updated from 2.4.26 to 2.4.27,
> > but tg3.o became not working with IBM BladeCenter.
> > 
> > I think tg3.o in 2.4.27 is generating something broken arp.
> > When I run 'arp' in the DHCP server (who doesn't use tg3.o),
> > the entry with <incomplete> status appears.
> > The IP address which has the <incomplete> status is
> > the DHCP client's (who is using tg3.o in 2.4.27).
> > 
> > The workaround I took is to replace tg3.h and tg3.c
> > in 2.4.27 with the files in 2.4.26, and it seems working fine.
> 
> David Miller is the tg3 maintainer, he will help you.

Does manual IP configuration work?

Re: TG3 doesn't work in kernel 2.4.27

[Tetsuo Handa]

Hello, David.

On Mon, 16 Aug 2004 14:38:24 -0700
"David S. Miller" wrote:
> On Mon, 16 Aug 2004 20:51:03 +0900
> Tetsuo Handa <a5497108@anet.ne.jp> wrote:
> 
> >  From 2.4.26 till 2.4.27-rc3 were all OK.
> > This trouble happens with 2.4.27-rc4 and later.
> 
> It's Sun's buggy 5704 Fiber auto-negotiation changes.
> 
> Here is a hacky possible fix, can you try it?
> 
> ===== drivers/net/tg3.c 1.190 vs edited =====

I compiled as a UP kernel but it wasn't the cause.
Also, I patched the above fix on 2.4.27-rc4 and
compiled as a UP kernel, but didn't work.

Now I would like to mail to Marcelo Tosatti, too.

Current Status:
 * The driver is tg3.o
 * The problem began with 2.4.27-rc4.
 * DHCP fails.
 * Static IP assignment shows no error, but network is still unreachable.
   ('ifconfig' and 'route add default gw' by hand)
 * DHCP packets are unreachable to DHCP server.
   (I don't know which options are used for tcpdump.
    I hear that 2.4.26 worked but 2.4.27 didn't. PXE's DHCP works. )
 * 'arp' at DHCP server shows a line with <incomplete> status.
 * Compiling as a UP kernel didn't solve the problem.
   (The box is Xeon 3.2 GHz / 1 CPU with hyper-threading)
 * Compiling as a UP kernel with the above fix didn't solve.
 * The error message "HW autoneg failed" appears.
   (But (I think) this message doesn't exist in tg3.c for 2.4.26,
    so I don't know autoneg were successful or not in 2.4.26,
    if helpful I will insert printk() and try.
    The link ups (if 2.4.26) 1000Mbps full duplex, off for TX and off for RX.)

Regards...
---
Tetsuo Handa

Re: TG3 doesn't work in kernel 2.4.27

[John W. Linville]

Tetsuo Handa wrote:

> I compiled as a UP kernel but it wasn't the cause.
> Also, I patched the above fix on 2.4.27-rc4 and
> compiled as a UP kernel, but didn't work.
> 

Testsuo,

Please send the output of `lspci -vv`.  It might be helpful to know 
exactly which device is involved.

David,

FWIW, I think I'm seeing the same problem with the BCM5751 built-in to 
my HP xw4200...

John

Re: TG3 doesn't work in kernel 2.4.27

[David S. Miller]

On Tue, 17 Aug 2004 10:03:34 -0400
"John W. Linville" <linville@tuxdriver.com> wrote:

> Tetsuo Handa wrote:
> 
> > I compiled as a UP kernel but it wasn't the cause.
> > Also, I patched the above fix on 2.4.27-rc4 and
> > compiled as a UP kernel, but didn't work.
> > 
> 
> Testsuo,
> 
> Please send the output of `lspci -vv`.  It might be helpful to know 
> exactly which device is involved.
> 
> David,
> 
> FWIW, I think I'm seeing the same problem with the BCM5751 built-in to 
> my HP xw4200...

This bug can only affect 5704 chips with fiber interfaces.

Re: TG3 doesn't work in kernel 2.4.27

[David S. Miller]

On Tue, 17 Aug 2004 21:30:06 +0900
Tetsuo Handa <a5497108@anet.ne.jp> wrote:

> I compiled as a UP kernel but it wasn't the cause.
> Also, I patched the above fix on 2.4.27-rc4 and
> compiled as a UP kernel, but didn't work.

Just add this patch, it will fix the problem but it
is not the nicest fix.  I'll work on a better one.

It just disables the new Sun code, and uses the old
fiber handling for 5704.

===== drivers/net/tg3.c 1.193 vs edited =====
--- 1.193/drivers/net/tg3.c	2004-08-16 18:01:30 -07:00
+++ edited/drivers/net/tg3.c	2004-08-16 18:46:58 -07:00
@@ -5415,13 +5398,13 @@
 	 * is enabled.
 	 */
 	tw32_f(MAC_LOW_WMARK_MAX_RX_FRAME, 2);
-
+#if 0
 	if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5704 &&
 	    (tp->tg3_flags2 & TG3_FLG2_PHY_SERDES)) {
 		/* Use hardware link auto-negotiation */
 		tp->tg3_flags2 |= TG3_FLG2_HW_AUTONEG;
 	}
-
+#endif
 	err = tg3_setup_phy(tp, 1);
 	if (err)
 		return err;

Re: TG3 doesn't work in kernel 2.4.27 (Resolved!)

[Tetsuo Handa]

Hello, David.

In message <20040817110248.4c2f7cd8.davem@redhat.com>
   "Re: TG3 doesn't work in kernel 2.4.27"
   ""David S. Miller" <davem@redhat.com>" wrote:

> On Tue, 17 Aug 2004 21:30:06 +0900
> Tetsuo Handa <a5497108@anet.ne.jp> wrote:
> 
> > I compiled as a UP kernel but it wasn't the cause.
> > Also, I patched the above fix on 2.4.27-rc4 and
> > compiled as a UP kernel, but didn't work.
> 
> Just add this patch, it will fix the problem but it
> is not the nicest fix.  I'll work on a better one.
> 
> It just disables the new Sun code, and uses the old
> fiber handling for 5704.
> 
It resolved the problem. It works on SMP kernel, too.
The line position was different for tg3.c in 2.4.27 .

-----START-----
--- tg3.c.org	2004-08-08 08:26:05.000000000 +0900
+++ tg3.c	2004-08-18 09:27:58.000000000 +0900
@@ -5264,7 +5264,7 @@
 	 * is enabled.
 	 */
 	tw32_f(MAC_LOW_WMARK_MAX_RX_FRAME, 2);
-
+#if 0
 	if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5704 &&
 	    tp->phy_id == PHY_ID_SERDES) {
 		/* Enable hardware link auto-negotiation */
@@ -5284,7 +5284,7 @@
 
 		tp->tg3_flags2 |= TG3_FLG2_HW_AUTONEG;
 	}
-
+#endif
 	err = tg3_setup_phy(tp, 1);
 	if (err)
 		return err;
-----END-----

Also, as you have mentioned that this bug can only affect 5704 chips with fiber interfaces,
I found that the box has BCM5704S. The following is the result of 'lspci -vv'.

-----START-----
00:00.0 Host bridge: ServerWorks CNB20-HE Host Bridge (rev 33)
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-

00:00.1 Host bridge: ServerWorks CNB20-HE Host Bridge
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-

00:00.2 Host bridge: ServerWorks CNB20-HE Host Bridge
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-

00:01.0 VGA compatible controller: ATI Technologies Inc Rage XL (rev 27) (prog-if 00 [VGA])
	Subsystem: IBM: Unknown device 0240
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping+ SERR- FastB2B-
	Status: Cap+ 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 64 (2000ns min), cache line size 08
	Interrupt: pin A routed to IRQ 20
	Region 0: Memory at fd000000 (32-bit, non-prefetchable) [size=16M]
	Region 1: I/O ports at 2200 [size=256]
	Region 2: Memory at febff000 (32-bit, non-prefetchable) [size=4K]
	Expansion ROM at <unassigned> [disabled] [size=128K]
	Capabilities: [5c] Power Management version 2
		Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
		Status: D0 PME-Enable- DSel=0 DScale=0 PME-

00:0f.0 Host bridge: ServerWorks CSB6 South Bridge (rev b0)
	Subsystem: ServerWorks: Unknown device 0201
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr+ Stepping- SERR+ FastB2B-
	Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort+ >SERR- <PERR-
	Latency: 64

00:0f.1 IDE interface: ServerWorks CSB6 RAID/IDE Controller (rev b0) (prog-if 8a [Master SecP PriP])
	Subsystem: ServerWorks: Unknown device 0212
	Control: I/O+ Mem- BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr+ Stepping- SERR+ FastB2B-
	Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 64, cache line size 08
	Region 0: I/O ports at <ignored>
	Region 1: I/O ports at <ignored>
	Region 2: I/O ports at <ignored>
	Region 3: I/O ports at <ignored>
	Region 4: I/O ports at 0700 [size=16]

00:0f.2 USB Controller: ServerWorks CSB6 OHCI USB Controller (rev 05) (prog-if 10 [OHCI])
	Subsystem: ServerWorks: Unknown device 0220
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr+ Stepping- SERR+ FastB2B-
	Status: Cap- 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 64 (20000ns max), cache line size 08
	Interrupt: pin A routed to IRQ 7
	Region 0: Memory at febfe000 (32-bit, non-prefetchable) [size=4K]

00:0f.3 ISA bridge: ServerWorks GCLE-2 Host Bridge
	Subsystem: ServerWorks: Unknown device 0230
	Control: I/O- Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr+ Stepping- SERR+ FastB2B-
	Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 0

00:10.0 Host bridge: ServerWorks: Unknown device 0110 (rev 12)
	Control: I/O- Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr+ Stepping- SERR+ FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort+ >SERR- <PERR-
	Capabilities: [60] PCI-X non-bridge device.
		Command: DPERE- ERO- RBC=0 OST=4
		Status: Bus=0 Dev=0 Func=0 64bit- 133MHz- SCD- USC-, DC=simple, DMMRBC=0, DMOST=0, DMCRS=0, RSCEM-
00:10.2 Host bridge: ServerWorks: Unknown device 0110 (rev 12)
	Control: I/O- Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr+ Stepping- SERR+ FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort+ >SERR- <PERR-
	Capabilities: [60] PCI-X non-bridge device.
		Command: DPERE- ERO- RBC=0 OST=4
		Status: Bus=0 Dev=0 Func=0 64bit- 133MHz- SCD- USC-, DC=simple, DMMRBC=0, DMOST=0, DMCRS=0, RSCEM-
01:00.0 Ethernet controller: Broadcom Corporation NetXtreme BCM5704S Gigabit Ethernet (rev 02)
	Subsystem: IBM: Unknown device 029c
	Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr+ Stepping- SERR+ FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 64 (16000ns min), cache line size 08
	Interrupt: pin A routed to IRQ 16
	Region 0: Memory at fbff0000 (64-bit, non-prefetchable) [size=64K]
	Region 2: Memory at fbfe0000 (64-bit, non-prefetchable) [size=64K]
	Capabilities: [40] PCI-X non-bridge device.
		Command: DPERE- ERO- RBC=2 OST=0
		Status: Bus=0 Dev=0 Func=0 64bit- 133MHz- SCD- USC-, DC=simple, DMMRBC=0, DMOST=0, DMCRS=0, RSCEM-	Capabilities: [48] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot+,D3cold+)
		Status: D0 PME-Enable+ DSel=0 DScale=1 PME-
	Capabilities: [50] Vital Product Data
	Capabilities: [58] Message Signalled Interrupts: 64bit+ Queue=0/3 Enable-
		Address: 0000000100000000  Data: c065

01:00.1 Ethernet controller: Broadcom Corporation NetXtreme BCM5704S Gigabit Ethernet (rev 02)
	Subsystem: IBM: Unknown device 029c
	Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr+ Stepping- SERR+ FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 64 (16000ns min), cache line size 08
	Interrupt: pin B routed to IRQ 17
	Region 0: Memory at fbfd0000 (64-bit, non-prefetchable) [size=64K]
	Region 2: Memory at fbfc0000 (64-bit, non-prefetchable) [size=64K]
	Capabilities: [40] PCI-X non-bridge device.
		Command: DPERE- ERO- RBC=2 OST=0
		Status: Bus=0 Dev=0 Func=0 64bit- 133MHz- SCD- USC-, DC=simple, DMMRBC=0, DMOST=0, DMCRS=0, RSCEM-	Capabilities: [48] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot+,D3cold+)
		Status: D0 PME-Enable+ DSel=0 DScale=1 PME-
	Capabilities: [50] Vital Product Data
	Capabilities: [58] Message Signalled Interrupts: 64bit+ Queue=0/3 Enable-
		Address: 0000000100000000  Data: bef5

02:02.0 Fibre Channel: QLogic Corp. QLA2312 Fibre Channel Adapter (rev 02)
	Subsystem: IBM: Unknown device 027d
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr+ Stepping- SERR+ FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 64 (16000ns min), cache line size 08
	Interrupt: pin A routed to IRQ 18
	Region 0: I/O ports at 2400 [size=256]
	Region 1: Memory at f9ffe000 (64-bit, non-prefetchable) [size=4K]
	Expansion ROM at <unassigned> [disabled] [size=128K]
	Capabilities: [44] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
		Status: D0 PME-Enable- DSel=0 DScale=0 PME-
	Capabilities: [4c] PCI-X non-bridge device.
		Command: DPERE- ERO+ RBC=0 OST=2
		Status: Bus=0 Dev=0 Func=0 64bit- 133MHz- SCD- USC-, DC=simple, DMMRBC=0, DMOST=0, DMCRS=0, RSCEM-	Capabilities: [54] Message Signalled Interrupts: 64bit+ Queue=0/3 Enable-
		Address: 0000000000000000  Data: 0000
	Capabilities: [64] #06 [0080]

02:02.1 Fibre Channel: QLogic Corp. QLA2312 Fibre Channel Adapter (rev 02)
	Subsystem: IBM: Unknown device 027d
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr+ Stepping- SERR+ FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 64 (16000ns min), cache line size 08
	Interrupt: pin B routed to IRQ 19
	Region 0: I/O ports at 2600 [size=256]
	Region 1: Memory at f9ffc000 (64-bit, non-prefetchable) [size=4K]
	Expansion ROM at <unassigned> [disabled] [size=128K]
	Capabilities: [44] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
		Status: D0 PME-Enable- DSel=0 DScale=0 PME-
	Capabilities: [4c] PCI-X non-bridge device.
		Command: DPERE- ERO+ RBC=0 OST=2
		Status: Bus=0 Dev=0 Func=0 64bit- 133MHz- SCD- USC-, DC=simple, DMMRBC=0, DMOST=0, DMCRS=0, RSCEM-	Capabilities: [54] Message Signalled Interrupts: 64bit+ Queue=0/3 Enable-
		Address: 0000000000000000  Data: 0000
	Capabilities: [64] #06 [0080]

-----END-----

Thank you very much.
-----
Tetsuo Handa