Patchwork tulip : fix 21142 with 10Mbps without negotiation

login
register
mail settings
Submitter Philippe De Muyter
Date Jan. 27, 2009, 4:57 p.m.
Message ID <20090127165727.GA25748@frolo.macqel>
Download mbox | patch
Permalink /patch/20475/
State Accepted
Delegated to: David Miller
Headers show

Comments

Philippe De Muyter - Jan. 27, 2009, 4:57 p.m.
Hello list

with current kernels, tulip 21142 ethernet controllers fail to connect
to a 10Mbps only (i.e. without negotiation-partner) network.  It used
to work in 2.4 kernels.  Fix that.  Tested on a 21142 Rev 0x11.

Signed-off-by: Philippe De Muyter <phdm@macqel.be>
David Miller - Jan. 30, 2009, 9:49 p.m.
From: Philippe De Muyter <phdm@macqel.be>
Date: Tue, 27 Jan 2009 17:57:27 +0100

> with current kernels, tulip 21142 ethernet controllers fail to connect
> to a 10Mbps only (i.e. without negotiation-partner) network.  It used
> to work in 2.4 kernels.  Fix that.  Tested on a 21142 Rev 0x11.
> 
> Signed-off-by: Philippe De Muyter <phdm@macqel.be>

Applied, thanks Philippe.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Philippe De Muyter - Nov. 8, 2009, 10:33 p.m.
Hello,

I have just installed 2.6.31 (from opensuse 11.2) one a tulip-equipped
computer and I get the following error message from the kernel :

[ 2495.526390] ------------[ cut here ]------------
[ 2495.526390] kernel BUG at /usr/src/packages/BUILD/kernel-default-2.6.31.5/linux-2.6.31/include/linux/netdevice.h:439!
[ 2495.526390] invalid opcode: 0000 [#1] SMP 
[ 2495.526390] last sysfs file: /sys/devices/pci0000:00/0000:00:07.1/host0/target0:0:0/0:0:0:0/block/sda/uevent
[ 2495.526390] Modules linked in: ohci_hcd raw1394 ohci1394 ieee1394 acpi_cpufreq speedstep_lib processor thermal_sys hwmon edd ipv6 af_packet fuse loop dm_mod rtc_cmos rtc_core rtc_lib apm pcspkr sg tulip uhci_hcd ehci_hcd reiserfs ata_piix ahci libata
[ 2495.526390] 
[ 2495.526390] Pid: 339, comm: kapmd Not tainted (2.6.31.5-0.1-default #1) 
[ 2495.526390] EIP: 0060:[<c3d6045d>] EFLAGS: 00010246 CPU: 0
[ 2495.526390] EIP is at tulip_up+0xa2d/0xa80 [tulip]
[ 2495.526390] EAX: 00000000 EBX: c1cd7000 ECX: c022af50 EDX: 0001ec00
[ 2495.526390] ESI: c1cd7340 EDI: 00000000 EBP: c20bde44 ESP: c20bddfc
[ 2495.526390]  DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
[ 2495.526390] Process kapmd (pid: 339, ti=c20bc000 task=c25a32c0 task.ti=c20bc000)
[ 2495.526390] Stack:
[ 2495.526390]  0000000b c20bde44 c02acf76 2caa9a94 c2481800 c20bde38 c043ca4a c20bde27
[ 2495.526390] <0> 069ee44b c20ec120 c087c020 0001ec00 c1cd7000 c3d5c720 2caa9a94 c1cd7000
[ 2495.526390] <0> c2480000 00000000 c20bde68 c3d60555 00000080 c1cd7000 c1cd7000 2caa9a94
[ 2495.526390] Call Trace:
[ 2495.526390]  [<c3d60555>] tulip_resume+0xa5/0xd0 [tulip]
[ 2495.526390]  [<c043dd65>] pci_legacy_resume+0x35/0x60
[ 2495.526390]  [<c043df2f>] pci_pm_resume+0x7f/0xb0
[ 2495.526390]  [<c04d82f2>] pm_op+0xd2/0x180
[ 2495.526390]  [<c04d91ee>] device_resume+0x5e/0x1a0
[ 2495.526390]  [<c04d93dd>] dpm_resume+0xad/0x140
[ 2495.526390]  [<c04d948b>] dpm_resume_end+0x1b/0x40
[ 2495.526390]  [<c3db1978>] check_events+0x148/0x240 [apm]
[ 2495.526390]  [<c3db23a2>] apm_mainloop+0x82/0x130 [apm]
[ 2495.526390]  [<c3db28fe>] apm+0x10e/0x3d0 [apm]
[ 2495.526390]  [<c026bef4>] kthread+0x84/0x90
[ 2495.526390]  [<c0204db7>] kernel_thread_helper+0x7/0x10
[ 2495.526390] Code: 45 e4 e8 37 ce 6c fc 8b 4d e8 89 5c 24 10 89 7c 24 0c 89 4c 24 04 89 44 24 08 c7 04 24 4c 4e d6 c3 e8 86 f7 89 fc e9 f4 f8 ff ff <0f> 0b eb fe 0f be 96 16 09 00 00 b9 01 00 00 00 8b 45 e8 e8 fb 
[ 2495.526390] EIP: [<c3d6045d>] tulip_up+0xa2d/0xa80 [tulip] SS:ESP 0068:c20bddfc
[ 2495.534162] ---[ end trace 609ed25c95a75fa1 ]---

This comes from a BUG_ON in napi_enable in netdevice.h.

napi_enable itself is called by tulip_up as such :

#ifdef CONFIG_TULIP_NAPI
	napi_enable(&tp->napi);
#endif

At first reading, a matching napi_disable is called in tulip_down.

Does someone know what could be wrong and have a fix or should I look myself ?

Thanks in advance

Philippe
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
David Miller - Nov. 21, 2009, 7:10 p.m.
From: Philippe De Muyter <phdm@macqel.be>
Date: Sun, 8 Nov 2009 23:33:05 +0100

> I have just installed 2.6.31 (from opensuse 11.2) one a tulip-equipped
> computer and I get the following error message from the kernel :

I took some looks at this bug.

I simply can't figure out how this is possible.  As both suspend
and resume make sure to make the necessary napi_disable() and
napi_enable() calls, and therefore they should match up and not
trigger this BUG().

I'll try to study the code some more when I get a chance.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch

diff -r 4e8869c1d518 drivers/net/tulip/21142.c
--- a/drivers/net/tulip/21142.c	Sat Jan 17 23:00:27 2009 +0000
+++ b/drivers/net/tulip/21142.c	Tue Jan 27 16:44:04 2009 +0100
@@ -9,6 +9,11 @@ 
 
 	Please refer to Documentation/DocBook/tulip-user.{pdf,ps,html}
 	for more information on this driver.
+
+	DC21143 manual "21143 PCI/CardBus 10/100Mb/s Ethernet LAN Controller
+	Hardware Reference Manual" is currently available at :
+	http://developer.intel.com/design/network/manuals/278074.htm
+	
 	Please submit bugs to http://bugzilla.kernel.org/ .
 */
 
@@ -32,7 +37,11 @@  void t21142_media_task(struct work_struc
 	int csr12 = ioread32(ioaddr + CSR12);
 	int next_tick = 60*HZ;
 	int new_csr6 = 0;
-
+	int csr14 = ioread32(ioaddr + CSR14);
+
+	/* CSR12[LS10,LS100] are not reliable during autonegotiation */
+	if ((csr14 & 0x80) && (csr12 & 0x7000) != 0x5000)
+		csr12 |= 6;
 	if (tulip_debug > 2)
 		printk(KERN_INFO"%s: 21143 negotiation status %8.8x, %s.\n",
 			   dev->name, csr12, medianame[dev->if_port]);
@@ -76,7 +85,7 @@  void t21142_media_task(struct work_struc
 			new_csr6 = 0x83860000;
 			dev->if_port = 3;
 			iowrite32(0, ioaddr + CSR13);
-			iowrite32(0x0003FF7F, ioaddr + CSR14);
+			iowrite32(0x0003FFFF, ioaddr + CSR14);
 			iowrite16(8, ioaddr + CSR15);
 			iowrite32(1, ioaddr + CSR13);
 		}
@@ -132,10 +141,14 @@  void t21142_lnk_change(struct net_device
 	struct tulip_private *tp = netdev_priv(dev);
 	void __iomem *ioaddr = tp->base_addr;
 	int csr12 = ioread32(ioaddr + CSR12);
-
+	int csr14 = ioread32(ioaddr + CSR14);
+
+	/* CSR12[LS10,LS100] are not reliable during autonegotiation */
+	if ((csr14 & 0x80) && (csr12 & 0x7000) != 0x5000)
+		csr12 |= 6;
 	if (tulip_debug > 1)
 		printk(KERN_INFO"%s: 21143 link status interrupt %8.8x, CSR5 %x, "
-			   "%8.8x.\n", dev->name, csr12, csr5, ioread32(ioaddr + CSR14));
+			   "%8.8x.\n", dev->name, csr12, csr5, csr14);
 
 	/* If NWay finished and we have a negotiated partner capability. */
 	if (tp->nway  &&  !tp->nwayset  &&  (csr12 & 0x7000) == 0x5000) {
@@ -143,7 +156,9 @@  void t21142_lnk_change(struct net_device
 		int negotiated = tp->sym_advertise & (csr12 >> 16);
 		tp->lpar = csr12 >> 16;
 		tp->nwayset = 1;
-		if (negotiated & 0x0100)		dev->if_port = 5;
+		/* If partner cannot negotiate, it is 10Mbps Half Duplex */
+		if (!(csr12 & 0x8000))		dev->if_port = 0;
+		else if (negotiated & 0x0100)	dev->if_port = 5;
 		else if (negotiated & 0x0080)	dev->if_port = 3;
 		else if (negotiated & 0x0040)	dev->if_port = 4;
 		else if (negotiated & 0x0020)	dev->if_port = 0;
@@ -214,7 +229,7 @@  void t21142_lnk_change(struct net_device
 			tp->timer.expires = RUN_AT(3*HZ);
 			add_timer(&tp->timer);
 		} else if (dev->if_port == 5)
-			iowrite32(ioread32(ioaddr + CSR14) & ~0x080, ioaddr + CSR14);
+			iowrite32(csr14 & ~0x080, ioaddr + CSR14);
 	} else if (dev->if_port == 0  ||  dev->if_port == 4) {
 		if ((csr12 & 4) == 0)
 			printk(KERN_INFO"%s: 21143 10baseT link beat good.\n",