Skip to contentSkip to conversation

kernel 2.6.19 and RealTek RTL8139 interrupt

DISCUSSION Oleg Mikheev 2 participants 2 messages Dec 5, 2006 – Dec 6, 2006

lore
Share & tools
lore
#0 lore raw

Hi guys,

I'm using FC5 with vanilla kernels.
Everything worked great until 2.6.19 was released. Not each time I'm trying to up my eth0 kernel produces this:

eth0: link up, 100Mbps, full-duplex, lpa 0x45E1 irq 18: nobody cared (try booting with the "irqpoll" option)

 [<c01391aa>] __report_bad_irq+0x36/0x7d
 [<c0139379>] note_interrupt+0x188/0x1c3
 [<c0138a19>] handle_IRQ_event+0x1a/0x3f
 [<c01396ca>] handle_fasteoi_irq+0x61/0x73
 [<c0104e34>] do_IRQ+0x6b/0x83
 [<c0120c76>] sys_rt_sigprocmask+0x80/0x93
 [<c01034a2>] common_interrupt+0x1a/0x20
 =======================

handlers:
[<c0257416>] (rtl8139_interrupt+0x0/0x3aa)
Disabling IRQ #18

I'm posting it here b/c I don't think it's a normal behavior. Here is the diff of dmesg output between 2.6.18.3 and 2.6.19:

140,142c141,143
< 8139too Fast Ethernet driver 0.9.27
< ACPI: PCI Interrupt 0000:01:06.0[A] -> GSI 20 (level, low) -> IRQ 20 < eth0: RealTek RTL8139 at 0xf8806c00, 00:40:45:28:2c:ae, IRQ 20
---

Quoted 3 lines

8139too Fast Ethernet driver 0.9.28
ACPI: PCI Interrupt 0000:01:06.0[A] -> GSI 20 (level, low) -> IRQ 18
eth0: RealTek RTL8139 at 0xf8806c00, 00:40:45:28:2c:ae, IRQ 18

147c148
< ACPI: PCI Interrupt 0000:00:1f.1[A] -> GSI 18 (level, low) -> IRQ 18
---

Quoted 1 line

ACPI: PCI Interrupt 0000:00:1f.1[A] -> GSI 18 (level, low) -> IRQ 19

197,198c198,200
< TCP bic registered
< Initializing IPsec netlink socket
---

Quoted 3 lines

input: AT Translated Set 2 keyboard as /class/input/input0 TCP cubic registered
Initializing XFRM netlink socket

294,306c305,310
< Non-volatile memory driver v1.2
< usb 4-2: new full speed USB device using uhci_hcd and address 2 < PM: Adding info for usb:4-2
< PM: Adding info for No Bus:usbdev4.2_ep00
< usb 4-2: configuration #1 chosen from 1 choice
< PM: Adding info for usb:4-2:1.0
< PM: Adding info for No Bus:usbdev4.2_ep81
< PM: Adding info for No Bus:usbdev4.2_ep02
< PM: Adding info for No Bus:usbdev4.2_ep82
< PM: Adding info for usb:4-2:1.1
< PM: Adding info for No Bus:usbdev4.2_ep03
< PM: Adding info for No Bus:usbdev4.2_ep83
< Bluetooth: Core ver 2.10
---

Quoted 6 lines

PM: Adding info for No Bus:usbdev3.2_ep02
PM: Adding info for No Bus:usbdev3.2_ep82
PM: Adding info for usb:3-2:1.1
PM: Adding info for No Bus:usbdev3.2_ep03
PM: Adding info for No Bus:usbdev3.2_ep83
Bluetooth: Core ver 2.11

312,315c316,319
< PM: Removing info for No Bus:usbdev4.2_ep03
< PM: Removing info for No Bus:usbdev4.2_ep83
< PM: Adding info for No Bus:usbdev4.2_ep03
< PM: Adding info for No Bus:usbdev4.2_ep83
---

Quoted 4 lines

PM: Removing info for No Bus:usbdev3.2_ep03
PM: Removing info for No Bus:usbdev3.2_ep83
PM: Adding info for No Bus:usbdev3.2_ep03
PM: Adding info for No Bus:usbdev3.2_ep83

317,320c321
< usbcore: registered new driver hci_usb
< PM: Adding info for ieee1394:00012f650000d9b3
< ieee1394: Host added: ID:BUS[0-00:1023] GUID[00012f650000d9b3] < PM: Adding info for ieee1394:00012f650000d9b3-0
---

Quoted 1 line

usbcore: registered new interface driver hci_usb

322c323
< device-mapper: ioctl: 4.7.0-ioctl (2006-06-24) initialised: [email protected]
---

Quoted 1 line

device-mapper: ioctl: 4.10.0-ioctl (2006-09-14) initialised: [email protected]

#1 lore raw
Quoted 31 lines · On Tue, 05 Dec 2006 21:36:37 +0300 Oleg Mikheev <[email protected]> wrote:

Hi guys,

I'm using FC5 with vanilla kernels.
Everything worked great until 2.6.19 was released. Not each time I'm trying to up my eth0 kernel produces this:

eth0: link up, 100Mbps, full-duplex, lpa 0x45E1 irq 18: nobody cared (try booting with the "irqpoll" option)

  [<c01391aa>] __report_bad_irq+0x36/0x7d
  [<c0139379>] note_interrupt+0x188/0x1c3
  [<c0138a19>] handle_IRQ_event+0x1a/0x3f
  [<c01396ca>] handle_fasteoi_irq+0x61/0x73
  [<c0104e34>] do_IRQ+0x6b/0x83
  [<c0120c76>] sys_rt_sigprocmask+0x80/0x93
  [<c01034a2>] common_interrupt+0x1a/0x20
  =======================

handlers:
[<c0257416>] (rtl8139_interrupt+0x0/0x3aa)
Disabling IRQ #18

I'm posting it here b/c I don't think it's a normal behavior. Here is the diff of dmesg output between 2.6.18.3 and 2.6.19:

140,142c141,143
< 8139too Fast Ethernet driver 0.9.27
< ACPI: PCI Interrupt 0000:01:06.0[A] -> GSI 20 (level, low) -> IRQ 20 < eth0: RealTek RTL8139 at 0xf8806c00, 00:40:45:28:2c:ae, IRQ 20
---

Quoted 7 lines
Quoted 3 lines

8139too Fast Ethernet driver 0.9.28
ACPI: PCI Interrupt 0000:01:06.0[A] -> GSI 20 (level, low) -> IRQ 18
eth0: RealTek RTL8139 at 0xf8806c00, 00:40:45:28:2c:ae, IRQ 18

147c148
< ACPI: PCI Interrupt 0000:00:1f.1[A] -> GSI 18 (level, low) -> IRQ 18
---

Quoted 1 line

ACPI: PCI Interrupt 0000:00:1f.1[A] -> GSI 18 (level, low) -> IRQ 19

This looks like http://bugzilla.kernel.org/show_bug.cgi?id=7601

Please see http://bugzilla.kernel.org/show_bug.cgi?id=7601#c10 for a possible fix.

That fix is in mainline now, so testing Linus's current tree would suit too, thanks.