Bug 16112

Summary: sis190 driver 'hangs' randomly on data spikes; NETDEV WATCHER dump in dmesg
Product: Drivers Reporter: Pawel Jasnos (pawel.jasnos)
Component: NetworkAssignee: Francois Romieu (romieu)
Status: NEW ---    
Severity: normal CC: akpm, alan, arutemus, brandon, szg00000
Priority: P1    
Hardware: i386   
OS: Linux   
Kernel Version: 3.5.4 Subsystem:
Regression: No Bisected commit-id:
Attachments: Dmesg after a 'hang'
dmesg info(part with info of card loading and crashing)
lspci info of my sis191 card

Description Pawel Jasnos 2010-06-02 22:08:02 UTC
Created attachment 26629 [details]
Dmesg after a 'hang'

the driver seems to 'freeze' by not transmitting any further
packets - pings or any connection attempts fail, wireshark is not able to
capture any passing packets. Bringing the interface down and then up doesn't
work, but reloading the driver (modprobe -r sis190; modprobe sis190 ) resolves
the problem. 


Steps to Reproduce:
1.Use a network card serviced by the driver
2.Download a lot of data in short period of time (few Gigabytes)
Actual Results:  
Driver stops servicing any further packets.

Expected Results:  
Driver works as normal.

Card details:
lspci -vvv (edited):
00:04.0 Ethernet controller: Silicon Integrated Systems [SiS] 191 Gigabit
Ethernet Adapter (rev 02)
        Subsystem: Elitegroup Computer Systems Device 5a00                      
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B- DisINTx-
        Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 0                                                              
        Interrupt: pin A routed to IRQ 19                                       
        Region 0: Memory at d4307000 (32-bit, non-prefetchable) [size=128]      
        Region 1: I/O ports at 1000 [size=128]                                  
        Capabilities: [40] Power Management version 2                           
                Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=0mA
PME(D0+,D1+,D2+,D3hot+,D3cold+)                    
                Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-          
        Kernel driver in use: sis190

dmesg |grep sis190: (fresh boot)
[    8.183248] sis190 Gigabit Ethernet driver 1.3 loaded.
[    8.183395] sis190 0000:00:04.0: PCI INT A -> GSI 19 (level, low) -> IRQ 19
[    8.183506] sis190 0000:00:04.0: setting latency timer to 64

Dmesg after a 'hang' is attached - I removed some MAC addresses/serial numbers for privacy.
Comment 1 Arteom 2012-10-01 10:44:28 UTC
I have the same problem on the same card sis191.
Easy way to reprosuce this problem to do something like this:
"iperf -c $IP -i 5 -t 100000 -d  -P 2"
5-10 sec and network is hangs.
After i just need to make reup interface(ifconfig down && ifconfig up) even without reloading driver.

Any advices to solve this problem?
Comment 2 Arteom 2012-10-01 10:56:05 UTC
Created attachment 81741 [details]
dmesg info(part with info of card loading and crashing)

adding dmesg info and also forgot to say, my curret kernel is 
Linux art 3.5.4-3-ARCH #1 SMP PREEMPT x86_64 GNU/Linux
Comment 3 Arteom 2012-10-01 10:57:51 UTC
Created attachment 81751 [details]
lspci info of my sis191 card

also adding lspci info