[#4120] Ping flood test fails for USB gadget ethernet of pio mode
Submitted By: Vivi Li
Open Date
2008-05-22 03:54:37 Close Date
2009-06-18 00:42:51
Priority:
Medium High Assignee:
Cliff Cai
Status:
Closed Fixed In Release:
N/A
Found In Release:
2008R1-RC8 Release:
Category:
N/A Board:
N/A
Processor:
ALL Silicon Revision:
Is this bug repeatable?:
Yes Resolution:
Fixed
Uboot version or rev.:
Toolchain version or rev.:
08r1-rc8
App binary format:
N/A
Summary: Ping flood test fails for USB gadget ethernet of pio mode
Details:
Ping sometimes fail for gadget ethernet on BF527-EZKIT/BF548-EZKIT. A full log is attached.
Bellow is part of the log:
--
root@adam-desktop2:~# ping -f -l50 -c 10000 192.168.1.66
PING 192.168.1.66 (192.168.1.66) 56(84) bytes of data.
...............................................
--- 192.168.1.66 ping statistics ---
10000 packets transmitted, 9953 received, 0% packet loss, time 1740ms
rtt min/avg/max/mdev = 1.079/8.719/9.493/0.473 ms, pipe 50, ipg/ewma 0.174/7.165 ms
root@adam-desktop2:~#
root@adam-desktop2:~#
root@adam-desktop2:~# ping -f -l50 -c 1000 192.168.1.66
PING 192.168.1.66 (192.168.1.66) 56(84) bytes of data.
..............EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE
--- 192.168.1.66 ping statistics ---
1000 packets transmitted, 876 received, 12% packet loss, time 13029ms
rtt min/avg/max/mdev = 0.570/6.985/9.106/1.557 ms, pipe 50, ipg/ewma 13.042/6.699 ms
root@adam-desktop2:~#
--
Another try, bellow is the log:
--
root@adam-desktop2:~# ping -f -l50 -c 1000 192.168.1.66
PING 192.168.1.66 (192.168.1.66) 56(84) bytes of data.
...............................................
--- 192.168.1.66 ping statistics ---
1000 packets transmitted, 953 received, 4% packet loss, time 153ms
rtt min/avg/max/mdev = 0.240/7.266/9.236/1.636 ms, pipe 50, ipg/ewma 0.153/7.024 ms
root@adam-desktop2:~#
root@adam-desktop2:~#
root@adam-desktop2:~# ping -f -l50 -c 1000 192.168.1.66
PING 192.168.1.66 (192.168.1.66) 56(84) bytes of data.
--- 192.168.1.66 ping statistics ---
1000 packets transmitted, 1000 received, 0% packet loss, time 156ms
rtt min/avg/max/mdev = 0.524/7.770/9.327/1.296 ms, pipe 50, ipg/ewma 0.156/7.161 ms
root@adam-desktop2:~#
root@adam-desktop2:~#
root@adam-desktop2:~# ping -f -l50 -c 1000 192.168.1.66
PING 192.168.1.66 (192.168.1.66) 56(84) bytes of data.
--- 192.168.1.66 ping statistics ---
1000 packets transmitted, 1000 received, 0% packet loss, time 149ms
rtt min/avg/max/mdev = 0.559/7.437/8.983/1.229 ms, pipe 50, ipg/ewma 0.149/6.803 ms
root@adam-desktop2:~#
root@adam-desktop2:~#
root@adam-desktop2:~# ping -f -l50 -c 1000 192.168.1.66
PING 192.168.1.66 (192.168.1.66) 56(84) bytes of data.
--- 192.168.1.66 ping statistics ---
1000 packets transmitted, 1000 received, 0% packet loss, time 149ms
rtt min/avg/max/mdev = 0.581/7.524/8.978/1.139 ms, pipe 50, ipg/ewma 0.149/6.887 ms
root@adam-desktop2:~#
root@adam-desktop2:~#
root@adam-desktop2:~# ping -f -l50 -c 1000 192.168.1.66
PING 192.168.1.66 (192.168.1.66) 56(84) bytes of data.
--- 192.168.1.66 ping statistics ---
1000 packets transmitted, 1000 received, 0% packet loss, time 148ms
rtt min/avg/max/mdev = 0.489/7.343/9.359/1.242 ms, pipe 50, ipg/ewma 0.149/6.378 ms
root@adam-desktop2:~#
root@adam-desktop2:~#
root@adam-desktop2:~# ping -f -l50 -c 1000 192.168.1.66
PING 192.168.1.66 (192.168.1.66) 56(84) bytes of data.
--- 192.168.1.66 ping statistics ---
1000 packets transmitted, 1000 received, 0% packet loss, time 149ms
rtt min/avg/max/mdev = 0.409/7.426/8.984/1.224 ms, pipe 50, ipg/ewma 0.149/6.892 ms
root@adam-desktop2:~#
root@adam-desktop2:~#
root@adam-desktop2:~# ping -f -l50 -c 1000 192.168.1.66
PING 192.168.1.66 (192.168.1.66) 56(84) bytes of data.
--- 192.168.1.66 ping statistics ---
1000 packets transmitted, 1000 received, 0% packet loss, time 148ms
rtt min/avg/max/mdev = 1.120/7.366/8.859/1.252 ms, pipe 50, ipg/ewma 0.148/6.829 ms
root@adam-desktop2:~#
root@adam-desktop2:~#
root@adam-desktop2:~# ping -f -l50 -c 1000 192.168.1.66
PING 192.168.1.66 (192.168.1.66) 56(84) bytes of data.
..............EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE
--- 192.168.1.66 ping statistics ---
1000 packets transmitted, 304 received, 69% packet loss, time 19877ms
rtt min/avg/max/mdev = 0.838/6.491/7.610/1.352 ms, pipe 50, ipg/ewma 19.896/5.677 ms
root@adam-desktop2:~#
--
Follow-ups
--- Bryan Wu 2008-05-29 00:17:44
fixed in the branch by enabling the DMA for read_fifo operation.
Please test again.
-Bryan
--- Bryan Wu 2008-05-29 05:36:43
Also need a patch I added before. We have to read out all the fifo data when
PKT_READY flag is set in the RXCSR.
I will sent it out for upstream review.
-Bryan
--- Vivi Li 2008-06-27 00:34:34
I can still see this bug in 527.
--
root:/>
PHY: 0:03 - Link is Up - 100/Full^M
modprobe g_ether host_addr=00:dc:c8:f7:75:05 dev_addr=00:dd:dc:eb:6d:f1^M
usb0: Ethernet Gadget, version: May Day 2005^M
usb0: using musb_hdrc, OUT ep6out IN ep5in STATUS ep1in^M
usb0: MAC 00:dd:dc:eb:6d:f1^M
usb0: HOST MAC 00:dc:c8:f7:75:05^M
usb0: RNDIS ready^M
root:/> usb0: high speed config #1: 100 mA, Ethernet Gadget, using CDC
Ethernet^M
lsmod^M
Module Size Used by^M
g_ether 17708 0 ^M
root:/>
Case 1 ...PASS
Case 1 ...PASS
ifconfig usb0 192.168.1.66 up^M
root:/>
root:/> ifconfig usb0
usb0 Link encap:Ethernet HWaddr 00:DD:DC:EB:6D:F1 ^M
inet addr:192.168.1.66 Bcast:192.168.1.255 Mask:255.255.255.0^M
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1^M
RX packets:0 errors:0 dropped:0 overruns:0 frame:0^M
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0^M
collisions:0 txqueuelen:1000
root:/>
Case 2 ...PASS
Case 2 ...PASS
su -
^[]0;test@adam-desktop2:
~/work/cruise/checkouts/uclinux-dist/testsuites/usbdev^Gtest@adam-desktop2:~/work/cruise/checkouts/uclinux-dist/testsuites/usbdev>
su -
Password:
^[]0;root@adam-desktop2: ~^Groot@adam-desktop2:~#
Case 3 ...PASS
Case 3 ...PASS
ifconfig eth2 192.168.1.56^M
^[]0;root@adam-desktop2: ~^Groot@adam-desktop2:~#
Case 4 ...PASS
Case 4 ...PASS
ping -f -l50 -c 1000 192.168.1.66^M
PING 192.168.1.66 (192.168.1.66) 56(84) bytes of data.
...................................................................
...................................................................
...................................................................
...................................................................
........EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE............
...................................................................
...................................................................
.....................
--- 192.168.1.66 ping statistics ---
1000 packets transmitted, 662 received, 33% packet loss, time 14761ms
rtt min/avg/max/mdev = 0.350/4.223/6.189/1.348 ms, pipe 50, ipg/ewma
14.776/4.645 ms
^[]0;root@adam-desktop2: ~^Groot@adam-desktop2:~#
--
--- Vivi Li 2008-07-16 06:53:21
This problem also exists on BF548-EZKIT.
--- zc liang 2008-10-20 00:55:29
bf527-v0.2 has come out.Does it fix the bug?
--- Vivi Li 2009-03-13 02:32:03
It still fails for bf548 0.2 for both DMA/PIO mode.
bf527 0.2 also fails.
--- Cliff Cai 2009-05-25 04:57:26
This problem occurs, only if DMA complete interrupt polling mode is enabled.
After start a DMA transfer,sometimes,DMA complete interrupt is never
asserted,then the polling becomes a dead loop.
Cliff
--- Cliff Cai 2009-06-11 05:43:03
Fixed by disabling double buffer mode for RX.
Cliff
--- Vivi Li 2009-06-16 22:38:51
This problem is ok now with BF548-EZKIT/BF527-EZKIT of pio mode.
For dma mode, I hardly to see such bug on BF548-EZKIT; but for BF527-EZKIT,
it's easy to find such bug.
--- Cliff Cai 2009-06-17 06:20:57
it's another problem,So close this and open new one.
Cliff
--- Vivi Li 2009-06-18 00:42:51
This problem is fixed for PIO mode.
DMA mode still have problem.
Close this one and open a new bug [#5267] about dma mode.
Files
Changes
Commits
Dependencies
Duplicates
Associations
Tags
File Name File Type File Size Posted By
gadget_ethernet_test_log.fail text/plain 16141 Vivi Li