[#4739] netperf_tcp_rr through bfin_mac is almost zero with set buffer packets

Submitted By: Mingquan Pan

Open Date

2008-12-22 21:47:32 Close Date

2009-02-10 23:05:05

Priority:

Medium Assignee:

Michael Hennerich

Status:

Closed Fixed In Release:

N/A

Found In Release:

N/A Release:

Category:

N/A Board:

N/A

Processor:

BF537 Silicon Revision:

Is this bug repeatable?:

Yes Resolution:

Duplicate

Uboot version or rev.:

Toolchain version or rev.:

Nov 11

App binary format:

N/A

Summary: netperf_tcp_rr through bfin_mac is almost zero with set buffer packets

Details:

netperf_tcp_rr on bf527 is almost zero with set buffer packets on trunk now.

config:

Ethernet (10 or 100Mbit) (NET_ETHERNET) [Y/n/?] ^M

Generic Media Independent Interface device support (MII) [Y/?] y^M

Blackfin on-chip MAC support (BFIN_MAC) [Y/n/m/?] ^M

Use L1 memory for rx/tx packets (BFIN_MAC_USE_L1) [Y/n/?] N^M

Number of transmit buffer packets (BFIN_TX_DESC_NUM) [10] 100^M

Number of receive buffer packets (BFIN_RX_DESC_NUM) [20] 500^M

RMII PHY Interface (EXPERIMENTAL) (BFIN_MAC_RMII) [Y/n/?] ^M

log:

Loading .text @ 0x00001000 (1196304 bytes)

Loading .rodata @ 0x00125110 (299720 bytes)

Loading .init.rodata @ 0x0016e3d8 (156 bytes)

Loading __ksymtab @ 0x0016e474 (16528 bytes)

Loading __ksymtab_gpl @ 0x00172504 (5752 bytes)

Loading __ksymtab_strings @ 0x00173b7c (50054 bytes)

Loading __init_rodata @ 0x0017ff04 (172 bytes)

Loading __param @ 0x0017ffb0 (640 bytes)

Clearing .bss @ 0x00180230 (66432 bytes)

Loading .data @ 0x001905b0 (72272 bytes)

Loading .init.text @ 0x001a2000 (102020 bytes)

Loading .init.data @ 0x001bae84 (20708 bytes)

Loading .init.setup @ 0x001bff68 (608 bytes)

Loading .initcall.init @ 0x001c01c8 (656 bytes)

Loading .con_initcall.init @ 0x001c0458 (4 bytes)

Loading .init.ramfs @ 0x001c045c (3084968 bytes)

Loading .text_l1 @ 0xffa00000 (5500 bytes)

sh_addr: FFA00000, p_paddr: 004B1704

Loading from: 024A2000 to 004B1704, size: 5500

Loading .data_l1 @ 0xff800000 (192 bytes)

sh_addr: FF800000, p_paddr: 004B2C80

Loading from: 024A4000 to 004B2C80, size: 192

## Starting application at 0x001af490 ...

Linux version 2.6.28-rc2-ADI-2009R1-pre-svn5830 (test@uclinux60-bf527-std) (gcc version 4.1.2 (ADI svn)) #68 Sat Dec 6 13:23:10 CST 2008

console [early_BFuart0] enabled

early printk enabled on early_BFuart0

Board Memory: 64MB

Kernel Managed Memory: 64MB

Memory map:

fixedcode = 0x00000400-0x00000490

text = 0x00001000-0x00125110

rodata = 0x00125110-0x00180230

bss = 0x00180230-0x001905b0

data = 0x001905b0-0x001a2000

stack = 0x001a0000-0x001a2000

init = 0x001a2000-0x004b3000

available = 0x004b3000-0x03eff000

DMA Zone = 0x03f00000-0x04000000

Hardware Trace Active and Enabled

Reset caused by Software reset

Compiled for ADSP-BF527 Rev 0.0

ERROR: Not running on ADSP-BF527: unknown CPUID 0x27e0 Rev 0.0

Blackfin Linux support by blackfin.uclinux.org/

Processor Speed: 525 MHz core clock and 131 MHz System Clock

NOMPU: setting up cplb tables for global access

Instruction Cache Enabled for CPU0

Data Cache Enabled for CPU0 (write-through)

Built 1 zonelists in Zone order, mobility grouping off. Total pages: 16001

Kernel command line: root=/dev/mtdblock0 rw earlyprintk=serial,uart0,57600 ip=10.100.4.50:10.100.4.174:10.100.4.174:255.255.255.0:bf527-ezkit:eth0:off

Configuring Blackfin Priority Driven Interrupts

PID hash table entries: 256 (order: 8, 1024 bytes)

console handover: boot [early_BFuart0] -> real [ttyBF0]

Dentry cache hash table entries: 8192 (order: 3, 32768 bytes)

Inode-cache hash table entries: 4096 (order: 2, 16384 bytes)

Memory available: 59116k/65536k RAM, (3140k init code, 1168k kernel code, 500k data, 1024k dma, 584k reserved)

Calibrating delay loop... 1044.48 BogoMIPS (lpj=2088960)

Security Framework initialized

Mount-cache hash table entries: 512

Blackfin Scratchpad data SRAM: 4 KB

Blackfin L1 Data A SRAM: 16 KB (15 KB free)

Blackfin L1 Data B SRAM: 16 KB (16 KB free)

Blackfin L1 Instruction SRAM: 48 KB (42 KB free)

PDA for CPU0 reserved at 00181228

net_namespace: 288 bytes

NET: Registered protocol family 16

Blackfin GPIO Controller

Blackfin DMA Controller

stamp_init(): registering device resources

usbcore: registered new interface driver usbfs

usbcore: registered new interface driver hub

usbcore: registered new device driver usb

musb_hdrc: version 6.0, pio, host, debug=0

musb_hdrc: USB Host mode controller at ffc03800 using PIO, IRQ 59

musb_hdrc musb_hdrc.0: MUSB HDRC host driver

musb_hdrc musb_hdrc.0: new USB bus registered, assigned bus number 1

usb usb1: configuration #1 chosen from 1 choice

hub 1-0:1.0: USB hub found

hub 1-0:1.0: 1 port detected

NET: Registered protocol family 2

IP route cache hash table entries: 1024 (order: 0, 4096 bytes)

TCP established hash table entries: 2048 (order: 2, 16384 bytes)

TCP bind hash table entries: 2048 (order: 1, 8192 bytes)

TCP: Hash tables configured (established 2048 bind 2048)

TCP reno registered

NET: Registered protocol family 1

msgmni has been set to 115

io scheduler noop registered

io scheduler anticipatory registered (default)

io scheduler cfq registered

bfin-t350mcqb: 320x240 24-bit RGB FrameBuffer initializing...

dma_alloc_init: dma_page @ 0x03a00000 - 256 pages at 0x03f00000

bfin-otp: initialized

Serial: Blackfin serial driver

bfin-uart.1: ttyBF0 at MMIO 0xffc02000 (irq = 31) is a BFIN-UART

brd: module loaded

bfin_mac_mdio: probed

bfin_mac: attached PHY driver [Generic PHY] (mii_bus:phy_addr=0:03, irq=-1, mdc_clk=2500000Hz(mdc_div=25)@sclk=131MHz)

bfin_mac bfin_mac.0: Blackfin on-chip Ethernet MAC driver, Version 1.1

m25p80 spi0.1: m25p16 (2048 Kbytes)

Creating 2 MTD partitions on "m25p80":

0x00000000-0x00040000 : "bootloader(spi)"

0x00040000-0x00200000 : "linux kernel(spi)"

bfin-spi bfin-spi.0: Blackfin on-chip SPI Controller Driver, Version 1.0, regs_base@ffc00500, dma channel@7

rtc-bfin rtc-bfin: rtc core: registered rtc-bfin as rtc0

bfin-wdt: initialized: timeout=20 sec (nowayout=0)

hid_tmff: Unknown symbol _input_ff_create_memless

usbcore: registered new interface driver usbhid

usbhid: v2.6:USB HID core driver

TCP cubic registered

NET: Registered protocol family 17

rtc-bfin rtc-bfin: setting system clock to 2004-06-01 03:57:55 UTC (1086062275)

IP-Config: Complete:

device=eth0, addr=10.100.4.50, mask=255.255.255.0, gw=10.100.4.174,

host=bf527-ezkit, domain=, nis-domain=(none),

bootserver=10.100.4.174, rootserver=10.100.4.174, rootpath=

Freeing unused kernel memory: 3140k freed

_____________________________________

a8888b. / Welcome to the uClinux distribution \

d888888b. / _ _ \

8P"YP"Y88 / | | |_| __ __ (TM) |

8|o||o|88 _____/ | | _ ____ _ _ \ \/ / |

8' .88 \ | | | | _ \| | | | \ / |

8`._.' Y8. \ | |__ | | | | | |_| | / \ |

d/ `8b. \ \____||_|_| |_|\____|/_/\_\ |

dP . Y8b. \ For embedded processors including |

d8:' " `::88b \ the Analog Devices Blackfin /

d8" 'Y88b \___________________________________/

:8P ' :888

8a. : _a88P For further information, check out:

._/"Yaa_: .| 88P| - blackfin.uclinux.org/

\ YP" `| 8P `. - docs.blackfin.uclinux.org/

/ \.___.d| .' - www.uclinux.org/

`--..__)8888P`._.' jgs/a:f - www.analog.com/blackfin

Have a lot of fun...

BusyBox v1.13.1 (2008-12-06 11:48:09 CST) built-in shell (msh)

Enter 'help' for a list of built-in commands.

root:/>

************** STEP 3: Start testing.

ifconfig eth0 10.100.4.50

root:/> PHY: 0:03 - Link is Up - 100/Full

test@uclinux60-bf527-std:~/work/cruise/checkouts/uclinux-dist/testsuites/ethernet> sudo ./netserver_x86

root's password:

Starting netserver at port 12865

test@uclinux60-bf527-std:~/work/cruise/checkouts/uclinux-dist/testsuites/ethernet> ps aux| grep netserver_x86 | head -1

root 16086 0.0 0.0 1816 240 ? Ss 13:24 0:00 ./netserver_x86

test@uclinux60-bf527-std:~/work/cruise/checkouts/uclinux-dist/testsuites/ethernet> usb 1-1: new high speed USB device using musb_hdrc and address 2

usb 1-1: configuration #1 chosen from 1 choice

cd /

root:/> chmod 777 netperf_script

root:/> ./netperf_script

root:/> sudo pkill netserver_x86

root's password:

test@uclinux60-bf527-std:~/work/cruise/checkouts/uclinux-dist/testsuites/ethernet> cat netperf_tcp_rr

TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.100.4.174 (10.100.4.174) port 0 AF_INET

Local /Remote

Socket Size Request Resp. Elapsed Trans.

Send Recv Size Size Time Rate

bytes Bytes bytes bytes secs. per sec

16384 87380 1 1 600.00 0.00

16384 87380

root:/> cat netperf_tcp_stream

TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.100.4.174 (10.100.4.174) port 0 AF_INET

netperf: receive_response: no response received. errno 110 counter 0

root:/>

Case 1 ...FAIL

while on bf537, with the same config it still has 3585.21 rate per sec.

Follow-ups

--- Sonic Zhang 2008-12-31 01:20:40

Can't replicated on kernel svn 5932.

root:/bin> netperf -H 10.100.4.174 -t TCP_STREAM -l 60

TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.100.4.174

(10.100.4.174) port 0 AF_INET

Recv Send Send

Socket Socket Message Elapsed

Size Size Size Time Throughput

bytes bytes bytes secs. 10^6bits/sec

87380 16384 16384 60.00 59.33

root:/bin> netperf -H 10.100.4.174 -t TCP_RR -l 60

TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.100.4.174

(10.100.4.174) port 0T

Local /Remote

Socket Size Request Resp. Elapsed Trans.

Send Recv Size Size Time Rate

bytes Bytes bytes bytes secs. per sec

16384 87380 1 1 60.00 3536.00

16384 87380

--- Mingquan Pan 2009-01-14 22:48:41

It looks still has this problem on svn head kernel on Linux version

2.6.28-ADI-2009R1-pre-svn6015 (test@uclinux60-bf527-std) (gcc version 4.1.2 (ADI

svn)) on board: CPU: ADSP bf527-0.0 (Detected Rev: 0.0).

And from the log with the set TX_DESC and RX_DESC value, it even can't rcp the

script over to the board while ping is through but have packets loss.

test@uclinux60-bf527-std:~/work/cruise/checkouts/uclinux-dist/testsuites/ethernet>

rcp netperf_script root@10.100.4.50:/

test@uclinux60-bf527-std:~/work/cruise/checkouts/uclinux-dist/testsuites/ethernet>

ping 10.100.4.50

PING 10.100.4.50 (10.100.4.50) 56(84) bytes of data.

64 bytes from 10.100.4.50: icmp_seq=2 ttl=64 time=0.201 ms

64 bytes from 10.100.4.50: icmp_seq=3 ttl=64 time=0.183 ms

64 bytes from 10.100.4.50: icmp_seq=7 ttl=64 time=0.197 ms

64 bytes from 10.100.4.50: icmp_seq=8 ttl=64 time=0.216 ms

64 bytes from 10.100.4.50: icmp_seq=9 ttl=64 time=0.199 ms

64 bytes from 10.100.4.50: icmp_seq=10 ttl=64 time=0.261 ms

64 bytes from 10.100.4.50: icmp_seq=11 ttl=64 time=0.192 ms

64 bytes from 10.100.4.50: icmp_seq=22 ttl=64 time=0.148 ms

64 bytes from 10.100.4.50: icmp_seq=24 ttl=64 time=0.135 ms

--- 10.100.4.50 ping statistics ---

26 packets transmitted, 9 received, 65% packet loss, time 25041ms

rtt min/avg/max/mdev = 0.135/0.192/0.261/0.036 ms

The kernel config file is attached.

--- Mingquan Pan 2009-01-16 02:32:16

It is now seen on bf537 stamp board too with Linux version

2.6.28-ADI-2009R1-pre-svn6019 (test@uclinux54-adv7393-ad1981-usb1362) (gcc

version 4.1.2 (ADI svn)).

sh-3.00$ sudo ./netserver_x86

Password:

Starting netserver at port 12865

sh-3.00$ ps aux| grep netserver_x86 | head -1

root 28179 0.0 0.0 1692 436 ? Ss 14:13 0:00

./netserver_x86

sh-3.00$ cd /

root:/> chmod 777 netperf_script

root:/> ./netperf_script

root:/> sudo pkill netserver_x86

Password:

sh-3.00$ cat netperf_tcp_rr

TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.100.4.174

(10.100.4.174) port 0 AF_INET

Local /Remote

Socket Size Request Resp. Elapsed Trans.

Send Recv Size Size Time Rate

bytes Bytes bytes bytes secs. per sec

16384 87380 1 1 600.00 0.00

16384 87380

root:/> cat netperf_tcp_stream

TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.100.4.174

(10.100.4.174) port 0 AF_INET

netperf: receive_response: no response received. errno 110 counter 0

root:/>

Case 1 ...FAIL

--- Mingquan Pan 2009-02-03 21:57:43

This is seen under the defauft clocks:

Processor Speed: 500 MHz core clock and 100 MHz System Clock^M.

If the clock is set at:

Processor Speed: 500 MHz core clock and 125 MHz System Clock

then the results is ok.

TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.100.4.174

(10.100.4.174) port 0 AF_INET^M

Local /Remote^M

Socket Size Request Resp. Elapsed Trans.^M

Send Recv Size Size Time Rate ^M

bytes Bytes bytes bytes secs. per sec ^M

16384 87380 1 1 600.00 4371.53 ^M

16384 87380 ^M

root:/> cat netperf_tcp_stream^M

TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.100.4.174

(10.100.4.174) port 0 AF_INET^M

Recv Send Send ^M

Socket Socket Message Elapsed ^M

Size Size Size Time Throughput ^M

bytes bytes bytes secs. 10^6bits/sec ^M

87380 16384 16384 600.00 93.77 ^M

--- Sonic Zhang 2009-02-03 22:26:14

So, this bug occurs when sclk=100 and cclk=500, while disappears when sclk=125

cclk=500. And the TX and RX buffer sizes are not the default values. TX buffer

size is set to 100 and RX to 500. This could be a bug in bfin_mac driver.

--- Michael Hennerich 2009-02-04 03:59:41

Grace,

is it also only seen with "set buffer packets" as described in the

earlier bug description? Regardless of SCLK?

-Michael

--- Sonic Zhang 2009-02-04 04:23:31

The latest result shows it only occur in 100/500 with Tx=100 and Rx=500.

--- Sonic Zhang 2009-02-08 21:49:17

duplicated bug of 4836

--- Mingquan Pan 2009-02-10 23:05:05

Yeah, this issue goes awag now and can get better results, like :

Starting Kernel at = 1ad48c

Linux version 2.6.28.3-ADI-2009R1-pre-svn6083 (test@uclinux60-bf527-std) (gcc

version 4.1.2 (ADI svn)) #30 Tue Feb 10 19:06:46 CST 2009

console [early_BFuart0] enabled

early printk enabled on early_BFuart0

Board Memory: 64MB

Kernel Managed Memory: 64MB

Memory map:

fixedcode = 0x00000400-0x00000490

text = 0x00001000-0x001229b0

rodata = 0x001229b0-0x0017e08c

bss = 0x0017e090-0x0018d07c

data = 0x0018d07c-0x001a0000

stack = 0x0019e000-0x001a0000

init = 0x001a0000-0x0077a000

available = 0x0077a000-0x03eff000

DMA Zone = 0x03f00000-0x04000000

Hardware Trace Active and Enabled

Boot Mode: 1

...

test@uclinux60-bf527-std:~/work/cruise/checkouts/uclinux-dist/testsuites/ethernet>

cat netperf_tcp_rr^M

TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.100.4.174

(10.100.4.174) port 0 AF_INET^M

Local /Remote^M

Socket Size Request Resp. Elapsed Trans.^M

Send Recv Size Size Time Rate ^M

bytes Bytes bytes bytes secs. per sec ^M

16384 87380 1 1 600.00 3674.37 ^M

16384 87380 ^M

root:/> cat netperf_tcp_stream^M

TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.100.4.174

(10.100.4.174) port 0 AF_INET^M

netperf: receive_response: no response received. errno 110 counter 0^M

root:/>

Case 1 ...FAIL

Since the error when catting netperf_tcp_stream is tracked in bug [1433]. So

close this.

Files

Changes

Commits

Dependencies

Duplicates

Associations

Partner Forums

Partner Libraries

[#4739] netperf_tcp_rr through bfin_mac is almost zero with set buffer packets