What sort of transfer speeds are to be expected throught the USB-net protocol on DM365?

Hi,

Can you elaborate more on the USB network setup ? Does DM365 act as RNDIS device or acts as USB Host connecting to a USB network dongle ?

Please refer to the pdf, where USB performance numbers are discussed for DM365 : www.ti.com/.../sprs566a.pdf

Jmfras,

DM365 BSP package contains a performance guide inside the document section, which might help you in further analysis. 

However, please note that 50MBps for USB 2.0 on DM365 are the very high expectations :)

This may not be feasible to reach this numbers on practical end-applications due to high CPU load. Please find the below link FYR for tested speeds. 

Good part is,  there is definite scope for improvement in USB throughput by optimizing the software including BSP and your applications :)

Based on the use case requirements, we can improve the throughput :)

I don't know, the DM365 is built in a DLP3000, but I can't find in the documentation how it acts. I just know it goes USB->Net from the GUI source code. Currently I'm using the following code (should it help) to connect with the DLP:

/* Socket Structure */
struct sockaddr_in sa;
sa.sin_family = AF_INET;
sa.sin_addr.s_addr = inet_addr(ip);
sa.sin_port = htons(0x5555);
int dm365 = socket(AF_INET,SOCK_STREAM,0);

/* Attempt to Establish Connection */
if (connect(dm365,(sockaddr*) &sa,sizeof(sa))!=-1){
cout<<"Connected."<<endl;
cout<<" "<<endl;}
else{
cout<<"Connection failed."<<endl;
return -1;}

I wasn't really expecting such high rate of transfer, but certainly not having to wait 8s for 5MB. I've been clocking all functions I'm using and on my end it's taking 0.05s per sequence (of 96 patterns with 52kB each), the problem is the clocking only accounts cpu tics, so it isn't considering hang time, so I'm not sure exactly what the problem is.

Hi,

I would suggest you to perform iperf test, You can cross compile and run to check. Also what is the turn around time when you give a ping command in shell ?
Also it would be helpful if you can give your dmesg logs and ifconfig output

ping:
--- 192.168.1.100 ping statistics ---
26 packets transmitted, 26 received, 0% packet loss, time 24998ms
rtt min/avg/max/mdev = 0.482/0.534/0.607/0.035 ms

ifconfing:
usb0 Link encap:Ethernet HWaddr 0a:87:55:02:f8:6a
inet addr:192.168.1.99 Bcast:192.168.1.255 Mask:255.255.255.0
inet6 addr: fe80::887:55ff:fe02:f86a/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:174 errors:0 dropped:0 overruns:0 frame:0
TX packets:1745 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:10464 (10.4 KB) TX bytes:2063727 (2.0 MB)

I'll look into that iperf test, see if I can manage to do that. As for dmesg I have no clue.

Hi,

Just type in command line like ifconfig
# dmesg

Your ping takes really long time. Also if can share the kernel config being used. What version of kernel ? How it is connected to network(may be any cable like ethernet connected to DM36x ?)

So, I guess it's acting as RNDIS. It's connected through USB.

dmesg:
[61606.464243] usb 3-2: new full-speed USB device number 19 using xhci_hcd
[61606.592424] usb 3-2: not running at top speed; connect to a high speed hub
[61606.593490] usb 3-2: New USB device found, idVendor=0525, idProduct=a4a2
[61606.593492] usb 3-2: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[61606.593494] usb 3-2: Product: RNDIS/Ethernet Gadget
[61606.593495] usb 3-2: Manufacturer: Linux 2.6.32.17-davinci1 with musb_hdrc
[61606.600062] cdc_ether 3-2:1.0 usb0: register 'cdc_ether' at usb-0000:00:14.0-2, CDC Ethernet Device, 0a:87:55:02:f8:6a
[61606.674317] cdc_ether 3-2:1.0 usb0: kevent 12 may have been dropped
[61888.627052] systemd-hostnamed[7812]: Warning: nss-myhostname is not installed. Changing the local hostname might make it unresolveable. Please install nss-myhostname!
[62060.446749] usb 3-2: USB disconnect, device number 19
[62060.446822] cdc_ether 3-2:1.0 usb0: unregister 'cdc_ether' usb-0000:00:14.0-2, CDC Ethernet Device
[62357.364379] usb 3-2: new full-speed USB device number 20 using xhci_hcd
[62357.492567] usb 3-2: not running at top speed; connect to a high speed hub
[62357.493469] usb 3-2: New USB device found, idVendor=0525, idProduct=a4a2
[62357.493472] usb 3-2: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[62357.493474] usb 3-2: Product: RNDIS/Ethernet Gadget
[62357.493477] usb 3-2: Manufacturer: Linux 2.6.32.17-davinci1 with musb_hdrc
[62357.499962] cdc_ether 3-2:1.0 usb0: register 'cdc_ether' at usb-0000:00:14.0-2, CDC Ethernet Device, 0a:87:55:02:f8:6a
[62357.570131] cdc_ether 3-2:1.0 usb0: kevent 12 may have been dropped
[62641.085674] usb 3-2: USB disconnect, device number 20
[62641.085739] cdc_ether 3-2:1.0 usb0: unregister 'cdc_ether' usb-0000:00:14.0-2, CDC Ethernet Device
[63187.368544] usb 3-2: new full-speed USB device number 21 using xhci_hcd
[63187.496732] usb 3-2: not running at top speed; connect to a high speed hub
[63187.497662] usb 3-2: New USB device found, idVendor=0525, idProduct=a4a2
[63187.497665] usb 3-2: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[63187.497667] usb 3-2: Product: RNDIS/Ethernet Gadget
[63187.497668] usb 3-2: Manufacturer: Linux 2.6.32.17-davinci1 with musb_hdrc
[63187.504067] cdc_ether 3-2:1.0 usb0: register 'cdc_ether' at usb-0000:00:14.0-2, CDC Ethernet Device, 0a:87:55:02:f8:6a
[63187.580074] cdc_ether 3-2:1.0 usb0: kevent 12 may have been dropped

kernel:

4.4.0-97-generic

Hi,

This looks to be dmesg from PC side. Can you share the dmesg from the DM36x side too ? Looks like it is configured as full speed(USB 1.1).

Same as done in PC. If you are getting DM36x shell, type
# dmesg

Oh, sorry the ping info I left before was also done on my end not on the DM36x shell. How do I access it?

Sorry, but how do I access DM365's shell?

please?

Finally I have some results:

ifconfig:
usb0 Link encap:Ethernet HWaddr 2A:C2:28:8B:90:6B
inet addr:192.168.1.100 Bcast:192.168.1.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:67750 errors:37 dropped:0 overruns:0 frame:37
TX packets:66769 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:4402136 (4.1 MiB) TX bytes:3624886 (3.4 MiB)

dmesg:
Linux version 2.6.32.17-davinci1 (sanjeev@sanjeev-VirtualBox) (gcc version 4.4.1 (Sourcery G++ Lite 2010q1-202) ) #5 PREEMPT Fri Sep 20 16:14:26 IST 2013
CPU: ARM926EJ-S [41069265] revision 5 (ARMv5TEJ), cr=00053177
CPU: VIVT data cache, VIVT instruction cache
Machine: DaVinci DM365 LCr
Memory policy: ECC disabled, Data cache writeback
On node 0 totalpages: 29696
free_area_init_node: node 0, pgdat c0342804, node_mem_map c035a000
DMA zone: 232 pages used for memmap
DMA zone: 0 pages reserved
DMA zone: 29464 pages, LIFO batch:7
DaVinci dm36x_rev1.2 variant 0x8
Built 1 zonelists in Zone order, mobility grouping on. Total pages: 29464
Kernel command line: mem=116M console=ttyS0,115200n8 root=/dev/ram0 rw initrd=0x82000000,5M ramdisk_size=15360 ip=off video=davincifb:vid0=1216x684x16,4050K:vid1=off:osd0=0x0x8,0K:osd1=0x0x8,0K dm365_imp.oper_mode=0 davinci_enc_mngr.ch0_output=LCD davinci_enc_mngr.ch0_mode=1216x684YUV vpfe_capture.interface=1
PID hash table entries: 512 (order: -1, 2048 bytes)
Dentry cache hash table entries: 16384 (order: 4, 65536 bytes)
Inode-cache hash table entries: 8192 (order: 3, 32768 bytes)
Memory: 116MB = 116MB total
Memory: 108956KB available (3024K code, 243K data, 124K init, 0K highmem)
SLUB: Genslabs=11, HWalign=32, Order=0-3, MinObjects=0, CPUs=1, Nodes=1
Hierarchical RCU implementation.
NR_IRQS:245
Console: colour dummy device 80x30
Calibrating delay loop... 147.86 BogoMIPS (lpj=739328)
Mount-cache hash table entries: 512
CPU: Testing write buffer coherency: ok
DaVinci: 8 gpio irqs
NET: Registered protocol family 16
davinci_serial_init:97: failed to get UART2 clock
bio: create slab <bio-0> at 0
DM365 IPIPE initialized in Continuous mode
vpss vpss: dm365_vpss vpss probed
vpss vpss: dm365_vpss vpss probe success
dm365_afew_hw_init
ch0 default output "LCD", mode "1216x684YUV"
be_encoder_setoutput>
Setting output to Composite
Start of vpbe_encoder_setmode..
pbe_encoder_setmode>
pbe_encoder_setoutput>
VPBE Encoder initialized
be_encoder_enumoutput>
pbe_encoder_enumoutput>
be_encoder_enumoutput>
pbe_encoder_enumoutput>
be_encoder_enumoutput>
pbe_encoder_enumoutput>
be_encoder_getoutput>
pbe_encoder_getoutput>
be_encoder_getmode>
be_encoder_getmode/>
VPBE Encoder Initialized
VPBE Encoder de-initialized
gicpd_encoder_setoutput>
Start of logicpd_encoder_setmode..
ogicpd_encoder_setoutput>
LogicPD Encoder initialized
gicpd_encoder_enumoutput>
ogicpd_encoder_enumoutput>
gicpd_encoder_setoutput>
Start of logicpd_encoder_setmode..
ogicpd_encoder_setoutput>
Start of logicpd_encoder_setmode..
gicpd_encoder_getoutput>
ogicpd_encoder_getoutput>
gicpd_encoder_getmode>
ogicpd_encoder_getmode>
LogicPD encoder initialized
Switching to clocksource timer0_1
musb_hdrc: version 6.0, cppi-dma, peripheral, debug=0
musb_hdrc: ConfigData=0x06 (UTMI-8, dyn FIFOs, SoftConn)
musb_hdrc: MHDRC RTL version 1.500
musb_hdrc: setup fifo_mode 2
musb_hdrc: 9/9 max ep, 4032/4096 memory
musb_hdrc: USB Peripheral mode controller at fec64000 using DMA, IRQ 12
NET: Registered protocol family 2
IP route cache hash table entries: 1024 (order: 0, 4096 bytes)
TCP established hash table entries: 4096 (order: 3, 32768 bytes)
TCP bind hash table entries: 4096 (order: 2, 16384 bytes)
TCP: Hash tables configured (established 4096 bind 4096)
TCP reno registered
NET: Registered protocol family 1
Trying to unpack rootfs image as initramfs...
rootfs image is not initramfs (no cpio magic); looks like an initrd
Freeing initrd memory: 5120K
JFFS2 version 2.2. (NAND) © 2001-2006 Red Hat, Inc.
msgmni has been set to 223
alg: No test for stdrng (krng)
io scheduler noop registered
io scheduler anticipatory registered (default)
gicpd_encoder_getmode>
ogicpd_encoder_getmode>
gicpd_encoder_getmode>
ogicpd_encoder_getmode>
gicpd_encoder_getmode>
ogicpd_encoder_getmode>
davincifb davincifb.0: dm_vid0_fb: 1216x684x16@0,0 with framebuffer size 4050KB
DM365 IPIPEIF probed
imp serializer initialized
davinci_previewer initialized
davinci_resizer initialized
Serial: 8250/16550 driver, 2 ports, IRQ sharing disabled
serial8250.0: ttyS0 at MMIO 0x1c20000 (irq = 40) is a 16550A
console [ttyS0] enabled
serial8250.0: ttyS1 at MMIO 0x1d06000 (irq = 41) is a 16550A
brd: module loaded
loop: module loaded
NAND device: Manufacturer ID: 0xec, Chip ID: 0xf1 (Samsung NAND 128MiB 3,3V 8-bit)
Bad block table found at page 65472, version 0x01
Bad block table found at page 65408, version 0x01
nand_read_bbt: Bad block at 0x000000040000
nand_read_bbt: Bad block at 0x0000007a0000
nand_read_bbt: Bad block at 0x000002480000
nand_read_bbt: Bad block at 0x0000039c0000
nand_read_bbt: Bad block at 0x000004a60000
nand_read_bbt: Bad block at 0x000004f40000
nand_read_bbt: Bad block at 0x000007700000
Creating 4 MTD partitions on "davinci_nand.0":
0x000000000000-0x000000400000 : "bootloader"
0x000000400000-0x000000800000 : "kernel"
0x000000800000-0x000000c00000 : "rootfs"
0x000000c00000-0x000008000000 : "data"
davinci_nand davinci_nand.0: controller rev. 2.3
spi_davinci spi_davinci.0: Controller at 0xfec66000
g_ether gadget: using random self ethernet address
g_ether gadget: using random host ethernet address
usb0: MAC 2a:c2:28:8b:90:6b
usb0: HOST MAC aa:28:9a:26:ac:40
g_ether gadget: Ethernet Gadget, version: Memorial Day 2008
g_ether gadget: g_ether ready
i2c /dev entries driver
Linux video capture interface: v2.00
vpfe_init
vpfe-capture: vpss clock vpss_master enabled
vpfe-capture vpfe-capture: v4l2 device registered
vpfe-capture vpfe-capture: video device registered
IndPico : switch to HD imager-MT9P031 video input
Sanjeev::MT9P031-> Probing for the camera device...
Sanjeev::MT9P031-> Camera device not detected!!!
vpfe-capture vpfe-capture: v4l2 sub device mt9p031 register fails
suresh:mt9m112_probe()
Sanjeev::MT9M112-> Probing for the camera device...
suresh:mt9m112 : f0 <= 0
mt9m112 : page reg write failed !!!!!!!
mt9m112 1-0048: No MT9M11x chip detected, register read ffffff87
Sanjeev::MT9M112-> Camera device not detected!!!
vpfe-capture vpfe-capture: v4l2 sub device mt9m112 register fails
mt9v113 1-003c: No MT9V113 chip detected, register read ffffff87
vpfe-capture vpfe-capture: v4l2 sub device mt9v113 register fails
No sub devices registered
vpfe_register_ccdc_device: DM365 ISIF
dm365_isif: probe of dm365_isif failed with error -22
af major#: 252, minor# 0
AF Driver initialized
aew major#: 251, minor# 0
AEW Driver initialized
vinci_display_init>
Trying to register davinci display video device.
layer=c669a000,layer->video_dev=c669a164
Trying to register davinci display video device.
layer=c669a400,layer->video_dev=c669a564
davinci_init:DaVinci V4L2 Display Driver V1.0 loaded
avinci_init>
watchdog watchdog: heartbeat 60 sec
davinci_mmc davinci_mmc.0: Using DMA, 4-bit mode
TCP cubic registered
NET: Registered protocol family 17
Clocks: disable unused mmcsd1
Clocks: disable unused spi1
Clocks: disable unused spi2
Clocks: disable unused spi3
Clocks: disable unused spi4
Clocks: disable unused pwm0
Clocks: disable unused pwm1
Clocks: disable unused pwm2
Clocks: disable unused pwm3
Clocks: disable unused timer1
Clocks: disable unused timer3
Clocks: disable unused emac
Clocks: disable unused voice_codec
Clocks: disable unused asp0
Clocks: disable unused rto
Clocks: disable unused mjcp
RAMDISK: gzip image found at block 0
g_ether gadget: full speed config #1: CDC Ethernet (ECM)
VFS: Mounted root (ext2 filesystem) on device 1:0.
Freeing init memory: 124K
g_ether gadget: full speed config #1: CDC Ethernet (ECM)

In the pdf you linked to I couldn't find such maximum referenced, it was something of the order of 5-10MB/s, could you point me to where that value is mentioned?
What sort of tuning do you recommend?

That is the standard, which you can find online for Full Speed USB. I am talking about tuning the kernel buffer sizes like read and write like in this link: wwwx.cs.unc.edu/.../network_tuning.php

Thanks.