DOS Stamp Internet Made Easy (DIME) v1.0b User`s Manual
Contents
1. C programming language How to use your C compiler DOS usage and application development Reading simple wiring diagrams An addition the following items are required e DOS Stamp with the standard second serial port configuration RS 232 and the DiskOnChip option e Cables to connect the DOS Stamp to your PC and a modem e External modem with RS 232 interface Borland Turbo C compiler v3 0 or higher Other DOS targeting C compilers may be used but source code changes may be required Two ISP Internet accounts at least one of which is dial up access A browser for testing An e mail account for testing eTCP WATTCP manual The book uC OS II The Real Time Kernel by Jean J Labrosse Serial Ports RS 232 and Modem Interfacing Most DIME applications will connect to the Internet using dial up phone service with regular modems or over the air with radio modems To use a modem with DIME it must be interfaced to the DOS Stamp s COM2 serial port RS 232 signals swing both positive 12V and negative 12V with respect to ground PC serial ports have the following RS 232 signals DIME_manual doc Page 1 DIME v1 0b User s Manual Name Purpose Details Transmit Data TX Outgoing data Output Positive voltage represents a 0 Negative voltage represent a 1 Receive Data RX Incoming data Input Positive voltage represents a 0 Negative voltage represent a 1 Clear To2. The maximum recommended baud rate is 38 4K The second option local specifies what type of physical link is used direct connection local or modem modem Use the local option even if you are using a modem because we do not want DSPPPD to monitor the modem DCD signal your application does that The third option crtscts specifies that hardware handshaking is not to be used DIME does not currently support hardware handshaking which is not necessary since the protocols are so robust The fourth option asyncmap a0000 specifies which values are to be escaped over the PPP link by setting the ACCM during LCP negotiation An ACCM of 000A0000 is commonly used by most ISPs The fifth option user lt username gt is your ISP account username The sixth option passwd lt password gt is your ISP account password When DSPPPD is invoked it initializes the UART goes resident returns to the DOS prompt and waits for a signal from your application that the physical layer is up 2 Invoke your application at the DOS prompt Your application establishes a connection by bringing the physical layer up The physical layer is managed by ConnectionTask which should be inside your application 3 Send a dialing command to the modem in ConnectionTask case CONN_NO see if we are signaled to establish a connection if openconn send modem dialing commands WriteSerialString ATZ r n OSTimeDIyHMSM 0 0 1 0 wait 1
3. Modem Connections Since the modem s DCD signal is RS 232 it swings way beyond the voltage limits of the PIO pin We need to limit its voltage range before passing it on to the PIO pin Resistor R1 works in conjunction with the PIO pin s internal chip protection diodes not shown to clamp the voltage swing to a safe range DIME does not use hardware handshaking so the modem s need for a CTS signal is satisfied by looping back its RTS output DIME_manual doc Page 2 DIME v1 0b User s Manual Modems are typically controlled using the AT command set When you initialize the modem be sure to include commands for turning off all types of handshaking both hardware and software They are not necessary for DIME Internet Standards and RFCs The Internet owes its incredible growth in large part to the open standards upon which it is based In fact the main reason the Internet came to be was the mandate from the U S government Defense Advanced Research Projects Agency DARPA to unify all the proprietary information systems the U S military had acquired over the years This required the design of standard protocols for networking of computers Internet standards are now presided over by the Internet Engineering Task Force IETF The standards are called Request for Comments REC and are available for free download from http www ietf org There are thousands of RFCs covering a wide variety of networking topics but here we
4. PC connects to your ISP over a Plain Old Telephone System POTS regular phone line your PC is probably using PPP Most likely your embedded Internet application will need to use PPP to connect to an ISP POTS modems radio modems and RS 232 serial data links are subject to noise and electromagnetic interference EMI which can corrupt data and produce extraneous data PPP has built in mechanisms for validating data PPP does not have mechanisms for retransmitting corrupted data that is the responsibility of the IP transport protocol used i e TCP IP In the event of a data error PPP discards the erroneous data PPP also ignores any data that is not encapsulated in a valid PPP frame PPP cannot work with 7 bit serial data links The link must be 8 bit clean so do not try to use 7 data bits in your UART configuration You must also have full duplex separate transmit and receive lines since PPP does not support half duplex shared transmit and receive serial data links To establish a PPP connection it is essential to understand how PPP works PPP is peer oriented This means that both machines are equal in status they are not master and slave Although technically both machines are peers the other machine is frequently referred to as the peer For example when the DOS Stamp negotiates with the peer the DOS Stamp is negotiating with the machine on the other end of the link PPP Link Phases There are five phases the PPP li
5. To accommodate these needs the packet header contains the destination and source IP addresses a sequence number a checksum and other fields Some of these header areas are part of IP and others are part of TCP The low level details are not important here WATTCP takes care of these details for you TCP is considered a reliable protocol because if packets are lost or corrupted they are retransmitted within reason Obviously a TCP connection cannot be expected to eternally retransmit packets because the entire Internet would become bogged down with old packets still trying to get through So there is a Time To Live TTL field in each TCP IP packet that tells TCP when to give up It is possible to have multiple TCP connections concurrently over a single Internet connection This is done with ports Ports are used to steer the data to various network applications that may be running concurrently For example the DIME demo unit has a web server that runs on TCP Port 80 and an e mail sender that runs on TCP Port 25 Both of these TCP applications run concurrently and the port numbers are used to steer the data from the Internet connection to each application UDP IP Protocol UDP IP is an acronym for User Datagram Protocol Internet Protocol The IP indicates that it uses the IP addressing scheme It is frequently referred to as just UDP The software that provides UDP IP capability is frequently referred to as a UDP IP
6. are interested in only the ones that directly affect how we connect an embedded system to the Internet Internet Protocol 1P and IP Addresses The most fundamental part of the Internet Protocol IP is the concept of how a machine is identified on the Internet The IP address is a way of uniquely identifying individual machines When your PC connects to the Internet it uses an IP address to identify itself to other computers on the Internet Why can t your PC be anonymous you ask The IP address is used by other machines on the Internet routers to route data to and from your PC Without knowing the IP address of your PC the routers would not know where to send the data For the same reason your embedded Internet application must have an IP address to connect to the Internet An IP address is a 32 bit unsigned binary number When IP addresses are written they are expressed as a decimal number with periods written between digit groups For example the IP address 192 168 1 9 The largest value of any of the digit groups is 255 and the smallest value is 0 For example the IP address 255 0 34 127 Each digit group represents the decimal equivalent of eight of the binary bits in the IP address For example the IP address 192 168 1 9 is equivalent to the 32 bit binary form 11000000 10101000 00000001 00001001 which is also equivalent to the hexadecimal form C0 A8 01 09 There are 2 or 4 294 967 296 over 4 billion possible IP add
7. method appending the form data as a query string Also the GET method automatically refreshes the browser window after sending the form data therefore the results are immediately viewable The POST method can send more data than a query string but it does not automatically refresh the browser window The POST method is capable of creating a file and appending new form data to it The GET method cannot do this This can be a liability however You would not want a website user to crash your embedded system by filling up the disk with appended data from repeated POSTs For this reason the DIME web server has the POST method disabled by default DIME_manual doc Page 15 DIME v1 0b User s Manual HEAD Method Part of HTTP RFC 1945 The HEAD method is identical to GET except that the HTTP server does not actually send the requested information The HEAD method is frequently used for testing the validity of hyperlinks DIME does not implement HEAD and returns the HTTP error code 501 Not Implemented if a HEAD request is received Hyper Text Markup Language HTML Web pages are typically written in Hyper Text Markup Language HTML HTML combines displayed text with hidden formatting codes When your browser shows a web page it interprets the HTML formatting codes and positions the displayed text accordingly To see what the HTML representation of a web page looks like click on View gt Source in your browser Writing HTML manually i
8. reason to choose DHCP over BOOTP for PPP links The PPP link keeps the same IP address for the duration of the connection so DHCP leases are unnecessary Static IP Address If you will be using a static IP address with PPP you may still use the BOOTP and DHCP methods This works because PPP has already established the link and has been assigned the static IP address by the peer If you prefer to not use BOOTP or DHCP you may specify a static IP address in the WATTCP CFG file like this my_ip 192 168 1 9 substitute your actual IP address Gateway IP Address A good definition for gateway is the machine nearest to your machine in the chain that relays your IP traffic to the network When you use an ISP the ISP is your gateway to the Internet The ISP s gateway has a static IP address WATTCP must know the gateway s IP address In the WATTCP CFG file specify the gateway IP address like this gateway 192 168 34 1 substitute the gateway s actual IP address If you are connecting to the Internet through a local gateway in your facility you must specify its IP address for the gateway not your ISP s gateway Net Mask A net mask also known as subnet mask is a way of qualifying IP addresses over a range Any IP address not matching the pattern of the net mask is considered to not be on the subnet Subnets are used to manage DIME_manual doc Page 11 DIME v1 0b User s Manual network traffic by confining the traffic on a s
9. resident and waits for your application When your application decides to actually establish the connection it sends the modem initialization string to DSPPPD which passes the string directly through to the modem Your application then queries DSPPPD for any response strings from the modem Once your application detects a modem carrier signal it then signals DSPPPD that the physical layer is up DSPPPD then begins the PPP process Once PPP has entered the Network phase DSPPPD signals your application that the IP layer is up Your application can then begin using network applications and must monitor the status of the physical layer concurrently When the physical layer goes down your application must signal DSPPPD so that it can move the PPP link into the Dead phase DSPPPD is a Terminate and Stay Resident TSR program It is invoked once at the DOS prompt before your application is invoked Command line options are required when DSPPPD is invoked These options are explained in the pppd man file on the DIME distribution disk DIME_manual doc Page 8 DIME v1 0b User s Manual How to Establish a PPP Link Establishing a PPP link involves the following process refer to source code in DIMEDEMO C 1 Invoke the PPP packet driver at the DOS prompt with the necessary command line options For example dspppd 38400 local crtscts asyncmap a0000 user lt username gt passwd lt password gt The first option 38400 is the baud rate to use
10. stack It is not a stack in the same sense of a CPU stack but instead refers to the layering of protocols looking like a stack UDP IP is an alternative method for transmitting data across IP networks Like TCP the data is placed in packets with information about the data This information goes into the packet header The actual data is placed after the header UDP is considered an unreliable protocol because it is one way A UDP sender sends a packet and no confirmation of reception is expected or accepted If the UDP packet got lost discarded or corrupted the sender would not know Why would we want to use such an unreliable protocol There exist some network services that are best served with one way transmission For example a time server would broadcast the time over UDP periodically and would not care if any machine were listening Also the processing overhead for UDP is much less than TCP This can make a significant difference in real time applications Like TCP UDP also uses ports to steer the data to various network applications Physical Upper and Lower Layers Protocols are arranged in layers Each layer has a purpose A layer either does something with the data or establishes and maintains the link DIME_manual doc Page 4 DIME v1 0b User s Manual Upper and lower are relative terms A layer is the upper layer if it is built on the layer you are talking about For example if you are
11. the upper layer network protocols that authentication was successful If authentication fails it moves the PPP link to the Terminate phase by signaling the lower layer LCP with a command to close the link The Network phase is when the actual data transfer to and from the Internet happens Each network protocol IP IPX AppleTalk etc must configure itself with the peer before data encapsulated in that protocol can be transferred This configuration is done using Network Control Protocol NCP The PPP link stays in the Network phase until either 1 LCP receives a signal to close the link or 2 a signal is received from the physical layer that the connection has been broken In the case of 1 a terminate packet may have been received from the peer or our own application may have sent the signal In the case of 2 the modem may have lost the carrier or other some other hardware problem may have occurred The PPP link then moves to the Terminate phase The Terminate phase is when the PPP link is closed gracefully During Termination LCP is used to close the link by exchanging terminate packets with the peer Upon receipt of these packets LCP signals the lower layer physical that LCP is down closed The physical layer should then disconnect hang up The PPP link then moves to the Dead phase If the physical layer is already down not ready LCP will not bother to send terminate packets because there is no connection avai
12. A HTTP server also known as a web server listens for incoming information requests on TCP Port 80 When you use a browser you are sending HTTP requests to a HTTP server When a request is received the server parses the Uniform Resource Identifier URD string and then performs the requested action There are three actions also known as methods a HTTP server can perform e GET method e POST method e HEAD method DIME_manual doc Page 14 DIME v1 0b User s Manual GET Method Part of HTTP RFC 1945 The GET method sends the requested information to the requester Browsers send GET commands to HTTP servers to get web pages For example in your browser if you type http www bagotronix com dimedemo htm the browser sends this request to the HTTP server GET dimedemo htm HTTP 1 0 This tells the server to get the data in the file dimedemo htm and send it to the browser Notice that http www bagotronix com is not in the browser request That information is not necessary for the server but the browser uses it to determine e That the request is to be sent in the form of a HTTP request e The domain name www bagotronix com is where the request is to be sent The GET method can also be used to send information to the HTTP server This is done with a query string appended to the GET request For example in your browser if you type http www bagotronix com dimedemo htm WhatsUpDoc the browser sends this
13. DOS Stamp Internet Made Easy DIME v1 0b User s Manual revision 1 0 2000 Bagotronix Inc All rights reserved This page intentionally left blank DIME v1 0b User s Manual Overview This document explains how the Internet works and how to interface an embedded computing application to it The examples given show how to use Bagotronix products with the Internet but the principles are similar for many other computing devices Previous knowledge of Internet protocols is not necessary but is helpful The DOS Stamp Internet Made Easy DIME software makes embedded Internet easy but there are several concepts you need to understand to apply DIME to your embedded Internet application These concepts are explained in this document This document is designed to be read sequentially from beginning to end The concepts are presented in a linear fashion with the most fundamental concepts first Prerequisites It is assumed that you are already familiar with using the Internet on your PC That is how to dial into your Internet Service Provider ISP start up a browser Internet Explorer Netscape etc visit websites and send and receive e mail You must also be familiar with how to upload programs and interact with the DOS Stamp This topic is covered in the DOS Stamp User s Manual which you should read before you read this document To customize DIME for your embedded Internet application you need to possess the following skills
14. HTTP SMTP AHDLC CRC PPP DIME PPP Implementation DIME implements PPP using a modified version of DOSPPPD v0 6 which is a DOS port of the Linux pppd program Bagotronix has modified DOSPPPD to e Make it work on the DOS Stamp e Make it capable of automatic unattended operation DSPPPD is the modified version of DOSPPPD that runs on the DOS Stamp DSPPPD uses the packet driver software interface which is an open standard for interfacing network applications with network physical layer drivers The packet driver specification is available for free download from http www crynwr com DSPPPD differs from most packet drivers in two important ways 1 Packet drivers typically perform only physical layer functions DSPPPD also performs the higher layer functions of PPP LCP authentication IPCP etc 2 Packet drivers typically are solely responsible for managing the physical layer With DSPPPD management of the physical layer is shared with your application The only part of the physical layer that DSPPPD manages is the UART DSPPPD initializes the UART and receives interrupts from it The remainder of the physical layer is managed by your application This division of labor makes it easy to accommodate different physical layer hardware such as POTS modems radio modems direct serial connections etc DSPPPD initializes the UART with a configuration specified on the DOS command line when DSPPPD is invoked DSPPPD then goes
15. Send CTS A signal from the other machine or our Input Not used by DIME modem that says it is OK for this machine to transmit Loop back to modem s RTS output Request To Send RTS A signal to the other machine or our modem that says it is OK for it to transmit to this machine Not used by DIME Loop back to modem s CTS input Data Terminal Ready DTR A signal to the other machine or our modem that says this machine is ready Output Not used by DIME Data Set Ready DSR A signal from the other machine or our modem that says it is ready Input Not used by DIME Data Carrier Detect DCD A signal from the modem that says it has detected a valid carrier Input Positive voltage when a carrier is present Negative voltage when no carrier is present Ring Indicator RI A signal from the modem that pulses each time phone line rings Input Not used by DIME Ground GND Ground reference for all other signals Connect DOS Stamp GND to modem GND Table 1 RS 232 Serial Port Signals Notice that the DOS Stamp COM2 port supports only two signals TX and RX To use a modem we need a third signal DCD To accommodate DCD use one of the DOS Stamp s general purpose I O pins PIO to monitor DCD See the connection diagram in Figure 1 DOS Stamp DIME Modem Connection To Modem Hi To DOS STAMP gt JP2 20 5 AA Figure 1 DOS Stamp DIME
16. ation need to run concurrently with the protocol layers and sockets The RTOS allows us to schedule the CPU s time between the protocol layers sockets and the application The RTOS does not replace DOS but adds the ability to multi task to your application The RTOS does not add the ability to multi task to DOS itself so there still can be only one DOS using task running concurrently DIME uses the uC OS II real time kernel as the RTOS The uC OS II kernel is pre emptive This means the kernel can interrupt a task in progress Cooperative kernels cannot do this and are not appropriate for our needs For more information on uC OS II refer to the book uC OS II The Real Time Kernel by Jean J Labrosse The Application Layer Above the TCP and UDP transport layers there is the application layer This is where data originates and is consumed In this context application does not necessarily refer to the entire application program but to the parts of it that interact with the network layer So even though DIME is an application it actually has two network applications built into it a HTTP web server and a SMTP e mail sender A network application uses sockets to send and or receive data over the TCP or UDP transport layer Hyper Text Transport Protocol HTTP A Network Application RFC 1945 Hyper Text Transport Protocol HTTP is a network application HTTP uses the TCP IP transport layer to send and receive web pages
17. ct files if you are using a linker that is compatible with Turbo C v3 0 If you experience linker problems you should recompile all of the source files except os_cpu_a obj This is an 80186 assembly language source file that has been assembled with TASM v3 0 If you are using Turbo C v3 0 or Borland C v3 x you may use the makefile DIMEDEMO MAK to automate this step You may also use the project file DIMEDEMO PRJ to compile and link within the IDE When compiling make sure the compiler options are set for Large memory model No floating point math Standard stack frame No debugging information Generate 80186 instructions 4 Upload these files to the DOS Stamp e DIMEDEMO EXE e DIMEDEMO HTM e DSPPPD EXE 5 On the DOS Stamp create the WATTCP CFG configuration file as described in the Sockets section At the DOS prompt type copy con wattcp cfg Type each line of the WATTCP configuration information When finished type lt CTRL Z gt to close the file Now test your DIMEDEMO setup 6 Follow step 1 in the How to Establish a PPP Link section 7 When the DOS prompt returns run DIMEDEMO The modem should dial within a few seconds After the modems negotiate the PPP exchange will occur If the ISP hangs up it is probably because the username or password is incorrect If the ISP does not hang up your connection is successful After about 60 seconds press the space bar to refresh the terminal window You shou
18. d If you have chosen a static IP address for your ISP account the IP address will be the same each time Asynchronous High level Data Link Control AHDLC Framing Part of PPP RFC 1662 When data is transmitted over asynchronous serial links using PPP it is transmitted in AHDLC frames A frame consists of a start flag also known as a frame delimiter an address field a control field a variable length of data and a CRC The start flag is a single byte 7E The address field is a single byte FF The control field is a single byte 03 Note that since 7E is used as the start flag it cannot be used elsewhere in the AHDLC frame Any bytes in the data or CRC with the value 7E must be converted to an equivalent two byte representation This is known as escaping The escaped representation is the value is 7D followed by another byte that is the original value exclusive OR ed with the value 20 This is illustrated below 12 3 4 5 TE FF 03 68 32 90 7E 34 61 original byte sequence where 1 is the start flag 2 is the address field 3 is the control field and 4 is where the data begins The first data byte that requires escaping is 5 but we are also going to escape 3 for reasons explained later After escaping the byte sequence becomes 1 2 3 4 5 7E FF 7D 23 68 32 90 7D 5E 34 61 resulting byte sequence Notice how the 03 control field was changed into the two byte sequence of 7D 23 and the 7E data was changed into th
19. e two byte sequence of 7D 5E The resulting byte sequence is what actually gets transmitted over the link The 7D value is the escape flag and tells the peer that the next byte is to be converted back to its original value The peer does this by discarding the 7D value and exclusive OR ing the escaped value with the value 20 Of course the escape flag 7D also cannot appear elsewhere in the data stream So to send a value of 7D a two byte sequence of 7D 5D is sent It is also possible to escape other values By default PPP escapes the values 00 1F 7D and 7E This explains why we escaped the value 03 in the example above Values in the 00 1F range can be individually escaped or not escaped by setting the Asynchronous Control Character Map ACCM parameter ACCM is negotiated with the peer during LCP It is desirable to escape as few values as the peer will support because sending escape flags lowers the effective throughput of the link DIME_manual doc Page 7 DIME v1 0b User s Manual The CRC is calculated based on the original un escaped values beginning with the address field The resulting CRC is itself escaped if necessary When data passes over the link it has been subjected to encapsulation at each layer For example data sent by TCP IP over a PPP link looks like this AHDLC frame start flag PPP AHDLC address field PPP AHDLC control field PPP IP header IP TCP header TCP Actual data Application i e
20. eer This is done transparently All your application has to do is make a BOOTP request after the physical layer is up and PPP is in the Network phase refer to ConnectionTask in DIMDEMO C case CONN_NETWORK_WAIT wait for packet driver s network phase DIME_manual doc Page 10 DIME v1 0b User s Manual if asyfuncs asyf_phase PHASE_NETWORK opencomn lt omitted for clarity gt _dobootp lt omitted for clarity gt At this point we should be able to use the connection connstate CONN_YES else lt omitted for clarity gt break In the WATTCP CFG file specify IP address assignment via BOOTP like this my_ip bootp Dynamic Host Configuration Protocol DHCP RFC 1112 1122 1123 1534 Dynamic Host Configuration Protocol DHCP is another way for a machine to find out what IP address it should use When a machine sends a DHCP request a DHCP server hands out an IP address to the machine DHCP works like BOOTP with one important difference DHCP leases IP address for a limited time When the lease expires the IP address is taken back by the DHCP server and another request must be made to obtain a new one When using PPP the DHCP request is actually serviced by the PPP packet driver it is not sent to the peer Although WATTCP supports DHCP it has not yet been tested with DSPPPD In the WATTCP CFG file specify IP address assignment via DHCP like this my_ip dhcp There is no compelling
21. he program if the WV represents something that is affected by real world events or program control OSSemPend WVSem 0 amp semerr wait for exclusive access to WV s sprintf wvSomeWebVar wvval 081X BiosTicks update WV OSSemPost WVSem release WV s If the WV value is to be changed by the web user with a browser the program must poll the WV to see if it has changed and act accordingly OSSemPend WVSem 0 amp semerr wait for exclusive access to WV s if wvSomeWebVar webchanged lt act accordingly gt wvSomeWebVar webchanged 0 reset flag DIME_manual doc Page 16 DIME v1 0b User s Manual OSSemPost WVSem release WV s It is essential to guarantee exclusive access to the WV Otherwise a higher priority task that uses the WV could interfere with a WV update already in progress in an interrupted lower priority task 3 Create a HTML web page with references to the WV name embedded in text forms or controls lt p gt The current BIOS tick counter shows amp nbsp lt b gt SomeWebVar lt b gt h amp nbsp ticks since midnight lt p gt This is a fragment of the DIMEDEMO HTM web page The WV is delimited by HH The HP is a signal to the DIME HTTP server that a WV value is to be substituted in place of everything between the delimiters The HTTP server is case sensitive and white space sensitive Be sure to use the exact WV name here with no leading or trailing space
22. lable to carry them Link Configuration Protocol LCP Part of PPP Link Configuration Protocol LCP is a way for peers to automatically agree upon many characteristics of the link such as packet size limits which authentication protocol to use header field compression etc LCP can also detect if the serial link is looped back echoing the data it receives If you are having trouble establishing a PPP link but the modem is dialing correctly it could be due to an unsupported LCP option required by one peer but not supported by the other LCP can usually establish a compatible link with almost any peer In difficult situations the resulting link is the lowest common denominator that each peer can accommodate In impossible situations the link will terminate after futile negotiations During LCP the authentication protocol is negotiated If the peers cannot agree on the same authentication protocol or if the username and password are not valid the link will be terminated immediately Password Authentication Protocol PAP Part of LCP Password Authentication Protocol PAP is the most commonly used PPP authentication protocol PAP is a simple protocol that sends the username and password to the peer in clear text not encrypted At this time DIME supports only PAP Challenge Handshake Authentication Protocol CHAP Part of LCP Challenge Handshake Authentication Protocol CHAP is an authentication protocol that sends
23. ld then see the IP address assigned to this link session If the IP address is still 0 0 0 0 something went wrong with the PPP exchange resulting in no IP address assignment from the ISP You may now try to bring the DIMEDEMO HTM web page up in your browser by typing DIME_manual doc Page 13 DIME v1 0b User s Manual http lt assigned IP address gt dimedemo htm into your browser If it does not work check to make sure you have entered the IP address and the filename correctly Refresh the web page to see the BIOS tick counter advance You may now try interacting with the web page buttons and fields Try sending an e mail message to yourself Try changing the state of GPIO pins by clicking on the GPIO buttons Real Time Operating System RTOS A Real Time Operating System RTOS is software that enables multiple tasks to run concurrently on a single CPU Each task is given a time slice on the CPU When that time slice is up the RTOS 1 Interrupts the current task 2 Saves the state of the current task 3 Decides which task to run next 4 Loads the state of the next task 5 Makes the next task into the current task 6 Runs the current task The DOS Stamp already has DOS on it so why do we need an RTOS Because DIME must run multiple tasks concurrently but DOS can only run one task at a time The protocol layers and sockets must be serviced in a regular timely manner However the non Internet parts of your embedded applic
24. modem to the DOS Stamp according to the diagram shown in the Serial Ports RS 232 and Modem Interfacing section 2 Edit DIMEDEMO C to make the following changes in boldface In ConnectionTask send modem dialing commands WriteSerialString ATZ r n OSTimeDIyHMSM 0 0 1 0 wait 1 s for modem to reset put your ISP s phone number in here and any special modem commands WriteSerialString AT amp KDT5551234 r n In SmtpTask Substitute the name of your SMTP e mail server here hostip resolve mail yourisp com get IP address of SMTP server lt omitted for clarity gt if hostip if we got the IP address OK send the message lt omitted for clarity gt Substitute the name of your e mail account here merr SMTP_SendHello pSmtpSock tmpmsg sizeof tmpmsg somebody someplace com lt omitted for clarity gt Substitute the name of your e mail account here merr SMTP_SendMailFrom pSmtpSock tmpmsg sizeof tmpmsg somebody someplace com lt omitted for clarity gt Substitute the name of your e mail account here merr SMTP_SendMsgHeader pSmtpSock now DIME_manual doc Page 12 DIME v1 0b User s Manual somebody someplace com wvMailRecipient wvval wvMailSubject wvval 3 Compile DIMEDEMO C and link it with these object files dosstamp obj os_cpu_a obj os_cpu_c obj ucos_ii obj ds obj dimedemo obj There is no need to recompile the source files for these obje
25. nk can be in Dead Establish Authenticate Network Terminate The Dead phase is when the PPP link is not ready to send or receive data This is usually because the physical layer does not have a connection established Once the physical layer has established a connection it must signal the upper layer LCP that the physical layer is up ready When LCP receives a signal that the lower layer physical is up the PPP link moves to the Establish phase DIME_manual doc Page 5 DIME v1 0b User s Manual The Establish phase is when the PPP link negotiates with the peer over the configuration of the link Both machines must come to an agreement on the configuration before the link can do useful work These negotiations are complex and are carried out by the Link Configuration Protocol LCP which runs during the Establish phase If LCP negotiation is successful LCP moves the PPP link into the Authenticate phase by signaling the upper layer authentication protocol that LCP is open If both machines cannot agree on the configuration LCP moves the PPP link back to the Dead phase by signaling the lower layer physical with a command to break the connection The Authenticate phase is when the PPP link authenticates itself to the peer using the chosen authentication protocol This is how the username and password are transmitted to the peer If the username and password are valid authentication moves the PPP link to the Network phase by signaling
26. ns These are documented in the WATTCP manual What is a socket The definition varies slightly depending on whom you ask but for our purposes WATTCP defines a socket as a combination of the transport type IP address and port number This completely describes which IP format the packets are in who the packets are coming from or going to and for which application the packets are intended DIME uses a slightly modified version of WATTCP which is a public domain open source sockets library created by Erick Engelke WATTCP must be configured before you can use sockets in your application WATTCP reads the configuration file WATTCP CFG when the sock_init function is called In the WATTCP CFG file you must specify these options IP address assignment method BOOTP DHCP or a static IP address Gateway IP address Net mask Name server IP address at least one There are other options for WATTCP but they are usually not necessary Boot Protocol BOOTP RFC 951 Boot Protocol BOOTP is a way for a machine to find out what IP address it should use When a machine sends a BOOTP request a BOOTP server hands out an IP address to the machine When using PPP the BOOTP request is actually serviced by the PPP packet driver it is not sent to the peer This works because PPP has already established the link and has been assigned an IP address by the peer Since PPP already has the IP address there is no need to pass the BOOTP request on to the p
27. request to the HTTP server GET dimedemo htm WhatsUpDoc HTTP 1 0 The query string is everything after the up to the separating space before the HTTP 1 0 What the server does with the query string depends on how the server is designed POST Method Part of HTTP RFC 1945 The POST method is used to send information to a HTTP server Browsers frequently use the POST method to send the results of on screen forms to a server When you buy something from a website you are asked to enter your name address credit card information etc into an on screen form When you click on the send or submit button the contents of the form are sent to the server For example the browser might send this request to the HTTP server POST someinfo txt HTTP 1 0 lt the info follows gt What the server does with the information depends on how the server is designed Typically the information is appended to an existing file of the same name If the file does not already exist it is created Some servers use the information to invoke CGI scripts which cause other actions to occur on the server The DIME web server does not support CGI but does create files and appends to existing files Either GET or POST methods can be used to send information to a HTTP server Which method is better For interactive web pages you will probably want to use the GET method Web pages can be easily designed with forms that send form data using the GET
28. ress will do just fine The only reason you might need a static IP address is if your embedded Internet application will be referred to by a hyperlink from some other web page The DIME demo web page has a static IP address since it is referred to by a hyperlink from the Bagotronix website DIME_manual doc Page 3 DIME v1 0b User s Manual TCP IP Protocol TCP IP is an acronym for Transmission Control Protocol Internet Protocol The IP indicates that it uses the IP addressing scheme It is frequently referred to as just TCP The software that provides TCP IP capability is frequently referred to as a TCP IP stack It is not a stack in the same sense of a CPU stack but instead refers to the layering of protocols looking like a stack TCP IP is a method for reliably transmitting data across IP networks When you bring up a website in your browser the browser sends data to the website and the website sends data back to your browser This exchange of data is done using TCP IP The data is placed in packets with information about the data This information goes into the packet header The actual data is placed after the header Why go through all this trouble making packets out of the data Why not just send the data and be done with it The routers on the Internet need to know where to send the data and who sent it They also need to know when it was sent in what order and if the data is still valid or has been corrupted
29. resses That may seem like a lot but as the Internet continues to grow the supply of unused IP addresses shrinks Eventually it may be necessary to use a new type of IP address with more digits But for now we can use 32 bit IP addresses if we are not wasteful How do we conserve the limited supply of IP addresses How are we assigned an available IP address The single answer to both questions dynamic IP addressing The concept of dynamic IP addressing is that when your machine establishes a connection to the ISP it is assigned an available IP address for the duration of the connection When the connection is terminated that IP address is taken back by the ISP and handed out to another machine as needed Your ISP reserves a block of IP addresses and hands them out as its customer s call in For example XYZ ISP Inc reserved 256 IP addresses but has 2000 customers As customers call in each one is assigned an IP address taken from the pool of reserved IP addresses When a customer s connection is terminated the IP address is taken back and returned to the pool of reserved IP addresses In this way 2000 Internet users can get by with only 256 IP addresses What if we don t like dynamic IP addressing What if we want to reserve our own IP address Then ask your ISP to rent you a static IP address You will pay more for a static IP address and it is usually not necessary For most embedded Internet applications a dynamic IP add
30. rk Application RFC 821 Simple Mail Transfer Protocol SMTP is a network application SMTP uses the TCP IP transport layer to send but not receive e mail messages A SMTP sender sends e mail on TCP Port 25 When you use an e mail program on your PC to send e mail it talks to the ISP s SMTP server When you use the e mail program on your PC to receive e mail it talks to the ISP s POP3 server not their SMTP server This is an important distinction When you set up DIME to talk to your ISP make sure you specify the ISP s SMTP server not its POP3 server DIME_manual doc Page 17 DIME v1 0b User s Manual The DIME SMTP sender implements the following SMTP commands HELO Hello MAIL RCPT Recipient DATA QUIT These commands must be sent in this order Between the DATA and QUIT commands you send the actual e mail message See SmtpTask in DIMEDEMO C for an example Recommended for Further Reading Sams Teach Yourself TCP IP in 24 Hours Joe Casad and Bob Willsey First Printing 1998 Sams Publishing ISBN 0 672 31248 4 uC OS II The Real Time Kernel Jean J Labrosse 1999 R amp D Books ISBN 0 87930 543 6 Illustrated TCP IP Matthew Naugle 1999 John Wiley amp Sons Inc ISBN 0 471 19656 8 PPP Design and Debugging James Carlson 1998 Addison Wesley Longman ISBN 0 201 18539 3 DIME_manual doc Page 18
31. s Your application is responsible for converting WV value text from to suitable data types There are several suitable C library functions for this purpose i e atoi sprintf After data type conversion your application must also validate the data to ensure it is within reasonable limits DIME HTTP Server Implementation The DIME HTTP server implements a subset of the HTTP 1 0 standard and supports GET method POST method with file creation and append Query strings HTTP 0 9 simple requests Proper content length for fixed length information responses i e static web pages The DIME HTTP server supports these features that are not part of the HTTP 1 0 standard e Dynamic HTML pages e Web variables The DIME HTTP server does not support e Time and date stamps e HEAD method e Content length for variable length information responses content length is omitted e CGI More than one instance of the DIME HTTP server can run concurrently This allows several users to access a web page on the same DOS Stamp simultaneously Browsers often make multiple requests to the same server concurrently For example a web page containing text and a graphic can cause a browser to make two simultaneous requests one for the text and one for the graphic If there is no available server socket the browser may not wait for one and put the broken graphic icon where the graphic should be Simple Mail Transfer Protocol SMTP A Netwo
32. s for modem to reset put your ISP s phone number in here and any special modem commands WriteSerialString AT amp KDT5761242 r n tick time OSTimeGet begin our carrier wait connstate CONN_CONNECT_WAIT break 4 Wait for the modem s DCD signal then signal DSPPPD that the physical layer is up in ConnectionTask case CONN_CONNECT_WATT wait for CD if PIOget PIO_CD amp amp openconn OSTimeDlyHMSM O 0 0 250 delay 250 ms so modem is really ready signal packet driver that lower layer is up asyfuncs asyf_lowerup ticktime OSTimeGet begin our network wait connstate CONN_NETWORK_WAIT DIME_manual doc Page 9 DIME v1 0b User s Manual else if OSTimeGet gt ticktime MODEM_WAIT OS_TICKS_PER_SEC Il openconn connstate CONN_CLOSING break 5 Wait for DSPPPD to move to the Network phase in ConnectionTask case CONN_NETWORK_WAIT wait for packet driver s network phase if asyfuncs asyf_phase PHASE_NETWORK amp amp openconn lt omitted for clarity gt At this point we should be able to use the connection connstate CONN_YES break At this point the PPP link is established and is available for use by the sockets network layers and your application Sockets Before you can use TCP or UDP to transfer data you need to open a socket or listen to a socket This is done with the tcp_open udp_open or tcp_listen functio
33. s tedious The easiest way to write HTML web pages is to use HTML editing software Static vs Dynamic Web Pages A static web page is one whose content does not change A dynamic web page i e the DIME demo is one whose content may be updated in real time due to real world events user input or program control Static web pages are not adequate for embedded Internet applications because they cannot be updated in real time To support dynamic web pages a mechanism for selectively updating the HTML content in real time is needed This may be done by embedding in the HTML references to variables that exist in the application program These variables are known as Web Variables Web Variables WV Web Variables WV are actual variables that reside in your application program Each WV is given a unique name a text string and can be assigned a value a text string Text is the only data type supported for the WV value Using a WV is a three step process 1 Create a WV like this static WV wvSomeWebVar WVCreate amp wvSomeWebVar SomeWebVar 10 SomeValue TRUE where the text name of the WV is SomeWebVar 10 bytes are allocated to store the text value the initial value of the WV is SomeValue and the initial value of the webchanged flag is TRUE The text name of a WV does not have to be the same as its C declaration name However similar names do make the source code easier to follow 2 Update the WV value inside t
34. talking about IP TCP is an upper layer because it is built on IP A layer is the lower layer 1f it underlies the layer you are talking about For example if you are talking about TCP IP is the lower layer The actual hardware and low level software drivers that interact with the hardware are called the physical layer and is the lowest layer of all The UART modem cables phone lines and physical layer software are all part of the physical layer as far as the protocols are concerned When someone says the physical layer is down it means that any of the following could be true The modem is not ready The modem is not connected to the remote modem the carrier was lost or was never established The UART is not working or has not been properly initialized The cables are not connected or improperly connected The physical layer software is low level software that initializes configures and directly interacts with the hardware For example software that sends an AT dialing command to the modem looks for the response and signals the upper layer the protocols that the hardware is ready to use Point To Point Protocol PPP RFC 1661 Point To Point Protocol PPP is actually a name for a suite of protocols It does not have an IP after it because it is not part of the Internet Protocol PPP is an additional layer of encapsulation that facilitates transmission of data over error prone serial data links If your
35. the username and password in encrypted form to the peer At this time DIME does not support CHAP DIME_manual doc Page 6 DIME v1 0b User s Manual Network Control Protocol NCP Part of PPP Network Control Protocol NCP is actually not a protocol NCP is a generic name given to any control protocol that configures PPP links for use with a network protocol For example if we want to use IP on the PPP link we need to configure the link using an NCP called IPCP If we want to use IPX on the PPP link we need to configure the link using an NCP called IPXCP IPCP and IPXCP are both classified as NCPs There are many other network protocols each one with its own NCP It is possible to use more than one network protocol concurrently on the same PPP link Since we are only using IP on the link we are only interested in IPCP and ignore all other NCPs Internet Protocol Control Protocol IPCP Part of PPP RFC 1332 Internet Protocol Control Protocol IPCP is used to configure the PPP link for IP IPCP negotiates options with the peer just like LCP does There are few options for IPCP Van Jacobson VJ compression is one of the options and is supported by DIME Even if the peer does not support VJ DIME can still fall back to non VJ operation through negotiation During IPCP the IP address is assigned by the peer In most cases the IP address is dynamically assigned This means it will be different each time the link is establishe
36. ubnet to only those machines on that subnet Subnetting is not really necessary for point to point links but the subnet mask mechanism is built into WATTCP so it must be configured properly In the WATTCP CFG file you can just use this netmask 255 255 255 0 Name Server IP Address A good definition for name server also known as Domain Name Server DNS is a machine that turns a domain name into an IP address When you use a browser and type in a domain name i e www bagotronix com the browser sends a DNS request to the name server The name server looks up the actual IP address of that domain and sends it to the browser This is how the browser finds out which IP address to send a web page request to You will probably want to use DNS services in your embedded Internet application Therefore WATTCP must know the name server IP address In the WATTCP CFG file specify the name server IP address like this nameserver 192 168 34 1 substitute the name server s actual IP address You may specify more than one name server assignment each on a separate line Itis a good idea to have at least two since any given name server may be unavailable at certain times How to Get DIMEDEMO C Up and Running You should get DIMEDEMO running as is before you start making changes for your application At this point you have learned enough concepts to configure DIMEDEMO and get it up and running Starting from the beginning 1 Connect the
Download Pdf Manuals
Related Search