ozymandias

I've had a Makerbot Cupcake CNC for a while now and have been wanting to convert it from the current arduino-bobino-fee-fi-sanguino controller to lightweight and modular MSP430 subsystems that would be useful in other control and robotics projects.
 
currently i have beta software that takes in UART data from a computer (i'm using GTKterm in linux) to the launchpad, where a G2231 controls chip select bits and resends the byte to (currently 3) other 2231's.
 
What i'm planning to do soon is modify the Makerbot to "close the loop" as far as control. Currently, you only theorectically know what state (where each bit is) by knowing what you've told it to do already; the "open loop" control (usually bad).
 
I've put up beta code in the code vault and i should :!: have proof of concept up and running by the end of the week.

I've had a Makerbot Cupcake CNC for a while now and have been wanting to convert it from the current arduino-bobino-fee-fi-sanguino controller to lightweight and modular MSP430 subsystems that would be useful in other control and robotics projects.
 
currently i have beta software that takes in UART data from a computer (i'm using GTKterm in linux) to the launchpad, where a G2231 controls chip select bits and resends the byte to (currently 3) other 2231's.
 
What i'm planning to do soon is modify the Makerbot to "close the loop" as far as control. Currently, you only theorectically know what state (where each bit is) by knowing what you've told it to do already; the "open loop" control (usually bad).
 
I've put up beta code in the code vault and i should :!: have proof of concept up and running by the end of the week.

Project: control 3 MSP430G2231's using a single G2231 as a bridge/converter/demultiplexer that recieves data through the USB
 
the bridge takes bytes and puts alternately sends them to recievers 1-3.
 
the recievers' P1.0 and P1.6 go to stepper motor controllers to step on rising edge and control direction, respectively.
 
"bridge" part:
-recieves bytes from 2400 8N1 input to P1.1 (requires launchpad 'RXD' jumper to be removed and the pin on the FTDI side connected to P1.1)
 
-after reception, transmits each byte out of P1.6 as MOSI and P1.5 as SCLK.
 
-cycles through bringing one of three output pins low as a chip select
 

#include #include // Conditions for 2400 Baud SW UART, ACLK = 32768 #define Bitime_5 0x06 // ~ 0.5 bit length + small adjustment #define Bitime 0x0E // 427us bit length ~ 2341 baud #define RXD 0x02 #define TXD 0x20 #define LED1 0x01 #define LED2 0x40 unsigned char RXTXData; unsigned char BitCnt; unsigned char temp; static void __inline__ delay(register unsigned int n); int main (void){ WDTCTL = WDTPW + WDTHOLD; // Stop watchdog timer CCTL0 = OUT; // TXD Idle as Mark TACTL = TASSEL_1 + MC_2; // ACLK, continuous mode P1SEL = TXD + RXD; // P1DIR = TXD + LED1 + BIT4 + BIT3; P1OUT = LED1 + BIT4 + BIT3; USICTL0 = USIPE7 | USIPE6 | USIPE5 | USIMST | USIOE | USISWRST; // Port, SPI Master USICTL1 = USICKPH | USIIE; USICKCTL = 0xE4;// /128 aclk? temp = 0; while(1){ P1OUT |= LED1 + BIT4 + BIT3; BitCnt = 0x8; // Load Bit counter CCTL0 = SCS + OUTMOD0 + CM1 + CAP + CCIE; // Sync, Neg Edge, Cap _BIS_SR(LPM3_bits + GIE); // Enter LPM3 w/ interr until char RXed if(temp == 2){ temp = 0; }else{ temp = temp +1; } } } // #pragma vector=TIMERA0_VECTOR interrupt(TIMERA0_VECTOR) Timer_A (void){ CCR0 += Bitime; // Add Offset to CCR0 if( CCTL0 & CAP ){ // Capture mode = start bit edge CCTL0 &= ~ CAP; // Switch from capture to compare mode CCR0 += Bitime_5; }else{ RXTXData = RXTXData >> 1; if (CCTL0 & SCCI) // Get bit waiting in receive latch RXTXData |= 0x80; BitCnt --; // All bits RXed? if ( BitCnt == 0){ USISRL = RXTXData; if(temp == 0){ P1OUT &= ~LED1; P1OUT |= BIT4 + BIT3; }else if(temp == 1){ P1OUT &= ~BIT4; P1OUT |= LED1 + BIT3; }else if(temp == 2){ P1OUT &= ~BIT3; P1OUT |= LED1 + BIT4; } USICNT |= 0x08; USICTL0 &= ~USISWRST; while (!(USIIFG & USICTL1)); USICTL0 |= USISWRST; delay(0); P1OUT |= LED1 + BIT4 + BIT3; CCTL0 &= ~ CCIE; // All bits RXed, disable interrupt _BIC_SR_IRQ(LPM3_bits + GIE); // Clear LPM3 bits from 0(SR) } } } static void __inline__ delay(register unsigned int n) { __asm__ __volatile__ ( "1: \n" " dec %[n] \n" " jne 1b \n" : [n] "+r"(n)); }
 
The reciever recieves the byte from the bridge and in state machine fashion puts the byte in:
-a counter for number of steps to take
-determines to set or clear a direction bit
-sets TACCR0 to control toggle rate
 

#include #include unsigned int bufff; unsigned char count; int main(void){ WDTCTL = WDTPW + WDTHOLD; // Stop watchdog timer P1OUT = 0x00; // P1.4 set, else reset P1DIR = 0x01 + BIT6; // P1.0 output, else input USICTL0 |= USIPE7 + USIPE5; // Port, SPI slave USICTL1 |= USIIE; // Counter interrupt, flag remains set USICTL0 &= ~USISWRST; // USI released for operation USICNT = 8; // init-load counter bufff = 0; count = 0; TACTL = TASSEL_1 | MC_1; //Set TimerA to use auxiliary clock in UP mode TACCTL0 = CCIE; //Enable the interrupt for TACCR0 match TACCR0 = 2; _BIS_SR(LPM0_bits + GIE); // Enter LPM0 w/ interrupt while(1){ } } // USI interrupt service routine interrupt(USI_VECTOR) universal_serial_interface(void){ if(count == 0){//bufff if(USISRL > 0x7A){ bufff = bufff + 0xFF00; }else{ bufff = bufff + USISRL; } }else if(count == 1){//dir if(USISRL > 0x3F){ P1OUT |= BIT6; }else{ P1OUT &= ~BIT6; } }else if(count == 2){//speed TACCR0 = USISRL; }else{ count = 0; } count = count + 1; if(count == 3){ count = 0; } USICNT = 8; // re-load counter } interrupt(TIMERA0_VECTOR) TIMERA0_ISR(void) { if( bufff > 0){ P1OUT ^= BIT0; bufff = bufff - 1; } }
 
PROBLEMS:
-the bridge can't pass data through it too fast. one must wait about ~200ms until the next byte is sent to prevent data loss. (the code isn't optimized and kinda "fatty" in all likelihood).
-since the TimerA assignments from USI data occur in the USI-ISR (and i haven't found a reliable way to do it otherwise), the data isn't 'pushed' into TimerA until the next byte is recieved.
-it would be cool to not have to modify the jumpers on the launchpad to do the UART, but its based on TI code that worked so i went with it.

I've had a Makerbot Cupcake CNC for a while now and have been wanting to convert it from the current arduino-bobino-fee-fi-sanguino controller to lightweight and modular MSP430 subsystems that would be useful in other control and robotics projects.
 
currently i have beta software that takes in UART data from a computer (i'm using GTKterm in linux) to the launchpad, where a G2231 controls chip select bits and resends the byte to (currently 3) other 2231's.
 
What i'm planning to do soon is modify the Makerbot to "close the loop" as far as control. Currently, you only theorectically know what state (where each bit is) by knowing what you've told it to do already; the "open loop" control (usually bad).
 
I've put up beta code in the code vault and i should :!: have proof of concept up and running by the end of the week.

I've had a Makerbot Cupcake CNC for a while now and have been wanting to convert it from the current arduino-bobino-fee-fi-sanguino controller to lightweight and modular MSP430 subsystems that would be useful in other control and robotics projects.
 
currently i have beta software that takes in UART data from a computer (i'm using GTKterm in linux) to the launchpad, where a G2231 controls chip select bits and resends the byte to (currently 3) other 2231's.
 
What i'm planning to do soon is modify the Makerbot to "close the loop" as far as control. Currently, you only theorectically know what state (where each bit is) by knowing what you've told it to do already; the "open loop" control (usually bad).
 
I've put up beta code in the code vault and i should :!: have proof of concept up and running by the end of the week.

Project: control 3 MSP430G2231's using a single G2231 as a bridge/converter/demultiplexer that recieves data through the USB
 
the bridge takes bytes and puts alternately sends them to recievers 1-3.
 
the recievers' P1.0 and P1.6 go to stepper motor controllers to step on rising edge and control direction, respectively.
 
"bridge" part:
-recieves bytes from 2400 8N1 input to P1.1 (requires launchpad 'RXD' jumper to be removed and the pin on the FTDI side connected to P1.1)
 
-after reception, transmits each byte out of P1.6 as MOSI and P1.5 as SCLK.
 
-cycles through bringing one of three output pins low as a chip select
 

#include #include // Conditions for 2400 Baud SW UART, ACLK = 32768 #define Bitime_5 0x06 // ~ 0.5 bit length + small adjustment #define Bitime 0x0E // 427us bit length ~ 2341 baud #define RXD 0x02 #define TXD 0x20 #define LED1 0x01 #define LED2 0x40 unsigned char RXTXData; unsigned char BitCnt; unsigned char temp; static void __inline__ delay(register unsigned int n); int main (void){ WDTCTL = WDTPW + WDTHOLD; // Stop watchdog timer CCTL0 = OUT; // TXD Idle as Mark TACTL = TASSEL_1 + MC_2; // ACLK, continuous mode P1SEL = TXD + RXD; // P1DIR = TXD + LED1 + BIT4 + BIT3; P1OUT = LED1 + BIT4 + BIT3; USICTL0 = USIPE7 | USIPE6 | USIPE5 | USIMST | USIOE | USISWRST; // Port, SPI Master USICTL1 = USICKPH | USIIE; USICKCTL = 0xE4;// /128 aclk? temp = 0; while(1){ P1OUT |= LED1 + BIT4 + BIT3; BitCnt = 0x8; // Load Bit counter CCTL0 = SCS + OUTMOD0 + CM1 + CAP + CCIE; // Sync, Neg Edge, Cap _BIS_SR(LPM3_bits + GIE); // Enter LPM3 w/ interr until char RXed if(temp == 2){ temp = 0; }else{ temp = temp +1; } } } // #pragma vector=TIMERA0_VECTOR interrupt(TIMERA0_VECTOR) Timer_A (void){ CCR0 += Bitime; // Add Offset to CCR0 if( CCTL0 & CAP ){ // Capture mode = start bit edge CCTL0 &= ~ CAP; // Switch from capture to compare mode CCR0 += Bitime_5; }else{ RXTXData = RXTXData >> 1; if (CCTL0 & SCCI) // Get bit waiting in receive latch RXTXData |= 0x80; BitCnt --; // All bits RXed? if ( BitCnt == 0){ USISRL = RXTXData; if(temp == 0){ P1OUT &= ~LED1; P1OUT |= BIT4 + BIT3; }else if(temp == 1){ P1OUT &= ~BIT4; P1OUT |= LED1 + BIT3; }else if(temp == 2){ P1OUT &= ~BIT3; P1OUT |= LED1 + BIT4; } USICNT |= 0x08; USICTL0 &= ~USISWRST; while (!(USIIFG & USICTL1)); USICTL0 |= USISWRST; delay(0); P1OUT |= LED1 + BIT4 + BIT3; CCTL0 &= ~ CCIE; // All bits RXed, disable interrupt _BIC_SR_IRQ(LPM3_bits + GIE); // Clear LPM3 bits from 0(SR) } } } static void __inline__ delay(register unsigned int n) { __asm__ __volatile__ ( "1: \n" " dec %[n] \n" " jne 1b \n" : [n] "+r"(n)); }
 
The reciever recieves the byte from the bridge and in state machine fashion puts the byte in:
-a counter for number of steps to take
-determines to set or clear a direction bit
-sets TACCR0 to control toggle rate
 

#include #include unsigned int bufff; unsigned char count; int main(void){ WDTCTL = WDTPW + WDTHOLD; // Stop watchdog timer P1OUT = 0x00; // P1.4 set, else reset P1DIR = 0x01 + BIT6; // P1.0 output, else input USICTL0 |= USIPE7 + USIPE5; // Port, SPI slave USICTL1 |= USIIE; // Counter interrupt, flag remains set USICTL0 &= ~USISWRST; // USI released for operation USICNT = 8; // init-load counter bufff = 0; count = 0; TACTL = TASSEL_1 | MC_1; //Set TimerA to use auxiliary clock in UP mode TACCTL0 = CCIE; //Enable the interrupt for TACCR0 match TACCR0 = 2; _BIS_SR(LPM0_bits + GIE); // Enter LPM0 w/ interrupt while(1){ } } // USI interrupt service routine interrupt(USI_VECTOR) universal_serial_interface(void){ if(count == 0){//bufff if(USISRL > 0x7A){ bufff = bufff + 0xFF00; }else{ bufff = bufff + USISRL; } }else if(count == 1){//dir if(USISRL > 0x3F){ P1OUT |= BIT6; }else{ P1OUT &= ~BIT6; } }else if(count == 2){//speed TACCR0 = USISRL; }else{ count = 0; } count = count + 1; if(count == 3){ count = 0; } USICNT = 8; // re-load counter } interrupt(TIMERA0_VECTOR) TIMERA0_ISR(void) { if( bufff > 0){ P1OUT ^= BIT0; bufff = bufff - 1; } }
 
PROBLEMS:
-the bridge can't pass data through it too fast. one must wait about ~200ms until the next byte is sent to prevent data loss. (the code isn't optimized and kinda "fatty" in all likelihood).
-since the TimerA assignments from USI data occur in the USI-ISR (and i haven't found a reliable way to do it otherwise), the data isn't 'pushed' into TimerA until the next byte is recieved.
-it would be cool to not have to modify the jumpers on the launchpad to do the UART, but its based on TI code that worked so i went with it.

Project: control 3 MSP430G2231's using a single G2231 as a bridge/converter/demultiplexer that recieves data through the USB
 
the bridge takes bytes and puts alternately sends them to recievers 1-3.
 
the recievers' P1.0 and P1.6 go to stepper motor controllers to step on rising edge and control direction, respectively.
 
"bridge" part:
-recieves bytes from 2400 8N1 input to P1.1 (requires launchpad 'RXD' jumper to be removed and the pin on the FTDI side connected to P1.1)
 
-after reception, transmits each byte out of P1.6 as MOSI and P1.5 as SCLK.
 
-cycles through bringing one of three output pins low as a chip select
 

#include #include // Conditions for 2400 Baud SW UART, ACLK = 32768 #define Bitime_5 0x06 // ~ 0.5 bit length + small adjustment #define Bitime 0x0E // 427us bit length ~ 2341 baud #define RXD 0x02 #define TXD 0x20 #define LED1 0x01 #define LED2 0x40 unsigned char RXTXData; unsigned char BitCnt; unsigned char temp; static void __inline__ delay(register unsigned int n); int main (void){ WDTCTL = WDTPW + WDTHOLD; // Stop watchdog timer CCTL0 = OUT; // TXD Idle as Mark TACTL = TASSEL_1 + MC_2; // ACLK, continuous mode P1SEL = TXD + RXD; // P1DIR = TXD + LED1 + BIT4 + BIT3; P1OUT = LED1 + BIT4 + BIT3; USICTL0 = USIPE7 | USIPE6 | USIPE5 | USIMST | USIOE | USISWRST; // Port, SPI Master USICTL1 = USICKPH | USIIE; USICKCTL = 0xE4;// /128 aclk? temp = 0; while(1){ P1OUT |= LED1 + BIT4 + BIT3; BitCnt = 0x8; // Load Bit counter CCTL0 = SCS + OUTMOD0 + CM1 + CAP + CCIE; // Sync, Neg Edge, Cap _BIS_SR(LPM3_bits + GIE); // Enter LPM3 w/ interr until char RXed if(temp == 2){ temp = 0; }else{ temp = temp +1; } } } // #pragma vector=TIMERA0_VECTOR interrupt(TIMERA0_VECTOR) Timer_A (void){ CCR0 += Bitime; // Add Offset to CCR0 if( CCTL0 & CAP ){ // Capture mode = start bit edge CCTL0 &= ~ CAP; // Switch from capture to compare mode CCR0 += Bitime_5; }else{ RXTXData = RXTXData >> 1; if (CCTL0 & SCCI) // Get bit waiting in receive latch RXTXData |= 0x80; BitCnt --; // All bits RXed? if ( BitCnt == 0){ USISRL = RXTXData; if(temp == 0){ P1OUT &= ~LED1; P1OUT |= BIT4 + BIT3; }else if(temp == 1){ P1OUT &= ~BIT4; P1OUT |= LED1 + BIT3; }else if(temp == 2){ P1OUT &= ~BIT3; P1OUT |= LED1 + BIT4; } USICNT |= 0x08; USICTL0 &= ~USISWRST; while (!(USIIFG & USICTL1)); USICTL0 |= USISWRST; delay(0); P1OUT |= LED1 + BIT4 + BIT3; CCTL0 &= ~ CCIE; // All bits RXed, disable interrupt _BIC_SR_IRQ(LPM3_bits + GIE); // Clear LPM3 bits from 0(SR) } } } static void __inline__ delay(register unsigned int n) { __asm__ __volatile__ ( "1: \n" " dec %[n] \n" " jne 1b \n" : [n] "+r"(n)); }
 
The reciever recieves the byte from the bridge and in state machine fashion puts the byte in:
-a counter for number of steps to take
-determines to set or clear a direction bit
-sets TACCR0 to control toggle rate
 

#include #include unsigned int bufff; unsigned char count; int main(void){ WDTCTL = WDTPW + WDTHOLD; // Stop watchdog timer P1OUT = 0x00; // P1.4 set, else reset P1DIR = 0x01 + BIT6; // P1.0 output, else input USICTL0 |= USIPE7 + USIPE5; // Port, SPI slave USICTL1 |= USIIE; // Counter interrupt, flag remains set USICTL0 &= ~USISWRST; // USI released for operation USICNT = 8; // init-load counter bufff = 0; count = 0; TACTL = TASSEL_1 | MC_1; //Set TimerA to use auxiliary clock in UP mode TACCTL0 = CCIE; //Enable the interrupt for TACCR0 match TACCR0 = 2; _BIS_SR(LPM0_bits + GIE); // Enter LPM0 w/ interrupt while(1){ } } // USI interrupt service routine interrupt(USI_VECTOR) universal_serial_interface(void){ if(count == 0){//bufff if(USISRL > 0x7A){ bufff = bufff + 0xFF00; }else{ bufff = bufff + USISRL; } }else if(count == 1){//dir if(USISRL > 0x3F){ P1OUT |= BIT6; }else{ P1OUT &= ~BIT6; } }else if(count == 2){//speed TACCR0 = USISRL; }else{ count = 0; } count = count + 1; if(count == 3){ count = 0; } USICNT = 8; // re-load counter } interrupt(TIMERA0_VECTOR) TIMERA0_ISR(void) { if( bufff > 0){ P1OUT ^= BIT0; bufff = bufff - 1; } }
 
PROBLEMS:
-the bridge can't pass data through it too fast. one must wait about ~200ms until the next byte is sent to prevent data loss. (the code isn't optimized and kinda "fatty" in all likelihood).
-since the TimerA assignments from USI data occur in the USI-ISR (and i haven't found a reliable way to do it otherwise), the data isn't 'pushed' into TimerA until the next byte is recieved.
-it would be cool to not have to modify the jumpers on the launchpad to do the UART, but its based on TI code that worked so i went with it.

Here is the example with LED1 and LED2 modification so you can see that it is actually working. Each character RX/TX will toggle the active LED back and forth.
 

//****************************************************************************** // MSP430G2xx1 Demo - Timer_A, Ultra-Low Pwr UART 2400 Echo, 32kHz ACLK // // Description: Use Timer_A CCR0 hardware output modes and SCCI data latch // to implement UART function @ 2400 baud. Software does not directly read and // write to RX and TX pins, instead proper use of output modes and SCCI data // latch are demonstrated. Use of these hardware features eliminates ISR // latency effects as hardware insures that output and input bit latching and // timing are perfectly synchronised with Timer_A regardless of other // software activity. In the Mainloop the UART function readies the UART to // receive one character and waits in LPM3 with all activity interrupt driven. // After a character has been received, the UART receive function forces exit // from LPM3 in the Mainloop which echo's back the received character. // ACLK = TACLK = LFXT1 = 32768Hz, MCLK = SMCLK = default DCO // //* An external watch crystal is required on XIN XOUT for ACLK *// // // MSP430G2xx1 // ----------------- // /|\| XIN|- // | | | 32kHz // --|RST XOUT|- // | | // | CCI0B/TXD/P1.5|--------> // | | 2400 8N1 // | CCI0A/RXD/P1.1|<-------- // #define RXD 0x02 // RXD on P1.1 #define TXD 0x20 // TXD on P1.5 #define LED1 0x01 #define LED2 0x40 // Conditions for 2400 Baud SW UART, ACLK = 32768 #define Bitime_5 0x06 // ~ 0.5 bit length + small adjustment #define Bitime 0x0E // 427us bit length ~ 2341 baud unsigned int RXTXData; unsigned char BitCnt; void TX_Byte (void); void RX_Ready (void); // D. Dang // Texas Instruments Inc. // October 2010 // Built with CCS Version 4.2.0 and IAR Embedded Workbench Version: 5.10 //****************************************************************************** #include int main (void) { WDTCTL = WDTPW + WDTHOLD; // Stop watchdog timer CCTL0 = OUT; // TXD Idle as Mark TACTL = TASSEL_1 + MC_2; // ACLK, continuous mode P1SEL = TXD + RXD; // P1DIR = TXD + LED1 + LED2; // // Mainloop for (; { RX_Ready(); // UART ready to RX one Byte P1OUT ^= LED1; _BIS_SR(LPM3_bits + GIE); // Enter LPM3 w/ interr until char RXed TX_Byte(); // TX Back RXed Byte Received P1OUT ^= LED2; } } // Function Transmits Character from RXTXData Buffer void TX_Byte (void) { BitCnt = 0xA; // Load Bit counter, 8data + ST/SP while (CCR0 != TAR) // Prevent async capture CCR0 = TAR; // Current state of TA counter CCR0 += Bitime; // Some time till first bit RXTXData |= 0x100; // Add mark stop bit to RXTXData RXTXData = RXTXData << 1; // Add space start bit CCTL0 = CCIS0 + OUTMOD0 + CCIE; // TXD = mark = idle while ( CCTL0 & CCIE ); // Wait for TX completion } // Function Readies UART to Receive Character into RXTXData Buffer void RX_Ready (void) { BitCnt = 0x8; // Load Bit counter CCTL0 = SCS + OUTMOD0 + CM1 + CAP + CCIE; // Sync, Neg Edge, Cap } // Timer A0 interrupt service routine #pragma vector=TIMERA0_VECTOR interrupt void Timer_A (void) { CCR0 += Bitime; // Add Offset to CCR0 // TX if (CCTL0 & CCIS0) // TX on CCI0B? { if ( BitCnt == 0) CCTL0 &= ~ CCIE; // All bits TXed, disable interrupt else { CCTL0 |= OUTMOD2; // TX Space if (RXTXData & 0x01) CCTL0 &= ~ OUTMOD2; // TX Mark RXTXData = RXTXData >> 1; BitCnt --; } } // RX else { if( CCTL0 & CAP ) // Capture mode = start bit edge { CCTL0 &= ~ CAP; // Switch from capture to compare mode CCR0 += Bitime_5; } else { RXTXData = RXTXData >> 1; if (CCTL0 & SCCI) // Get bit waiting in receive latch RXTXData |= 0x80; BitCnt --; // All bits RXed? if ( BitCnt == 0) //>>>>>>>>>> Decode of Received Byte Here <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< { CCTL0 &= ~ CCIE; // All bits RXed, disable interrupt _BIC_SR_IRQ(LPM3_bits); // Clear LPM3 bits from 0(SR) } //>>>>>>>>>> Decode of Received Byte Here <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< } } }

It would work if the size and offsets of the flash and RAM segments, and the addresses of all the peripherals you use, are the same between the MSP430G2231 and the MSP430F2012.
 
For any chip, you should simply use the lower-cased version of the MCU name. E.g., -mmcu=msp430g2231
 
While I decided to keep support for the legacy genericized MCU identifiers along with the preferred ones in LTS20110716 and its patches, those aliases were removed in the current development series, starting around 20111205.

msp430-gcc -mmcu=msp430fr5739 works for me just fine. This, naturally, will depend on whether or not
you have one of the newer versions of msp430-gcc.
 
It's available from the Fraunchpad webpage, http://www.ti.com/tool/msp-exp430fr5739
 
-Rusty-