Hey, Peter, thanks for that answer. I just tested it on my system the
snippet compiles just fine if using logf().

Now, sorry to derail the conversation, but I am still having issues with
the library I wrote.

Compiling the library:

[tron@HAL9000 libhd44780]$ make
Making release build...
msp430-gcc -Os -c hd44780.c -o hd44780.o -mmcu=msp430g2231
-DF_CPU=1000000UL -std=gnu99 -Wall
msp430-ar -r libhd44780-4bit.a hd44780.o
msp430-size hd44780.o
   text    data     bss     dec     hex filename
    572       0       2     574     23e hd44780.o
Done.

Compiling the core file:

[tron@HAL9000 hd-test]$ make
Making release build...
msp430-gcc -Os -c core.c -I../libhd44780 -mmcu=msp430g2231
-DF_CPU=1000000 -std=gnu99 -Wall
core.c:39:6: warning: return type of 'main' is not 'int' [-Wmain]
msp430-gcc -L../libhd44780 -lhd44780-4bit -mmcu=msp430g2231 core.o -o
hd-test.elf
core.o: In function `main':
core.c:(.init9+0x8): undefined reference to `hd_init'
core.c:(.init9+0x16): undefined reference to `putchar'
collect2: ld returned 1 exit status
make: *** [link] Error 1

Amusingly, if I link directly to the .o file generated when compiling
the library, it works ok:

[tron@HAL9000 hd-test]$ make debug
Making debug build...
msp430-gcc -g -c core.c -I../libhd44780 -mmcu=msp430g2231
-DF_CPU=1000000 -std=gnu99 -Wall
core.c:39:6: warning: return type of 'main' is not 'int' [-Wmain]
msp430-gcc -mmcu=msp430g2231 ../libhd44780/hd44780.o core.o -o
hd-test.elf
>>> FIRMWARE SIZE <<<
msp430-size hd-test.elf
   text    data     bss     dec     hex filename
    782      44       4     830     33e hd-test.elf


Is there anything I am missing?

On Sun, 2012-09-16 at 19:48 -0500, Peter Bigot wrote:
> There is no double (64-bit) support in the mspgcc toolchain, only
> float (32-bit), with double being equivalent.  While most math
> libraries have both float and double entrypoints (e.g., sin() is
> simply a wrapper around sinf()), it appears log() only exists as
> logf().  Please file this as a bug on the tracker, but it won't be
> high priority since it appears to have always been this way.
> 
> Since double isn't supported, it's probably best to not use it, rather
> than mislead the maintainer into thinking it's actually available.
> Use "float", and then use the corresponding functions like "logf".
> 
> Peter

Attachment: signature.asc
Description: This is a digitally signed message part

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Mspgcc-users mailing list
Mspgcc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mspgcc-users

Reply via email to