mspgcc-users — Discussion list for users

Re: [Mspgcc-users] gcc .-current

[Dmitry <di...@ei...>]

 
> I'm a little confused here - C does not normally support nested
> functions, does it?  Is this part of the C99 standard?  If so, then it's
> great that

C does not normally support nested functions, this is a gcc extension.

> msp430-gcc supports them.  However, are they actually any real use?  I
> have

Yes, very convenient mechanism for local reaction on an interrupt - you do
not have to allocate global vars and such. Any cleanup can be done on the
local stack.

> done many years programming in Pascal, a language that does support
> nested functions, and I have not found any need for nested functions -
> indeed, I have found them to be confusing and difficult to maintain, as
> well as being less efficient in use (in Pascal, anyway).  Obviously they
> don't do any harm - if I don't want to use them, I don't have to.  But
> I'm just trying to see if there are circumstances where code can really
> be written better by using them.  I suppose if you have optomisation
> turned on, then the compiler will probably in-line nested functions
> where possible, so if you like that style of coding, it will give you
> quite a powerful mechanism for splitting a function into seperate parts
> without any overhead (other than compiler effort!).
> 
> Why does the code get enlarged by 48 bytes?

Only if you export your local nested function outside the parents scope,
the code will be enlarged - 8 bytes on local stack is allocated for
function call (in order to call nested function from outside) . 48 bytes
are - 8 bytes trampoline template, 8 bytes on stack and 24 bytes of code,
which copies context and funct address to stack.

If you're using nested call within a parent's scope, this does not involve
any overhead (but only 1 insn/cycle - mov r1,r6 or mov r4,r6 before funct
call). If you declare it as inline - this will be (hopefully) inlined.

> 
> >
> > Anyway, this is a supported feature, which msp430-gcc has (avr does
> > not ;)
> >
> > 4. attribute 'reentrant' added. (Not in includes)
> >
> > This attribute forces gcc to issue eint/dint on function entry/exit
> > irrespective to the status register state.
> >
> 
> Is there any nice way to add an eint/dint to a function that *is*
> dependant on the status register state?  I very often need to be able to
> turn off interrupts, then restore them to the previous state - it is
> extremly useful for accessing hardware or data structures from either
> interrupt routines or background routines.
> 

Within ISR interrups are disabled unless you reenable it (with
enablenested attribute).

For non-IRS:
Yes, there is an attribute 'critical' (read doc.txt) which does as
follows:

funct:
	push r2	; save status
	dint	; disable ints
	...pushes
	.... funct body follows
	...pops
	reti	; restore r2 and return.
	

So,

type critical funct(argtype arg)
{
	type var;
	...
	return var;
}

will be reentrant under any circumstances.


cheers,
~d


> mvh.
> 
> David
> 
> 
> >
> >
> > Have fun
> > ~d
> >
> 
> 
> 
> 
> -------------------------------------------------------
> In remembrance
> www.osdn.com/911/
> _______________________________________________
> Mspgcc-users mailing list
> Msp...@li...
> https://lists.sourceforge.net/lists/listinfo/mspgcc-users
> 
> 


/********************************************************************
     ("`-''-/").___..--''"`-._     (\   Dimmy the Wild      UA1ACZ
      `6_ 6  )   `-.  (     ).`-.__.`)  Enterprise Information Sys 
      (_Y_.)'  ._   )  `._ `. ``-..-'   Nevsky prospekt,   20 / 44
    _..`--'_..-_/  /--'_.' ,'           Saint Petersburg,   Russia
   (il),-''  (li),'  ((!.-'             +7 (812)  3468202, 5585314
 ********************************************************************/

Re: [Mspgcc-users] gcc .-current

[David B. <da...@we...>]

> Fellows,
> I will probably stick on gcc-3.-current version of the compiler and will
> add new feateres and such to this branch. I checked the code generated
> here and find it reasonable. In many cases current versin generates better
> code than 3.0...
>
> So, new I committed new version of gcc with some improvements:
> 1. Improved movstr() insn. (movstr is a built-in compiler command)
> 2. Added trampolines (aka nested functions)
>
> Several words about it:
> Nested functions are function declared within a function declaration and
> can refer to local variables. This is usefull for some cleanups due to
> interrupted call or external alarm. Another advantage - almost everything
> can be done on stack. For example:
>
< Skip example >
>
> Here, all data transfers can be done on buffer allocated on stack =>
> no .data waste.
>
> Disadvantages:
>  - Every nested declaration enlarges code by ~48 bytes.
>  - Every nested call preceeds by 2 instructions:
> mov #SOMECONST, r6
> br #nestedfunct
>  - Something else :)


I'm a little confused here - C does not normally support nested functions,
does it?  Is this part of the C99 standard?  If so, then it's great that
msp430-gcc supports them.  However, are they actually any real use?  I have
done many years programming in Pascal, a language that does support nested
functions, and I have not found any need for nested functions - indeed, I
have found them to be confusing and difficult to maintain, as well as being
less efficient in use (in Pascal, anyway).  Obviously they don't do any
harm - if I don't want to use them, I don't have to.  But I'm just trying to
see if there are circumstances where code can really be written better by
using them.  I suppose if you have optomisation turned on, then the compiler
will probably in-line nested functions where possible, so if you like that
style of coding, it will give you quite a powerful mechanism for splitting a
function into seperate parts without any overhead (other than compiler
effort!).

Why does the code get enlarged by 48 bytes?

>
> Anyway, this is a supported feature, which msp430-gcc has (avr does not ;)
>
> 4. attribute 'reentrant' added. (Not in includes)
>
> This attribute forces gcc to issue eint/dint on function entry/exit
> irrespective to the status register state.
>

Is there any nice way to add an eint/dint to a function that *is* dependant
on the status register state?  I very often need to be able to turn off
interrupts, then restore them to the previous state - it is extremly useful
for accessing hardware or data structures from either interrupt routines or
background routines.

mvh.

David


>
>
> Have fun
> ~d
>

Re: Fw: [Mspgcc-users] running linux on msp?

[Matthias W. <ms...@pe...>]

On Tue, Sep 10, 2002 at 10:13:15AM +0400, Dmitry wrote:

> Fellows, if you want to have ucos-ii (www.ucos-ii.com) ported fro msp430,
> I can do it. Just let me know, cause I do not want to do it without
> demand.

Would be great - a port for the IAR C is already available and small in size:
http://www.ucos-ii.com/contents/products/ucos-ii/downloads/ti-msp430.zip
Maybe it can be used with Mspgcc too?

        Matthias

Re: Fw: [Mspgcc-users] running linux on msp?

[Garst R. R. <re...@is...>]

Martin Theobald wrote:
> 
> > >  I think "OS" is the wrong question. What are components you need?
> 
> I need Multitasking Support and standard libaries to develop software for
> different Processor architektures.
> 
> Aditionally I need Irda and Compact Flash Card access.
uc/OSII promises Compact Flash soon. I'm working on Irda and will now
concentrate my efforts there. When I get something solid I will post it.
Garst

RE: Fw: [Mspgcc-users] running linux on msp?

[Sylvain B. <sbe...@re...>]

There already is an MSP port of ucos-ii:
http://www.ucos-ii.com/contents/products/ucos-ii/ports-misc.html

Anybody, knows of a good open source TCP/IP stack for the MSP that would
easily integrate with ucos-ii?  thanks.

About uCLinux, I've run into a site about a port to a Hitachi 16-bit MCU
(H8S):
http://www.azpower.com/H8-uClinux/

Sylvain

-----Original Message-----
From: msp...@li...
[mailto:msp...@li...]On Behalf Of Garst R.
Reese
Sent: Tuesday, September 10, 2002 5:54 AM
To: Msp...@li...
Subject: Re: Fw: [Mspgcc-users] running linux on msp?



> Fellows, if you want to have ucos-ii (www.ucos-ii.com) ported fro msp430,
I can do it.
> Just let me know, cause I do not want to do it without demand.
>
They claim they will soon have drivers for MMC and CF cards, which use
FAT FS that they already have, so yes. This would be extremely useful
for me, and it sounds like Martin could use it also.
Thanks,
   Garst


-------------------------------------------------------
This sf.net email is sponsored by: OSDN - Tired of that same old
cell phone?  Get a new here for FREE!
https://www.inphonic.com/r.asp?r=sourceforge1&refcode1=vs3390
_______________________________________________
Mspgcc-users mailing list
Msp...@li...
https://lists.sourceforge.net/lists/listinfo/mspgcc-users

[Mspgcc-users] gcc .-current

[Dmitry <di...@ei...>]

Fellows,
I will probably stick on gcc-3.-current version of the compiler and will
add new feateres and such to this branch. I checked the code generated
here and find it reasonable. In many cases current versin generates better
code than 3.0...

So, new I committed new version of gcc with some improvements:
1. Improved movstr() insn. (movstr is a built-in compiler command)
2. Added trampolines (aka nested functions)

Several words about it:
Nested functions are function declared within a function declaration and
can refer to local variables. This is usefull for some cleanups due to
interrupted call or external alarm. Another advantage - almost everything
can be done on stack. For example:

void dummy_example(int circumstance)   
{
        int stop = 0, res;
        char buf[256];
        void st(int a)
        {
                stop = a;
                // device_read() interrupted
                // clean up input buffer
                cleanup(buf);
        }
	void st1()
	{
		stop = 1;
	}

        // register internal callback (for any event like IO error
        // or external interrupt)
        register_callback(c?st:st1, 5);

        while(stop == 0)
        {
                if((res = device_read (buf)) == -1)
                        st(1);	// break
                else if(res == 0)
                        // no data in buffer
                        break;
                else
                        // transfer data to user space
                        uiomove(buf, res);
        }

        unregister_callback(c?st:st1);
}

Here, all data transfers can be done on buffer allocated on stack => 
no .data waste.

Disadvantages:
 - Every nested declaration enlarges code by ~48 bytes.
 - Every nested call preceeds by 2 instructions:
	mov	#SOMECONST, r6
	br	#nestedfunct
 - Something else :)

Anyway, this is a supported feature, which msp430-gcc has (avr does not ;)

4. attribute 'reentrant' added. (Not in includes)

This attribute forces gcc to issue eint/dint on function entry/exit
irrespective to the status register state.



Have fun
~d



/********************************************************************
     ("`-''-/").___..--''"`-._     (\   Dimmy the Wild      UA1ACZ
      `6_ 6  )   `-.  (     ).`-.__.`)  Enterprise Information Sys 
      (_Y_.)'  ._   )  `._ `. ``-..-'   Nevsky prospekt,   20 / 44
    _..`--'_..-_/  /--'_.' ,'           Saint Petersburg,   Russia
   (il),-''  (li),'  ((!.-'             +7 (812)  3468202, 5585314
 ********************************************************************/

Re: Fw: [Mspgcc-users] running linux on msp?

[Garst R. R. <re...@is...>]

> Fellows, if you want to have ucos-ii (www.ucos-ii.com) ported fro msp430, I can do it.
> Just let me know, cause I do not want to do it without demand.
> 
They claim they will soon have drivers for MMC and CF cards, which use
FAT FS that they already have, so yes. This would be extremely useful
for me, and it sounds like Martin could use it also.
Thanks,
   Garst

Re: Fw: [Mspgcc-users] running linux on msp?

[Martin T. <m.t...@az...>]

> >  I think "OS" is the wrong question. What are components you need? 

I need Multitasking Support and standard libaries to develop software for
different Processor architektures.

Aditionally I need Irda and Compact Flash Card access.

Martin

Re: Fw: [Mspgcc-users] running linux on msp?

[Dmitry <di...@ei...>]

>  I think "OS" is the wrong question. What are components you need? I

You at least need memory manager and process scheduler (well, interprocess stack as well). If you need to interract with some external devices, you need device drivers. Everything else (including device drivers by the way) can be considered as user applications (qnx is example).

In Linux, from my experience (I give up with it a couple of years ago though), these things are huge and exhausting, therefore cannot be used with msp430 arch.

But if you need any drivers and 'inter-driver-interraction', this can be considered as a part of a real OS...
For example, uc/os-ii will be Ok here.

Fellows, if you want to have ucos-ii (www.ucos-ii.com) ported fro msp430, I can do it.
Just let me know, cause I do not want to do it without demand.

~d

> need an Irda stack and a compact file system for Multimedia cards.
> Working on both using the SPI bus connected to a MAX3100,
> MultiMediaCard, LCD driver ...
>   Garst

/********************************************************************
     ("`-''-/").___..--''"`-._     (\   Dimmy the Wild      UA1ACZ
      `6_ 6  )   `-.  (     ).`-.__.`)  Enterprise Information Sys 
      (_Y_.)'  ._   )  `._ `. ``-..-'   Nevsky prospekt,   20 / 44
    _..`--'_..-_/  /--'_.' ,'           Saint Petersburg,   Russia
   (il),-''  (li),'  ((!.-'             +7 (812)  3468202, 5585314
 ********************************************************************/

Re: Fw: [Mspgcc-users] running linux on msp?

[Garst R. R. <re...@is...>]

David Brown wrote:
 
> If you are thinking of projects like ucLinux, then they are all (to my
> knowledge - having never used ucLinux, but having looked longingly at it)
> based on 32-bit CPUs.  The most common targets are Coldfire and Arm, both of
> which are 32-bit.  Remember, Linux is not unix - it was originally dedicated
> to the 32-bit x86 with MMU, and has mostly been adapted up to more powerful
> architectures, not down to smaller ones.  Theoretically, it could be ported
> to suitably powerful 16-bit CPUs, but I don't think anyone has done so.
> 
Eons ago there were long discussions on the linux list about ports to a
286. The summary was "ain't gonna happen."
 I think "OS" is the wrong question. What are components you need? I
need an Irda stack and a compact file system for Multimedia cards.
Working on both using the SPI bus connected to a MAX3100,
MultiMediaCard, LCD driver ...
  Garst

Re: Fw: [Mspgcc-users] running linux on msp?

[David B. <da...@we...>]

> David Brown wrote:
>
> >I would think the short answer to that is it can't be done - far too much
in
> >Linux assumes that you've got at least a 32-bit CPU.  Other problems
include
> >memory (IIRC, the biggest msp430 has about 4K ram - Linux is not happy
with
> >less than about 4 MB), processor speed, etc.
> >
> >The long answer is why on earth do you want to run Linux on the msp430?
> >
> >
> It amazes me how many people waffle on about Linux or Embedded XP or
> some such OS getting a big slice of the embedded market. Anyone who ever
> developed embedded software knows No-OS will be the big winner. No-OS is
> compact, it runs on pretty much any CPU, and the licencing conditions
> are very favourable to both free and commercial development. :-)

I use No-OS on almost all my embedded systems, so I can vouch for its
benifits.

>
> I don't think Linux demands a 32-bit CPU, though. I think some of the
> microcontroller versions run on 16 bit hardware. Remember, Unix is a
> 16-bit OS that was only adapted to 32-bit CPUs well into its commercial
> life.
>

If you are thinking of projects like ucLinux, then they are all (to my
knowledge - having never used ucLinux, but having looked longingly at it)
based on 32-bit CPUs.  The most common targets are Coldfire and Arm, both of
which are 32-bit.  Remember, Linux is not unix - it was originally dedicated
to the 32-bit x86 with MMU, and has mostly been adapted up to more powerful
architectures, not down to smaller ones.  Theoretically, it could be ported
to suitably powerful 16-bit CPUs, but I don't think anyone has done so.

Re: Fw: [Mspgcc-users] running linux on msp?

[Steve U. <st...@co...>]

David Brown wrote:

>I would think the short answer to that is it can't be done - far too much in
>Linux assumes that you've got at least a 32-bit CPU.  Other problems include
>memory (IIRC, the biggest msp430 has about 4K ram - Linux is not happy with
>less than about 4 MB), processor speed, etc.
>
>The long answer is why on earth do you want to run Linux on the msp430?
>  
>
It amazes me how many people waffle on about Linux or Embedded XP or 
some such OS getting a big slice of the embedded market. Anyone who ever 
developed embedded software knows No-OS will be the big winner. No-OS is 
compact, it runs on pretty much any CPU, and the licencing conditions 
are very favourable to both free and commercial development. :-)

I don't think Linux demands a 32-bit CPU, though. I think some of the 
microcontroller versions run on 16 bit hardware. Remember, Unix is a 
16-bit OS that was only adapted to 32-bit CPUs well into its commercial 
life.

Regards,
Steve

Re: [Mspgcc-users] running linux on msp?

[Steve U. <st...@co...>]

Andreas Schwarz wrote:

>Monday, September 09, 2002, 2:08:11 PM, you wrote:
>
>  
>
>>has anyone of you tried to run embedded Linux on Msp 430?
>>Does anybody know some urls with informations about that?
>>    
>>
>
>You need at least 0.5MB ROM and 1 MB Ram -> no chance with the MSP430.
>  
>
But no embedded software ever fits within the available memory until you 
start tweaking :-)

Regards,
Steve

Fw: [Mspgcc-users] running linux on msp?

[David B. <da...@we...>]

I would think the short answer to that is it can't be done - far too much in
Linux assumes that you've got at least a 32-bit CPU.  Other problems include
memory (IIRC, the biggest msp430 has about 4K ram - Linux is not happy with
less than about 4 MB), processor speed, etc.

The long answer is why on earth do you want to run Linux on the msp430?


Incidently, is there any chance of the "reply" address for this mailing list
being set to the list, rather than the poster?  Most replies should be sent
to the list for the benifit of all, but it is easy to forget and send a
reply to the o.p.'s own address.

mvh.

 David



>
> ----- Original Message -----
> From: Martin Theobald
> To: msp...@li...
> Sent: Monday, September 09, 2002 2:08 PM
> Subject: [Mspgcc-users] running linux on msp?
>
>
> Hello,
>
> has anyone of you tried to run embedded Linux on Msp 430?
> Does anybody know some urls with informations about that?
>
> Thanks for Answers
>
> Martin
>

Re: [Mspgcc-users] running linux on msp?

[Andreas S. <and...@we...>]

Monday, September 09, 2002, 2:08:11 PM, you wrote:

> has anyone of you tried to run embedded Linux on Msp 430?
> Does anybody know some urls with informations about that?

You need at least 0.5MB ROM and 1 MB Ram -> no chance with the MSP430.

[Mspgcc-users] running linux on msp?

[Martin T. <m.t...@az...>]

Hello,

has anyone of you tried to run embedded Linux on Msp 430?
Does anybody know some urls with informations about that?

Thanks for Answers

Martin

[Mspgcc-users] gcc-3.current

[Dmitry <di...@ei...>]

Fellows,
I submitted a port of gcc3.-current (I played with version of 2002 09 02).

This seems to produce a slightly better (just a few instruction less) code for big 
projects and I think the run-time performance is slightly better (but not obviously seen from assembly file though)

So, you can find files in gcc/gcc-3.3/ directory.

~d

/********************************************************************
     ("`-''-/").___..--''"`-._     (\   Dimmy the Wild      UA1ACZ
      `6_ 6  )   `-.  (     ).`-.__.`)  Enterprise Information Sys 
      (_Y_.)'  ._   )  `._ `. ``-..-'   Nevsky prospekt,   20 / 44
    _..`--'_..-_/  /--'_.' ,'           Saint Petersburg,   Russia
   (il),-''  (li),'  ((!.-'             +7 (812)  3468202, 5585314
 ********************************************************************/

Re: [Mspgcc-users] mspgcc and c-spy debug as c?

[Chris L. <cli...@gm...>]

Am 08.09.2002 12:16:34, schrieb "David Brown" <da...@we...>:
>(is there an archive available?).

yes. here:
http://sourceforge.net/mail/?group_id=42303

chris

Re: [Mspgcc-users] mspgcc and c-spy debug as c?

[David B. <da...@we...>]

> Am 06.09.2002 19:44:35, schrieb Szabó András <sz...@sa...>:
> >Hi!
>
> welcome!
>
> >My problem is:
> > 1) I write my c program
> > 2) I compile it to hex
> > 3) C-spy debugs is as bytecode/asm not as c.
>
> that's all what is in a hex file - no sources.
>
> >How can i see c language lines in my c-spy?
>
> currently not possible, there are only ways to get some symbols in the
code.
> however GDB (gnu debugger) support is on it's way.
>
> chris
>

It is quite possible to see the C language lines for c-spy debugging.  It's
a bit inelegant, and it doesn't include data segment (variable) information,
but it works well enough until gdb support is completed.  The basic trick is
to use objdump to generate an IAR-assembler compatible disassembly including
the original C code as assembler comments.  This is then assembled and
linked using the (free) IAR tools, giving you something that looks to C-Spy
as a large assembly file, which includes the C code as comments.  There
should be reasonable instructions in some past posts to this list (is there
an archive available?).

David

Re: [Mspgcc-users] sfrb ?

[dimmy <di...@ei...>]

basically, this comes from TI includes.

sfrb means thet this defenition of SpecialFunctionRegisterByte.

the sfrb is a magic macro, which should translate, for example ME1_ to ME1 
defenition like

volatile unsigned char ME1 asm("0x0004");

Obviously, thai can be done as:

#define ME1	(*(insigned char *)0x0004))

But then gcc issues some stupid code for read-modify-write instructions 
(sometimes)

~d


On Saturday 07 September 2002 01:35, Garst R. Reese wrote:
> Can someone enlighten me on the meaning of sfrb in the following ?
> #define ME1_                0x0004  /* Module Enable 1 */
> sfrb(ME1,ME1_);
>
> Thanks,
>    Garst
>
>
> -------------------------------------------------------
> This sf.net email is sponsored by: OSDN - Tired of that same old
> cell phone?  Get a new here for FREE!
> https://www.inphonic.com/r.asp?r=sourceforge1&refcode1=vs3390
> _______________________________________________
> Mspgcc-users mailing list
> Msp...@li...
> https://lists.sourceforge.net/lists/listinfo/mspgcc-users

-- 
*********************************************************************
   ("`-''-/").___..--''"`-._     (\       Dimmy the Wild      UA1ACZ
    `6_ 6  )   `-.  (     ).`-.__.`)      Enterprise Information Sys 
    (_Y_.)'  ._   )  `._ `. ``-..-'       Nevsky prospekt,   20 / 44
  _..`--'_..-_/  /--'_.' ,'               Saint Petersburg,   Russia
 (il),-''  (li),'  ((!.-'                 +7 (812) 314-8860, 5585314
*********************************************************************

Re: [Mspgcc-users] sfrb ?

[Garst R. R. <re...@is...>]

Thanks,
Also for the serJTAG stuff. My thanks as well to TI.
Chris Liechti wrote:
> 
> Am 06.09.2002 23:35:14, schrieb "Garst R. Reese" <re...@is...>:
> 
> >Can someone enlighten me on the meaning of sfrb in the following ?
> >#define ME1_                0x0004  /* Module Enable 1 */
> >sfrb(ME1,ME1_);
> 
> noooo dont ask ;-)
> 
> the macro is defined in <iomacros.h> it's a rather tricky macro to convince GCC that it can access that address as variable and that it's
> volatile (must not be cached in a register). (just defining a volatile variable at address x does generate less efficient code, e.g. "bic"
> and "bis" asm insn are not used directly, rather a its a read, modify, store sequence)
> 
> i don't know the real the reason that it's split in two parts, a define and ans sfrb(), our files are just based on the TI ones, and those had
> this like that. maybe they thought that you could use the underline verison if you use the address for calculations or so.
> 
> chris

Re: [Mspgcc-users] sfrb ?

[Chris L. <cli...@gm...>]

Am 06.09.2002 23:35:14, schrieb "Garst R. Reese" <re...@is...>:

>Can someone enlighten me on the meaning of sfrb in the following ?
>#define ME1_                0x0004  /* Module Enable 1 */
>sfrb(ME1,ME1_);

noooo dont ask ;-)

the macro is defined in <iomacros.h> it's a rather tricky macro to convince GCC that it can access that address as variable and that it's 
volatile (must not be cached in a register). (just defining a volatile variable at address x does generate less efficient code, e.g. "bic" 
and "bis" asm insn are not used directly, rather a its a read, modify, store sequence)

i don't know the real the reason that it's split in two parts, a define and ans sfrb(), our files are just based on the TI ones, and those had 
this like that. maybe they thought that you could use the underline verison if you use the address for calculations or so.

chris

[Mspgcc-users] sfrb ?

[Garst R. R. <re...@is...>]

Can someone enlighten me on the meaning of sfrb in the following ?
#define ME1_                0x0004  /* Module Enable 1 */
sfrb(ME1,ME1_);

Thanks,
   Garst

Re: [Mspgcc-users] Strange strings

[Chris L. <cli...@gm...>]

Am 06.09.2002 17:21:54, schrieb Paul Burke <pa...@sc...>:

>I have just had a little problem when outputting strings.
>
>If a string has 2 or more commas in it, the compiler gets upset. For
>example:
>
> puts("O for a muse of fire, that would ascend\r\n"); 
> puts("The brightest Heaven of invention\r\n"); 
> puts("A kingdom for a stage, princes to act,\r\n");
> puts("And monarchs to behold the swelling scene\r\n\r\n");
>
>Line 1 is fine. Line 3 causes "internal error: unsupported location
>error", but if I remove either of the commas it too is OK.
>
>Is this normal for GCC, and if so, why?

no thats not normal:
if i compile this:

#include <stdlib.h>
int main() {
    char buf[80];
    snprintf(buf, 80, "hello, coma\n");
    snprintf(buf, 80, "hello, coma,\n");
    snprintf(buf, 80, "A kingdom for a stage, princes to act,\r\n");
}

with "msp430-gcc -mmcu=msp430x1121 tt.c" i don't get any
warning or error.

you problem is probably related to the definition of "puts". is it a
macro or function?

are you using a recent msp430-gcc version?

chris

Re: [Mspgcc-users] Strange strings

[dimmy <di...@ei...>]

Oops,
could not reproduce it.

looks like gcc works fine here...

Hm... can you send me the rest?

~d


On Friday 06 September 2002 19:21, Paul Burke wrote:
> I have just had a little problem when outputting strings.
>
> If a string has 2 or more commas in it, the compiler gets upset. For
> example:
>
>  puts("O for a muse of fire, that would ascend\r\n");
>  puts("The brightest Heaven of invention\r\n");
>  puts("A kingdom for a stage, princes to act,\r\n");
>  puts("And monarchs to behold the swelling scene\r\n\r\n");
>
> Line 1 is fine. Line 3 causes "internal error: unsupported location
> error", but if I remove either of the commas it too is OK.
>
> Is this normal for GCC, and if so, why?
>
> Thanks
>
> Paul Burke
>
>
> -------------------------------------------------------
> This sf.net email is sponsored by: OSDN - Tired of that same old
> cell phone?  Get a new here for FREE!
> https://www.inphonic.com/r.asp?r=sourceforge1&refcode1=vs3390
> _______________________________________________
> Mspgcc-users mailing list
> Msp...@li...
> https://lists.sourceforge.net/lists/listinfo/mspgcc-users

-- 
*********************************************************************
   ("`-''-/").___..--''"`-._     (\       Dimmy the Wild      UA1ACZ
    `6_ 6  )   `-.  (     ).`-.__.`)      Enterprise Information Sys 
    (_Y_.)'  ._   )  `._ `. ``-..-'       Nevsky prospekt,   20 / 44
  _..`--'_..-_/  /--'_.' ,'               Saint Petersburg,   Russia
 (il),-''  (li),'  ((!.-'                 +7 (812) 314-8860, 5585314
*********************************************************************