Sign in
Search msp430
Subscribe to msp430
Ads
Discussion Groups
See Also
Discussion Groups | MSP430 | PLACING CODE
The purpose of this group is to foster exchange of information on the Texas Instruments MSP430 family of microcontrollers and related tools. Everyone welcome, all levels of familiarity/expertise.
PLACING CODE - B - Oct 29 9:10:08 2009
HI!
I have a system using MSP430F169 and my firmware is huge. We are near to
runout of flash memory for any updates.
The system does lots of floating point calculations, and now we would like
to add some features (for example a configurable way of doing those
calculations).
---> PROBLEM
I thought that one way of changing the calculations algorithms could be
defining a flash zone for all the floating point routines, so I could send
a chunk of binay data remotely and store there. Of course I should be
careful with the addressing zones (mainly the rountine called from outside,
i mean the code that won't be updated)
In that case, i could define memory sectors and that all, but, how can I
place the build in routines as pow() or exp(). The compiler place them as
its wish. (I am using IAR).
If you think another way of do what I want I will be pleased of hearing
ideas.
Thanks in advance.
B
[Non-text portions of this message have been removed]
------------------------------------
(You need to be a member of msp430 -- send a blank email to msp430-subscribe@yahoogroups.com )
I have a system using MSP430F169 and my firmware is huge. We are near to
runout of flash memory for any updates.
The system does lots of floating point calculations, and now we would like
to add some features (for example a configurable way of doing those
calculations).
---> PROBLEM
I thought that one way of changing the calculations algorithms could be
defining a flash zone for all the floating point routines, so I could send
a chunk of binay data remotely and store there. Of course I should be
careful with the addressing zones (mainly the rountine called from outside,
i mean the code that won't be updated)
In that case, i could define memory sectors and that all, but, how can I
place the build in routines as pow() or exp(). The compiler place them as
its wish. (I am using IAR).
If you think another way of do what I want I will be pleased of hearing
ideas.
Thanks in advance.
B
[Non-text portions of this message have been removed]
------------------------------------
(You need to be a member of msp430 -- send a blank email to msp430-subscribe@yahoogroups.com )
Re: PLACING CODE - Michael - Oct 29 9:24:45 2009
Hi Bukitoo,
Have you checked IAR's documentation? If I'm not mistaken, you can use either the #pragma vector directive (just as for interrupts) or you can use the @ operand just as you would with fixed address RAM variables.
Have you searched this forum? It's not the first time something like this has been asked.
What about your other problems like size? How about using a bigger MSP430F16x with more FLASH or RAM for storing/executing functions? Maybe you could consider using an external SPI FLASH memory (1Mbit?) for storing functions and loading them into RAM when execution is needed.
Regards,
Michael K.
--- In m...@yahoogroups.com, B wrote:
>
> HI!
>
> I have a system using MSP430F169 and my firmware is huge. We are near to
> runout of flash memory for any updates.
> The system does lots of floating point calculations, and now we would like
> to add some features (for example a configurable way of doing those
> calculations).
> ---> PROBLEM
>
> I thought that one way of changing the calculations algorithms could be
> defining a flash zone for all the floating point routines, so I could send
> a chunk of binay data remotely and store there. Of course I should be
> careful with the addressing zones (mainly the rountine called from outside,
> i mean the code that won't be updated)
>
> In that case, i could define memory sectors and that all, but, how can I
> place the build in routines as pow() or exp(). The compiler place them as
> its wish. (I am using IAR).
>
> If you think another way of do what I want I will be pleased of hearing
> ideas.
>
> Thanks in advance.
>
> B
> [Non-text portions of this message have been removed]
>
------------------------------------
(You need to be a member of msp430 -- send a blank email to msp430-subscribe@yahoogroups.com )
Have you checked IAR's documentation? If I'm not mistaken, you can use either the #pragma vector directive (just as for interrupts) or you can use the @ operand just as you would with fixed address RAM variables.
Have you searched this forum? It's not the first time something like this has been asked.
What about your other problems like size? How about using a bigger MSP430F16x with more FLASH or RAM for storing/executing functions? Maybe you could consider using an external SPI FLASH memory (1Mbit?) for storing functions and loading them into RAM when execution is needed.
Regards,
Michael K.
--- In m...@yahoogroups.com, B
>
> HI!
>
> I have a system using MSP430F169 and my firmware is huge. We are near to
> runout of flash memory for any updates.
> The system does lots of floating point calculations, and now we would like
> to add some features (for example a configurable way of doing those
> calculations).
> ---> PROBLEM
>
> I thought that one way of changing the calculations algorithms could be
> defining a flash zone for all the floating point routines, so I could send
> a chunk of binay data remotely and store there. Of course I should be
> careful with the addressing zones (mainly the rountine called from outside,
> i mean the code that won't be updated)
>
> In that case, i could define memory sectors and that all, but, how can I
> place the build in routines as pow() or exp(). The compiler place them as
> its wish. (I am using IAR).
>
> If you think another way of do what I want I will be pleased of hearing
> ideas.
>
> Thanks in advance.
>
> B
> [Non-text portions of this message have been removed]
>
------------------------------------
(You need to be a member of msp430 -- send a blank email to msp430-subscribe@yahoogroups.com )
Re: Re: PLACING CODE - B - Oct 29 9:43:08 2009
In the past I needed to place a function within a fixed address, but a
function
made by me, no a build in as those included in comiler libraries.
On Thu, Oct 29, 2009 at 10:22 AM, Michael wrote:
> Hi Bukitoo,
> Have you checked IAR's documentation? If I'm not mistaken, you can use
> either the #pragma vector directive (just as for interrupts) or you can use
> the @ operand just as you would with fixed address RAM variables.
> Have you searched this forum? It's not the first time something like this
> has been asked.
>
> What about your other problems like size? How about using a bigger
> MSP430F16x with more FLASH or RAM for storing/executing functions? Maybe you
> could consider using an external SPI FLASH memory (1Mbit?) for storing
> functions and loading them into RAM when execution is needed.
>
> Regards,
> Michael K.
> --- In m...@yahoogroups.com, B
> wrote:
> >
> > HI!
> >
> > I have a system using MSP430F169 and my firmware is huge. We are near to
> > runout of flash memory for any updates.
> > The system does lots of floating point calculations, and now we would
> like
> > to add some features (for example a configurable way of doing those
> > calculations).
> > ---> PROBLEM
> >
> > I thought that one way of changing the calculations algorithms could be
> > defining a flash zone for all the floating point routines, so I could
> send
> > a chunk of binay data remotely and store there. Of course I should be
> > careful with the addressing zones (mainly the rountine called from
> outside,
> > i mean the code that won't be updated)
> >
> > In that case, i could define memory sectors and that all, but, how can I
> > place the build in routines as pow() or exp(). The compiler place them as
> > its wish. (I am using IAR).
> >
> > If you think another way of do what I want I will be pleased of hearing
> > ideas.
> >
> > Thanks in advance.
> >
> > B
> >
> >
> > [Non-text portions of this message have been removed]
> >
>
[Non-text portions of this message have been removed]
------------------------------------
(You need to be a member of msp430 -- send a blank email to msp430-subscribe@yahoogroups.com )
function
made by me, no a build in as those included in comiler libraries.
On Thu, Oct 29, 2009 at 10:22 AM, Michael
> Hi Bukitoo,
> Have you checked IAR's documentation? If I'm not mistaken, you can use
> either the #pragma vector directive (just as for interrupts) or you can use
> the @ operand just as you would with fixed address RAM variables.
> Have you searched this forum? It's not the first time something like this
> has been asked.
>
> What about your other problems like size? How about using a bigger
> MSP430F16x with more FLASH or RAM for storing/executing functions? Maybe you
> could consider using an external SPI FLASH memory (1Mbit?) for storing
> functions and loading them into RAM when execution is needed.
>
> Regards,
> Michael K.
> --- In m...@yahoogroups.com
> wrote:
> >
> > HI!
> >
> > I have a system using MSP430F169 and my firmware is huge. We are near to
> > runout of flash memory for any updates.
> > The system does lots of floating point calculations, and now we would
> like
> > to add some features (for example a configurable way of doing those
> > calculations).
> > ---> PROBLEM
> >
> > I thought that one way of changing the calculations algorithms could be
> > defining a flash zone for all the floating point routines, so I could
> send
> > a chunk of binay data remotely and store there. Of course I should be
> > careful with the addressing zones (mainly the rountine called from
> outside,
> > i mean the code that won't be updated)
> >
> > In that case, i could define memory sectors and that all, but, how can I
> > place the build in routines as pow() or exp(). The compiler place them as
> > its wish. (I am using IAR).
> >
> > If you think another way of do what I want I will be pleased of hearing
> > ideas.
> >
> > Thanks in advance.
> >
> > B
> >
> >
> > [Non-text portions of this message have been removed]
> >
>
[Non-text portions of this message have been removed]
------------------------------------
(You need to be a member of msp430 -- send a blank email to msp430-subscribe@yahoogroups.com )
Re: PLACING CODE - p_murayama - Oct 29 20:13:41 2009
Hello!
Apparently, size does matter.
In the 16x series, I think the 169 has the largest flash or at least one of the largest.
By the way, if you need more flash and more CPU power, one way to do it is to use
a 2618. I replaced a 169 by a 2618 on one board, and it worls.
NB: I am not sure everything will work, not sure that every port, serial, etc will be
at the same pin, but in my case it works.
So you can upgrade form 8 MHZ/ 60 k flash / 2K RAM to 16 MHz / 120k flash / 8k RAM.
Might be worth trying.
Another thing: are the floating point calculations ABSOLUTELY unavoidable?
Pascal
--- In m...@yahoogroups.com, "Michael" wrote:
> What about your other problems like size? How about using a bigger MSP430F16x with more FLASH or RAM for storing/executing functions? Maybe you could consider using an external SPI FLASH memory (1Mbit?) for storing functions and loading them into RAM when execution is needed.
>
> Regards,
> Michael K.
------------------------------------
______________________________
Stellaris® MCU Family: New Parts, New Package, New Price.
(You need to be a member of msp430 -- send a blank email to msp430-subscribe@yahoogroups.com )
Apparently, size does matter.
In the 16x series, I think the 169 has the largest flash or at least one of the largest.
By the way, if you need more flash and more CPU power, one way to do it is to use
a 2618. I replaced a 169 by a 2618 on one board, and it worls.
NB: I am not sure everything will work, not sure that every port, serial, etc will be
at the same pin, but in my case it works.
So you can upgrade form 8 MHZ/ 60 k flash / 2K RAM to 16 MHz / 120k flash / 8k RAM.
Might be worth trying.
Another thing: are the floating point calculations ABSOLUTELY unavoidable?
Pascal
--- In m...@yahoogroups.com, "Michael"
> What about your other problems like size? How about using a bigger MSP430F16x with more FLASH or RAM for storing/executing functions? Maybe you could consider using an external SPI FLASH memory (1Mbit?) for storing functions and loading them into RAM when execution is needed.
>
> Regards,
> Michael K.
------------------------------------
______________________________
Stellaris® MCU Family: New Parts, New Package, New Price.
(You need to be a member of msp430 -- send a blank email to msp430-subscribe@yahoogroups.com )
Re: Re: PLACING CODE - b - Oct 29 23:46:43 2009
YES. FLOATING POINT CALCULATIONS ARE THE HART OF MY SYSTEM.
I HAVE TRIED TO DEFINE SECTOR FOR ONE OF THE FUNCTIONS OF MATH.H (POW).
USING SECTORS I COULD PLACE UNA PART OF THE POW FUNCTION WHERE I WANTED,
BUT THEN THE REST OF THE REST OF THE FUNCTIONS THAT POW NEEDS WHERE
PLACED WHEN THE COMPILER WANTED!
BY THE WAY. MY SYSTEM DOESN'T ALLOW HARDWARE CHANGES RIGHT NOW. THE
PROJECTS HAS STARTED (BACK IN 2006) I CHOSE F169 BECAUS IT WAS THE MOST
POWERFUL REGARDING MEMORY SPACE. I KNOW THAT NOW MSP FAMILY GREW.
p_murayama escribió:
>
>
> Hello!
>
> Apparently, size does matter.
> In the 16x series, I think the 169 has the largest flash or at least
> one of the largest.
> By the way, if you need more flash and more CPU power, one way to do
> it is to use
> a 2618. I replaced a 169 by a 2618 on one board, and it worls.
> NB: I am not sure everything will work, not sure that every port,
> serial, etc will be
> at the same pin, but in my case it works.
> So you can upgrade form 8 MHZ/ 60 k flash / 2K RAM to 16 MHz / 120k
> flash / 8k RAM.
>
> Might be worth trying.
>
> Another thing: are the floating point calculations ABSOLUTELY unavoidable?
>
> Pascal
>
> --- In m...@yahoogroups.com,
> "Michael" wrote:
>
> > What about your other problems like size? How about using a bigger
> MSP430F16x with more FLASH or RAM for storing/executing functions?
> Maybe you could consider using an external SPI FLASH memory (1Mbit?)
> for storing functions and loading them into RAM when execution is needed.
> >
> > Regards,
> > Michael K.
> >
> >
------------------------------------
______________________________
Stellaris® MCU Family: New Parts, New Package, New Price.
(You need to be a member of msp430 -- send a blank email to msp430-subscribe@yahoogroups.com )
I HAVE TRIED TO DEFINE SECTOR FOR ONE OF THE FUNCTIONS OF MATH.H (POW).
USING SECTORS I COULD PLACE UNA PART OF THE POW FUNCTION WHERE I WANTED,
BUT THEN THE REST OF THE REST OF THE FUNCTIONS THAT POW NEEDS WHERE
PLACED WHEN THE COMPILER WANTED!
BY THE WAY. MY SYSTEM DOESN'T ALLOW HARDWARE CHANGES RIGHT NOW. THE
PROJECTS HAS STARTED (BACK IN 2006) I CHOSE F169 BECAUS IT WAS THE MOST
POWERFUL REGARDING MEMORY SPACE. I KNOW THAT NOW MSP FAMILY GREW.
p_murayama escribió:
>
>
> Hello!
>
> Apparently, size does matter.
> In the 16x series, I think the 169 has the largest flash or at least
> one of the largest.
> By the way, if you need more flash and more CPU power, one way to do
> it is to use
> a 2618. I replaced a 169 by a 2618 on one board, and it worls.
> NB: I am not sure everything will work, not sure that every port,
> serial, etc will be
> at the same pin, but in my case it works.
> So you can upgrade form 8 MHZ/ 60 k flash / 2K RAM to 16 MHz / 120k
> flash / 8k RAM.
>
> Might be worth trying.
>
> Another thing: are the floating point calculations ABSOLUTELY unavoidable?
>
> Pascal
>
> --- In m...@yahoogroups.com
> "Michael"
>
> > What about your other problems like size? How about using a bigger
> MSP430F16x with more FLASH or RAM for storing/executing functions?
> Maybe you could consider using an external SPI FLASH memory (1Mbit?)
> for storing functions and loading them into RAM when execution is needed.
> >
> > Regards,
> > Michael K.
> >
> >
------------------------------------
______________________________
Stellaris® MCU Family: New Parts, New Package, New Price.
(You need to be a member of msp430 -- send a blank email to msp430-subscribe@yahoogroups.com )