MSPM33-SDK: Issue on length configure in linker file ?

Sebastien Le Diouris

Prodigy 120 points

Part Number: MSPM33-SDK
Other Parts Discussed in Thread: MSPM33C321A

Hello,

I'm working on MSPM33C321A eval board, and I use the "empty" exemple on Code Composer Studio.

When I read the linker file (device_linker.cmd) automatically generated, I see that :

MEMORY

{

FLASH (RX) : origin = 0x10000000, length = 0x000FFFFF

SRAM (RWX) : origin = 0x30000000, length = 0x0003FFFF

DATA (R) : origin = 0x80000000, length = 0x00007FFF

BCR_CONFIG (R) : origin = 0x80101800, length = 0x000003FF

BSL_CONFIG (R) : origin = 0x80101C00, length = 0x000003FF

}

Are there any errors, is it not ?

MEMORY

{

FLASH (RX) : origin = 0x10000000, length = 0x000100000

SRAM (RWX) : origin = 0x30000000, length = 0x00040000

DATA (R) : origin = 0x80000000, length = 0x00008000

BCR_CONFIG (R) : origin = 0x80101800, length = 0x00000400

BSL_CONFIG (R) : origin = 0x80101C00, length = 0x00000400

}

Thanks

24 days ago

0 Eric Rentschler 24 days ago

TI__Intellectual 2010 points

Hi Sebastien,

Are your seeing an error when compiling the empty project? Those are the correct numbers.

Thanks,

-Eric Rentschler

0 Sebastien Le Diouris 23 days ago in reply to Eric Rentschler

Prodigy 120 points

Hello Eric,

what I mean, is (with line FLASH for example):

FLASH (RX) : origin = 0x10000000, length = 0x000FFFFF

So the flash starts at 0x10000000 with length at 0x000FFFFF, that means the end flash is 0x100FFFFE but end flash is 0x100FFFFF

0x00100000 = 1Mo

0x000FFFFF = less than 1Mo

Thanks

0 Eric Rentschler 23 days ago in reply to Sebastien Le Diouris

TI__Intellectual 2010 points

Hi Sebastien,

The length is base 0 so there is a 1 bit offset.

Thanks,

-Eric Rentschler

0 Sebastien Le Diouris 23 days ago in reply to Eric Rentschler

Prodigy 120 points

Hello Eric,

something is not clear, if the length is based on 0, that means it is not good, let me show an short example:

FLASH (RX) : origin = 0x00000000, length = 0x0000000F (15 bytes)

Adr 0 : Data 1

Adr 1 : Data 2

Adr 2 : Data 3

Adr 3 : Data 4

....

Adr 14 (0x0000000E) : Data 15

Adr 15 (0x0000000F) : No Data because exceed the length.

Is my analyze correct or not ?

Thanks

0 Eric Rentschler 23 days ago in reply to Sebastien Le Diouris

TI__Intellectual 2010 points

Hi Sebastien,

Your analysis is missing Data16. A length of 0x0000000F corresponds to 16 bytes not 15. This is because the full flash memory address will go from 0x0000.0000 to 0x0000.000F with 0x0000.000F being the last usable memory address.

Thanks,

-Eric Rentschler

0 Sebastien Le Diouris 22 days ago in reply to Eric Rentschler

Prodigy 120 points

Hello Eric,

It is always not clear, let me explain what I have in mind.

To take the same example

So 0x0F = 15, right ? 16 = 0x10

so if I define a length of 0 --> no address

If I define a length of 1 --> just 1 address 0x00000000

If I define a length of 2 --> 2 addresses 0x00000000 & 0x00000001

If I define a length of 3 --> 3 addresses 0x00000000 & 0x00000001 & 0x00000002

....

If I define a length of 15 (0x0F) --> 15 addresses 0x00000000 & 0x00000001 & 0x00000002 & .... & 0x0000000E

So it misses 1 address, so length should be equal to 16 (0x10)

Do not calculate like this 0x00000000 + 0x0F = 0x0000000F (do not forget, you start at 0 not 1)

Thanks

0 Eric Rentschler 22 days ago in reply to Sebastien Le Diouris

TI__Intellectual 2010 points

Hi Sebastien,

The length is 0 based, meaning that a length of 0 would result in an addressable memory of 1 byte. A length of 0x0F would result in addresses from 0x00 to 0x0F with 0x0F being the last usable address.

Thanks,

-Eric Rentschler

0 Sebastien Le Diouris 22 days ago in reply to Eric Rentschler

Prodigy 120 points

Hello Eric,

ok thanks

Code Composer Studio™︎

Code Composer Studio forum

MSPM33-SDK: Issue on length configure in linker file ?