EmbeddedRelated.com
Forums
Memfault State of IoT Report

First Flash segment after INFO Blocks documented incorrectly?

Started by reallygene October 18, 2006
Greetings.

I'm using a '149 with 60K flash.
I needed to store more parameters than would fit in the INFO_A and
INFO_B blocks. I chose to grab the next sector in flash (referred to
in the docs as "Segment n".

A careful examination of the flash map on page 16 of the SLAS272F
datasheet for the '149 shows that Segment n is only 256 bytes in
size, which contradicts the text: "Each segment in main memory is 512
bytes in size."

All other segments are 512 bytes, and are aligned on 512-byte
boundaries.

This only shows up in the 60K part; in all smaller parts, Segment n
is 512 bytes...

Has anyone seen this documented? Or have a guess as to why it occurs?

Thanks,

--Gene

Beginning Microcontrollers with the MSP430

Take a look at the memory organization on page 15. (Last colmn of the
table.) The CPU can only access 64KB. They may have 512 byte in that
segment but the CPU (or JTAG) simple run out of addreees to access
half of that segment.

--- In m..., "reallygene" wrote:
>
> Greetings.
>
> I'm using a '149 with 60K flash.
> I needed to store more parameters than would fit in the INFO_A and
> INFO_B blocks. I chose to grab the next sector in flash (referred
to
> in the docs as "Segment n".
>
> A careful examination of the flash map on page 16 of the SLAS272F
> datasheet for the '149 shows that Segment n is only 256 bytes in
> size, which contradicts the text: "Each segment in main memory is
512
> bytes in size."
>
> All other segments are 512 bytes, and are aligned on 512-byte
> boundaries.
>
> This only shows up in the 60K part; in all smaller parts, Segment n
> is 512 bytes...
>
> Has anyone seen this documented? Or have a guess as to why it
occurs?
>
> Thanks,
>
> --Gene
>
Hi Gene, just saw your post.

I posted this email back in November 2005, but never got a TI
comment. I suppose we all just accept it as obvious.

"Here is something curious:

MSP430F Flash consists of 2 (or 1 for some devices) 128-byte segments
of info memory located at 0x1000-0x10FF. There are also 512-byte
segments stacked up to the end address 0xFFFF, the number depending
on the processor version. For the MSP430F149 that means the lowest
segment (so-called segment-N) must therefore only be 256 bytes in
length, from 0x1100-0x11FF.

Curious; I wonder why this isn't mentioned in the data sheet. One
assumes that erasing this segment-N (where N0) only erases
256-bytes. Segment N-1 (119) and onwards are presumably all 512
bytes. This might affect some download algorithms."

Hugh

At 02:28 PM 10/18/2006, you wrote:
Greetings.

I'm using a '149 with 60K flash.
I needed to store more parameters than would fit in the INFO_A and
INFO_B blocks. I chose to grab the next sector in flash (referred to
in the docs as "Segment n".

A careful examination of the flash map on page 16 of the SLAS272F
datasheet for the '149 shows that Segment n is only 256 bytes in
size, which contradicts the text: "Each segment in main memory is 512
bytes in size."

All other segments are 512 bytes, and are aligned on 512-byte
boundaries.

This only shows up in the 60K part; in all smaller parts, Segment n
is 512 bytes...

Has anyone seen this documented? Or have a guess as to why it occurs?

Thanks,

--Gene
If you look at the data sheet for the 149 on page 16 there is a
picture of the memory map and it shows that the sgment n runs from
1100 to 11FF (256 bytes) and segment n-1 runs from 1200 to 13FF.

Ian

On 29/10/06, Hugh Molesworth wrote:
> Hi Gene, just saw your post.
>
> I posted this email back in November 2005, but never got a TI
> comment. I suppose we all just accept it as obvious.
>
> "Here is something curious:
>
> MSP430F Flash consists of 2 (or 1 for some devices) 128-byte segments
> of info memory located at 0x1000-0x10FF. There are also 512-byte
> segments stacked up to the end address 0xFFFF, the number depending
> on the processor version. For the MSP430F149 that means the lowest
> segment (so-called segment-N) must therefore only be 256 bytes in
> length, from 0x1100-0x11FF.
>
> Curious; I wonder why this isn't mentioned in the data sheet. One
> assumes that erasing this segment-N (where N0) only erases
> 256-bytes. Segment N-1 (119) and onwards are presumably all 512
> bytes. This might affect some download algorithms."
>
> Hugh
>
> At 02:28 PM 10/18/2006, you wrote:
> Greetings.
>
> I'm using a '149 with 60K flash.
> I needed to store more parameters than would fit in the INFO_A and
> INFO_B blocks. I chose to grab the next sector in flash (referred to
> in the docs as "Segment n".
>
> A careful examination of the flash map on page 16 of the SLAS272F
> datasheet for the '149 shows that Segment n is only 256 bytes in
> size, which contradicts the text: "Each segment in main memory is 512
> bytes in size."
>
> All other segments are 512 bytes, and are aligned on 512-byte
> boundaries.
>
> This only shows up in the 60K part; in all smaller parts, Segment n
> is 512 bytes...
>
> Has anyone seen this documented? Or have a guess as to why it occurs?
>
> Thanks,
>
> --Gene

Memfault State of IoT Report