Forums

S-record and the Sreccvt with Cosmic

Started by Frank August 10, 2005
Hi,

I am using a DP256 device with the Cosmic tool chain. My program works fine when downloaded via Zap tools.

However i need a even boundary paged based srecord for manual programming. My problem is, I used the cosmic chex to generate the srecord as follows,

chex -p -o out.s in.o I used the above file with SRECCVT as follows

sreccvt -m 0xC0000 0xFFFFF 32 -lp -o out.s in.s

However the above comes up with mem segment out side memory map as follows..

S2 24 0C 7BBC.............

I basically would just like to "reformat" the srecord on even boundaries. However it seems that my CHEX included code areas that are outside the specified 0xC0000 - 0xFFFFF

It is clear that EEPROM based code should not be there, however unpaged code in the 64k base address should be part of the srecord somehow.

Right now I just very confused.............

Frank


Frank

S-Records are defined to have linear physical addresses. For the dp256,
the physical linear
space is $C0000-$FFFFF. All the 16k pages of flash, even those in the
64k paged memory
space are mapped into $C0000-$FFFFF for S-Record purposes. Only Logical
Resords
(which by the way are not S-Records) have addresses in the 64k paged space.

For the dp256, the 64k paged space is mapped as:

64k paged page physical linear
___________ ___ __________________
$4000-$7FFF 3E $F8000-$FBFFF
$8000-$BFFF depends on the ppage register
$C000-$FFFF 3F $FC000-$FFFFF

Dave Frank wrote:

> Hi,
>
> I am using a DP256 device with the Cosmic tool chain. My program works
> fine when downloaded via Zap tools.
>
> However i need a even boundary paged based srecord for manual
> programming. My problem is, > I used the cosmic chex to generate the srecord as follows,
>
> chex -p -o out.s in.o > I used the above file with SRECCVT as follows
>
> sreccvt -m 0xC0000 0xFFFFF 32 -lp -o out.s in.s
>
> However the above comes up with mem segment out side memory map as
> follows..
>
> S2 24 0C 7BBC.............
>
> I basically would just like to "reformat" the srecord on even
> boundaries. However it seems that my CHEX included code areas that are
> outside the specified 0xC0000 - 0xFFFFF
>
> It is clear that EEPROM based code should not be there, however
> unpaged code in the 64k base address should be part of the srecord
> somehow.
>
> Right now I just very confused.............
>
> Frank >
> >. >
>




You should ask them for help. I gather you paid them plenty already.

You don't want -lp. That's doing much more than adjust alignment. Are
you going to load it in banked s19 format, or linear, and which is it
in now? --- In 68HC12@68HC..., Frank <linktek@s...> wrote:
> Hi,
>
> I am using a DP256 device with the Cosmic tool chain. My program
works fine when downloaded via Zap tools.
>
> However i need a even boundary paged based srecord for manual
programming. My problem is,
>
>
> I used the cosmic chex to generate the srecord as follows,
>
> chex -p -o out.s in.o > I used the above file with SRECCVT as follows
>
> sreccvt -m 0xC0000 0xFFFFF 32 -lp -o out.s in.s
>
> However the above comes up with mem segment out side memory map as
follows..
>
> S2 24 0C 7BBC.............
>
> I basically would just like to "reformat" the srecord on even
boundaries. However it seems that my CHEX included code areas that are
outside the specified 0xC0000 - 0xFFFFF
>
> It is clear that EEPROM based code should not be there, however
unpaged code in the 64k base address should be part of the srecord
somehow.
>
> Right now I just very confused.............
>
> Frank >


Hi,

Yes, I excluded -lp so all pages will be visible through address range 0x0C
0000 - 0x0F FFFF.

I have, at this point, successfully generated a linear output srecord file
after removing all the conflicting memory map areas (eeprom stuff).

Looking at the S2 output file, all seems to be in order with all records
starting on even addresses etc. However this is a linear S2 format.

I ideally would like to have a paged S2 format.

Thanks very much for the assistance

Frank ----- Original Message -----
From: "Jefferson Smith" <imajeffs@imaj...>
To: <68HC12@68HC...>
Sent: Wednesday, August 10, 2005 1:46 PM
Subject: [68HC12] Re: S-record and the Sreccvt with Cosmic You should ask them for help. I gather you paid them plenty already.

You don't want -lp. That's doing much more than adjust alignment. Are
you going to load it in banked s19 format, or linear, and which is it
in now? --- In 68HC12@68HC..., Frank <linktek@s...> wrote:
> Hi,
>
> I am using a DP256 device with the Cosmic tool chain. My program
works fine when downloaded via Zap tools.
>
> However i need a even boundary paged based srecord for manual
programming. My problem is,
>
>
> I used the cosmic chex to generate the srecord as follows,
>
> chex -p -o out.s in.o > I used the above file with SRECCVT as follows
>
> sreccvt -m 0xC0000 0xFFFFF 32 -lp -o out.s in.s
>
> However the above comes up with mem segment out side memory map as
follows..
>
> S2 24 0C 7BBC.............
>
> I basically would just like to "reformat" the srecord on even
boundaries. However it seems that my CHEX included code areas that are
outside the specified 0xC0000 - 0xFFFFF
>
> It is clear that EEPROM based code should not be there, however
unpaged code in the 64k base address should be part of the srecord
somehow.
>
> Right now I just very confused.............
>
> Frank >
Yahoo! Groups Links



--- In 68HC12@68HC..., Frank <linktek@s...> wrote:

> Looking at the S2 output file, all seems to be in order with all
records
> starting on even addresses etc. However this is a linear S2 format.
>
> I ideally would like to have a paged S2 format.
>
> Thanks very much for the assistance
>
> Frank

You need to read SRecCvtRG.pdf

The -ob would put the output in banked format, if that's what you mean.