EmbeddedRelated.com
Forums
The 2024 Embedded Online Conference

Serial Flash Protected Area

Started by Vladimir Vassilevsky October 28, 2007
  Is there any reason why the SPI flash chips always have the protected 
area in the upper addresses at the end of the available memory?

  All of the CPUs that I know start booting from the SPI flash by 
reading some information which begins at the address 0. So if the 
beginning of the flash gets corrupt, then the CPU is dead and we can't 
reprogram the flash by software. Therefore it seems logical to protect 
the beginning of the flash, not the end.

But why the protection is always done in the opposite way? Is there any 
logical reason for that?

Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com
Vladimir Vassilevsky wrote:
> > Is there any reason why the SPI flash chips always have the protected > area in the upper addresses at the end of the available memory? > > All of the CPUs that I know start booting from the SPI flash by reading > some information which begins at the address 0. So if the beginning of > the flash gets corrupt, then the CPU is dead and we can't reprogram the > flash by software. Therefore it seems logical to protect the beginning > of the flash, not the end. > > But why the protection is always done in the opposite way? Is there any > logical reason for that?
I think it carries over from Serial EEPROMs, and pre-dates booting from Serial memory. In Serial EE's it allowed some small protected area, above the normal data-storage area, and given the widely varying sizes of Serial EE (but all with common 0000 address point) to protect the top area, away from the data space. I'm waiting for 'run from serial memory', as the next step, for code that does not need to be fast :) -jg

Jim Granville wrote:


> I'm waiting for 'run from serial memory', as the next step, for > code that does not need to be fast :)
Oh, that would be marvelous! Also, it would be great to have a CPU instruction which executes the contents of a register as an opcode :) Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com

The 2024 Embedded Online Conference