This is the mail archive of the
ecos-discuss@sources.redhat.com
mailing list for the eCos project.
Re: minor bug in flash_28fxxx.inl
On Wed, Jun 20, 2001 at 11:43:19AM +0200, Jesper Skov wrote:
> >>>>> "Christoph" == Christoph Csebits <christoph.csebits@frequentis.com> writes:
>
> Christoph> hi maybe i found a bug in flash_28fxxx.inl.
>
> Christoph> a 28F160 have a block size of 32k ( 0x8000 ) the whole
> Christoph> flash block size is multiplied with the number of
> Christoph> interleaved devices. (thats clear)
>
Another minor deviation:
I investigated the intel manual and found that
block size for a 28f160 is 64K, so:
# define FLASH_BLOCK_SIZE (0x10000*CYGNUM_FLASH_INTERLEAVE)
> The latter should be CYGNUM_FLASH_SERIES, I think. I hope to get time
> to rewrite that driver to match what I did for the AMD driver soonish
> - but I'll commit this fix for now. Thanks.
I dont think that the block size of the whole flash memory
depends on consecutively flash devices.
(depending on interleaved flash devices is ok)
i would prefer to define
# define FLASH_BLOCK_SIZE (0x10000)
and leave the other code like it is.
nevertheless the driver is working, thanks
Christoph