Dear Wolfgang Denk,
On Tue, Jun 1, 2010 at 6:20 PM, Wolfgang Denk wd@denx.de wrote:
Has this been tested on NOR flash devices, too?
No, I'm sorry it hasn't -- but I'm glad you asked. On a closer look I think the changes in this patch series could result in the dereference of a null pointer when using a NOR flash device. The features are centered around calls to mtd->block_isbad() and I don't think this function pointer is set for NOR flash devices. I will add 'if(mtd->block_isbad())' checks to the code to prevent the potential dereference of a null pointer and resubmit.
On Tue, Jun 1, 2010 at 6:34 PM, Wolfgang Denk wd@denx.de wrote:
Can we have a configuration that is neutral to the memory footprint for such boards that don't want to use any of the new features? It seems the patches always add ~ 200 bytes to the code size (on PowerPC).
Yes, absolutely. I think that it is patch 2/4 which does not have the surrounding ifdefs of the other patches. I will make the changes in that patch conditional on a CONFIG_ define and resubmit shortly.
I was supposed to double check the image sizes before and after the patch -- as is directed in your wiki instructions, sorry for the omission. I will include a summary of the size differences in powerpc images with the resubmission of this patch series -- along with a './MAKEALL powerpc' output -- also as directed in your wiki instructions.
Best Regards,
Ben Gardiner
--- Nanometrics Inc. +1 (613) 592-6776 x239 http://www.nanometrics.ca