Dear Mike Frysinger, All:)
> On Tuesday, September 14, 2010 10:50:09 Reinhard Meyer wrote:
>> JEDEC types
>> non JEDEC types
> 
> this changelog is useless ... either make it say something, or just leave it 
> empty
> 
>> +static int ramtron_common(struct spi_flash *flash,
>> +            u32 offset, size_t len, void *buf, u8 command)
>> +{
>> +
>> +    if (sn->params->addr_len == 3 && sn->params->merge_cmd == 0) {
>> +            cmd[0] = command;
>> +            cmd[1] = offset >> 16;
>> +            cmd[2] = offset >> 8;
>> +            cmd[3] = offset;
>> +            cmd_len = 4;
>> +    }
>> +    else if (sn->params->addr_len == 2 && sn->params->merge_cmd == 0) {
>> +            cmd[0] = command;
>> +            cmd[1] = offset >> 8;
>> +            cmd[2] = offset;
>> +            cmd_len = 3;
>> +    }
>> +    else {
> 
> i thought i had mentioned this before, but those else statements need cuddling
> 
>> +    /* claim the bus */
>> +    ret = spi_claim_bus(flash->spi);
>> +    if (ret) {
>> +            debug("SF: Unable to claim SPI bus\n");
>> +            return ret;
>> +    }
> 
> i'm thinking we should push the bus claim from the read/write funcs into the 
> common spi flash layer ... not that this is specific to your flash.  we can 
> do 
> this after yours gets merged.
> 
>> +    sn = malloc(sizeof(struct ramtron_spi_fram));
> 
> sizeof(*sn)
> 
> otherwise, this looks good
> -mike

I am unsure whether currently I should put further effort into this, because 
this
won't work on architectures with incomplete relocation. The pointers in the
"probe" table would need relocation fixups otherwise.

Reinhard

<<attachment: reinhard_meyer.vcf>>

_______________________________________________
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

Reply via email to