On Fri, May 15, 2009 at 6:59 AM, Sujit Karataparambil <sjt.kar@xxxxxxxxx> wrote:
> Neil,
>
> I donot understand why this needs to be done in raid. Why should be the raid
where else ?
> taking care of 4k*n data sets.
> What I see of this patch is that it specifically checks for some sort
> of 4K Boundaries.
instead of 4K*2^n where n>=0
correct and recompure IO posistion.
> Thanks,
> Sujit
>
> On Thu, May 14, 2009 at 10:42 PM, raz ben yehuda <raziebe@xxxxxxxxx> wrote:
>> Neil.
>> I do not know what is policy for mdadm patches. I would thankful for some instructions.
>> The bellow is the patch I applied to support raid0 chunk sizes 4K*n.
>> ---
>> --- mdadm.c (revision 16152)
>> +++ mdadm.c (working copy)
>> @@ -314,7 +314,8 @@
>> exit(2);
>> }
>> chunk = strtol(optarg, &c, 10);
>> - if (!optarg[0] || *c || chunk<4 || (chunk%4)) {
>> + if (!optarg[0] || *c || chunk<4 || (((chunk-1)&chunk) && level !=0)
>> + || (((chunk%4) && level ==0) )) {
>> fprintf(stderr, Name ": invalid chunk/rounding value: %s\n",
>> optarg);
>> exit(2);
>>
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-raid" in
>> the body of a message to majordomo@xxxxxxxxxxxxxxx
>> More majordomo info at http://vger.kernel.org/majordomo-info.html
>>
>
>
>
> --
> -- Sujit K M
>
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
