[parted-devel] [PATCH 6/6] libparted: fix optimal IO alignment

[Jim Meyering]

Phillip Susi wrote:

> On 12/19/2012 3:41 PM, Jim Meyering wrote:
>> Ok, that makes sense. Then what do you think about lowering the bar
>> from 1MiB to 128KiB? Maybe even 64KiB, I suppose.  Better safe than
>> sorry. Just look at how many people are still reporting problems
>> with parted 1.8, so many years later.
>
> Wasn't the problem there just that it did not know about optimal
> alignment, not that it was getting incorrect values from the kernel?

The current code filters minimal as well as optimal alignment values,
so I presume it was for a good reason.

> The facility was added to the kernel so that tools like parted could
> be told the appropriate value to use rather than having to guess, so
> I'm inclined to use it.  Why have a minimum?  If the kernel says it
> knows this device should use a 48 KiB alignment, who are we to assume
> that is wrong?

I hope Brian has details.

>> I'm pretty sure those tests do actually serve some purpose. Maybe
>> Brian (Cc'd) knows for sure? (Brian, this is about the use of
>> PED_DEFAULT_ALIGNMENT, aka 1MiB in commit
>> c749046a54d983f74f8156c0aea71b0995b9477d)
>
> Did you forget to Cc him?

Yes.  Thanks.  This time, I have.