Re: [PATCH 2/2] hugetlb: synchronize alloc with page cache insert

From: Adam Litke
Date: Wed Jan 11 2006 - 18:02:04 EST


On Wed, 2006-01-11 at 14:52 -0800, William Lee Irwin III wrote:
> On Wed, Jan 11, 2006 at 04:24:23PM -0600, Adam Litke wrote:
> > My only concern is if I am using the correct lock for the job here.
>
> ->i_lock looks rather fishy. It may have been necessary when ->i_blocks
> was used for nefarious purposes, but without ->i_blocks fiddling, it's
> not needed unless I somehow missed the addition of some custom fields
> to hugetlbfs inodes and their modifications by any of these functions.

Nope, all the i_blocks stuff is gone. I was just looking for a
spin_lock for serializing all allocations for a particular hugeltbfs
file and i_lock seemed to fit that bill. It could be said, however,
that the locking strategy used in the patch protects a section of code,
not a data structure (which can be a bad idea). Any thoughts on a less
"fishy" locking strategy for this case?

--
Adam Litke - (agl at us.ibm.com)
IBM Linux Technology Center

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/