On 2/25/20 2:05 AM, Dave Chinner wrote:
On Sat, Feb 22, 2020 at 07:06:07PM -0700, Allison Collins wrote:
This patch adds a new helper function xfs_attr_node_shrink used to shrink an
attr name into an inode if it is small enough. This helps to modularize
the greater calling function xfs_attr_node_removename.
Signed-off-by: Allison Collins <allison.henderson@xxxxxxxxxx>
Can you move this helper function up to early in the patch set?
That way the code gets simpler and less tangled before adding
all the new state machine gubbins?
I suspect that you should do this for all the functions the state
machine breaks up into gotos, too. THat way adding the state machine
is really just changing how the functions that do the work are
called, rather than jumping into the middle of long functions....
I know, it turns it into a longer series, but it also means that all
the refactoring work (which needs to be done anyway) can be
separated and merged while we are still reviewing and working on the
state machine based operations, thereby reducing the size of the
patchset you have to manage and keep up to date over time....
Cheers,
Dave.
Sure, looking ahead it sounds like that works better for others too, so
I'll move them back down. Thanks!
Allison
