On Sunday 31 August 2014 at 13:24:35, Edward Shishkin wrote:
>
> On 08/18/2014 01:31 PM, Ivan Shapovalov wrote:
> > - Deferred (BA_DEFER) deallocations do not make use of target stage and
> > other flags, so remove them to avoid confusion.
> > - mark final deallocations in wandering log code as deferred.
> >
> > With the last point, there is a clear semantic distinction between
> > deferred and immediate deallocations.
> > Deferred mode is used to deallocate blocks that were previously written,
> > immediate mode is used to deallocate "just allocated" blocks in error
> > paths or if some previously allocated blocks turn out to be unneeded.
> >
> > This way, we may get rid of discard-related hack in the deallocation routine,
> > i. e. treating immediate deallocations as deferred (which messes up
> > block accounting). This is done in the next commit.
> >
> > Signed-off-by: Ivan Shapovalov <intelfx100@xxxxxxxxx>
> > ---
> > fs/reiser4/plugin/txmod.c | 18 ++++++------------
> > fs/reiser4/tree.c | 3 +--
> > fs/reiser4/wander.c | 5 ++---
> > 3 files changed, 9 insertions(+), 17 deletions(-)
> >
> > diff --git a/fs/reiser4/plugin/txmod.c b/fs/reiser4/plugin/txmod.c
> > index 63d461f..d8a99b7 100644
> > --- a/fs/reiser4/plugin/txmod.c
> > +++ b/fs/reiser4/plugin/txmod.c
> > @@ -287,8 +287,7 @@ static int forward_relocate_unformatted(flush_pos_t *flush_pos,
> > * on relocating - free nodes which are going to be
> > * relocated
> > */
> > - reiser4_dealloc_blocks(&start, &allocated,
> > - BLOCK_ALLOCATED, BA_DEFER);
> > + reiser4_dealloc_blocks(&start, &allocated, 0, BA_DEFER);
> >
> > /* assign new block numbers to protected nodes */
> > assign_real_blocknrs(flush_pos, oid, index, allocated, first_allocated);
> > @@ -386,16 +385,13 @@ static squeeze_result squeeze_relocate_unformatted(znode *left,
> > result = put_unit_to_end(left, key, ©_extent);
> >
> > if (result == -E_NODE_FULL) {
> > - int target_block_stage;
> > /*
> > * free blocks which were just allocated
> > */
> > - target_block_stage =
> > - (state ==
> > - ALLOCATED_EXTENT) ? BLOCK_FLUSH_RESERVED :
> > - BLOCK_UNALLOCATED;
> > reiser4_dealloc_blocks(&first_allocated, &allocated,
> > - target_block_stage,
> > + (state == ALLOCATED_EXTENT)
> > + ? BLOCK_FLUSH_RESERVED
> > + : BLOCK_UNALLOCATED,
> > BA_PERMANENT);
> > /*
> > * rewind the preceder
> > @@ -408,8 +404,7 @@ static squeeze_result squeeze_relocate_unformatted(znode *left,
> > /*
> > * free nodes which were relocated
> > */
> > - reiser4_dealloc_blocks(&start, &allocated,
> > - BLOCK_ALLOCATED, BA_DEFER);
> > + reiser4_dealloc_blocks(&start, &allocated, 0, BA_DEFER);
> > }
> > /*
> > * assign new block numbers to protected nodes
> > @@ -689,8 +684,7 @@ static int forward_try_defragment_locality(znode * node,
> > goto exit;
> >
> > if (!ZF_ISSET(node, JNODE_CREATED) &&
> > - (ret = reiser4_dealloc_block(znode_get_block(node), 0,
> > - BA_DEFER | BA_FORMATTED)))
> > + (ret = reiser4_dealloc_block(znode_get_block(node), 0, BA_DEFER)))
>
>
> Here you drop BA_FORMATTED flag for the node, which is really formatted.
> It makes a mess. Do you have more neat solution?
I've dropped BA_FORMATTED flag for one reason: if BA_DEFER is specified,
other flags are not taken into account. So it is redundant to specify those.
By the same reason 0 is already passed instead of target block stage.
Actually, only two pieces of this patch are not no-op: I've added BA_DEFER
in dealloc_tx_list() and in dealloc_wmap_actor().
If you think that BA_FORMATTED must remain despite it is unused --
then ok, so it shall be.
Thanks,
--
Ivan Shapovalov / intelfx /Attachment:
signature.asc
Description: This is a digitally signed message part.
