bcache: bugfix - moving_gc now moves only correct buckets

Removed gc_move_threshold because picking buckets only by
threshold could lead moving extra buckets (ei. if there are
buckets at the threshold that aren't supposed to be moved
do to space considerations).

This is replaced by a GC_MOVE bit in the gc_mark bitmask.
Now only marked buckets get moved.

Signed-off-by: Nicholas Swenson <nks@daterainc.com>
Signed-off-by: Kent Overstreet <kmo@daterainc.com>

diff --git a/drivers/md/bcache/alloc.c b/drivers/md/bcache/alloc.c
index 2b46bf1..4c9852d 100644 (file)
--- a/drivers/md/bcache/alloc.c
+++ b/drivers/md/bcache/alloc.c
@@ -421,9 +421,11 @@ out:
 
        if (watermark <= WATERMARK_METADATA) {
                SET_GC_MARK(b, GC_MARK_METADATA);
+               SET_GC_MOVE(b, 0);
                b->prio = BTREE_PRIO;
        } else {
                SET_GC_MARK(b, GC_MARK_RECLAIMABLE);
+               SET_GC_MOVE(b, 0);
                b->prio = INITIAL_PRIO;
        }
 

diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h
index 4beb55a..a7b1a76 100644 (file)
--- a/drivers/md/bcache/bcache.h
+++ b/drivers/md/bcache/bcache.h
@@ -197,7 +197,7 @@ struct bucket {
        uint8_t         disk_gen;
        uint8_t         last_gc; /* Most out of date gen in the btree */
        uint8_t         gc_gen;
-       uint16_t        gc_mark;
+       uint16_t        gc_mark; /* Bitfield used by GC. See below for field */
 };
 
 /*
@@ -209,7 +209,8 @@ BITMASK(GC_MARK,     struct bucket, gc_mark, 0, 2);
 #define GC_MARK_RECLAIMABLE    0
 #define GC_MARK_DIRTY          1
 #define GC_MARK_METADATA       2
-BITMASK(GC_SECTORS_USED, struct bucket, gc_mark, 2, 14);
+BITMASK(GC_SECTORS_USED, struct bucket, gc_mark, 2, 13);
+BITMASK(GC_MOVE, struct bucket, gc_mark, 15, 1);
 
 #include "journal.h"
 #include "stats.h"
@@ -445,7 +446,6 @@ struct cache {
         * call prio_write() to keep gens from wrapping.
         */
        uint8_t                 need_save_prio;
-       unsigned                gc_move_threshold;
 
        /*
         * If nonzero, we know we aren't going to find any buckets to invalidate

diff --git a/drivers/md/bcache/movinggc.c b/drivers/md/bcache/movinggc.c
index 46c9523..30f347d 100644 (file)
--- a/drivers/md/bcache/movinggc.c
+++ b/drivers/md/bcache/movinggc.c
@@ -25,10 +25,9 @@ static bool moving_pred(struct keybuf *buf, struct bkey *k)
        unsigned i;
 
        for (i = 0; i < KEY_PTRS(k); i++) {
-               struct cache *ca = PTR_CACHE(c, k, i);
                struct bucket *g = PTR_BUCKET(c, k, i);
 
-               if (GC_SECTORS_USED(g) < ca->gc_move_threshold)
+               if (GC_MOVE(g))
                        return true;
        }
 
@@ -227,9 +226,8 @@ void bch_moving_gc(struct cache_set *c)
                        sectors_to_move -= GC_SECTORS_USED(b);
                }
 
-               ca->gc_move_threshold = bucket_heap_top(ca);
-
-               pr_debug("threshold %u", ca->gc_move_threshold);
+               while (heap_pop(&ca->heap, b, bucket_cmp))
+                       SET_GC_MOVE(b, 1);
        }
 
        mutex_unlock(&c->bucket_lock);