qcow2: Assign the L2 cache relatively to the image size

Sufficient L2 cache can noticeably improve the performance when using
large images with frequent I/O.

Previously, unless 'cache-size' was specified and was large enough, the
L2 cache was set to a certain size without taking the virtual image size
into account.

Now, the L2 cache assignment is aware of the virtual size of the image,
and will cover the entire image, unless the cache size needed for that is
larger than a certain maximum. This maximum is set to 1 MB by default
(enough to cover an 8 GB image with the default cluster size) but can
be increased or decreased using the 'l2-cache-size' option. This option
was previously documented as the *maximum* L2 cache size, and this patch
makes it behave as such, instead of as a constant size. Also, the
existing option 'cache-size' can limit the sum of both L2 and refcount
caches, as previously.

Signed-off-by: Leonid Bloch <lbloch@janustech.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block/qcow2.c

@@ -777,29 +777,35 @@ static void read_cache_sizes(BlockDriverState *bs, QemuOpts *opts,
                              uint64_t *refcount_cache_size, Error **errp)
 {
     BDRVQcow2State *s = bs->opaque;
-    uint64_t combined_cache_size;
+    uint64_t combined_cache_size, l2_cache_max_setting;
     bool l2_cache_size_set, refcount_cache_size_set, combined_cache_size_set;
     int min_refcount_cache = MIN_REFCOUNT_CACHE_SIZE * s->cluster_size;
+    uint64_t virtual_disk_size = bs->total_sectors * BDRV_SECTOR_SIZE;
+    uint64_t max_l2_cache = virtual_disk_size / (s->cluster_size / 8);
 
     combined_cache_size_set = qemu_opt_get(opts, QCOW2_OPT_CACHE_SIZE);
     l2_cache_size_set = qemu_opt_get(opts, QCOW2_OPT_L2_CACHE_SIZE);
     refcount_cache_size_set = qemu_opt_get(opts, QCOW2_OPT_REFCOUNT_CACHE_SIZE);
 
     combined_cache_size = qemu_opt_get_size(opts, QCOW2_OPT_CACHE_SIZE, 0);
-    *l2_cache_size = qemu_opt_get_size(opts, QCOW2_OPT_L2_CACHE_SIZE, 0);
+    l2_cache_max_setting = qemu_opt_get_size(opts, QCOW2_OPT_L2_CACHE_SIZE,
+                                             DEFAULT_L2_CACHE_MAX_SIZE);
     *refcount_cache_size = qemu_opt_get_size(opts,
                                              QCOW2_OPT_REFCOUNT_CACHE_SIZE, 0);
 
     *l2_cache_entry_size = qemu_opt_get_size(
         opts, QCOW2_OPT_L2_CACHE_ENTRY_SIZE, s->cluster_size);
 
+    *l2_cache_size = MIN(max_l2_cache, l2_cache_max_setting);
+
     if (combined_cache_size_set) {
         if (l2_cache_size_set && refcount_cache_size_set) {
             error_setg(errp, QCOW2_OPT_CACHE_SIZE ", " QCOW2_OPT_L2_CACHE_SIZE
                        " and " QCOW2_OPT_REFCOUNT_CACHE_SIZE " may not be set "
                        "at the same time");
             return;
-        } else if (*l2_cache_size > combined_cache_size) {
+        } else if (l2_cache_size_set &&
+                   (l2_cache_max_setting > combined_cache_size)) {
             error_setg(errp, QCOW2_OPT_L2_CACHE_SIZE " may not exceed "
                        QCOW2_OPT_CACHE_SIZE);
             return;
@@ -814,9 +820,6 @@ static void read_cache_sizes(BlockDriverState *bs, QemuOpts *opts,
         } else if (refcount_cache_size_set) {
             *l2_cache_size = combined_cache_size - *refcount_cache_size;
         } else {
-            uint64_t virtual_disk_size = bs->total_sectors * BDRV_SECTOR_SIZE;
-            uint64_t max_l2_cache = virtual_disk_size / (s->cluster_size / 8);
-
             /* Assign as much memory as possible to the L2 cache, and
              * use the remainder for the refcount cache */
             if (combined_cache_size >= max_l2_cache + min_refcount_cache) {
@@ -828,12 +831,6 @@ static void read_cache_sizes(BlockDriverState *bs, QemuOpts *opts,
                 *l2_cache_size = combined_cache_size - *refcount_cache_size;
             }
         }
-    } else {
-        if (!l2_cache_size_set) {
-            *l2_cache_size = MAX(DEFAULT_L2_CACHE_SIZE,
-                                 (uint64_t)DEFAULT_L2_CACHE_CLUSTERS
-                                 * s->cluster_size);
-        }
     }
     /* l2_cache_size and refcount_cache_size are ensured to have at least
      * their minimum values in qcow2_update_options_prepare() */

block/qcow2.h

@@ -74,9 +74,7 @@
 /* Must be at least 4 to cover all cases of refcount table growth */
 #define MIN_REFCOUNT_CACHE_SIZE 4 /* clusters */
 
-/* Whichever is more */
+#define DEFAULT_L2_CACHE_MAX_SIZE S_1MiB
-#define DEFAULT_L2_CACHE_CLUSTERS 8 /* clusters */
-#define DEFAULT_L2_CACHE_SIZE S_1MiB
 
 #define DEFAULT_CLUSTER_SIZE S_64KiB
 

docs/qcow2-cache.txt

@@ -125,8 +125,12 @@ There are a few things that need to be taken into account:
  - Both caches must have a size that is a multiple of the cluster size
    (or the cache entry size: see "Using smaller cache sizes" below).
 
- - The default L2 cache size is 8 clusters or 1MB (whichever is more),
+ - The maximum L2 cache size is 1 MB by default (enough for full coverage
-   and the minimum is 2 clusters (or 2 cache entries, see below).
+   of 8 GB images, with the default cluster size). This value can be
+   modified using the "l2-cache-size" option. QEMU will not use more memory
+   than needed to hold all of the image's L2 tables, regardless of this max.
+   value. The minimal L2 cache size is 2 clusters (or 2 cache entries, see
+   below).
 
  - The default (and minimum) refcount cache size is 4 clusters.
 
@@ -184,9 +188,10 @@ Some things to take into account:
    always uses the cluster size as the entry size.
 
  - If the L2 cache is big enough to hold all of the image's L2 tables
-   (as explained in the "Choosing the right cache sizes" section
+   (as explained in the "Choosing the right cache sizes" and "How to
-   earlier in this document) then none of this is necessary and you
+   configure the cache sizes" sections in this document) then none of
-   can omit the "l2-cache-entry-size" parameter altogether.
+   this is necessary and you can omit the "l2-cache-entry-size"
+   parameter altogether.
 
 
 Reducing the memory usage

qemu-options.hx

@@ -736,9 +736,9 @@ The maximum total size of the L2 table and refcount block caches in bytes
 
 @item l2-cache-size
 The maximum size of the L2 table cache in bytes
-(default: if cache-size is not defined - 1048576 bytes or 8 clusters, whichever
+(default: if cache-size is not specified - 1M; otherwise, as large as possible
-is larger; otherwise, as large as possible or needed within the cache-size,
+within the cache-size, while permitting the requested or the minimal refcount
-while permitting the requested or the minimal refcount cache size)
+cache size)
 
 @item refcount-cache-size
 The maximum size of the refcount block cache in bytes

tests/qemu-iotests/137

@@ -109,7 +109,6 @@ $QEMU_IO \
     -c "reopen -o cache-size=1M,l2-cache-size=64k,refcount-cache-size=64k" \
     -c "reopen -o cache-size=1M,l2-cache-size=2M" \
     -c "reopen -o cache-size=1M,refcount-cache-size=2M" \
-    -c "reopen -o l2-cache-size=256T" \
     -c "reopen -o l2-cache-entry-size=33k" \
     -c "reopen -o l2-cache-entry-size=128k" \
     -c "reopen -o refcount-cache-size=256T" \
@@ -119,6 +118,13 @@ $QEMU_IO \
     -c "reopen -o cache-clean-interval=-1" \
     "$TEST_IMG" | _filter_qemu_io
 
+IMGOPTS="cluster_size=256k" _make_test_img 32P
+$QEMU_IO \
+    -c "reopen -o l2-cache-entry-size=512,l2-cache-size=1T" \
+    "$TEST_IMG" | _filter_qemu_io
+
+_make_test_img 64M
+
 echo
 echo === Test transaction semantics ===
 echo

tests/qemu-iotests/137.out

@@ -19,7 +19,6 @@ Parameter 'lazy-refcounts' expects 'on' or 'off'
 cache-size, l2-cache-size and refcount-cache-size may not be set at the same time
 l2-cache-size may not exceed cache-size
 refcount-cache-size may not exceed cache-size
-L2 cache size too big
 L2 cache entry size must be a power of two between 512 and the cluster size (65536)
 L2 cache entry size must be a power of two between 512 and the cluster size (65536)
 Refcount cache size too big
@@ -27,6 +26,9 @@ Conflicting values for qcow2 options 'overlap-check' ('constant') and 'overlap-c
 Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all
 Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all
 Cache clean interval too big
+Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=36028797018963968
+L2 cache size too big
+Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
 
 === Test transaction semantics ===