| Message ID | 0b8f83b4f26071753dc409645ad8c6f6e215468a.1520952419.git.berto@igalia.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On 03/13/2018 10:02 AM, Alberto Garcia wrote: > The L2 and refcount caches have default sizes that can be overriden > using the l2-cache-size and refcount-cache-size (an additional > parameter named cache-size sets the combined size of both caches). > > Unless forced by one of the aforementioned parameters, QEMU will set > the unspecified sizes so that the L2 cache is 4 times larger than the > refcount cache. > > > However this patch takes a completely different approach and instead > of keeping a ratio between both cache sizes it assigns as much as > possible to the L2 cache and the remainder to the refcount cache. > > The reason is that L2 tables are used for every single I/O request > from the guest and the effect of increasing the cache is significant > and clearly measurable. Refcount blocks are however only used for > cluster allocation and internal snapshots and in practice are accessed > sequentially in most cases, so the effect of increasing the cache is > negligible (even when doing random writes from the guest). > > So, make the refcount cache as small as possible unless the user > explicitly asks for a larger one. I like the reasoning given here. I'd count this as a bugfix, safe even during freeze (but it's ultimately the maintainer's call) > > Signed-off-by: Alberto Garcia <berto@igalia.com> > --- > block/qcow2.c | 31 +++++++++++++++++++------------ > block/qcow2.h | 4 ---- > tests/qemu-iotests/137.out | 2 +- > 3 files changed, 20 insertions(+), 17 deletions(-) > +++ b/block/qcow2.c > @@ -802,23 +802,30 @@ 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 { > - *refcount_cache_size = combined_cache_size > - / (DEFAULT_L2_REFCOUNT_SIZE_RATIO + 1); > - *l2_cache_size = combined_cache_size - *refcount_cache_size; In the old code, refcount_cache_size and l2_cache_size are both set to fractions of the combined size, so both are positive (even if combined_cache_size is too small for the minimums required) > + uint64_t virtual_disk_size = bs->total_sectors * BDRV_SECTOR_SIZE; > + uint64_t max_l2_cache = virtual_disk_size / (s->cluster_size / 8); > + uint64_t min_refcount_cache = > + (uint64_t) MIN_REFCOUNT_CACHE_SIZE * s->cluster_size; > + > + /* 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) { > + *l2_cache_size = max_l2_cache; > + *refcount_cache_size = combined_cache_size - *l2_cache_size; > + } else { > + *refcount_cache_size = > + MIN(combined_cache_size, min_refcount_cache); but here, if combined_cache_size is smaller than min_refcount_cache, > + *l2_cache_size = combined_cache_size - *refcount_cache_size; then l2_cache_size is set to a negative value. I think you're missing bounds validations that the combined cache size is large enough for the minimums required. Or maybe a slight tweak, if it is okay for one of the two caches to be sized at 0 (that is, if combined_cache_size is too small for refcount, can it instead be given 100% to the l2 cache and let refcount be uncached)? > + } > } > } else { > - if (!l2_cache_size_set && !refcount_cache_size_set) { > + if (!l2_cache_size_set) { > *l2_cache_size = MAX(DEFAULT_L2_CACHE_BYTE_SIZE, > (uint64_t)DEFAULT_L2_CACHE_CLUSTERS > * s->cluster_size); > - *refcount_cache_size = *l2_cache_size > - / DEFAULT_L2_REFCOUNT_SIZE_RATIO; > - } else if (!l2_cache_size_set) { > - *l2_cache_size = *refcount_cache_size > - * DEFAULT_L2_REFCOUNT_SIZE_RATIO; > - } else if (!refcount_cache_size_set) { > - *refcount_cache_size = *l2_cache_size > - / DEFAULT_L2_REFCOUNT_SIZE_RATIO; > + } > + if (!refcount_cache_size_set) { > + *refcount_cache_size = MIN_REFCOUNT_CACHE_SIZE * s->cluster_size; > } > } >
On Tue 13 Mar 2018 07:23:36 PM CET, Eric Blake wrote: >> + *refcount_cache_size = >> + MIN(combined_cache_size, min_refcount_cache); > > but here, if combined_cache_size is smaller than min_refcount_cache, > >> + *l2_cache_size = combined_cache_size - *refcount_cache_size; > > then l2_cache_size is set to a negative value. No, it's set to 0. If combined == 4k and min_refcount == 256, then refcount_cache_size = MIN(4k, 256k) // 4k l2_cache_size = 4k - 4k; // 0 Then the caller ensures that it's always set to the minimum (as it did with the previous code). Berto
On 03/13/2018 01:48 PM, Alberto Garcia wrote: > On Tue 13 Mar 2018 07:23:36 PM CET, Eric Blake wrote: >>> + *refcount_cache_size = >>> + MIN(combined_cache_size, min_refcount_cache); >> >> but here, if combined_cache_size is smaller than min_refcount_cache, >> >>> + *l2_cache_size = combined_cache_size - *refcount_cache_size; >> >> then l2_cache_size is set to a negative value. > > No, it's set to 0. > > If combined == 4k and min_refcount == 256, then > > refcount_cache_size = MIN(4k, 256k) // 4k > l2_cache_size = 4k - 4k; // 0 Ah. Mental breakdown on my part in trying to compute (x - MIN()). > > Then the caller ensures that it's always set to the minimum (as it did > with the previous code). So the caller will use larger than the requested limits if the requested limits are too small, and we are okay with calculations resulting in 0 here. All right, thanks for stepping me through my error; you're good to go with: Reviewed-by: Eric Blake <eblake@redhat.com>
diff --git a/block/qcow2.c b/block/qcow2.c index 7472af6931..8342b0186f 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -802,23 +802,30 @@ 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 { - *refcount_cache_size = combined_cache_size - / (DEFAULT_L2_REFCOUNT_SIZE_RATIO + 1); - *l2_cache_size = combined_cache_size - *refcount_cache_size; + uint64_t virtual_disk_size = bs->total_sectors * BDRV_SECTOR_SIZE; + uint64_t max_l2_cache = virtual_disk_size / (s->cluster_size / 8); + uint64_t min_refcount_cache = + (uint64_t) MIN_REFCOUNT_CACHE_SIZE * s->cluster_size; + + /* 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) { + *l2_cache_size = max_l2_cache; + *refcount_cache_size = combined_cache_size - *l2_cache_size; + } else { + *refcount_cache_size = + MIN(combined_cache_size, min_refcount_cache); + *l2_cache_size = combined_cache_size - *refcount_cache_size; + } } } else { - if (!l2_cache_size_set && !refcount_cache_size_set) { + if (!l2_cache_size_set) { *l2_cache_size = MAX(DEFAULT_L2_CACHE_BYTE_SIZE, (uint64_t)DEFAULT_L2_CACHE_CLUSTERS * s->cluster_size); - *refcount_cache_size = *l2_cache_size - / DEFAULT_L2_REFCOUNT_SIZE_RATIO; - } else if (!l2_cache_size_set) { - *l2_cache_size = *refcount_cache_size - * DEFAULT_L2_REFCOUNT_SIZE_RATIO; - } else if (!refcount_cache_size_set) { - *refcount_cache_size = *l2_cache_size - / DEFAULT_L2_REFCOUNT_SIZE_RATIO; + } + if (!refcount_cache_size_set) { + *refcount_cache_size = MIN_REFCOUNT_CACHE_SIZE * s->cluster_size; } } diff --git a/block/qcow2.h b/block/qcow2.h index ccb92a9696..cdf41055ae 100644 --- a/block/qcow2.h +++ b/block/qcow2.h @@ -77,10 +77,6 @@ #define DEFAULT_L2_CACHE_CLUSTERS 8 /* clusters */ #define DEFAULT_L2_CACHE_BYTE_SIZE 1048576 /* bytes */ -/* The refblock cache needs only a fourth of the L2 cache size to cover as many - * clusters */ -#define DEFAULT_L2_REFCOUNT_SIZE_RATIO 4 - #define DEFAULT_CLUSTER_SIZE 65536 diff --git a/tests/qemu-iotests/137.out b/tests/qemu-iotests/137.out index e28e1eadba..96724a6c33 100644 --- a/tests/qemu-iotests/137.out +++ b/tests/qemu-iotests/137.out @@ -22,7 +22,7 @@ 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) -L2 cache size too big +Refcount cache size too big Conflicting values for qcow2 options 'overlap-check' ('constant') and 'overlap-check.template' ('all') 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
The L2 and refcount caches have default sizes that can be overriden using the l2-cache-size and refcount-cache-size (an additional parameter named cache-size sets the combined size of both caches). Unless forced by one of the aforementioned parameters, QEMU will set the unspecified sizes so that the L2 cache is 4 times larger than the refcount cache. This is based on the premise that the refcount metadata needs to be only a fourth of the L2 metadata to cover the same amount of disk space. This is incorrect for two reasons: a) The amount of disk covered by an L2 table depends solely on the cluster size, but in the case of a refcount block it depends on the cluster size *and* the width of each refcount entry. The 4/1 ratio is only valid with 16-bit entries (the default). b) When we talk about disk space and L2 tables we are talking about guest space (L2 tables map guest clusters to host clusters), whereas refcount blocks are used for host clusters (including L1/L2 tables and the refcount blocks themselves). On a fully populated (and uncompressed) qcow2 file, image size > virtual size so there are more refcount entries than L2 entries. Problem (a) could be fixed by adjusting the algorithm to take into account the refcount entry width. Problem (b) could be fixed by increasing a bit the refcount cache size to account for the clusters used for qcow2 metadata. However this patch takes a completely different approach and instead of keeping a ratio between both cache sizes it assigns as much as possible to the L2 cache and the remainder to the refcount cache. The reason is that L2 tables are used for every single I/O request from the guest and the effect of increasing the cache is significant and clearly measurable. Refcount blocks are however only used for cluster allocation and internal snapshots and in practice are accessed sequentially in most cases, so the effect of increasing the cache is negligible (even when doing random writes from the guest). So, make the refcount cache as small as possible unless the user explicitly asks for a larger one. Signed-off-by: Alberto Garcia <berto@igalia.com> --- block/qcow2.c | 31 +++++++++++++++++++------------ block/qcow2.h | 4 ---- tests/qemu-iotests/137.out | 2 +- 3 files changed, 20 insertions(+), 17 deletions(-)