Суббота, 4 июля 2020, 0:49 +03:00 от Vladislav Shpilevoy <v.shpilevoy@tarantool.org>:
 

Hi! Thanks for the patch!

See 2 comments below.

On 03/07/2020 15:32, Ilya Kosarev wrote:
> Since 9fba29abb4e05babb9b23b4413bd8083f0fba933 (memtx: introduce tuple
> compare hint) index_get uses key_hint to get a comparison hint for the
> key. It requires key type to be validated. However in
> space_before_replace index_get was used without key validation on some
> execution paths. Now it is fixed and validation is being performed if
> needed. Corresponding test case is introduced.
>
> Closes #5093
> ---
> Branch: https://github.com/tarantool/tarantool/tree/i.kosarev/gh-5093-before_replace-key-validation
> Issue: https://github.com/tarantool/tarantool/issues/5093
>
> @ChangeLog:
> * Add needed key validation to space_before_replace.
>
> src/box/space.c | 6 +++---
> .../gh-5093-before_replace-key-type.result | 21 +++++++++++++++++++
> .../gh-5093-before_replace-key-type.test.lua | 7 +++++++
> 3 files changed, 31 insertions(+), 3 deletions(-)
> create mode 100644 test/engine/gh-5093-before_replace-key-type.result
> create mode 100644 test/engine/gh-5093-before_replace-key-type.test.lua
>
> diff --git a/src/box/space.c b/src/box/space.c
> index eecbde7fa7..66768208b3 100644
> --- a/src/box/space.c
> +++ b/src/box/space.c
> @@ -359,9 +359,6 @@ space_before_replace(struct space *space, struct txn *txn,
> return -1;
> key = request->key;
> part_count = mp_decode_array(&key);
> - if (exact_key_validate(index->def->key_def,
> - key, part_count) != 0)
> - return -1;
> break;
> case IPROTO_INSERT:
> case IPROTO_REPLACE:
> @@ -380,6 +377,9 @@ space_before_replace(struct space *space, struct txn *txn,
> /* Unknown request type, nothing to do. */
> return 0;
> }
> + if (key != NULL &&
> + exact_key_validate(index->def->key_def, key, part_count) != 0)
> + return -1;

1. The patch is technically correct. But I don't like that you
add key != NULL check to the hot path. Even though it is always != NULL
except blackhole engine space, which doesn't have indexes either.

Consider the refactoring below. I know it causes bigger diff, but
I consider it ok for the sake performance. As we learned recently,
even +1 or 2 ifs in a hot path may cost.

====================
diff --git a/src/box/space.c b/src/box/space.c
index 66768208b..73172ddd2 100644
--- a/src/box/space.c
+++ b/src/box/space.c
@@ -347,6 +347,7 @@ space_before_replace(struct space *space, struct txn *txn,
  const char *key = NULL;
  uint32_t part_count = 0;
  struct index *index = NULL;
+ struct tuple *old_tuple = NULL;
 
  /*
  * Lookup the old tuple.
@@ -364,7 +365,7 @@ space_before_replace(struct space *space, struct txn *txn,
  case IPROTO_REPLACE:
  case IPROTO_UPSERT:
  if (pk == NULL)
- break;
+ goto after_getting_old;
  index = pk;
  key = tuple_extract_key_raw(request->tuple, request->tuple_end,
  index->def->key_def, MULTIKEY_NONE,
@@ -377,15 +378,11 @@ space_before_replace(struct space *space, struct txn *txn,
  /* Unknown request type, nothing to do. */
  return 0;
  }
- if (key != NULL &&
- exact_key_validate(index->def->key_def, key, part_count) != 0)
+ if (exact_key_validate(index->def->key_def, key, part_count) != 0)
  return -1;
-
- struct tuple *old_tuple = NULL;
- if (index != NULL &&
- index_get(index, key, part_count, &old_tuple) != 0)
+ if (index_get(index, key, part_count, &old_tuple) != 0)
  return -1;
-
+after_getting_old:;
  /*
  * Create the new tuple.
  */
====================

> diff --git a/test/engine/gh-5093-before_replace-key-type.test.lua b/test/engine/gh-5093-before_replace-key-type.test.lua
> new file mode 100644
> index 0000000000..9cba3c1597
> --- /dev/null
> +++ b/test/engine/gh-5093-before_replace-key-type.test.lua
> @@ -0,0 +1,7 @@
> +test_run = require('test_run').new()
> +trigger = function() end
> +
> +s = box.schema.space.create('gh-5093', {engine=test_run:get_cfg('engine')})
> +_ = s:create_index('value', {parts={{1, type='decimal'}}})
> +_ = s:before_replace(trigger)
> +s:replace{'1111.1111'}

2. Please, cleanup. Do s:drop().