[tarantool-patches] Re[2]: [tarantool-patches] Re: [PATCH 1/3] sql: EXPLAIN through net.box leads to SEGFAULT
Мерген Имеев
imeevma at tarantool.org
Thu Nov 22 23:55:22 MSK 2018
After discussion we decided that this patch will serve us
part of patch-set that should close newly filed bug:
https://github.com/tarantool/tarantool/issues/3832
The main Idea behind this ticket is that any statement
should return column type for when it returns column name.
And this ticket allow us to find an error without falling
into SEGMENTATION FAULT or returning less metadata.
Четверг, 22 ноября 2018, 21:31 +03:00 от Vladislav Shpilevoy <v.shpilevoy at tarantool.org>:
>
>It was just a question, "why did you decide to do not
>send type". It was not necessary to start sending UNKNOWN. I
>think, that unknown type should be sent never.
>
>On 22/11/2018 21:08, Imeev Mergen wrote:
>> Hi! Thank you for review! Diff between versions and new patch
>> below.
>>
>> On 11/19/18 8:27 PM, n.pettik wrote:
>>
>>>> EXPLAIN envokes SEGMENTATION FAULT when being executed through
>>> Typo: invokes.
>>>
>>>> net.box. It happens due to column type of the result of this
>>>> function being NULL.
>>> I’ve checked and not only EXPLAIN results, but also EXPLAIN QUERY PLAN,
>>> several pragmas (for instance, pragma table_info()) etc. So, initially problem
>>> was not only in EXPLAIN statement. Please exposed commit message and
>>> add tests on other cases.
>>>
>>> Also, why did you decide to avoid sending type at all, instead of sending
>>> UNKNOWN for example?
>>
>> *Diff between versions:*
>>
>> diff --git a/src/box/execute.c b/src/box/execute.c
>> index 5b747cf..cd809f8 100644
>> --- a/src/box/execute.c
>> +++ b/src/box/execute.c
>> @@ -529,6 +529,9 @@ sql_get_description(struct sqlite3_stmt *stmt, struct obuf *out,
>> return -1;
>>
>> for (int i = 0; i < column_count; ++i) {
>> + size_t size = mp_sizeof_map(2) +
>> + mp_sizeof_uint(IPROTO_FIELD_NAME) +
>> + mp_sizeof_uint(IPROTO_FIELD_TYPE);
>> const char *name = sqlite3_column_name(stmt, i);
>> const char *type = sqlite3_column_datatype(stmt, i);
>> /*
>> @@ -537,27 +540,20 @@ sql_get_description(struct sqlite3_stmt *stmt, struct obuf *out,
>> * column_name simply returns them.
>> */
>> assert(name != NULL);
>> - size_t size = mp_sizeof_uint(IPROTO_FIELD_NAME) +
>> - mp_sizeof_str(strlen(name));
>> - if (type != NULL) {
>> - size += mp_sizeof_map(2) +
>> - mp_sizeof_uint(IPROTO_FIELD_TYPE) +
>> - mp_sizeof_str(strlen(type));
>> - } else {
>> - size += mp_sizeof_map(1);
>> - }
>> + if (type == NULL)
>> + type = "UNKNOWN";
>> + size += mp_sizeof_str(strlen(name));
>> + size += mp_sizeof_str(strlen(type));
>> char *pos = (char *) obuf_alloc(out, size);
>> if (pos == NULL) {
>> diag_set(OutOfMemory, size, "obuf_alloc", "pos");
>> return -1;
>> }
>> - pos = mp_encode_map(pos, 1 + (type != NULL));
>> + pos = mp_encode_map(pos, 2);
>> pos = mp_encode_uint(pos, IPROTO_FIELD_NAME);
>> pos = mp_encode_str(pos, name, strlen(name));
>> - if (type != NULL) {
>> - pos = mp_encode_uint(pos, IPROTO_FIELD_TYPE);
>> - pos = mp_encode_str(pos, type, strlen(type));
>> - }
>> + pos = mp_encode_uint(pos, IPROTO_FIELD_TYPE);
>> + pos = mp_encode_str(pos, type, strlen(type));
>> }
>> return 0;
>> }
>> diff --git a/src/box/lua/net_box.c b/src/box/lua/net_box.c
>> index 342c6a7..c7063d9 100644
>> --- a/src/box/lua/net_box.c
>> +++ b/src/box/lua/net_box.c
>> @@ -644,7 +644,8 @@ netbox_decode_metadata(struct lua_State *L, const char **data)
>> lua_createtable(L, count, 0);
>> for (uint32_t i = 0; i < count; ++i) {
>> uint32_t map_size = mp_decode_map(data);
>> - assert(map_size == 1 || map_size == 2);
>> + assert(map_size == 2);
>> + (void) map_size;
>> uint32_t key = mp_decode_uint(data);
>> assert(key == IPROTO_FIELD_NAME);
>> (void) key;
>> @@ -653,13 +654,11 @@ netbox_decode_metadata(struct lua_State *L, const char **data)
>> const char *str = mp_decode_str(data, &len);
>> lua_pushlstring(L, str, len);
>> lua_setfield(L, -2, "name");
>> - if (map_size == 2) {
>> - key = mp_decode_uint(data);
>> - assert(key == IPROTO_FIELD_TYPE);
>> - const char *type = mp_decode_str(data, &len);
>> - lua_pushlstring(L, type, len);
>> - lua_setfield(L, -2, "type");
>> - }
>> + key = mp_decode_uint(data);
>> + assert(key == IPROTO_FIELD_TYPE);
>> + const char *type = mp_decode_str(data, &len);
>> + lua_pushlstring(L, type, len);
>> + lua_setfield(L, -2, "type");
>> lua_rawseti(L, -2, i + 1);
>> }
>> }
>> diff --git a/test/sql/iproto.result b/test/sql/iproto.result
>> index 31e5d2e..f229427 100644
>> --- a/test/sql/iproto.result
>> +++ b/test/sql/iproto.result
>> @@ -774,9 +774,9 @@ cn:close()
>> ---
>> ...
>> --
>> --- EXPLAIN could lead to segfault because it does not send column
>> --- data type, but looks like DQL. Netbox had been trying to decode
>> --- type for any DQL.
>> +-- Some commands, e.g. EXPLAIN, could lead to segfault because it
>> +-- does not send column data type, but looks like DQL. Netbox had
>> +-- been trying to decode type for any DQL.
>> --
>> cn = remote.connect(box.cfg.listen)
>> ---
>> @@ -785,6 +785,32 @@ cn = remote.connect(box.cfg.listen)
>> _ = cn:execute("EXPLAIN SELECT 1;")
>> ---
>> ...
>> +-- Segmentation fault when PRAGMA TABLE_INFO() executed using
>> +-- net.box.
>> +box.sql.execute('CREATE TABLE test (id INT PRIMARY KEY)')
>> +---
>> +...
>> +cn:execute("PRAGMA TABLE_INFO(test);")
>> +---
>> +- metadata:
>> + - name: cid
>> + type: UNKNOWN
>> + - name: name
>> + type: UNKNOWN
>> + - name: type
>> + type: UNKNOWN
>> + - name: notnull
>> + type: UNKNOWN
>> + - name: dflt_value
>> + type: UNKNOWN
>> + - name: pk
>> + type: UNKNOWN
>> + rows:
>> + - [0, 'ID', 'integer', 1, null, 1]
>> +...
>> +box.sql.execute('DROP TABLE test')
>> +---
>> +...
>> cn:close()
>> ---
>> ...
>> diff --git a/test/sql/iproto.test.lua b/test/sql/iproto.test.lua
>> index a0f0f15..9bcc7ef 100644
>> --- a/test/sql/iproto.test.lua
>> +++ b/test/sql/iproto.test.lua
>> @@ -248,15 +248,21 @@ cn:execute('drop table test')
>> cn:close()
>>
>> --
>> --- EXPLAIN could lead to segfault because it does not send column
>> --- data type, but looks like DQL. Netbox had been trying to decode
>> --- type for any DQL.
>> +-- Some commands, e.g. EXPLAIN, could lead to segfault because it
>> +-- does not send column data type, but looks like DQL. Netbox had
>> +-- been trying to decode type for any DQL.
>> --
>> cn = remote.connect(box.cfg.listen)
>>
>> -- Segmentation fault when EXPLAIN executed using net.box.
>> _ = cn:execute("EXPLAIN SELECT 1;")
>>
>> +-- Segmentation fault when PRAGMA TABLE_INFO() executed using
>> +-- net.box.
>> +box.sql.execute('CREATE TABLE test (id INT PRIMARY KEY)')
>> +cn:execute("PRAGMA TABLE_INFO(test);")
>> +box.sql.execute('DROP TABLE test')
>> +
>> cn:close()
>>
>> box.schema.user.revoke('guest', 'read,write,execute', 'universe')
>>
>>
>> *New version:*
>>
>> commit 29dde0365922bfcec0d159b36c6454854adf34f7
>> Author: Mergen Imeev < imeevma at gmail.com >
>> Date: Fri Nov 16 13:34:46 2018 +0300
>>
>> sql: Print column type even if it is NULL
>>
>> Some commands, e.g. EXPLAIN, invokes SEGMENTATION FAULT when being
>> executed through net.box. It happens due to column type of the
>> result of this function being NULL. This patch adds checks for
>> received column type.
>>
>> diff --git a/src/box/execute.c b/src/box/execute.c
>> index e450444..cd809f8 100644
>> --- a/src/box/execute.c
>> +++ b/src/box/execute.c
>> @@ -540,6 +540,8 @@ sql_get_description(struct sqlite3_stmt *stmt, struct obuf *out,
>> * column_name simply returns them.
>> */
>> assert(name != NULL);
>> + if (type == NULL)
>> + type = "UNKNOWN";
>> size += mp_sizeof_str(strlen(name));
>> size += mp_sizeof_str(strlen(type));
>> char *pos = (char *) obuf_alloc(out, size);
>> diff --git a/test/sql/iproto.result b/test/sql/iproto.result
>> index d077ee8..f229427 100644
>> --- a/test/sql/iproto.result
>> +++ b/test/sql/iproto.result
>> @@ -773,6 +773,47 @@ cn:execute('drop table test')
>> cn:close()
>> ---
>> ...
>> +--
>> +-- Some commands, e.g. EXPLAIN, could lead to segfault because it
>> +-- does not send column data type, but looks like DQL. Netbox had
>> +-- been trying to decode type for any DQL.
>> +--
>> +cn = remote.connect(box.cfg.listen)
>> +---
>> +...
>> +-- Segmentation fault when EXPLAIN executed using net.box.
>> +_ = cn:execute("EXPLAIN SELECT 1;")
>> +---
>> +...
>> +-- Segmentation fault when PRAGMA TABLE_INFO() executed using
>> +-- net.box.
>> +box.sql.execute('CREATE TABLE test (id INT PRIMARY KEY)')
>> +---
>> +...
>> +cn:execute("PRAGMA TABLE_INFO(test);")
>> +---
>> +- metadata:
>> + - name: cid
>> + type: UNKNOWN
>> + - name: name
>> + type: UNKNOWN
>> + - name: type
>> + type: UNKNOWN
>> + - name: notnull
>> + type: UNKNOWN
>> + - name: dflt_value
>> + type: UNKNOWN
>> + - name: pk
>> + type: UNKNOWN
>> + rows:
>> + - [0, 'ID', 'integer', 1, null, 1]
>> +...
>> +box.sql.execute('DROP TABLE test')
>> +---
>> +...
>> +cn:close()
>> +---
>> +...
>> box.schema.user.revoke('guest', 'read,write,execute', 'universe')
>> ---
>> ...
>> diff --git a/test/sql/iproto.test.lua b/test/sql/iproto.test.lua
>> index 1470eda..9bcc7ef 100644
>> --- a/test/sql/iproto.test.lua
>> +++ b/test/sql/iproto.test.lua
>> @@ -247,10 +247,27 @@ cn:execute('drop table test')
>>
>> cn:close()
>>
>> +--
>> +-- Some commands, e.g. EXPLAIN, could lead to segfault because it
>> +-- does not send column data type, but looks like DQL. Netbox had
>> +-- been trying to decode type for any DQL.
>> +--
>> +cn = remote.connect(box.cfg.listen)
>> +
>> +-- Segmentation fault when EXPLAIN executed using net.box.
>> +_ = cn:execute("EXPLAIN SELECT 1;")
>> +
>> +-- Segmentation fault when PRAGMA TABLE_INFO() executed using
>> +-- net.box.
>> +box.sql.execute('CREATE TABLE test (id INT PRIMARY KEY)')
>> +cn:execute("PRAGMA TABLE_INFO(test);")
>> +box.sql.execute('DROP TABLE test')
>> +
>> +cn:close()
>> +
>> box.schema.user.revoke('guest', 'read,write,execute', 'universe')
>> box.schema.user.revoke('guest', 'create', 'space')
>> space = nil
>>
>> -- Cleanup xlog
>> box.snapshot()
>> -
>>
-- <br/>Мерген Имеев
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.tarantool.org/pipermail/tarantool-patches/attachments/20181122/a20b79ca/attachment.html>
More information about the Tarantool-patches
mailing list