As found by @KochetovNicolai before this patch, lambda in
ThreadFromGlobalPool() ctor assigns value only to a copy of the
thread_id value, and so check in joinable() had been working
incorrectly, fix this by changing the value not the shared_ptr itself.
Also it is not safe to assign thread_id w/o atomics, since this can be
racy, so wrap id with std::atomic<>
Fixes: #28431
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
joinable() should be used only outside, since internally it is enough to
know `state` to know that something is wrong.
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
A simple HelloWorld program with zero includes except iostream triggers
a build of ca. 2000 source files. The reason is that ClickHouse's
top-level CMakeLists.txt overrides "add_executable()" to link all
binaries against "clickhouse_new_delete". This links against
"clickhouse_common_io", which in turn has lots of 3rd party library
dependencies ... Without linking "clickhouse_new_delete", the number of
compiled files for "HelloWorld" goes down to ca. 70.
As an example, the self-extracting-executable needs none of its current
dependencies but other programs may also benefit.
In order to restore access to the original "add_executable()", the
overriding version is now prefixed. There is precedence for a
"clickhouse_" prefix (as opposed to "ch_"), for example
"clickhouse_split_debug_symbols". In general prefixing makes sense also
because overriding CMake commands relies on undocumented behavior and is
considered not-so-great practice (*).
(*) https://crascit.com/2018/09/14/do-not-redefine-cmake-commands/
When WRITE lock attemp fails (exclusive lock for ALTER/DELETE), and
there are multiple READ locks (shared lock for SELECT/INSERT), i.e. one
INSERT is in progress and one SELECT is queued after ALTER/DELETE
started but before it fails, this SELECT will wait until INSERT will
finishes.
This happens because in case of WRITE lock failure it does not notify
the next READ lock that can be acquired.
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
SQL functions countSubstrings(), countSubstringsCaseInsensitive(),
countSubstringsUTF8(), position(), positionCaseInsensitive(),
positionUTF8() with non-const pattern argument use fallback sorters
LibCASCIICaseSensitiveStringSearcher and LibCASCIICaseInsensitiveStringSearcher
which call ::strstr(), resp. ::strcasestr(). These functions assume that
the haystack is 0-terminated and they even document that. However, the
callers did not check if the haystack contains 0-byte (perhaps because
its sort of expensive). As a consequence, if the haystack contained a
zero byte in it's payload, matches behind this zero byte were ignored.
create table t (id UInt32, pattern String) engine = MergeTree() order by id;
insert into t values (1, 'x');
select countSubstrings('aaaxxxaa\0xxx', pattern) from t;
We returned 3 before this commit, now we return 6
