Not actually implemented, though. It does print out some jit-compiled stuff,
but that's about it. For example, this query:
select number from system.numbers where something(cast(number as Float64)) == 4
results in this on server's stderr:
define double @"something(CAST(number, 'Float64'))"(void**, i8*, void*) {
"something(CAST(number, 'Float64'))":
ret double 1.234500e+04
}
(and an exception, because that's what the non-jitted method does.)
As one may notice, this function neither reads the input (first argument;
tuple of arrays) nor writes the output (third argument; array), instead
returning some general nonsense.
In addition, `#if USE_EMBEDDED_COMPILER` doesn't work for some reason,
including LLVM headers requires -Wno-unused-parameter, this probably only
works on LLVM 5.0 due to rampant API instability, and I'm definitely
no expert on CMake. In short, there's still a long way to go.
* Macos test fix
* Fix freebsd build
* CLICKHOUSE-3553 Rebuild loggers on every config change
* rename clickhouse_common_configprocessor -> clickhouse_common_config; move Server/ConfigReloader.cpp -> Common/Config/
* CLICKHOUSE-3553 Set new level to all loggers on reload
* Dont use hash
* Update BaseDaemon.h
* Allow build without rt library
* fix
* Cmake misc
* CLICKHOUSE-3600 Use correct lld version
* CLICKHOUSE-3600 Use correct lld version
* Use libllvm same version as clang
* fix
* Fix build with iodbc
* Fix iodbc build
* Fix gold find
* Fix empty lines
* Make separate libs Common/ZooKeeper and Common/ConfigProcessor
renamed: dbms/src/Common/StringUtils.h -> dbms/src/Common/StringUtils/StringUtils.h
* Fix space
* Fix data in system.build_options
* merge
* merge
Interpreters/Compiler.cpp contained hard-coded paths for system's
includes needed for query compiler. These paths were not portable
between different Linux distros and gcc/clang versions. For example,
Debian/Ubuntu use /usr/lib/gcc/x86_64-linux-gnu/*/include,
RHEL/Fedora use /usr/lib/gcc/x86_64-redhat-linux/*/include,
others use /usr/lib/gcc/*/include (without x86_64-XXX triplet).
Patch 68850012b "Embedded compiler fixes" attempted to fix this problem
by adding CMAKE_LIBRARY_ARCHITECTURE after /usr/lib. Unfortunally,
CMAKE_LIBRARY_ARCHITECTURE is not defined on RHEL/Fedora because someone
decided to omit "-gnu" from x86_64-redhat-linux (see RHBZ#1531678).
Patch 70e35d0bc "Build fixes (#1718)" added a workaround for
undefined CMAKE_LIBRARY_ARCHITECTURE on RHEL/Fedora, but hasn't fixed
problem with missing /usr/lib/gcc/x86_64-redhat-linux/*/include/
in the list of hardcoded paths.
Remove hard-coded paths and get the list of `-isystem` includes directly
from bundled clickhouse-clang.
Other changes:
- Enable RPATH for the build directory to get working binaries
without installing them by `make install`.
Cap'n Proto is a binary message format.
Like Protocol Buffers and Thrift (but unlike JSON or MessagePack), Cap'n Proto messages are strongly-typed and not self-describing. Due to this, it requires a schema setting to specify schema file and the root object. The schema is parsed on runtime and cached for each SQL statement.
* fix segfault on ATTACH MATERIALIZED VIEW
* allow only TEMPORARY TABLEs (not databases or views)
* forbid choosing ENGINE for TEMPORARY tables
* fix segfault on CREATE TABLE t1 AS t2 where t2 is a View
* Log query id in executeQuery; Better type mismatch error; change format in report tool
* Better log query_id
* fix message
* Use c++11 thread_local instaed of gcc's __thread
* lock mutex before notifying waiting thread in sync insertion into distributed [#CLICKHOUSE-3379]
* Cmake: fix build without downloaded submodules (#1379)
* fix
* ZooKeeper: fixed stack smashing with tryGet()
The tryGet() operation creates a 1MB buffer on stack. This may or
may not work depending on the default stack size for threads,
whether the stack protector is enabled or not, recursion depth,
and the actual value size.
This is probably going to slow down some ZK operations, but I don't
see how else this could work reliably with the existing API.
* increased timeout for test_insertion_sync_fails_with_timeout
* Update CHANGELOG_RU.md
* Update ZooKeeper.cpp
* Fix warnings
* Fixes
* Dont strip debug info from asan, tsan and other builds except releases
* Fix asan error causd by test 00144
* Fix empty log message (#CLICKHOUSE-3378)
In the current version of Poco, unsigned long no longer aliases to
UInt64 with LP64. The size_t aliases to unsigned long with clang,
so all the uses of size_t instead of UInt64 when interacting with
Poco interfaces are gone. I replaced uses with UInt64 where it makes
sense, and added an overloaded function for readVarUInt() to support size_t.
