Since the removal of jerry-libc, only jerry-ext uses the
`FEATURE_INIT_FINI` optional feature. Thus, it's better to move it
from the global cmakelists to that of jerry-ext.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
Rationale:
- There is no port under targets/ that would use it. All of them
turn it off when building.
- That's no surprise, as jerry-libc supports no barebone MCUs but
posix targets with syscalls only. Actually, that's Linux only,
because macOS builds have turned off the use of jerry-libc a
while ago.
- And there is no point in maintaining a highly restricted set of
libc functions: as soon as someone wants to use JerryScript in a
scenario that needs more functions than jerry-main, they have to
choose a different libc (most problably the compiler's default
one).
I think that we should not keep supporting an otherwise unused
library for the purposes of jerry-main on arm/x86/x64-linux only.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
The patch also ensures that all components access only the public
headers of other components (except for unit tests, which are
allowed to use private headers, too).
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
In `tools/build.py`:
- For the sake of readability, group CLI arguments as general build
options, options to control the building of components, and
component-specific options.
- To prevent duplications, remove the defaults from those CLI
arguments that correspond to CMake options and have defaults in
any of the CMakeLists. Should any of the defaults change, they
will have to be changed at a single place only. (Those options
that are not set on the command line of `tools/build.py` are not
passed as options to `cmake` either.)
- Convert `--unittests` and `--doctests` to ON/OFF options like the
rest of the component switches.
- Touch on some of the help messages of the CLI arguments.
Other changes:
- The change in `--unittests` and `--doctests` is a slightly CLI-
breaking change of `tools/build.py`. Thus, follow up on this in
`tools/run-tests.py`.
- Move `ENABLE_ALL_IN_ONE` into `jerry-core` as it is not a general
option but specific to that component.
- Remove the forcing of `ENABLE_ALL_IN_ONE` for some compilers/
platforms as it is still an option, not a hard requirement.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
- Add `--install` option to `tools/build.py`.
- Make use of `--install` in `tools/run-tests.py` by testing the
installed the executables instead of those in the build tree.
Related changes:
- Collect unit test binaries in the `tests` subdir of the build
tree instead of `bin`.
- The `ls`-based collection of the unit test binaries had some
shortcomings hitherto unrevealed (it didn't filter for files so
it could potentially "collect" dictionaries, too), which has now
been replaced with a more stable `find`-based solution.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
The cmakelists contained various compiler-specific parts scattered
across the file. This patch moves them to as few conditional blocks
as possible.
Additional changes:
- `-Werror` does not depend on whether jerry-libc is enabled.
- Some stylistic fixes.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
- Status messages in CMakeLists.txt files got misaligned, fixing.
- An extra space sneaked in before an `#ifdef`, fixing.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
Remove TARGET_HOST defines from the jerry-libc module and replace with compiler provided macros.
JerryScript-DCO-1.0-Signed-off-by: Istvan Miklos imiklos2@inf.u-szeged.hu
Some benchmark suites contain test cases that have nonreproducible
behaviour. This is mostly caused by relying on "now" when dealing
with dates or timestamps, instead of using a fixed moment. (A
notorious example is the crypto-aes.js test case of the sunspider
bechmark suite, where the heap memory consumption can vary between
34K-41K heap because of using `(new Date()).getTime()`.)
This commit renames the jerry-minimal command line tool to
jerry-test (to better reflect its purpose) and adds extra code,
which intercepts some calls to libc (`gettimeofday`, `rand`) and
pins their results to some fixed values. This makes the tool
useless in a general case but ensures stable results when
benchmarking -- for which it is mostly used.
As a side effect, the commit also changes jerry-libc by making all
libc functions weak symbols to allow their override from
application code.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
This patch adds an extension to snapshots which allows storing
multiple position independent primary functions in a single
snapshot data. A new application called jerry-snapshot is
added to the project to manage snapshots. Currently the only
option is merging snapshots.
JerryScript-DCO-1.0-Signed-off-by: Zoltan Herczeg zherczeg.u-szeged@partner.samsung.com
By using constructors/destructors we unify the case of static linking with
that of dynamic linking, and we reuse the build flag FEATURE_INIT_FINI. Using
constructors/destructors also allows us to cover the case where library
constructor/destructor functionality is unavailable, because we can expose the
module registration/unregistration functions as global symbols, to be called
explicitly from within the application.
Fixes https://github.com/jerryscript-project/jerryscript/issues/1952
JerryScript-DCO-1.0-Signed-off-by: Gabriel Schulhof gabriel.schulhof@intel.com
Markdown files in the docs/ directory can now be annotated to turn
fenced C code blocks into unit tests. The recognized syntax is:
[doctest]: # (name="test.c", test="run")
```c
// unit test code
```
The commit also fixes the issues revealed during the initial
annotation.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
are not accepted. This is a bug in gcc 7.
For now disable the fallthrough comment detection.
Disabling the check does not introduce any risk
as previously it was not enabled by default and
vera++ already check these kind of comments.
JerryScript-DCO-1.0-Signed-off-by: Peter Gal pgal.u-szeged@partner.samsung.com
Removed various superfluous declarations, and rewrote some parts to
be nicer or more consistent.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
Added `handler` module to `jerry-ext` to contain implementation of
commonly used external function handlers: `assert`, `gc`, and
`print`.
Also adapted jerry-main to use jerry-ext/handler
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
It provides some APIs for binding developers, so that
they can validate the type of the js argument and convert/assign them
to the native argument.
Related Issue: #1716
JerryScript-DCO-1.0-Signed-off-by: Martijn The martijn.the@intel.com
JerryScript-DCO-1.0-Signed-off-by: Zidong Jiang zidong.jiang@intel.com
PR #1505 added support for TI compiler. It explicitly added a
message to notify the user that static linking is forced. PR #1755
added a more generic approach to signal such forced settings and
adapted the TI-specific static linking notification to this
approach. However, it turned out that TI forcibly changed another
setting, too: it disabled release binary stripping, but without
notification. This patch fixes this by moving the setting override
to a consistent place and adding a notification.
PR #1505 also added some source code changes, most importantly a
complex struct initialization for a variable in
`ecma-objects-general.c`. However, that initialization was coded
as a macro to trick the style checker. This patch gets rid of that
macro and uses proper C99 struct initializer with designators.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
This commit changes the concept of JerryScript port implementations
from a simple directory of C source files (which get injected among
the sources of `jerry-core`) into a proper static library (which
may be linked to an application together with `jerry-core`). As a
consequence, this commit introduces a new library to the
JerryScript component architecture: the sources of the default port
implementation form `jerry-port-default`.
Changes in more detail:
- The sources in `targets/default` are moved to `jerry-port/default`
and are turned into a proper static library.
- Actually, the default port implementation has two library
variants, one that implements the bare minimum only
(`jerry-port-default-minimal`) and one that has some extra
functionalities specific to this implementation (the "full"
`jerry-port-default`).
- The new libraries have an interface header in
`jerry-port/default/include`, which extends the common
`jerryscript-port.h` API with functions specific to these
libraries.
- All non-standard port functions have now the
`jerry_port_default_` prefix (this affects `jobqueue_init` and
`jobqueue_run`).
- The jobqueue implementation functions became config macro
independent: it is now the responsibility of the linker to pick
the needed objects from the library, and omit those (e.g.,
jobqueue-related code) that are not referenced.
- Build of the libraries can be controlled with the new
`JERRY_PORT_DEFAULT` cmake option.
- The cmake option `PORT_DIR` is dropped, and `PORT_DIR/*.c` is not
appended to `jerry-core` sources.
- Instead, the `jerry` tool of `jerry-main` links to
`jerry-port-default`, while `jerry-minimal` links to
`jerry-port-default-minimal`.
- `tests/unit-core` tests are also linked to
`jerry-port-default-minimal`.
- Tools adapted.
- `build.py` has `--jerry-port-default` instead of `--port-dir`.
- `check-*.sh` have paths updated (`jerry-port/default` instead
of `targets/default`).
- Miscellaneous.
- Dropped `#ifndef`s from `jerryscript-port.h`. It is a public
header of the `jerry-core` library, which means that it must
not contain configuration-dependent parts (once the library is
built with some config macros and the archive and the headers
are installed, there is no way for the header to tell what
those config macrose were).
- Added documentation comments to the JobQueue Port API (in
`jerryscript-port.h`) and to several default port
implementation functions (in `jerry-port/default`).
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
The unit tests should follow the component structure, so this patch
moves all `jerry-core` unit tests under `tests/unit-core` and the
`jerry-libm` unit tests under `tests/unit-libm`.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
Currently, the Darwin platform and the TI compiler force some of
the settings to predefined values. This patch ensures that these
overrides are highlighted during build.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
Disable release binary stripping on OSX: The linker of OSX does not
support `-s` for link-time stripping. (The option does not cause an
error but clutters build output with warnings.)
No OSX-dependent LTO flags: LTO is disabled on OSX.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
Function definitions with no parameters should always use the void keyword to allow the compiler to catch invalid calls of those functions at compile time.
Enable -Wold-style-definition to catch this early in the future.
Fixes#1649.
JerryScript-DCO-1.0-Signed-off-by: Tilmann Scheller t.scheller@samsung.com
Jerryscript. The changes include:
CMakeLists.txt: we added conditionals around GCC-specific flags, added
support for TI flags, and removed the always-on debugging flag (-g)
/toolchain_mcu_tim4f.cmake: new toolchain file uses TI-specific parameters
jerry-api.h: the sys/types.h file was #include'd but never used, so we
removed it
jrt-types.h: ditto
jerry-port-default-date.c: the TI toolchain doesn't include
sys/time.h, so we guarded uses of the package
ecma-objects-general.c: added initialization that Travis (the
auto-checking tool) required
JerryScript-DCO-1.0-Signed-off-by: Timothy Harvey t-harvey@ti.com
Since the project is now hosted at the JS Foundation we can move to unified copyright notices for the project.
Starting with this commit all future contributions to the project should only carry the following copyright notice (except for third-party code which requires copyright information to be preserved):
"Copyright JS Foundation and other contributors, http://js.foundation" (without the quotes)
This avoids cluttering the codebase with contributor-specific copyright notices which have a higher maintenance overhead and tend to get outdated quickly. Also dropping the year from the copyright notices helps to avoid yearly code changes just to update the copyright notices.
Note that each contributor still retains full copyright ownership of his/her contributions and the respective authorship is tracked very accurately via Git.
JerryScript-DCO-1.0-Signed-off-by: Tilmann Scheller t.scheller@samsung.com
It now has the same effect as defining _DEFAULT_SOURCE, but generates a
compile-time warning (unless _DEFAULT_SOURCE is also defined). To allow
code that requires _BSD_SOURCE in glibc 2.19 and earlier and _DEFAULT_SOURCE
in glibc 2.20 and later to compile without warnings, define both _BSD_SOURCE and _DEFAULT_SOURCE.
JerryScript-DCO-1.0-Signed-off-by: Robert Sipka rsipka.uszeged@partner.samsung.com
`-g` is enough (most modern compilers use DWARF4 as default, at
least, e.g., GCC has switched from DWARF2 in 4.8), more future
proof (if a compiler moves to a newer version, it wont get forced
to a fixed older version), and less restrictive (some compilers
handle the "standard" `-g` only but don't understand its variants).
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
* Cmake had LTO OFF, build.py had LTO ON by default. ON became the
common setting.
* Cmake had snapshot save/exec OFF, build.py had them ON by
default. OFF became the common setting.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
Although both jerry-libc and jerry-libm have configuration options
that enable/disable their build, in practice, only jerry-libc can be
replaced with the system (compiler-default) libc. If jerry-libm is
disabled, the build of jerry-main fails, as there is no way to
instruct the linker to link the system libm to the binary. (The
build system does have a way to pass flags to the linker, but those
flags are listed before the linked objects. For the references to
get resolved correctly, the libraries to be linked have to be
specified _after_ the objects.)
This patch adds the EXTERNAL_LINK_LIBS configuration option to
CMakeLists, which ensures that the specified libraries get
correctly passed to the linker. (E.g, replacing jerry-libm with
system libm becomes possible with
`JERRY_LIBM=OFF EXTERNAL_LINK_LIBS='-lm'`.)
Additionally, the patch also makes the following related changes:
* Removes the COMPILER_DEFAULT_LIBC configuration option, as it is
(almost) always the opposite of JERRY_LIBC. Moreover, its name is
misleading: its only role is to add `-nostdlib` to the linker
flags.
* Makes use of transitive library dependencies: if a library has
another library as dependency, and it is linked to a binary, its
dependency is linked as well. Thus, jerry-libc, jerry-libm, and
any external libraries are added to jerry-core as dependency, and
then only jerry-core is linked to executables (cmake will take
care of the rest).
* build.py and run-tests.py follow up the changes, along with some
minor syntax changes.
* Moves static linking option to global CMakeLists, as unit test
binaries should be linked the same way as jerry-main.
* Adds EXTERNAL_COMPILER_FLAGS and EXTERNAL_LINKER_FLAGS as last to
the flag list, to allow user override of (nearly) anything.
The patch speculatively follows up the build system changes in the
mbed, riot-stm32f4, and zephyr targets.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
OS X build regularly reports some 39 warnings falling in 3 major
categories:
* "static function '...' is used in an inline function with
external linkage [-Wstatic-in-inline]": Some semantics around
`inline` have changed between C89 and C99, and gcc and clang seem
to disagree on how strict they should be about them. Solution
chosen is to use `-Wnostatic-in-inline` command line option for
clang.
* "implicit conversion turns floating-point number into integer:
'double' to 'bool' [-Wfloat-conversion]": `if (fmod (..., ...))`
was used at different places, which is not nice anyway, thus the
return value is compared explicitly against `ECMA_NUMBER_ZERO`.
* "format string is not a string literal [-Wformat-nonliteral]":
Console and log port I/O functions have a printf-like interface,
and the default implementations actually pass both format string
and the remaining arguments to a vfprintf. However, clang is
strict about the format string parameter of vfprintf and expects
a literal there. By annotating the port I/O functions with
`__attribute__ ((format (printf, ..., ...)))`, clang will check
the format string being a literal string earlier, when the port
functions are called, and will not complain within them when
vfprintf is called.
(Actually, this has revealed an incorrect format string, which
has been fixed as well.)
(There were also some single conversion errors not listed above.)
The patch was tested on OS X (where all warnings disappeared), but
it should help clang compilation on other OS's as well.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
* Remove JERRY_CORE CMake option: the building of the core
JerryScript library should not be optional.
* Fix wording of comments, status and error messages.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
Apple does not support staticaly built applications on OSX, and all
dynamically built apps have to be linked against the System lib
(i.e., `-static` should not be used and `-lSystem` is a must).
As System contains all libc and libm functions, building (and
linking) the minimal jerry-libc and jerry-libm libs makes no sense.
Moreover, if JERRY_LIBC is ON, the compiler will use the jerry-libc
headers but will link the libc functions from System, which causes
heavy confusion and segfaults at run time.
Thus, this patch changes the build system to disable the building
of jerry-libc and jerry-libm, and enables the use of system
libraries when building on OSX.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
CMakeLists already contains macros to ease adding compilation and
warning flags. This patch:
* Ensures that they are used whereever possible.
* Adds more macros to help with other flags as well.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
We removed that implementation where the build directory isn't set up to build with exactly one
configuration of the project but potentially several variants: the same build directory
can/must be used for debug and release builds, for full or compact profile versions, etc.
So we reworked the CMakeLists, and now one build dir deal with exactly one configuration
of the project's libraries and tools.
JerryScript-DCO-1.0-Signed-off-by: Zsolt Borbély zsborbely.u-szeged@partner.samsung.com
JerryScript-DCO-1.0-Signed-off-by: Robert Sipka rsipka.uszeged@partner.samsung.com
Building with -std=c99 is not sufficient to enforce a pure C99 build as it still allows GNU extensions which don't conflict with C99 to be used. Add -pedantic to the build options to ensure that the codebase is only using C99 constructs and the build will fail whenever any compiler extension is used. This helps to ensure that JerryScript remains portable and can be built with any C99-compatible C compiler.
Removing -Wpedantic, as any warnings of that type will already trigger an error when building with -pedantic.
JerryScript-DCO-1.0-Signed-off-by: Tilmann Scheller t.scheller@samsung.com
The GCC 4.7 check is a leftover from the times when JerryScript was a C++11 codebase, it was there to ensure that JerryScript is compiled with a GCC version which has sufficient support for C++11.
Today JerryScript is a pure C99 codebase (enforced by building all the sources with -std=c99) and this check is no longer necessary. JerryScript should now build just fine even with rather old versions of GCC.
JerryScript-DCO-1.0-Signed-off-by: Tilmann Scheller t.scheller@samsung.com
* Moved the error codes to jerry-port.h and declared port function
`jerry_port_fatal`.
* Moved "exit or abort on fail" functionality to the newly added
jerry-port-default-fatal.c.
* This implied that a default port-specific API had to be introduced:
functions `jerry_port_default_set_abort_on_fail` and
`jerry_port_default_is_abort_on_fail` declared in jerry-port-default.h
control the fatal exit behaviour.
* For the sake of clarity, renamed jerry-port.c to
jerry-port-default-io.c.
* Adapted CMakeLists to deal with port implementations consisting of
more then one source file and exposing headers. This also required
the renaming of `EXTERNAL_PORT_FILE` cmake option to
`EXTERNAL_PORT_DIR`.
* Adapted main sources to use the default port header for the
abort-on-fail functionality, as that is not part of the core jerry
API anymore.
* Added default port implementation to the static source code checker
tools.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
* Rename modified fdlibm to jerry-libm
* Move third-party/fdlibm to jerry-libm
* Rename original fdlibm.h to jerry-libm-internal.h
* And remove it from the public headers.
* Rename jerry-libm's public header to math.h
* This also makes jerry-core sources include `<math.h>`. Therefore,
should anyone want to use a different libm implementation with
jerry, it becomes possible. (The same way as we provide a minimal
libc with standard headers, but should it be insufficient or
conflicting for someone, it can be replaced.)
* Drop `s_` prefix from jerry-libm sources
* The original fdlibm implementation had various prefixes (e.g., `k_`
for sources of kernel routines, and `w_` for wrapper routines), but
after the specialization of fdlibm to jerry, only `s_` remained.
Since it does not encode anything anymore, it can be dropped.
* Stylistic edits to jerry-libm's CMakeLists
* Align project name with other CMakeLists in the code base
* Move Jerry-LibM under Apache License
* Using the same approach as was used by linux-wireless when ath5k
driver license needed clarification. Solution was proposed by SFLC.
External mail for future reference: http://lwn.net/Articles/247806/
* Tests & checks
* Remove FD from the name of libm unit test-related files
* Make vera++ and cppcheck check jerry-libm
* Targets
* Speculative update of targets to use jerry-libm
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
Add new build target: test-buildoptions
Now every build option is tested on the proper targets. These are the native
and the MCU targets in release mode and unittests.
Therefore the USE_COMPILER_DEFAULT_LIBC build option is refactored.
JerryScript-DCO-1.0-Signed-off-by: Zsolt Borbély zsborbely.u-szeged@partner.samsung.com
MEM_HEAP_PTR_64 is duplicating existing information: stdint.h, which
header is already used by the project, defines various _MAX macros
for upper limits of integer types. The comparison of UINTPTR_MAX and
UINT32_MAX can give the same info as encoded in MEM_HEAP_PTR_64.
The stdint.h-based approach has the benefit that jerry can support
any 64-bit architecture without the need for editing the build
system. (With the existing approach, CMakeLists has to know about
every 64-bit architecture to work properly.)
Thus, removing the extraneous macro from the code.
The patch also changes the mem_pools_chunk_t struct, as it turned
out that the struct does not have to be padded to MEM_POOL_CHUNK_SIZE.
(The padding also depended on MEM_HEAP_PTR_64.) It is enough if the
size of the struct is smaller than (or equal to) MEM_POOL_CHUNK_SIZE.
JerryScript-DCO-1.0-Signed-off-by: Akos Kiss akiss@inf.u-szeged.hu
