| Age | Commit message (Collapse) | Author |
|---|
|
would not recompile if the LLVM C compiler flags changed. This could
happen if the user did something like
```
make -f Makefile.cmake.bitcode LLVMCC.ExtraFlags=-Wall
```
|
|
With the old buildsystem we could pass CFLAGS when building runtime
libs. Support passing some additional flags to cmake-based system too.
We need this to build 32 and 64bit runtime libs separately (but not
whole klee).
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
|
|
Remove support for reporting the approximate git tag.
|
|
|
|
Moving to KLEE 1.3.0
|
|
This was confusing because it would emit something like
`v1.0.0-290-g08d4716` because the 1.1.0 and 1.2.0 releases
didn't have a tag on the master branch so `git describe --tags`
would just find the `v1.0.0` tag and report based on that tag.
|
|
|
|
|
|
Fixing current version of STP in Dockerfile (see #505) to 2.1.2
|
|
|
|
|
|
Fix Expr::compare
|
|
* Making `Expr::compre(const Expr&, ExprEquivSet)` private and moving
its implementation into `Expr.cpp`.
* Document `Expr::compare(const Expr&)`.
This partially addresses #515 .
|
|
and make it a pure virtual method. Also make it protected rather than
public because it is an implementation detail of `Expr::compare()`.
This means that sub classes must implement it so this commit also
provides implementations. The comparision behaves as before except
for `ConcatExpr` where its `width` attribute is now used (previously
it wasn't).
This commit also documents the semantics of `Expr::compareContents(const Expr&)`
which was clearly needed because it has been incorrectly implemented
in the past and has gone unnoticed for several years.
This partially addresses #515 .
|
|
`compareContents()`.
This bug would not have affected correctness but it would have
affected performance because if `Expr::compare()` was called on
two structually equal `NotExpr` then its children would be
checked for structually equality twice - once in
NotExpr::compareContents()` and once in `Expr::compare()`.
This partially addresses #515 .
|
|
|
|
|
|
to see if we can find klee-uclibc's C library.
|
|
`llvm-config` when using LLVM 3.5 and newer.
In newer versions of `llvm-config`, `--ldflags` doesn't give
the system libraries anymore. Instead we need to use `--system-libs`.
Issue reported by @ryosa .
|
|
Reported by @jirislaby in #507.
|
|
Now LLVM libraries are added as imported targets and their link dependencies
are explicitly stated so CMake can get the link order correct.
|
|
This has shown that there is another circular dependency
(added by me! sigh...) between `kleeCore` and `kleeModule`.
|
|
This fixes issue #507.
We can't build the components as shared libraries right now due
to cyclic dependencies (see #502 for a fix) and there are a few
other patches needed too (see #507).
Building the components as shared libraries isn't really desirable
anyway because it would require us to ship KLEE's libraries which
don't have a stable API.
|
|
|
|
interfering with CMake build.
We now raise a fatal error if we detect certain files in the source
tree. This is a heuristic but it's good enough for now.
The order of includes has also been changed so that the build tree
include directory has priority over the source tree include directory so
that even if the added check was missing the right header file (i.e.
`${CMAKE_BINARY_DIR}/include/klee/Config/config.h` would be picked up at
build time.
|
|
These were changes that I forgot to make in
dda296e09ee53ed85ccf1c3f08e7e809adce612e .
|
|
implement it in the solver.
|
|
|
|
|
|
MartinNowack-fix_bfs2
|
|
add nicer error messages for --use-merge and add explanation why it currently cannot be used with random-path
|
|
|
|
|
|
transitive dependencies on KLEE's libraries rather than on the final
binaries. This is better because it means we can build
other tools that use KLEE's libraries and not need to express the
needed LLVM dependencies.
It also makes it clearer what the dependencies are between KLEE
libraries. This has illustrated a problem with the `kleeBasic`
library. It contains `ConstructSolverChain.cpp` which clearly
belongs in `kleaverSolver` not in `kleeBasic`. This will be fixed
later.
|
|
Links in the correct LLVM libraries when using the MCJIT. No effect for LLVM
versions less than 3.6
|
|
For performance reasons, if KLEE branches, one state is reused
and it is progressed by adding new constraints.
Make sure both new states end up at the end of the BFS searcher queue.
|
|
currently cannot be used with random-path
|
|
This is done as a separate commit because it imports
third party code. It's under the Boost license though
so it "should be" fine.
|
|
don't trigger unnecessary rebuilds. Also make the Dockerfile ignore
Vim source files anywhere in the tree.
|
|
This is based off intial work by @jirislaby in #481. However it
has been substantially modified.
Notably it includes a separate build sytem to build the runtimes which
is inspired by the old build system. The reason for doing this is
because CMake is not well suited for building the runtime:
* CMake is configured to use the host compiler, not the bitcode
compiler. These are not the same thing.
* Building the runtime using `add_custom_command()` is flawed
because we can't automatically get transitive depencies (i.e.
header file dependencies) unless the CMake generator is makefiles.
(See `IMPLICIT_DEPENDS` of `add_custom_command()` in CMake).
So for now we have a very simple build system for building the runtimes.
In the future we can replace this with something more sophisticated if
we need it.
Support for all features of the old build system are implemented apart
from recording the git revision and showing it in the output of
`klee --help`.
Another notable change is the CMake build system works much better with
LLVM installs which don't ship with testing tools. The build system
will download the sources for `FileCheck` and `not` tools if the
corresponding binaries aren't available and will build them. However
`lit` (availabe via `pip install lit`) and GTest must already be
installed.
Apart from better support for testing a significant advantage of the
new CMake build system compared to the existing "Autoconf/Makefile"
build system is that it is **not** coupled to LLVM's build system
(unlike the existing build system). This means that LLVM's
autoconf/Makefiles don't need to be installed somewhere on the system.
Currently all tests pass.
Support has been implemented in TravisCI and the Dockerfile for
building with CMake.
The existing "Autoconf/Makefile" build system has been left intact
and so both build systems can coexist for a short while. We should
remove the old build system as soon as possible though because it
creates an unnecessary maintance burden.
|
|
Support for Darwin platform in RaiseAsm pass
|
|
|
|
Upgraded to boolector-2.2.0 and some cleanup and refactoring of the metaSMT support.
|
|
|
|
|
|
|
|
runAndGetCexForked())
|
|
|
|
|
|
Fix `-Wmisleading-indentation` warning and also correctly set the `dirty` flag
|
