tracing code.
Managed static was just insane overhead for this. We took memory fences
and external function calls in every path that pushed a pretty stack
frame. This includes a multitude of layers setting up and tearing down
passes, the parser in Clang, everywhere. For the regression test suite
or low-overhead JITs, this was contributing to really significant
overhead.
Even the LLVM ThreadLocal is really overkill here because it uses
pthread_{set,get}_specific logic, and has careful code to both allocate
and delete the thread local data. We don't actually want any of that,
and this code in particular has problems coping with deallocation. What
we want is a single TLS pointer that is valid to use during global
construction and during global destruction, any time we want. That is
exactly what every host compiler and OS we use has implemented for
a long time, and what was standardized in C++11. Even though not all of
our host compilers support the thread_local keyword, we can directly use
the platform-specific keywords to get the minimal functionality needed.
Provided this limited trial survives the build bots, I will move this to
Compiler.h so it is more widely available as a light weight if limited
alternative to the ThreadLocal class. Many thanks to David Majnemer for
helping me think through the implications across platforms and craft the
MSVC-compatible syntax.
The end result is *substantially* faster. When running llc in a tight
loop over a small IR file targeting the aarch64 backend, this improves
its performance by over 10% for me. It also seems likely to fix the
remaining regressions seen by JIT users with threading enabled.
This may actually have more impact on real-world compile times due to
the use of the pretty stack tracing utility throughout the rest of Clang
or LLVM, but I've not collected any detailed measurements.
llvm-svn: 227300
Design Of lib/System
====================
The software in this directory is designed to completely shield LLVM from any
and all operating system specific functionality. It is not intended to be a
complete operating system wrapper (such as ACE), but only to provide the
functionality necessary to support LLVM.
The software located here, of necessity, has very specific and stringent design
rules. Violation of these rules means that cracks in the shield could form and
the primary goal of the library is defeated. By consistently using this library,
LLVM becomes more easily ported to new platforms since the only thing requiring
porting is this library.
Complete documentation for the library can be found in the file:
llvm/docs/SystemLibrary.html
or at this URL:
http://llvm.org/docs/SystemLibrary.html
While we recommend that you read the more detailed documentation, for the
impatient, here's a high level summary of the library's requirements.
1. No system header files are to be exposed through the interface.
2. Std C++ and Std C header files are okay to be exposed through the interface.
3. No exposed system-specific functions.
4. No exposed system-specific data.
5. Data in lib/System classes must use only simple C++ intrinsic types.
6. Errors are handled by returning "true" and setting an optional std::string
7. Library must not throw any exceptions, period.
8. Interface functions must not have throw() specifications.
9. No duplicate function impementations are permitted within an operating
system class.
To accomplish these requirements, the library has numerous design criteria that
must be satisfied. Here's a high level summary of the library's design criteria:
1. No unused functionality (only what LLVM needs)
2. High-Level Interfaces
3. Use Opaque Classes
4. Common Implementations
5. Multiple Implementations
6. Minimize Memory Allocation
7. No Virtual Methods
16b670ec20