diff options
| -rw-r--r-- | docs/ReleaseNotes-2.6.html | 258 |
1 files changed, 138 insertions, 120 deletions
diff --git a/docs/ReleaseNotes-2.6.html b/docs/ReleaseNotes-2.6.html index 1ac46e8..5549914 100644 --- a/docs/ReleaseNotes-2.6.html +++ b/docs/ReleaseNotes-2.6.html @@ -52,26 +52,8 @@ current one. To see the release notes for a specific release, please see the </div> - Many new papers added to /pubs/ - Machine LICM, hoists things like constant pool loads, loads from readonly stubs, vector constant synthesization code, etc. - Machine Sinking - Regalloc improvements for commuting, various spiller peephole optimizations, cross-class coalescing. - Support for debug line numbers when optimization enabled - gold lto plugin - - target-specific intrinsics (r63765) - <tt>llc -enable-value-prop</tt>, propagation of value info (sign/zero ext info) from one MBB to another - interpreter + libffi - LLVMContext - Preliminary support for addrspace 256 -> GS, 257 -> FS, known problems: CodeGenerator.html#x86_memory - -Add support for the PowerPC 64-bit SVR4 ABI. - - NSW/NUW/exact div - Inbounds for GEP - SRoA improvements for vector unions, memset, arbitrary weird bitfield accesses etc. It now produces "strange" sized integers. - pre-alloc splitter?? - X86: Support for softfloat modes, typically used by OS kernels. + + MC: MCSection, MCAsmInfo MCInstPrinter did it make it in? @@ -83,94 +65,23 @@ Add support for the PowerPC 64-bit SVR4 ABI. ELF Writer? How stable? - LSR promotes int induction variables to 64-bit on 64-bit targets, major perf boost for numerical code. - LSR now analyzes pointer expressions (e.g. getelementptrs), not just integers. - Stack slot coloring for register spills (denser stack frames) - SelectionDAGS: New BuildVectorSDNode (r65296), and ISD::VECTOR_SHUFFLE (r69952 / PR2957) - New PrettyStackTrace, crashes of llvm tools should give some indication of what the compiler was doing at the time of the crash (e.g. running a pass), and print out command line arguments. - new linkage types linkonce_odr, weak_odr, linker_private, and available_externally. - Inliner reuse stack space when inlining arrays? - - Regalloc hints for allocation stuff: Evan r73381/r73671. Finished/enabled? - API Cleanup: - no use of hash_set/hash_map, no more llvm::OStream - Use raw_ostream for everything, killed off llvm/Streams.h and DOUT - - Mips now supports O32 Calling Convention - - StringRef class, Twine class. - New BlackFin backend. - +APIs: + Shrink wrapping support in PEI, what is the state of it? - X86-64: better modeling of implicit zero extensions, eliminates a lot of redundant zexts - X86-64 TLS support for local exec and initial exec. - X86 - Better modeling of H registerts as subregs. - Getelementpr instruction now allows any integer type for array/pointer indexes. - - include/llvm/Analysis/LiveValues.h => dead?? - lib/Analysis/LoopVR.cpp ==> dead?? - include/llvm/CodeGen/LazyLiveness.h ==> dead? - lib/Transforms/IPO/MergeFunctions.cpp ==> dead? - llvm/Analysis/PointerTracking.h ==> dead?? - PRedSimplify -> Ask vmkit if it is still useful and for testcases. - - ARM calling convention code is now tblgen generated instead of manual. - ARM: NEON support. neonfp for doing single precision fp with neon instead of VFP. - Tblgen now supports multiclass inheritance. - Unladen swallow as user? - Loop index split disabled by default? - - New WeakVH and AssertingVH and CallbackVH classes. - New llvm/ADT/Triple class. - Removed the IA64 backend. - Profile info improvements by Andreas Neustifter. - PostRA scheduler improvements David Goodwin. +<!-- +Almost dead code. + include/llvm/Analysis/LiveValues.h => dan + lib/Analysis/LoopVR.cpp ==> dan/nick/owen + lib/Transforms/IPO/MergeFunctions.cpp + llvm/Analysis/PointerTracking.h +--> - New MSP430 and SystemZ backends. - llvm-gcc now supports a new TCE target. - klee web page at klee.llvm.org - New llvm/System/Atomic.h, llvm/System/RWMutex.h for portable atomic ops, rw locks. - llvm_start_multithreaded: ProgrammersMAnual.html#threading - - Tablegen now supports a number of new string and list operations like - !(subst), !(foreach), !car, !cdr, !null, !if, !cast. - New fadd, fsub, fmul instructions and classes. - New MachineVerifier pass. - Enabled GVN Load PRE. - ARM AAPCS-VFP hard float ABI is supported. - LLVM build now builds all libraries as .a files instead of some libraries as relinked .o files. This requires some APIs like InitializeAllTargets.h. TargetRegistry! - - ARM Thumb2 support: status? - CBE status: not part of the release criteria. - - New SourceMgr, SMLoc classes for simple parsers with caret diagnostics and #include support, (used by - tablegen, llvm-mc, the .ll parser, FileCheck, etc) - FileCheck! + CHECK-NEXT - New compiler-rt project. - New Static Single Information (SSI) construction pass (not used by anything yet, experimental). - llvm_report_error() error handling API (llvm/Support/ErrorHandling.h) - - x86: Vector icmp/fcmp now work with SSE codegen. - X86: all global variable reference logic is now in ClassifyGlobalReference. - JIT support for oprofile (r75279), configure with --with-oprofile. Now we get line # and function info for JIT'd functions. - Mention gcc plugin. - - New EngineBuilder class for creating JITs: r76276 Reid Kleckner <reid at kleckner.net> - - -asm-verbose now prints location info (with -g) and loop nest info. - JIT now supports generating more than 16M of code. - -removed the BigBlock register allocator, it had bitrotted. -Target intrinsics can now return multiple results. - - SSE 4.2 support. - Ada bindings for LLVM IR. - Many extensions to the C APIs. + <!-- Unfinished features in 2.6: Mention gcc plugin. @@ -185,6 +96,10 @@ Target intrinsics can now return multiple results. Logo web page. llvm devmtg compiler_rt + klee web page at klee.llvm.org + Many new papers added to /pubs/ + Mention gcc plugin. + --> <!-- *********************************************************************** --> @@ -406,6 +321,7 @@ machine code using LLVM's JIT.</p> minor improvements. Some of the major improvements and new features are listed in this section. </p> + </div> <!--=========================================================================--> @@ -418,7 +334,15 @@ in this section. <p>LLVM 2.6 includes several major new capabilities:</p> <ul> -<li>Something wonderful!</li> +<li>Support for debug line numbers when optimization enabled</li> +<li>gold lto plugin</li> +<li>New MSP430 and SystemZ backends.</li> +<li>New BlackFin backend.</li> +<li>LLVMContext, llvm_start_multithreaded: ProgrammersManual.html#threading</li> +<li>Unladen swallow as user?</li> +<li>klee web page at klee.llvm.org</li> +<li>FileCheck</li> +<li>New compiler-rt project.</li> <li>LLVM 2.6 includes a brand new experimental LLVM bindings to the Ada2005 programming language.</li> </ul> @@ -453,7 +377,14 @@ includes support for the C, C++, Objective-C, Ada, and Fortran front-ends.</p> can be useful if you are writing a front-end for LLVM:</p> <ul> -<li>Something wonderful!</li> +<li>Getelementpr instruction now allows any integer type for array/pointer indexes.</li> +<li>Inbounds for GEP</li> +<li>NSW/NUW/exact div</li> +<li>LSR promotes int induction variables to 64-bit on 64-bit targets, major perf boost for numerical code.</li> +<li>LSR now analyzes pointer expressions (e.g. getelementptrs), not just integers.</li> +<li>new linkage types linkonce_odr, weak_odr, linker_private, and available_externally.</li> +<li>New fadd, fsub, fmul instructions and classes. </li> +<li>Target intrinsics can now return multiple results.</li> </ul> </div> @@ -470,7 +401,11 @@ release includes a few major enhancements and additions to the optimizers:</p> <ul> -<li>Something wonderful!</li> +<li>SRoA improvements for vector unions, memset, arbitrary weird bitfield accesses etc. It now produces "strange" sized integers.</li> +<li>Inliner reuse stack space when inlining arrays?</li> +<li>Enabled GVN Load PRE.</li> +<li>New Static Single Information (SSI) construction pass (not used by anything yet, experimental).</li> +</li> </ul> @@ -489,7 +424,22 @@ it run faster:</p> <ul> -<li>Something wonderful!</li> +<li> -asm-verbose now prints location info (with -g) and loop nest info.</li> +<li>Tblgen now supports multiclass inheritance and a number of new string and + list operations like !(subst), !(foreach), !car, !cdr, !null, !if, !cast. + These make the .td files more expressive and allow more aggressive factoring + of duplication across instruction patterns.</li> +<li>New MachineVerifier pass.</li> +<li>Machine LICM, hoists things like constant pool loads, loads from readonly stubs, vector constant synthesization code, etc.</li> +<li>Machine Sinking</li> +<li>target-specific intrinsics (r63765)</li> +<li>Regalloc improvements for commuting, various spiller peephole optimizations, cross-class coalescing.</li> +<li><tt>llc -enable-value-prop</tt>, propagation of value info (sign/zero ext info) from one MBB to another</li> +<li>Regalloc hints for allocation stuff: Evan r73381/r73671. Finished/enabled?</li> +<li>Stack slot coloring for register spills (denser stack frames)</li> +<li>SelectionDAGS: New BuildVectorSDNode (r65296), and ISD::VECTOR_SHUFFLE (r69952 / PR2957)</li> +<li>PostRA scheduler improvements David Goodwin.</li> +</li> </ul> </div> @@ -504,7 +454,16 @@ it run faster:</p> <ul> -<li>Something wonderful!</li> +<li>Preliminary support for addrspace 256 -> GS, 257 -> FS, known problems: CodeGenerator.html#x86_memory</li> +<li>Support for softfloat modes, typically used by OS kernels.</li> + +<li>X86-64: better modeling of implicit zero extensions, eliminates a lot of redundant zexts</li> +<li>X86-64 TLS support for local exec and initial exec.</li> +<li>Better modeling of H registerts as subregs.</li> +<li>Vector icmp/fcmp now work with SSE codegen.</li> +<li>SSE 4.2 support.</li> +<li>all global variable reference logic is now in ClassifyGlobalReference.</li> +</li> </ul> </div> @@ -519,18 +478,18 @@ it run faster:</p> </p> <ul> -<li>Something wonderful!</li> +<li>Support for floating-point, indirect function calls, and + passing/returning aggregate types to functions. +<li>The code generator is able to generate debug info into output COFF files. +<li>Support for placing an object into a specific section or at a specific + address in memory.</li> </ul> <p>Things not yet supported:</p> <ul> -<li>Floating point.</li> -<li>Passing/returning aggregate types to and from functions.</li> <li>Variable arguments.</li> -<li>Indirect function calls.</li> <li>Interrupts/programs.</li> -<li>Debug info.</li> </ul> </div> @@ -554,21 +513,74 @@ AAPCS-VFP "hard float" calling conventions are also supported with the and subject to change. The Neon intrinsics, in particular, may change in future releases of LLVM. </li> + + ARM AAPCS-VFP hard float ABI is supported. + ARM calling convention code is now tblgen generated instead of manual. + ARM: NEON support. neonfp for doing single precision fp with neon instead of VFP. + </ul> </div> +<!--=========================================================================--> +<div class="doc_subsection"> +<a name="OtherTarget">Other Target Specific Improvements</a> +</div> + +<div class="doc_text"> +<p>New features of other targets include: +</p> + +<ul> +<li>Add support for the PowerPC 64-bit SVR4 ABI.</li> +<li>Mips now supports O32 Calling Convention.</li> +</ul> + +</div> <!--=========================================================================--> <div class="doc_subsection"> -<a name="llvmc">Improvements in LLVMC</a> +<a name="newapis">New Useful APIs</a> </div> <div class="doc_text"> -<p>New features include:</p> <ul> -<li>Something wonderful!</li> +<li>New EngineBuilder class for creating JITs: r76276</li> + New PrettyStackTrace, crashes of llvm tools should give some indication of what the compiler was doing at the time of the crash (e.g. running a pass), and print out command line arguments. + StringRef class, Twine class. + New WeakVH and AssertingVH and CallbackVH classes. + New llvm/ADT/Triple class. + llvm_report_error() error handling API (llvm/Support/ErrorHandling.h) + New llvm/System/Atomic.h, llvm/System/RWMutex.h for portable atomic ops, rw locks. + New SourceMgr, SMLoc classes for simple parsers with caret diagnostics and #include support, (used by + tablegen, llvm-mc, the .ll parser, FileCheck, etc) + +<ul> + + +</div> + +<!--=========================================================================--> +<div class="doc_subsection"> +<a name="otherimprovements">Other Improvements and New Features</a> +</div> + +<div class="doc_text"> +<p>Other miscellaneous features include:</p> + +<ul> +<li>interpreter + libffi</li> +<li> JIT now supports generating more than 16M of code.</li> +<li>Users can now <a +href="http://llvm.org/doxygen/classllvm_1_1ExecutionEngine.html#fcd2b4b92ca38812ca31640b0da14927">register</a> +a <a href="http://llvm.org/doxygen/classllvm_1_1JITEventListener.html">JITEventListener</a> +to receive callbacks when the JIT emits or frees machine code. The +OProfile support uses this mechanism.</li> + JIT support for oprofile (r75279), configure with --with-oprofile. Now we get line # and function info for JIT'd functions. + +<li> Profile info improvements by Andreas Neustifter.</li> +<li> Many extensions to the C APIs.</li> </ul> </div> @@ -587,14 +599,23 @@ from the previous release.</p> <ul> -<li>Something horrible!</li> - +<li>The Itanium (IA64) backend has been removed. It was not supported and + bitrotted.</li> +<li>The BigBlock register allocator has been removed, it also bitrotted.</li> +<li>The C Backend (-march=c) is no longer considered part of the LLVM release +criteria. We still want it to work, but no one is maintaining it and it lacks +support for arbitrary precision integers and other important IR features.</li> </ul> <p>In addition, many APIs have changed in this release. Some of the major LLVM API changes are:</p> + + API Cleanup: + no use of hash_set/hash_map, no more llvm::OStream + Use raw_ostream for everything, killed off llvm/Streams.h and DOUT + <ul> <li>LLVM's global uniquing tables for <tt>Type</tt>s and <tt>Constant</tt>s have been privatized into members of an <tt>LLVMContext</tt>. A number of APIs @@ -724,7 +745,7 @@ using the stock GCC 3.x.x series 'out the box', See: <a href="#brokengcc">Broken versions of GCC and other tools</a>. However, A <a href="http://pkg.auroraux.org/GCC">Modern GCC Build</a> for x86/x64 has been made available from the third party AuroraUX Project -that has been meticulously tested for bootstrapping LLVM & Clang.</li> +that has been meticulously tested for bootstrapping LLVM & Clang.</li> </ul> </div> @@ -809,9 +830,6 @@ processors, thumb programs can crash or produce wrong results (<a href="http://llvm.org/PR1388">PR1388</a>).</li> <li>Compilation for ARM Linux OABI (old ABI) is supported but not fully tested. </li> -<li>There is a bug in QEMU-ARM (<= 0.9.0) which causes it to incorrectly - execute -programs compiled with LLVM. Please use more recent versions of QEMU.</li> </ul> </div> |