Permalink
Commits on Apr 15, 2012
  1. fix pr12559: mark unavailable win32 math libcalls

    jckarter committed Apr 14, 2012
    also fix SimplifyLibCalls to use TLI rather than compile-time conditionals to enable optimizations on floor, ceil, round, rint, and nearbyint
Commits on Mar 31, 2012
  1. Add a workaround for building with old versions of clang.

    Rafael Espindola
    Rafael Espindola committed Mar 31, 2012
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153820 91177308-0d34-0410-b5e6-96231b3b80d8
  2. Add a triple to the test.

    Rafael Espindola
    Rafael Espindola committed Mar 31, 2012
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153818 91177308-0d34-0410-b5e6-96231b3b80d8
  3. Teach CodeGen's version of computeMaskedBits to understand the range …

    Rafael Espindola
    Rafael Espindola committed Mar 31, 2012
    …metadata.
    
    This is the CodeGen equivalent of r153747. I tested that there is not noticeable
    performance difference with any combination of -O0/-O2 /-g when compiling
    gcc as a single compilation unit.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153817 91177308-0d34-0410-b5e6-96231b3b80d8
  4. Fix dynamic linking on PPC64.

    hfinkel committed Mar 31, 2012
    Dynamic linking on PPC64 has had problems since we had to move the top-down
    hazard-detection logic post-ra. For dynamic linking to work there needs to be
    a nop placed after every call. It turns out that it is really hard to guarantee
    that nothing will be placed in between the call (bl) and the nop during post-ra
    scheduling. Previous attempts at fixing this by placing logic inside the
    hazard detector only partially worked.
    
    This is now fixed in a different way: call+nop codegen-only instructions. As far
    as CodeGen is concerned the pair is now a single instruction and cannot be split.
    This solution works much better than previous attempts.
    
    The scoreboard hazard detector is also renamed to be more generic, there is currently
    no cpu-specific logic in it.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153816 91177308-0d34-0410-b5e6-96231b3b80d8
  5. Give the always-inliner its own custom filter. It shouldn't have to pay

    chandlerc committed Mar 31, 2012
    the very high overhead of the complex inline cost analysis when all it
    wants to do is detect three patterns which must not be inlined. Comment
    the code, clean it up, and leave some hints about possible performance
    improvements if this ever shows up on a profile.
    
    Moving this off of the (now more expensive) inline cost analysis is
    particularly important because we have to run this inliner even at -O0.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153814 91177308-0d34-0410-b5e6-96231b3b80d8
  6. Remove a bunch of empty, dead, and no-op methods from all of these

    chandlerc committed Mar 31, 2012
    interfaces. These methods were used in the old inline cost system where
    there was a persistent cache that had to be updated, invalidated, and
    cleared. We're now doing more direct computations that don't require
    this intricate dance. Even if we resume some level of caching, it would
    almost certainly have a simpler and more narrow interface than this.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153813 91177308-0d34-0410-b5e6-96231b3b80d8
  7. Initial commit for the rewrite of the inline cost analysis to operate

    chandlerc committed Mar 31, 2012
    on a per-callsite walk of the called function's instructions, in
    breadth-first order over the potentially reachable set of basic blocks.
    
    This is a major shift in how inline cost analysis works to improve the
    accuracy and rationality of inlining decisions. A brief outline of the
    algorithm this moves to:
    
    - Build a simplification mapping based on the callsite arguments to the
      function arguments.
    - Push the entry block onto a worklist of potentially-live basic blocks.
    - Pop the first block off of the *front* of the worklist (for
      breadth-first ordering) and walk its instructions using a custom
      InstVisitor.
    - For each instruction's operands, re-map them based on the
      simplification mappings available for the given callsite.
    - Compute any simplification possible of the instruction after
      re-mapping, and store that back int othe simplification mapping.
    - Compute any bonuses, costs, or other impacts of the instruction on the
      cost metric.
    - When the terminator is reached, replace any conditional value in the
      terminator with any simplifications from the mapping we have, and add
      any successors which are not proven to be dead from these
      simplifications to the worklist.
    - Pop the next block off of the front of the worklist, and repeat.
    - As soon as the cost of inlining exceeds the threshold for the
      callsite, stop analyzing the function in order to bound cost.
    
    The primary goal of this algorithm is to perfectly handle dead code
    paths. We do not want any code in trivially dead code paths to impact
    inlining decisions. The previous metric was *extremely* flawed here, and
    would always subtract the average cost of two successors of
    a conditional branch when it was proven to become an unconditional
    branch at the callsite. There was no handling of wildly different costs
    between the two successors, which would cause inlining when the path
    actually taken was too large, and no inlining when the path actually
    taken was trivially simple. There was also no handling of the code
    *path*, only the immediate successors. These problems vanish completely
    now. See the added regression tests for the shiny new features -- we
    skip recursive function calls, SROA-killing instructions, and high cost
    complex CFG structures when dead at the callsite being analyzed.
    
    Switching to this algorithm required refactoring the inline cost
    interface to accept the actual threshold rather than simply returning
    a single cost. The resulting interface is pretty bad, and I'm planning
    to do lots of interface cleanup after this patch.
    
    Several other refactorings fell out of this, but I've tried to minimize
    them for this patch. =/ There is still more cleanup that can be done
    here. Please point out anything that you see in review.
    
    I've worked really hard to try to mirror at least the spirit of all of
    the previous heuristics in the new model. It's not clear that they are
    all correct any more, but I wanted to minimize the change in this single
    patch, it's already a bit ridiculous. One heuristic that is *not* yet
    mirrored is to allow inlining of functions with a dynamic alloca *if*
    the caller has a dynamic alloca. I will add this back, but I think the
    most reasonable way requires changes to the inliner itself rather than
    just the cost metric, and so I've deferred this for a subsequent patch.
    The test case is XFAIL-ed until then.
    
    As mentioned in the review mail, this seems to make Clang run about 1%
    to 2% faster in -O0, but makes its binary size grow by just under 4%.
    I've looked into the 4% growth, and it can be fixed, but requires
    changes to other parts of the inliner.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153812 91177308-0d34-0410-b5e6-96231b3b80d8
  8. Add support to the InstVisitor for visiting a generic callsite. The

    chandlerc committed Mar 31, 2012
    visitor will now visit a CallInst and an InvokeInst with
    instruction-specific visitors, then visit a generic CallSite visitor,
    then delegate back to the Instruction visitor and the TerminatorInst
    visitors depending on whether a call or an invoke originally. This will
    be used in the soon-to-land inline cost rewrite.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153811 91177308-0d34-0410-b5e6-96231b3b80d8
  9. Move trivial functions into the class definition.

    isanbard committed Mar 31, 2012
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153810 91177308-0d34-0410-b5e6-96231b3b80d8
  10. Trim headers.

    isanbard committed Mar 31, 2012
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153809 91177308-0d34-0410-b5e6-96231b3b80d8
  11. Indent according to LLVM's style guide.

    isanbard committed Mar 31, 2012
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153808 91177308-0d34-0410-b5e6-96231b3b80d8
  12. Cleanup whitespace and trim some of the #includes.

    isanbard committed Mar 31, 2012
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153807 91177308-0d34-0410-b5e6-96231b3b80d8
  13. Internalize: Remove reference of @llvm.noinline, it was replaced with…

    d0k committed Mar 31, 2012
    … the noinline attribute a long time ago.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153806 91177308-0d34-0410-b5e6-96231b3b80d8
  14. These strings aren't 'const char *' but 'char *'.

    isanbard committed Mar 31, 2012
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153805 91177308-0d34-0410-b5e6-96231b3b80d8
  15. Cleanup whitespace.

    isanbard committed Mar 31, 2012
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153804 91177308-0d34-0410-b5e6-96231b3b80d8
  16. Cleanup whitespace and remove unneeded 'extern' keyword on function d…

    isanbard committed Mar 31, 2012
    …efinitions.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153802 91177308-0d34-0410-b5e6-96231b3b80d8
  17. Clean up the naming in this test. Someone pointed this out in review at

    chandlerc committed Mar 31, 2012
    one point, and I forgot to go back and clean it up. Sorry about that. =/
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153801 91177308-0d34-0410-b5e6-96231b3b80d8
  18. FileCheck-ize this test, and generally tidy it up prior to changing

    chandlerc committed Mar 31, 2012
    things around.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153799 91177308-0d34-0410-b5e6-96231b3b80d8
  19. I noticed in passing that the Metadata getIfExists method was creatin…

    CunningBaldrick committed Mar 31, 2012
    …g a new
    
    node and returning it if one didn't exist.
    
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153798 91177308-0d34-0410-b5e6-96231b3b80d8
  20. Correctly vectorize powi.

    hfinkel committed Mar 31, 2012
    The powi intrinsic requires special handling because it always takes a single
    integer power regardless of the result type. As a result, we can vectorize
    only if the powers are equal. Fixes PR12364.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153797 91177308-0d34-0410-b5e6-96231b3b80d8
  21. comment typo

    atrick committed Mar 31, 2012
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153796 91177308-0d34-0410-b5e6-96231b3b80d8
  22. Select static relocation model if it is jitting.

    Akira Hatanaka
    Akira Hatanaka committed Mar 31, 2012
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153795 91177308-0d34-0410-b5e6-96231b3b80d8
  23. Introduce Register Units: Give each leaf register a number.

    atrick committed Mar 31, 2012
    First small step toward modeling multi-register multi-pressure. In the
    future, register units can also be used to model liveness and
    aliasing.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153794 91177308-0d34-0410-b5e6-96231b3b80d8
  24. Add a 2 byte safety margin in offset computations.

    stoklund committed Mar 31, 2012
    ARMConstantIslandPass still has bugs where jump table compression can
    cause constant pool entries to go out of range.
    
    Add a safety margin of 2 bytes when placing constant islands, but use
    the real max displacement for verification.
    
    <rdar://problem/11156595>
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153789 91177308-0d34-0410-b5e6-96231b3b80d8
  25. Add more debugging output to ARMConstantIslandPass.

    stoklund committed Mar 31, 2012
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153788 91177308-0d34-0410-b5e6-96231b3b80d8
Commits on Mar 30, 2012
  1. * Set the scope attributes for the ASM symbol we added to be the valu…

    isanbard committed Mar 30, 2012
    …e passed
    
      into the function.
    * Reorder some header files.
    
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153783 91177308-0d34-0410-b5e6-96231b3b80d8
  2. Rip out emission of the regIsInRegClass function for the asm printer.

    d0k committed Mar 30, 2012
    It's slow, bloated and completely redundant with MCRegisterClass::contains.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153782 91177308-0d34-0410-b5e6-96231b3b80d8
  3. ARM fix encoding fixup resolution for ldrd and friends.

    Jim Grosbach
    Jim Grosbach committed Mar 30, 2012
    The 8-bit payload is not contiguous in the opcode. Move the upper nibble
    over 4 bits into the correct place.
    
    rdar://11158641
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153780 91177308-0d34-0410-b5e6-96231b3b80d8
  4. Use SequenceToOffsetTable in emitRegisterNameString.

    stoklund committed Mar 30, 2012
    This allows suffix sharing in register names. (AX is a suffix of EAX).
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153777 91177308-0d34-0410-b5e6-96231b3b80d8
  5. Reapply 153764 and 153761 with a fix.

    stoklund committed Mar 30, 2012
    Use an explicit comparator instead of the default.
    
    The sets are sorted, but not using the default comparator. Hopefully,
    this will unbreak the Linux builders.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153772 91177308-0d34-0410-b5e6-96231b3b80d8
  6. Revert 153764 and 153761. They broke a --enable-optimized --enable-as…

    Rafael Espindola
    Rafael Espindola committed Mar 30, 2012
    …sertions
    
    --enable-expensive-checks build.
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153771 91177308-0d34-0410-b5e6-96231b3b80d8