OSchip
/
llvm-project
mirror of https://github.com/THU-DSP-LAB/llvm-project.git
13
0
Fork 6
Code Issues Packages Projects Releases Wiki Activity
257,842 Commits 32 Branches 0 Tags 2.2 GiB
Commit Graph

79 Commits

Author SHA1 Message Date
Sanjay Patel 8c5f236197 [InstCombine] enable (X <<nsw C1) >>s C2 --> X <<nsw (C1 - C2) for vectors with splat constants 
llvm-svn: 293570
 2017-01-30 23:35:52 +00:00
Sanjay Patel abbb118a78 [InstCombine] add vector test for (X <<nsw C1) >>s C2 --> X <<nsw (C1 - C2); NFC 
llvm-svn: 293566
 2017-01-30 23:26:17 +00:00
Sanjay Patel 0c39d56a60 [InstCombine] enable more lshr(shl X, C1), C2 folds for vectors with splat constants 
llvm-svn: 293562
 2017-01-30 23:01:05 +00:00
Sanjay Patel 98cc841421 [InstCombine] add tests for more shift-shift patterns; NFC 
llvm-svn: 293555
 2017-01-30 22:24:36 +00:00
Sanjay Patel 373db5ba6c [InstCombine] enable (X >>?exact C1) << C2 --> X >>?exact (C1-C2) for vectors with splat constants 
llvm-svn: 293524
 2017-01-30 18:40:23 +00:00
Sanjay Patel 1a86607d38 [InstCombine] add vector splat tests for (X >>?exact C1) << C2 --> X >>?exact (C1-C2); NFC 
llvm-svn: 293517
 2017-01-30 18:17:14 +00:00
Sanjay Patel 77732d5033 [InstCombine] enable (X <<nsw C1) >>s C2 --> X <<nsw (C1-C2) for vectors with splat constants 
llvm-svn: 293507
 2017-01-30 17:19:32 +00:00
Sanjay Patel 8e644c08ee [InstCombine] fixed to propagate 'exact' on lshr 
The original shift is bigger, so this may qualify as 'obvious', 
but here's an attempt at an Alive-based proof:

Name: exact
Pre: (C1 u< C2)
%a = shl i8 %x, C1
%b = lshr exact i8 %a, C2 
  =>
%c = lshr exact i8 %x, C2 - C1
%b = and i8 %c, ((1 << width(C1)) - 1) u>> C2

Optimization is correct!

llvm-svn: 293498
 2017-01-30 16:53:03 +00:00
Sanjay Patel 5d6687da99 [InstCombine] add 'exact' to lshr to show that it got dropped; NFC 
llvm-svn: 293496
 2017-01-30 16:38:49 +00:00
Sanjay Patel 1196d7cd7f [InstCombine] enable lshr(shl X, C1), C2 folds for vectors with splat constants 
llvm-svn: 293489
 2017-01-30 16:11:40 +00:00
Sanjay Patel 127d64065a [InstCombine] add tests for shift-shift patterns; NFC 
llvm-svn: 293487
 2017-01-30 15:54:50 +00:00
Sanjay Patel 062adaab83 [InstCombine] enable (X >>?,exact C1) << C2 --> X << (C2 - C1) for vectors with splats 
llvm-svn: 293435
 2017-01-29 17:11:18 +00:00
Sanjay Patel c00574830f [InstCombine] add tests for shl(shr X, C1), C2 transforms; NFC 
llvm-svn: 293434
 2017-01-29 16:52:59 +00:00
Sanjay Patel ab8b32de71 [InstCombine] use m_APInt to allow shift-shift folds for vectors with splat constants 
Some existing 'FIXME' tests are still not folded because of splat holes in value tracking.

llvm-svn: 292151
 2017-01-16 19:35:45 +00:00
Sanjay Patel 21347ffddf [InstCombine] add tests to show missed vector folds; NFC 
Also, add comments and remove bogus comment.

llvm-svn: 292082
 2017-01-15 23:45:03 +00:00
Sanjay Patel 5f8451afad [InstCombine] use m_APInt to allow ashr folds for vectors with splat constants 
llvm-svn: 292064
 2017-01-15 16:38:19 +00:00
Sanjay Patel b22f6c5f26 [InstCombine] use m_APInt to allow shl folds for vectors with splat constants 
llvm-svn: 291934
 2017-01-13 18:39:09 +00:00
Sanjay Patel bbc1c1e46b [InstCombine] add tests to show missing transforms for vector shl; NFC 
llvm-svn: 291926
 2017-01-13 18:27:23 +00:00
David Majnemer cb892e9066 [InstCombine] Move casts around shift operations 
It is possible to perform a left shift before zero extending if the
shift would only shift out zeros.

llvm-svn: 290928
 2017-01-04 02:21:34 +00:00
Sanjay Patel f5887f1fbd [InstCombine] use m_APInt to allow icmp X, C folds for splat constant vectors 
isSignBitCheck could be changed to take a pointer param to avoid the 'UnusedBit' ugliness.

llvm-svn: 281231
 2016-09-12 16:25:41 +00:00
Sanjay Patel db400baa80 [InstCombine] add tests to show missing vector folds 
llvm-svn: 281219
 2016-09-12 15:51:42 +00:00
Sanjay Patel 5c5311f4e5 [InstCombine] use m_APInt to allow icmp (and X, Y), C folds for splat constant vectors 
llvm-svn: 279937
 2016-08-28 18:18:00 +00:00
Sanjay Patel d398d4a39e [InstCombine] use m_APInt to allow icmp eq/ne (shr X, C2), C folds for splat constant vectors 
llvm-svn: 279677
 2016-08-24 22:22:06 +00:00
Sanjay Patel 163a5ab799 remove FIXME comment; fixed by previous commit 
llvm-svn: 279400
 2016-08-21 16:40:42 +00:00
Sanjay Patel 7ffcde7422 [InstCombine] use m_APInt to allow icmp (shl X, Y), C folds for splat constant vectors, part 3 
This is a partial enablement (move the ConstantInt guard down).

llvm-svn: 279399
 2016-08-21 16:35:34 +00:00
Sanjay Patel 46a68ba618 [InstCombine] add tests for missing vector icmp folds 
llvm-svn: 278768
 2016-08-16 00:48:38 +00:00
Sanjay Patel aaf34d1bfc [InstCombine] add test for missing vector icmp fold 
llvm-svn: 278708
 2016-08-15 18:39:54 +00:00
Sanjay Patel 195eb9340a minimize test 
llvm-svn: 278707
 2016-08-15 18:35:44 +00:00
Sanjay Patel 3f506daf8c remove unnecessary IR comments about uses 
llvm-svn: 278705
 2016-08-15 18:32:50 +00:00
Sanjay Patel d391b0d69e [InstCombine] add tests for missing vector icmp folds 
llvm-svn: 278704
 2016-08-15 18:26:56 +00:00
Sanjay Patel beebe05af1 [InstCombine] add test for missing vector icmp fold 
llvm-svn: 278635
 2016-08-14 22:29:27 +00:00
Sanjay Patel e2e89ef936 [ValueTracking, InstCombine] extend isKnownToBeAPowerOfTwo() to handle vector splat constants 
We could try harder to handle non-splat vector constants too, 
but that seems much rarer to me.

Note that the div test isn't resolved because there's a check
for isIntegerTy() guarding that transform.

Differential Revision: http://reviews.llvm.org/D20497

llvm-svn: 270369
 2016-05-22 15:41:53 +00:00
Sanjay Patel 312c9afd90 add test for vector shift 
llvm-svn: 270284
 2016-05-20 22:08:16 +00:00
Sanjay Patel ebc0faa8d4 [InstCombine] regenerate checks 
llvm-svn: 268245
 2016-05-02 15:32:10 +00:00
Rafael Espindola d1beb07d39 Have a single way for creating unique value names. 
We had two code paths. One would create names like "foo.1" and the other
names like "foo1".

For globals it is important to use "foo.1" to help C++ name demangling.
For locals there is no strong reason to go one way or the other so I
kept the most common mangling (foo1).

llvm-svn: 253804
 2015-11-22 00:16:24 +00:00
Sunil Srivastava d79dfcbc37 Changed renaming of local symbols by inserting a dot vefore the numeric suffix. 
One code change and several test changes to match that
details in http://reviews.llvm.org/D9481

llvm-svn: 237150
 2015-05-12 16:47:30 +00:00
David Majnemer 5a7717e498 ConstantFold, InstSimplify: undef >>a x can be either -1 or 0, choose 0 
Zero is usually a nicer constant to have than -1.

llvm-svn: 223969
 2014-12-10 21:58:15 +00:00
Ankur Garg 51eeba70da [test/Transforms/InstCombine/shift.ll] Removed duplicate test cases. NFC. 
Removed some duplicate test cases from the file /test/Transforms/InstCombine/shift.ll.

test54 and test57 were duplicates of each other.
test55 and test58 were duplicates of each other.

(Removed test57 and test58)

llvm-svn: 223767
 2014-12-09 10:35:19 +00:00
Sonam Kumari 90d266c0a9 Removal Of Duplicate Test Case from shift.ll file 
llvm-svn: 223648
 2014-12-08 09:40:43 +00:00
Matt Arsenault fed3dc8dc6 Revert "Revert r206045, "Fix shift by constants for vector."" 
Fix cases where the Value itself is used, and not the constant value.

llvm-svn: 206214
 2014-04-14 21:50:37 +00:00
NAKAMURA Takumi 58ad0c87f8 Whitespace. 
llvm-svn: 206154
 2014-04-14 07:03:13 +00:00
NAKAMURA Takumi 26afa982ec Revert r206045, "Fix shift by constants for vector." 
It broke some builders, at least, i686.

llvm-svn: 206153
 2014-04-14 07:02:57 +00:00
Matt Arsenault 173a1e577c Fix shift by constants for vector. 
ashr <N x iM>, <N x iM> M -> undef

llvm-svn: 206045
 2014-04-11 17:57:53 +00:00
Benjamin Kramer 010f108382 InstCombine: Check for zero shift amounts before subtracting one causing integer overflow. 
PR17026. Also avoid undefined shifts and shift amounts larger than 64 bits
(those are always undef because we can't represent integer types that large).

llvm-svn: 189672
 2013-08-30 14:35:35 +00:00
Stephen Lin c1c7a1309c Update Transforms tests to use CHECK-LABEL for easier debugging. No functionality change. 
This update was done with the following bash script:

  find test/Transforms -name "*.ll" | \
  while read NAME; do
    echo "$NAME"
    if ! grep -q "^; *RUN: *llc" $NAME; then
      TEMP=`mktemp -t temp`
      cp $NAME $TEMP
      sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \
      while read FUNC; do
        sed -i '' "s/;\(.*\)\([A-Za-z0-9_]*\):\( *\)@$FUNC\([( ]*\)\$/;\1\2-LABEL:\3@$FUNC(/g" $TEMP
      done
      mv $TEMP $NAME
    fi
  done

llvm-svn: 186268
 2013-07-14 01:42:54 +00:00
David Majnemer bafa537eb7 Commit r185909 was a misapplied patch, fix it 
llvm-svn: 185910
 2013-07-09 07:58:32 +00:00
Nadav Rotem 4349f6963e Revert r174152. The shift amount may overflow and in that case this transformation is illegal. 
llvm-svn: 174156
 2013-02-01 07:59:33 +00:00
Nadav Rotem 1d584029ae Optimize shift lefts of a constant by a value plus constant into a single shift. 
llvm-svn: 174152
 2013-02-01 06:45:40 +00:00
Shuxin Yang 81b3678564 - Fix a problematic way in creating all-the-1 APInt. 
- Propagate "exact" bit of [l|a]shr instruction.

llvm-svn: 169942
 2012-12-12 00:29:03 +00:00
Shuxin Yang 73285933c9 For rdar://12329730, last piece. 
This change attempts to simplify (X^Y) -> X or Y in the user's context if we know that
only bits from X or Y are demanded.

  A minimized case is provided bellow. This change will simplify "t>>16" into "var1 >>16".

  =============================================================
  unsigned foo (unsigned val1, unsigned val2) {
    unsigned t = val1 ^ 1234;
    return (t >> 16) | t; // NOTE: t is used more than once.
  }
  =============================================================

  Note that if the "t" were used only once, the expression would be finally optimized as well.
However, with with this change, the optimization will take place earlier.

  Reviewed by Nadav, Thanks a lot!

llvm-svn: 169317
 2012-12-04 22:15:32 +00:00