1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-21 20:12:56 +02:00
Commit Graph

286 Commits

Author SHA1 Message Date
Erik Verbruggen
11e61b79e5 Revert "InstCombine: merge constants in both operands of icmp."
This reverts commit r204912, and follow-up commit r204948.

This introduced a performance regression, and the fix is not completely
clear yet.

llvm-svn: 205010
2014-03-28 14:50:57 +00:00
Erik Verbruggen
5e4efd4306 InstCombine: merge constants in both operands of icmp.
Transform:
    icmp X+Cst2, Cst
into:
    icmp X, Cst-Cst2
when Cst-Cst2 does not overflow, and the add has nsw.

llvm-svn: 204912
2014-03-27 11:16:05 +00:00
Shuxin Yang
6dd1fc357c Remove two popcount patterns which we are already able to recognize.
llvm-svn: 170158
2012-12-13 23:16:19 +00:00
Micah Villmow
bb1a25cd67 Move TargetData to DataLayout.
llvm-svn: 165402
2012-10-08 16:38:25 +00:00
Benjamin Kramer
dca12ad159 Fabs folding is implemented.
llvm-svn: 162186
2012-08-19 09:51:44 +00:00
Micah Villmow
759c8c46ac Test revert of test changes.
llvm-svn: 160632
2012-07-23 16:42:45 +00:00
Micah Villmow
0482f60db4 Test commit.
llvm-svn: 160631
2012-07-23 16:37:24 +00:00
Benjamin Kramer
35d213e1c2 Add a microoptimization note.
llvm-svn: 159082
2012-06-23 15:19:31 +00:00
Benjamin Kramer
c672ae3ee2 Remove README entry obsoleted by register masks.
llvm-svn: 154588
2012-04-12 12:47:29 +00:00
Benjamin Kramer
b3055f03e1 Add another note about a missed compare with nsw arithmetic instcombine.
llvm-svn: 153574
2012-03-28 10:50:18 +00:00
Benjamin Kramer
4be5a4bca4 Add a note about a cute little fabs optimization.
llvm-svn: 153543
2012-03-27 22:42:42 +00:00
Benjamin Kramer
cc6159f7d9 Add two missed instcombines related to compares with nsw arithmetic.
llvm-svn: 153542
2012-03-27 22:03:19 +00:00
Benjamin Kramer
f4e9cbfc05 Add two notes for correlated-expression optimizations.
llvm-svn: 139263
2011-09-07 22:49:26 +00:00
Benjamin Kramer
1cab6179ab Don't emit a bit test if there is only one case the test can yield false. A simple SETNE is sufficient.
llvm-svn: 135126
2011-07-14 01:38:42 +00:00
Benjamin Kramer
5079b61657 InstCombine: Fold A-b == C --> b == A-C if A and C are constants.
The backend already knew this trick.

llvm-svn: 132915
2011-06-13 15:24:24 +00:00
Chris Lattner
f58c2c02b5 clarify this, apparently it is confusing :)
llvm-svn: 131916
2011-05-23 20:17:44 +00:00
Chris Lattner
3066a7e5ad add a note.
llvm-svn: 131863
2011-05-22 18:28:46 +00:00
Chris Lattner
9359dea740 move PR9408 here.
llvm-svn: 131841
2011-05-22 05:45:06 +00:00
Chris Lattner
01ceb99a05 Transform: "icmp eq (trunc (lshr(X, cst1)), cst" to "icmp (and X, mask), cst"
when X has multiple uses.  This is useful for exposing secondary optimizations,
but the X86 backend isn't ready for this when X has a single use.  For example,
this can disable load folding.

This is inching towards resolving PR6627.

llvm-svn: 130238
2011-04-26 20:18:20 +00:00
Chris Lattner
35166f3b22 add a missed bitfield instcombine.
llvm-svn: 130137
2011-04-25 18:44:26 +00:00
Benjamin Kramer
f6eab5f86e DAGCombine: fold "(zext x) == C" into "x == (trunc C)" if the trunc is lossless.
On x86 this allows to fold a load into the cmp, greatly reducing register pressure.
  movzbl	(%rdi), %eax
  cmpl	$47, %eax
->
  cmpb	$47, (%rdi)

This shaves 8k off gcc.o on i386. I'll leave applying the patch in README.txt to Chris :)

llvm-svn: 130005
2011-04-22 18:47:44 +00:00
Chris Lattner
5fa2544086 add a minor missed dag combine that is blocking mid-level optimization
improvements, that will lead to fixing PR6627.

llvm-svn: 129504
2011-04-14 04:21:42 +00:00
Benjamin Kramer
12ea727553 Add a note.
llvm-svn: 128286
2011-03-25 17:32:40 +00:00
Eli Friedman
f9d5010081 A bit more analysis of a memset-related README entry.
llvm-svn: 128107
2011-03-22 20:49:53 +00:00
Eli Friedman
0da0b4aed0 This README entry was fixed recently.
llvm-svn: 127982
2011-03-21 01:33:03 +00:00
Chris Lattner
871d62dc5b add a note
llvm-svn: 126719
2011-03-01 00:24:51 +00:00
Benjamin Kramer
412ffed4f0 Add some DAGCombines for (adde 0, 0, glue), which are useful to optimize legalized code for large integer arithmetic.
1. Inform users of ADDEs with two 0 operands that it never sets carry
2. Fold other ADDs or ADDCs into the ADDE if possible

It would be neat if we could do the same thing for SETCC+ADD eventually, but we can't do that in target independent code.

llvm-svn: 126557
2011-02-26 22:48:07 +00:00
Chris Lattner
c373140c8b add a missed loop deletion case.
llvm-svn: 126103
2011-02-21 02:13:39 +00:00
Chris Lattner
8760c28fe1 add an idiom that loop idiom could theoretically catch.
llvm-svn: 126101
2011-02-21 01:33:38 +00:00
Duncan Sands
e0ece264ba This has been implemented.
llvm-svn: 125738
2011-02-17 08:16:56 +00:00
Chris Lattner
035876162f add some notes on compares + binops. Remove redundant entries.
llvm-svn: 125702
2011-02-17 01:43:46 +00:00
Chris Lattner
9f4e529571 Add a few missed xforms from GCC PR14753
llvm-svn: 125681
2011-02-16 19:16:34 +00:00
Eli Friedman
b409f8da64 Remove outdated README entry.
llvm-svn: 125660
2011-02-16 07:41:19 +00:00
Eli Friedman
5f848d70fa Remove outdated README entry.
llvm-svn: 125659
2011-02-16 07:18:18 +00:00
Eli Friedman
30a64ae1b9 Update README entry.
llvm-svn: 125658
2011-02-16 07:17:44 +00:00
Anders Carlsson
61f2126479 Remove a virtual inheritance case that clang can devirtualize fully now.
llvm-svn: 124989
2011-02-06 20:16:49 +00:00
Benjamin Kramer
b739613711 SimplifyCFG: Turn switches into sub+icmp+branch if possible.
This makes the job of the later optzn passes easier, allowing the vast amount of
icmp transforms to chew on it.

We transform 840 switches in gcc.c, leading to a 16k byte shrink of the resulting
binary on i386-linux.

The testcase from README.txt now compiles into
  decl  %edi
  cmpl  $3, %edi
  sbbl  %eax, %eax
  andl  $1, %eax
  ret

llvm-svn: 124724
2011-02-02 15:56:22 +00:00
Chris Lattner
1d534245fc add a note, progress unblocked by PR8575 being fixed.
llvm-svn: 124599
2011-01-31 20:23:28 +00:00
Benjamin Kramer
6b3c3de09a Teach DAGCombine to fold fold (sra (trunc (sr x, c1)), c2) -> (trunc (sra x, c1+c2) when c1 equals the amount of bits that are truncated off.
This happens all the time when a smul is promoted to a larger type.

On x86-64 we now compile "int test(int x) { return x/10; }" into
  movslq  %edi, %rax
  imulq $1717986919, %rax, %rax
  movq  %rax, %rcx
  shrq  $63, %rcx
  sarq  $34, %rax <- used to be "shrq $32, %rax; sarl $2, %eax"
  addl  %ecx, %eax

This fires 96 times in gcc.c on x86-64.

llvm-svn: 124559
2011-01-30 16:38:43 +00:00
Chris Lattner
c025bec9e1 this isn't a memset, we do convert dest[i] to one though :)
llvm-svn: 124097
2011-01-24 02:32:00 +00:00
Chris Lattner
1f9ed3b437 with recent work, we now optimize this into:
define i32 @foo(i32 %x) nounwind readnone ssp {
entry:
  %tobool = icmp eq i32 %x, 0
  %tmp5 = select i1 %tobool, i32 2, i32 1
  ret i32 %tmp5
}

llvm-svn: 124091
2011-01-24 01:12:18 +00:00
Anders Carlsson
ba7114ef4c Add a memset loop that LoopIdiomRecognize doesn't recognize.
llvm-svn: 124082
2011-01-23 20:31:00 +00:00
Chris Lattner
08e1bf567f add a note
llvm-svn: 123752
2011-01-18 07:47:48 +00:00
Anders Carlsson
d0103ebf92 Update README.txt to remove the DAE enhancement.
llvm-svn: 123597
2011-01-16 21:26:15 +00:00
Chris Lattner
91f1b21cf1 add some commentary
llvm-svn: 123572
2011-01-16 06:39:44 +00:00
Chandler Carruth
a3261fcca5 Simplify a README.txt entry significantly to expose the core issue.
llvm-svn: 123556
2011-01-16 01:40:23 +00:00
Chris Lattner
d2d217dc46 typo
llvm-svn: 123406
2011-01-13 22:11:56 +00:00
Chris Lattner
6745cd150c memcpy + metadata = bliss :)
llvm-svn: 123405
2011-01-13 22:08:15 +00:00
Chandler Carruth
250dce460c Teach constant folding to perform conversions from constant floating
point values to their integer representation through the SSE intrinsic
calls. This is the last part of a README.txt entry for which I have real
world examples.

llvm-svn: 123206
2011-01-11 01:07:24 +00:00
Owen Anderson
4479341626 Fix a random missed optimization by making InstCombine more aggressive when determining which bits are demanded by
a comparison against a constant.

llvm-svn: 123203
2011-01-11 00:36:45 +00:00