Commit Graph

283 Commits

Author SHA1 Message Date
Nikolai Bozhenov 1925594ea0 [NFC] Use stdin for some tests instead of positional argument.
Summary: Otherwise unexpected matches with the path to the tests might happen.

Reviewers: rengolin, spatel, efriedma, RKSimon

Reviewed By: spatel

Subscribers: n.bozhenov, javed.absar, llvm-commits

Patch by Andrei Elovikov <andrei.elovikov@intel.com>

Differential Revision: https://reviews.llvm.org/D32994

llvm-svn: 306684
2017-06-29 14:51:54 +00:00
Matthias Braun 537d039104 RegScavenging: Add scavengeRegisterBackwards()
Re-apply r276044/r279124/r305516. Fixed a problem where we would refuse
to place spills as the very first instruciton of a basic block and thus
artifically increase pressure (test in
test/CodeGen/PowerPC/scavenging.mir:spill_at_begin)

This is a variant of scavengeRegister() that works for
enterBasicBlockEnd()/backward(). The benefit of the backward mode is
that it is not affected by incomplete kill flags.

This patch also changes
PrologEpilogInserter::doScavengeFrameVirtualRegs() to use the register
scavenger in backwards mode.

Differential Revision: http://reviews.llvm.org/D21885

llvm-svn: 305625
2017-06-17 02:08:18 +00:00
Matthias Braun 35530d7129 Revert "RegScavenging: Add scavengeRegisterBackwards()"
Revert because of reports of some PPC input starting to spill when it
was predicted that it wouldn't and no spillslot was reserved.

This reverts commit r305516.

llvm-svn: 305566
2017-06-16 17:48:08 +00:00
Matthias Braun a42c537912 RegScavenging: Add scavengeRegisterBackwards()
Re-apply r276044/r279124. Trying to reproduce or disprove the ppc64
problems reported in the stage2 build last time, which I cannot
reproduce right now.

This is a variant of scavengeRegister() that works for
enterBasicBlockEnd()/backward(). The benefit of the backward mode is
that it is not affected by incomplete kill flags.

This patch also changes
PrologEpilogInserter::doScavengeFrameVirtualRegs() to use the register
scavenger in backwards mode.

Differential Revision: http://reviews.llvm.org/D21885

llvm-svn: 305516
2017-06-15 22:14:55 +00:00
Krzysztof Parzyszek 6a0005d1b4 Move machine-cse-physreg.mir to test/CodeGen/Thumb
llvm-svn: 303778
2017-05-24 17:20:47 +00:00
Nirav Dave da8f221273 Elide stores which are overwritten without being observed.
Summary:
In SelectionDAG, when a store is immediately chained to another store
to the same address, elide the first store as it has no observable
effects. This is causes small improvements dealing with intrinsics
lowered to stores.

Test notes:

* Many testcases overwrite store addresses multiple times and needed
  minor changes, mainly making stores volatile to prevent the
  optimization from optimizing the test away.

* Many X86 test cases optimized out instructions associated with
  associated with va_start.

* Note that test_splat in CodeGen/AArch64/misched-stp.ll no longer has
  dependencies to check and can probably be removed and potentially
  replaced with another test.

Reviewers: rnk, john.brawn

Subscribers: aemerson, rengolin, qcolombet, jyknight, nemanjai, nhaehnle, javed.absar, llvm-commits

Differential Revision: https://reviews.llvm.org/D33206

llvm-svn: 303198
2017-05-16 19:43:56 +00:00
Artyom Skrobov 53cf1897cc [ARM] ScheduleDAGRRList::DelayForLiveRegsBottomUp must consider OptionalDefs
Summary:
D30400 has enabled tADC and tSBC instructions to be unglued, thereby allowing CPSR to remain live between Thumb1 scheduling units.

Most Thumb1 instructions have an OptionalDef for CPSR; but the scheduler ignored the OptionalDefs, and could unwittingly insert a flag-setting instruction in between an ADDS and the corresponding ADC.

Reviewers: javed.absar, atrick, MatzeB, t.p.northover, jmolloy, rengolin

Reviewed By: javed.absar

Subscribers: rogfer01, efriedma, aemerson, rengolin, llvm-commits, MatzeB

Differential Revision: https://reviews.llvm.org/D31081

llvm-svn: 301106
2017-04-23 06:58:08 +00:00
Artyom Skrobov 8d9643009f [Thumb1] The recently added tADCS and tSBCS pseudo-instructions were missing `Uses = [CPSR]`
Summary: Thanks to Oliver Stannard for helping catch this.

Reviewers: olista01, efriedma

Subscribers: llvm-commits, rengolin

Differential Revision: https://reviews.llvm.org/D31815

llvm-svn: 300951
2017-04-21 07:35:21 +00:00
Matt Arsenault f10061ec70 Add address space mangling to lifetime intrinsics
In preparation for allowing allocas to have non-0 addrspace.

llvm-svn: 299876
2017-04-10 20:18:21 +00:00
Artyom Skrobov 92c0653095 Reapply r298417 "[ARM] Recommit the glueless lowering of addc/adde in Thumb1"
The UB in t2_so_imm_neg conversion has been addressed under D31242 / r298512

This reverts commit r298482.

llvm-svn: 298562
2017-03-22 23:35:51 +00:00
Artyom Skrobov 50a066b313 [ARM] t2_so_imm_neg had a subtle bug in the conversion, and could trigger UB by negating (int)-2147483648. By pure luck, none of the pre-existing tests triggered this; so I'm adding one.
Summary: Thanks to Vitaly Buka for helping catch this.

Reviewers: rengolin, jmolloy, efriedma, vitalybuka

Subscribers: llvm-commits, aemerson

Differential Revision: https://reviews.llvm.org/D31242

llvm-svn: 298512
2017-03-22 15:09:30 +00:00
Vitaly Buka e69c137f90 Revert "[ARM] Recommit the glueless lowering of addc/adde in Thumb1, including the amended (no UB anymore) fix for adding/subtracting -2147483648."
Fails check-llvm with ubsan

This reverts commit r298417.

llvm-svn: 298482
2017-03-22 05:07:44 +00:00
Artyom Skrobov 40a4f40679 [ARM] Recommit the glueless lowering of addc/adde in Thumb1,
including the amended (no UB anymore) fix for adding/subtracting -2147483648.

This reverts r298328 "[ARM] Revert r297443 and r297820."
and partially reverts r297842 "Revert "[Thumb1] Fix the bug when adding/subtracting -2147483648""

llvm-svn: 298417
2017-03-21 18:39:41 +00:00
Eli Friedman 76732acc23 [ARM] Revert r297443 and r297820.
The glueless lowering of addc/adde in Thumb1 has known serious
miscompiles (see https://reviews.llvm.org/D31081), and r297820
causes an infinite loop for certain constructs.  It's not
clear when they will be fixed, so let's just take them out
of the tree for now.

(I resolved a small conflict with r297453.)

llvm-svn: 298328
2017-03-21 00:26:39 +00:00
Artyom Skrobov e72e1ba434 Revert "[Thumb1] Fix the bug when adding/subtracting -2147483648"
This reverts r297820 which apparently fails on A15 hosts.

llvm-svn: 297842
2017-03-15 14:50:43 +00:00
Artyom Skrobov 3fa5fd1dd2 [Thumb1] Fix the bug when adding/subtracting -2147483648
Differential Revision: https://reviews.llvm.org/D30829

llvm-svn: 297820
2017-03-15 10:19:16 +00:00
Nirav Dave 54e22f33d9 In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled.
Recommiting with compiler time improvements

    Recommitting after fixup of 32-bit aliasing sign offset bug in DAGCombiner.

    * Simplify Consecutive Merge Store Candidate Search

    Now that address aliasing is much less conservative, push through
    simplified store merging search and chain alias analysis which only
    checks for parallel stores through the chain subgraph. This is cleaner
    as the separation of non-interfering loads/stores from the
    store-merging logic.

    When merging stores search up the chain through a single load, and
    finds all possible stores by looking down from through a load and a
    TokenFactor to all stores visited.

    This improves the quality of the output SelectionDAG and the output
    Codegen (save perhaps for some ARM cases where we correctly constructs
    wider loads, but then promotes them to float operations which appear
    but requires more expensive constant generation).

    Some minor peephole optimizations to deal with improved SubDAG shapes (listed below)

    Additional Minor Changes:

      1. Finishes removing unused AliasLoad code

      2. Unifies the chain aggregation in the merged stores across code
         paths

      3. Re-add the Store node to the worklist after calling
         SimplifyDemandedBits.

      4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is
         arbitrary, but seems sufficient to not cause regressions in
         tests.

      5. Remove Chain dependencies of Memory operations on CopyfromReg
         nodes as these are captured by data dependence

      6. Forward loads-store values through tokenfactors containing
          {CopyToReg,CopyFromReg} Values.

      7. Peephole to convert buildvector of extract_vector_elt to
         extract_subvector if possible (see
         CodeGen/AArch64/store-merge.ll)

      8. Store merging for the ARM target is restricted to 32-bit as
         some in some contexts invalid 64-bit operations are being
         generated. This can be removed once appropriate checks are
         added.

    This finishes the change Matt Arsenault started in r246307 and
    jyknight's original patch.

    Many tests required some changes as memory operations are now
    reorderable, improving load-store forwarding. One test in
    particular is worth noting:

      CodeGen/PowerPC/ppc64-align-long-double.ll - Improved load-store
      forwarding converts a load-store pair into a parallel store and
      a memory-realized bitcast of the same value. However, because we
      lose the sharing of the explicit and implicit store values we
      must create another local store. A similar transformation
      happens before SelectionDAG as well.

    Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle

llvm-svn: 297695
2017-03-14 00:34:14 +00:00
Artyom Skrobov bf19d4bc29 [Thumb1] combine ADDC/SUBC with a negative immediate
Summary: This simple optimization has been split out of https://reviews.llvm.org/D30400

Reviewers: efriedma, jmolloy

Subscribers: llvm-commits, rengolin

Differential Revision: https://reviews.llvm.org/D30829

llvm-svn: 297682
2017-03-13 22:36:14 +00:00
Artyom Skrobov 0c93ceb5d8 For Thumb1, lower ADDC/ADDE/SUBC/SUBE via the glueless ARMISD nodes,
same as already done for ARM and Thumb2.

Reviewers: jmolloy, rogfer01, efriedma

Subscribers: aemerson, llvm-commits, rengolin

Differential Revision: https://reviews.llvm.org/D30400

llvm-svn: 297443
2017-03-10 07:40:27 +00:00
Artyom Skrobov 1388e2f792 In Thumb1, materialize a move between low registers as a `movs`, if CPSR isn't live.
Summary: Previously, it had always been materialized as a push/pop sequence.

Reviewers: labrinea, jroelofs

Reviewed By: jroelofs

Subscribers: llvm-commits, rengolin

Differential Revision: https://reviews.llvm.org/D30648

llvm-svn: 297134
2017-03-07 09:38:16 +00:00
Chandler Carruth ce52b80744 [SDAG] Revert r296476 (and r296486, r296668, r296690).
This patch causes compile times for some patterns to explode. I have
a (large, unreduced) test case that slows down by more than 20x and
several test cases slow down by 2x. I'm sending some of the test cases
directly to Nirav and following up with more details in the review log,
but this should unblock anyone else hitting this.

llvm-svn: 296862
2017-03-03 10:02:25 +00:00
Eli Friedman 36795239f5 [ARM] Don't generate deprecated T1 STM.
This prevents generating stm r1!, {r0, r1} on Thumb1, where value
stored for r1 is UNKONWN.

Patch by Zhaoshi Zheng.

Differential Revision: https://reviews.llvm.org/D27910

llvm-svn: 296538
2017-02-28 23:32:55 +00:00
Nirav Dave f830dec3f2 In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled.
Recommiting after fixup of 32-bit aliasing sign offset bug in DAGCombiner.

    * Simplify Consecutive Merge Store Candidate Search

    Now that address aliasing is much less conservative, push through
    simplified store merging search and chain alias analysis which only
    checks for parallel stores through the chain subgraph. This is cleaner
    as the separation of non-interfering loads/stores from the
    store-merging logic.

    When merging stores search up the chain through a single load, and
    finds all possible stores by looking down from through a load and a
    TokenFactor to all stores visited.

    This improves the quality of the output SelectionDAG and the output
    Codegen (save perhaps for some ARM cases where we correctly constructs
    wider loads, but then promotes them to float operations which appear
    but requires more expensive constant generation).

    Some minor peephole optimizations to deal with improved SubDAG shapes (listed below)

    Additional Minor Changes:

      1. Finishes removing unused AliasLoad code

      2. Unifies the chain aggregation in the merged stores across code
         paths

      3. Re-add the Store node to the worklist after calling
         SimplifyDemandedBits.

      4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is
         arbitrary, but seems sufficient to not cause regressions in
         tests.

      5. Remove Chain dependencies of Memory operations on CopyfromReg
         nodes as these are captured by data dependence

      6. Forward loads-store values through tokenfactors containing
          {CopyToReg,CopyFromReg} Values.

      7. Peephole to convert buildvector of extract_vector_elt to
         extract_subvector if possible (see
         CodeGen/AArch64/store-merge.ll)

      8. Store merging for the ARM target is restricted to 32-bit as
         some in some contexts invalid 64-bit operations are being
         generated. This can be removed once appropriate checks are
         added.

    This finishes the change Matt Arsenault started in r246307 and
    jyknight's original patch.

    Many tests required some changes as memory operations are now
    reorderable, improving load-store forwarding. One test in
    particular is worth noting:

      CodeGen/PowerPC/ppc64-align-long-double.ll - Improved load-store
      forwarding converts a load-store pair into a parallel store and
      a memory-realized bitcast of the same value. However, because we
      lose the sharing of the explicit and implicit store values we
      must create another local store. A similar transformation
      happens before SelectionDAG as well.

    Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle

llvm-svn: 296476
2017-02-28 14:24:15 +00:00
Artyom Skrobov 24a593fd20 Relate the CHECK: lines to the functions that they're checking [NFC]
llvm-svn: 296450
2017-02-28 08:58:40 +00:00
Nirav Dave 73cd0194cf Revert "In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled."
This reverts commit r296252 until 256-bit operations are more efficiently generated in X86.

llvm-svn: 296279
2017-02-26 01:27:32 +00:00
Nirav Dave beabf456df In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled.
Recommiting after fixup of 32-bit aliasing sign offset bug in DAGCombiner.

    * Simplify Consecutive Merge Store Candidate Search

    Now that address aliasing is much less conservative, push through
    simplified store merging search and chain alias analysis which only
    checks for parallel stores through the chain subgraph. This is cleaner
    as the separation of non-interfering loads/stores from the
    store-merging logic.

    When merging stores search up the chain through a single load, and
    finds all possible stores by looking down from through a load and a
    TokenFactor to all stores visited.

    This improves the quality of the output SelectionDAG and the output
    Codegen (save perhaps for some ARM cases where we correctly constructs
    wider loads, but then promotes them to float operations which appear
    but requires more expensive constant generation).

    Some minor peephole optimizations to deal with improved SubDAG shapes (listed below)

    Additional Minor Changes:

      1. Finishes removing unused AliasLoad code

      2. Unifies the chain aggregation in the merged stores across code
         paths

      3. Re-add the Store node to the worklist after calling
         SimplifyDemandedBits.

      4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is
         arbitrary, but seems sufficient to not cause regressions in
         tests.

      5. Remove Chain dependencies of Memory operations on CopyfromReg
         nodes as these are captured by data dependence

      6. Forward loads-store values through tokenfactors containing
          {CopyToReg,CopyFromReg} Values.

      7. Peephole to convert buildvector of extract_vector_elt to
         extract_subvector if possible (see
         CodeGen/AArch64/store-merge.ll)

      8. Store merging for the ARM target is restricted to 32-bit as
         some in some contexts invalid 64-bit operations are being
         generated. This can be removed once appropriate checks are
         added.

    This finishes the change Matt Arsenault started in r246307 and
    jyknight's original patch.

    Many tests required some changes as memory operations are now
    reorderable, improving load-store forwarding. One test in
    particular is worth noting:

      CodeGen/PowerPC/ppc64-align-long-double.ll - Improved load-store
      forwarding converts a load-store pair into a parallel store and
      a memory-realized bitcast of the same value. However, because we
      lose the sharing of the explicit and implicit store values we
      must create another local store. A similar transformation
      happens before SelectionDAG as well.

    Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle

llvm-svn: 296252
2017-02-25 11:43:58 +00:00
Roger Ferrer Ibanez 56db97d4de [ARM] Fix constant islands pass.
The pass tries to fix a spill of LR that turns out to be unnecessary.
So it removes the tPOP but forgets to remove tPUSH.
This causes the stack be misaligned upon returning the function.

Thus, remove the tPUSH as well in this case.

Differential Revision: https://reviews.llvm.org/D30207

llvm-svn: 295816
2017-02-22 09:06:21 +00:00
Artyom Skrobov 4592f6206c In Thumb1 mode, the custom lowering for ARMISD::CMPZ could never emit tADDi3
Reviewers: jmolloy, t.p.northover

Reviewed By: t.p.northover

Subscribers: t.p.northover, aemerson, rengolin, llvm-commits

Differential Revision: https://reviews.llvm.org/D30097

llvm-svn: 295478
2017-02-17 18:59:16 +00:00
James Molloy 92497542e7 [Thumb-1] TBB generation: spot redefinitions of index register
We match a sequence of 3-4 instructions into a tTBB pseudo. One of our checks is that
a particular register in that sequence is killed (so it can be clobbered by the pseudo).

We weren't noticing if an errant MOV or other instruction had infiltrated the
sequence we were walking. If it had, and it defined the register we've already
identified as killed, it makes it live across the tBR_JT and thus unclobberable.

Notice this case and bail out.

llvm-svn: 294949
2017-02-13 14:07:39 +00:00
Sanne Wouda 57b63d6ade [LLC] Add an inline assembly diagnostics handler.
Summary:
llc would hit a fatal error for errors in inline assembly. The
diagnostics message is now printed.

Reviewers: rengolin, MatzeB, javed.absar, anemet

Reviewed By: anemet

Subscribers: jyknight, nemanjai, llvm-commits

Differential Revision: https://reviews.llvm.org/D29408

llvm-svn: 293999
2017-02-03 11:14:39 +00:00
Nirav Dave 93f9d5ce04 Revert "In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled."
This reverts commit r293893 which is miscompiling lua on ARM and
bootstrapping for x86-windows.

llvm-svn: 293915
2017-02-02 18:24:55 +00:00
Nirav Dave 4442667fc5 In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled.
Recommiting after fixing X86 inc/dec chain bug.

    * Simplify Consecutive Merge Store Candidate Search

    Now that address aliasing is much less conservative, push through
    simplified store merging search and chain alias analysis which only
    checks for parallel stores through the chain subgraph. This is cleaner
    as the separation of non-interfering loads/stores from the
    store-merging logic.

    When merging stores search up the chain through a single load, and
    finds all possible stores by looking down from through a load and a
    TokenFactor to all stores visited.

    This improves the quality of the output SelectionDAG and the output
    Codegen (save perhaps for some ARM cases where we correctly constructs
    wider loads, but then promotes them to float operations which appear
    but requires more expensive constant generation).

    Some minor peephole optimizations to deal with improved SubDAG shapes (listed below)

    Additional Minor Changes:

      1. Finishes removing unused AliasLoad code

      2. Unifies the chain aggregation in the merged stores across code
         paths

      3. Re-add the Store node to the worklist after calling
         SimplifyDemandedBits.

      4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is
         arbitrary, but seems sufficient to not cause regressions in
         tests.

      5. Remove Chain dependencies of Memory operations on CopyfromReg
         nodes as these are captured by data dependence

      6. Forward loads-store values through tokenfactors containing
          {CopyToReg,CopyFromReg} Values.

      7. Peephole to convert buildvector of extract_vector_elt to
         extract_subvector if possible (see
         CodeGen/AArch64/store-merge.ll)

      8. Store merging for the ARM target is restricted to 32-bit as
         some in some contexts invalid 64-bit operations are being
         generated. This can be removed once appropriate checks are
         added.

    This finishes the change Matt Arsenault started in r246307 and
    jyknight's original patch.

    Many tests required some changes as memory operations are now
    reorderable, improving load-store forwarding. One test in
    particular is worth noting:

      CodeGen/PowerPC/ppc64-align-long-double.ll - Improved load-store
      forwarding converts a load-store pair into a parallel store and
      a memory-realized bitcast of the same value. However, because we
      lose the sharing of the explicit and implicit store values we
      must create another local store. A similar transformation
      happens before SelectionDAG as well.

    Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle

llvm-svn: 293893
2017-02-02 14:39:42 +00:00
Kyle Butt b15c06677c CodeGen: Allow small copyable blocks to "break" the CFG.
When choosing the best successor for a block, ordinarily we would have preferred
a block that preserves the CFG unless there is a strong probability the other
direction. For small blocks that can be duplicated we now skip that requirement
as well, subject to some simple frequency calculations.

Differential Revision: https://reviews.llvm.org/D28583

llvm-svn: 293716
2017-01-31 23:48:32 +00:00
Nirav Dave d32a421f75 Revert "In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled."
This reverts commit r293184 which is failing in LTO builds

llvm-svn: 293188
2017-01-26 16:46:13 +00:00
Nirav Dave de6516c466 In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled.
* Simplify Consecutive Merge Store Candidate Search

    Now that address aliasing is much less conservative, push through
    simplified store merging search and chain alias analysis which only
    checks for parallel stores through the chain subgraph. This is cleaner
    as the separation of non-interfering loads/stores from the
    store-merging logic.

    When merging stores search up the chain through a single load, and
    finds all possible stores by looking down from through a load and a
    TokenFactor to all stores visited.

    This improves the quality of the output SelectionDAG and the output
    Codegen (save perhaps for some ARM cases where we correctly constructs
    wider loads, but then promotes them to float operations which appear
    but requires more expensive constant generation).

    Some minor peephole optimizations to deal with improved SubDAG shapes (listed below)

    Additional Minor Changes:

      1. Finishes removing unused AliasLoad code

      2. Unifies the chain aggregation in the merged stores across code
         paths

      3. Re-add the Store node to the worklist after calling
         SimplifyDemandedBits.

      4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is
         arbitrary, but seems sufficient to not cause regressions in
         tests.

      5. Remove Chain dependencies of Memory operations on CopyfromReg
         nodes as these are captured by data dependence

      6. Forward loads-store values through tokenfactors containing
          {CopyToReg,CopyFromReg} Values.

      7. Peephole to convert buildvector of extract_vector_elt to
         extract_subvector if possible (see
         CodeGen/AArch64/store-merge.ll)

      8. Store merging for the ARM target is restricted to 32-bit as
         some in some contexts invalid 64-bit operations are being
         generated. This can be removed once appropriate checks are
         added.

    This finishes the change Matt Arsenault started in r246307 and
    jyknight's original patch.

    Many tests required some changes as memory operations are now
    reorderable, improving load-store forwarding. One test in
    particular is worth noting:

      CodeGen/PowerPC/ppc64-align-long-double.ll - Improved load-store
      forwarding converts a load-store pair into a parallel store and
      a memory-realized bitcast of the same value. However, because we
      lose the sharing of the explicit and implicit store values we
      must create another local store. A similar transformation
      happens before SelectionDAG as well.

    Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle

llvm-svn: 293184
2017-01-26 16:02:24 +00:00
Kyle Butt efe56fed12 Revert "CodeGen: Allow small copyable blocks to "break" the CFG."
This reverts commit ada6595a526d71df04988eb0a4b4fe84df398ded.

This needs a simple probability check because there are some cases where it is
not profitable.

llvm-svn: 291695
2017-01-11 19:55:19 +00:00
Kyle Butt df27aa8c89 CodeGen: Allow small copyable blocks to "break" the CFG.
When choosing the best successor for a block, ordinarily we would have preferred
a block that preserves the CFG unless there is a strong probability the other
direction. For small blocks that can be duplicated we now skip that requirement
as well.

Differential revision: https://reviews.llvm.org/D27742

llvm-svn: 291609
2017-01-10 23:04:30 +00:00
Sjoerd Meijer 96e10b5a9e [Thumb] Teach ISel how to lower compares of AND bitmasks efficiently
This is essentially a recommit of r285893, but with a correctness fix. The
problem of the original commit was that this:

bic r5, r7, #31
cbz r5, .LBB2_10

got rewritten into:

lsrs  r5, r7, #5
beq .LBB2_10

The result in destination register r5 is not the same and this is incorrect
when r5 is not dead. So this fix includes checking the uses of the AND
destination register. And also, compared to the original commit, some regression
tests didn't need changing anymore because of this extra check.

For completeness, this was the original commit message:

For the common pattern (CMPZ (AND x, #bitmask), #0), we can do some more
efficient instruction selection if the bitmask is one consecutive sequence of
set bits (32 - clz(bm) - ctz(bm) == popcount(bm)).

1) If the bitmask touches the LSB, then we can remove all the upper bits and
set the flags by doing one LSLS.
2) If the bitmask touches the MSB, then we can remove all the lower bits and
set the flags with one LSRS.
3) If the bitmask has popcount == 1 (only one set bit), we can shift that bit
into the sign bit with one LSLS and change the condition query from NE/EQ to
MI/PL (we could also implement this by shifting into the carry bit and
branching on BCC/BCS).
4) Otherwise, we can emit a sequence of LSLS+LSRS to remove the upper and lower
zero bits of the mask.

1-3 require only one 16-bit instruction and can elide the CMP. 4 requires two
16-bit instructions but can elide the CMP and doesn't require materializing a
complex immediate, so is also a win.

Differential Revision: https://reviews.llvm.org/D27761

llvm-svn: 289794
2016-12-15 09:38:59 +00:00
Nirav Dave f5bf03c7ef Revert "In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled."
Reverting due to ARM MCJIT and MIPS LLD error.

This reverts commit r289659.

llvm-svn: 289667
2016-12-14 16:43:44 +00:00
Nirav Dave 8527ab0ad2 In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled.
Retrying after fixing after removing load-store factoring through
token factors in favor of improved token factor operand pruning

Simplify Consecutive Merge Store Candidate Search

Now that address aliasing is much less conservative, push through
simplified store merging search which only checks for parallel stores
through the chain subgraph. This is cleaner as the separation of
non-interfering loads/stores from the store-merging logic.

Whem merging stores, search up the chain through a single load, and
finds all possible stores by looking down from through a load and a
TokenFactor to all stores visited. This improves the quality of the
output SelectionDAG and generally the output CodeGen (with some
exceptions).

Additional Minor Changes:

   1. Finishes removing unused AliasLoad code
   2. Unifies the the chain aggregation in the merged stores across
      code paths
   3. Re-add the Store node to the worklist after calling
      SimplifyDemandedBits.
   4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is
      arbitrary, but seemed sufficient to not cause regressions in
      tests.

This finishes the change Matt Arsenault started in r246307 and
jyknight's original patch.

Many tests required some changes as memory operations are now
reorderable. Some tests relying on the order were changed to use
volatile memory operations

Noteworthy tests:

    CodeGen/AArch64/argument-blocks.ll -
      It's not entirely clear what the test_varargs_stackalign test is
      supposed to be asserting, but the new code looks right.

    CodeGen/AArch64/arm64-memset-inline.lli -
    CodeGen/AArch64/arm64-stur.ll -
    CodeGen/ARM/memset-inline.ll -

      The backend now generates *worse* code due to store merging
      succeeding, as we do do a 16-byte constant-zero store efficiently.

    CodeGen/AArch64/merge-store.ll -
      Improved, but there still seems to be an extraneous vector insert
      from an element to itself?

    CodeGen/PowerPC/ppc64-align-long-double.ll -
      Worse code emitted in this case, due to the improved store->load
      forwarding.

    CodeGen/X86/dag-merge-fast-accesses.ll -
    CodeGen/X86/MergeConsecutiveStores.ll -
    CodeGen/X86/stores-merging.ll -
    CodeGen/Mips/load-store-left-right.ll -
      Restored correct merging of non-aligned stores

    CodeGen/AMDGPU/promote-alloca-stored-pointer-value.ll -
      Improved. Correctly merges buffer_store_dword calls

    CodeGen/AMDGPU/si-triv-disjoint-mem-access.ll -
      Improved. Sidesteps loading a stored value and
      merges two stores

    CodeGen/X86/pr18023.ll -
      This test has been removed, as it was asserting incorrect
      behavior. Non-volatile stores *CAN* be moved past volatile loads,
      and now are.

    CodeGen/X86/vector-idiv.ll -
    CodeGen/X86/vector-lzcnt-128.ll -
      It's basically impossible to tell what these tests are actually
      testing. But, looks like the code got better due to the memory
      operations being recognized as non-aliasing.

    CodeGen/X86/win32-eh.ll -
      Both loads of the securitycookie are now merged.

Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle

Subscribers: wdng, nhaehnle, nemanjai, arsenm, weimingz, niravd, RKSimon, aemerson, qcolombet, dsanders, resistor, tstellarAMD, t.p.northover, spatel

Differential Revision: https://reviews.llvm.org/D14834

llvm-svn: 289659
2016-12-14 15:44:26 +00:00
Nirav Dave bedb5d906c Revert "In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled."
This reverts commit r289221 which appears to be triggering an assertion

llvm-svn: 289226
2016-12-09 17:18:24 +00:00
Nirav Dave fd51ff4fd8 In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled.
Retrying after fixing overly aggressive load-store forwarding optimization.

Simplify Consecutive Merge Store Candidate Search

Now that address aliasing is much less conservative, push through
simplified store merging search which only checks for parallel stores
through the chain subgraph. This is cleaner as the separation of
non-interfering loads/stores from the store-merging logic.

Whem merging stores, search up the chain through a single load, and
finds all possible stores by looking down from through a load and a
TokenFactor to all stores visited. This improves the quality of the
output SelectionDAG and generally the output CodeGen (with some
exceptions).

Additional Minor Changes:

   1. Finishes removing unused AliasLoad code
   2. Unifies the the chain aggregation in the merged stores across
      code paths
   3. Re-add the Store node to the worklist after calling
      SimplifyDemandedBits.
   4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is
      arbitrary, but seemed sufficient to not cause regressions in
      tests.

This finishes the change Matt Arsenault started in r246307 and
jyknight's original patch.

Many tests required some changes as memory operations are now
reorderable. Some tests relying on the order were changed to use
volatile memory operations

Noteworthy tests:

    CodeGen/AArch64/argument-blocks.ll -
      It's not entirely clear what the test_varargs_stackalign test is
      supposed to be asserting, but the new code looks right.

    CodeGen/AArch64/arm64-memset-inline.lli -
    CodeGen/AArch64/arm64-stur.ll -
    CodeGen/ARM/memset-inline.ll -

      The backend now generates *worse* code due to store merging
      succeeding, as we do do a 16-byte constant-zero store efficiently.

    CodeGen/AArch64/merge-store.ll -
      Improved, but there still seems to be an extraneous vector insert
      from an element to itself?

    CodeGen/PowerPC/ppc64-align-long-double.ll -
      Worse code emitted in this case, due to the improved store->load
      forwarding.

    CodeGen/X86/dag-merge-fast-accesses.ll -
    CodeGen/X86/MergeConsecutiveStores.ll -
    CodeGen/X86/stores-merging.ll -
    CodeGen/Mips/load-store-left-right.ll -
      Restored correct merging of non-aligned stores

    CodeGen/AMDGPU/promote-alloca-stored-pointer-value.ll -
      Improved. Correctly merges buffer_store_dword calls

    CodeGen/AMDGPU/si-triv-disjoint-mem-access.ll -
      Improved. Sidesteps loading a stored value and
      merges two stores

    CodeGen/X86/pr18023.ll -
      This test has been removed, as it was asserting incorrect
      behavior. Non-volatile stores *CAN* be moved past volatile loads,
      and now are.

    CodeGen/X86/vector-idiv.ll -
    CodeGen/X86/vector-lzcnt-128.ll -
      It's basically impossible to tell what these tests are actually
      testing. But, looks like the code got better due to the memory
      operations being recognized as non-aliasing.

    CodeGen/X86/win32-eh.ll -
      Both loads of the securitycookie are now merged.

Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle

Subscribers: wdng, nhaehnle, nemanjai, arsenm, weimingz, niravd, RKSimon, aemerson, qcolombet, dsanders, resistor, tstellarAMD, t.p.northover, spatel

Differential Revision: https://reviews.llvm.org/D14834

llvm-svn: 289221
2016-12-09 16:15:12 +00:00
James Molloy e7d97368f2 Revert "[Thumb] Teach ISel how to lower compares of AND bitmasks efficiently"
This reverts commit r285893. It caused (probably) http://lab.llvm.org:8011/builders/clang-cmake-thumbv7-a15-full-sh/builds/83 .

llvm-svn: 285912
2016-11-03 14:08:01 +00:00
James Molloy b60d8b1987 [Thumb] Teach ISel how to lower compares of AND bitmasks efficiently
This recommits r281323, which was backed out for two reasons. One, a selfhost failure, and two, it apparently caused Chromium failures. Actually, the latter was a red herring. The log has expired from the former, but I suspect that was a red herring too (actually caused by another problematic patch of mine). Therefore reapplying, and will watch the bots like a hawk.

For the common pattern (CMPZ (AND x, #bitmask), #0), we can do some more efficient instruction selection if the bitmask is one consecutive sequence of set bits (32 - clz(bm) - ctz(bm) == popcount(bm)).

1) If the bitmask touches the LSB, then we can remove all the upper bits and set the flags by doing one LSLS.
2) If the bitmask touches the MSB, then we can remove all the lower bits and set the flags with one LSRS.
3) If the bitmask has popcount == 1 (only one set bit), we can shift that bit into the sign bit with one LSLS and change the condition query from NE/EQ to MI/PL (we could also implement this by shifting into the carry bit and branching on BCC/BCS).
4) Otherwise, we can emit a sequence of LSLS+LSRS to remove the upper and lower zero bits of the mask.

1-3 require only one 16-bit instruction and can elide the CMP. 4 requires two 16-bit instructions but can elide the CMP and doesn't require materializing a complex immediate, so is also a win.

llvm-svn: 285893
2016-11-03 10:18:20 +00:00
Nirav Dave a81682aad4 Revert "In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled."
This reverts commit r284151 which appears to be triggering a LTO
failures on Hexagon

llvm-svn: 284157
2016-10-13 20:23:25 +00:00
Nirav Dave 4b36957243 In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled.
Retrying after upstream changes.

   Simplify Consecutive Merge Store Candidate Search

   Now that address aliasing is much less conservative, push through
   simplified store merging search which only checks for parallel stores
   through the chain subgraph. This is cleaner as the separation of
   non-interfering loads/stores from the store-merging logic.

   Whem merging stores, search up the chain through a single load, and
   finds all possible stores by looking down from through a load and a
   TokenFactor to all stores visited. This improves the quality of the
   output SelectionDAG and generally the output CodeGen (with some
   exceptions).

   Additional Minor Changes:

       1. Finishes removing unused AliasLoad code
       2. Unifies the the chain aggregation in the merged stores across
       code paths
       3. Re-add the Store node to the worklist after calling
       SimplifyDemandedBits.
       4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is
       arbitrary, but seemed sufficient to not cause regressions in
       tests.

   This finishes the change Matt Arsenault started in r246307 and
   jyknight's original patch.

   Many tests required some changes as memory operations are now
   reorderable. Some tests relying on the order were changed to use
   volatile memory operations

   Noteworthy tests:

    CodeGen/AArch64/argument-blocks.ll -
      It's not entirely clear what the test_varargs_stackalign test is
      supposed to be asserting, but the new code looks right.

    CodeGen/AArch64/arm64-memset-inline.lli -
    CodeGen/AArch64/arm64-stur.ll -
    CodeGen/ARM/memset-inline.ll -

      The backend now generates *worse* code due to store merging
      succeeding, as we do do a 16-byte constant-zero store efficiently.

    CodeGen/AArch64/merge-store.ll -
      Improved, but there still seems to be an extraneous vector insert
      from an element to itself?

    CodeGen/PowerPC/ppc64-align-long-double.ll -
      Worse code emitted in this case, due to the improved store->load
      forwarding.

    CodeGen/X86/dag-merge-fast-accesses.ll -
    CodeGen/X86/MergeConsecutiveStores.ll -
    CodeGen/X86/stores-merging.ll -
    CodeGen/Mips/load-store-left-right.ll -
      Restored correct merging of non-aligned stores

    CodeGen/AMDGPU/promote-alloca-stored-pointer-value.ll -
      Improved. Correctly merges buffer_store_dword calls

    CodeGen/AMDGPU/si-triv-disjoint-mem-access.ll -
      Improved. Sidesteps loading a stored value and
      merges two stores

    CodeGen/X86/pr18023.ll -
      This test has been removed, as it was asserting incorrect
      behavior. Non-volatile stores *CAN* be moved past volatile loads,
      and now are.

    CodeGen/X86/vector-idiv.ll -
    CodeGen/X86/vector-lzcnt-128.ll -
      It's basically impossible to tell what these tests are actually
      testing. But, looks like the code got better due to the memory
      operations being recognized as non-aliasing.

    CodeGen/X86/win32-eh.ll -
      Both loads of the securitycookie are now merged.

    CodeGen/AMDGPU/vgpr-spill-emergency-stack-slot-compute.ll -
      This test appears to work but no longer exhibits the spill behavior.

Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle

Subscribers: wdng, nhaehnle, nemanjai, arsenm, weimingz, niravd, RKSimon, aemerson, qcolombet, dsanders, resistor, tstellarAMD, t.p.northover, spatel

Differential Revision: https://reviews.llvm.org/D14834

llvm-svn: 284151
2016-10-13 19:20:16 +00:00
Reid Kleckner bdfc05ff93 Re-land "[Thumb] Save/restore high registers in Thumb1 pro/epilogues"
Reverts r283938 to reinstate r283867 with a fix.

The original change had an ArrayRef referring to a destroyed temporary
initializer list. Use plain C arrays instead.

llvm-svn: 283942
2016-10-11 21:14:03 +00:00
Reid Kleckner f4876beb2b Revert "[Thumb] Save/restore high registers in Thumb1 pro/epilogues"
This reverts r283867.

This appears to be an infinite loop:

    while (HiRegToSave != AllHighRegs.end() && CopyReg != AllCopyRegs.end()) {
      if (HiRegsToSave.count(*HiRegToSave)) {
        ...

        CopyReg = findNextOrderedReg(++CopyReg, CopyRegs, AllCopyRegs.end());
        HiRegToSave =
            findNextOrderedReg(++HiRegToSave, HiRegsToSave, AllHighRegs.end());
      }
    }

llvm-svn: 283938
2016-10-11 20:54:41 +00:00
Oliver Stannard d2083fb356 [Thumb] Save/restore high registers in Thumb1 pro/epilogues
The high registers are not allocatable in Thumb1 functions, but they
could still be used by inline assembly, so we need to save and restore
the callee-saved high registers (r8-r11) in the prologue and epilogue.

This is complicated by the fact that the Thumb1 push and pop
instructions cannot access these registers. Therefore, we have to move
them down into low registers before pushing, and move them back after
popping into low registers.

In most functions, we will have low registers that are also being
pushed/popped, which we can use as the temporary registers for
saving/restoring the high registers. However, this is not guaranteed, so
we may need to push some extra low registers to ensure that the high
registers can be saved/restored. For correctness, it would be sufficient
to use just one low register, but if we have enough low registers
available then we only need one push/pop instruction, rather than one
per high register.

We can also use the argument/return registers when they are not live,
and the link register when saving (but not restoring), reducing the
number of extra registers we need to push.

There are still a few extreme edge cases where we need two push/pop
instructions, because not enough low registers can be made live in the
prologue or epilogue.

In addition to the regression tests included here, I've also tested this
using a script to generate functions which clobber different
combinations of registers, have different numbers of argument and return
registers (including variadic arguments), allocate different fixed sized
objects on the stack, and do or don't use variable sized allocas and the
__builtin_return_address intrinsic (all of which affect the available
registers in the prologue and epilogue). I ran these functions in a test
harness which verifies that all of the callee-saved registers are
correctly preserved.

Differential Revision: https://reviews.llvm.org/D24228

llvm-svn: 283867
2016-10-11 10:12:25 +00:00
Nirav Dave e524f50882 Revert "In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled."
This reverts commit r282600 due to test failues with MCJIT

llvm-svn: 282604
2016-09-28 16:37:50 +00:00