Commit Graph

14791 Commits

Author SHA1 Message Date
Bill Wendling 965bd58902 Reset some of the target options which affect code generation.
This doesn't reset all of the target options within the TargetOptions
object. This is because some of those are ABI-specific and must be determined if
it's okay to change those on the fly.

llvm-svn: 176986
2013-03-13 22:26:59 +00:00
Manman Ren 11fec384de No functionality change. Use unreachable in getCUOffset.
llvm-svn: 176963
2013-03-13 18:41:27 +00:00
Manman Ren 723956fe1a No functionality change. Use unreachable in getCompileUnit.
llvm-svn: 176962
2013-03-13 18:33:41 +00:00
Jakob Stoklund Olesen c1b566db1b Check register classes also when changing them.
We have the same assertion in createVirtualRegister.

llvm-svn: 176959
2013-03-13 18:11:17 +00:00
Manman Ren 14a029d9f3 Debug Info: use DW_FORM_ref_addr instead of DW_FORM_ref4 if the referenced DIE
belongs to a different compile unit.

DW_FORM_ref_addr should be used for cross compile-unit reference.

When compiling a large application, we got a dwarfdump verification error where
abstract_origin points to nowhere.

This error can't be reproduced on any testing case in MultiSource.
We may have other cases where we use DW_FORM_ref4 unconditionally.

rdar://problem/13370501

llvm-svn: 176882
2013-03-12 18:27:15 +00:00
Richard Relph 61046a9727 Avoid generating ISD::SELECT for vector operands to SIGN_EXTEND
llvm-svn: 176881
2013-03-12 18:17:18 +00:00
David Blaikie dc69ebb1fc Remove support for versioned debug info.
Versioned debug info support has been a burden to maintain & also compromised
current debug info verification by causing test cases testing old debug info to
remain rather than being updated to the latest. It also makes it hard to add or
change the metadata schema by requiring various backwards-compatibility in the
DI* hierarchy.

So it's being removed in preparation for new changes to the schema to tidy up
old/unnecessary fields and add new fields needed for new debug info (well, new
to LLVM at least).

The more surprising part of this is the changes to DI*::Verify - this became
necessary due to the changes to AsmWriter. AsmWriter was relying on the version
test to decide which bits of metadata were actually debug info when printing
the comment annotations. Without the version information the tag numbers were
too common & it would print debug info on random metadata that happened to
start with an integer that matched a tag number. Instead this change makes the
Verify functions more precise (just adding "number of operands" checks - not
type checking those operands yet) & relies on that to decide which metadata is
debug info metadata.

llvm-svn: 176838
2013-03-11 23:39:23 +00:00
Nick Lewycky 48beb21185 Fix a crasher newly introduced in r176659/r176649, where fast-isel tries to
lower an expect intrinsic that is a constant expression.

llvm-svn: 176830
2013-03-11 21:44:37 +00:00
Jakub Staszak df17ddd56b Cleanup #includes.
llvm-svn: 176787
2013-03-10 13:11:23 +00:00
Jakob Stoklund Olesen 1001673865 Remove wrong and unnecessary assertion.
PHIs are allowed to have multiple operand pairs per predecessor, and
this code works just fine when it happens.

llvm-svn: 176734
2013-03-08 23:00:13 +00:00
Jan Wen Voung 7857a64909 Disable statistics on Release builds and move tests that depend on -stats.
Summary:
Statistics are still available in Release+Asserts (any +Asserts builds),
and stats can also be turned on with LLVM_ENABLE_STATS.

Move some of the FastISel stats that were moved under DEBUG()
back out of DEBUG(), since stats are disabled across the board now.

Many tests depend on grepping "-stats" output.  Move those into
a orig_dir/Stats/. so that they can be marked as unsupported
when building without statistics.

Differential Revision: http://llvm-reviews.chandlerc.com/D486

llvm-svn: 176733
2013-03-08 22:56:31 +00:00
Jakob Stoklund Olesen 8d1aaf21cf Rewrite the physreg part of findLastUseBefore().
To find the last use of a register unit, start from the bottom and scan
upwards until a user is found.

<rdar://problem/13353090>

llvm-svn: 176706
2013-03-08 18:08:57 +00:00
Benjamin Kramer 0f12a5c2a4 Remove default from fully covered switch.
llvm-svn: 176703
2013-03-08 17:03:19 +00:00
Tom Stellard d93ef7afaa LegalizeDAG: Respect the result of TLI.getBooleanContents() when expanding SETCC
llvm-svn: 176695
2013-03-08 15:37:02 +00:00
Tom Stellard b1588fc057 DAGCombiner: Use correct value type for checking legality of BR_CC v3
LegalizeDAG.cpp uses the value of the comparison operands when checking
the legality of BR_CC, so DAGCombiner should do the same.

v2:
  - Expand more BR_CC value types for NVPTX

v3:
  - Expand correct BR_CC value types for Hexagon, Mips, and XCore.

llvm-svn: 176694
2013-03-08 15:36:57 +00:00
Andrew Trick 48f2a72435 Add -verify-misched option.
This verifies live intervals both before and after scheduling. It's
useful for anyone hacking on live interval update.

Note that we don't yet pass verification all the time. We don't yet
handle updating nonallocatable live intervals perfectly.

llvm-svn: 176685
2013-03-08 05:40:34 +00:00
Bill Wendling 2d915e2c15 Revert r176154 in favor of a better approach.
Code generation makes some basic assumptions about the IR it's been given. In
particular, if there is only one 'invoke' in the function, then that invoke
won't be going away. However, with the advent of the `llvm.donothing' intrinsic,
those invokes may go away. If all of them go away, the landing pad no longer has
any users. This confuses the back-end, which asserts.

This happens with SjLj exceptions, because that's the model that modifies the IR
based on there being invokes, etc. in the function.

Remove any invokes of `llvm.donothing' during SjLj EH preparation. This will
give us a CFG that the back-end won't be confused about. If all of the invokes
in a function are removed, then the SjLj EH prepare pass won't insert the bogus
code the relies upon the invokes being there.
<rdar://problem/13228754&13316637>

llvm-svn: 176677
2013-03-08 02:21:08 +00:00
Jakob Stoklund Olesen bed165422f Add a getPassName() method.
llvm-svn: 176669
2013-03-07 23:55:51 +00:00
Jakob Stoklund Olesen 299cedc7ca Rename isEarlierInSameTrace to isUsefulDominator.
In very rare cases caused by irreducible control flow, the dominating
block can have the same trace head without actually being part of the
trace.

As long as such a dominator still has valid instruction depths, it is OK
to use it for computing instruction depths.

Rename the function to avoid lying, and add a check that instruction
depths are computed for the dominator.

llvm-svn: 176668
2013-03-07 23:55:49 +00:00
Chad Rosier 3a200e1faf [fast-isel] Seriously, add support for the expect intrinsic.
rdar://13370942

llvm-svn: 176659
2013-03-07 21:38:33 +00:00
Chad Rosier 9c1796f877 [fast-isel] Add support for the expect intrinsic.
rdar://13370942

llvm-svn: 176649
2013-03-07 20:42:17 +00:00
Benjamin Kramer fdf362bd69 ArrayRefize some code. No functionality change.
llvm-svn: 176648
2013-03-07 20:33:29 +00:00
Andrew Trick 0f23b763a9 pre-RA-sched debug-only fix
llvm-svn: 176638
2013-03-07 19:21:08 +00:00
Andrew Trick b2ab8a732c pre-RA-sched assertion fix. This bug was exposed by r176037.
rdar:13370002 [pre-RA-sched] assertion: released too many times

I tracked this down to an earlier hack that is no longer applicable
and interfered with normal scheduler logic. With the changes in
r176037, it was causing an instruction to be scheduled multiple times.

I have an external test case that I tried hard to reduce and
failed. I can't even reproduce with llc.

llvm-svn: 176636
2013-03-07 19:07:57 +00:00
Nadav Rotem 40cda80af8 No need to go through int64 and APInt when generating a new constant.
llvm-svn: 176615
2013-03-07 06:34:49 +00:00
Jim Grosbach 48a91abc10 SDAG: Handle scalarizing an extend of a <1 x iN> vector.
Just scalarize the element and rebuild a vector of the result type
from that.

rdar://13281568

llvm-svn: 176614
2013-03-07 05:47:54 +00:00
Manman Ren 1e4272085d Debug Info: store the files and directories for each compile unit.
We now emit a line table for each compile unit. To reduce the prologue size
of each line table, the files and directories used by each compile unit are
stored in std::map<unsigned, std::vector< > > instead of std::vector< >.

The prologue for a lto'ed image can be as big as 93K. Duplicating 93K for each
compile unit causes a huge increase of debug info. With this patch, each
prologue will only emit the files required by the compile unit.

rdar://problem/13342023

llvm-svn: 176605
2013-03-07 01:42:00 +00:00
Bill Wendling a69d0aaa71 Remove unused #includes.
llvm-svn: 176467
2013-03-05 01:00:45 +00:00
Eli Bendersky b1caf3c30e Remove duplicate line and move another closer to its actual use
llvm-svn: 176391
2013-03-01 23:32:40 +00:00
Akira Hatanaka 3d055580a9 Set properties for f128 type.
llvm-svn: 176378
2013-03-01 21:11:44 +00:00
Chad Rosier b3864609cf Generate an error message instead of asserting or segfaulting when we can't
handle indirect register inputs.
rdar://13322011

llvm-svn: 176367
2013-03-01 19:12:05 +00:00
Michael Liao 6af16fc3b7 Fix PR10475
- ISD::SHL/SRL/SRA must have either both scalar or both vector operands
  but TLI.getShiftAmountTy() so far only return scalar type. As a
  result, backend logic assuming that breaks.
- Rename the original TLI.getShiftAmountTy() to
  TLI.getScalarShiftAmountTy() and re-define TLI.getShiftAmountTy() to
  return target-specificed scalar type or the same vector type as the
  1st operand.
- Fix most TICG logic assuming TLI.getShiftAmountTy() a simple scalar
  type.

llvm-svn: 176364
2013-03-01 18:40:30 +00:00
Andrew Trick f9669beacd Scheduler diagnostics. Print the register name.
llvm-svn: 176316
2013-03-01 00:19:14 +00:00
Andrew Trick 2a8edef70b Instructions schedulers should report correct height/depth.
We avoided computing DAG height/depth during Node printing because it
shouldn't depend on an otherwise valid DAG. But this has become far
too annoying for the common case of a valid DAG where we want to see
valid values. If doing the computation on-the-fly turns out to be a
problem in practice, then I'll add a mode to the diagnostics to only
force it when we're likely to have a valid DAG, otherwise explicitly
print INVALID instead of bogus numbers. For now, just go for it all
the time.

llvm-svn: 176314
2013-03-01 00:19:09 +00:00
Eli Bendersky 33ebf836bc A small refactoring + adding comments.
SelectionDAGIsel::LowerArguments needs a function, not a basic block. So it
makes sense to pass it the function instead of extracting a basic-block from
the function and then tossing it. This is also more self-documenting (functions
have arguments, BBs don't).

In addition, added comments to a couple of Select* methods.

llvm-svn: 176305
2013-02-28 23:09:18 +00:00
Benjamin Kramer f7cfac7a14 Cost model support for lowered math builtins.
We make the cost for calling libm functions extremely high as emitting the
calls is expensive and causes spills (on x86) so performance suffers. We still
vectorize important calls like ceilf and friends on SSE4.1. and fabs.

Differential Revision: http://llvm-reviews.chandlerc.com/D466

llvm-svn: 176287
2013-02-28 19:09:33 +00:00
Eli Bendersky d0c6e7b038 Put some per-instruction statistics of fast isel under NDEBUG, together with
other per-instruction statistics.

llvm-svn: 176273
2013-02-28 18:05:12 +00:00
Eric Christopher 5958b46053 Remove unnecessary check against isGlobalVariable. We check it
a few lines above.

llvm-svn: 176224
2013-02-27 23:49:50 +00:00
Eric Christopher 3f79b8c374 Rework comment slightly and fix a few typos.
llvm-svn: 176223
2013-02-27 23:49:47 +00:00
Eric Christopher 10d35e9065 Remove unnecessary cast to void.
llvm-svn: 176222
2013-02-27 23:49:45 +00:00
Manman Ren 584e4c0eda Debug Info: for static member variables, always put AT_MIPS_linkage_name to the
definition DIE (TAG_variable), and put AT_MIPS_linkage_name to TAG_member when
DarwinGDBCompat is true.

Darwin GDB needs AT_MIPS_linkage_name at both places to work.

Follow-up patch to r176143.
rdar://problem/13291234

llvm-svn: 176220
2013-02-27 23:21:02 +00:00
Nadav Rotem c29095fb50 Silence the unused variable warning.
llvm-svn: 176218
2013-02-27 22:52:54 +00:00
Nadav Rotem 00b75dd3c4 The FastISEL should be fast. But when we record statistics we use atomic operations to increment the counters.
This patch disables the counters on non-debug builds. This reduces the runtime of SelectionDAGISel::SelectCodeCommon by ~5%.

llvm-svn: 176214
2013-02-27 21:59:43 +00:00
Michael Ilseman ba8446c80e Reverted: r176136 - Have a way for a target to opt-out of target-independent fast isel
llvm-svn: 176204
2013-02-27 19:54:00 +00:00
Manman Ren 683f59b36c SelectionDAG: If llvm.donothing has a landingpad, we should clear
CurrentCallSite to avoid an assertion failure:
assert(MMI.getCurrentCallSite() == 0 && "Overlapping call sites!");

rdar://problem/13228754

llvm-svn: 176154
2013-02-27 02:11:57 +00:00
Manman Ren 5ae44d2b75 Debug Info: for static member variables, add AT_MIPS_linkage_name to the
definition DIE, to make old GDB happy.

We have a regression for old GDB when Clang uses DW_TAG_member to declare
static members inside a class, instead of DW_TAG_variable. This patch will fix
this regression.

rdar://problem/13291234

llvm-svn: 176143
2013-02-27 00:02:32 +00:00
Michael Ilseman 846c6f0a32 Have a way for a target to opt-out of target-independent fast isel
llvm-svn: 176136
2013-02-26 23:15:23 +00:00
Manman Ren fe494749e4 Revert r176120 as it caused a failure at static-member.cpp
llvm-svn: 176129
2013-02-26 22:35:53 +00:00
Manman Ren 5222195831 Debug Info: for static member variables, move AT_MIPS_linkage_name from
TAG_member inside a class to the specification DIE.

Having AT_MIPS_linkage_name on TAG_member caused old gdb (GNU 6.3.50) to
error out. Also gcc 4.7 has AT_MIPS_linkage_name on the specification DIE.

rdar://problem/13291234

llvm-svn: 176120
2013-02-26 20:48:29 +00:00
Chad Rosier 0587597fb8 Fix wording.
llvm-svn: 176055
2013-02-25 22:20:00 +00:00