Commit Graph

25951 Commits

Author SHA1 Message Date
Eli Friedman e16c0ff1d3 Moving potential optimizations out of PR2330 into lib/Target/README.txt.
Hopefully this isn't too much stuff to dump into this file.

llvm-svn: 60285
2008-11-30 07:36:04 +00:00
Eli Friedman c8228d263b Followup to r60283: optimize arbitrary width signed divisions as well
as unsigned divisions.  Same caveats as before.

llvm-svn: 60284
2008-11-30 06:35:39 +00:00
Eli Friedman 1b7fc154a5 Fix for PR2164: allow transforming arbitrary-width unsigned divides into
multiplies.

Some more cleverness would be nice, though. It would be nice if we 
could do this transformation on illegal types.  Also, we would 
prefer a narrower constant when possible so that we can use a narrower
multiply, which can be cheaper.

llvm-svn: 60283
2008-11-30 06:02:26 +00:00
Bill Wendling 7abf352f44 Don't make TwoToExp signed by default.
llvm-svn: 60279
2008-11-30 05:29:33 +00:00
Bill Wendling af200e9237 From Hacker's Delight:
"For signed integers, the determination of overflow of x*y is not so simple. If
x and y have the same sign, then overflow occurs iff xy > 2**31 - 1. If they
have opposite signs, then overflow occurs iff xy < -2**31."

In this case, x == -1.

llvm-svn: 60278
2008-11-30 05:01:05 +00:00
Eli Friedman bd0f57821a APIntify a test which is potentially unsafe otherwise, and fix the
nearby FIXME.

I'm not sure what the right way to fix the Cell test was; if the 
approach I used isn't okay, please let me know.

llvm-svn: 60277
2008-11-30 04:59:26 +00:00
Bill Wendling 70635adea3 Instcombine was illegally transforming -X/C into X/-C when either X or C
overflowed on negation. This commit checks to make sure that neithe C nor X
overflows. This requires that the RHS of X (a subtract instruction) be a
constant integer.

llvm-svn: 60275
2008-11-30 03:42:12 +00:00
Chris Lattner 441042796d Two changes: Make getDependency remove QueryInst for a dirty record's
ReverseLocalDeps when we update it.  This fixes a regression test
failure from my last commit.

Second, for each non-local cached information structure, keep a bit that
indicates whether it is dirty or not.  This saves us a scan over the whole
thing in the common case when it isn't dirty.

llvm-svn: 60274
2008-11-30 02:52:26 +00:00
Chris Lattner fc678e2af5 introduce a typedef, no functionality change.
llvm-svn: 60272
2008-11-30 02:30:50 +00:00
Chris Lattner 1b810bd5e6 Change NonLocalDeps to be a densemap of pointers to densemap
instead of containing them by value.  This increases the density
(!) of NonLocalDeps as well as making the reallocation case 
faster.  This speeds up gvn on 403.gcc by 2% and makes room for
future improvements.

I'm not super thrilled with having to explicitly manage the new/delete
of the map, but it is necesary for the next change.

llvm-svn: 60271
2008-11-30 02:28:25 +00:00
Chris Lattner ff862c4e88 calls never depend on allocations.
llvm-svn: 60268
2008-11-30 01:44:00 +00:00
Chris Lattner 3ff6d01586 Fix a fixme by making memdep's handling of allocations more logical.
If we see that a load depends on the allocation of its memory with no
intervening stores, we now return a 'None' depedency instead of "Normal".
This tweaks GVN to do its optimization with the new result.

llvm-svn: 60267
2008-11-30 01:39:32 +00:00
Chris Lattner 60444f8aa5 implement a fixme by introducing a new getDependencyFromInternal
method that returns its result as a DepResultTy instead of as a
MemDepResult.  This reduces conversion back and forth.

llvm-svn: 60266
2008-11-30 01:26:32 +00:00
Chris Lattner 2059753e66 Move the getNonLocalDependency method to a more logical place in
the file, no functionality change.

llvm-svn: 60265
2008-11-30 01:18:27 +00:00
Chris Lattner 3d5d5f2c6d REmove an old fixme, resolve another fixme by adding liberal
comments about what this class does.

llvm-svn: 60264
2008-11-30 01:17:08 +00:00
Chris Lattner ada1f87988 remove a bit of incorrect code that tried to be tricky about speeding up
dependencies.  The basic situation was this: consider if we had:

  store1
  ...
  store2
  ...
  store3

Where memdep thinks that store3 depends on store2 and store2 depends 
on store1.  The problem happens when we delete store2: The code in 
question was updating dep info for store3 to be store1.  This is a
spiffy optimization, but is not safe at all, because aliasing isn't
transitive.  This bug isn't exposed today with DSE because DSE will only
zap store2 if it is identifical to store 3, and in this case, it is 
safe to update it to depend on store1.  However, memcpyopt is not so
fortunate, which is presumably why the "dropInstruction" code used to
exist.

Since this doesn't actually provide a speedup in practice, just rip the
code out.

llvm-svn: 60263
2008-11-30 01:09:30 +00:00
Chris Lattner 63bd586d35 Eliminate the dropInstruction method, which is not needed any more.
Fix a subtle iterator invalidation bug I introduced in the last commit.

llvm-svn: 60258
2008-11-29 23:30:39 +00:00
Chris Lattner e7d7e13bf7 implement some fixme's: when deleting an instruction with
an entry in the nonlocal deps map, don't reset entries
referencing that instruction to [dirty, null], instead, set
them to [dirty,next] where next is the instruction after the
deleted one.  Use this information in the non-local deps
code to avoid rescanning entire blocks.

This speeds up GVN slightly by avoiding pointless work.  On
403.gcc this makes GVN 1.5% faster. 

llvm-svn: 60256
2008-11-29 22:02:15 +00:00
Chris Lattner 1c6b62eb4d Change MemDep::getNonLocalDependency to return its results as
a smallvector instead of a DenseMap.  This speeds up GVN by 5%
on 403.gcc.

llvm-svn: 60255
2008-11-29 21:33:22 +00:00
Chris Lattner b8ec75bc35 move MemoryDependenceAnalysis::verifyRemoved to the end of the file,
no functionality/code change.

llvm-svn: 60254
2008-11-29 21:25:10 +00:00
Chris Lattner f280b0c729 reimplement getNonLocalDependency with a simpler worklist
formulation that is faster and doesn't require nonLazyHelper.
Much less code.

llvm-svn: 60253
2008-11-29 21:22:42 +00:00
Chris Lattner 8c5ff516c6 Fix a thinko that manifested as a crash on clamav last night.
llvm-svn: 60251
2008-11-29 20:29:04 +00:00
Chris Lattner 9f1988ab6c rename some maps.
llvm-svn: 60242
2008-11-29 09:20:15 +00:00
Chris Lattner 5cd1cfad11 rename some variables.
llvm-svn: 60241
2008-11-29 09:15:21 +00:00
Chris Lattner 80c081828f eliminate a bunch of code in favor of using AliasAnalysis::getModRefInfo.
Put a some code back to handle buggy behavior that GVN expects: it wants
loads to depend on each other, and accesses to depend on their allocations.

llvm-svn: 60240
2008-11-29 09:09:48 +00:00
Chris Lattner 81f19e9aa4 simplify some code and rename some variables. Reduce nesting.
Use getTypeStoreSize instead of ABITypeSize for in-memory size
in a couple places.

llvm-svn: 60238
2008-11-29 08:51:16 +00:00
Chris Lattner 51ba8d0630 Split getDependency into getDependency and getDependencyFrom, the
former does caching, the later doesn't.  This dramatically simplifies
the logic in getDependency and getDependencyFrom.

llvm-svn: 60234
2008-11-29 03:47:00 +00:00
Bill Wendling 469e3aa696 Temporarily revert r60195. It's causing an optimized bootstrap of llvm-gcc to fail.
llvm-svn: 60233
2008-11-29 03:43:04 +00:00
Chris Lattner e4d32791ef Now that DepType is private, we can start cleaning up some of its uses:
Document the Dirty value more precisely, use it for the uninitialized
DepResultTy value.  Change reverse mappings to be from an instruction*
instead of DepResultTy, and stop tracking other forms.  This makes it more
clear that we only care about the instruction cases.

Eliminate a DepResultTy,bool pair by using Dirty in the local case as well,
shrinking the map and simplifying the code.

This speeds up GVN by ~3% on 403.gcc.

llvm-svn: 60232
2008-11-29 03:22:12 +00:00
Chris Lattner 7f9c8a0f05 Introduce and use a new MemDepResult class to hold the results of a memdep
query.  This makes it crystal clear what cases can escape from MemDep that
the clients have to handle.  This also gives the clients a nice simplified
interface to it that is easy to poke at.

This patch also makes DepResultTy and MemoryDependenceAnalysis::DepType
private, yay.

llvm-svn: 60231
2008-11-29 02:29:27 +00:00
Chris Lattner de04e1173a Reimplement the internal abstraction used by MemDep in terms
of a pointer/int pair instead of a manually bitmangled pointer.
This forces clients to think a little more about checking the 
appropriate pieces and will be useful for internal 
implementation improvements later.

I'm not particularly happy with this.  After going through this
I don't think that the clients of memdep should be exposed to
the internal type at all.  I'll fix this in a subsequent commit.

This has no functionality change.

llvm-svn: 60230
2008-11-29 01:43:36 +00:00
Chris Lattner d3d9111ede Fix PR3141 by ensuring that MemoryDependenceAnalysis::removeInstruction
properly updates the reverse dependency map when it installs updated 
dependencies for instructions that depend on the removed instruction.

llvm-svn: 60222
2008-11-28 22:51:08 +00:00
Chris Lattner f3f6a801cc don't revisit instructions off the beginning of the block.
llvm-svn: 60221
2008-11-28 22:50:08 +00:00
Chris Lattner 73c254593e more cleanups for MemoryDependenceAnalysis::removeInstruction,
no functionality change.

llvm-svn: 60219
2008-11-28 22:28:27 +00:00
Chris Lattner a25d3952c6 random cleanups, no functionality change.
llvm-svn: 60218
2008-11-28 22:04:47 +00:00
Chris Lattner 554d1221aa Run verifyRemoved from removeInstruction when -debug is specified.
This shows the root problem behind PR3141.

llvm-svn: 60216
2008-11-28 21:45:17 +00:00
Chris Lattner e5fd5c29de rename "ping" to "verifyRemoved". I don't know why 'ping' what chosen,
but it doesn't make any sense at all.

Also make the method const, private, and fit in 80 cols while we're at it.

llvm-svn: 60215
2008-11-28 21:42:09 +00:00
Chris Lattner f2a8ba4cf0 simplify some code, remove escaped newline.
llvm-svn: 60213
2008-11-28 21:29:52 +00:00
Chris Lattner dca2cd3562 remove mysterious escaped newlines.
llvm-svn: 60211
2008-11-28 21:16:44 +00:00
Chris Lattner 8a172daa55 don't call MergeBasicBlockIntoOnlyPred on a block whose only
predecessor is itself.  This doesn't make sense, and this is
a dead infinite loop anyway.

llvm-svn: 60210
2008-11-28 19:54:49 +00:00
Duncan Sands 71ecd67b5d Add include files needed when building with
gcc 4.4 (due to use of sprintf).

llvm-svn: 60209
2008-11-28 10:20:03 +00:00
Duncan Sands 595a4423dc Fix build with gcc-4.4: it doesn't like PICStyle
being both a namespace and a variable name.

llvm-svn: 60208
2008-11-28 09:29:37 +00:00
Chris Lattner e9f6c355bf rewrite RecursivelyDeleteTriviallyDeadInstructions to use a more efficient
formulation that doesn't require set lookups or scanning a set.

llvm-svn: 60203
2008-11-28 01:20:46 +00:00
Chris Lattner d4b5ba615e remove some weirdness that came from the LSR code that has
nothing to do with dead instruction elimination.  No tests in
dejagnu depend on this, so I don't know what it was needed for.

llvm-svn: 60202
2008-11-28 00:58:15 +00:00
Chris Lattner 1adb6759ef rewrite a big chunk of how DSE does recursive dead operand
elimination to use more modern infrastructure.  Also do a bunch
of small cleanups.

llvm-svn: 60201
2008-11-28 00:27:14 +00:00
Chris Lattner 8e84c129ce delete ErasePossiblyDeadInstructionTree, replacing uses of it with
RecursivelyDeleteTriviallyDeadInstructions.

llvm-svn: 60196
2008-11-27 23:25:44 +00:00
Chris Lattner c077a2a535 Simplify LoopStrengthReduce::DeleteTriviallyDeadInstructions by
making it use RecursivelyDeleteTriviallyDeadInstructions to do
the heavy lifting.

llvm-svn: 60195
2008-11-27 23:23:35 +00:00
Chris Lattner a1bbdff933 enhance RecursivelyDeleteTriviallyDeadInstructions to make
PHIs dead if they are single-value.

llvm-svn: 60194
2008-11-27 23:18:11 +00:00
Chris Lattner 1cb4f72706 Enhance RecursivelyDeleteTriviallyDeadInstructions to optionally
return a list of deleted instructions.

llvm-svn: 60193
2008-11-27 23:14:34 +00:00
Chris Lattner 96e2dbe008 use continue to reduce indentation
llvm-svn: 60192
2008-11-27 23:00:20 +00:00