aboutsummaryrefslogtreecommitdiff
path: root/lib
AgeCommit message (Collapse)Author
2008-01-27Implement some dag combines that allow doing fneg/fabs/fcopysign in integerChris Lattner
registers if used by a bitconvert or using a bitconvert. This allows us to avoid constant pool loads and use cheaper integer instructions when the values come from or end up in integer regs anyway. For example, we now compile CodeGen/X86/fp-in-intregs.ll to: _test1: movl $2147483648, %eax xorl 4(%esp), %eax ret _test2: movl $1065353216, %eax orl 4(%esp), %eax andl $3212836864, %eax ret Instead of: _test1: movss 4(%esp), %xmm0 xorps LCPI2_0, %xmm0 movd %xmm0, %eax ret _test2: movss 4(%esp), %xmm0 andps LCPI3_0, %xmm0 movss LCPI3_1, %xmm1 andps LCPI3_2, %xmm1 orps %xmm0, %xmm1 movd %xmm1, %eax ret bitconverts can happen due to various calling conventions that require fp values to passed in integer regs in some cases, e.g. when returning a complex. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46414 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-27add a noteChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46413 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-27Use fldz and fld1 for long double constants instead of a constant pool load.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46411 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-27The CorrelatedExpressionElimination pass is known to be buggy. Remove it.Bill Wendling
This fixes PR1769. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46408 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-27For long double constants, print an approximation of their value to the .s ↵Chris Lattner
file to make it easier to read. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46407 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-27Fold fptrunc(add (fpextend x), (fpextend y)) -> add(x,y), as GCC does.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46406 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-26Add some notes.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46405 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-26Remove some code for inferring alignment info from the x86 backend Chris Lattner
now that the dag combiner does it. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46404 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-26Infer alignment of loads and increase their alignment when we can tell they are Chris Lattner
from the stack. This allows us to compile stack-align.ll to: _test: movsd LCPI1_0, %xmm0 movapd %xmm0, %xmm1 *** andpd 4(%esp), %xmm1 andpd _G, %xmm0 addsd %xmm1, %xmm0 movl 20(%esp), %eax movsd %xmm0, (%eax) ret instead of: _test: movsd LCPI1_0, %xmm0 ** movsd 4(%esp), %xmm1 ** andpd %xmm0, %xmm1 andpd _G, %xmm0 addsd %xmm1, %xmm0 movl 20(%esp), %eax movsd %xmm0, (%eax) ret git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46401 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-26If there's no instructions being emitted on X86 for a function, emit aBill Wendling
nop. Emit the nop directly for PPC. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46398 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-26If there are no machine instructions emitted for a function, then insertBill Wendling
a "nop" instruction so that we don't have the function's label associated with something that it's not supposed to be associated with. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46394 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-26Create an explicit copy for byval parameters evenDuncan Sands
when inlining a readonly function. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46393 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-26If we have a function like this:Bill Wendling
void bork() { int *address = 0; *address = 0; } It's compiled into LLVM code that looks like this: define void @bork() noreturn nounwind { entry: unreachable } This is bad on some platforms (like PPC) because it will generate the label for the function but no body. The label could end up being associated with some non-code related stuff, like a section. This places a "trap" instruction if the SimplifyCFG pass removed all code from the function leaving only one "unreachable" instruction. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46387 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-26Fix some bugs in SimplifyNodeWithTwoResults where it would call deletenode to Chris Lattner
delete a node even if it was not dead in some cases. Instead, just add it to the worklist. Also, make sure to use the CombineTo methods, as it was doing things that were unsafe: the top level combine loop could touch dangling memory. This fixes CodeGen/Generic/2008-01-25-dag-combine-mul.ll git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46384 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-26don't bother making x&-1 only to simplify it in dag combine. This commonly ↵Chris Lattner
occurs expanding i64 ops. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46383 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-25reduce indentationChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46377 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-25Do this more neatly.Duncan Sands
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46369 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-25fix long lines.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46355 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-25JITEmitter.cpp was trying to sync the icache for function stubs, butChris Lattner
was actually passing a completely incorrect size to sys_icache_invalidate. Instead of having the JITEmitter do this (which doesn't have the correct size), just make the target sync its own stubs. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46354 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-25DeadStoreElimination can treat byval parameters as if there were alloca's ↵Owen Anderson
for the purpose of removing end-of-function stores. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46351 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-25Add skeletal code to increase the alignment of loads and stores whenChris Lattner
we can infer it. This will eventually help stuff, though it doesn't do much right now because all fixed FI's have an alignment of 1. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46349 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-25move MachineFrameInfo::CreateFixedObject out of line, give MachineFrameInfoChris Lattner
a reference to TargetFrameInfo. Rearrange order of fields in StackObject to save a word. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46348 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-25include alignment and volatility information in -view-*-dags outputChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46347 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-25optimize fxor like forChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46345 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-25Add target-specific dag combines for FAND(x,0) and FOR(x,0). This allowsChris Lattner
us to compile: double test(double X) { return copysign(0.0, X); } into: _test: andpd LCPI1_0(%rip), %xmm0 ret instead of: _test: pxor %xmm1, %xmm1 andpd LCPI1_0(%rip), %xmm1 movapd %xmm0, %xmm2 andpd LCPI1_1(%rip), %xmm2 movapd %xmm1, %xmm0 orpd %xmm2, %xmm0 ret git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46344 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-25Provide correct DWARF register numbering for debug information emission on ↵Anton Korobeynikov
x86-32/Darwin. This should fix bunch of issues. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46337 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-24Don't dump the function!Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46320 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-24getUnderlyingObject can return null, handle this.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46318 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-24Teach basicaa that 'byval' arguments define a new memory location thatChris Lattner
can't be aliased to other known objects. This allows us to know that byval pointer args don't alias globals, etc. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46315 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-24Add hasByValAttr() and hasNoAliasAttr() methods to the Argument class.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46314 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-24clarify a comment, thanks Duncan.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46313 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-24Significantly simplify and improve handling of FP function results on x86-32.Chris Lattner
This case returns the value in ST(0) and then has to convert it to an SSE register. This causes significant codegen ugliness in some cases. For example in the trivial fp-stack-direct-ret.ll testcase we used to generate: _bar: subl $28, %esp call L_foo$stub fstpl 16(%esp) movsd 16(%esp), %xmm0 movsd %xmm0, 8(%esp) fldl 8(%esp) addl $28, %esp ret because we move the result of foo() into an XMM register, then have to move it back for the return of bar. Instead of hacking ever-more special cases into the call result lowering code we take a much simpler approach: on x86-32, fp return is modeled as always returning into an f80 register which is then truncated to f32 or f64 as needed. Similarly for a result, we model it as an extension to f80 + return. This exposes the truncate and extensions to the dag combiner, allowing target independent code to hack on them, eliminating them in this case. This gives us this code for the example above: _bar: subl $12, %esp call L_foo$stub addl $12, %esp ret The nasty aspect of this is that these conversions are not legal, but we want the second pass of dag combiner (post-legalize) to be able to hack on them. To handle this, we lie to legalize and say they are legal, then custom expand them on entry to the isel pass (PreprocessForFPConvert). This is gross, but less gross than the code it is replacing :) This also allows us to generate better code in several other cases. For example on fp-stack-ret-conv.ll, we now generate: _test: subl $12, %esp call L_foo$stub fstps 8(%esp) movl 16(%esp), %eax cvtss2sd 8(%esp), %xmm0 movsd %xmm0, (%eax) addl $12, %esp ret where before we produced (incidentally, the old bad code is identical to what gcc produces): _test: subl $12, %esp call L_foo$stub fstpl (%esp) cvtsd2ss (%esp), %xmm0 cvtss2sd %xmm0, %xmm0 movl 16(%esp), %eax movsd %xmm0, (%eax) addl $12, %esp ret Note that we generate slightly worse code on pr1505b.ll due to a scheduling deficiency that is unrelated to this patch. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46307 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-24Fix this buggy transformation. Two observations:Chris Lattner
1. we already know the value is dead, so don't bother replacing it with undef. 2. The very case the comment describes actually makes the load live which asserts in deletenode. If we do the replacement and the node becomes live, just treat it as new. This fixes a failure on X86/2008-01-16-InvalidDAGCombineXform.ll with some local changes in my tree. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46306 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-24The dag combiner is missing revisiting nodes that it really should, and thus ↵Chris Lattner
leaving dead stuff around. This gets fed into the isel pass and causes certain foldings from happening because nodes have extraneous uses floating around. For example, if we turned foo(bar(x)) -> baz(x), we sometimes left bar(x) around. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46305 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-24fold fp_round(fp_round(x)) -> fp_round(x).Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46304 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-24Fix potential buffer overflowAnton Korobeynikov
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46296 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-24Move some functionality for adding flags to MachineInstr's into methods on ↵Owen Anderson
MachineInstr rather than LiveVariables. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46295 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-24Forgot these.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46292 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-23Let each target decide byval alignment. For X86, it's 4-byte unless the ↵Evan Cheng
aggregare contains SSE vector(s). For x86-64, it's max of 8 or alignment of the type. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46286 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-23The last pieces needed for loading arbitraryDuncan Sands
precision integers. This won't actually work (and most of the code is dead) unless the new legalization machinery is turned on. While there, I rationalized the handling of i1, and removed some bogus (and unused) sextload patterns. For i1, this could result in microscopically better code for some architectures (not X86). It might also result in worse code if annotating with AssertZExt nodes turns out to be more harmful than helpful. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46280 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-23Honor explicit section information on Darwin.Dale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46267 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-22Fix an iterator invalidation issue.Owen Anderson
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46263 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-22SSE varargs arguments are passed in memory.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46262 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-22Simplify SelectionDAG::getNode so that a big switch stmt is not #ifndef Chris Lattner
NDEBUG. This is in response to a really nasty bug I introduced that Dale tracked down, hopefully this won't happen in the future. Many thanks Dale. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46254 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-22The final piece needed for storing arbitrary precisionDuncan Sands
integers. Handle truncstore of a legal type to an unusual number of bits. Most of this code is not reachable unless the new legalize infrastructure is turned on. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46249 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-22Enable the fix I just checked in, silly me.Nick Lewycky
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46247 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-22Multiply can be evaluated in a different type, so long as the target type hasNick Lewycky
a smaller bitwidth. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46244 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-22Trivial patch to fix two warnings, please pull into llvm 2.2Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46243 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-21Clarify a deviation from the original algorithm.Owen Anderson
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46218 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-21Improve a few comments.Owen Anderson
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46217 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.2.3-18-g5258 (git 2.32.0) at 2025-08-05 03:21:15 +0000