mirror of
https://github.com/c64scene-ar/llvm-6502.git
synced 2025-02-08 21:32:39 +00:00
TargetTransformInfo rather than TargetLowering, removing one of the primary instances of the layering violation of Transforms depending directly on Target. This is a really big deal because LSR used to be a "special" pass that could only be tested fully using llc and by looking at the full output of it. It also couldn't run with any other loop passes because it had to be created by the backend. No longer is this true. LSR is now just a normal pass and we should probably lift the creation of LSR out of lib/CodeGen/Passes.cpp and into the PassManagerBuilder. =] I've not done this, or updated all of the tests to use opt and a triple, because I suspect someone more familiar with LSR would do a better job. This change should be essentially without functional impact for normal compilations, and only change behvaior of targetless compilations. The conversion required changing all of the LSR code to refer to the TTI interfaces, which fortunately are very similar to TargetLowering's interfaces. However, it also allowed us to *always* expect to have some implementation around. I've pushed that simplification through the pass, and leveraged it to simplify code somewhat. It required some test updates for one of two things: either we used to skip some checks altogether but now we get the default "no" answer for them, or we used to have no information about the target and now we do have some. I've also started the process of removing AddrMode, as the TTI interface doesn't use it any longer. In some cases this simplifies code, and in others it adds some complexity, but I think it's not a bad tradeoff even there. Subsequent patches will try to clean this up even further and use other (more appropriate) abstractions. Yet again, almost all of the formatting changes brought to you by clang-format. =] git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@171735 91177308-0d34-0410-b5e6-96231b3b80d8
//===---------------------------------------------------------------------===// Common register allocation / spilling problem: mul lr, r4, lr str lr, [sp, #+52] ldr lr, [r1, #+32] sxth r3, r3 ldr r4, [sp, #+52] mla r4, r3, lr, r4 can be: mul lr, r4, lr mov r4, lr str lr, [sp, #+52] ldr lr, [r1, #+32] sxth r3, r3 mla r4, r3, lr, r4 and then "merge" mul and mov: mul r4, r4, lr str lr, [sp, #+52] ldr lr, [r1, #+32] sxth r3, r3 mla r4, r3, lr, r4 It also increase the likelihood the store may become dead. //===---------------------------------------------------------------------===// bb27 ... ... %reg1037 = ADDri %reg1039, 1 %reg1038 = ADDrs %reg1032, %reg1039, %NOREG, 10 Successors according to CFG: 0x8b03bf0 (#5) bb76 (0x8b03bf0, LLVM BB @0x8b032d0, ID#5): Predecessors according to CFG: 0x8b0c5f0 (#3) 0x8b0a7c0 (#4) %reg1039 = PHI %reg1070, mbb<bb76.outer,0x8b0c5f0>, %reg1037, mbb<bb27,0x8b0a7c0> Note ADDri is not a two-address instruction. However, its result %reg1037 is an operand of the PHI node in bb76 and its operand %reg1039 is the result of the PHI node. We should treat it as a two-address code and make sure the ADDri is scheduled after any node that reads %reg1039. //===---------------------------------------------------------------------===// Use local info (i.e. register scavenger) to assign it a free register to allow reuse: ldr r3, [sp, #+4] add r3, r3, #3 ldr r2, [sp, #+8] add r2, r2, #2 ldr r1, [sp, #+4] <== add r1, r1, #1 ldr r0, [sp, #+4] add r0, r0, #2 //===---------------------------------------------------------------------===// LLVM aggressively lift CSE out of loop. Sometimes this can be negative side- effects: R1 = X + 4 R2 = X + 7 R3 = X + 15 loop: load [i + R1] ... load [i + R2] ... load [i + R3] Suppose there is high register pressure, R1, R2, R3, can be spilled. We need to implement proper re-materialization to handle this: R1 = X + 4 R2 = X + 7 R3 = X + 15 loop: R1 = X + 4 @ re-materialized load [i + R1] ... R2 = X + 7 @ re-materialized load [i + R2] ... R3 = X + 15 @ re-materialized load [i + R3] Furthermore, with re-association, we can enable sharing: R1 = X + 4 R2 = X + 7 R3 = X + 15 loop: T = i + X load [T + 4] ... load [T + 7] ... load [T + 15] //===---------------------------------------------------------------------===// It's not always a good idea to choose rematerialization over spilling. If all the load / store instructions would be folded then spilling is cheaper because it won't require new live intervals / registers. See 2003-05-31-LongShifts for an example. //===---------------------------------------------------------------------===// With a copying garbage collector, derived pointers must not be retained across collector safe points; the collector could move the objects and invalidate the derived pointer. This is bad enough in the first place, but safe points can crop up unpredictably. Consider: %array = load { i32, [0 x %obj] }** %array_addr %nth_el = getelementptr { i32, [0 x %obj] }* %array, i32 0, i32 %n %old = load %obj** %nth_el %z = div i64 %x, %y store %obj* %new, %obj** %nth_el If the i64 division is lowered to a libcall, then a safe point will (must) appear for the call site. If a collection occurs, %array and %nth_el no longer point into the correct object. The fix for this is to copy address calculations so that dependent pointers are never live across safe point boundaries. But the loads cannot be copied like this if there was an intervening store, so may be hard to get right. Only a concurrent mutator can trigger a collection at the libcall safe point. So single-threaded programs do not have this requirement, even with a copying collector. Still, LLVM optimizations would probably undo a front-end's careful work. //===---------------------------------------------------------------------===// The ocaml frametable structure supports liveness information. It would be good to support it. //===---------------------------------------------------------------------===// The FIXME in ComputeCommonTailLength in BranchFolding.cpp needs to be revisited. The check is there to work around a misuse of directives in inline assembly. //===---------------------------------------------------------------------===// It would be good to detect collector/target compatibility instead of silently doing the wrong thing. //===---------------------------------------------------------------------===// It would be really nice to be able to write patterns in .td files for copies, which would eliminate a bunch of explicit predicates on them (e.g. no side effects). Once this is in place, it would be even better to have tblgen synthesize the various copy insertion/inspection methods in TargetInstrInfo. //===---------------------------------------------------------------------===// Stack coloring improvements: 1. Do proper LiveStackAnalysis on all stack objects including those which are not spill slots. 2. Reorder objects to fill in gaps between objects. e.g. 4, 1, <gap>, 4, 1, 1, 1, <gap>, 4 => 4, 1, 1, 1, 1, 4, 4 //===---------------------------------------------------------------------===// The scheduler should be able to sort nearby instructions by their address. For example, in an expanded memset sequence it's not uncommon to see code like this: movl $0, 4(%rdi) movl $0, 8(%rdi) movl $0, 12(%rdi) movl $0, 0(%rdi) Each of the stores is independent, and the scheduler is currently making an arbitrary decision about the order. //===---------------------------------------------------------------------===// Another opportunitiy in this code is that the $0 could be moved to a register: movl $0, 4(%rdi) movl $0, 8(%rdi) movl $0, 12(%rdi) movl $0, 0(%rdi) This would save substantial code size, especially for longer sequences like this. It would be easy to have a rule telling isel to avoid matching MOV32mi if the immediate has more than some fixed number of uses. It's more involved to teach the register allocator how to do late folding to recover from excessive register pressure.