similar to: [LLVMdev] How to infer sign information from LLVM IR?

Today I ran into another aspect of the poison problem... Basically, SimplifyCFG wants to take expr1 && expr2 and flatten it into x = expr1 y = expr2 x&y This isn't safe when expr2 might execute UB. The consequence is that no LLVM shift instruction is safe to speculatively execute, nor is any nsw/nuw/exact variant, unless the operands can be proven to be in

Hello, I'd like to offer an alternative solution to the "poison problem": remove it. What follows is rather informal. I'd happily write up a nicer document if this RFC stands up to scrutiny. The idea was born from two observations: - undef was introduced to model a load of uninitialized memory, a form of undefined behavior. - poison was introduced to model integer overflow,

Hi, Recently 10% performance regression on an important benchmark showed up after we integrated https://reviews.llvm.org/rL318299. The analysis showed that rL318299 triggered loop rotation on an multi exits loop, and the loop rotation introduced code layout issue. The performance regression is a side-effect of rL318299. I got two testcases a.ll and b.ll attached to illustrate the problem. a.ll

On Mon, Dec 18, 2017 at 5:46 PM Xinliang David Li <davidxl at google.com> wrote: > The introduction of cleanup.cond block in b.ll without loop-rotation > already makes the layout worse than a.ll. > > > Without introducing cleanup.cond block, the layout out is > > entry->while.cond -> while.body->ret > > All the arrows are hot fall through edges which is

Hi, InstCombine canonicalizes index operands (unless they are into struct types) to pointer size. The comment says: "If we are using a wider index than needed for this platform, shrink it to what we need. If narrower, sign-extend it to what we need. This explicit cast can make subsequent optimizations more obvious.". For our architecture, the canonicalization is a bit

I am investigating changing BoundsChecking to use address-based rather than size- & offset-based tests. To explain, here is a short code sample cribbed from one of the tests: %mem = tail call i8* @calloc(i64 1, i64 %elements) %memobj = bitcast i8* %mem to i64* %ptr = getelementptr inbounds i64* %memobj, i64 %index %4 = load i64* %ptr, align 8 Currently, the IR for bounds checking

Hello all, I just noticed that LLVM has trouble recognizing nested simple select operations. My testcase is as follows: -------------------------- snip ------------------------------ #include <stdio.h> int main(int argc, char **argv) { int a; scanf("%d", &a); a = a > 255 ? 255 : a < 0 ? 0 : a; printf("A: %d\n", a); return 0; }

Hi, We recently found a testcase showing that simplifications in instcombine sometimes change the instruction without reducing the instruction cost, but causing problems in TwoAddressInstruction pass. And it looks like the problem is generic and other simplification may have the same issue. I want to get some ideas about what is the best way to fix such kind of problem. The testcase:

I'd missed the fact that j wasn't just being decremented. This isn't as easy as I said. Philip On 03/03/2016 02:36 PM, Philip Reames via llvm-dev wrote: > SCEV should be able to easily prove that j is positive here. I'm not > sure where the right place to use that information would be in this > case. Sanjoy, can you comment? > > Philip > > On 03/03/2016

The LRM (http://llvm.org/docs/LangRef.html#llvm-lifetime-start-intrinsic) essentially states that: - ptr is dead before a call to "lifetime.start size, ptr" - ptr is dead after a call to "lifetime.end size, ptr" This is all good and fine, and the expected use case is that all "lifetime.end size, ptr" markers are matched with a preceding "lifetime.start

Hi Kevin, Thanks for your interest and for your deep analysis. Unfortunately, your approach doesn't catch all bugs and is vulnerable to an attack. Consider the following case: ...................... | ----- obj --- | | end ^ ptr ^ ^ end-of-memory The scenario is as follows: - an object is allocated in the last page of the address space - obj is byte

Hi, is there any standard way to figure out the label of the basic block where a conditional branch merges? If we have a branch statement without an else part this is straightforward: br i1 %cmp, label %if.then, label %if.end, !dbg !24 We only need to check the operand names whether they contain "end". However things are getting more complex when there is an else branch and in

On Wed, Aug 2, 2017 at 3:36 PM, Matthias Braun <mbraun at apple.com> wrote: > So to write this in a more condensed form, you have: > > %v0 = ... > %v1 = and %v0, 255 > %v2 = and %v1, 31 > use %v1 > use %v2 > > and transform this to > %v0 = ... > %v1 = and %v0, 255 > %v2 = and %v0, 31 > ... > > This is a classical problem with instruction

The IR after LSR is: *** IR Dump After Loop Strength Reduction *** ; Preheader: entry: tail call void @fill_array(i32* getelementptr inbounds ([10 x i32], [10 x i32]* @buffer, i32 0, i32 0)) #2 br label %while.body ; Loop: while.body: ; preds = %while.body, %entry %lsr.iv = phi i32 [ %lsr.iv.next, %while.body ], [ 0, %entry ] %uglygep = getelementptr

Hello to everybody, I'm working on some improvements on trip count computation with ScalarEvolution analysis. Considering the following test ;----------------------------------------------------------------------------; define void @foo(i32 %a, i32 %b, i32 %s) #0 { entry: %cmp = icmp sgt i32 %s, 0 %cmp15 = icmp sgt i32 %a, %b %or.cond = and i1 %cmp, %cmp15 br i1 %or.cond, label

On Thu, Jul 6, 2017 at 7:00 AM, Davide Italiano <davide at freebsd.org> wrote: > On Wed, Jul 5, 2017 at 3:59 PM, Johan Engelen via llvm-dev > <llvm-dev at lists.llvm.org> wrote: > > Hi all, > > I just found an optimization regression regarding simple > > dataflow/constprop analysis: > > https://godbolt.org/g/Uz8P7t > > > > This code >

Just moving this branch of the thread out of the review because I don't want to derail the review thread... Kostya - why are these two cases not optimization bugs in general? (why do they only affect sanitizers?) On Mon, Nov 18, 2013 at 8:37 PM, Kostya Serebryany <kcc at google.com> wrote: > And we've been just informed by the mozilla folks about yet another case > of

Dear R mailing list, I have a dataset with genotypes from trios and I would like to infer haplotypes for each mother, father and child. The package that I could find that can do this is tdthap. But when the mother is homozygous (e.g., 2/2) the haplotype is called as not possible to infer (0); I would prefer for it to call the genotype (2). From what I understand it is doing what I would like

The instruction mnemonics for the immediate forms of bit-test instructions including bt, btr and bts, btc do not work. llvm-mc barfs with: error: ambiguous instructions require an explicit suffix This is highly user-unfriendly, since we can easily infer what the user meant by inspecting $imm and translating the instruction appropriately. Do it exactly as the Intel manual describes. Note that

On 3/30/2016 4:42 PM, Rail Shafigulin via llvm-dev wrote: > i'm getting a > > Could not infer all types in pattern! > > error in my backend. it is happening on the following instruction: > > VGETITEM: (set GPR:{i32:f32}:$rD, (extractelt:{i32:f32} > VR:{v4i32:v4f32}:$rA, GPR:i32:$rB)). > > how do i make it use appropriate types? in other words if it is f32 then