search for: call9

Displaying 12 results from an estimated 12 matches for "call9".

Did you mean: call
2012 Feb 27
2
[LLVMdev] Alias in LLVM 3.0
...LVM 3.0... extern __attribute__((overloadable, weak, alias("__SH1I422"))) float2 shuffle(float2, uint2); In LLVM 2.9 and LLVM 3.0, our front-end generates: @__shuffle_2f32_2u32 = alias weak <2 x i32> (<2 x i32>, <2 x i32>)* @4 And the calls, before linking, look like: %call9 = call <2 x float> @__shuffle_2f32_2u32(<2 x float> %tmp7, <2 x i32> %tmp8) nounwind After linking with LLVM 3.0, the call looks like: %call9 = call <2 x float> bitcast (<2 x i32> (<2 x i32>, <2 x i32>)* @__shuffle_2f32_2u32 to <2 x float> (<2 x fl...
2012 Feb 28
0
[LLVMdev] Alias in LLVM 3.0
Hi Richard, > In LLVM 2.9 and LLVM 3.0, our front-end generates: > > @__shuffle_2f32_2u32 = alias weak <2 x i32> (<2 x i32>, <2 x i32>)* @4 > > And the calls, before linking, look like: > > %call9 = call <2 x float> @__shuffle_2f32_2u32(<2 x float> %tmp7, <2 x i32> > %tmp8) nounwind I don't see how this is possible - it should be rejected by the verifier. Given the above definition of @__shuffle_2f32_2u32, its type is <2 x i32> (<2 x i32>, <2 x i32&gt...
2011 Oct 06
2
[LLVMdev] A potential bug
...nation.cpp. This pass eliminates stores backwards aggressively in an end BB. It does not check dependencies on stores in an end BB though. For example, in this code snippet: ... 1. %sum.safe_r47.pre-phi = phi i64* [ %sum.safe_r47.pre, %entry.for.end_crit_edge ], [ %sum.safe_r42, %for.body ] 2. %call9 = call i32 @gettimeofday(%struct.timeval* %end, %struct.timeval* null) nounwind 3. %0 = bitcast %struct.timeval* %start to i64* // eliminated by HandleEndBlock in DeadStoreElimination.cpp 4. %1 = bitcast %struct.timeval* %agg.tmp to i64* // eliminated ... 5. %tmp49 = load i64* %0,...
2011 Oct 06
2
[LLVMdev] A potential bug
...ards aggressively in an end BB. It does not check dependencies >> on stores in an end BB though. For example, in this code snippet: >>   ... >> 1.  %sum.safe_r47.pre-phi = phi i64* [ %sum.safe_r47.pre, >> %entry.for.end_crit_edge ], [ %sum.safe_r42, %for.body ] >> 2.  %call9 = call i32 @gettimeofday(%struct.timeval* %end, %struct.timeval* >> null) nounwind >> 3.  %0 = bitcast %struct.timeval* %start to i64*           // eliminated by >> HandleEndBlock in DeadStoreElimination.cpp >> 4.  %1 = bitcast %struct.timeval* %agg.tmp to i64*     // elimin...
2011 Oct 06
0
[LLVMdev] A potential bug
...es > stores backwards aggressively in an end BB. It does not check dependencies > on stores in an end BB though. For example, in this code snippet: >   ... > 1.  %sum.safe_r47.pre-phi = phi i64* [ %sum.safe_r47.pre, > %entry.for.end_crit_edge ], [ %sum.safe_r42, %for.body ] > 2.  %call9 = call i32 @gettimeofday(%struct.timeval* %end, %struct.timeval* > null) nounwind > 3.  %0 = bitcast %struct.timeval* %start to i64*           // eliminated by > HandleEndBlock in DeadStoreElimination.cpp > 4.  %1 = bitcast %struct.timeval* %agg.tmp to i64*     // eliminated ... > 5....
2011 Oct 06
0
[LLVMdev] A potential bug
.... It does not check > dependencies > >> on stores in an end BB though. For example, in this code snippet: > >> ... > >> 1. %sum.safe_r47.pre-phi = phi i64* [ %sum.safe_r47.pre, > >> %entry.for.end_crit_edge ], [ %sum.safe_r42, %for.body ] > >> 2. %call9 = call i32 @gettimeofday(%struct.timeval* %end, > %struct.timeval* > >> null) nounwind > >> 3. %0 = bitcast %struct.timeval* %start to i64* // eliminated > by > >> HandleEndBlock in DeadStoreElimination.cpp > >> 4. %1 = bitcast %struct.timeval*...
2011 Oct 06
1
[LLVMdev] A potential bug
...> dependencies >> >> on stores in an end BB though. For example, in this code snippet: >> >>   ... >> >> 1.  %sum.safe_r47.pre-phi = phi i64* [ %sum.safe_r47.pre, >> >> %entry.for.end_crit_edge ], [ %sum.safe_r42, %for.body ] >> >> 2.  %call9 = call i32 @gettimeofday(%struct.timeval* %end, >> >> %struct.timeval* >> >> null) nounwind >> >> 3.  %0 = bitcast %struct.timeval* %start to i64*           // >> >> eliminated by >> >> HandleEndBlock in DeadStoreElimination.cpp >> &...
2013 Aug 16
2
[LLVMdev] [Polly] Analysis of extra compile-time overhead for simple nested loops
...der ], [ %inc, %for.body4 >> ] >> %inc7 = add nsw i32 %a.018, 1 >> %cmp1 = icmp slt i32 %inc7, %cond22 >> br i1 %cmp1, label %for.cond2.preheader, label %for.end8 >> for.end8: >> %x.0.lcssa = phi i32 [ 0, %cond.end ], [ %x.1.lcssa, %for.inc6 ] >> %call9 = tail call i32 (i8*, ...)* @printf(i8* getelementptr inbounds ([4 >> x i8]* @.str, i64 0, i64 0), i32 %x.0.lcssa) #3 >> ret i32 0 >> } >> declare i32 @atoi(...) >> declare i32 @printf(i8* nocapture readonly, ...) >> >> Such code is very simple and there...
2013 Aug 15
0
[LLVMdev] [Polly] Analysis of extra compile-time overhead for simple nested loops
...[ %x.019, %for.cond2.preheader ], [ %inc, %for.body4 > ] > %inc7 = add nsw i32 %a.018, 1 > %cmp1 = icmp slt i32 %inc7, %cond22 > br i1 %cmp1, label %for.cond2.preheader, label %for.end8 > for.end8: > %x.0.lcssa = phi i32 [ 0, %cond.end ], [ %x.1.lcssa, %for.inc6 ] > %call9 = tail call i32 (i8*, ...)* @printf(i8* getelementptr inbounds ([4 > x i8]* @.str, i64 0, i64 0), i32 %x.0.lcssa) #3 > ret i32 0 > } > declare i32 @atoi(...) > declare i32 @printf(i8* nocapture readonly, ...) > > Such code is very simple and there is no memory instruction at...
2013 Aug 16
0
[LLVMdev] [Polly] Analysis of extra compile-time overhead for simple nested loops
...t;>> ] >>> %inc7 = add nsw i32 %a.018, 1 >>> %cmp1 = icmp slt i32 %inc7, %cond22 >>> br i1 %cmp1, label %for.cond2.preheader, label %for.end8 >>> for.end8: >>> %x.0.lcssa = phi i32 [ 0, %cond.end ], [ %x.1.lcssa, %for.inc6 ] >>> %call9 = tail call i32 (i8*, ...)* @printf(i8* getelementptr inbounds >>> ([4 >>> x i8]* @.str, i64 0, i64 0), i32 %x.0.lcssa) #3 >>> ret i32 0 >>> } >>> declare i32 @atoi(...) >>> declare i32 @printf(i8* nocapture readonly, ...) >>> >&gt...
2013 Aug 15
4
[LLVMdev] [Polly] Analysis of extra compile-time overhead for simple nested loops
....019, %for.cond2.preheader ], [ %inc, %for.body4 ] %inc7 = add nsw i32 %a.018, 1 %cmp1 = icmp slt i32 %inc7, %cond22 br i1 %cmp1, label %for.cond2.preheader, label %for.end8 for.end8: %x.0.lcssa = phi i32 [ 0, %cond.end ], [ %x.1.lcssa, %for.inc6 ] %call9 = tail call i32 (i8*, ...)* @printf(i8* getelementptr inbounds ([4 x i8]* @.str, i64 0, i64 0), i32 %x.0.lcssa) #3 ret i32 0 } declare i32 @atoi(...) declare i32 @printf(i8* nocapture readonly, ...) Such code is very simple and there is no memory instruction at all, so the polly-dependence pass...
2013 Aug 16
0
[LLVMdev] [Polly] Analysis of extra compile-time overhead for simple nested loops
...[ %x.019, %for.cond2.preheader ], [ %inc, %for.body4 ] > %inc7 = add nsw i32 %a.018, 1 > %cmp1 = icmp slt i32 %inc7, %cond22 > br i1 %cmp1, label %for.cond2.preheader, label %for.end8 > for.end8: > %x.0.lcssa = phi i32 [ 0, %cond.end ], [ %x.1.lcssa, %for.inc6 ] > %call9 = tail call i32 (i8*, ...)* @printf(i8* getelementptr inbounds ([4 x i8]* @.str, i64 0, i64 0), i32 %x.0.lcssa) #3 > ret i32 0 > } > declare i32 @atoi(...) > declare i32 @printf(i8* nocapture readonly, ...) This test case misses the target data. If you would attach the original fil...