search for: host226227232

...like: "assert succeeded165": ; preds = %"assert succeeded146" %buf_host181 = getelementptr inbounds %struct.buffer_t, %struct.buffer_t* %error_op_pcmpeqq_272.buffer, i64 0, i32 1 %23 = bitcast i8** %buf_host181 to double** %error_op_pcmpeqq_272.host226227232 = load double*, double** %23, align 8 %24 = icmp eq %struct.buffer_t* %error_op_pcmpeqq_272.buffer, null br i1 %24, label %"assert failed183", label %"assert succeeded184", !prof !4 Here you have as check for nullptr at %24, but you already loaded %error_op_pcmpeqq_272.host...

...ssert succeeded165": ; preds = %"assert succeeded146" > %buf_host181 = getelementptr inbounds %struct.buffer_t, %struct.buffer_t* %error_op_pcmpeqq_272.buffer, i64 0, i32 1 > %23 = bitcast i8** %buf_host181 to double** > %error_op_pcmpeqq_272.host226227232 = load double*, double** %23, align 8 > %24 = icmp eq %struct.buffer_t* %error_op_pcmpeqq_272.buffer, null > br i1 %24, label %"assert failed183", label %"assert succeeded184", !prof !4 > > Here you have as check for nullptr at %24, but you already loaded %error...

...eeded165": ; preds = %"assert > succeeded146" > %buf_host181 = getelementptr inbounds %struct.buffer_t, > %struct.buffer_t* %error_op_pcmpeqq_272.buffer, i64 0, i32 1 > %23 = bitcast i8** %buf_host181 to double** > %error_op_pcmpeqq_272.host226227232 = load double*, double** %23, align 8 > %24 = icmp eq %struct.buffer_t* %error_op_pcmpeqq_272.buffer, null > br i1 %24, label %"assert failed183", label %"assert succeeded184", > !prof !4 > > Here you have as check for nullptr at %24, but you already > load...

...; preds = %"assert >> succeeded146" >> %buf_host181 = getelementptr inbounds %struct.buffer_t, >> %struct.buffer_t* %error_op_pcmpeqq_272.buffer, i64 0, i32 1 >> %23 = bitcast i8** %buf_host181 to double** >> %error_op_pcmpeqq_272.host226227232 = load double*, double** %23, align >> 8 >> %24 = icmp eq %struct.buffer_t* %error_op_pcmpeqq_272.buffer, null >> br i1 %24, label %"assert failed183", label %"assert succeeded184", >> !prof !4 >> >> Here you have as check for nullptr at %...

...; preds = %"assert succeeded146" >>>> %buf_host181 = getelementptr inbounds %struct.buffer_t, %struct.buffer_t* %error_op_pcmpeqq_272.buffer, i64 0, i32 1 >>>> %23 = bitcast i8** %buf_host181 to double** >>>> %error_op_pcmpeqq_272.host226227232 = load double*, double** %23, align 8 >>>> %24 = icmp eq %struct.buffer_t* %error_op_pcmpeqq_272.buffer, null >>>> br i1 %24, label %"assert failed183", label %"assert succeeded184", !prof !4 >>>> >>>> Here you have as check f...

...> > > %struct.buffer_t* %error_op_pcmpeqq_272.buffer, i64 0, i32 1 > > > > > > > > > > > > > > > %23 = bitcast i8** %buf_host181 to double** > > > > > > > > > > > > > > > %error_op_pcmpeqq_272.host226227232 = load double*, double** > > > > > %23, > > > > > align 8 > > > > > > > > > > > > > > > %24 = icmp eq %struct.buffer_t* %error_op_pcmpeqq_272.buffer, > > > > > null > > > > > > > &gt...

...ssert succeeded165": ; preds = %"assert succeeded146" > %buf_host181 = getelementptr inbounds %struct.buffer_t, %struct.buffer_t* %error_op_pcmpeqq_272.buffer, i64 0, i32 1 > %23 = bitcast i8** %buf_host181 to double** > %error_op_pcmpeqq_272.host226227232 = load double*, double** %23, align 8 > %24 = icmp eq %struct.buffer_t* %error_op_pcmpeqq_272.buffer, null > br i1 %24, label %"assert failed183", label %"assert succeeded184", !prof !4 > > Here you have as check for nullptr at %24, but you already loaded %error...

> On Feb 29, 2016, at 11:16 AM, Alina Sbirlea via llvm-dev <llvm-dev at lists.llvm.org> wrote: > > All, > > To get the discussion going in a focused manner, here is an initial patch with a running test. The test is from the Halide suite and is checking the correctness of several simd operations. > (Notes: the patch is large due to the number of operations being tested;

...; preds = %"assert >>> succeeded146" >>> %buf_host181 = getelementptr inbounds %struct.buffer_t, >>> %struct.buffer_t* %error_op_pcmpeqq_272.buffer, i64 0, i32 1 >>> %23 = bitcast i8** %buf_host181 to double** >>> %error_op_pcmpeqq_272.host226227232 = load double*, double** %23, >>> align 8 >>> %24 = icmp eq %struct.buffer_t* %error_op_pcmpeqq_272.buffer, null >>> br i1 %24, label %"assert failed183", label %"assert succeeded184", >>> !prof !4 >>> >>> Here you have a...

...ded165": ; preds = %"assert succeeded146" >> %buf_host181 = getelementptr inbounds %struct.buffer_t, %struct.buffer_t* %error_op_pcmpeqq_272.buffer, i64 0, i32 1 >> %23 = bitcast i8** %buf_host181 to double** >> %error_op_pcmpeqq_272.host226227232 = load double*, double** %23, align 8 >> %24 = icmp eq %struct.buffer_t* %error_op_pcmpeqq_272.buffer, null >> br i1 %24, label %"assert failed183", label %"assert succeeded184", !prof !4 >> >> Here you have as check for nullptr at %24, but you alrea...

All, To get the discussion going in a focused manner, here is an initial patch with a running test. The test is from the Halide suite and is checking the correctness of several simd operations. (Notes: the patch is large due to the number of operations being tested; I expect a lot of changes before actually landing it, this is simply to continue the discussion using a concrete example.)