search for: omp_fork_call

...ntrinsic_a? > > — > Mehdi > > > >> 1. create i32* priv_val = alloca i32 %priv_i = ...in the region, and replace all %val with %prv_val in the region. >> 2. perform outlining. >> >> Caller code >> .... >> omp_push_num_threads(4) >> omp_fork_call( .... outline_par_region....) .... >> >> Callee code: >> Outlined_par_rgion { >> I32* priv_val = alloca 32 >> I32* priv_i = .... >> >> Ret >> } >> >> For OpenMP, we do support it at -O0, -O1, -O2 and -O3. We had to make sure it ru...

...gnostic supposed to behave in face of llvm.experimental.intrinsic_a? — Mehdi > 1. create i32* priv_val = alloca i32 %priv_i = ...in the region, and replace all %val with %prv_val in the region. > 2. perform outlining. > > Caller code > .... > omp_push_num_threads(4) > omp_fork_call( .... outline_par_region....) .... > > Callee code: > Outlined_par_rgion { > I32* priv_val = alloca 32 > I32* priv_i = .... > > Ret > } > > For OpenMP, we do support it at -O0, -O1, -O2 and -O3. We had to make sure it runs correctly w/ and w/o optimizations...

...ntrinsic_a? > > — > Mehdi > > > >> 1. create i32* priv_val = alloca i32 %priv_i = ...in the region, and replace all %val with %prv_val in the region. >> 2. perform outlining. >> >> Caller code >> .... >> omp_push_num_threads(4) >> omp_fork_call( .... outline_par_region....) .... >> >> Callee code: >> Outlined_par_rgion { >> I32* priv_val = alloca 32 >> I32* priv_i = .... >> >> Ret >> } >> >> For OpenMP, we do support it at -O0, -O1, -O2 and -O3. We had to make sure it ru...

...> >> >> >>> 1. create i32* priv_val = alloca i32 %priv_i = ...in the region, and replace all %val with %prv_val in the region. >>> 2. perform outlining. >>> >>> Caller code >>> .... >>> omp_push_num_threads(4) >>> omp_fork_call( .... outline_par_region....) .... >>> >>> Callee code: >>> Outlined_par_rgion { >>> I32* priv_val = alloca 32 >>> I32* priv_i = .... >>> >>> Ret >>> } >>> >>> For OpenMP, we do support it at -O0, -O1, -O...

...> >> >> >>> 1. create i32* priv_val = alloca i32 %priv_i = ...in the region, and replace all %val with %prv_val in the region. >>> 2. perform outlining. >>> >>> Caller code >>> .... >>> omp_push_num_threads(4) >>> omp_fork_call( .... outline_par_region....) .... >>> >>> Callee code: >>> Outlined_par_rgion { >>> I32* priv_val = alloca 32 >>> I32* priv_i = .... >>> >>> Ret >>> } >>> >>> For OpenMP, we do support it at -O0, -O1, -O...

Hi Sanjoy, Yes, that's exactly what we have been looking at recently here, but the region tags seem to make it possible to express the control flow as well, so I think we could start with reg ions+metadata, as Hal and Xinmin proposed, and then figure out what needs to be first class instructions. --Vikram Adve > On Jan 19, 2017, at 11:03 PM, Sanjoy Das <sanjoy at

...t;>>> 1. create i32* priv_val = alloca i32 %priv_i = ...in the region, and replace all %val with %prv_val in the region. >>>> 2. perform outlining. >>>> >>>> Caller code >>>> .... >>>> omp_push_num_threads(4) >>>> omp_fork_call( .... outline_par_region....) .... >>>> >>>> Callee code: >>>> Outlined_par_rgion { >>>> I32* priv_val = alloca 32 >>>> I32* priv_i = .... >>>> >>>> Ret >>>> } >>>> >>>> For O...

...c_a? > > — > Mehdi > > > > >> 1. create i32* priv_val = alloca i32 %priv_i = ...in the region, and >> replace all %val with %prv_val in the region. >> 2. perform outlining. >> >> Caller code >> .... >> omp_push_num_threads(4) >> omp_fork_call( .... outline_par_region....) .... >> >> Callee code: >> Outlined_par_rgion { >> I32* priv_val = alloca 32 >> I32* priv_i = .... >> >> Ret >> } >> >> For OpenMP, we do support it at -O0, -O1, -O2 and -O3. We had to make >> su...

...i32* priv_val = alloca i32 %priv_i = ...in the region, and replace all %val with %prv_val in the region. >>>>> 2. perform outlining. >>>>> >>>>> Caller code >>>>> .... >>>>> omp_push_num_threads(4) >>>>> omp_fork_call( .... outline_par_region....) .... >>>>> >>>>> Callee code: >>>>> Outlined_par_rgion { >>>>> I32* priv_val = alloca 32 >>>>> I32* priv_i = .... >>>>> >>>>> Ret >>>>> } >>...

...it, the question is how it an openmp agnostic supposed to behave in face of llvm.experimental.intrinsic_a? — Mehdi 1. create i32* priv_val = alloca i32 %priv_i = ...in the region, and replace all %val with %prv_val in the region. 2. perform outlining. Caller code .... omp_push_num_threads(4) omp_fork_call( .... outline_par_region....) .... Callee code: Outlined_par_rgion { I32* priv_val = alloca 32 I32* priv_i = .... Ret } For OpenMP, we do support it at -O0, -O1, -O2 and -O3. We had to make sure it runs correctly w/ and w/o optimizations and advanced analysis. So we need to preserve all source...

..._a? > > — > Mehdi > > > > >> 1. create i32* priv_val = alloca i32 %priv_i = ...in the region, and >> replace all %val with %prv_val in the region. >> 2. perform outlining. >> >> Caller code >> .... >> omp_push_num_threads(4) >> omp_fork_call( .... outline_par_region....) .... >> >> Callee code: >> Outlined_par_rgion { >> I32* priv_val = alloca 32 >> I32* priv_i = .... >> >> Ret >> } >> >> For OpenMP, we do support it at -O0, -O1, -O2 and -O3. We had to >> make s...

...e of llvm.experimental.intrinsic_a? > > — > Mehdi > > > > 1. create i32* priv_val = alloca i32 %priv_i = ...in the region, and replace all %val with %prv_val in the region. > 2. perform outlining. > > Caller code > .... > omp_push_num_threads(4) > omp_fork_call( .... outline_par_region....) .... > > Callee code: > Outlined_par_rgion { > I32* priv_val = alloca 32 > I32* priv_i = .... > > Ret > } > > For OpenMP, we do support it at -O0, -O1, -O2 and -O3. We had to make sure it runs correctly w/ and w/o optimizations and...

...it, the question is how it an openmp agnostic supposed to behave in face of llvm.experimental.intrinsic_a? — Mehdi 1. create i32* priv_val = alloca i32 %priv_i = ...in the region, and replace all %val with %prv_val in the region. 2. perform outlining. Caller code .... omp_push_num_threads(4) omp_fork_call( .... outline_par_region....) .... Callee code: Outlined_par_rgion { I32* priv_val = alloca 32 I32* priv_i = .... Ret } For OpenMP, we do support it at -O0, -O1, -O2 and -O3. We had to make sure it runs correctly w/ and w/o optimizations and advanced analysis. So we need to preserve all source...

...e of llvm.experimental.intrinsic_a? > > — > Mehdi > > > > 1. create i32* priv_val = alloca i32 %priv_i = ...in the region, and replace all %val with %prv_val in the region. > 2. perform outlining. > > Caller code > .... > omp_push_num_threads(4) > omp_fork_call( .... outline_par_region....) .... > > Callee code: > Outlined_par_rgion { > I32* priv_val = alloca 32 > I32* priv_i = .... > > Ret > } > > For OpenMP, we do support it at -O0, -O1, -O2 and -O3. We had to make sure it runs correctly w/ and w/o optimizations and...