o Td@sddlmZddlmZmZddlmZddlmZddl m Z ddl m Z ddl mZddlmZd d Zed d Zed dZeddZe eedddddZeddZddZeddZeddZeddZd S)!)ir)cudatypes)cgutils)RequireLiteralValue) signature)overload_attribute) nvvmutils) intrinsiccCsB|j}|dkr tj}n|dvrttj|}ntdt|tjS)N)zargument can only be 1, 2, 3) literal_valuerint32UniTuple ValueErrorr)ndimvalrestyperh/home/ncw/WWW/www-new/content/articles/pi-bbp/venv/lib/python3.10/site-packages/numba/cuda/intrinsics.py_type_grid_functions rcCs,t|tjs t|t|}dd}||fS)agrid(ndim) Return the absolute position of the current thread in the entire grid of blocks. *ndim* should correspond to the number of dimensions declared when instantiating the kernel. If *ndim* is 1, a single integer is returned. If *ndim* is 2 or 3, a tuple of the given number of integers is returned. Computation of the first integer is as follows:: cuda.threadIdx.x + cuda.blockIdx.x * cuda.blockDim.x and is similar for the other two indices, but using the ``y`` and ``z`` attributes. cSsJ|j}|tjkrtj|ddSt|tjr#tj||jd}t ||SdS)Nr )dim) return_typerrr get_global_id isinstancercountr pack_array)contextbuildersigargsridsrrrcodegen1s   zgrid..codegenrrIntegerLiteralrr typingctxrr r#rrrgrids r(cs8t|tjs t|t|}ddfdd}||fS)agridsize(ndim) Return the absolute size (or shape) in threads of the entire grid of blocks. *ndim* should correspond to the number of dimensions declared when instantiating the kernel. If *ndim* is 1, a single integer is returned. If *ndim* is 2 or 3, a tuple of the given number of integers is returned. Computation of the first integer is as follows:: cuda.blockDim.x * cuda.gridDim.x and is similar for the other two indices, but using the ``y`` and ``z`` attributes. cSs0t|d|}t|d|}|||S)Nzntid.znctaid.)r call_sregmul)rrntidnctaidrrr_nthreads_for_dimRs z#gridsize.._nthreads_for_dimcs||j}|d}|tjkr|St|tjr:|d}|jdkr't|||fS|jdkr<|d}t||||fSdSdS)Nxyr r z)rrrrrrrr)rrr r!rnxnynzr-rrr#Ws       zgridsize..codegenr$r&rr4rgridsize<s  r5cCttj}dd}||fS)NcSs t|dS)Nwarpsize)r r))rrr r!rrrr#ms z_warpsize..codegen)rrrr'r r#rrr _warpsizeis r9r7r)targetcCs dd}|S)z_ The size of a warp. All architectures implemented to date have a warp size of 32. cSstS)N)r9)modrrrgetyszcuda_warpsize..getr)r;r<rrr cuda_warpsizessr=cCr6)a Synchronize all threads in the same thread block. This function implements the same pattern as barriers in traditional multi-threaded programming: this function waits until all threads in the block call it, at which point it returns control to all its callers. cSs<d}|j}ttd}t|||}||d|S)Nzllvm.nvvm.barrier0r)moduler FunctionTypeVoidTyperget_or_insert_functioncallget_dummy_value)rrr r!fnamelmodfntysyncrrrr#s  zsyncthreads..codegen)rrnoner8rrr syncthreadss rIcs2t|tjsdSttjtj}fdd}||fS)Ncs6ttdtdf}t|j|}|||S)N )rr?IntTyperrAr>rB)rrr r!rFrGrDrrr#s z'_syncthreads_predicate..codegen)rrIntegerri4)r' predicaterDr r#rrLr_syncthreads_predicates  rPcCd}t|||S)z syncthreads_count(predicate) An extension to numba.cuda.syncthreads where the return value is a count of the threads where predicate is true. zllvm.nvvm.barrier0.popcrPr'rOrDrrrsyncthreads_count rTcCrQ)z syncthreads_and(predicate) An extension to numba.cuda.syncthreads where 1 is returned if predicate is true for all threads or 0 otherwise. zllvm.nvvm.barrier0.andrRrSrrrsyncthreads_andrUrVcCrQ)z syncthreads_or(predicate) An extension to numba.cuda.syncthreads where 1 is returned if predicate is true for any thread or 0 otherwise. zllvm.nvvm.barrier0.orrRrSrrrsyncthreads_orrUrWN)llvmliternumbarr numba.corernumba.core.errorsrnumba.core.typingrnumba.core.extendingr numba.cudar numba.cuda.extendingr rr(r5r9Moduler=rIrPrTrVrWrrrrs4         ,