完善时间剪枝功能

benchmark_contract_sliced.py

@@ -3,19 +3,19 @@ import time, pickle, os
 import numpy as np
 from mpi4py import MPI
 
-NQUBITS, NLAYERS, NCORES = 25, 10, 96
+NQUBITS, NLAYERS, NCORES = 25, 10, 48
 
 comm = MPI.COMM_WORLD
 rank, size = comm.Get_rank(), comm.Get_size()
 
-os.environ['OMP_NUM_THREADS'] = str(max(1, NCORES // size))
-os.environ['MKL_NUM_THREADS'] = str(max(1, NCORES // size))
+os.environ['OMP_NUM_THREADS'] = str(NCORES)
+os.environ['MKL_NUM_THREADS'] = str(NCORES)
 
 import torch
 import qibo, quimb as qu
 from qibotn.observables import build_random_circuit
 
-torch.set_num_threads(max(1, NCORES // size))
+torch.set_num_threads(NCORES)
 
 circuit = build_random_circuit(NQUBITS, NLAYERS)
 qibo.set_backend("qibotn", platform="quimb")

benchmark_search.py

@@ -6,7 +6,7 @@ import qibo, quimb as qu
 
 from mpi4py import MPI
 
-NQUBITS, NLAYERS, WORKERS = 30, 10, 96
+NQUBITS, NLAYERS, WORKERS = 20, 10, 96
 
 comm = MPI.COMM_WORLD
 rank, size = comm.Get_rank(), comm.Get_size()
@@ -23,7 +23,7 @@ if rank == 0:
     print(f"Searching {NQUBITS}q {NLAYERS}l, method={method}, ranks={size}, workers/rank={WORKERS}...")
 t0 = time.time()
 tree = parallel_path_search(tn, tn.outer_inds(), method=method,
-    total_repeats=1024, max_time=300, n_workers=WORKERS)
+    total_repeats=1024, max_time=300, n_workers=WORKERS,trial_timeout=60)
 t_search = time.time() - t0
 
 if rank == 0:

hostfile

@@ -1,2 +1,2 @@
-10.20.6.74:1
-10.20.6.102:1
+10.20.6.74
+#10.20.6.102

src/qibotn/parallel.py

@@ -13,18 +13,15 @@ except ImportError:
     MPI = None
 
 
-def _serial_search(tn_bytes, output_inds, repeats, seed, max_time, slicing_opts=None):
-    """Single-process path search with cotengra."""
-    import random
-    import cotengra as ctg
+def _run_single_trial(tn_bytes, output_inds, seed, slicing_opts):
+    import random, cotengra as ctg
     random.seed(seed)
     tn = pickle.loads(tn_bytes)
     opt = ctg.HyperOptimizer(
         methods=["kahypar", "kahypar-agglom", "spinglass"],
-        max_repeats=repeats,
+        max_repeats=1,
         parallel=False,
         minimize="combo-256",
-        max_time=max_time,
         optlib="random",
         slicing_opts=slicing_opts,
         progbar=False,
@@ -33,13 +30,62 @@ def _serial_search(tn_bytes, output_inds, repeats, seed, max_time, slicing_opts=
     return tree.combo_cost(factor=256), tree
 
 
-def _processpool_search(tn, output_inds, total_repeats, n_workers, max_time, slicing_opts=None):
-    """ProcessPool-based parallel search."""
+def _kill_pool(pool):
+    for pid in list(pool._processes.keys()):
+        try:
+            os.kill(pid, signal.SIGKILL)
+        except ProcessLookupError:
+            pass
+    pool.shutdown(wait=False)
+
+
+def _serial_search(tn_bytes, output_inds, repeats, seed, max_time, slicing_opts=None, trial_timeout=None):
+    import time
+
+    if trial_timeout is None:
+        import random, cotengra as ctg
+        random.seed(seed)
+        tn = pickle.loads(tn_bytes)
+        opt = ctg.HyperOptimizer(
+            methods=["kahypar", "kahypar-agglom", "spinglass"],
+            max_repeats=repeats,
+            parallel=False,
+            minimize="combo-256",
+            max_time=max_time,
+            optlib="random",
+            slicing_opts=slicing_opts,
+            progbar=False,
+        )
+        tree = tn.contraction_tree(optimize=opt, output_inds=output_inds)
+        return tree.combo_cost(factor=256), tree
+
+    deadline = time.time() + max_time
+    best_cost, best_tree = float("inf"), None
+
+    for i in range(repeats):
+        if time.time() >= deadline:
+            break
+        timeout = min(trial_timeout, deadline - time.time())
+        pool = ProcessPoolExecutor(max_workers=1)
+        fut = pool.submit(_run_single_trial, tn_bytes, output_inds, seed * 10000 + i, slicing_opts)
+        try:
+            cost, tree = fut.result(timeout=timeout)
+            if cost < best_cost:
+                best_cost, best_tree = cost, tree
+        except Exception:
+            pass
+        finally:
+            _kill_pool(pool)
+
+    return best_cost, best_tree
+
+
+def _processpool_search(tn, output_inds, total_repeats, n_workers, max_time, slicing_opts=None, trial_timeout=None):
     tn_bytes = pickle.dumps(tn)
     repeats_per = max(1, total_repeats // n_workers)
     pool = ProcessPoolExecutor(max_workers=n_workers)
     futures = [
-        pool.submit(_serial_search, tn_bytes, output_inds, repeats_per, seed, max_time, slicing_opts)
+        pool.submit(_serial_search, tn_bytes, output_inds, repeats_per, seed, max_time, slicing_opts, trial_timeout)
         for seed in range(n_workers)
     ]
     best_cost, best_tree = float("inf"), None
@@ -56,109 +102,78 @@ def _processpool_search(tn, output_inds, total_repeats, n_workers, max_time, sli
     finally:
         for fut in futures:
             fut.cancel()
-        for pid in list(pool._processes.keys()):
-            try:
-                os.kill(pid, signal.SIGKILL)
-            except ProcessLookupError:
-                pass
-        pool.shutdown(wait=False)
+        _kill_pool(pool)
     return best_tree
 
 
-def _mpi_search(tn, output_inds, total_repeats, max_time, n_workers=None, slicing_opts=None):
-    """MPI+ProcessPool hybrid search.
-
-    Each MPI rank uses a local ProcessPool for parallel search,
-    then the best tree is gathered and broadcast.
-    """
+def _mpi_search(tn, output_inds, total_repeats, max_time, n_workers=None, slicing_opts=None, trial_timeout=None):
     comm = MPI.COMM_WORLD
-    rank = comm.Get_rank()
-    size = comm.Get_size()
+    rank, size = comm.Get_rank(), comm.Get_size()
     tn_bytes = pickle.dumps(tn)
     repeats_per = max(1, total_repeats // size)
 
     if n_workers and n_workers > 1:
-        # Hybrid: each MPI rank uses ProcessPool
         local_tree = _processpool_search(
-            tn, output_inds, repeats_per, n_workers, max_time, slicing_opts
+            tn, output_inds, repeats_per, n_workers, max_time, slicing_opts, trial_timeout
         )
+        local_cost = local_tree.combo_cost(factor=256) if local_tree else float("inf")
     else:
-        # Pure MPI: each rank runs serial
         local_cost, local_tree = _serial_search(
-            tn_bytes, output_inds, repeats_per, rank, max_time, slicing_opts
+            tn_bytes, output_inds, repeats_per, rank, max_time, slicing_opts, trial_timeout
         )
 
-    all_results = comm.gather((local_tree.combo_cost(factor=256), local_tree), root=0)
+    all_results = comm.gather((local_cost, local_tree), root=0)
     best_tree = None
     if rank == 0:
         best_cost = float("inf")
         for cost, tree in all_results:
             if tree is not None and cost < best_cost:
                 best_cost, best_tree = cost, tree
-    best_tree = comm.bcast(best_tree, root=0)
-    return best_tree
+    return comm.bcast(best_tree, root=0)
 
 
 def parallel_path_search(tn, output_inds, method='processpool', total_repeats=1024,
-                         max_time=300, n_workers=48, slicing_opts=None):
+                         max_time=300, n_workers=48, slicing_opts=None, trial_timeout=None):
     """Parallel contraction path search.
 
     Args:
-        tn: Tensor network (quimb TensorNetwork)
-        output_inds: Output indices
         method: 'processpool' | 'mpi' | 'serial'
-        total_repeats: Total optimization repeats
-        max_time: Timeout per worker (seconds)
-        n_workers: Number of workers (processpool only, or per-MPI-rank if MPI)
-        slicing_opts: dict for cotengra slicing_opts (memory control)
-
-    Returns:
-        Best contraction tree
+        total_repeats: Total optimization repeats across all workers
+        max_time: Global timeout per worker (seconds)
+        n_workers: Workers per MPI rank (or total for processpool)
+        slicing_opts: cotengra slicing options for memory control
+        trial_timeout: Per-trial timeout (seconds); kills and skips hung trials
     """
     if method == 'serial':
         tn_bytes = pickle.dumps(tn)
-        _, tree = _serial_search(tn_bytes, output_inds, total_repeats, 0, max_time, slicing_opts)
+        _, tree = _serial_search(tn_bytes, output_inds, total_repeats, 0, max_time, slicing_opts, trial_timeout)
         return tree
     elif method == 'mpi':
         if not _HAVE_MPI:
             raise ImportError("mpi4py not available")
-        return _mpi_search(tn, output_inds, total_repeats, max_time, n_workers, slicing_opts)
+        return _mpi_search(tn, output_inds, total_repeats, max_time, n_workers, slicing_opts, trial_timeout)
     elif method == 'processpool':
-        return _processpool_search(tn, output_inds, total_repeats, n_workers, max_time, slicing_opts)
+        return _processpool_search(tn, output_inds, total_repeats, n_workers, max_time, slicing_opts, trial_timeout)
     else:
         raise ValueError(f"Unknown method: {method}")
 
 
 def parallel_contract(tree, arrays, method='mpi', comm=None):
-    """Parallel sliced contraction.
-
-    Args:
-        tree: Contraction tree
-        arrays: List of tensor arrays
-        method: 'mpi' (only MPI supported for now)
-        comm: MPI communicator
-
-    Returns:
-        Contracted result (on root rank for MPI, otherwise for all)
-    """
     if method == 'mpi':
         if not _HAVE_MPI or comm is None:
             raise ValueError("MPI method requires mpi4py and comm")
         return _contract_mpi(tree, arrays, comm)
-    else:
-        raise ValueError(f"Unknown method: {method}")
+    raise ValueError(f"Unknown method: {method}")
 
 
 def _contract_mpi(tree, arrays, comm, root=0):
-    """Distribute contraction slices across MPI ranks with Reduce."""
-    rank = comm.Get_rank()
-    size = comm.Get_size()
+    rank, size = comm.Get_rank(), comm.Get_size()
     is_torch = type(arrays[0]).__module__.startswith("torch")
 
     result_np = None
     for i in range(rank, tree.multiplicity, size):
         x = tree.contract_slice(arrays, i)
-        x_np = np.asfortranarray(x.detach().cpu().numpy() if is_torch else np.asarray(x))
+        x_np = np.asarray(x.detach().cpu().numpy() if is_torch else x).reshape(-1)
         result_np = x_np if result_np is None else result_np + x_np
 
     if result_np is None:
@@ -166,8 +181,4 @@ def _contract_mpi(tree, arrays, comm, root=0):
 
     result = np.zeros_like(result_np) if rank == root else None
     comm.Reduce(result_np, result, root=root)
-
-    if rank == root and is_torch:
-        import torch
-        return torch.from_numpy(np.asarray(result))
     return result

vtune_contract/config/analysis_type.cfg

@@ -0,0 +1,268 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<bag xmlns:boolean="http://www.w3.org/2001/XMLSchema#boolean" xmlns:double="http://www.intel.com/2001/XMLSchema#double" xmlns:exsl="http://exslt.org/common" xmlns:int="http://www.w3.org/2001/XMLSchema#int" xmlns:str="http://exslt.org/strings" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
+ <internal>
+  <name>%HotspotsAtypeName</name>
+  <shortName>%HotspotsAtypeShortName</shortName>
+  <description>%HotspotsAtypeDescription</description>
+  <shortDescription>%HotspotsAtypeShortDescription</shortDescription>
+  <abbreviation>hs</abbreviation>
+  <alias>hotspots</alias>
+  <int:schemaVersion>1</int:schemaVersion>
+  <int:contentVersion>1</int:contentVersion>
+  <iconClass>fire solid</iconClass>
+  <property name="helpId">configs.analysis_type-hotspots_f1101</property>
+ </internal>
+ <prerequisites xsl:version="1.0" exsl:keep_exsl_namespace="" syntax="norules">
+  <xsl:copy-of select="document('config://analysis_type/include/require_connection.xsl?connections=group_generic,tcp')"/>
+ </prerequisites>
+ <knobs xsl:version="1.0" exsl:keep_exsl_namespace="" syntax="norules">
+  <xsl:variable name="pmuConditionsExtended" select="document('config://include/pmu_variables_extended.xsl')"/>
+  <xsl:variable name="isExtendedIsaSupport" select="$pmuConditionsExtended//variables/isExtendedIsaSupport"/>
+  <stringKnob id="initialViewpoint" displayName="%InitialViewpoint" boolean:visible="false">
+   <defaultValue>%HotspotsByCPUUsageViewpointName</defaultValue>
+  </stringKnob>
+  <stringKnob id="allowedViewpoints" boolean:visible="false">
+   <defaultValue>%HotspotsByCPUUsageViewpointName,%TasksOverviewViewpointName</defaultValue>
+  </stringKnob>
+  <xsl:variable name="knobsParams">
+   <params samplingIntervalApplyKnob="samplingMode" samplingIntervalApplyKnobValue="hw" pmuSamplingIntervalDescription="SamplingIntervalDescriptionOnHotspots"/>
+  </xsl:variable>
+  <xsl:variable name="knobsParamsName">
+   <xsl:text>config://analysis_type/include/knobs.xsl?</xsl:text>
+   <xsl:for-each select="exsl:node-set($knobsParams)//@*">
+    <xsl:value-of select="concat(name(), '=', .)"/>
+    <xsl:text>&amp;</xsl:text>
+   </xsl:for-each>
+  </xsl:variable>
+  <xsl:variable name="commonKnobs" select="document($knobsParamsName)"/>
+  <xsl:choose>
+   <xsl:when test="exsl:ctx('targetOS', '') = 'QNX' or exsl:ctx('targetOS', '') = 'MacOSX'">
+    <enumKnob id="samplingMode" displayName="%SamplingMode" cliName="sampling-mode" boolean:visible="false">
+     <description>%SamplingModeDescription</description>
+     <values>
+      <value displayName="%HardwareSampling" cliName="hw">hw</value>
+      <defaultValue>hw</defaultValue>
+     </values>
+    </enumKnob>
+    <xsl:copy-of select="$commonKnobs//knobs/doubleKnob[@id='samplingInterval']"/>
+    <booleanKnob id="enableStackCollect" displayName="%EnableStackCollection" cliName="enable-stack-collection">
+     <xsl:if test="not(exsl:is_experimental('sep-lbr'))">
+      <xsl:attribute name="boolean:visible">false</xsl:attribute>
+     </xsl:if>
+     <description>%EnableStackCollectionDescription</description>
+     <boolean:defaultValue>false</boolean:defaultValue>
+    </booleanKnob>
+   </xsl:when>
+   <xsl:when test="$isExtendedIsaSupport='true'">
+    <enumKnob id="samplingMode" cliName="sampling-mode" boolean:visible="true">
+     <description>%HardwareSamplingDescription</description>
+     <values>
+      <value displayName="%HardwareSampling" cliName="hw">hw</value>
+      <defaultValue>hw</defaultValue>
+     </values>
+    </enumKnob>
+    <xsl:copy-of select="$commonKnobs//knobs/doubleKnob[@id='samplingInterval']"/>
+    <booleanKnob id="enableStackCollect" displayName="%EnableStackCollection" cliName="enable-stack-collection">
+     <xsl:attribute name="boolean:visible">false</xsl:attribute>
+     <description>%EnableStackCollectionDescription</description>
+     <boolean:defaultValue>true</boolean:defaultValue>
+    </booleanKnob>
+   </xsl:when>
+   <xsl:otherwise>
+    <groupKnob id="groupForCustomControl">
+     <knobProperty name="knob_control_id">hotspotsGroup</knobProperty>
+     <knobs>
+      <enumKnob id="samplingMode" displayName="%SamplingMode" cliName="sampling-mode" boolean:visible="true">
+       <description>%SamplingModeDescription</description>
+       <values>
+        <value displayName="%SoftwareSampling" cliName="sw">sw</value>
+        <value displayName="%HardwareSampling" cliName="hw">hw</value>
+        <defaultValue>sw</defaultValue>
+       </values>
+      </enumKnob>
+      <xsl:copy-of select="$commonKnobs//knobs/doubleKnob[@id='samplingInterval']"/>
+      <booleanKnob id="enableStackCollect" displayName="%EnableStackCollection" cliName="enable-stack-collection">
+       <xsl:if test="not(exsl:is_experimental('sep-lbr')) and exsl:ctx('targetOS', '') = 'MacOSX'">
+        <xsl:attribute name="boolean:visible">false</xsl:attribute>
+       </xsl:if>
+       <description>%EnableStackCollectionDescription</description>
+       <boolean:defaultValue>false</boolean:defaultValue>
+      </booleanKnob>
+      <xsl:copy-of select="$commonKnobs//knobs/enumKnob[@id='stackSizeCollect']"/>
+     </knobs>
+    </groupKnob>
+   </xsl:otherwise>
+  </xsl:choose>
+  <doubleKnob id="slowGoodFrameThreshold" displayName="%SlowGoodFrameThreshold" cliName="slow-frames-threshold" boolean:visible="false">
+   <description>%SlowGoodFrameThresholdDescription</description>
+   <double:min>0.01</double:min>
+   <double:max>1024000</double:max>
+   <double:defaultValue>40</double:defaultValue>
+  </doubleKnob>
+  <doubleKnob id="goodFastFrameThreshold" displayName="%GoodFastFrameThreshold" cliName="fast-frames-threshold" boolean:visible="false">
+   <description>%GoodFastFrameThresholdDescription</description>
+   <double:min>0.01</double:min>
+   <double:max>1024000</double:max>
+   <double:defaultValue>100</double:defaultValue>
+  </doubleKnob>
+  <xsl:if test="exsl:ctx('targetOS', '') != 'Android'">
+   <booleanKnob id="enableCharacterizationInsights" displayName="%EnableCharacterizationInsights" cliName="enable-characterization-insights">
+    <boolean:defaultValue>true</boolean:defaultValue>
+    <description>%EnableCharacterizationInsightsDescription</description>
+   </booleanKnob>
+  </xsl:if>
+ </knobs>
+ <analysis xsl:version="1.0" str:keep_str_namespace="" exsl:keep_exsl_namespace="" syntax="norules">
+  <xsl:variable name="minMajorVersion" select="number(5)"/>
+  <xsl:variable name="minMinorVersion" select="number(10)"/>
+  <xsl:variable name="sepVersion" select="string(exsl:ctx('SEPDriverVersion', ''))"/>
+  <xsl:variable name="currentSepVersions" select="str:tokenize($sepVersion, '.')"/>
+  <xsl:variable name="pmuConditions" select="document('config://include/pmu_variables.xsl')"/>
+  <xsl:variable name="isHybridPMU" select="$pmuConditions//variables/isHybridPMU"/>
+  <xsl:variable name="useEventBasedCounts" select="(exsl:ctx('samplingMode', 'sw')='sw' or (number($currentSepVersions[1]) &gt; $minMajorVersion) or        (number($currentSepVersions[1]) = $minMajorVersion and number($currentSepVersions[2]) &gt; $minMinorVersion)) and ($isHybridPMU != 'true')"/>
+  <xsl:variable name="events" select="document(concat('config://analysis_type/include/common_events.xsl?', 'useEventBasedCounts=', $useEventBasedCounts, '&amp;isHotspots=true'))"/>
+  <xsl:variable name="mainEvents" select="$events//events/cpi"/>
+  <xsl:variable name="fpuEvents" select="$events//events/fpu"/>
+  <xsl:variable name="retiredEvents" select="$events//events/retired"/>
+  <xsl:variable name="retiredEventsPM" select="$events//events/retired_perf_metrics"/>
+  <xsl:variable name="pmuCommon" select="document('config://include/pmu_common.xsl')"/>
+  <xsl:variable name="perfMetricsPossible" select="$pmuCommon//variables/perfMetricsPossible"/>
+  <collector id="runss">
+   <xsl:choose>
+    <xsl:when test="exsl:ctx('samplingMode', 'sw')='sw'">
+     <xsl:if test="not(exsl:ctx('isTPSSAvailable', 0)) and not(exsl:ctx('isPtraceAvailable', 0)) and not(exsl:ctx('targetOS', '') = 'QNX')">
+      <xsl:value-of select="exsl:error('%RunssHotspotsNotSupported')"/>
+     </xsl:if>
+     <xsl:if test="exsl:ctx('isPtraceScopeLimited', 0)">
+      <xsl:value-of select="exsl:error('%RunssPtraceScopeLimited')"/>
+     </xsl:if>
+     <xsl:if test="exsl:ctx('targetOS', '') = 'MacOSX'">
+      <xsl:value-of select="exsl:error('%ThisAnalysisTypeIsNotApplicable')"/>
+     </xsl:if>
+     <collectorKnob knob="collectSamplesMode">stack</collectorKnob>
+     <collectorKnob knob="samplingInterval">10</collectorKnob>
+     <collectorKnob knob="collectUserTasksEventsCountersMode">true</collectorKnob>
+     <xsl:if test="exsl:ctx('isPytraceAvailable', 0)">
+      <xsl:choose>
+       <xsl:when test="exsl:ctx('targetOS', '') = 'Windows'">
+        <collectorKnob knob="mrteType">java,dotnet,python</collectorKnob>
+       </xsl:when>
+       <xsl:when test="exsl:ctx('targetOS', '') = 'Linux'">
+        <collectorKnob knob="mrteType">java,python</collectorKnob>
+       </xsl:when>
+      </xsl:choose>
+     </xsl:if>
+     <collectorKnob knob="collectOpenMPRegions">
+      <xsl:value-of select="exsl:ctx('analyzeOpenMPRegions', 0)"/>
+     </collectorKnob>
+     <xsl:if test="exsl:ctx('enableCharacterizationInsights', 0)">
+      <xsl:variable name="isSEPFlow" select="exsl:ctx('isSEPDriverAvailable', 0) and (not(exsl:ctx('usePerf', 0)) or exsl:ctx('targetOS', '') = 'Windows' or exsl:ctx('targetOS', '') = 'MacOSX' or exsl:ctx('targetOS', '') = 'FreeBSD')"/>
+      <xsl:variable name="isPerfFlow" select="(exsl:ctx('targetOS', '') = 'Linux' or exsl:ctx('targetOS', '') = 'Android') and exsl:ctx('LinuxPerfCredentials', 'NotAvailable')!='NotAvailable'                        and contains(exsl:ctx('LinuxPerfCapabilities', ''), 'format')"/>
+      <xsl:variable name="isPMUAvailable" select="exsl:ctx('PerfmonVersion', '4') &gt; '1' and ($isSEPFlow or $isPerfFlow)"/>
+      <xsl:variable name="isPermissionInsufficient" select="exsl:ctx('targetOS')='Windows' and not(exsl:ctx('AdministratorPrivileges', 'false'))"/>
+      <xsl:choose>
+       <xsl:when test="not($isPMUAvailable) or $isPermissionInsufficient">
+        <xsl:if test="not($isPMUAvailable)">
+         <xsl:value-of select="exsl:warning('%HWInsightsNotAvailableWarningTpss')"/>
+        </xsl:if>
+        <xsl:if test="$isPermissionInsufficient">
+         <xsl:value-of select="exsl:warning('%HWInsightsNotAvailablePermissionWarningTpss')"/>
+        </xsl:if>
+       </xsl:when>
+       <xsl:otherwise>
+        <xsl:variable name="fullEventList">
+         <xsl:value-of select="$mainEvents"/>
+         <xsl:if test="$fpuEvents and $fpuEvents != ''">
+          <xsl:text>,</xsl:text>
+          <xsl:value-of select="$fpuEvents"/>
+         </xsl:if>
+         <xsl:choose>
+          <xsl:when test="$perfMetricsPossible = 'true'">
+           <xsl:if test="$retiredEventsPM and $retiredEventsPM != ''">
+            <xsl:text>,</xsl:text>
+            <xsl:value-of select="$retiredEventsPM"/>
+           </xsl:if>
+          </xsl:when>
+          <xsl:otherwise>
+           <xsl:if test="$retiredEvents and $retiredEvents != ''">
+            <xsl:text>,</xsl:text>
+            <xsl:value-of select="$retiredEvents"/>
+           </xsl:if>
+          </xsl:otherwise>
+         </xsl:choose>
+        </xsl:variable>
+        <collectorKnob knob="pmuEventConfig">
+         <xsl:value-of select="exsl:merge_pmu_events($fullEventList)"/>
+        </collectorKnob>
+        <xsl:choose>
+         <xsl:when test="$perfMetricsPossible = 'true'">
+          <boolean:collectorKnob knob="useAggregatedCounting">true</boolean:collectorKnob>
+          <boolean:collectorKnob knob="usePerfMetrics">true</boolean:collectorKnob>
+         </xsl:when>
+         <xsl:otherwise>
+          <boolean:collectorKnob knob="useCountingMode">true</boolean:collectorKnob>
+         </xsl:otherwise>
+        </xsl:choose>
+       </xsl:otherwise>
+      </xsl:choose>
+     </xsl:if>
+     <boolean:collectorKnob knob="collectSWHotspots">true</boolean:collectorKnob>
+    </xsl:when>
+    <xsl:otherwise>
+     <xsl:attribute name="id">runsa</xsl:attribute>
+     <xsl:if test="exsl:ctx('PMU', '') = ''">
+      <xsl:value-of select="exsl:error('%UnknownPMUForAT')"/>
+     </xsl:if>
+     <collectorKnob knob="useEventBasedCounts">
+      <xsl:value-of select="$useEventBasedCounts"/>
+     </collectorKnob>
+     <collectorKnob knob="collectUserTasksEventsCountersMode">true</collectorKnob>
+     <collectorKnob knob="eventMode">all</collectorKnob>
+     <collectorKnob knob="enableStackCollection">
+      <xsl:value-of select="exsl:ctx('enableStackCollect', 0)"/>
+     </collectorKnob>
+     <xsl:copy-of select="document('config://analysis_type/include/knobs.xsl')//knobs/stackSizeKnobLogic/*"/>
+     <boolean:collectorKnob knob="enableCSwitch">false</boolean:collectorKnob>
+     <boolean:collectorKnob knob="cpuGpuUsageData">true</boolean:collectorKnob>
+     <xsl:variable name="usePerfMetrics" select="$perfMetricsPossible = 'true' and not(exsl:ctx('enableStackCollect', 0))"/>
+     <xsl:variable name="fullEventList">
+      <xsl:value-of select="$mainEvents"/>
+      <xsl:choose>
+       <xsl:when test="$usePerfMetrics">
+        <xsl:if test="$retiredEventsPM and $retiredEventsPM != ''">
+         <xsl:text>,</xsl:text>
+         <xsl:value-of select="$retiredEventsPM"/>
+        </xsl:if>
+       </xsl:when>
+       <xsl:otherwise>
+        <xsl:if test="$retiredEvents and $retiredEvents != ''">
+         <xsl:text>,</xsl:text>
+         <xsl:value-of select="$retiredEvents"/>
+        </xsl:if>
+       </xsl:otherwise>
+      </xsl:choose>
+      <xsl:if test="exsl:ctx('enableCharacterizationInsights', 0)">
+       <xsl:if test="$fpuEvents and $fpuEvents != ''">
+        <xsl:text>,</xsl:text>
+        <xsl:value-of select="$fpuEvents"/>
+       </xsl:if>
+      </xsl:if>
+     </xsl:variable>
+     <collectorKnob knob="pmuEventConfig">
+      <xsl:value-of select="exsl:merge_pmu_events($fullEventList)"/>
+     </collectorKnob>
+     <collectorKnob knob="pmuSamplingInterval">
+      <xsl:value-of select="format-number(exsl:ctx('samplingInterval', 1), '#.####')"/>
+     </collectorKnob>
+     <boolean:collectorKnob knob="isUArchUsageAvailable">true</boolean:collectorKnob>
+     <collectorKnob knob="usePerfMetrics">
+      <xsl:value-of select="$usePerfMetrics"/>
+     </collectorKnob>
+     <xsl:if test="exsl:is_experimental('ehfi')">
+      <collectorKnob knob="analyzeEHFIClasses">true</collectorKnob>
+     </xsl:if>
+    </xsl:otherwise>
+   </xsl:choose>
+  </collector>
+ </analysis>
+</bag>

vtune_contract/config/collection.cfg

@@ -0,0 +1,34 @@
+<?xml version='1.0' encoding='UTF-8'?>
+
+<bag xmlns:boolean="http://www.w3.org/2001/XMLSchema#boolean" xmlns:double="http://www.intel.com/2001/XMLSchema#double" xmlns:int="http://www.w3.org/2001/XMLSchema#int" xmlns:null="http://www.intel.com/2009/BagSchema#null">
+ <pointer:workload pointer:type_id="dasID_cctrl2::ApplicationWorkload" pointer:ptr_type="dasID_cctrl2::IWorkload">
+  <workload targetName="launch" launch_app.app_to_launch="mpirun" launch_app.app_parameters="&quot;-np&quot; &quot;2&quot; &quot;--hostfile&quot; &quot;/home/yx/qibotn/hostfile&quot; &quot;-perhost&quot; &quot;2&quot; &quot;--map-by&quot; &quot;numa&quot; &quot;python&quot; &quot;benchmark_contract_sliced.py&quot; " result_directory.path="/home/yx/qibotn/vtune_contract">
+   <search_directories/>
+   <context>
+    <contextValue id="allowMultipleRuns" boolean:value="false"/>
+    <contextValue id="analyzeKvmGuest" boolean:value="false"/>
+    <contextValue id="analyzeSystemWide" boolean:value="false"/>
+    <contextValue id="cpuMask" value=""/>
+    <contextValue id="customCollector" value=""/>
+    <contextValue id="dataLimit" int:value="1000"/>
+    <contextValue id="enableRing" boolean:value="false"/>
+    <contextValue id="finalizationMode" value="fast"/>
+    <contextValue id="followChild" boolean:value="true"/>
+    <contextValue id="followChildGroup" null:value=""/>
+    <contextValue id="followChildStrategy" value=""/>
+    <contextValue id="groupForFinalizationControl" null:value=""/>
+    <contextValue id="kvmGuestKallsyms" value=""/>
+    <contextValue id="kvmGuestModules" value=""/>
+    <contextValue id="kvmProfileGuest" null:value=""/>
+    <contextValue id="mrteMode" value="auto"/>
+    <contextValue id="targetDurationType" value="short"/>
+    <contextValue id="targetRingBuffer" double:value="0"/>
+    <contextValue id="targetType" value="launch"/>
+    <contextValue id="traceMpi" boolean:value="false"/>
+    <contextValue id="tracingMode" null:value=""/>
+    <contextValue id="wrapperScriptContent" value=""/>
+    <contextValue id="wrapperScriptPath" value=""/>
+   </context>
+  </workload>
+ </pointer:workload>
+</bag>

vtune_contract/config/context_values.cfg

@@ -0,0 +1,369 @@
+<?xml version='1.0' encoding='UTF-8'?>
+
+<bag xmlns:boolean="http://www.w3.org/2001/XMLSchema#boolean" xmlns:double="http://www.intel.com/2001/XMLSchema#double" xmlns:int="http://www.w3.org/2001/XMLSchema#int" xmlns:long="http://www.w3.org/2001/XMLSchema#long" xmlns:null="http://www.intel.com/2009/BagSchema#null" xmlns:unsignedByte="http://www.w3.org/2001/XMLSchema#unsignedByte" xmlns:unsignedInt="http://www.w3.org/2001/XMLSchema#unsignedInt" xmlns:unsignedLong="http://www.w3.org/2001/XMLSchema#unsignedLong" xmlns:unsignedShort="http://www.w3.org/2001/XMLSchema#unsignedShort">
+ <contextValue id="CLIENT_ID" value="CLI"/>
+ <contextValue id="CPU_NAME" value="Intel(R) Xeon(R) Processor code named Emeraldrapids"/>
+ <contextValue id="GPUUserGroups" value=""/>
+ <contextValue id="Hypervisor" value="None"/>
+ <contextValue id="HypervisorType" value="None"/>
+ <contextValue id="IsNUMANodeWithoutCPUsPresent" boolean:value="false"/>
+ <contextValue id="L2CATDetails" value="COS=8;ways=16"/>
+ <contextValue id="L3CATDetails" value="COS=15;ways=20"/>
+ <contextValue id="LLCSize" unsignedLong:value="272629760"/>
+ <contextValue id="LinuxPerfCapabilities" value="breakpoint:raw;cpu:raw,format,events,ldlat,frontend,offcore_rsp;cstate_core:raw,format,events;cstate_pkg:raw,format,events;dsa0:raw,format;dsa2:raw,format;iax1:raw,format;iax3:raw,format;intel_bts:raw;intel_pt:raw,format;kprobe:raw,format;msr:raw,format,events;power:raw,format,events;software:raw;tracepoint:raw;uncore_cha:52,raw,format;uncore_iio:10,raw,format;uncore_iio_free_running:10,raw,format,events;uncore_imc:8,raw,format,events;uncore_imc_free_running:4,raw,format,events;uncore_irp:10,raw,format;uncore_m2m:4,raw,format;uncore_m2pcie:10,raw,format;uncore_m3upi:4,raw,format;uncore_mdf:14,raw,format;uncore_pcu:raw,format;uncore_upi:3,raw,format;uprobe:raw,format"/>
+ <contextValue id="LinuxPerfCredentials" value="User"/>
+ <contextValue id="LinuxPerfMuxIntervalMs" int:value="1"/>
+ <contextValue id="LinuxPerfStackCapabilities" value="fp,dwarf,lbr"/>
+ <contextValue id="LinuxPerfVersion" value="6.6.99-vtune"/>
+ <contextValue id="LinuxRelease" value="6.12.0-124.52.1.el10_1.x86_64"/>
+ <contextValue id="OS" value="Linux"/>
+ <contextValue id="OS@version" int:value="1"/>
+ <contextValue id="OSBitness" value="64"/>
+ <contextValue id="OSBuildNumber" unsignedShort:value="0"/>
+ <contextValue id="PMU" value="emeraldrapids_server"/>
+ <contextValue id="PerfmonVersion" unsignedByte:value="5"/>
+ <contextValue id="RootPrivileges" boolean:value="false"/>
+ <contextValue id="accurateCPUTimeDetection" boolean:value="false"/>
+ <contextValue id="adjustCollectionBoundsByOMPApps" boolean:value="false"/>
+ <contextValue id="advancedLoopAnalysis" boolean:value="false"/>
+ <contextValue id="allowMultipleRuns" boolean:value="false"/>
+ <contextValue id="allowedViewpoints" value="%HotspotsByCPUUsageViewpointName,%TasksOverviewViewpointName"/>
+ <contextValue id="analyzeActivePowerConsumption" boolean:value="false"/>
+ <contextValue id="analyzeDgfxBandwidth" boolean:value="false"/>
+ <contextValue id="analyzeEHFIClasses" boolean:value="false"/>
+ <contextValue id="analyzeEnergyConsumption" boolean:value="false"/>
+ <contextValue id="analyzeFPU" boolean:value="false"/>
+ <contextValue id="analyzeFullProcTrace" boolean:value="false"/>
+ <contextValue id="analyzeIdlePowerConsumption" boolean:value="false"/>
+ <contextValue id="analyzeKvmGuest" boolean:value="false"/>
+ <contextValue id="analyzeMemoryConsumption" boolean:value="false"/>
+ <contextValue id="analyzePersistentMemory" boolean:value="false"/>
+ <contextValue id="analyzeSystemWide" boolean:value="false"/>
+ <contextValue id="appRunsCount" unsignedInt:value="1"/>
+ <contextValue id="apsMode" boolean:value="false"/>
+ <contextValue id="areKernelPtrsRestricted" value="yes"/>
+ <contextValue id="atraceEventConfig" value=""/>
+ <contextValue id="basicBlockAnalysis" boolean:value="true"/>
+ <contextValue id="cacheMonitoringUpscalingFactor" unsignedLong:value="106496"/>
+ <contextValue id="calleeAttributionMode" value="undefined"/>
+ <contextValue id="chipsetEventConfig" value=""/>
+ <contextValue id="collectCAT" boolean:value="false"/>
+ <contextValue id="collectCPUGPUBandwidth" boolean:value="false"/>
+ <contextValue id="collectCacheOccupancy" boolean:value="false"/>
+ <contextValue id="collectCallCounts" boolean:value="false"/>
+ <contextValue id="collectFPGAOpenCl" boolean:value="false"/>
+ <contextValue id="collectFpgaBlueStreamEvents" boolean:value="false"/>
+ <contextValue id="collectFramesMode" boolean:value="true"/>
+ <contextValue id="collectFullProcTrace" boolean:value="false"/>
+ <contextValue id="collectGpuCm" boolean:value="false"/>
+ <contextValue id="collectGpuMetal" boolean:value="false"/>
+ <contextValue id="collectGpuOpenCl" value=""/>
+ <contextValue id="collectGpuOpenClArg" value=""/>
+ <contextValue id="collectHwTrace" boolean:value="false"/>
+ <contextValue id="collectIOMMU" boolean:value="false"/>
+ <contextValue id="collectIoMode" value="off"/>
+ <contextValue id="collectIoWaits" boolean:value="false"/>
+ <contextValue id="collectL3ExternalBW" boolean:value="false"/>
+ <contextValue id="collectMemBandwidth" boolean:value="false"/>
+ <contextValue id="collectMemObjects" boolean:value="false"/>
+ <contextValue id="collectMmioAccess" boolean:value="false"/>
+ <contextValue id="collectOpenMPRegions" boolean:value="false"/>
+ <contextValue id="collectOsCounters" boolean:value="false"/>
+ <contextValue id="collectPCIeBandwidth" boolean:value="false"/>
+ <contextValue id="collectPCIeP2PBandwidth" boolean:value="false"/>
+ <contextValue id="collectPStateData" boolean:value="true"/>
+ <contextValue id="collectPTforTSX" boolean:value="false"/>
+ <contextValue id="collectPreciseClockticks" boolean:value="false"/>
+ <contextValue id="collectSWHotspots" boolean:value="true"/>
+ <contextValue id="collectSamplesMode" value="stack"/>
+ <contextValue id="collectSignalsMode" value="off"/>
+ <contextValue id="collectTSXCycles" boolean:value="false"/>
+ <contextValue id="collectThrottlingReasons" boolean:value="false"/>
+ <contextValue id="collectTopology" boolean:value="false"/>
+ <contextValue id="collectTripCounts" boolean:value="false"/>
+ <contextValue id="collectUserDataAllMode" boolean:value="false"/>
+ <contextValue id="collectUserSyncMode" boolean:value="false"/>
+ <contextValue id="collectUserTasksEventsCountersMode" boolean:value="true"/>
+ <contextValue id="collectWaitsMode" value="off"/>
+ <contextValue id="collectXpuCompute" value="true"/>
+ <contextValue id="collectionStart" value="15:57:39 07/05/2026 UTC"/>
+ <contextValue id="collectionStart@version" int:value="2"/>
+ <contextValue id="collectionStartTsc" unsignedLong:value="5790238599043820950"/>
+ <contextValue id="collectionStop" value="15:58:09 07/05/2026 UTC"/>
+ <contextValue id="collectionStop@version" int:value="2"/>
+ <contextValue id="collectionStopTsc" unsignedLong:value="5790238892163798430"/>
+ <contextValue id="collectorTypeDecription" value="Driverless Perf per-process counting,User-mode sampling and tracing"/>
+ <contextValue id="collectorTypeDecription@version" int:value="1"/>
+ <contextValue id="commandLine" value="mpirun &quot;-np&quot; &quot;2&quot; &quot;--hostfile&quot; &quot;/home/yx/qibotn/hostfile&quot; &quot;-perhost&quot; &quot;2&quot; &quot;--map-by&quot; &quot;numa&quot; &quot;python&quot; &quot;benchmark_contract_sliced.py&quot; "/>
+ <contextValue id="commandLine@version" int:value="1"/>
+ <contextValue id="computerName" value="node-3"/>
+ <contextValue id="computerName@version" int:value="1"/>
+ <contextValue id="connectionType" value="localhost"/>
+ <contextValue id="coreFrequencies" value="2699999873"/>
+ <contextValue id="coreFrequencies@version" int:value="1"/>
+ <contextValue id="cpuByIoWaits" boolean:value="false"/>
+ <contextValue id="cpuFamily" unsignedInt:value="6"/>
+ <contextValue id="cpuFamily@version" int:value="1"/>
+ <contextValue id="cpuGpuUsageData" boolean:value="false"/>
+ <contextValue id="cpuMask" value=""/>
+ <contextValue id="cpuModel" unsignedInt:value="207"/>
+ <contextValue id="cpuModel@version" int:value="1"/>
+ <contextValue id="createGPUQueueFrames" boolean:value="false"/>
+ <contextValue id="criticalRankOpenmpThreadCount" unsignedInt:value="0"/>
+ <contextValue id="criticalRankOpenmpThreadCount@version" int:value="1"/>
+ <contextValue id="cswitchMode" value="inactive"/>
+ <contextValue id="customCollector" value=""/>
+ <contextValue id="dataLimit" int:value="1000"/>
+ <contextValue id="disableGPUSysinfo" boolean:value="false"/>
+ <contextValue id="disableRetCompression" boolean:value="false"/>
+ <contextValue id="dramBandwidthLimits" boolean:value="false"/>
+ <contextValue id="dumpComputeTaskBinaries" boolean:value="false"/>
+ <contextValue id="emonCSVTraceFormat" boolean:value="false"/>
+ <contextValue id="emonMachineReadableVersionFile" boolean:value="false"/>
+ <contextValue id="emonPreset" value=""/>
+ <contextValue id="enableCStateCollection" boolean:value="false"/>
+ <contextValue id="enableCSwitch" boolean:value="false"/>
+ <contextValue id="enableCharacterizationInsights" boolean:value="true"/>
+ <contextValue id="enableCycleAccurateMode" boolean:value="false"/>
+ <contextValue id="enableDramBandwidthLimitsWarning" boolean:value="false"/>
+ <contextValue id="enableHWBasedCSCollection" boolean:value="false"/>
+ <contextValue id="enableInterruptsCollection" boolean:value="false"/>
+ <contextValue id="enableLBRCollection" boolean:value="false"/>
+ <contextValue id="enableMemoryObjectCorrelation" boolean:value="false"/>
+ <contextValue id="enableMpiTracing" boolean:value="false"/>
+ <contextValue id="enableOpenglesInstrumentation" boolean:value="false"/>
+ <contextValue id="enablePEBSCollection" boolean:value="false"/>
+ <contextValue id="enableParallelFsCollection" boolean:value="false"/>
+ <contextValue id="enableRing" boolean:value="false"/>
+ <contextValue id="enableRing0ProfilingMode" boolean:value="false"/>
+ <contextValue id="enableStackCollect" boolean:value="false"/>
+ <contextValue id="enableStackCollection" boolean:value="false"/>
+ <contextValue id="enableThreadAffinity" boolean:value="false"/>
+ <contextValue id="enableTimedPEBSCollection" boolean:value="false"/>
+ <contextValue id="enableVTSSCollection" boolean:value="false"/>
+ <contextValue id="energyProfilingMode" value="none"/>
+ <contextValue id="environmentVars" null:value=""/>
+ <contextValue id="environmentVars@version" int:value="1"/>
+ <contextValue id="errorsAsWarnings" boolean:value="false"/>
+ <contextValue id="eventInfo" boolean:value="false"/>
+ <contextValue id="eventMode" value="all"/>
+ <contextValue id="eventMuxFactor" int:value="5"/>
+ <contextValue id="explicitUncoreEventsConfig" boolean:value="false"/>
+ <contextValue id="fileRequiestLogic" value="local"/>
+ <contextValue id="finalizationMode" value="fast"/>
+ <contextValue id="followChild" boolean:value="true"/>
+ <contextValue id="followChildGroup" null:value=""/>
+ <contextValue id="followChildStrategy" value=""/>
+ <contextValue id="forceMuxOff" boolean:value="false"/>
+ <contextValue id="forceShowInlines" boolean:value="false"/>
+ <contextValue id="forceSystemWide" boolean:value="false"/>
+ <contextValue id="fpgaAocxOrBinaryFile" value=""/>
+ <contextValue id="fpgaNoMemTransfers" boolean:value="false"/>
+ <contextValue id="fpgaNoTemporal" boolean:value="false"/>
+ <contextValue id="fpgaOnBoard" value="None"/>
+ <contextValue id="fpgaPeriod" int:value="0"/>
+ <contextValue id="fpgaSourceFile" value=""/>
+ <contextValue id="fpuVersion" value="1_0"/>
+ <contextValue id="ftraceEventConfig" value=""/>
+ <contextValue id="genArchOnBoard" int:value="0"/>
+ <contextValue id="goodFastFrameThreshold" double:value="100"/>
+ <contextValue id="gpuAdapterNames" value=""/>
+ <contextValue id="gpuCounters" value=""/>
+ <contextValue id="gpuDisplayPrefix" value="GPU"/>
+ <contextValue id="gpuFixedMetricsSelected" value=""/>
+ <contextValue id="gpuHWProfiling" value=""/>
+ <contextValue id="gpuLLCSize" unsignedLong:value="272629760"/>
+ <contextValue id="gpuMetricsSelected" value=""/>
+ <contextValue id="gpuPerformanceMode" null:value=""/>
+ <contextValue id="gpuPlatform" null:value=""/>
+ <contextValue id="gpuPlatformIndex" null:value=""/>
+ <contextValue id="gpuPlatformIndex@version" int:value="1"/>
+ <contextValue id="gpuProfilingMode" value=""/>
+ <contextValue id="gpuSamplingInterval" value=""/>
+ <contextValue id="gpuUsage" boolean:value="false"/>
+ <contextValue id="groupForCustomControl" null:value=""/>
+ <contextValue id="groupForFinalizationControl" null:value=""/>
+ <contextValue id="groupForGPUCustomCollection" null:value=""/>
+ <contextValue id="groupForTraceEventConfig" null:value=""/>
+ <contextValue id="handleLostEvents" boolean:value="false"/>
+ <contextValue id="hideSystemByDefault" boolean:value="false"/>
+ <contextValue id="hideWarningInPerfsnapshot" boolean:value="false"/>
+ <contextValue id="hostName" value="node-3"/>
+ <contextValue id="hostOS" value="Linux"/>
+ <contextValue id="hostsCount" int:value="1"/>
+ <contextValue id="hostsCount@version" int:value="1"/>
+ <contextValue id="i915Status" value="MissingDriver"/>
+ <contextValue id="ignorePowerData" boolean:value="false"/>
+ <contextValue id="inKernelProfilingAnalysis" boolean:value="false"/>
+ <contextValue id="initialViewpoint" value="%HotspotsByCPUUsageViewpointName"/>
+ <contextValue id="iptCollectEvents" boolean:value="false"/>
+ <contextValue id="iptRegionsToLoad" int:value="0"/>
+ <contextValue id="is3DXP2LMMode" boolean:value="false"/>
+ <contextValue id="is3DXPAppDirectMode" boolean:value="false"/>
+ <contextValue id="is3DXPPresent" boolean:value="false"/>
+ <contextValue id="isAOCLAvailable" boolean:value="false"/>
+ <contextValue id="isCATSupportedByCPU" boolean:value="true"/>
+ <contextValue id="isCPUSupportedBySocwatch" boolean:value="true"/>
+ <contextValue id="isCSwitchAvailable" value="no"/>
+ <contextValue id="isCpuThrottlingAvailable" boolean:value="false"/>
+ <contextValue id="isDeviceOrCredentialGuardEnabled" boolean:value="false"/>
+ <contextValue id="isEHFIAvailable" boolean:value="false"/>
+ <contextValue id="isEnergyCollectionSupported" boolean:value="true"/>
+ <contextValue id="isFtraceAvailable" value="ftraceAccessError,debugfsNotAccessible"/>
+ <contextValue id="isFtraceAvailableKnob" value="ftraceAccessError,debugfsNotAccessible"/>
+ <contextValue id="isFunctionTracingAvailable" value="no"/>
+ <contextValue id="isGpuBusynessAvailable" value="unsupportedHardware"/>
+ <contextValue id="isGpuMultiRunRequired" boolean:value="false"/>
+ <contextValue id="isGpuWaitAvailable" value="no"/>
+ <contextValue id="isHTEnabled" boolean:value="false"/>
+ <contextValue id="isHybridCPU" boolean:value="false"/>
+ <contextValue id="isHybridCPU@version" int:value="1"/>
+ <contextValue id="isIowaitTracingAvailable" value="no"/>
+ <contextValue id="isL2CATAvailable" boolean:value="true"/>
+ <contextValue id="isL3CATAvailable" boolean:value="true"/>
+ <contextValue id="isL3CacheOccupancyAvailable" boolean:value="true"/>
+ <contextValue id="isL3LocalBWAvailable" boolean:value="true"/>
+ <contextValue id="isL3MonitoringSupportedByCPU" boolean:value="true"/>
+ <contextValue id="isL3TotalBWAvailable" boolean:value="true"/>
+ <contextValue id="isMaxDRAMBandwidthMeasurementSupported" boolean:value="true"/>
+ <contextValue id="isMdfEtwAvailable" boolean:value="false"/>
+ <contextValue id="isNMIWatchDogTimerRunning" boolean:value="true"/>
+ <contextValue id="isPAXDriverLoaded" boolean:value="false"/>
+ <contextValue id="isPStateAvailable" boolean:value="true"/>
+ <contextValue id="isPTAvailable" boolean:value="true"/>
+ <contextValue id="isPerfPCIeMappingAvailable" boolean:value="false"/>
+ <contextValue id="isPowerVRDataAvailable" value="no"/>
+ <contextValue id="isPtraceAvailable" boolean:value="true"/>
+ <contextValue id="isPtraceScopeLimited" boolean:value="false"/>
+ <contextValue id="isPytraceAvailable" boolean:value="true"/>
+ <contextValue id="isSEPDriverAvailable" boolean:value="false"/>
+ <contextValue id="isSGXAvailable" boolean:value="true"/>
+ <contextValue id="isSocwatchDriverLoaded" boolean:value="false"/>
+ <contextValue id="isTPSSAvailable" boolean:value="true"/>
+ <contextValue id="isTSXAvailable" boolean:value="false"/>
+ <contextValue id="isUArchUsageAvailable" boolean:value="false"/>
+ <contextValue id="isVSyncAvailable" value="no"/>
+ <contextValue id="isVTSSPPDriverAvailable" boolean:value="true"/>
+ <contextValue id="isXelinkAvailable" boolean:value="false"/>
+ <contextValue id="kernelStackWarningCheck" boolean:value="false"/>
+ <contextValue id="kernelsToProfile" value="*#1#1#4294967295"/>
+ <contextValue id="kvmGuestKallsyms" value=""/>
+ <contextValue id="kvmGuestModules" value=""/>
+ <contextValue id="kvmProfileGuest" null:value=""/>
+ <contextValue id="l0DevicesAvailable" boolean:value="false"/>
+ <contextValue id="l0GPUDevicesAvailable" boolean:value="false"/>
+ <contextValue id="l0LoaderStatus" value="LibNotFound"/>
+ <contextValue id="l0MetricConfig" value=""/>
+ <contextValue id="l0SamplingInterval" double:value="1"/>
+ <contextValue id="l0SamplingType" value="time"/>
+ <contextValue id="l0SysmanConfig" value=""/>
+ <contextValue id="l0VPUDevicesAvailable" boolean:value="false"/>
+ <contextValue id="loadLbrStackToDb" boolean:value="true"/>
+ <contextValue id="loadPebsData" boolean:value="true"/>
+ <contextValue id="loadRawLbrData" boolean:value="false"/>
+ <contextValue id="logicalCPUCount" int:value="96"/>
+ <contextValue id="logicalCPUCount@version" int:value="1"/>
+ <contextValue id="maxLocalBandwidth" unsignedLong:value="0"/>
+ <contextValue id="maxRegionDuration" double:value="100"/>
+ <contextValue id="memoryAccessBandwidthMeasuring" boolean:value="false"/>
+ <contextValue id="memoryObjectMinSize" int:value="1024"/>
+ <contextValue id="memoryType" value="Unknown"/>
+ <contextValue id="mpiRank" null:value=""/>
+ <contextValue id="mpiRank@version" int:value="2"/>
+ <contextValue id="mrteMode" value="auto"/>
+ <contextValue id="mrteType" value="java,python"/>
+ <contextValue id="muxGroupCount" int:value="3"/>
+ <contextValue id="nameThreadsAsCreationModule" boolean:value="false"/>
+ <contextValue id="noSignalAltstack" boolean:value="false"/>
+ <contextValue id="numaVersion" value="1_0"/>
+ <contextValue id="numaVersion@version" int:value="1"/>
+ <contextValue id="numaVersionCurrent" value="1_0"/>
+ <contextValue id="omniPathOnBoard" value="None"/>
+ <contextValue id="openclSourceAsm" boolean:value="true"/>
+ <contextValue id="openmpProcessCount" unsignedInt:value="0"/>
+ <contextValue id="openmpProcessCount@version" int:value="1"/>
+ <contextValue id="openmpThreadCount" unsignedInt:value="2"/>
+ <contextValue id="openmpThreadCount@version" int:value="2"/>
+ <contextValue id="osDetailedName" value="6.12.0-124.52.1.el10_1.x86_64 \S&#10;Kernel \r on \m&#10;"/>
+ <contextValue id="osDetailedName@version" int:value="1"/>
+ <contextValue id="osKernelRelease" value="6.12.0-124.52.1.el10_1.x86_64"/>
+ <contextValue id="osKernelRelease@version" int:value="1"/>
+ <contextValue id="osName" value="Linux"/>
+ <contextValue id="osName@version" int:value="1"/>
+ <contextValue id="packageCount" int:value="2"/>
+ <contextValue id="packageCount@version" int:value="1"/>
+ <contextValue id="pausedTime" double:value="0"/>
+ <contextValue id="pausedTime@version" int:value="2"/>
+ <contextValue id="pciClassParts" value=""/>
+ <contextValue id="perfForceSystemWide" boolean:value="false"/>
+ <contextValue id="physicalCoreCount" int:value="96"/>
+ <contextValue id="physicalCoreCount@version" int:value="1"/>
+ <contextValue id="platformType" value="157"/>
+ <contextValue id="pmuEventConfig" value="CPU_CLK_UNHALTED.THREAD,CPU_CLK_UNHALTED.REF_TSC:sample:sa=2700000,INST_RETIRED.ANY:sample:sa=2700000,CPU_CLK_UNHALTED.DISTRIBUTED,FP_ARITH_INST_RETIRED.SCALAR_SINGLE,FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE,FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE,FP_ARITH_INST_RETIRED.SCALAR_DOUBLE,FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE,FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE,UOPS_EXECUTED.X87,UOPS_RETIRED.SLOTS:sample:sa=2000003,UOPS_EXECUTED.THREAD,FP_ARITH_INST_RETIRED.512B_PACKED_DOUBLE,FP_ARITH_INST_RETIRED.512B_PACKED_SINGLE,FP_ARITH_INST_RETIRED2.SCALAR_HALF,FP_ARITH_INST_RETIRED2.128B_PACKED_HALF,FP_ARITH_INST_RETIRED2.256B_PACKED_HALF,FP_ARITH_INST_RETIRED2.512B_PACKED_HALF,TOPDOWN.SLOTS:sample:sa=2000003,TOPDOWN.SLOTS_P:sample:sa=2000003"/>
+ <contextValue id="pmuSamplingInterval" double:value="1"/>
+ <contextValue id="populatedIoParts" value=""/>
+ <contextValue id="populatedIoUnits" value=""/>
+ <contextValue id="populatedTidValuesForIO" value=""/>
+ <contextValue id="preciseMultiplexing" boolean:value="false"/>
+ <contextValue id="preferDriverlessCollection" boolean:value="true"/>
+ <contextValue id="preferedGpuAdapter" value="none"/>
+ <contextValue id="preferredEbsCollectorOrder" value="perf,vtss,sep"/>
+ <contextValue id="processKernelBinaries" boolean:value="false"/>
+ <contextValue id="referenceFrequency" unsignedLong:value="2699999873"/>
+ <contextValue id="referenceFrequency@version" int:value="2"/>
+ <contextValue id="resolveCallsites" boolean:value="true"/>
+ <contextValue id="restrictPCIeBandwidthByClass" value="None"/>
+ <contextValue id="resultSize" long:value="87587138"/>
+ <contextValue id="resultSize@version" int:value="1"/>
+ <contextValue id="ringBuffer" int:value="0"/>
+ <contextValue id="ringBufferSizeInSeconds" double:value="0"/>
+ <contextValue id="runss:enable" boolean:value="true"/>
+ <contextValue id="samplingInterval" double:value="5"/>
+ <contextValue id="samplingMode" value="sw"/>
+ <contextValue id="scPids" value="1035901"/>
+ <contextValue id="scStartTscs" value="5790238597824248360"/>
+ <contextValue id="shortCollectionMux" boolean:value="false"/>
+ <contextValue id="shortCollectionMux@version" int:value="1"/>
+ <contextValue id="showGPUBandwidthHistogram" boolean:value="true"/>
+ <contextValue id="showInlinesByDefault" boolean:value="true"/>
+ <contextValue id="slowGoodFrameThreshold" double:value="40"/>
+ <contextValue id="stackSize" int:value="0"/>
+ <contextValue id="stackSizeCollect" value="0"/>
+ <contextValue id="stackStitching" boolean:value="true"/>
+ <contextValue id="stackTypeCollect" value="software"/>
+ <contextValue id="stackUnwindLimit" int:value="8388608"/>
+ <contextValue id="stackwalk" value="offline"/>
+ <contextValue id="stallReasonsSamplingInterval" value=""/>
+ <contextValue id="supportedTargetTypes" value="all"/>
+ <contextValue id="suppressCSVSyntaxWarnings" boolean:value="false"/>
+ <contextValue id="systemCollectorConfig" value=""/>
+ <contextValue id="systemWideContextSwitch" boolean:value="false"/>
+ <contextValue id="systemWideDiskIO" boolean:value="false"/>
+ <contextValue id="targetConcurrency" int:value="96"/>
+ <contextValue id="targetConcurrency@version" int:value="2"/>
+ <contextValue id="targetDurationType" value="short"/>
+ <contextValue id="targetGPU" value="none"/>
+ <contextValue id="targetOS" value="Linux"/>
+ <contextValue id="targetOS@version" int:value="1"/>
+ <contextValue id="targetOption" value="localhost"/>
+ <contextValue id="targetRingBuffer" double:value="0"/>
+ <contextValue id="targetType" value="launch"/>
+ <contextValue id="tidValuesForIO" value=""/>
+ <contextValue id="tmamVersion" value="5_1"/>
+ <contextValue id="tmamVersion@version" int:value="1"/>
+ <contextValue id="tmamVersionCurrent" value="5_1"/>
+ <contextValue id="totalElapsedTime" double:value="29.311997748100001"/>
+ <contextValue id="totalElapsedTime@version" int:value="3"/>
+ <contextValue id="traceMpi" boolean:value="false"/>
+ <contextValue id="tracingMode" null:value=""/>
+ <contextValue id="uncoreSamplingInterval" int:value="10"/>
+ <contextValue id="useAOCLProfile" boolean:value="false"/>
+ <contextValue id="useAggregatedCounting" boolean:value="false"/>
+ <contextValue id="useCountingMode" boolean:value="true"/>
+ <contextValue id="useEventBasedCounts" boolean:value="false"/>
+ <contextValue id="useGpuCounting" boolean:value="false"/>
+ <contextValue id="usePerf" boolean:value="false"/>
+ <contextValue id="usePerfMetrics" boolean:value="false"/>
+ <contextValue id="wrapperScriptContent" value=""/>
+ <contextValue id="wrapperScriptPath" value=""/>
+</bag>

vtune_contract/config/log.cfg

@@ -0,0 +1,71 @@
+<?xml version='1.0' encoding='UTF-8'?>
+
+<bag xmlns:int="http://www.w3.org/2001/XMLSchema#int" xmlns:long="http://www.w3.org/2001/XMLSchema#long">
+ <message_entry_t int:status="0" cap="Collection failed" msg="Collection failed. The data cannot be displayed. " long:timeStamp="1778169489297">
+  <message msg="Only user space will be profiled due to credentials lack. Consider changing /proc/sys/kernel/perf_event_paranoid file for enabling kernel space profiling." int:severity="2"/>
+  <message msg="Collection started. To stop the collection, either press CTRL-C or enter from another console window: vtune -r /home/yx/qibotn/vtune_contract -command stop." int:severity="1"/>
+  <message msg="Unauthorized control server connection." int:severity="4"/>
+  <message msg="Collection stopped." int:severity="1"/>
+ </message_entry_t>
+ <message_entry_t int:status="1" cap="Finalization completed with warnings" msg="Result finalization has completed with warnings that may affect the representation of analysis data. Please see details below." long:timeStamp="1778169531570">
+  <message msg="Using result path `/home/yx/qibotn/vtune_contract'" int:severity="1"/>
+  <message msg="The database has been cleared, elapsed time is 0.295 seconds." int:severity="16"/>
+  <message msg="Raw data has been loaded to the database, elapsed time is 14.977 seconds." int:severity="16"/>
+  <message msg="Data transformations have been finished, elapsed time is 0.017 seconds." int:severity="16"/>
+  <message msg="Cannot locate file `&lt;frozen importlib._bootstrap&gt;'." int:severity="2"/>
+  <message msg="Cannot locate file `&lt;frozen importlib._bootstrap_external&gt;'." int:severity="2"/>
+  <message msg="Cannot locate file `&lt;string&gt;.ko'." int:severity="2"/>
+  <message msg="Cannot locate file `&lt;frozen abc&gt;.ko'." int:severity="2"/>
+  <message msg="Cannot locate file `&lt;frozen posixpath&gt;.ko'." int:severity="2"/>
+  <message msg="Cannot locate file `/usr/lib64/python3.12/encodings/cp775.py'." int:severity="2"/>
+  <message msg="Cannot locate file `&lt;shim&gt;.ko'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/lib64/libibverbs.so.1'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/usr/bin/grep'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/usr/bin/bash'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/intel/oneapi/mpi/2021.17/opt/mpi/libfabric/lib/prov/libmlx-fi.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/lib64/libucp.so.0'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/lib64/libuct.so.0'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/lib64/libucs.so.0'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/intel/oneapi/vtune/2025.7/lib64/libtpsstool.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/lib64/libucm.so.0'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/usr/lib64/libibverbs/libmlx5-rdmav57.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/lib64/ucx/libuct_ib.so.0'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/usr/bin/lscpu'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/lib64/libc.so.6'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/lib64/ucx/libuct_ib_mlx5.so.0'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/torch/lib/libgomp.so.1'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/torch/lib/libtorch_python.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/usr/bin/python3.12'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/lib64/libpython3.12.so.1.0'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/usr/bin/whoami'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/scipy/linalg/_decomp_interpolative.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/usr/bin/uname'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/usr/lib64/python3.12/lib-dynload/_pickle.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/lib64/libstdc++.so.6'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/scipy/spatial/_ckdtree.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/usr/lib64/python3.12/lib-dynload/_ctypes.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/numpy/_core/_multiarray_umath.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/lib64/libffi.so.8'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/mpi4py/MPI.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/scipy/interpolate/_rgi_cython.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/intel/oneapi/vtune/2025.7/lib64/runtime/libittnotify_collector.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/numpy/random/bit_generator.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/numpy/random/_pcg64.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/numpy/random/_generator.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/torch/lib/libc10.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/scipy/_cyutility.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/81d243bd2c585b0f4821__mypyc.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/intel/oneapi/mpi/2021.17/opt/mpi/libfabric/lib/libfabric.so.1'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/scipy/linalg/_flapack.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/scipy/optimize/_highspy/_core.cpython-312-x86_64-linux-gnu.so'." int:severity="2"/>
+  <message msg="Cannot locate debugging information for file `/home/yx/qibotn/.venv/lib64/python3.12/site-packages/torch/lib/libtorch_cpu.so'." int:severity="2"/>
+  <message msg="Symbol resolution has been finished, elapsed time is 22.559 seconds." int:severity="16"/>
+  <message msg="Deferred data transformations have been finished, elapsed time is 0.526 seconds." int:severity="16"/>
+  <message msg="Data model parameters have been set, elapsed time is 0.017 seconds." int:severity="16"/>
+  <message msg="Precomputing frequently used data has been finished, elapsed time is 0.342 seconds." int:severity="16"/>
+  <message msg="Redundant overtime data has been discarded, elapsed time is 0.002 seconds." int:severity="16"/>
+  <message msg="Raw collector data has been discarded, elapsed time is 0.000 seconds." int:severity="16"/>
+  <message msg="Finalizing the result took 41.182 seconds." int:severity="16"/>
+  <message msg="Knob values have been set, elapsed time is 0.000 seconds." int:severity="16"/>
+ </message_entry_t>
+</bag>

vtune_contract/config/runss.options

@@ -0,0 +1,31 @@
+-r
+/home/yx/qibotn/vtune_contract
+--stack-stitching
+--data-limit-mb=1000
+--disk-space-limit=0
+--mrte-type=java,python
+--stack-unwind-limit=8388608
+--itt-config=frame
+--itt-config=task,event,counter
+--stackwalk=offline
+--mrte-mode=auto
+--type=cpu:counters:nostack
+--type=cpu:stack
+--interval=10
+--platform-type=157
+--pmu-type=emeraldrapids_server
+--event-config=CPU_CLK_UNHALTED.THREAD,CPU_CLK_UNHALTED.REF_TSC:sample:sa=2700000,INST_RETIRED.ANY:sample:sa=2700000,CPU_CLK_UNHALTED.DISTRIBUTED,FP_ARITH_INST_RETIRED.SCALAR_SINGLE,FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE,FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE,FP_ARITH_INST_RETIRED.SCALAR_DOUBLE,FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE,FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE,UOPS_EXECUTED.X87,UOPS_RETIRED.SLOTS:sample:sa=2000003,UOPS_EXECUTED.THREAD,FP_ARITH_INST_RETIRED.512B_PACKED_DOUBLE,FP_ARITH_INST_RETIRED.512B_PACKED_SINGLE,FP_ARITH_INST_RETIRED2.SCALAR_HALF,FP_ARITH_INST_RETIRED2.128B_PACKED_HALF,FP_ARITH_INST_RETIRED2.256B_PACKED_HALF,FP_ARITH_INST_RETIRED2.512B_PACKED_HALF,TOPDOWN.SLOTS:sample:sa=2000003,TOPDOWN.SLOTS_P:sample:sa=2000003
+--collector=perf
+--count
+--
+mpirun
+-np
+2
+--hostfile
+/home/yx/qibotn/hostfile
+-perhost
+2
+--map-by
+numa
+python
+benchmark_contract_sliced.py

vtune_contract/config/search_dir.cfg

@@ -0,0 +1,19 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<root>
+ <All_Configurations-All_Platforms>
+  <searchDirs>
+   <searchCategory>
+    <category type="u8_t">1</category>
+   </searchCategory>
+   <searchCategory>
+    <category type="u8_t">2</category>
+   </searchCategory>
+   <searchCategory>
+    <category type="u8_t">3</category>
+   </searchCategory>
+   <searchCategory>
+    <category type="u8_t">4</category>
+   </searchCategory>
+  </searchDirs>
+ </All_Configurations-All_Platforms>
+</root>

vtune_contract/config/state.cfg

@@ -0,0 +1,4 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<root>
+ <guiEngine type="bag"/>
+</root>

vtune_contract/data.0/1035901.stat.perf

@@ -0,0 +1,129 @@
+# AMPLXE_PERF_STAT
+# hostname : node-3
+# os release : 6.12.0-124.52.1.el10_1.x86_64
+# arch : x86_64
+# cpuid : GenuineIntel,6,207,2
+# cpudesc : INTEL(R) XEON(R) PLATINUM 8558P
+# total memory : 527171272 kB
+# nrcpus online : 96
+# nrcpus avail : 96
+# sibling cores : 0-47
+# sibling cores : 48-95
+# sibling threads : 0
+# sibling threads : 1
+# sibling threads : 2
+# sibling threads : 3
+# sibling threads : 4
+# sibling threads : 5
+# sibling threads : 6
+# sibling threads : 7
+# sibling threads : 8
+# sibling threads : 9
+# sibling threads : 10
+# sibling threads : 11
+# sibling threads : 12
+# sibling threads : 13
+# sibling threads : 14
+# sibling threads : 15
+# sibling threads : 16
+# sibling threads : 17
+# sibling threads : 18
+# sibling threads : 19
+# sibling threads : 20
+# sibling threads : 21
+# sibling threads : 22
+# sibling threads : 23
+# sibling threads : 24
+# sibling threads : 25
+# sibling threads : 26
+# sibling threads : 27
+# sibling threads : 28
+# sibling threads : 29
+# sibling threads : 30
+# sibling threads : 31
+# sibling threads : 32
+# sibling threads : 33
+# sibling threads : 34
+# sibling threads : 35
+# sibling threads : 36
+# sibling threads : 37
+# sibling threads : 38
+# sibling threads : 39
+# sibling threads : 40
+# sibling threads : 41
+# sibling threads : 42
+# sibling threads : 43
+# sibling threads : 44
+# sibling threads : 45
+# sibling threads : 46
+# sibling threads : 47
+# sibling threads : 48
+# sibling threads : 49
+# sibling threads : 50
+# sibling threads : 51
+# sibling threads : 52
+# sibling threads : 53
+# sibling threads : 54
+# sibling threads : 55
+# sibling threads : 56
+# sibling threads : 57
+# sibling threads : 58
+# sibling threads : 59
+# sibling threads : 60
+# sibling threads : 61
+# sibling threads : 62
+# sibling threads : 63
+# sibling threads : 64
+# sibling threads : 65
+# sibling threads : 66
+# sibling threads : 67
+# sibling threads : 68
+# sibling threads : 69
+# sibling threads : 70
+# sibling threads : 71
+# sibling threads : 72
+# sibling threads : 73
+# sibling threads : 74
+# sibling threads : 75
+# sibling threads : 76
+# sibling threads : 77
+# sibling threads : 78
+# sibling threads : 79
+# sibling threads : 80
+# sibling threads : 81
+# sibling threads : 82
+# sibling threads : 83
+# sibling threads : 84
+# sibling threads : 85
+# sibling threads : 86
+# sibling threads : 87
+# sibling threads : 88
+# sibling threads : 89
+# sibling threads : 90
+# sibling threads : 91
+# sibling threads : 92
+# sibling threads : 93
+# sibling threads : 94
+# sibling threads : 95
+# process id : 1035901
+;;1852162590870;0
+;;1954556299080;1
+;;4241120946016;2
+;;1852848155581;3
+;;66657;4
+;;0;5
+;;0;6
+;;5745804;7
+;;405292255;8
+;;296278351;9
+;;439721;10
+;;4247142966778;11
+;;4146749907087;12
+;;4563086821843;13
+;;201;14
+;;0;15
+;;0;16
+;;0;17
+;;0;18
+;;11112975545220;19
+;;11122858880946;20