[PATCH] D156530: [MISched] Do not erase resource booking history for subunits.

[Francesco Petrogalli via Phabricator via llvm-commits]

fpetrogalli created this revision.
fpetrogalli added a reviewer: andreadb.
Herald added subscribers: javed.absar, hiraditya, MatzeB.
Herald added a project: All.
fpetrogalli requested review of this revision.
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.

When dealing with the subunits of a resource group, we should reset
the subunits availability at the first avaiable cycle of the resource
that contains the subunits. Previously, the reset operation was
returning cycle 0, effectively erasing the booking history of the
subunits.

This change has effects only for schedule models that set
EnableIntervals = true.

For the existing scheduling models in open source LLVM, this is a NFC.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D156530

Files:
  llvm/lib/CodeGen/MachineScheduler.cpp


Index: llvm/lib/CodeGen/MachineScheduler.cpp
===================================================================
--- llvm/lib/CodeGen/MachineScheduler.cpp
+++ llvm/lib/CodeGen/MachineScheduler.cpp
@@ -2331,20 +2331,24 @@
          "Cannot have zero instances of a ProcResource");
 
   if (isUnbufferedGroup(PIdx)) {
-    // If any subunits are used by the instruction, report that the resource
-    // group is available at 0, effectively removing the group record from
-    // hazarding and basing the hazarding decisions on the subunit records.
-    // Otherwise, choose the first available instance from among the subunits.
-    // Specifications which assign cycles to both the subunits and the group or
-    // which use an unbuffered group with buffered subunits will appear to
-    // schedule strangely. In the first case, the additional cycles for the
-    // group will be ignored.  In the second, the group will be ignored
-    // entirely.
+    // If any subunits are used by the instruction, report that the
+    // subunits of the resource group are available at the first cycle
+    // in which the unit is available, effectively removing the group
+    // record from hazarding and basing the hazarding decisions on the
+    // subunit records. Otherwise, choose the first available instance
+    // from among the subunits.  Specifications which assign cycles to
+    // both the subunits and the group or which use an unbuffered
+    // group with buffered subunits will appear to schedule
+    // strangely. In the first case, the additional cycles for the
+    // group will be ignored.  In the second, the group will be
+    // ignored entirely.
     for (const MCWriteProcResEntry &PE :
          make_range(SchedModel->getWriteProcResBegin(SC),
                     SchedModel->getWriteProcResEnd(SC)))
       if (ResourceGroupSubUnitMasks[PIdx][PE.ProcResourceIdx])
-        return std::make_pair(0u, StartIndex);
+        return std::make_pair(
+            getNextResourceCycleByInstance(StartIndex, Cycles, StartAtCycle),
+            StartIndex);
 
     auto SubUnits = SchedModel->getProcResource(PIdx)->SubUnitsIdxBegin;
     for (unsigned I = 0, End = NumberOfInstances; I < End; ++I) {


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D156530.545126.patch
Type: text/x-patch
Size: 2218 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230728/c5d35e7a/attachment.bin>