rts: Add --internal-counters RTS flag and several counters

The existing internal counters:
* gc_alloc_block_sync
* whitehole_spin
* gen[g].sync
* gen[1].sync

are now not shown in the -s report unless --internal-counters is also passed.

If --internal-counters is passed we now show the counters above, reformatted, as
well as several other counters. In particular, we now count the yieldThread()
calls that SpinLocks do as well as their spins.

The added counters are:
* gc_spin (spin and yield)
* mut_spin (spin and yield)
* whitehole_threadPaused (spin only)
* whitehole_executeMessage (spin only)
* whitehole_lockClosure (spin only)
* waitForGcThreadsd (spin and yield)

As well as the following, which are not SpinLock-like things:
* any_work
* do_work
* scav_find_work

See the Note for descriptions of what these counters are.

We add busy_wait_nops in these loops along with the counter increment where it
was absent.

Old internal counters output:
```
gc_alloc_block_sync: 0
whitehole_gc_spin: 0
gen[0].sync: 0
gen[1].sync: 0
```

New internal counters output:
```
Internal Counters:
                                           Spins        Yields
    gc_alloc_block_sync                      323             0
    gc_spin                              9016713           752
    mut_spin                            57360944         47716
    whitehole_gc                               0           n/a
    whitehole_threadPaused                     0           n/a
    whitehole_executeMessage                   0           n/a
    whitehole_lockClosure                      0             0
    waitForGcThreads                           2           415
    gen[0].sync                                6             0
    gen[1].sync                                1             0

    any_work                                2017
    no_work                                 2014
    scav_find_work                          1004
```

Test Plan:
./validate

Check it builds with #define PROF_SPIN removed from includes/rts/Config.h

Reviewers: bgamari, erikd, simonmar, hvr

Reviewed By: simonmar

Subscribers: rwbarton, thomie, carter

GHC Trac Issues: #3553, #9221

Differential Revision: https://phabricator.haskell.org/D4302

4 files changed, 30 insertions, 2 deletions

diff --git a/includes/RtsAPI.h b/includes/RtsAPI.h
index 27a5080220..6f011cbf6e 100644
--- a/includes/RtsAPI.h
+++ b/includes/RtsAPI.h
@@ -210,6 +210,30 @@ typedef struct _RTSStats {
 
   GCDetails gc;
 
+  // -----------------------------------
+  // Internal Counters
+
+  // The number of times a GC thread spun on its 'gc_spin' lock.
+  // Will be zero if the rts was not built with PROF_SPIN
+  uint64_t gc_spin_spin;
+  // The number of times a GC thread yielded on its 'gc_spin' lock.
+  // Will be zero if the rts was not built with PROF_SPIN
+  uint64_t gc_spin_yield;
+  // The number of times a GC thread spun on its 'mut_spin' lock.
+  // Will be zero if the rts was not built with PROF_SPIN
+  uint64_t mut_spin_spin;
+  // The number of times a GC thread yielded on its 'mut_spin' lock.
+  // Will be zero if the rts was not built with PROF_SPIN
+  uint64_t mut_spin_yield;
+  // The number of times a GC thread has checked for work across all parallel
+  // GCs
+  uint64_t any_work;
+  // The number of times a GC thread has checked for work and found none across
+  // all parallel GCs
+  uint64_t no_work;
+  // The number of times a GC thread has iterated it's outer loop across all
+  // parallel GCs
+  uint64_t scav_find_work;
 } RTSStats;
 
 void getRTSStats (RTSStats *s);
diff --git a/includes/rts/Flags.h b/includes/rts/Flags.h
index aed4dca384..6487947749 100644
--- a/includes/rts/Flags.h
+++ b/includes/rts/Flags.h
@@ -195,6 +195,7 @@ typedef struct _MISC_FLAGS {
     bool generate_dump_file;
     bool generate_stack_trace;
     bool machineReadable;
+    bool internalCounters;       /* See Note [Internal Counter Stats] */
     StgWord linkerMemBase;       /* address to ask the OS for memory
                                   * for the linker, NULL ==> off */
 } MISC_FLAGS;
diff --git a/includes/rts/SpinLock.h b/includes/rts/SpinLock.h
index 6530a3a2f0..1dca02f795 100644
--- a/includes/rts/SpinLock.h
+++ b/includes/rts/SpinLock.h
@@ -27,7 +27,8 @@
 typedef struct SpinLock_
 {
     StgWord   lock;
-    StgWord64 spin; // DEBUG version counts how much it spins
+    StgWord64 spin;  // incremented every time we spin in ACQUIRE_SPIN_LOCK
+    StgWord64 yield; // incremented every time we yield in ACQUIRE_SPIN_LOCK
 } SpinLock;
 #else
 typedef StgWord SpinLock;
@@ -49,6 +50,7 @@ INLINE_HEADER void ACQUIRE_SPIN_LOCK(SpinLock * p)
             p->spin++;
             busy_wait_nop();
         }
+        p->yield++;
         yieldThread();
     } while (1);
 }
@@ -66,6 +68,7 @@ INLINE_HEADER void initSpinLock(SpinLock * p)
     write_barrier();
     p->lock = 1;
     p->spin = 0;
+    p->yield = 0;
 }
 
 #else
diff --git a/includes/rts/storage/GC.h b/includes/rts/storage/GC.h
index 2aed7c57ee..d4182dd7f9 100644
--- a/includes/rts/storage/GC.h
+++ b/includes/rts/storage/GC.h
@@ -120,7 +120,7 @@ typedef struct generation_ {
     // stats information
     uint32_t collections;
     uint32_t par_collections;
-    uint32_t failed_promotions;
+    uint32_t failed_promotions;         // Currently unused
 
     // ------------------------------------
     // Fields below are used during GC only