diff --git a/services/audioflinger/AudioFlinger.cpp b/services/audioflinger/AudioFlinger.cpp
index eb20019..9353e70 100644
--- a/services/audioflinger/AudioFlinger.cpp
+++ b/services/audioflinger/AudioFlinger.cpp
@@ -4186,7 +4186,7 @@
         mCblk(NULL),
         // mBuffer
         // mBufferEnd
-        mFrameCount(0),
+        mStepCount(0),
         mState(IDLE),
         mSampleRate(sampleRate),
         mFormat(format),
@@ -4277,7 +4277,7 @@
 void AudioFlinger::ThreadBase::TrackBase::releaseBuffer(AudioBufferProvider::Buffer* buffer)
 {
     buffer->raw = NULL;
-    mFrameCount = buffer->frameCount;
+    mStepCount = buffer->frameCount;
     // FIXME See note at getNextBuffer()
     (void) step();      // ignore return value of step()
     buffer->frameCount = 0;
@@ -4287,7 +4287,7 @@
     bool result;
     audio_track_cblk_t* cblk = this->cblk();
 
-    result = cblk->stepServer(mFrameCount, isOut());
+    result = cblk->stepServer(mStepCount, isOut());
     if (!result) {
         ALOGV("stepServer failed acquiring cblk mutex");
         mStepServerFailed = true;
@@ -4435,7 +4435,7 @@
 
 /*static*/ void AudioFlinger::PlaybackThread::Track::appendDumpHeader(String8& result)
 {
-    result.append("   Name Client Type Fmt Chn mask   Session mFrCnt fCount S M F SRate  "
+    result.append("   Name Client Type Fmt Chn mask   Session StpCnt fCount S M F SRate  "
                   "L dB  R dB    Server      User     Main buf    Aux Buf  Flags Underruns\n");
 }
 
@@ -4506,7 +4506,7 @@
             mFormat,
             mChannelMask,
             mSessionId,
-            mFrameCount,
+            mStepCount,
             mCblk->frameCount,
             stateChar,
             mMute,
@@ -5499,7 +5499,7 @@
 
 /*static*/ void AudioFlinger::RecordThread::RecordTrack::appendDumpHeader(String8& result)
 {
-    result.append("   Clien Fmt Chn mask   Session Buf  S SRate  Serv     User   FrameCount\n");
+    result.append("   Clien Fmt Chn mask   Session Step S SRate  Serv     User   FrameCount\n");
 }
 
 void AudioFlinger::RecordThread::RecordTrack::dump(char* buffer, size_t size)
@@ -5509,7 +5509,7 @@
             mFormat,
             mChannelMask,
             mSessionId,
-            mFrameCount,
+            mStepCount,
             mState,
             mCblk->sampleRate,
             mCblk->server,
diff --git a/services/audioflinger/AudioFlinger.h b/services/audioflinger/AudioFlinger.h
index 38744d0..8cf58b1 100644
--- a/services/audioflinger/AudioFlinger.h
+++ b/services/audioflinger/AudioFlinger.h
@@ -453,7 +453,7 @@
                 return mState == TERMINATED;
             }
 
-            bool step();
+            bool step();    // mStepCount is an implicit input
             void reset();
 
             virtual bool isOut() const = 0; // true for Track and TimedTrack, false for RecordTrack,
@@ -466,7 +466,8 @@
             void*               mBuffer;    // start of track buffer, typically in shared memory
             void*               mBufferEnd; // &mBuffer[mFrameCount * frameSize], where frameSize
                                             //   is based on mChannelCount and 16-bit samples
-            uint32_t            mFrameCount;
+            uint32_t            mStepCount; // saves AudioBufferProvider::Buffer::frameCount as of
+                                            // time of releaseBuffer() for later use by step()
             // we don't really need a lock for these
             track_state         mState;
             const uint32_t      mSampleRate;    // initial sample rate only; for tracks which