runtime: allow up to 128 GB of allocated memory

Incorporates code from CL 6828055.

Fixes #2142.

R=golang-dev, iant, devon.odell
CC=golang-dev
https://golang.org/cl/6826088

2 files changed, 17 insertions, 19 deletions

diff --git a/src/pkg/runtime/malloc.goc b/src/pkg/runtime/malloc.goc
index a96372451c..9353653acd 100644
--- a/src/pkg/runtime/malloc.goc
+++ b/src/pkg/runtime/malloc.goc
@@ -323,32 +323,30 @@ runtime·mallocinit(void)
 	// enough to hold 4 bits per allocated word.
 	if(sizeof(void*) == 8 && (limit == 0 || limit > (1<<30))) {
 		// On a 64-bit machine, allocate from a single contiguous reservation.
-		// 16 GB should be big enough for now.
+		// 128 GB (MaxMem) should be big enough for now.
 		//
 		// The code will work with the reservation at any address, but ask
-		// SysReserve to use 0x000000f800000000 if possible.
-		// Allocating a 16 GB region takes away 36 bits, and the amd64
+		// SysReserve to use 0x000000c000000000 if possible.
+		// Allocating a 128 GB region takes away 37 bits, and the amd64
 		// doesn't let us choose the top 17 bits, so that leaves the 11 bits
-		// in the middle of 0x00f8 for us to choose.  Choosing 0x00f8 means
-		// that the valid memory addresses will begin 0x00f8, 0x00f9, 0x00fa, 0x00fb.
-		// None of the bytes f8 f9 fa fb can appear in valid UTF-8, and
-		// they are otherwise as far from ff (likely a common byte) as possible.
-		// Choosing 0x00 for the leading 6 bits was more arbitrary, but it
-		// is not a common ASCII code point either.  Using 0x11f8 instead
+		// in the middle of 0x00c0 for us to choose.  Choosing 0x00c0 means
+		// that the valid memory addresses will begin 0x00c0, 0x00c1, ..., 0x0x00df.
+		// In little-endian, that's c0 00, c1 00, ..., df 00. None of those are valid
+		// UTF-8 sequences, and they are otherwise as far away from 
+		// ff (likely a common byte) as possible. An earlier attempt to use 0x11f8 
 		// caused out of memory errors on OS X during thread allocations.
 		// These choices are both for debuggability and to reduce the
 		// odds of the conservative garbage collector not collecting memory
 		// because some non-pointer block of memory had a bit pattern
 		// that matched a memory address.
 		//
-		// Actually we reserve 17 GB (because the bitmap ends up being 1 GB)
-		// but it hardly matters: fc is not valid UTF-8 either, and we have to
-		// allocate 15 GB before we get that far.
+		// Actually we reserve 136 GB (because the bitmap ends up being 8 GB)
+		// but it hardly matters: e0 00 is not valid UTF-8 either.
 		//
 		// If this fails we fall back to the 32 bit memory mechanism
-		arena_size = 16LL<<30;
+		arena_size = MaxMem;
 		bitmap_size = arena_size / (sizeof(void*)*8/4);
-		p = runtime·SysReserve((void*)(0x00f8ULL<<32), bitmap_size + arena_size);
+		p = runtime·SysReserve((void*)(0x00c0ULL<<32), bitmap_size + arena_size);
 	}
 	if (p == nil) {
 		// On a 32-bit machine, we can't typically get away
diff --git a/src/pkg/runtime/malloc.h b/src/pkg/runtime/malloc.h
index 765cd02eb2..916b473a00 100644
--- a/src/pkg/runtime/malloc.h
+++ b/src/pkg/runtime/malloc.h
@@ -114,12 +114,12 @@ enum
 	HeapAllocChunk = 1<<20,		// Chunk size for heap growth
 
 	// Number of bits in page to span calculations (4k pages).
-	// On 64-bit, we limit the arena to 16G, so 22 bits suffices.
-	// On 32-bit, we don't bother limiting anything: 20 bits for 4G.
+	// On 64-bit, we limit the arena to 128GB, or 37 bits.
+	// On 32-bit, we don't bother limiting anything, so we use the full 32-bit address.
 #ifdef _64BIT
-	MHeapMap_Bits = 22,
+	MHeapMap_Bits = 37 - PageShift,
 #else
-	MHeapMap_Bits = 20,
+	MHeapMap_Bits = 32 - PageShift,
 #endif
 
 	// Max number of threads to run garbage collection.
@@ -133,7 +133,7 @@ enum
 // This must be a #define instead of an enum because it
 // is so large.
 #ifdef _64BIT
-#define	MaxMem	(16ULL<<30)	/* 16 GB */
+#define	MaxMem	(1ULL<<(MHeapMap_Bits+PageShift))	/* 128 GB */
 #else
 #define	MaxMem	((uintptr)-1)
 #endif