Merge branch 'valgrind' of https://github.com/daurnimator/zig into da…

…urnimator-valgrind

CMakeLists.txt

@@ -674,6 +674,9 @@ set(ZIG_STD_FILES
     "std.zig"
     "testing.zig"
     "unicode.zig"
+    "valgrind.zig"
+    "valgrind/callgrind.zig"
+    "valgrind/memcheck.zig"
     "zig.zig"
     "zig/ast.zig"
     "zig/parse.zig"

std/heap.zig

@@ -273,6 +273,10 @@ pub const FixedBufferAllocator = struct {
     buffer: []u8,
 
     pub fn init(buffer: []u8) FixedBufferAllocator {
+        // This loop gets optimized out in ReleaseFast mode
+        for (buffer) |*byte| {
+            byte.* = undefined;
+        }
         return FixedBufferAllocator{
             .allocator = Allocator{
                 .allocFn = alloc,

std/mem.zig

@@ -49,6 +49,7 @@ pub const Allocator = struct {
     pub fn destroy(self: *Allocator, ptr: var) void {
         const non_const_ptr = @intToPtr([*]u8, @ptrToInt(ptr));
         self.freeFn(self, non_const_ptr[0..@sizeOf(@typeOf(ptr).Child)]);
+        _ = std.valgrind.freeLikeBlock(non_const_ptr, 0);
     }
 
     pub fn alloc(self: *Allocator, comptime T: type, n: usize) ![]T {
@@ -62,6 +63,7 @@ pub const Allocator = struct {
         const byte_count = math.mul(usize, @sizeOf(T), n) catch return Error.OutOfMemory;
         const byte_slice = try self.allocFn(self, byte_count, alignment);
         assert(byte_slice.len == byte_count);
+        _ = std.valgrind.mallocLikeBlock(byte_slice, 0, false);
         // This loop gets optimized out in ReleaseFast mode
         for (byte_slice) |*byte| {
             byte.* = undefined;
@@ -86,6 +88,12 @@ pub const Allocator = struct {
         const byte_count = math.mul(usize, @sizeOf(T), n) catch return Error.OutOfMemory;
         const byte_slice = try self.reallocFn(self, old_byte_slice, byte_count, alignment);
         assert(byte_slice.len == byte_count);
+        if (byte_slice.ptr == old_byte_slice.ptr) {
+            _ = std.valgrind.resizeInPlaceBlock(old_byte_slice, byte_count, 0);
+        } else {
+            _ = std.valgrind.freeLikeBlock(old_byte_slice.ptr, 0);
+            _ = std.valgrind.mallocLikeBlock(byte_slice, 0, false);
+        }
         if (n > old_mem.len) {
             // This loop gets optimized out in ReleaseFast mode
             for (byte_slice[old_byte_slice.len..]) |*byte| {
@@ -114,8 +122,15 @@ pub const Allocator = struct {
         // n <= old_mem.len and the multiplication didn't overflow for that operation.
         const byte_count = @sizeOf(T) * n;
 
-        const byte_slice = self.reallocFn(self, @sliceToBytes(old_mem), byte_count, alignment) catch unreachable;
+        const old_byte_slice = @sliceToBytes(old_mem);
+        const byte_slice = self.reallocFn(self, old_byte_slice, byte_count, alignment) catch unreachable;
         assert(byte_slice.len == byte_count);
+        if (byte_slice.ptr == old_byte_slice.ptr) {
+            _ = std.valgrind.resizeInPlaceBlock(old_byte_slice, byte_count, 0);
+        } else {
+            _ = std.valgrind.freeLikeBlock(old_byte_slice.ptr, 0);
+            _ = std.valgrind.mallocLikeBlock(byte_slice, 0, false);
+        }
         return @bytesToSlice(T, @alignCast(alignment, byte_slice));
     }
 
@@ -124,6 +139,7 @@ pub const Allocator = struct {
         if (bytes.len == 0) return;
         const non_const_ptr = @intToPtr([*]u8, @ptrToInt(bytes.ptr));
         self.freeFn(self, non_const_ptr[0..bytes.len]);
+        _ = std.valgrind.freeLikeBlock(non_const_ptr, 0);
     }
 };
 

std/std.zig

@@ -45,6 +45,7 @@ pub const rb = @import("rb.zig");
 pub const sort = @import("sort.zig");
 pub const testing = @import("testing.zig");
 pub const unicode = @import("unicode.zig");
+pub const valgrind = @import("valgrind.zig");
 pub const zig = @import("zig.zig");
 
 test "std" {
@@ -91,5 +92,6 @@ test "std" {
     _ = @import("sort.zig");
     _ = @import("testing.zig");
     _ = @import("unicode.zig");
+    _ = @import("valgrind.zig");
     _ = @import("zig.zig");
 }

std/valgrind/callgrind.zig

@@ -0,0 +1,87 @@
+const std = @import("../index.zig");
+const valgrind = std.valgrind;
+
+pub const CallgrindClientRequest = extern enum {
+    DumpStats = valgrind.ToolBase("CT"),
+    ZeroStats,
+    ToggleCollect,
+    DumpStatsAt,
+    StartInstrumentation,
+    StopInstrumentation,
+};
+
+fn doCallgrindClientRequestExpr(default: usize, request: CallgrindClientRequest,
+        a1: usize, a2: usize, a3: usize, a4: usize, a5: usize
+    ) usize
+{
+    return valgrind.doClientRequest(
+        default,
+        @intCast(usize, @enumToInt(request)),
+        a1, a2, a3, a4, a5);
+}
+
+fn doCallgrindClientRequestStmt(request: CallgrindClientRequest,
+        a1: usize, a2: usize, a3: usize, a4: usize, a5: usize
+    ) void
+{
+    _ = doCallgrindClientRequestExpr(0, request, a1, a2, a3, a4, a5);
+}
+
+
+
+/// Dump current state of cost centers, and zero them afterwards
+pub fn dumpStats() void {
+    doCallgrindClientRequestStmt(CallgrindClientRequest.DumpStats,
+        0, 0, 0, 0, 0);
+}
+
+
+/// Dump current state of cost centers, and zero them afterwards.
+/// The argument is appended to a string stating the reason which triggered
+/// the dump. This string is written as a description field into the
+/// profile data dump.
+pub fn dumpStatsAt(pos_str: [*]u8) void {
+    doCallgrindClientRequestStmt(CallgrindClientRequest.DumpStatsAt,
+        @ptrToInt(pos_str),
+        0, 0, 0, 0);
+}
+
+
+/// Zero cost centers
+pub fn zeroStats() void {
+    doCallgrindClientRequestStmt(CallgrindClientRequest.ZeroStats,
+        0, 0, 0, 0, 0);
+}
+
+
+/// Toggles collection state.
+/// The collection state specifies whether the happening of events
+/// should be noted or if they are to be ignored. Events are noted
+/// by increment of counters in a cost center
+pub fn toggleCollect() void {
+    doCallgrindClientRequestStmt(CallgrindClientRequest.ToggleCollect,
+        0, 0, 0, 0, 0);
+}
+
+
+/// Start full callgrind instrumentation if not already switched on.
+/// When cache simulation is done, it will flush the simulated cache;
+/// this will lead to an artificial cache warmup phase afterwards with
+/// cache misses which would not have happened in reality.
+pub fn startInstrumentation() void {
+    doCallgrindClientRequestStmt(CallgrindClientRequest.StartInstrumentation,
+        0, 0, 0, 0, 0);
+}
+
+
+/// Stop full callgrind instrumentation if not already switched off.
+/// This flushes Valgrinds translation cache, and does no additional
+/// instrumentation afterwards, which effectivly will run at the same
+/// speed as the "none" tool (ie. at minimal slowdown).
+/// Use this to bypass Callgrind aggregation for uninteresting code parts.
+/// To start Callgrind in this mode to ignore the setup phase, use
+/// the option "--instr-atstart=no".
+pub fn stopInstrumentation() void {
+    doCallgrindClientRequestStmt(CallgrindClientRequest.StopInstrumentation,
+        0, 0, 0, 0, 0);
+}