diff --git a/vlib/builtin/linux_bare/libc_impl.v b/vlib/builtin/linux_bare/libc_impl.v
index 7c7fbb9cdf..b5724b7351 100644
--- a/vlib/builtin/linux_bare/libc_impl.v
+++ b/vlib/builtin/linux_bare/libc_impl.v
@@ -111,12 +111,6 @@ fn memcmp(a &C.void, b &C.void, n usize) int {
 [export: 'free']
 [unsafe]
 fn __free(ptr &C.void) {
-	/*
-	err := mm_free(ptr)
-	if err != .enoerror {
-		eprintln('free error:')
-		panic(err)
-	}*/
 	unsafe {
 		global_allocator.free_(ptr)
 	}
diff --git a/vlib/builtin/linux_bare/memory_managment.v b/vlib/builtin/linux_bare/memory_managment.v
index e17520df06..5973c35a02 100644
--- a/vlib/builtin/linux_bare/memory_managment.v
+++ b/vlib/builtin/linux_bare/memory_managment.v
@@ -38,7 +38,7 @@ fn system_alloc(_ voidptr, size usize) (voidptr, usize, u32) {
 	map_flags := MapFlags(int(MapFlags.map_private) | int(MapFlags.map_anonymous))
 	// END CONSTS
 
-	a, e := sys_mmap(&byte(0), u64(size + sizeof(u64)), mem_prot, map_flags, -1, 0)
+	a, e := sys_mmap(&byte(0), u64(size), mem_prot, map_flags, -1, 0)
 
 	if e == .enoerror {
 		return a, size, 0
diff --git a/vlib/builtin/wasm_bare/libc_impl.v b/vlib/builtin/wasm_bare/libc_impl.v
new file mode 100644
index 0000000000..bbde0a5055
--- /dev/null
+++ b/vlib/builtin/wasm_bare/libc_impl.v
@@ -0,0 +1,173 @@
+module builtin
+
+import dlmalloc
+
+__global global_allocator dlmalloc.Dlmalloc
+
+[unsafe]
+pub fn memcpy(dest &C.void, src &C.void, n usize) &C.void {
+	dest_ := unsafe { &byte(dest) }
+	src_ := unsafe { &byte(src) }
+	unsafe {
+		for i in 0 .. int(n) {
+			dest_[i] = src_[i]
+		}
+	}
+	return unsafe { dest }
+}
+
+[export: 'malloc']
+[unsafe]
+fn __malloc(n usize) &C.void {
+	return unsafe { global_allocator.malloc(n) }
+}
+
+[unsafe]
+fn strlen(_s &C.void) usize {
+	s := unsafe { &byte(_s) }
+	mut i := 0
+	for ; unsafe { s[i] } != 0; i++ {}
+	return usize(i)
+}
+
+[unsafe]
+fn realloc(old_area &C.void, new_size usize) &C.void {
+	if old_area == 0 {
+		return unsafe { malloc(int(new_size)) }
+	}
+	if new_size == usize(0) {
+		unsafe { free(old_area) }
+		return 0
+	}
+	old_size := unsafe { *(&u64(old_area - sizeof(u64))) }
+	if u64(new_size) <= old_size {
+		return unsafe { old_area }
+	} else {
+		new_area := unsafe { malloc(int(new_size)) }
+		unsafe { memmove(new_area, old_area, usize(old_size)) }
+		unsafe { free(old_area) }
+		return new_area
+	}
+}
+
+[unsafe]
+fn memset(s &C.void, c int, n usize) &C.void {
+	mut s_ := unsafe { &char(s) }
+	for i in 0 .. int(n) {
+		unsafe {
+			s_[i] = char(c)
+		}
+	}
+	return unsafe { s }
+}
+
+[unsafe]
+fn memmove(dest &C.void, src &C.void, n usize) &C.void {
+	dest_ := unsafe { &byte(dest) }
+	src_ := unsafe { &byte(src) }
+	mut temp_buf := unsafe { malloc(int(n)) }
+	for i in 0 .. int(n) {
+		unsafe {
+			temp_buf[i] = src_[i]
+		}
+	}
+
+	for i in 0 .. int(n) {
+		unsafe {
+			dest_[i] = temp_buf[i]
+		}
+	}
+	unsafe { free(temp_buf) }
+	return unsafe { dest }
+}
+
+[export: 'calloc']
+[unsafe]
+fn __calloc(nmemb usize, size usize) &C.void {
+	new_area := unsafe { malloc(int(nmemb) * int(size)) }
+	unsafe { memset(new_area, 0, nmemb * size) }
+	return new_area
+}
+
+fn getchar() int {
+	return 0
+}
+
+fn memcmp(a &C.void, b &C.void, n usize) int {
+	a_ := unsafe { &byte(a) }
+	b_ := unsafe { &byte(b) }
+	for i in 0 .. int(n) {
+		if unsafe { a_[i] != b_[i] } {
+			unsafe {
+				return a_[i] - b_[i]
+			}
+		}
+	}
+	return 0
+}
+
+[export: 'free']
+[unsafe]
+fn __free(ptr &C.void) {
+	unsafe {
+		global_allocator.free_(ptr)
+	}
+}
+
+fn vsprintf(str &char, format &char, ap &byte) int {
+	panic('vsprintf(): string interpolation is not supported in `-freestanding`')
+}
+
+fn vsnprintf(str &char, size usize, format &char, ap &byte) int {
+	panic('vsnprintf(): string interpolation is not supported in `-freestanding`')
+}
+
+enum Errno {
+	enoerror
+	eerror
+}
+
+// not really needed
+fn bare_read(buf &byte, count u64) (i64, Errno) {
+	return 0, Errno.eerror
+}
+
+pub fn bare_print(buf &byte, len u64) {
+}
+
+fn bare_eprint(buf &byte, len u64) {
+}
+
+pub fn write(_fd i64, _buf &byte, _count u64) i64 {
+	return -1
+}
+
+[noreturn]
+fn bare_panic(msg string) {
+	println('V panic' + msg)
+	exit(1)
+}
+
+fn bare_backtrace() string {
+	return 'backtraces are not available with `-freestanding`'
+}
+
+[export: 'exit']
+[noreturn]
+fn __exit(code int) {
+	unsafe {
+		// the only way to abort process execution in WASM
+		mut x := &int(voidptr(0))
+		*x = code
+	}
+	for {}
+}
+
+[export: 'qsort']
+fn __qsort(base voidptr, nmemb usize, size usize, sort_cb FnSortCB) {
+	panic('qsort() is not yet implemented in `-freestanding`')
+}
+
+fn init_global_allocator() {
+	global_allocator = dlmalloc.new(get_wasm_allocator())
+}
diff --git a/vlib/builtin/wasm_bare/memory_management.v b/vlib/builtin/wasm_bare/memory_management.v
new file mode 100644
index 0000000000..65f400d2f3
--- /dev/null
+++ b/vlib/builtin/wasm_bare/memory_management.v
@@ -0,0 +1,68 @@
+module builtin
+
+import dlmalloc
+
+// Corresponding intrinsic to wasm’s `memory.grow` instruction
+//
+// This function, when called, will attempt to grow the default linear memory by the specified delta of pages.
+// The current WebAssembly page size is 65536 bytes (64 KB). If memory is successfully grown then the previous size of memory, in pages, is returned.
+// If memory cannot be grown then -1 is returned.
+//
+// The argument mem is the numerical index of which memory to return the size of. Note that currently the WebAssembly specification only supports one memory,
+// so it is required that zero is passed in. The argument is present to be forward-compatible with future WebAssembly revisions.
+// If a nonzero argument is passed to this function it will currently unconditionally abort
+fn C.__builtin_wasm_memory_grow(mem u32, delta usize) usize
+
+/// Corresponding intrinsic to wasm's `memory.size` instruction
+///
+/// This function, when called, will return the current memory size in units of
+/// pages. The current WebAssembly page size is 65536 bytes (64 KB).
+fn C.__builtin_wasm_memory_size(mem u32) usize
+
+const page_size = 65536
+
+fn system_alloc(_ voidptr, size usize) (voidptr, usize, u32) {
+	pages := size / page_size
+	prev := C.__builtin_wasm_memory_grow(0, pages)
+	if prev == -1 {
+		return voidptr(0), 0, 0
+	}
+	return voidptr(prev * page_size), pages * page_size, 0
+}
+
+fn system_remap(_ voidptr, _ voidptr, _ usize, _ usize, _ bool) voidptr {
+	return voidptr(0)
+}
+
+fn system_free_part(_ voidptr, _ voidptr, _ usize, _ usize) bool {
+	return false
+}
+
+fn system_free(_ voidptr, _ voidptr, _ usize) bool {
+	return false
+}
+
+fn system_allocates_zeros(_ voidptr) bool {
+	return false
+}
+
+fn system_page_size(_ voidptr) usize {
+	return page_size
+}
+
+fn system_can_release_part(_ voidptr, _ u32) bool {
+	return false
+}
+
+fn get_wasm_allocator() dlmalloc.Allocator {
+	return dlmalloc.Allocator{
+		alloc: system_alloc
+		remap: system_remap
+		free_part: system_free_part
+		free_: system_free
+		can_release_part: system_can_release_part
+		allocates_zeros: system_allocates_zeros
+		page_size: system_page_size
+		data: voidptr(0)
+	}
+}
diff --git a/vlib/dlmalloc/dlmalloc.v b/vlib/dlmalloc/dlmalloc.v
index eb8bda8f10..4fa2b7f9f7 100644
--- a/vlib/dlmalloc/dlmalloc.v
+++ b/vlib/dlmalloc/dlmalloc.v
@@ -16,6 +16,14 @@ module dlmalloc
 
 import math.bits
 
+$if debug ? {
+	#include "valgrind.h"
+}
+/*
+fn C.VALGRIND_MALLOCLIKE_BLOCK(addr voidptr, size usize, rzb usize,is_zeroed bool)
+fn C.VALGRIND_FREELIKE_BLOCK(addr voidptr, rzB usize)
+fn C.VALGRIND_MAKE_MEM_UNDEFINED(addr voidptr, size usize)
+*/
 pub const (
 	n_small_bins           = 32
 	n_tree_bins            = 32
@@ -144,7 +152,11 @@ fn small_index(size usize) u32 {
 }
 
 fn align_up(a usize, alignment usize) usize {
-	return (a + (alignment - 1)) & ~(alignment - 1)
+	if a % alignment == 0 {
+		return a
+	} else {
+		return a - (a % alignment) + alignment
+	}
 }
 
 fn left_bits(x u32) u32 {
@@ -467,7 +479,7 @@ fn (mut dl Dlmalloc) unlink_small_chunk(chunk_ &Chunk, size usize) {
 	mut b := chunk.next
 	idx := small_index(size)
 
-	if b == f {
+	if voidptr(b) == voidptr(f) {
 		unsafe { dl.clear_smallmap(idx) }
 	} else {
 		f.next = b
@@ -514,7 +526,7 @@ fn (mut dl Dlmalloc) unlink_large_chunk(chunk_ &TreeChunk) {
 		}
 
 		mut h := dl.treebin_at(chunk.index)
-		if chunk == *h {
+		if voidptr(chunk) == voidptr(*h) {
 			*h = r
 			if isnil(r) {
 				dl.clear_treemap(chunk.index)
@@ -549,7 +561,7 @@ fn (mut dl Dlmalloc) unlink_first_small_chunk(head_ &Chunk, next_ &Chunk, idx u3
 	mut head := head_
 
 	mut ptr := next.prev
-	if head == ptr {
+	if voidptr(head) == voidptr(ptr) {
 		unsafe { dl.clear_smallmap(idx) }
 	} else {
 		ptr.next = head
@@ -574,9 +586,11 @@ pub fn (mut dl Dlmalloc) calloc(size usize) voidptr {
 [unsafe]
 pub fn (mut dl Dlmalloc) free_(mem voidptr) {
 	unsafe {
+		// C.VALGRIND_FREELIKE_BLOCK(mem, 0)
 		mut p := chunk_from_mem(mem)
 
 		mut psize := p.size()
+
 		next := p.plus_offset(psize)
 
 		if !p.pinuse() {
@@ -589,6 +603,7 @@ pub fn (mut dl Dlmalloc) free_(mem voidptr) {
 				{
 					dl.footprint -= psize
 				}
+
 				return
 			}
 
@@ -600,6 +615,7 @@ pub fn (mut dl Dlmalloc) free_(mem voidptr) {
 			} else if (next.head & dlmalloc.inuse) == dlmalloc.inuse {
 				dl.dvsize = psize
 				p.set_free_with_pinuse(psize, next)
+
 				return
 			}
 		}
@@ -620,18 +636,21 @@ pub fn (mut dl Dlmalloc) free_(mem voidptr) {
 				if dl.should_trim(tsize) {
 					dl.sys_trim(0)
 				}
+
 				return
 			} else if voidptr(next) == voidptr(dl.dv) {
 				dl.dvsize += psize
 				dsize := dl.dvsize
 				dl.dv = p
 				p.set_size_and_pinuse_of_free_chunk(dsize)
+
 				return
 			} else {
 				nsize := next.size()
 				psize += nsize
 				dl.unlink_chunk(next, nsize)
 				p.set_size_and_pinuse_of_free_chunk(psize)
+
 				if voidptr(p) == voidptr(dl.dv) {
 					dl.dvsize = psize
 					return
@@ -892,11 +911,20 @@ fn (mut dl Dlmalloc) treemap_is_marked(idx u32) bool {
 	return dl.treemap & (1 << idx) != 0
 }
 
+pub fn (mut dl Dlmalloc) malloc(size usize) voidptr {
+	unsafe {
+		p := dl.malloc_real(size)
+		if !isnil(p) {
+			// C.VALGRIND_MALLOCLIKE_BLOCK(p, size, 0,false)
+		}
+		return p
+	}
+}
+
 /// malloc behaves as libc malloc, but operates within the given space
 [unsafe]
-pub fn (mut dl Dlmalloc) malloc(size usize) voidptr {
+fn (mut dl Dlmalloc) malloc_real(size usize) voidptr {
 	mut nb := usize(0)
-
 	unsafe {
 		if size <= max_small_request() {
 			nb = request_2_size(size)
@@ -907,10 +935,14 @@ pub fn (mut dl Dlmalloc) malloc(size usize) voidptr {
 
 				b := dl.smallbin_at(idx)
 				mut p := b.prev
-				dl.unlink_first_small_chunk(b, p, idx)
 				smallsize := small_index2size(idx)
+
+				dl.unlink_first_small_chunk(b, p, idx)
+
 				p.set_inuse_and_pinuse(smallsize)
+
 				ret := p.to_mem()
+
 				return ret
 			}
 
@@ -934,7 +966,9 @@ pub fn (mut dl Dlmalloc) malloc(size usize) voidptr {
 						r.set_size_and_pinuse_of_free_chunk(size)
 						dl.replace_dv(r, rsize)
 					}
+
 					ret := p.to_mem()
+
 					return ret
 				} else if dl.treemap != 0 {
 					mem := dl.tmalloc_small(nb)
@@ -972,8 +1006,9 @@ pub fn (mut dl Dlmalloc) malloc(size usize) voidptr {
 				dl.dv = voidptr(0)
 				p.set_inuse_and_pinuse(dvs)
 			}
+			ret := p.to_mem()
 
-			return p.to_mem()
+			return ret
 		}
 		// Split the top node if we can
 		if nb < dl.topsize {
@@ -984,8 +1019,9 @@ pub fn (mut dl Dlmalloc) malloc(size usize) voidptr {
 			mut r := dl.top
 			r.head = rsize | dlmalloc.pinuse
 			p.set_size_and_pinuse_of_inuse_chunk(nb)
+			ret := p.to_mem()
 
-			return p.to_mem()
+			return ret
 		}
 
 		return dl.sys_alloc(nb)
@@ -1011,18 +1047,22 @@ fn (mut dl Dlmalloc) init_top(ptr &Chunk, size_ usize) {
 	size := size_ - offset
 	dl.top = p
 	dl.topsize = size
+	// C.VALGRIND_MAKE_MEM_UNDEFINED(p.plus_offset(sizeof(usize)),sizeof(usize))
 	p.head = size | dlmalloc.pinuse
-
+	// C.VALGRIND_MAKE_MEM_UNDEFINED(p.plus_offset(size + sizeof(usize)),sizeof(usize))
 	p.plus_offset(size).head = top_foot_size()
 	dl.trim_check = u32(default_trim_threshold())
 }
 
 [unsafe]
 fn (mut dl Dlmalloc) sys_alloc(size usize) voidptr {
-	asize := align_up(size + top_foot_size() + malloc_alignment(), default_granularity())
+	page_size := dl.system_allocator.page_size(dl.system_allocator.data)
+	asize := align_up(align_up(size + top_foot_size() + malloc_alignment(), default_granularity()),
+		page_size)
 	unsafe {
-		tbase, mut tsize, flags := dl.system_allocator.alloc(dl.system_allocator.data,
-			asize)
+		alloc := dl.system_allocator.alloc
+		tbase, mut tsize, flags := alloc(dl.system_allocator.data, asize)
+
 		if isnil(tbase) {
 			return tbase
 		}
@@ -1085,7 +1125,9 @@ fn (mut dl Dlmalloc) sys_alloc(size usize) voidptr {
 			mut r := dl.top
 			r.head = rsize | dlmalloc.pinuse
 			p.set_size_and_pinuse_of_inuse_chunk(size)
-			return p.to_mem()
+			ret := p.to_mem()
+
+			return ret
 		}
 	}
 	return voidptr(0)
@@ -1122,6 +1164,7 @@ fn (mut dl Dlmalloc) tmalloc_small(size usize) voidptr {
 			rc.set_size_and_pinuse_of_free_chunk(rsize)
 			dl.replace_dv(rc, rsize)
 		}
+
 		return vc.to_mem()
 	}
 }
@@ -1382,6 +1425,8 @@ pub fn (mut dl Dlmalloc) memalign(alignment_ usize, bytes usize) voidptr {
 				dl.dispose_chunk(remainder, remainder_size)
 			}
 		}
+
+		// C.VALGRIND_MALLOCLIKE_BLOCK(p.to_mem(), bytes, 0, false)
 		return p.to_mem()
 	}
 }