liblzma: Avoid memcpy(NULL, foo, 0) because it is undefined behavior.

I should have always known this but I didn't. Here is an example
as a reminder to myself:

    int mycopy(void *dest, void *src, size_t n)
    {
        memcpy(dest, src, n);
        return dest == NULL;
    }

In the example, a compiler may assume that dest != NULL because
passing NULL to memcpy() would be undefined behavior. Testing
with GCC 8.2.1, mycopy(NULL, NULL, 0) returns 1 with -O0 and -O1.
With -O2 the return value is 0 because the compiler infers that
dest cannot be NULL because it was already used with memcpy()
and thus the test for NULL gets optimized out.

In liblzma, if a null-pointer was passed to memcpy(), there were
no checks for NULL *after* the memcpy() call, so I cautiously
suspect that it shouldn't have caused bad behavior in practice,
but it's hard to be sure, and the problematic cases had to be
fixed anyway.

Thanks to Jeffrey Walton.

3 files changed, 23 insertions, 5 deletions

diff --git a/src/liblzma/common/common.c b/src/liblzma/common/common.c
index 57e3f8eb..0053a6aa 100644
--- a/src/liblzma/common/common.c
+++ b/src/liblzma/common/common.c
@@ -99,7 +99,11 @@ lzma_bufcpy(const uint8_t *restrict in, size_t *restrict in_pos,
 	const size_t out_avail = out_size - *out_pos;
 	const size_t copy_size = my_min(in_avail, out_avail);
 
-	memcpy(out + *out_pos, in + *in_pos, copy_size);
+	// Call memcpy() only if there is something to copy. If there is
+	// nothing to copy, in or out might be NULL and then the memcpy()
+	// call would trigger undefined behavior.
+	if (copy_size > 0)
+		memcpy(out + *out_pos, in + *in_pos, copy_size);
 
 	*in_pos += copy_size;
 	*out_pos += copy_size;
diff --git a/src/liblzma/lz/lz_decoder.c b/src/liblzma/lz/lz_decoder.c
index bb21d0d0..6c9024e2 100644
--- a/src/liblzma/lz/lz_decoder.c
+++ b/src/liblzma/lz/lz_decoder.c
@@ -91,11 +91,17 @@ decode_buffer(lzma_coder *coder,
 				in, in_pos, in_size);
 
 		// Copy the decoded data from the dictionary to the out[]
-		// buffer.
+		// buffer. Do it conditionally because out can be NULL
+		// (in which case copy_size is always 0). Calling memcpy()
+		// with a null-pointer is undefined even if the third
+		// argument is 0.
 		const size_t copy_size = coder->dict.pos - dict_start;
 		assert(copy_size <= out_size - *out_pos);
-		memcpy(out + *out_pos, coder->dict.buf + dict_start,
-				copy_size);
+
+		if (copy_size > 0)
+			memcpy(out + *out_pos, coder->dict.buf + dict_start,
+					copy_size);
+
 		*out_pos += copy_size;
 
 		// Reset the dictionary if so requested by coder->lz.code().
diff --git a/src/liblzma/simple/simple_coder.c b/src/liblzma/simple/simple_coder.c
index 13ebabc7..4f499bef 100644
--- a/src/liblzma/simple/simple_coder.c
+++ b/src/liblzma/simple/simple_coder.c
@@ -118,7 +118,15 @@ simple_code(void *coder_ptr, const lzma_allocator *allocator,
 		// coder->pos and coder->size yet. This way the coder can be
 		// restarted if the next filter in the chain returns e.g.
 		// LZMA_MEM_ERROR.
-		memcpy(out + *out_pos, coder->buffer + coder->pos, buf_avail);
+		//
+		// Do the memcpy() conditionally because out can be NULL
+		// (in which case buf_avail is always 0). Calling memcpy()
+		// with a null-pointer is undefined even if the third
+		// argument is 0.
+		if (buf_avail > 0)
+			memcpy(out + *out_pos, coder->buffer + coder->pos,
+					buf_avail);
+
 		*out_pos += buf_avail;
 
 		// Copy/Encode/Decode more data to out[].