_chunker.c: fix some bugs

- better check return value of fd.read(n) and reject if it returns more bytes than requested.
- avoid giving len<=0 to posix_fadvise(), which could drop the rest of the file from cache.
- buzhash: check for len == 0 edge case
- correctly Py_DECREF in cases of errors
- check for malloc/calloc failures

src/borg/_chunker.c

@@ -87,6 +87,9 @@ buzhash(const unsigned char *data, size_t len, const uint32_t *h)
 {
     uint32_t i;
     uint32_t sum = 0, imod;
+    if (len == 0) {
+        return 0;
+    }
     for(i = len - 1; i > 0; i--)
     {
         imod = i & 0x1f;
@@ -118,12 +121,24 @@ static Chunker *
 chunker_init(size_t window_size, uint32_t chunk_mask, size_t min_size, size_t max_size, uint32_t seed)
 {
     Chunker *c = calloc(sizeof(Chunker), 1);
+    if(!c) {
+        return NULL;
+    }
     c->window_size = window_size;
     c->chunk_mask = chunk_mask;
     c->min_size = min_size;
     c->table = buzhash_init_table(seed);
+    if(!c->table) {
+        free(c);
+        return NULL;
+    }
     c->buf_size = max_size;
     c->data = malloc(c->buf_size);
+    if(!c->data) {
+        free(c->table);
+        free(c);
+        return NULL;
+    }
     c->fh = -1;
     return c;
 }
@@ -219,7 +234,9 @@ chunker_fill(Chunker *c)
             overshoot = 0;
         }
 
-        posix_fadvise(c->fh, offset & ~pagemask, length - overshoot, POSIX_FADV_DONTNEED);
+        if (length - overshoot > 0 || length == 0) {
+            posix_fadvise(c->fh, offset & ~pagemask, length - overshoot, POSIX_FADV_DONTNEED);
+        }
         #endif
 
         PyEval_RestoreThread(thread_state);
@@ -230,15 +247,21 @@ chunker_fill(Chunker *c)
         if(!data) {
             return 0;
         }
-        n = PyBytes_Size(data);
+        ssize_t read_bytes = PyBytes_Size(data);
         if(PyErr_Occurred()) {
             // we wanted bytes(), but got something else
+            Py_DECREF(data);
             return 0;
         }
-        if(n) {
-            memcpy(c->data + c->position + c->remaining, PyBytes_AsString(data), n);
-            c->remaining += n;
-            c->bytes_read += n;
+        if(read_bytes > n) {
+            Py_DECREF(data);
+            PyErr_SetString(PyExc_ValueError, "read() returned too many bytes");
+            return 0;
+        }
+        if(read_bytes) {
+            memcpy(c->data + c->position + c->remaining, PyBytes_AsString(data), read_bytes);
+            c->remaining += read_bytes;
+            c->bytes_read += read_bytes;
         }
         else {
             c->eof = 1;

src/borg/chunker.pyx

@@ -247,6 +247,8 @@ cdef class Chunker:
         assert hash_window_size + min_size + 1 <= max_size, "too small max_size"
         hash_mask = (1 << hash_mask_bits) - 1
         self.chunker = chunker_init(hash_window_size, hash_mask, min_size, max_size, seed & 0xffffffff)
+        if not self.chunker:
+            raise MemoryError('chunker_init failed')
 
     def chunkify(self, fd, fh=-1):
         """