Fix alignment issues and build on SPARC

nbd-tester-client, in the "integrity" test, was making some assumptions
that were incorrect on architectures such as SPARC which care about
alignment. This patch fixes those, by making sure we always access
64-bit values through a uint64_t, rather than through a char.

diff --git a/nbd-tester-client.c b/nbd-tester-client.c
index 26d1fd2..b503553 100644 (file)
--- a/nbd-tester-client.c
+++ b/nbd-tester-client.c
@@ -714,8 +714,8 @@ static inline void makebuf(char *buf, uint64_t seq, uint64_t blknum) {
 }
                
 static inline int checkbuf(char *buf, uint64_t seq, uint64_t blknum) {
-       char cmp[512];
-       makebuf(cmp, seq, blknum);
+       uint64_t cmp[64]; // 512/8 = 64
+       makebuf((char *)cmp, seq, blknum);
        return memcmp(cmp, buf, 512)?-1:0;
 }
 
@@ -1100,13 +1100,15 @@ int integrity_test(gchar* hostname, int port, char* name, int sock,
                                goto err_open;
                        }
                                
-                       prc = g_hash_table_lookup(handlehash, rep.handle);
+                       uint64_t handle;
+                       memcpy(&handle,rep.handle,8);
+                       prc = g_hash_table_lookup(handlehash, &handle);
                        if (!prc) {
                                retval=-1;
                                snprintf(errstr, errstr_len, "Unrecognised handle in reply: 0x%llX", *(long long unsigned int*)(rep.handle));
                                goto err_open;
                        }
-                       if (!g_hash_table_remove(handlehash, rep.handle)) {
+                       if (!g_hash_table_remove(handlehash, &handle)) {
                                retval=-1;
                                snprintf(errstr, errstr_len, "Could not remove handle from hash: 0x%llX", *(long long unsigned int*)(rep.handle));
                                goto err_open;