| commit | 0fa6db5149c6757f1a4613931fb84bb446733cd3 | |
| author | Dan Carpenter <dan.carpenter@oracle.com> | |
| Mon, 24 Nov 2014 12:39:25 +0000 (24 15:39 +0300) | ||
| committer | Dan Carpenter <dan.carpenter@oracle.com> | |
| Mon, 24 Nov 2014 12:39:25 +0000 (24 15:39 +0300) | ||
| tree | cea491e26760db551b1d6ee8264337601b51b4da | treesnapshot (tar.gz zip) |
| parent | 8de4a23cdbecf18af45506f3cd177e9acf111fd8 | commitdiff |
type_val: improve the type/value information in the database
Smatch tries to record the possible values of "(struct whatever)->foo" in
the database. The problem is that sometimes we store
"(struct whatever)->foo" into some other "void *bar". This makes a link
between ->foo and bar so that if we get a ->foo out of bar, then we're
going to assume it's the same thing that we put there.
This works ok in testing.
Also I'm putting "ignored" in the function_type_value table just so I can
see which things are affected and to help debug any problems.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Smatch tries to record the possible values of "(struct whatever)->foo" in
the database. The problem is that sometimes we store
"(struct whatever)->foo" into some other "void *bar". This makes a link
between ->foo and bar so that if we get a ->foo out of bar, then we're
going to assume it's the same thing that we put there.
This works ok in testing.
Also I'm putting "ignored" in the function_type_value table just so I can
see which things are affected and to help debug any problems.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>