| commit | 69136cd09bfb379fc7bed31165fb6db077c9a00e | |
| author | Dan Carpenter <dan.carpenter@oracle.com> | |
| Mon, 18 Jun 2012 08:05:55 +0000 (18 11:05 +0300) | ||
| committer | Dan Carpenter <dan.carpenter@oracle.com> | |
| Mon, 18 Jun 2012 08:05:55 +0000 (18 11:05 +0300) | ||
| tree | 8c795df4ef6abc85d0722f806cdddabd6466e103 | treesnapshot (tar.gz zip) |
| parent | 87e1db881a8f3e8e8a520fc107ca276e6066c277 | commitdiff |
buf_size: remove the &* and [0] from around the pointer earlier
The situation is that &p[0] is the same as 'p'. We handled this already
for regular arrays, but not if the array was dynamically allocated. So
I've moved it forward.
I had to make get_real_array_size() ignore EXPR_BINOP because now it
has to handle &array[5] which is not the same size as &array[0]. It's not
clear if we should return the size of the element or the size of the the
array minus the first 5 elements, but for now we just call it unknown.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
The situation is that &p[0] is the same as 'p'. We handled this already
for regular arrays, but not if the array was dynamically allocated. So
I've moved it forward.
I had to make get_real_array_size() ignore EXPR_BINOP because now it
has to handle &array[5] which is not the same size as &array[0]. It's not
clear if we should return the size of the element or the size of the the
array minus the first 5 elements, but for now we just call it unknown.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
| smatch_buf_size.c | diffblobblamehistory | |
| validation/sm_buf_size6.c | [new file with mode: 0644] | blob |