| commit | f7e183021293b4b457a0f706918b16728a0f22f6 | |
| author | Dan Carpenter <dan.carpenter@oracle.com> | |
| Thu, 20 Jun 2013 09:25:12 +0000 (20 12:25 +0300) | ||
| committer | Dan Carpenter <dan.carpenter@oracle.com> | |
| Thu, 20 Jun 2013 09:25:12 +0000 (20 12:25 +0300) | ||
| tree | 9f08d03c6d4ff347787a0aa77220535a347c4e98 | treesnapshot (tar.gz zip) |
| parent | d8b54f88fe53d937abc43c897d0c478553b63f3a | commitdiff |
math: fix get_fuzzy_max()
The problem here is if you have something like
val = 0-u32max (0-u32max, 4-6)
Then the fuzzy min would be 6 and the fuzzy max would be 4 when we call
alloc_rl() it says that that's inverted so it allocates a whole range.
Fuzzy min/max is weird that way. We need to flip them the right way around
before calling alloc_rl().
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
The problem here is if you have something like
val = 0-u32max (0-u32max, 4-6)
Then the fuzzy min would be 6 and the fuzzy max would be 4 when we call
alloc_rl() it says that that's inverted so it allocates a whole range.
Fuzzy min/max is weird that way. We need to flip them the right way around
before calling alloc_rl().
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
| smatch_math.c | diffblobblamehistory |