| commit | 4eff961d813a5a087ece3bf0a728a23f50337a2a | |
| author | Dan Carpenter <error27@gmail.com> | |
| Sat, 21 Mar 2009 11:19:27 +0000 (21 14:19 +0300) | ||
| committer | Dan Carpenter <error27@gmail.com> | |
| Sat, 21 Mar 2009 11:19:27 +0000 (21 14:19 +0300) | ||
| tree | 8569d3ffd611dc948560be84a806b63abc314f21 | treesnapshot (tar.gz zip) |
| parent | e1ff92e5b6230c4529d2f7c5728cda5552dc91a7 | commitdiff |
core: improve anding condition handling.
The handling of merged states was obviously wrong before and so this is
an improvement.
It's a bit tricky because sometimes mutually exclusive states are
legit.
if (!foo && !(foo = kmalloc())) {
Sometimes when it complains about mutually exclusive states that
really does mean a condition is always false. We set a lot of states
all the time because of implications. I may remove the
"mutually exclusive" print statement again.
Signed-off-by: Dan Carpenter <error27@gmail.com>
The handling of merged states was obviously wrong before and so this is
an improvement.
It's a bit tricky because sometimes mutually exclusive states are
legit.
if (!foo && !(foo = kmalloc())) {
Sometimes when it complains about mutually exclusive states that
really does mean a condition is always false. We set a lot of states
all the time because of implications. I may remove the
"mutually exclusive" print statement again.
Signed-off-by: Dan Carpenter <error27@gmail.com>
| smatch_implied.c | diffblobblamehistory | |
| smatch_slist.c | diffblobblamehistory |