| commit | 10159285d704540c48c6896082ac67375fcfc230 | |
| author | Dan Carpenter <error27@gmail.com> | |
| Wed, 28 Apr 2010 15:45:10 +0000 (28 17:45 +0200) | ||
| committer | Dan Carpenter <error27@gmail.com> | |
| Wed, 28 Apr 2010 15:45:10 +0000 (28 17:45 +0200) | ||
| tree | daeaf4e0282be6842e2322746805ee00453b4d2c | treesnapshot (tar.gz zip) |
| parent | 1a20b562d3bbc5490adf4e92c12f0799f10dcccf | commitdiff |
delete check_null_deref.c and clean up validation/
check_deref.c and check_check_deref.c basically reimpliment
check_null_deref.c
One thing that isn't they don't handle is cross function dereferences. But
those results were generally bogus so it's no loss. I plan to add support
for this again when Smatch gets proper cross function analysis.
The other thing that I noticed from looking through validation/sm_select.c
is that in check_null_deref.c if you dereference something with a state of
merged &ok and &undefined then it would complain. check_check_deref.c
doesn't do this.
Anyway, it feels really good to delete check_null_deref.c.
Signed-off-by: Dan Carpenter <error27@gmail.com>
check_deref.c and check_check_deref.c basically reimpliment
check_null_deref.c
One thing that isn't they don't handle is cross function dereferences. But
those results were generally bogus so it's no loss. I plan to add support
for this again when Smatch gets proper cross function analysis.
The other thing that I noticed from looking through validation/sm_select.c
is that in check_null_deref.c if you dereference something with a state of
merged &ok and &undefined then it would complain. check_check_deref.c
doesn't do this.
Anyway, it feels really good to delete check_null_deref.c.
Signed-off-by: Dan Carpenter <error27@gmail.com>