repo.or.cz
/
smatch.git
/
search
commit
grep
author
committer
pickaxe
?
search:
re
summary
|
log
|
graphiclog1
|
graphiclog2
|
commit
|
commitdiff
|
tree
|
refs
|
edit
|
fork
first
·
prev
·
next
Merge branch 'master' into null_deref
2009-02-17
Dan Carp
e
n
t
e
r
Merge branch '
m
as
t
e
r'
i
nto null_deref
commit
|
commitdiff
|
tree
2009-02-17
D
an Carpenter
Add validation
script for check_null
_
deref
.
c
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-16
Dan Ca
r
pen
t
er
Clean up
.
new->possibl
e
i
s alway
s
null here
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-14
Dan Carp
e
nter
Oops
.
->possibles wasn't getting set correctly
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-14
Dan Carpenter
Use
strip_e
x
p
r
() in sp
l
it_conditions()
i
nstead rep
r
od
u
cing
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-14
D
a
n
C
arpenter
Ad
d
some m
o
re
o
perations to get_value()
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-13
Dan Carpenter
Return &
m
e
r
ged
i
n
s
tead
of
&
undefined
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-13
Da
n
C
a
rpenter
m
a
ke get_value()
r
eturn a va
l
u
e for
b
i
t
w
ise or expr
e
s
s
i
o
n
s
.
commit
|
commitdiff
|
tree
2009-02-13
Dan Carpenter
Try a
g
ain re
m
ovi
n
g state
s
from pools
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-13
Dan Carpenter
R
evert "Remove a state from it
'
s
-
>pools when set_state
.
.
.
commit
|
commitdiff
|
tree
2009-02-11
Dan Carpenter
Eli
m
inate
a
warnin
g
.
Probably the wron
g
way to
d
o
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-11
Dan
C
arpenter
Clean
ups
.
Comme
n
ts
.
Rename slist_stack
t
o named_stack
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-11
Dan Carpenter
Ha ha
.
I still s
u
ck at pointers
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-11
Dan Carpent
e
r
C
leanups
.
Comme
n
ts
.
Remove
debu
g
g
i
n
g code from filter()
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-11
Da
n
Carpenter
Clean up
.
Also preserve th
e
->poo
l
s in one
p
ath of
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-11
Dan
C
ar
p
enter
Clean up get_state_slist()
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-11
Dan Carpenter
Rem
o
ve _
_
o
v
erwri
t
e_cur_slist()
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-11
Dan Carp
e
nter
Th
e
re
was
d
ebugging outpu
t
that ac
c
identally g
o
t c
o
mmitt
e
d
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-11
Dan Carpenter
Remove
a
s
t
ate fro
m
it's ->poo
l
s when set_state
(
)
is
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-11
Dan Carpenter
D
on'
t
d
elete the
->pools on merging
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-11
Da
n
C
ar
p
e
nt
e
r
Try understand
the implic
a
tions fro
m
compound conditions
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-11
Dan Carpenter
Create a state &
m
erg
e
d
.
When we merge stuff it
isn
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-10
Dan Carpe
n
t
er
Ok
.
This time smatch_implied _real
l
y_ do
e
s some
t
hing
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-10
Dan Carp
e
nter
Ignore cscope files
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-10
Dan Carpenter
Small cleanups
.
(Wh
i
te sp
a
ce
)
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-10
Dan Carpenter
Whi
t
espa
c
e cha
n
ge
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-10
Dan Carp
e
nter
Add some more
loc
k
i
ng functions
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-10
Dan
C
arpenter
Fix implied poo
l
s
so tha
t
at least it
d
oe
s
something
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-10
D
a
n
Carpent
e
r
It se
e
m
e
d l
i
ke tra
c
ker_list would be useful in a lot
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-05
Dan Car
p
enter
Add a README describing smatch
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-05
Dan C
a
rpenter
Add
s
cript to fi
n
d NUL
L
dereferences between
functions
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-05
Dan C
a
rp
e
n
ter
F
ix usage message
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-05
Dan Carpente
r
Check for ma
l
loc() as
well
a
s kmalloc
(
)
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-05
Dan Carpenter
Fix segfault wh
e
n handl
i
ng implic
i
tly declared
functio
n
s
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-04
Da
n
C
a
rpenter
Warn
w
he
n
we don't acquire lock
s
w
h
e
n
c
alling
c
ertain
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-04
D
a
n Carp
e
nter
T
r
y (and f
a
il)
to f
i
nd memory l
e
aks
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-03
Da
n
Carpenter
A
lmost print a mess
a
ge when
y
ou come to unrea
c
habl
e
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-03
Dan Carpenter
Add check to look
for double
f
rees
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-03
Dan Carpenter
Add checkin
g
f
o
r lock_ker
n
el()/unlock_kernel()
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-02
Dan C
a
rp
e
nter
Make
c
he
c
k_locki
n
g c
h
ec
k
both spinlo
c
ks and semaphores
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-02
Dan Car
p
enter
Add a n
e
w
function slist_has_
s
tat
e
()
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-02
Dan Carpenter
I
f there is a
n
i++ th
e
n th
e
value
o
f 'i
'
is un
d
e
f
ined
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-02
Dan Carpenter
A
d
d
a new ch
e
c
k
_locking
s
c
ript that
chec
k
s spinlocks
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-02-02
Dan Car
p
enter
Save
t
he poss
i
b
l
e states
i
nstead of just mer
g
ing them
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-01-28
D
a
n
Carpe
n
te
r
Ignore ema
c
s garbage
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-01-28
Dan
C
a
r
p
enter
Merge br
a
nch 'ma
s
te
r
' into pools5
commit
|
commitdiff
|
tree
2009-01-20
Dan Carp
e
nter
Fix
l
i
kely(
)
and u
n
likely() han
d
ling
.
Signed-off-by:
Dan Carpenter
<dcarpenter@bikeee.(none)>
commit
|
commitdiff
|
tree
2009-01-20
D
a
n
Ca
r
penter
Add
s
ingle quotes ar
o
und t
h
e
va
r
iable for con
s
is
t
en
c
y
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2009-01-16
Dan Carpenter
We coul
d
save some memory by not
allo
c
ati
n
g new m
e
mor
y
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-12-21
Dan C
a
rpenter
A
d
d a c
o
mment showi
n
g where
t
he
tests ar
e
supposed
.
.
.
commit
|
commitdiff
|
tree
2008-12-21
Dan Carpenter
This is left ove
r
from October as
part
o
f the implied
.
.
.
commit
|
commitdiff
|
tree
2008-12-21
Dan Car
p
ente
r
A
dd get_possible_states()
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-12-21
Dan Carpenter
S
tuff
left over from October
.
Implied
p
o
ols work
.
commit
|
commitdiff
|
tree
2008-10-19
Dan Carpenter
We
c
ould save s
o
m
e memo
r
y
b
y not
allocat
i
ng ne
w
memory
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-18
Dan Carpen
t
er
Fix memory lea
k
.
Add more
c
omments
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-17
Dan Car
p
ente
r
Add a merge
hook
.
Handle
the situ
a
tions lik
e
:
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-17
D
an Carp
e
nter
s
trcmp
(
)
d
o
e
sn't acce
p
t
NULL
a
r
gumen
t
s
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-17
D
an Carpenter
rename t
w
o
functi
o
ns: get_variable_f
r
om_expr_si
m
ple
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-17
D
an Carpente
r
get_
v
ariable_
f
rom_expr() uses a
s
tatic buffer and you
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-17
Dan
Carpenter
Cre
a
t
e a c
h
eck_
t
emplate
.
c ex
a
mpl
e
s
c
ri
p
t that look
s
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-15
Dan Carpent
e
r
Make
get_all
_
states() not segfault ever
y
time
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-15
Dan
Carpen
t
er
Fix get_value() to return n
e
gativ
e
v
a
l
ues
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-15
Da
n
Carpe
n
t
er
We onl
y
ever pass one parameter into get_value()
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-15
D
a
n Carpe
n
te
r
Fix get_all_states(
)
to match t
h
e functio
n
d
e
fi
n
iti
o
n
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-15
Dan Carpenter
Remove add_s
t
ate() and state_de
f
ined() definitio
n
s
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-14
D
a
n
Carpe
n
t
e
r
Clean up
.
Fix
le
a
k
.
My problem is th
a
t I do
n
't understand
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-14
Dan Carpenter
Initial
com
m
it of the "implied state poo
l
s" stu
f
f
.
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-14
Dan Ca
r
penter
Doh
.
We haven't been using smatc
h
_
e
xt
r
a fo
r
a
while
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-13
Dan Ca
r
penter
G
et rid o
f
path_history
.
It's nev
e
r go
i
ng
to
be u
s
ed
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-13
Dan
Carpenter
Make go
t
o_stac
k
static
.
commit
|
commitdiff
|
tree
2008-10-13
Dan
Carpente
r
Speed up merge_slis
t
n
ow that the slist
s
are
o
rde
r
ed
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-11
Dan Carpent
e
r
Insert stuff o
n
to lists in a
s
ort
e
d orde
r
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-11
Dan
Carpenter
Dele
t
e an u
n
used function m
e
rge_stat
e
_
stac
k
()
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-08
Dan Ca
r
penter
fix: if (
*
foo) { (*foo)->bar
;
}
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-07
Dan Car
p
enter
kmalloc() and friends do
n
't always return
a
n
on-null
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-06
D
a
n Carpe
n
t
er
The
c
h
ecks from her
e
got merg
e
d
into
che
c
k_null_d
e
ref
.
c
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-04
Dan C
a
rpenter
Mov
e
the st
u
ff fr
o
m chec
k
_derefed_p
a
rams
.
c
i
nto check_null_d
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-10-04
Dan Car
p
enter
Fix bug in or_slis
t
_stack()
.
if (
a
&& b) w
a
sn't
b
ein
g
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-30
Dan
C
arpent
e
r
A
dd a hook at the end of
a file
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-28
Dan Ca
r
penter
Forgot another strip_expr()
call
.
.
.
:
/
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-28
Dan Carpe
n
ter
H
andle assi
g
n
m
eents inside conditions
b
ett
e
r
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-28
D
a
n Car
p
enter
O
n
l
y
print some
t
hing for a->foo and
not
a
.
foo
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-25
D
an Carpenter
Ignore array d
e
claration
s
.
Those can't be NULL obviously
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-23
Da
n
Carpenter
Make cur_slist s
t
atic
.
It's only used in sma
t
ch_states
.
c
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-22
Dan Carpenter
White space cha
n
ges
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-22
D
an Carpenter
Use me
r
ge_sl
i
st()
i
nst
e
ad
of trying to handle it yoursel
f
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-22
Dan Ca
r
pente
r
__proc
e
ss
_
con
t
inu
e
s(
)
is
s
uppo
s
ed to
b
e basi
c
ally the
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-21
Dan Carpenter
New s
m
atch
_
ignore
.
c file
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-21
Da
n
Carpenter
(fo
o
)
= 1 is t
h
e same as foo = 1
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-21
Da
n
Ca
r
penter
Generalize st
i
p_
c
asts()
.
It's no
w
cal
l
ed strip_expr
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-20
Dan
Carpe
n
ter
Ign
o
re casts
.
fo
o
(
(int *)&bar) is t
h
e same
a
s foo
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-20
Dan Car
p
en
t
e
r
Created
s
how
_
st
a
t
e()
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-20
Dan Carpenter
I
gnor
e
s
tuff l
i
ke foo
.
b
ar
.
We o
n
ly care ab
o
ut foo
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-20
Dan Car
p
enter
What happens befor
e
is that par
a
meter
s
would get assig
n
ed
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-19
Dan Carpente
r
Missed a crashing bug in the int => struct conv
e
rsio
n
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-19
D
an C
a
rpent
e
r
I can't believe
I was doing stacks wrong this whole
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-19
Dan Carpe
n
t
er
C
h
anged smatch state fr
o
m being an int to bei
n
g point
e
r
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-17
Da
n
C
a
rpente
r
I want to re
u
se the name smatch_state
.
The current
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-17
Dan Carp
e
n
t
e
r
R
e
d
o
t
he
null path t
h
ing
.
S
o
long as ther
e
is one
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
2008-09-17
Dan Carpenter
Clean up
.
Com
p
licated i
s
just a
b
o
o
l
e
a
n
.
Set
it to
.
.
.
Signed-off-by:
Dan Carpenter
<error27@gmail.com>
commit
|
commitdiff
|
tree
next