repo.or.cz
/
official-gcc.git
/
search
commit
grep
author
committer
pickaxe
?
search:
re
summary
|
log
|
graphiclog1
|
graphiclog2
|
commit
|
commitdiff
|
tree
|
refs
|
edit
|
fork
first
·
prev
·
next
PR71951: Fix unwinding with -fomit-frame-pointer
2017-09-21
wilco
PR
7
1951: Fix unwinding with -
f
omit-frame-pointe
r
commit
|
commitdiff
|
tree
2017-09-13
wilco
Upd
a
te aarch
6
4/vmov_n_1
.
c
commit
|
commitdiff
|
tree
2017-09-12
w
i
lco
Add
DW_CFA_AARCH
6
4_negate_
r
a_state to dwarf2
.
d
ef/
h
.
.
.
commit
|
commitdiff
|
tree
2017-09-12
wil
c
o
Rem
o
ve '*' from mo
v
si/di/ti p
a
tterns
commit
|
commitdiff
|
tree
2017-09-06
wilco
PR78468
-
add alloca
a
l
ignment t
e
st
commit
|
commitdiff
|
tree
2017-09-05
w
i
l
c
o
Improve max_insns_skipped logic
commit
|
commitdiff
|
tree
2017-09-05
wilco
Improv
e
alloca al
i
gn
m
ent
commit
|
commitdiff
|
tree
2017-09-04
w
i
lco
F
ix
l
drd offset
s
commit
|
commitdiff
|
tree
2017-08-21
wilco
This
p
atch
s
implifie
s
pow
(
C, x)
i
nto exp (x * C1)
.
.
.
commit
|
commitdiff
|
tree
2017-08-14
wilc
o
[
A
Arch64
]
Fix
l
o
ngbranch tes
t
commit
|
commitdiff
|
tree
2017-08-14
wilco
Add c
h
eck_effective_targ
e
t_a
u
toincde
c
.
commit
|
commitdiff
|
tree
2017-08-07
w
i
lco
[
AA
r
ch
6
4] Use gen_frame_mem for c
a
llee-saves
commit
|
commitdiff
|
tree
2017-08-07
wilco
Fix un
r
esolved in gcc
.
dg/pr4
6
9
32
.
c
commit
|
commitdiff
|
tree
2017-07-26
wilc
o
Dis
a
ble p
r
79041-2
.
c with -mabi=ilp
3
2
.
commit
|
commitdiff
|
tree
2017-07-26
w
ilco
Fix P
R
46932: Block
a
uto inc
r
ement
on fr
a
me
p
ointer
commit
|
commitdiff
|
tree
2017-07-25
wilco
Fix PR7904
1
commit
|
commitdiff
|
tree
2017-07-24
wilco
Fix PR79041
commit
|
commitdiff
|
tree
2017-06-28
wilco
This
patch fix
e
s a failure in gcc
.
target/aarch64/r
e
load
.
.
.
commit
|
commitdiff
|
tree
2017-06-28
w
ilco
Improve Cortex-A53
s
hift bypa
s
s
commit
|
commitdiff
|
tree
2017-06-21
wilco
Emit
S
IMD moves as
mov
commit
|
commitdiff
|
tree
2017-06-21
wi
l
co
Imp
r
ove d
u
p pattern
commit
|
commitdiff
|
tree
2017-06-21
wi
l
co
Mark symbols as constant
commit
|
commitdiff
|
tree
2017-06-14
wilco
Im
p
rove Cortex-
A
53 F
P
s
c
h
e
d
ul
e
r
commit
|
commitdiff
|
tree
2017-05-30
w
il
c
o
Fix ARM bootstrap failure due
t
o an
o
dd w
a
rning:
commit
|
commitdiff
|
tree
2017-05-24
wilco
When
lr
a
-
r
emat rematerializes an
i
nstruction w
i
th a
.
.
.
commit
|
commitdiff
|
tree
2017-05-18
wilco
Add an assert to le
a
f_function_p to ens
u
re
it
i
s
n
ot
.
.
.
commit
|
commitdiff
|
tree
2017-05-10
wilco
Move an use-after
-
free ac
c
es
s
before
t
he dele
t
e
.
commit
|
commitdiff
|
tree
2017-05-08
wilco
This fi
x
es a few failur
e
s on ARM and AArch6
4
du
e
to
.
.
.
commit
|
commitdiff
|
tree
2017-05-05
wilco
Float to int
m
o
v
es
c
urr
e
ntly generate inef
f
ic
i
ent code
.
.
.
commit
|
commitdiff
|
tree
2017-05-05
wil
c
o
Code scheduling
for Cortex-A53 isn't as good as it
.
.
.
commit
|
commitdiff
|
tree
2017-05-04
wilco
Many sup
p
orted co
r
es use t
h
e
AUTOP
R
EFETCHE
R
_WEAK setting
.
.
.
commit
|
commitdiff
|
tree
2017-05-04
wilco
Set jump alignment to 4 for Corte
x
cores as it reduce
s
.
.
.
commit
|
commitdiff
|
tree
2017-05-04
wilco
Wi
t
h -mcpu=generic the loop alignment is
currently
.
.
.
commit
|
commitdiff
|
tree
2017-05-04
wilco
All c
o
res which ad
d
a
cpu_addrcost_tabl
e
us
e
a non
.
.
.
commit
|
commitdiff
|
tree
2017-03-22
w
ilco
R
e
cent
l
y we've put a lot of effort into improvin
g
ifcvt
.
.
.
commit
|
commitdiff
|
tree
2017-03-22
wilco
Many
sup
p
orted
c
ores
i
mple
m
ent fusio
n
of AE
S
i
nstructions
.
.
.
commit
|
commitdiff
|
tree
2017-01-18
w
i
l
c
o
S
H
A
1
H instructi
o
ns
m
ay b
e
scheduled after a SHA1C instruction
commit
|
commitdiff
|
tree
2017-01-17
wilco
Thi
s
patch simplifies the handling of EH ret
u
rn
.
We
.
.
.
commit
|
commitdiff
|
tree
2017-01-11
w
i
lco
My
previo
u
s change
t
o the Co
r
tex
-
A53 schedu
l
e
r
resulted
.
.
.
commit
|
commitdiff
|
tree
2016-12-16
wilco
The negdi2 pa
t
t
e
rn
s
for A
R
M and Th
u
mb-2 are duplicat
e
d
.
.
.
commit
|
commitdiff
|
tree
2016-12-16
wilco
Thumb uses a spec
i
a
l
regist
e
r allocation order
t
o increase
.
.
.
commit
|
commitdiff
|
tree
2016-12-15
wi
l
co
U
s
i
ng leaf_funct
i
on_p in
a
b
ackend is d
a
n
gerous a
s
.
.
.
commit
|
commitdiff
|
tree
2016-12-09
wilco
Add
the test this ti
m
e
.
.
.
commit
|
commitdiff
|
tree
2016-12-08
wilco
Th
i
s patch fixes an issue in aar
c
h64_clas
s
ify_address
.
.
.
commit
|
commitdiff
|
tree
2016-12-07
wi
l
co
GCC caches the w
h
ether a func
t
ion is a leaf in crtl
.
.
.
commit
|
commitdiff
|
tree
2016-12-07
wilco
Improve TI mode
add
r
ess
o
ffsets
-
these may eit
h
er
.
.
.
commit
|
commitdiff
|
tree
2016-11-16
wi
l
co
L
o
oking
at PR77308, one of the iss
u
es
is that th
e
bswap
.
.
.
commit
|
commitdiff
|
tree
2016-11-14
wilco
T
he sec
o
nd patch
updates the Cortex-
A
57 sched
u
l
er no
w
.
.
.
commit
|
commitdiff
|
tree
2016-11-14
wilco
Cu
r
rently th
e
S
BF
M
, UBF
M
and BFM instr
u
ction
s
al
l
u
se
.
.
.
commit
|
commitdiff
|
tree
2016-11-14
wilco
Th
e
exis
t
ing vect
o
r costs stop some
be
n
efi
c
ia
l
v
ecto
r
ization
.
.
.
commit
|
commitdiff
|
tree
2016-11-02
w
ilco
Add LE/BE SHA1H patte
r
ns
w
ith a V2SI input
.
This avoids
.
.
.
commit
|
commitdiff
|
tree
2016-11-02
wilco
T
he add expander still cont
a
ins some expansion
c
ode
.
.
.
commit
|
commitdiff
|
tree
2016-10-25
wilco
With -fpu=neon DI mode sh
i
ft
s
a
re
e
xp
a
nded a
f
t
er re
l
oad
.
.
.
commit
|
commitdiff
|
tree
2016-10-21
wilco
Real
l
y add th
e
test this t
i
me
.
.
.
commit
|
commitdiff
|
tree
2016-10-21
wil
c
o
Improve stack adjustment by reusing a te
m
porary m
o
ve
.
.
.
commit
|
commitdiff
|
tree
2016-10-21
wilco
If the number
of in
t
eger
c
allee-saves is odd, any FP
.
.
.
commit
|
commitdiff
|
tree
2016-10-05
w
i
l
co
Move
all existi
n
g
strchr an
d
strrchr folding from buil
t
ins
.
.
.
commit
|
commitdiff
|
tree
2016-09-28
wi
l
co
I
f strchr can't be
folded in
gimple-fol
d
, w
e
still
.
.
.
commit
|
commitdiff
|
tree
2016-09-28
wilco
Optimize st
r
ch
r
(s, 0) to s +
strlen (s)
.
strchr
.
.
.
commit
|
commitdiff
|
tree
2016-09-07
w
ilco
Im
p
rove
aarc
h
64_legitimize_addr
e
ss - avoid split
t
ing
.
.
.
commit
|
commitdiff
|
tree
2016-09-01
w
i
l
co
This pat
c
h adds legitim
i
ze_address_disp
l
acement ho
o
k
.
.
.
commit
|
commitdiff
|
tree
2016-08-01
wilco
This pa
t
c
h
o
pti
m
izes t
h
e prolog and
ep
i
l
og
code to
.
.
.
commit
|
commitdiff
|
tree
2016-07-28
wilco
On AArch64 the UXTB
and UXTH
i
nstructi
o
n
s are al
i
a
ses
.
.
.
commit
|
commitdiff
|
tree
2016-07-28
wilco
This pa
t
chset improves
z
ero e
x
tend costs a
n
d
code generatio
n
.
commit
|
commitdiff
|
tree
2016-07-28
wilco
This pat
c
h improv
e
s the
readability
of t
h
e pro
l
o
g
and
.
.
.
commit
|
commitdiff
|
tree
2016-07-06
w
i
lco
This p
a
t
c
h
improves
t
he accuracy of
the C
o
rtex-A53
.
.
.
commit
|
commitdiff
|
tree
2016-07-06
w
i
lco
Fix proto
t
ype in vst1Q_laneu6
4
-1
.
c to unsigne
d
c
har*
.
commit
|
commitdiff
|
tree
2016-06-30
wilc
o
T
h
i
s
p
atch sets the branch
c
o
st to th
e
same most optimal
.
.
.
commit
|
commitdiff
|
tree
2016-06-29
w
i
lco
Increase loop
alignment on Cortex
cores to 8
a
nd set
.
.
.
commit
|
commitdiff
|
tree
2016-06-28
wilco
This pat
c
h
f
i
xes
a
bug in the bsw
a
p pass
.
In big-
e
nd
i
an
.
.
.
commit
|
commitdiff
|
tree
2016-06-21
w
ilco
T
h
e
r
ece
n
tly
a
dded
g
cc
.
t
a
rget
/
a
a
rch64/adv
s
imd-intrinsics
.
.
.
commit
|
commitdiff
|
tree
2016-06-21
wilco
Fix tree-
s
sa/
a
ttr-hotcold-
2
.
c
f
ailures now that
t
he
.
.
.
commit
|
commitdiff
|
tree
2016-06-21
wilco
D
u
e to recen
t
imp
r
ovements to
t
he vectorize
r
,
the number
.
.
.
commit
|
commitdiff
|
tree
2016-06-20
w
i
l
c
o
This
p
atch cleans up the -
m
pc-relative-l
o
ads
optio
n
.
.
.
commit
|
commitdiff
|
tree
2016-06-20
wil
c
o
I
mprove mo
d
es_ti
e
able by returning tru
e
i
n
more cases
.
.
.
commit
|
commitdiff
|
tree
2016-06-20
wilco
The Cortex-A57 sche
d
uler is missing
f
cs
e
l, s
o
add it
.
commit
|
commitdiff
|
tree
2016-05-27
wilco
Remove aa
r
ch64_cannot_change_
m
ode_cl
a
ss as the underlying
.
.
.
commit
|
commitdiff
|
tree
2016-05-26
wilco
GCC expands switch statements i
n
a
very simplistic
.
.
.
commit
|
commitdiff
|
tree
2016-05-26
w
i
lco
SIMD operations
l
ike combine prefe
r
t
o
have t
h
eir o
p
eran
d
s
.
.
.
commit
|
commitdiff
|
tree
2016-05-16
wilc
o
Update d
o
cumentatio
n
of AArch64 options
for GCC6 to
.
.
.
commit
|
commitdiff
|
tree
2016-05-16
wilco
So
m
e patterns are using '%w2' for immediate
o
pe
r
ands
.
.
.
commit
|
commitdiff
|
tree
2016-05-16
wilco
This patch
f
ixes the
a
t
tri
b
ut
e
s
o
f in
t
eger immedi
a
te
.
.
.
commit
|
commitdiff
|
tree
2016-02-17
wi
l
co
As part of
t
he final cleanup
o
f aarch64_internal_
m
o
v_immedia
.
.
.
commit
|
commitdiff
|
tree
2016-02-10
w
ilco
E
n
able instruction fusion of
dependent AESE; AESMC
.
.
.
commit
|
commitdiff
|
tree
2016-02-10
wilco
E
n
able instruc
t
i
o
n
f
usion of
A
ES instr
u
c
t
i
ons on ARM
.
.
.
commit
|
commitdiff
|
tree
2016-02-04
wilco
Thi
s
patch fix
e
s an exponential issue in ccmp
.
c
.
When
.
.
.
commit
|
commitdiff
|
tree
2016-02-03
wi
l
co
F
ix the
ccmp_1
.
c test back to use '0'
a
s regula
r
expre
s
sions
.
.
.
commit
|
commitdiff
|
tree
2016-02-02
w
ilco
Improve
T
ARGET
_
IRA_CHA
N
G
E
_PSEUDO_A
L
L
O
C
NO_CLA
S
S
target
.
.
.
commit
|
commitdiff
|
tree
2016-02-02
w
il
c
o
This p
a
tch adds suppo
r
t
f
o
r t
h
e TARGET_IR
A
_
CHANGE
_
PSEUDO_ALL
.
.
.
commit
|
commitdiff
|
tree
2016-01-28
wilco
A recent chan
g
e
adde
d
UNSP
E
C to the CCMP pa
t
terns
to
.
.
.
commit
|
commitdiff
|
tree
2016-01-28
wil
c
o
Add support for vector permu
t
e cost
s
ince v
a
rious permutes
.
.
.
commit
|
commitdiff
|
tree
2016-01-28
wilco
S
eve
r
al instruct
i
ons disassemble a zero
immediate as
.
.
.
commit
|
commitdiff
|
tree
2016-01-27
w
i
lco
2015-01
-
27 Wi
l
c
o Dijkstra <wdijkstr@arm
.
com>
commit
|
commitdiff
|
tree