Presentation on code optimization for compiler design

Lecture 11: Code Optimization
CS 540
George Mason University

CS 540 Spring 2009 GMU 2
Code Optimization
REQUIREMENTS:
• Meaning must be preserved (correctness)
• Speedup must occur on average.
• Work done must be worth the effort.
OPPORTUNITIES:
• Programmer (algorithm, directives)
• Intermediate code
• Target code

Code Optimization
Scanner
(lexical
analysis)
Parser
(syntax
analysis)
Code
Optimizer
Semantic
Analysis
(IC generator)
Code
Generator
Symbol
Table
Source
language
tokens Syntactic
structure
Syntactic/semantic
structure
Target
language

Levels
• Window – peephole optimization
• Basic block
• Procedural – global (control flow graph)
• Program level – intraprocedural (program
dependence graph)

Peephole Optimizations
• Constant Folding
x := 32 becomes x := 64
x := x + 32
• Unreachable Code
goto L2
x := x + 1  unneeded
• Flow of control optimizations
goto L1 becomes goto L2
…
L1: goto L2

• Algebraic Simplification
x := x + 0  unneeded
• Dead code
x := 32  where x not used after statement
y := x + y  y := y + 32
• Reduction in strength
x := x * 2  x := x + x

• Local in nature
• Pattern driven
• Limited by the size of the window

Basic Block Level
• Common Subexpression elimination
• Constant Propagation
• Dead code elimination
• Plus many others such as copy propagation,
value numbering, partial redundancy
elimination, …

Simple example: a[i+1] = b[i+1]
• t1 = i+1
• t2 = b[t1]
• t3 = i + 1
• a[t3] = t2
• t1 = i + 1
• t2 = b[t1]
• t3 = i + 1  no longer live
• a[t1] = t2
Common expression can be eliminated

• i = 4
• t1 = i+1
• t2 = b[t1]
• a[t1] = t2
• i = 4
• t1 = 5
• t2 = b[t1]
• a[t1] = t2
Now, suppose i is a constant:
• i = 4
• t1 = 5
• t2 = b[5]
• a[5] = t2
• i = 4
• t2 = b[5]
• a[5] = t2
Final Code:

Control Flow Graph - CFG
CFG = < V, E, Entry >, where
V = vertices or nodes, representing an instruction or basic
block (group of statements).
E = (V x V) edges, potential flow of control
Entry is an element of V, the unique program entry
Two sets used in algorithms:
• Succ(v) = {x in V| exists e in E, e = v x}
• Pred(v) = {x in V| exists e in E, e = x v}
1 2 3 4 5

Definitions
• point - any location between adjacent statements
and before and after a basic block.
• A path in a CFG from point p1 to pn is a sequence
of points such that  j, 1 <= j < n, either pi is the
point immediately preceding a statement and pi+1
is the point immediately following that statement
in the same block, or pi is the end of some block
and pi+1 is the start of a successor block.

CFG
c = a + b
d = a * c
i = 1
f[i] = a + b
c = c * 2
if c > d
g = a * c g = d * d
i = i + 1
if i > 10
points
path

Optimizations on CFG
• Must take control flow into account
– Common Sub-expression Elimination
– Constant Propagation
– Dead Code Elimination
– Partial redundancy Elimination
– …
• Applying one optimization may create
opportunities for other optimizations.

Redundant Expressions
An expression x op y is redundant at a point p if it has
already been computed at some point(s) and no intervening
operations redefine x or y.
m = 2*y*z t0 = 2*y t0 = 2*y
m = t0*z m = t0*z
n = 3*y*z t1 = 3*y t1 = 3*y
n = t1*z n = t1*z
o = 2*y–z t2 = 2*y
o = t2-z o = t0-z
redundant

c = a + b
d = a * c
i = 1
f[i] = a + b
c = c * 2
if c > d
g = a * c g = d * d
i = i + 1
if i > 10
Candidates:
a + b
a * c
d * d
c * 2
i + 1
Definition site
Since a + b is
available here,
 redundant!

c = a + b
d = a * c
i = 1
f[i] = a + b
c = c * 2
if c > d
g = a * c g = d * d
i = i + 1
if i > 10
Candidates:
a + b
a * c
d * d
c * 2
i + 1
Definition site
Kill site
Not available
 Not redundant

• An expression e is defined at some point p in the
CFG if its value is computed at p. (definition site)
• An expression e is killed at point p in the CFG if
one or more of its operands is defined at p. (kill
site)
• An expression is available at point p in a CFG if
every path leading to p contains a prior definition
of e and e is not killed between that definition and
p.

Removing Redundant Expressions
t1 = a + b
c = t1
d = a * c
i = 1
f[i] = t1
c = c * 2
if c > d
g = a * c g = d*d
i = i + 1
if i > 10
Candidates:
a + b
a * c
d * d
c * 2
i + 1

Constant Propagation
b = 5
c = 4*b
c > b
d = b + 2
e = a + b
b = 5
c = 20
c > 5
d = 7
e = a + 5
e = a + b
t
f t
f
b = 5
c = 20
20 > 5
d = 7
e = a + 5
t
f

Constant Propagation
b = 5
c = 20
20 > 5
d = 7
e = a + 5
t
f
b = 5
c = 20
d = 7
e = a + 5

Copy Propagation
b = a
c = 4*b
c > b
d = b + 2
e = a + b
b = a
c = 4*a
c > a
d = a + 2
e = a + a
e = a + b

Simple Loop Optimizations: Code
Motion
while (i <= limit - 2)
t := limit - 2
while (i <= t)
L1:
t1 = limit – 2
if (i > t1) goto L2
body of loop
goto L1
L2:
t1 = limit – 2
L1:
if (i > t1) goto L2
body of loop
goto L1
L2:

Simple Loop Optimizations:
Strength Reduction
• Induction Variables control loop iterations
j = j – 1
t4 = 4 * j
t5 = a[t4]
if t5 > v
j = j – 1
t4 = t4 - 4
t5 = a[t4]
if t5 > v
t4 = 4*j

Simple Loop Optimizations
• Loop transformations are often used to expose
other optimization opportunities:
– Normalization
– Loop Interchange
– Loop Fusion
– Loop Reversal
– …

Consider Matrix Multiplication
for i = 1 to n do
for j = 1 to n do
for k = 1 to n do
C[i,j] = C[i,j] + A[i,k] + B[k,j]
end
end
end
= +
i i
j j
k
k
C B
A

Memory Usage
• For A: Elements are accessed across rows, spatial locality
is exploited for cache (assuming row major storage)
• For B: Elements are accessed along columns, unless
cache can hold all of B, cache will have problems.
• For C: Single element computed per loop – use register to
hold
= +
i i
j j
k
k
C B
A

Matrix Multiplication Version 2
for i = 1 to n do
for k = 1 to n do
for j = 1 to n do
C[i,j] = C[i,j] + A[i,k] + B[k,j]
end
end
end
= +
i i
j j
k
C B
A
loop interchange
k

Memory Usage
• For A: Single element loaded for loop body
• For B: Elements are accessed along rows to
exploit spatial locality.
• For C: Extra loading/storing, but across rows
= +
i i
j j
k
C B
A
k

Simple Loop Optimizations
• How to determine safety?
– Does the new multiply give the same answer?
– Can be reversed??
for (I=1 to N) a[I] = a[I+1] – can this loop be
safely reversed?

Data Dependencies
• Flow Dependencies - write/read
x := 4;
y := x + 1
• Output Dependencies - write/write
x := 4;
x := y + 1;
• Antidependencies - read/write
y := x + 1;
x := 4;

x := 4
y := 6
p := x + 2
z := y + p
x := z
y := p
x := 4 y := 6
p := x + 2
z := y + p
y := p x := z
Flow
Output
Anti

Global Data Flow Analysis
Collecting information about the way data is used in
a program.
• Takes control flow into account
• HL control constructs
– Simpler – syntax driven
– Useful for data flow analysis of source code
• General control constructs – arbitrary branching
Information needed for optimizations such as:
constant propagation, common sub-expressions,
partial redundancy elimination …

Dataflow Analysis: Iterative
Techniques
• First, compute local (block level)
information.
• Iterate until no changes
while change do
change = false
for each basic block
apply equations updating IN and OUT
if either IN or OUT changes, set change
to true
end

Live Variable Analysis
A variable x is live at a point p if there is
some path from p where x is used
before it is defined.
Want to determine for some variable x
and point p whether the value of x
could be used along some path starting
at p.
• Information flows backwards
• May – ‘along some path starting at p’
is x live
here?

Global Live Variable Analysis
Want to determine for some variable x and point p whether
the value of x could be used along some path starting at p.
• DEF[B] - set of variables assigned values in B prior to any
use of that variable
• USE[B] - set of variables used in B prior to any definition
of that variable
• OUT[B] - variables live immediately after the block
OUT[B] - IN[S] for all S in succ(B)
• IN[B] - variables live immediately before the block
IN[B] = USE[B] + (OUT[B] - DEF[B])

d1: a = 1
d2: b = 2
d3: c = a + b
d4: d = c - a
d8: b = a + b
d9: e = c - 1
d10: a = b * d
d22: b = a - d
d5: d = b * d
d6: d = a + b
d7: e = e + 1
B1
B2
B3
B4
B5
B6
DEF=a,b
USE =
DEF=c,d
USE = a,b
DEF=
USE = b,d
DEF=d
USE = a,b,e
DEF= e
USE = a,b,c
DEF= a
USE = b,d

Global Live Variable Analysis
Want to determine for some variable x and point p whether
the value of x could be used along some path starting at p.
• DEF[B] - set of variables assigned values in B prior to any
use of that variable
• USE[B] - set of variables used in B prior to any definition
of that variable
• OUT[B] - variables live immediately after the block
OUT[B] - IN[S] for all S in succ(B)
• IN[B] - variables live immediately before the block
IN[B] = USE[B]  (OUT[B] - DEF[B])

IN OUT IN OUT IN OUT
B1  a,b  a,b e a,b,e
B2 a,b a,b,c,d a,b,e a,b,c,d ,e a,b,e a,b,c,d,e
B3 a,b,c,d e a,b,c,e a,b,c,d,e a,b,c,d,e a,b,c,d,e a,b,c,d,e
B4 a,b,c,e a,b,c,d,e a,b,c,e a,b,c,d,e a,b,c,e a,b,c,d,e
B5 a,b,c,d a,b,d a,b,c,d a,b,d,e a,b,c,d a,b,d,e
B6 b,d  b,d  b,d 
OUT[B] =  IN[S] for all S in succ(B)
IN[B] = USE[B] + (OUT[B] - DEF[B])
Block DEF USE
B1 {a,b} { }
B2 {c,d} {a,b}
B3 { } {b,d}
B4 {d} {a,b,e}
B5 {e} {a,b,c}
B6 {a} {b,d}

{ }
{b,d}
{e}
{a,b,e}
{a,b,e}
{a,b,c,d,e}
{a,b,c,d,e}
{a,b,c,d,e}
{a,b,c,d}
{a,b,d,e}
{a,b,c,e}
{a,b,c,d,e}

Dataflow Analysis Problem #2:
Reachability
• A definition of a variable x is a statement that
may assign a value to x.
• A definition may reach a program point p if
there exists some path from the point
immediately following the definition to p such
that the assignment is not killed along that
path.
• Concept: relationship between definitions and
uses

What blocks do definitions d2 and
d4 reach?
d1 i = m – 1
d2 j = n
d3 i = i + 1
d4 j = j - 1
B1
B2
B3
B4 B5
d2
d4

Reachability Analysis:
Unstructured Input
1. Compute GEN and KILL at block—level
2. Compute IN[B] and OUT[B] for B
IN[B] = U OUT[P] where P is a predecessor of B
OUT[B] = GEN[B] U (IN[B] - KILL[B])
3. Repeat step 2 until there are no changes to
OUT sets

Reachability Analysis: Step 1
For each block, compute local (block level)
information = GEN/KILL sets
– GEN[B] = set of definitions generated by B
– KILL[B] = set of definitions that can not reach
the end of B
This information does not take control flow
between blocks into account.

Reasoning about Basic Blocks
Effect of single statement: a = b + c
• Uses variables {b,c}
• Kills all definitions of {a}
• Generates new definition (i.e. assigns a value) of
{a}
Local Analysis:
• Analyze the effect of each instruction
• Compose these effects to derive information about
the entire block

Example
d1 i = m – 1
d2 j = n
d3 a = u1
B1
B2
B3
B4
d4 i = i + 1
d5 j = j - 1
d6 a = u2
d7 i = u2
Gen = 4,5
Kill = 1,2,7
Gen = 1,2,3
Kill = 4,5,6,7
Gen = 7
Kill = 1,4
Gen = 6
Kill = 3

Reachability Analysis: Step 2
Compute IN/OUT for each block in a forward
direction. Start with IN[B] = 
– IN[B] = set of defns reaching the start of B
=  (out[P]) for all predecessor blocks in the CFG
– OUT[B] = set of defns reaching the end of B
= GEN[B]  (IN[B] – KILL[B])
Keep computing IN/OUT sets until a fixed point is
reached.

Reaching Definitions Algorithm
• Input: Flow graph with GEN and KILL for each block
• Output: in[B] and out[B] for each block.
For each block B do out[B] = gen[B], (true if in[B] = emptyset)
change := true;
while change do begin
change := false;
for each block B do begin
in[B] := U out[P], where P is a predecessor of B;
oldout = out[B];
out[B] := gen[B] U (in[B] - kill [B])
if out[B] != oldout then change := true;
end
end

IN OUT
B1  1,2,3
B2  4,5
B3  6
B4  7
IN[B] = (out[P]) for all predecessor
blocks in the CFG
OUT[B] = GEN[B]  (IN[B] – KILL[B])
d1 i = m – 1
d2 j = n
d3 a = u1
B1
B2
B3
B4
d4 i = i + 1
d5 j = j - 1
d6 a = u2
d7 i = u2
Gen = 4,5
Kill = 1,2,7
Gen = 1,2,3
Kill = 4,5,6,7
Gen = 7
Kill = 1,4
Gen = 6
Kill = 3

IN OUT IN OUT
B1  1,2,3  1,2,3
B2  4,5 OUT[1]+OUT[4]
= 1,2,3,7
4,5 + (1,2,3,7
– 1,2,7)
= 3,4,5
B3  6 OUT[2] = 3,4,5 6 + (3,4,5 – 3)
= 4,5,6
B4  7 OUT[2]+OUT[3]
= 3,4,5,6
7 + (3,4,5,6 – 1,4) = 3,5,6,7
blocks in the CFG
OUT[B] = GEN[B] + (IN[B] – KILL[B])

IN OUT IN OUT IN OUT
B1  1,2,3  1,2,3  1,2,3
B2  4,5 1,2,3,7 3,4,5 OUT[1] + OUT[4] =
1,2,3,5,6,7
4,5 + (1,2,3,5,6,7-1,2,7) =
3,4,5,6
B3  6 3,4,5 4,5,6 OUT[2] = 3,4,5,6 6 + (3,4,5,6 – 3)
= 4,5,6
B4  7 3,4,5,6 3,5,6,7 OUT[2] + OUT[3] =
3,4,5,6
7+(3,4,5,6 – 1,4)
= 3,5,6,7
blocks in the CFG
OUT[B] = GEN[B] + (IN[B] – KILL[B])

Forward vs. Backward
Forward flow vs. Backward flow
Forward: Compute OUT for given IN,GEN,KILL
– Information propagates from the predecessors of a
vertex.
– Examples: Reachability, available expressions, constant
propagation
Backward: Compute IN for given OUT,GEN,KILL
– Information propagates from the successors of a vertex.
– Example: Live variable Analysis

Forward vs. Backward Equations
Forward vs. backward
– Forward:
• IN[B] - process OUT[P] for all P in
predecessors(B)
• OUT[B] = local U (IN[B] – local)
– Backward:
• OUT[B] - process IN[S] for all S in
successor(B)
• IN[B] = local U (OUT[B] – local)

May vs. Must
May vs. Must
Must – true on all paths
Ex: constant propagation – variable must provably hold
appropriate constant on all paths in order to do a
substitution
May – true on some path
Ex: Live variable analysis – a variable is live if it could
be used on some path; reachability – a definition
reaches a point if it can reach it on some path

May vs. Must Equations
• May vs. Must
– May – IN[B] = (out[P]) for all
P in pred(B)
– Must – IN[B] = (out[P]) for
all P in pred(B)

• Reachability
– IN[B] = (out[P]) for all P in pred(B)
– OUT[B] = GEN[B] + (IN[B] – KILL[B])
• Live Variable Analysis
– OUT[B] = (IN[S]) for all S in succ(B)
– IN[B] = USE[B]  (OUT[B] - DEF[B])
• Constant Propagation
– IN[B] = (out[P]) for all P in pred(B)
– OUT[B] = DEF_CONST[B]  (IN[B] –
KILL_CONST[B])

Discussion
• Why does this work?
– Finite set – can be represented as bit vectors
– Theory of lattices
• Is this guaranteed to terminate?
– Sets only grow and since finite in size …
• Can we find ways to reduce the number of
iterations?

Choosing visit order for Dataflow
Analysis
In forward flow analysis situations, if we visit the blocks in
depth first order, we can reduce the number of iterations.
Suppose definition d follows block path 3  5  19  35 
16  23  45  4  10  17 where the block numbering
corresponds to the preorder depth-first numbering.
Then we can compute the reach of this definition in 3 iterations
of our algorithm.
3  5  19  35  16  23  45  4  10  17

Presentation on code optimization for compiler design

More Related Content

Similar to Presentation on code optimization for compiler design (20)

Recently uploaded (20)

Presentation on code optimization for compiler design