code review 4584046: big: removed some gratuitous +/-1's

big: removed some gratuitous +/-1's

[gri, 2011-06-08 18:02:41]

Hello mtj@google.com (cc: golang-dev@googlegroups.com),

I'd like you to review this change to
https://go.googlecode.com/hg/

[bradfitz, 2011-06-08 18:05:33]

http://codereview.appspot.com/4584046/diff/5001/src/pkg/big/nat.go
File src/pkg/big/nat.go (right):

http://codereview.appspot.com/4584046/diff/5001/src/pkg/big/nat.go#newcode772
src/pkg/big/nat.go:772: for k := 1; k < len(x); k++ {
Is this a performance CL or a visual-cleanup CL?

If the former, is it cheaper to avoid len each time through the loop?

for k, m := 1, len(x); k < m; k++ {

... perhaps?

[gri, 2011-06-08 18:16:08]

http://codereview.appspot.com/4584046/diff/5001/src/pkg/big/nat.go
File src/pkg/big/nat.go (right):

http://codereview.appspot.com/4584046/diff/5001/src/pkg/big/nat.go#newcode772
src/pkg/big/nat.go:772: for k := 1; k < len(x); k++ {
On 2011/06/08 18:05:33, bradfitz wrote:
> Is this a performance CL or a visual-cleanup CL?
> 
> If the former, is it cheaper to avoid len each time through the loop?
> 
> for k, m := 1, len(x); k < m; k++ {
> 
> ... perhaps?

x is a local variable (slice), and len(x) is as fast as accessing a local
variable (it's just accessing the len field of the slice). This is not better.

The change might make the code perhaps a tiny bit faster, but it's mostly
simpler code, which is better. k is simply always the index of the next word,
because that's the index we need.

[bradfitz, 2011-06-08 18:19:38]

LGTM

On Wed, Jun 8, 2011 at 11:16 AM, <gri@golang.org> wrote:

>
>
http://codereview.appspot.com/**4584046/diff/5001/src/pkg/big/**nat.go<http:/...
> File src/pkg/big/nat.go (right):
>
> http://codereview.appspot.com/**4584046/diff/5001/src/pkg/big/**
>
nat.go#newcode772<http://codereview.appspot.com/4584046/diff/5001/src/pkg/big/nat.go#newcode772>
> src/pkg/big/nat.go:772: for k := 1; k < len(x); k++ {
> On 2011/06/08 18:05:33, bradfitz wrote:
>
>> Is this a performance CL or a visual-cleanup CL?
>>
>
>  If the former, is it cheaper to avoid len each time through the loop?
>>
>
>  for k, m := 1, len(x); k < m; k++ {
>>
>
>  ... perhaps?
>>
>
> x is a local variable (slice), and len(x) is as fast as accessing a
> local variable (it's just accessing the len field of the slice). This is
> not better.
>
> The change might make the code perhaps a tiny bit faster, but it's
> mostly simpler code, which is better. k is simply always the index of
> the next word, because that's the index we need.
>
>
>
http://codereview.appspot.com/**4584046/<http://codereview.appspot.com/4584046/>
>

[gri, 2011-06-08 18:24:30]

*** Submitted as http://code.google.com/p/go/source/detail?r=fb6f033c562f ***

big: removed some gratuitous +/-1's

R=mtj, bradfitz
CC=golang-dev
http://codereview.appspot.com/4584046