| Message ID | 20161119010808.GF1200@avx2 |
|---|---|
| State | Accepted, archived |
| Delegated to: | David Miller |
| Headers | show |
From: Alexey Dobriyan <adobriyan@gmail.com> Date: Sat, 19 Nov 2016 04:08:08 +0300 > 1) cast to "int" is unnecessary: > u8 will be promoted to int before decrementing, > small positive numbers fit into "int", so their values won't be changed > during promotion. > > Once everything is int including loop counters, signedness doesn't > matter: 32-bit operations will stay 32-bit operations. > > But! Someone tried to make this loop smart by making everything of > the same type apparently in an attempt to optimise it. > Do the optimization, just differently. > Do the cast where it matters. :^) > > 2) frag size is unsigned entity and sum of fragments sizes is also > unsigned. > > Make everything unsigned, leave no MOVSX instruction behind. ... > Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Applied to net-next.
From: Alexey Dobriyan > Sent: 19 November 2016 01:08 ... > - for (i = (int)skb_shinfo(skb)->nr_frags - 1; i >= 0; i--) > + for (i = skb_shinfo(skb)->nr_frags - 1; (int)i >= 0; i--) > len += skb_frag_size(&skb_shinfo(skb)->frags[i]); Think I'd use: for (i = skb_shinfo(skb)->nr_frags; i-- != 0; ) David
On Wed, Nov 23, 2016 at 3:49 PM, David Laight <David.Laight@aculab.com> wrote: > From: Alexey Dobriyan >> Sent: 19 November 2016 01:08 > ... >> - for (i = (int)skb_shinfo(skb)->nr_frags - 1; i >= 0; i--) >> + for (i = skb_shinfo(skb)->nr_frags - 1; (int)i >= 0; i--) >> len += skb_frag_size(&skb_shinfo(skb)->frags[i]); > > Think I'd use: > for (i = skb_shinfo(skb)->nr_frags; i-- != 0; ) This kind of diverges from canonical loop form: for (init; temination condition: iterator) by shifting everything into termination part. A
--- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -1799,11 +1799,11 @@ static inline unsigned int skb_headlen(const struct sk_buff *skb) return skb->len - skb->data_len; } -static inline int skb_pagelen(const struct sk_buff *skb) +static inline unsigned int skb_pagelen(const struct sk_buff *skb) { - int i, len = 0; + unsigned int i, len = 0; - for (i = (int)skb_shinfo(skb)->nr_frags - 1; i >= 0; i--) + for (i = skb_shinfo(skb)->nr_frags - 1; (int)i >= 0; i--) len += skb_frag_size(&skb_shinfo(skb)->frags[i]); return len + skb_headlen(skb); } --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -581,7 +581,7 @@ int ip_do_fragment(struct net *net, struct sock *sk, struct sk_buff *skb, */ if (skb_has_frag_list(skb)) { struct sk_buff *frag, *frag2; - int first_len = skb_pagelen(skb); + unsigned int first_len = skb_pagelen(skb); if (first_len - hlen > mtu || ((first_len - hlen) & 7) || --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -625,7 +625,7 @@ int ip6_fragment(struct net *net, struct sock *sk, struct sk_buff *skb, hroom = LL_RESERVED_SPACE(rt->dst.dev); if (skb_has_frag_list(skb)) { - int first_len = skb_pagelen(skb); + unsigned int first_len = skb_pagelen(skb); struct sk_buff *frag2; if (first_len - hlen > mtu ||
1) cast to "int" is unnecessary: u8 will be promoted to int before decrementing, small positive numbers fit into "int", so their values won't be changed during promotion. Once everything is int including loop counters, signedness doesn't matter: 32-bit operations will stay 32-bit operations. But! Someone tried to make this loop smart by making everything of the same type apparently in an attempt to optimise it. Do the optimization, just differently. Do the cast where it matters. :^) 2) frag size is unsigned entity and sum of fragments sizes is also unsigned. Make everything unsigned, leave no MOVSX instruction behind. add/remove: 0/0 grow/shrink: 0/3 up/down: 0/-4 (-4) function old new delta skb_cow_data 835 834 -1 ip_do_fragment 2549 2548 -1 ip6_fragment 3130 3128 -2 Total: Before=154865032, After=154865028, chg -0.00% Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> --- include/linux/skbuff.h | 6 +++--- net/ipv4/ip_output.c | 2 +- net/ipv6/ip6_output.c | 2 +- 3 files changed, 5 insertions(+), 5 deletions(-)