Commit 747d6dab authored by Bart De Schuymer's avatar Bart De Schuymer Committed by Linus Torvalds

[PATCH] update comments in ip_tables.c

I believe the comments about locking in net/ipv4/netfilter/ip_tables.c
are outdated/misleading.  Please consider the following comments instead
(or something similar).
parent 0f3a9b1c
...@@ -68,13 +68,12 @@ static DECLARE_MUTEX(ipt_mutex); ...@@ -68,13 +68,12 @@ static DECLARE_MUTEX(ipt_mutex);
#define inline #define inline
#endif #endif
/* Locking is simple: we assume at worst case there will be one packet /*
in user context and one from bottom halves (or soft irq if Alexey's
softnet patch was applied).
We keep a set of rules for each CPU, so we can avoid write-locking We keep a set of rules for each CPU, so we can avoid write-locking
them; doing a readlock_bh() stops packets coming through if we're them in the softirq when updating the counters and therefore
in user context. only need to read-lock in the softirq; doing a write_lock_bh() in user
context stops packets coming through and allows user context to read
the counters or update the rules.
To be cache friendly on SMP, we arrange them like so: To be cache friendly on SMP, we arrange them like so:
[ n-entries ] [ n-entries ]
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment