Searched hist:cf86a086a18095e33e0637cb78cda1fcf5280852 (Results 1 – 3 of 3) sorted by relevance
| /openbmc/linux/include/net/ |
| H A D | dst_ops.h | diff cf86a086a18095e33e0637cb78cda1fcf5280852 Thu May 07 20:58:10 CDT 2020 Eric Dumazet <edumazet@google.com> net/dst: use a smaller percpu_counter batch for dst entries accounting
percpu_counter_add() uses a default batch size which is quite big
on platforms with 256 cpus. (2*256 -> 512)
This means dst_entries_get_fast() can be off by +/- 2*(nr_cpus^2)
(131072 on servers with 256 cpus)
Reduce the batch size to something more reasonable, and
add logic to ip6_dst_gc() to call dst_entries_get_slow()
before calling the _very_ expensive fib6_run_gc() function.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
| /openbmc/linux/net/core/ |
| H A D | dst.c | diff cf86a086a18095e33e0637cb78cda1fcf5280852 Thu May 07 20:58:10 CDT 2020 Eric Dumazet <edumazet@google.com> net/dst: use a smaller percpu_counter batch for dst entries accounting
percpu_counter_add() uses a default batch size which is quite big
on platforms with 256 cpus. (2*256 -> 512)
This means dst_entries_get_fast() can be off by +/- 2*(nr_cpus^2)
(131072 on servers with 256 cpus)
Reduce the batch size to something more reasonable, and
add logic to ip6_dst_gc() to call dst_entries_get_slow()
before calling the _very_ expensive fib6_run_gc() function.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
| /openbmc/linux/net/ipv6/ |
| H A D | route.c | diff cf86a086a18095e33e0637cb78cda1fcf5280852 Thu May 07 20:58:10 CDT 2020 Eric Dumazet <edumazet@google.com> net/dst: use a smaller percpu_counter batch for dst entries accounting
percpu_counter_add() uses a default batch size which is quite big
on platforms with 256 cpus. (2*256 -> 512)
This means dst_entries_get_fast() can be off by +/- 2*(nr_cpus^2)
(131072 on servers with 256 cpus)
Reduce the batch size to something more reasonable, and
add logic to ip6_dst_gc() to call dst_entries_get_slow()
before calling the _very_ expensive fib6_run_gc() function.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|