[PATCH bpf-next 0/2] Remove unused fields in cpumap & devmap

[PATCH bpf-next 0/2] Remove unused fields in cpumap & devmap

[Hou Tao]

From: Hou Tao <houtao1@huawei.com>

Hi,

Patchset "Simplify xdp_do_redirect_map()/xdp_do_flush_map() and XDP
maps" [0] changed per-map flush list to global per-cpu flush list
for cpumap, devmap and xskmap, but it forgot to remove these unused
fields from cpumap and devmap. So just remove these unused fields.

Comments and suggestions are always welcome.

Regards,
Tao

[0]: https://lore.kernel.org/bpf/20191219061006.21980-1-bjorn.topel@gmail.com

Hou Tao (2):
  bpf, cpumap: Remove unused cmap field from bpf_cpu_map_entry
  bpf, devmap: Remove unused dtab field from bpf_dtab_netdev

 kernel/bpf/cpumap.c | 3 ---
 kernel/bpf/devmap.c | 2 --
 2 files changed, 5 deletions(-)

-- 
2.29.2

[PATCH bpf-next 1/2] bpf, cpumap: Remove unused cmap field from bpf_cpu_map_entry

[Hou Tao]

From: Hou Tao <houtao1@huawei.com>

Since commit cdfafe98cabe ("xdp: Make cpumap flush_list common for all
map instances"), cmap is no longer used, so just remove it.

Signed-off-by: Hou Tao <houtao1@huawei.com>
---
 kernel/bpf/cpumap.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/kernel/bpf/cpumap.c b/kernel/bpf/cpumap.c
index 6ae02be7a48e..0a16e30b16ef 100644
--- a/kernel/bpf/cpumap.c
+++ b/kernel/bpf/cpumap.c
@@ -60,8 +60,6 @@ struct bpf_cpu_map_entry {
 	/* XDP can run multiple RX-ring queues, need __percpu enqueue store */
 	struct xdp_bulk_queue __percpu *bulkq;
 
-	struct bpf_cpu_map *cmap;
-
 	/* Queue with potential multi-producers, and single-consumer kthread */
 	struct ptr_ring *queue;
 	struct task_struct *kthread;
@@ -588,7 +586,6 @@ static long cpu_map_update_elem(struct bpf_map *map, void *key, void *value,
 		rcpu = __cpu_map_entry_alloc(map, &cpumap_value, key_cpu);
 		if (!rcpu)
 			return -ENOMEM;
-		rcpu->cmap = cmap;
 	}
 	rcu_read_lock();
 	__cpu_map_entry_replace(cmap, key_cpu, rcpu);
-- 
2.29.2

[PATCH bpf-next 2/2] bpf, devmap: Remove unused dtab field from bpf_dtab_netdev

[Hou Tao]

From: Hou Tao <houtao1@huawei.com>

Commit 96360004b862 ("xdp: Make devmap flush_list common for all map
instances") removes the use of bpf_dtab_netdev::dtab in bq_enqueue(),
so just remove dtab from bpf_dtab_netdev.

Signed-off-by: Hou Tao <houtao1@huawei.com>
---
 kernel/bpf/devmap.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
index 49cc0b5671c6..4d42f6ed6c11 100644
--- a/kernel/bpf/devmap.c
+++ b/kernel/bpf/devmap.c
@@ -65,7 +65,6 @@ struct xdp_dev_bulk_queue {
 struct bpf_dtab_netdev {
 	struct net_device *dev; /* must be first member, due to tracepoint */
 	struct hlist_node index_hlist;
-	struct bpf_dtab *dtab;
 	struct bpf_prog *xdp_prog;
 	struct rcu_head rcu;
 	unsigned int idx;
@@ -874,7 +873,6 @@ static struct bpf_dtab_netdev *__dev_map_alloc_node(struct net *net,
 	}
 
 	dev->idx = idx;
-	dev->dtab = dtab;
 	if (prog) {
 		dev->xdp_prog = prog;
 		dev->val.bpf_prog.id = prog->aux->id;
-- 
2.29.2

Re: [PATCH bpf-next 1/2] bpf, cpumap: Remove unused cmap field from bpf_cpu_map_entry

[Jesper Dangaard Brouer]

On 28/07/2023 03.49, Hou Tao wrote:
> From: Hou Tao <houtao1@huawei.com>
> 
> Since commit cdfafe98cabe ("xdp: Make cpumap flush_list common for all
> map instances"), cmap is no longer used, so just remove it.
> 
> Signed-off-by: Hou Tao <houtao1@huawei.com>
> ---

LGTM

Acked-by: Jesper Dangaard Brouer <hawk@kernel.org>


>   kernel/bpf/cpumap.c | 3 ---
>   1 file changed, 3 deletions(-)
> 
> diff --git a/kernel/bpf/cpumap.c b/kernel/bpf/cpumap.c
> index 6ae02be7a48e..0a16e30b16ef 100644
> --- a/kernel/bpf/cpumap.c
> +++ b/kernel/bpf/cpumap.c
> @@ -60,8 +60,6 @@ struct bpf_cpu_map_entry {
>   	/* XDP can run multiple RX-ring queues, need __percpu enqueue store */
>   	struct xdp_bulk_queue __percpu *bulkq;
>   
> -	struct bpf_cpu_map *cmap;
> -
>   	/* Queue with potential multi-producers, and single-consumer kthread */
>   	struct ptr_ring *queue;
>   	struct task_struct *kthread;
> @@ -588,7 +586,6 @@ static long cpu_map_update_elem(struct bpf_map *map, void *key, void *value,
>   		rcpu = __cpu_map_entry_alloc(map, &cpumap_value, key_cpu);
>   		if (!rcpu)
>   			return -ENOMEM;
> -		rcpu->cmap = cmap;
>   	}
>   	rcu_read_lock();
>   	__cpu_map_entry_replace(cmap, key_cpu, rcpu);

Re: [PATCH bpf-next 2/2] bpf, devmap: Remove unused dtab field from bpf_dtab_netdev

[Jesper Dangaard Brouer]

On 28/07/2023 03.49, Hou Tao wrote:
> From: Hou Tao <houtao1@huawei.com>
> 
> Commit 96360004b862 ("xdp: Make devmap flush_list common for all map
> instances") removes the use of bpf_dtab_netdev::dtab in bq_enqueue(),
> so just remove dtab from bpf_dtab_netdev.
> 
> Signed-off-by: Hou Tao <houtao1@huawei.com>

LGTM

Acked-by: Jesper Dangaard Brouer <hawk@kernel.org>

> ---
>   kernel/bpf/devmap.c | 2 --
>   1 file changed, 2 deletions(-)
> 
> diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
> index 49cc0b5671c6..4d42f6ed6c11 100644
> --- a/kernel/bpf/devmap.c
> +++ b/kernel/bpf/devmap.c
> @@ -65,7 +65,6 @@ struct xdp_dev_bulk_queue {
>   struct bpf_dtab_netdev {
>   	struct net_device *dev; /* must be first member, due to tracepoint */
>   	struct hlist_node index_hlist;
> -	struct bpf_dtab *dtab;
>   	struct bpf_prog *xdp_prog;
>   	struct rcu_head rcu;
>   	unsigned int idx;
> @@ -874,7 +873,6 @@ static struct bpf_dtab_netdev *__dev_map_alloc_node(struct net *net,
>   	}
>   
>   	dev->idx = idx;
> -	dev->dtab = dtab;
>   	if (prog) {
>   		dev->xdp_prog = prog;
>   		dev->val.bpf_prog.id = prog->aux->id;

Re: [PATCH bpf-next 1/2] bpf, cpumap: Remove unused cmap field from bpf_cpu_map_entry

[Jiri Olsa]

On Fri, Jul 28, 2023 at 09:49:41AM +0800, Hou Tao wrote:
> From: Hou Tao <houtao1@huawei.com>
> 
> Since commit cdfafe98cabe ("xdp: Make cpumap flush_list common for all
> map instances"), cmap is no longer used, so just remove it.

nit, should it have Fixes: cdfafe98cabe ?

same for the other patch, other than that for the patchset:

Acked-by: Jiri Olsa <jolsa@kernel.org>

jirka

> 
> Signed-off-by: Hou Tao <houtao1@huawei.com>
> ---
>  kernel/bpf/cpumap.c | 3 ---
>  1 file changed, 3 deletions(-)
> 
> diff --git a/kernel/bpf/cpumap.c b/kernel/bpf/cpumap.c
> index 6ae02be7a48e..0a16e30b16ef 100644
> --- a/kernel/bpf/cpumap.c
> +++ b/kernel/bpf/cpumap.c
> @@ -60,8 +60,6 @@ struct bpf_cpu_map_entry {
>  	/* XDP can run multiple RX-ring queues, need __percpu enqueue store */
>  	struct xdp_bulk_queue __percpu *bulkq;
>  
> -	struct bpf_cpu_map *cmap;
> -
>  	/* Queue with potential multi-producers, and single-consumer kthread */
>  	struct ptr_ring *queue;
>  	struct task_struct *kthread;
> @@ -588,7 +586,6 @@ static long cpu_map_update_elem(struct bpf_map *map, void *key, void *value,
>  		rcpu = __cpu_map_entry_alloc(map, &cpumap_value, key_cpu);
>  		if (!rcpu)
>  			return -ENOMEM;
> -		rcpu->cmap = cmap;
>  	}
>  	rcu_read_lock();
>  	__cpu_map_entry_replace(cmap, key_cpu, rcpu);
> -- 
> 2.29.2
> 

Re: [PATCH bpf-next 1/2] bpf, cpumap: Remove unused cmap field from bpf_cpu_map_entry

[Jakub Kicinski]

On Fri, 28 Jul 2023 14:04:28 +0200 Jiri Olsa wrote:
> nit, should it have Fixes: cdfafe98cabe ?

I don't think so, Fixes is for people backporting fixes.

Re: [PATCH bpf-next 0/2] Remove unused fields in cpumap & devmap

[patchwork-bot+netdevbpf]

Hello:

This series was applied to bpf/bpf-next.git (master)
by Martin KaFai Lau <martin.lau@kernel.org>:

On Fri, 28 Jul 2023 09:49:40 +0800 you wrote:
> From: Hou Tao <houtao1@huawei.com>
> 
> Hi,
> 
> Patchset "Simplify xdp_do_redirect_map()/xdp_do_flush_map() and XDP
> maps" [0] changed per-map flush list to global per-cpu flush list
> for cpumap, devmap and xskmap, but it forgot to remove these unused
> fields from cpumap and devmap. So just remove these unused fields.
> 
> [...]

Here is the summary with links:
  - [bpf-next,1/2] bpf, cpumap: Remove unused cmap field from bpf_cpu_map_entry
    https://git.kernel.org/bpf/bpf-next/c/2d20bfc315eb
  - [bpf-next,2/2] bpf, devmap: Remove unused dtab field from bpf_dtab_netdev
    https://git.kernel.org/bpf/bpf-next/c/1ea66e89f68c

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html