netdev - RE: [PATCH net-next v2 1/4] net: Register system page pool as an XDP memory model

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Wed, 03 Apr 2024 13:20:32 -0700
From: John Fastabend <john.fastabend@...il.com>
To: Toke Høiland-Jørgensen <toke@...hat.com>, 
 "David S. Miller" <davem@...emloft.net>, 
 Eric Dumazet <edumazet@...gle.com>, 
 Jakub Kicinski <kuba@...nel.org>, 
 Paolo Abeni <pabeni@...hat.com>, 
 Alexei Starovoitov <ast@...nel.org>, 
 Daniel Borkmann <daniel@...earbox.net>, 
 Jesper Dangaard Brouer <hawk@...nel.org>, 
 John Fastabend <john.fastabend@...il.com>
Cc: Toke Høiland-Jørgensen <toke@...hat.com>, 
 Alexander Lobakin <aleksander.lobakin@...el.com>, 
 netdev@...r.kernel.org, 
 bpf@...r.kernel.org
Subject: RE: [PATCH net-next v2 1/4] net: Register system page pool as an XDP
 memory model

Toke Høiland-Jørgensen wrote:
> To make the system page pool usable as a source for allocating XDP
> frames, we need to register it with xdp_reg_mem_model(), so that page
> return works correctly. This is done in preparation for using the system
> page pool for the XDP live frame mode in BPF_TEST_RUN; for the same
> reason, make the per-cpu variable non-static so we can access it from
> the test_run code as well.
> 
> Reviewed-by: Alexander Lobakin <aleksander.lobakin@...el.com>
> Tested-by: Alexander Lobakin <aleksander.lobakin@...el.com>
> Signed-off-by: Toke Høiland-Jørgensen <toke@...hat.com>
> ---
>  include/linux/netdevice.h |  1 +
>  net/core/dev.c            | 13 ++++++++++++-
>  2 files changed, 13 insertions(+), 1 deletion(-)
> 
> diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
> index c541550b0e6e..e1dfdf0c4075 100644
> --- a/include/linux/netdevice.h
> +++ b/include/linux/netdevice.h
> @@ -3345,6 +3345,7 @@ static inline void input_queue_tail_incr_save(struct softnet_data *sd,
>  }
>  
>  DECLARE_PER_CPU_ALIGNED(struct softnet_data, softnet_data);
> +DECLARE_PER_CPU_ALIGNED(struct page_pool *, system_page_pool);
>  
>  static inline int dev_recursion_level(void)
>  {
> diff --git a/net/core/dev.c b/net/core/dev.c
> index d8dd293a7a27..cdb916a647e7 100644
> --- a/net/core/dev.c
> +++ b/net/core/dev.c
> @@ -428,7 +428,7 @@ EXPORT_PER_CPU_SYMBOL(softnet_data);
>   * PP consumers must pay attention to run APIs in the appropriate context
>   * (e.g. NAPI context).
>   */
> -static DEFINE_PER_CPU_ALIGNED(struct page_pool *, system_page_pool);
> +DEFINE_PER_CPU_ALIGNED(struct page_pool *, system_page_pool);
>  
>  #ifdef CONFIG_LOCKDEP
>  /*
> @@ -11739,12 +11739,20 @@ static int net_page_pool_create(int cpuid)
>  		.pool_size = SYSTEM_PERCPU_PAGE_POOL_SIZE,
>  		.nid = NUMA_NO_NODE,
>  	};
> +	struct xdp_mem_info info;
>  	struct page_pool *pp_ptr;
> +	int err;
>  
>  	pp_ptr = page_pool_create_percpu(&page_pool_params, cpuid);
>  	if (IS_ERR(pp_ptr))
>  		return -ENOMEM;
>  
> +	err = xdp_reg_mem_model(&info, MEM_TYPE_PAGE_POOL, pp_ptr);
> +	if (err) {
> +		page_pool_destroy(pp_ptr);
> +		return err;
> +	}
> +
>  	per_cpu(system_page_pool, cpuid) = pp_ptr;
>  #endif
>  	return 0;
> @@ -11834,12 +11842,15 @@ static int __init net_dev_init(void)
>  out:
>  	if (rc < 0) {
>  		for_each_possible_cpu(i) {
> +			struct xdp_mem_info mem = { .type = MEM_TYPE_PAGE_POOL };
>  			struct page_pool *pp_ptr;
>  
>  			pp_ptr = per_cpu(system_page_pool, i);
>  			if (!pp_ptr)
>  				continue;
>  
> +			mem.id = pp_ptr->xdp_mem_id;
> +			xdp_unreg_mem_model(&mem);

Take it or leave it, a net_page_pool_destroy(int cpuid) would be
symmetric here.

>  			page_pool_destroy(pp_ptr);
>  			per_cpu(system_page_pool, i) = NULL;
>  		}

Acked-by: John Fastabend <john.fastabend@...il.com>