linux-kernel - Re: [PATCH bpf-next v4 1/4] fprobe: use rhltable for fprobe_ip

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]

Message-Id: <20250816235023.4dabfbc13a46a859de61cf4d@kernel.org>
Date: Sat, 16 Aug 2025 23:50:23 +0900
From: Masami Hiramatsu (Google) <mhiramat@...nel.org>
To: Menglong Dong <menglong8.dong@...il.com>
Cc: olsajiri@...il.com, rostedt@...dmis.org, mathieu.desnoyers@...icios.com,
 hca@...ux.ibm.com, revest@...omium.org, linux-kernel@...r.kernel.org,
 linux-trace-kernel@...r.kernel.org, bpf@...r.kernel.org
Subject: Re: [PATCH bpf-next v4 1/4] fprobe: use rhltable for
 fprobe_ip_table

Hi Menglong,

Sorry, one more thing.

> @@ -260,14 +263,12 @@ static int fprobe_entry(struct ftrace_graph_ent *trace, struct fgraph_ops *gops,
>  	if (WARN_ON_ONCE(!fregs))
>  		return 0;
>  
> -	first = node = find_first_fprobe_node(func);
> -	if (unlikely(!first))
> -		return 0;
> -
> +	rcu_read_lock();

Actually, we don't need these rcu_read_lock() in this function, because
the caller function_graph_enter_regs() uses ftrace_test_recursion_trylock()
which disables preemption. Thus we don't need to do this again here.

> +	head = rhltable_lookup(&fprobe_ip_table, &func, fprobe_rht_params);
>  	reserved_words = 0;
> -	hlist_for_each_entry_from_rcu(node, hlist) {
> +	rhl_for_each_entry_rcu(node, pos, head, hlist) {
>  		if (node->addr != func)
> -			break;
> +			continue;
>  		fp = READ_ONCE(node->fp);
>  		if (!fp || !fp->exit_handler)
>  			continue;
> @@ -278,17 +279,19 @@ static int fprobe_entry(struct ftrace_graph_ent *trace, struct fgraph_ops *gops,
>  		reserved_words +=
>  			FPROBE_HEADER_SIZE_IN_LONG + SIZE_IN_LONG(fp->entry_data_size);
>  	}
> -	node = first;
> +	rcu_read_unlock();
>  	if (reserved_words) {
>  		fgraph_data = fgraph_reserve_data(gops->idx, reserved_words * sizeof(long));
>  		if (unlikely(!fgraph_data)) {
> -			hlist_for_each_entry_from_rcu(node, hlist) {
> +			rcu_read_lock();

Ditto.

> +			rhl_for_each_entry_rcu(node, pos, head, hlist) {
>  				if (node->addr != func)
> -					break;
> +					continue;
>  				fp = READ_ONCE(node->fp);
>  				if (fp && !fprobe_disabled(fp))
>  					fp->nmissed++;
>  			}
> +			rcu_read_unlock();
>  			return 0;
>  		}
>  	}

Thank you,


-- 
Masami Hiramatsu (Google) <mhiramat@...nel.org>