netdev - Re: [PATCH v5 perf,bpf 12/15] perf, bpf: enable annotation of bpf program

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:   Mon, 4 Mar 2019 20:43:25 +0000
From:   Song Liu <songliubraving@...com>
To:     Jiri Olsa <jolsa@...hat.com>
CC:     Networking <netdev@...r.kernel.org>,
        linux-kernel <linux-kernel@...r.kernel.org>,
        "ast@...nel.org" <ast@...nel.org>,
        "daniel@...earbox.net" <daniel@...earbox.net>,
        Kernel Team <Kernel-team@...com>,
        "peterz@...radead.org" <peterz@...radead.org>,
        "acme@...hat.com" <acme@...hat.com>,
        "jolsa@...nel.org" <jolsa@...nel.org>,
        "namhyung@...nel.org" <namhyung@...nel.org>
Subject: Re: [PATCH v5 perf,bpf 12/15] perf, bpf: enable annotation of bpf
 program



> On Mar 4, 2019, at 12:37 PM, Jiri Olsa <jolsa@...hat.com> wrote:
> 
> On Mon, Mar 04, 2019 at 07:45:00PM +0000, Song Liu wrote:
>> 
>> 
>>> On Mar 4, 2019, at 5:53 AM, Jiri Olsa <jolsa@...hat.com> wrote:
>>> 
>>> On Wed, Feb 27, 2019 at 09:06:40PM -0800, Song Liu wrote:
>>> 
>>> SNIP
>>> 
>>>> +	fflush(s);
>>>> +	do {
>>>> +		const struct bpf_line_info *linfo = NULL;
>>>> +		struct disasm_line *dl;
>>>> +		size_t prev_buf_size;
>>>> +		const char *srcline;
>>>> +		u64 addr;
>>>> +
>>>> +		addr = pc + ((u64 *)(info_linear->info.jited_ksyms))[sub_id];
>>>> +		count = disassemble(pc, &info);
>>>> +
>>>> +		if (prog_linfo)
>>>> +			linfo = bpf_prog_linfo__lfind_addr_func(prog_linfo,
>>>> +								addr, sub_id,
>>>> +								nr_skip);
>>>> +
>>>> +		if (linfo && btf) {
>>>> +			srcline = btf__name_by_offset(btf, linfo->line_off);
>>>> +			nr_skip++;
>>>> +		} else
>>>> +			srcline = NULL;
>>>> +
>>>> +		fprintf(s, "\n");
>>>> +		prev_buf_size = buf_size;
>>>> +		fflush(s);
>>>> +
>>>> +		if (!opts->hide_src_code && srcline) {
>>>> +			args->offset = -1;
>>> 
>>> why's the offset not needed in here?
>> 
>> This is the source code line. Based on my understanding of annotation
>> code, we put -1 for source code line. Did I misunderstand something?
> 
> please double check, IIRC annotation_line__print will print
> just more info if offset is present, together with the source,
> but I haven't checked that code for some time, so I might be wrong
> 
> jirka

In symbol__parse_objdump_line(), default offset is -1. And it is only 
updated if line_ip is not -1. And line_ip for src code line is -1, so 
offset for the source line is -1. 

It works fine in the tests when I toggle source code on/off with key
's'. 

Thanks
Song