linux-kernel - Re: [PATCH 08/10] perf tool: Unify handling of features when writing feature section

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: <20111206133630.GI7059@infradead.org>
Date:	Tue, 6 Dec 2011 11:36:30 -0200
From:	Arnaldo Carvalho de Melo <acme@...hat.com>
To:	Robert Richter <robert.richter@....com>
Cc:	Ingo Molnar <mingo@...e.hu>, Peter Zijlstra <peterz@...radead.org>,
	Stephane Eranian <eranian@...gle.com>,
	Frederic Weisbecker <fweisbec@...il.com>,
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 08/10] perf tool: Unify handling of features when
	writing feature section

Em Tue, Dec 06, 2011 at 11:32:38AM +0100, Robert Richter escreveu:
> The features HEADER_TRACE_INFO and HEADER_BUILD_ID are handled
> different when writing the feature section. All other features are
> simply disabled on failure and writing the section goes on without
> returning an error. There is no reason for these special cases. This
> patch unifies handling of the features.
> 
> This should be ok since all features can be parsed independently.
> Offset and size of a feature's block is stored in struct perf_file_
> section right after the data block of perf.data (see perf_session__
> write_header()). Thus, if a feature does not exist then other features
> can be processed anyway.
> 
> Also moving special code for HEADER_BUILD_ID out to write_build_id().

For this one I just would add a big fat warning that if there are
build-ids on the system but the table can't be written, then extreme
care has to be taken when doing a perf report.

I.e. one has to be completely sure that the binaries hasn't changed if
not validating the build-ids.

If you fix that please add this warning as well when no build-ids are
found, which hopefully is the odd case these days as all distros I'm
aware of have build-ids in all DSOs.

- Arnaldo

> Cc: Frederic Weisbecker <fweisbec@...il.com>
> Cc: Stephane Eranian <eranian@...gle.com>
> Signed-off-by: Robert Richter <robert.richter@....com>
> ---
>  tools/perf/util/header.c |   14 +++++---------
>  1 files changed, 5 insertions(+), 9 deletions(-)
> 
> diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
> index dda3c5f..5cfe06a 100644
> --- a/tools/perf/util/header.c
> +++ b/tools/perf/util/header.c
> @@ -445,6 +445,9 @@ static int write_build_id(int fd, struct perf_header *h,
>  
>  	session = container_of(h, struct perf_session, header);
>  
> +	if (!perf_session__read_build_ids(session, true))
> +		return -1;
> +
>  	err = dsos__write_buildid_table(h, fd);
>  	if (err < 0) {
>  		pr_debug("failed to write buildid table\n");
> @@ -1413,10 +1416,6 @@ static int perf_header__adds_write(struct perf_header *header,
>  
>  	session = container_of(header, struct perf_session, header);
>  
> -	if (perf_header__has_feat(header, HEADER_BUILD_ID &&
> -	    !perf_session__read_build_ids(session, true)))
> -		perf_header__clear_feat(header, HEADER_BUILD_ID);
> -
>  	nr_sections = bitmap_weight(header->adds_features, HEADER_FEAT_BITS);
>  	if (!nr_sections)
>  		return 0;
> @@ -1432,13 +1431,11 @@ static int perf_header__adds_write(struct perf_header *header,
>  
>  	err = do_write_feat(fd, header, HEADER_TRACE_INFO, &p, evlist);
>  	if (err)
> -		goto out_free;
> +		perf_header__clear_feat(header, HEADER_TRACE_INFO);
>  
>  	err = do_write_feat(fd, header, HEADER_BUILD_ID, &p, evlist);
> -	if (err) {
> +	if (err)
>  		perf_header__clear_feat(header, HEADER_BUILD_ID);
> -		goto out_free;
> -	}
>  
>  	err = do_write_feat(fd, header, HEADER_HOSTNAME, &p, evlist);
>  	if (err)
> @@ -1496,7 +1493,6 @@ static int perf_header__adds_write(struct perf_header *header,
>  	err = do_write(fd, feat_sec, sec_size);
>  	if (err < 0)
>  		pr_debug("failed to write feature section\n");
> -out_free:
>  	free(feat_sec);
>  	return err;
>  }
> -- 
> 1.7.7
> 
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/