linux-kernel - Re: [PATCH v2 1/2] perf sdt: add scanning of sdt probles arguments

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: <20161125144002.GC16426@kernel.org>
Date:   Fri, 25 Nov 2016 11:40:02 -0300
From:   Arnaldo Carvalho de Melo <acme@...nel.org>
To:     Alexis Berlemont <alexis.berlemont@...il.com>
Cc:     linux-kernel@...r.kernel.org, peterz@...radead.org,
        mingo@...hat.com, alexander.shishkin@...ux.intel.com,
        hemant@...ux.vnet.ibm.com
Subject: Re: [PATCH v2 1/2] perf sdt: add scanning of sdt probles arguments

Em Sat, Nov 19, 2016 at 12:56:36AM +0100, Alexis Berlemont escreveu:
> During a "perf buildid-cache --add" command, the section
> ".note.stapsdt" of the "added" binary is scanned in order to list the
> available SDT markers available in a binary. The parts containing the
> probes arguments were left unscanned.
> 
> The whole section is now parsed; the probe arguments are extracted for
> later use.
> 
> Signed-off-by: Alexis Berlemont <alexis.berlemont@...il.com>
> ---
>  tools/perf/util/symbol-elf.c | 16 +++++++++++++++-
>  tools/perf/util/symbol.h     |  1 +
>  2 files changed, 16 insertions(+), 1 deletion(-)
> 
> diff --git a/tools/perf/util/symbol-elf.c b/tools/perf/util/symbol-elf.c
> index 99400b0..0fbe0b2 100644
> --- a/tools/perf/util/symbol-elf.c
> +++ b/tools/perf/util/symbol-elf.c
> @@ -1822,7 +1822,7 @@ void kcore_extract__delete(struct kcore_extract *kce)
>  static int populate_sdt_note(Elf **elf, const char *data, size_t len,
>  			     struct list_head *sdt_notes)
>  {
> -	const char *provider, *name;
> +	const char *provider, *name, *args;
>  	struct sdt_note *tmp = NULL;
>  	GElf_Ehdr ehdr;
>  	GElf_Addr base_off = 0;
> @@ -1881,6 +1881,20 @@ static int populate_sdt_note(Elf **elf, const char *data, size_t len,
>  		goto out_free_prov;
>  	}
>  
> +	args = (const char *)memchr(name, '\0', data + len - name);

Humm, no need for casting?

> +
> +	/*
> +	 * There is no argument if:
> +	 * - We reached the end of the note;
> +	 * - There is not enough room to hold a potential string;
> +	 * - The argument string is empty or just contains ':'.
> +	 */
> +	if (args == NULL || data + len - args < 2 ||
> +		args[1] == ':' || args[1] == '\0')
> +		tmp->args = NULL;
> +	else
> +		tmp->args = strdup(++args);

Shouldn't we check this and do error back propagation? I.e. if there are
args and we don't handle them, silently, that looks bad

> +
>  	if (gelf_getclass(*elf) == ELFCLASS32) {
>  		memcpy(&tmp->addr, &buf, 3 * sizeof(Elf32_Addr));
>  		tmp->bit32 = true;
> diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
> index 2d0a905..913be07 100644
> --- a/tools/perf/util/symbol.h
> +++ b/tools/perf/util/symbol.h
> @@ -347,6 +347,7 @@ int arch__choose_best_symbol(struct symbol *syma, struct symbol *symb);
>  struct sdt_note {
>  	char *name;			/* name of the note*/
>  	char *provider;			/* provider name */
> +	char *args;
>  	bool bit32;			/* whether the location is 32 bits? */
>  	union {				/* location, base and semaphore addrs */
>  		Elf64_Addr a64[3];
> -- 
> 2.10.2