Re: [BUGFIX PATCH v2.1] tracing: probeevent: Fix to support minus offset from symbol

[Masami Hiramatsu]

On Fri, 23 Feb 2018 09:15:42 +0900
Namhyung Kim <namhyung@xxxxxxxxxx> wrote:

> Hi Masami,
> 
> On Thu, Feb 22, 2018 at 04:45:01PM +0900, Masami Hiramatsu wrote:
> > In Documentation/trace/kprobetrace.txt, it says
> > 
> >  @SYM[+|-offs] : Fetch memory at SYM +|- offs (SYM should be a data symbol)
> > 
> > However, the parser doesn't parse minus offset correctly, since
> > commit 2fba0c8867af ("tracing/kprobes: Fix probe offset to be
> > unsigned") drops minus ("-") offset support for kprobe probe
> > address usage.
> > 
> > This fixes the traceprobe_split_symbol_offset() to parse minus
> > offset again with checking the offset range, and add a minus
> > offset check in kprobe probe address usage.
> > 
> > Fixes: 2fba0c8867af ("tracing/kprobes: Fix probe offset to be unsigned")
> > Signed-off-by: Masami Hiramatsu <mhiramat@xxxxxxxxxx>
> > ---
> >  Changes in v2.1:
> >   - Fix to ensure the offset value is less than LONG_MAX.
> >   - Fail if the offset is minus when traceprobe_split_symbol_offset()
> >     is used for the probe address.
> >   - Add Fixes tag.
> > ---
> >  kernel/trace/trace_kprobe.c |    4 ++--
> >  kernel/trace/trace_probe.c  |   15 ++++++++++-----
> >  kernel/trace/trace_probe.h  |    2 +-
> >  3 files changed, 13 insertions(+), 8 deletions(-)
> > 
> > diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c
> > index 5ce9b8cf7be3..b5b1d8aa47d6 100644
> > --- a/kernel/trace/trace_kprobe.c
> > +++ b/kernel/trace/trace_kprobe.c
> > @@ -667,7 +667,7 @@ static int create_trace_kprobe(int argc, char **argv)
> >  	char *symbol = NULL, *event = NULL, *group = NULL;
> >  	int maxactive = 0;
> >  	char *arg;
> > -	unsigned long offset = 0;
> > +	long offset = 0;
> >  	void *addr = NULL;
> >  	char buf[MAX_EVENT_NAME_LEN];
> >  
> > @@ -755,7 +755,7 @@ static int create_trace_kprobe(int argc, char **argv)
> >  		symbol = argv[1];
> >  		/* TODO: support .init module functions */
> >  		ret = traceprobe_split_symbol_offset(symbol, &offset);
> > -		if (ret) {
> > +		if (ret || offset < 0) {
> >  			pr_info("Failed to parse either an address or a symbol.\n");
> >  			return ret;
> >  		}
> > diff --git a/kernel/trace/trace_probe.c b/kernel/trace/trace_probe.c
> > index d59357308677..ec3856147fdd 100644
> > --- a/kernel/trace/trace_probe.c
> > +++ b/kernel/trace/trace_probe.c
> > @@ -320,21 +320,26 @@ static fetch_func_t get_fetch_size_function(const struct fetch_type *type,
> >  }
> >  
> >  /* Split symbol and offset. */
> > -int traceprobe_split_symbol_offset(char *symbol, unsigned long *offset)
> > +int traceprobe_split_symbol_offset(char *symbol, long *offset)
> >  {
> > +	unsigned long ul;
> >  	char *tmp;
> >  	int ret;
> >  
> >  	if (!offset)
> >  		return -EINVAL;
> >  
> > -	tmp = strchr(symbol, '+');
> > +	tmp = strpbrk(symbol, "+-");
> >  	if (tmp) {
> > -		/* skip sign because kstrtoul doesn't accept '+' */
> > -		ret = kstrtoul(tmp + 1, 0, offset);
> > +		ret = kstrtoul(tmp + 1, 0, &ul);
> 
> Why not just using kstrtol() which handles the sign correctly?

Oops, right!

Thank you!

> 
> Thanks,
> Namhyung
> 
> 
> >  		if (ret)
> >  			return ret;
> > -
> > +		if (ul > LONG_MAX)
> > +			return -E2BIG;
> > +		if (*tmp == '-')
> > +			*offset = -ul;
> > +		else
> > +			*offset = ul;
> >  		*tmp = '\0';
> >  	} else
> >  		*offset = 0;
> > diff --git a/kernel/trace/trace_probe.h b/kernel/trace/trace_probe.h
> > index 0745f895f780..75daff22ccea 100644
> > --- a/kernel/trace/trace_probe.h
> > +++ b/kernel/trace/trace_probe.h
> > @@ -365,7 +365,7 @@ extern int traceprobe_conflict_field_name(const char *name,
> >  extern void traceprobe_update_arg(struct probe_arg *arg);
> >  extern void traceprobe_free_probe_arg(struct probe_arg *arg);
> >  
> > -extern int traceprobe_split_symbol_offset(char *symbol, unsigned long *offset);
> > +extern int traceprobe_split_symbol_offset(char *symbol, long *offset);
> >  
> >  /* Sum up total data length for dynamic arraies (strings) */
> >  static nokprobe_inline int
> > 


-- 
Masami Hiramatsu <mhiramat@xxxxxxxxxx>