Re: [PATCH 2/3] tracing: Fix parse_synth_field() error handling

From: Axel Rasmussen
Date: Tue Sep 29 2020 - 17:57:24 EST


On Tue, Sep 29, 2020 at 1:33 PM Tom Zanussi <zanussi@xxxxxxxxxx> wrote:
>
> synth_field_size() returns either the size or an error. However, the
> code assigns the return val to ssize_t which is unsigned, and then
> tests whether it's less than 0, which it isn't so discards the error.

I think the patch is correct, but the commit message is not.
field->size is a size_t (unsigned), not an ssize_t (signed). I think
this should say instead something like:

synth_field_size() returns either a positive size or an error (zero or
a negative value). However, the existing code assumes the only error
value is 0. It doesn't handle negative error codes, as it assigns
directly to field->size (a size_t; unsigned), thereby interpreting the
error code as a valid size instead.

>
> Do the test before assignment to field->size.
>
> Signed-off-by: Tom Zanussi <zanussi@xxxxxxxxxx>
> ---
> kernel/trace/trace_events_synth.c | 6 ++++--
> 1 file changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/kernel/trace/trace_events_synth.c b/kernel/trace/trace_events_synth.c
> index a9cd7793f7ea..6e7282c7b530 100644
> --- a/kernel/trace/trace_events_synth.c
> +++ b/kernel/trace/trace_events_synth.c
> @@ -465,6 +465,7 @@ static struct synth_field *parse_synth_field(int argc, const char **argv,
> struct synth_field *field;
> const char *prefix = NULL, *field_type = argv[0], *field_name, *array;
> int len, ret = 0;
> + int size;

Why not make this an ssize_t

>
> if (field_type[0] == ';')
> field_type++;
> @@ -520,11 +521,12 @@ static struct synth_field *parse_synth_field(int argc, const char **argv,
> field->type[len - 1] = '\0';
> }
>
> - field->size = synth_field_size(field->type);
> - if (!field->size) {
> + size = synth_field_size(field->type);
> + if (size < 0) {
> ret = -EINVAL;
> goto free;
> }
> + field->size = size;
>
> if (synth_field_is_string(field->type))
> field->is_string = true;
> --
> 2.17.1
>