From: Steven Rostedt <rostedt@goodmis.org>
To: Linux Trace Devel <linux-trace-devel@vger.kernel.org>
Subject: [PATCH] libtraceevent: Add warnings if fields are outside the event
Date: Sun, 31 Jul 2022 12:02:47 -0400 [thread overview]
Message-ID: <20220731120247.71528711@rorschach.local.home> (raw)
From: "Steven Rostedt (Google)" <rostedt@goodmis.org>
If a field is referenced outside of the event that is being processed,
do not access that field and warn about it.
This was triggered when there was a bug in the kernel that showed the
wrong offset to the fields that were outside the size of the event. When
trace-cmd used this library to get a field of an event that was at the
end of a sub buffer, it read past the page that was not allocated and
crashed.
Not only warn when this happens, but also just return a zero value and
not read past the event.
Link: https://lore.kernel.org/all/20220731015928.7ab3a154@rorschach.local.home/
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
src/event-parse.c | 28 ++++++++++++++++++++++++++++
1 file changed, 28 insertions(+)
diff --git a/src/event-parse.c b/src/event-parse.c
index b252c4d..b3caff6 100644
--- a/src/event-parse.c
+++ b/src/event-parse.c
@@ -4084,6 +4084,18 @@ static inline void dynamic_offset_field(struct tep_handle *tep,
*offset += field->offset + field->size;
}
+static bool check_data_offset_size(struct tep_event *event, const char *field_name,
+ int data_size, int field_offset, int field_size)
+{
+ /* Check to make sure the field is within the data */
+ if (field_offset + field_size <= data_size)
+ return false;
+
+ tep_warning("Event '%s' field '%s' goes beyond the size of the event (%d > %d)",
+ event->name, field_name, field_offset + field_size, data_size);
+ return true;
+}
+
static unsigned long long
eval_num_arg(void *data, int size, struct tep_event *event, struct tep_print_arg *arg)
{
@@ -4110,6 +4122,12 @@ eval_num_arg(void *data, int size, struct tep_event *event, struct tep_print_arg
if (!arg->field.field)
goto out_warning_field;
}
+ if (check_data_offset_size(event, arg->field.name, size,
+ arg->field.field->offset,
+ arg->field.field->size)) {
+ val = 0;
+ break;
+ }
/* must be a number */
val = tep_read_number(tep, data + arg->field.field->offset,
arg->field.field->size);
@@ -4177,6 +4195,11 @@ eval_num_arg(void *data, int size, struct tep_event *event, struct tep_print_arg
default:
goto default_op; /* oops, all bets off */
}
+ if (check_data_offset_size(event, arg->field.name, size,
+ offset, field_size)) {
+ val = 0;
+ break;
+ }
val = tep_read_number(tep,
data + offset, field_size);
if (typearg)
@@ -4286,6 +4309,11 @@ eval_num_arg(void *data, int size, struct tep_event *event, struct tep_print_arg
/* Without [], we pass the address to the dynamic data */
dynamic_offset_field(tep, arg->dynarray.field, data, size,
&offset, NULL);
+ if (check_data_offset_size(event, arg->field.name, size,
+ offset, field_size)) {
+ val = (unsigned long)data;
+ break;
+ }
val = (unsigned long long)((unsigned long)data + offset);
val = (unsigned long)data + offset;
break;
--
2.35.1
reply other threads:[~2022-07-31 16:02 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20220731120247.71528711@rorschach.local.home \
--to=rostedt@goodmis.org \
--cc=linux-trace-devel@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).