[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <20220820134316.156058831@goodmis.org>
Date: Sat, 20 Aug 2022 09:43:16 -0400
From: Steven Rostedt <rostedt@...dmis.org>
To: linux-kernel@...r.kernel.org
Cc: Ingo Molnar <mingo@...nel.org>,
Andrew Morton <akpm@...ux-foundation.org>,
Masami Hiramatsu <mhiramat@...nel.org>,
Tzvetomir Stoyanov <tz.stoyanov@...il.com>,
Tom Zanussi <zanussi@...nel.org>
Subject: [PATCH v2 0/6] tracing/eprobes: Fixes for unexpected arguments
While using eprobes, I decided to entertain the thougth of what would
happen if I tried to get the instruction pointer "%rip", knowing full
well that eprobes do not have access to pt_regs. Well, I found out, and
it led me down a rabbit hole of bugs.
This series fixes those bugs, by not allowing register access for eprobes,
and also filling the holes of @symbol and @immediate argument.
Changes since v1: https://lore.kernel.org/all/20220820014035.531145719@goodmis.org/
- Fixed parenthesis warning
- Fixed comment about comm arguments
- Made kprobes and eprobes process "$COMM" as well as "$comm"
- Made filters consistent with histograms with "common_cpu"
Steven Rostedt (Google) (6):
tracing/eprobes: Do not allow eprobes to use $stack, or % for regs
tracing/eprobes: Do not hardcode $comm as a string
tracing/eprobes: Fix reading of string fields
tracing/eprobes: Have event probes be consistent with kprobes and uprobes
tracing/probes: Have kprobes and uprobes use $COMM too
tracing: Have filter accept "common_cpu" to be consistent
----
kernel/trace/trace_eprobe.c | 91 ++++++++++++++++++++++++++++++++++++++++++---
kernel/trace/trace_events.c | 1 +
kernel/trace/trace_probe.c | 29 +++++++++------
3 files changed, 104 insertions(+), 17 deletions(-)
Powered by blists - more mailing lists