linux-kernel - [PATCH] tracing: Fix sleeping while atomic in kdb ftdump

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>] [day] [month] [year] [list]

Message-Id: <20220708170919.1.I75844e5038d9425add2ad853a608cb44bb39df40@changeid>
Date:   Fri,  8 Jul 2022 17:09:52 -0700
From:   Douglas Anderson <dianders@...omium.org>
To:     Steven Rostedt <rostedt@...dmis.org>,
        Ingo Molnar <mingo@...hat.com>
Cc:     kgdb-bugreport@...ts.sourceforge.net,
        Daniel Thompson <daniel.thompson@...aro.org>,
        Douglas Anderson <dianders@...omium.org>,
        Masami Hiramatsu <mhiramat@...nel.org>,
        linux-kernel@...r.kernel.org
Subject: [PATCH] tracing: Fix sleeping while atomic in kdb ftdump

If you drop into kdb and type "ftdump" you'll get a sleeping while
atomic warning from memory allocation in trace_find_next_entry().

This appears to have been caused by commit ff895103a84a ("tracing:
Save off entry when peeking at next entry"), which added the
allocation in that path. The problematic commit was already fixed by
commit 8e99cf91b99b ("tracing: Do not allocate buffer in
trace_find_next_entry() in atomic") but that fix missed the kdb case.

The fix here is easy: just move the assignment of the static buffer to
the place where it should have been to begin with:
trace_init_global_iter(). That function is called in two places, once
is right before the assignment of the static buffer added by the
previous fix and once is in kdb.

Note that it appears that there's a second static buffer that we need
to assign that was added in commit efbbdaa22bb7 ("tracing: Show real
address for trace event arguments"), so we'll move that too.

Fixes: ff895103a84a ("tracing: Save off entry when peeking at next entry")
Fixes: efbbdaa22bb7 ("tracing: Show real address for trace event arguments")
Signed-off-by: Douglas Anderson <dianders@...omium.org>
---

 kernel/trace/trace.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index 2c95992e2c71..64700ad2866b 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -9866,6 +9866,12 @@ void trace_init_global_iter(struct trace_iterator *iter)
 	/* Output in nanoseconds only if we are using a clock in nanoseconds. */
 	if (trace_clocks[iter->tr->clock_id].in_ns)
 		iter->iter_flags |= TRACE_FILE_TIME_IN_NS;
+
+	/* Can not use kmalloc for iter.temp and iter.fmt */
+	iter->temp = static_temp_buf;
+	iter->temp_size = STATIC_TEMP_BUF_SIZE;
+	iter->fmt = static_fmt_buf;
+	iter->fmt_size = STATIC_FMT_BUF_SIZE;
 }

 void ftrace_dump(enum ftrace_dump_mode oops_dump_mode)
@@ -9898,11 +9904,6 @@ void ftrace_dump(enum ftrace_dump_mode oops_dump_mode)

 	/* Simulate the iterator */
 	trace_init_global_iter(&iter);
-	/* Can not use kmalloc for iter.temp and iter.fmt */
-	iter.temp = static_temp_buf;
-	iter.temp_size = STATIC_TEMP_BUF_SIZE;
-	iter.fmt = static_fmt_buf;
-	iter.fmt_size = STATIC_FMT_BUF_SIZE;

 	for_each_tracing_cpu(cpu) {
 		atomic_inc(&per_cpu_ptr(iter.array_buffer->data, cpu)->disabled);
-- 
2.37.0.rc0.161.g10f37bed90-goog