lkml.org 
[lkml]   [2012]   [Mar]   [23]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
Subject[PATCH] x86: stop recursive fault in print_context_stack after stack overflow
After printing out the first line of a stack backtrace,
print_context_stack() calls print_ftrace_graph_addr() to check if it's
making a graph of function calls, usually not the case. But unfortunate
ordering of assignments causes this to oops if an earlier stack overflow
corrupted threadinfo->task. Reorder to avoid that irritation.

(The fact that there was a stack overflow may often be more interesting
than the stack that can now be shown; but integrating that information
with this stacktrace is awkward, so leave it to overflow reporting.)

Signed-off-by: Hugh Dickins <hughd@google.com>
---

arch/x86/kernel/dumpstack.c | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)

--- linux.git/arch/x86/kernel/dumpstack.c 2012-03-18 16:15:34.000000000 -0700
+++ linux/arch/x86/kernel/dumpstack.c 2012-03-23 10:43:28.196084684 -0700
@@ -37,13 +37,16 @@ print_ftrace_graph_addr(unsigned long ad
const struct stacktrace_ops *ops,
struct thread_info *tinfo, int *graph)
{
- struct task_struct *task = tinfo->task;
+ struct task_struct *task;
unsigned long ret_addr;
- int index = task->curr_ret_stack;
+ int index;

if (addr != (unsigned long)return_to_handler)
return;

+ task = tinfo->task;
+ index = task->curr_ret_stack;
+
if (!task->ret_stack || index < *graph)
return;


\
 
 \ /
  Last update: 2012-03-23 22:05    [W:0.022 / U:0.040 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site