LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
* [PATCH] printk: implement printk_buf overflow warning
@ 2008-02-13  9:03 Tejun Heo
  2008-02-13 23:59 ` Andrew Morton
  0 siblings, 1 reply; 3+ messages in thread
From: Tejun Heo @ 2008-02-13  9:03 UTC (permalink / raw)
  To: Linus Torvalds, Andrew Morton, Randy Dunlap, Linux Kernel

printk silently truncates messages longer than 1024 - 1 bytes.
Implement overflow detection and append "$PRINTK_BUF_OVERFLOW$\n" to
truncated messages.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 kernel/printk.c |   15 ++++++++++++---
 1 files changed, 12 insertions(+), 3 deletions(-)

diff --git a/kernel/printk.c b/kernel/printk.c
index 074a3ea..419cd47 100644
--- a/kernel/printk.c
+++ b/kernel/printk.c
@@ -42,6 +42,7 @@ void __attribute__((weak)) early_printk(const char *fmt, ...)
 {
 }
 
+#define PRINTK_BUF_LEN	1024
 #define __LOG_BUF_LEN	(1 << CONFIG_LOG_BUF_SHIFT)
 
 /* printk's without a loglevel use this.. */
@@ -619,9 +620,10 @@ asmlinkage int vprintk(const char *fmt, va_list args)
 	static volatile unsigned int printk_cpu = UINT_MAX;
 	static const char printk_recursion_bug_msg [] =
 		KERN_CRIT "BUG: recent printk recursion!\n";
+	static const char overflow_tag[] = "$PRINTK_BUF_OVERFLOW$\n";
 	static int printk_recursion_bug;
 	static int log_level_unknown = 1;
-	static char printk_buf[1024];
+	static char printk_buf[PRINTK_BUF_LEN + sizeof(overflow_tag) - 1];
 
 	unsigned long flags;
 	int printed_len = 0;
@@ -663,8 +665,15 @@ asmlinkage int vprintk(const char *fmt, va_list args)
 		printed_len = sizeof(printk_recursion_bug_msg);
 	}
 	/* Emit the output into the temporary buffer */
-	printed_len += vscnprintf(printk_buf + printed_len,
-				  sizeof(printk_buf) - printed_len, fmt, args);
+	printed_len += vsnprintf(printk_buf + printed_len,
+				 PRINTK_BUF_LEN - printed_len, fmt, args);
+
+	/* Check for overflow and tag the message if overflowed */
+	if (printed_len >= PRINTK_BUF_LEN) {
+		memcpy(printk_buf + PRINTK_BUF_LEN - 1, overflow_tag,
+		       sizeof(overflow_tag));
+		printed_len = sizeof(printk_buf) - 1;
+	}
 
 	/*
 	 * Copy the output into log_buf.  If the caller didn't provide
-- 
1.5.2.4


^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] printk: implement printk_buf overflow warning
  2008-02-13  9:03 [PATCH] printk: implement printk_buf overflow warning Tejun Heo
@ 2008-02-13 23:59 ` Andrew Morton
  2008-02-14  0:14   ` Tejun Heo
  0 siblings, 1 reply; 3+ messages in thread
From: Andrew Morton @ 2008-02-13 23:59 UTC (permalink / raw)
  To: Tejun Heo; +Cc: torvalds, randy.dunlap, linux-kernel

On Wed, 13 Feb 2008 18:03:12 +0900
Tejun Heo <htejun@gmail.com> wrote:

> printk silently truncates messages longer than 1024 - 1 bytes.
> Implement overflow detection and append "$PRINTK_BUF_OVERFLOW$\n" to
> truncated messages.

I don't think I've ever heard of anyone managing to overflow the printk
buffer and it's hard to imagine how a sane person could manage to do this.

I assume this is happening because of changes in your other patchset, and
that this change is hence associated with those changes and should probably
be a part of the same patch series?

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] printk: implement printk_buf overflow warning
  2008-02-13 23:59 ` Andrew Morton
@ 2008-02-14  0:14   ` Tejun Heo
  0 siblings, 0 replies; 3+ messages in thread
From: Tejun Heo @ 2008-02-14  0:14 UTC (permalink / raw)
  To: Andrew Morton; +Cc: torvalds, randy.dunlap, linux-kernel

Andrew Morton wrote:
> On Wed, 13 Feb 2008 18:03:12 +0900
> Tejun Heo <htejun@gmail.com> wrote:
> 
>> printk silently truncates messages longer than 1024 - 1 bytes.
>> Implement overflow detection and append "$PRINTK_BUF_OVERFLOW$\n" to
>> truncated messages.
> 
> I don't think I've ever heard of anyone managing to overflow the printk
> buffer and it's hard to imagine how a sane person could manage to do this.
> 
> I assume this is happening because of changes in your other patchset, and
> that this change is hence associated with those changes and should probably
> be a part of the same patch series?

I thought this went well with the recursion warning.  I'll move it into
the mprintk patchset.

Thanks.

-- 
tejun

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2008-02-14  0:14 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2008-02-13  9:03 [PATCH] printk: implement printk_buf overflow warning Tejun Heo
2008-02-13 23:59 ` Andrew Morton
2008-02-14  0:14   ` Tejun Heo

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).