Linux MIPS Architecture development
 help / color / mirror / Atom feed
From: tsbogend@alpha.franken.de (Thomas Bogendoerfer)
To: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Cc: ralf@linux-mips.org, linux-mips@linux-mips.org
Subject: Re: Breakage in arch/mips/kernel/traps.c for 64bit
Date: Sun, 4 May 2008 00:48:49 +0200	[thread overview]
Message-ID: <20080503224849.GA2314@alpha.franken.de> (raw)
In-Reply-To: <20080504.011647.93019265.anemo@mba.ocn.ne.jp>

On Sun, May 04, 2008 at 01:16:47AM +0900, Atsushi Nemoto wrote:
> On Fri, 2 May 2008 11:11:13 +0100, Ralf Baechle <ralf@linux-mips.org> wrote:
> > It came as part of 39b8d5254246ac56342b72f812255c8f7a74dca9 which is a
> > patch amalgated from several other patches.  Below is the original patch
> > it came with.  I think the idea of the patch is valid but the idea needs a
> > bit of mending.
> 
> Then how about this fix?

hmm, why not simply use __get_user() when accessing the stack content ?
show_stacktrace() already does it for stack dumping ? This would
avoid any work for whatever sick stack mappings. Below is a patch,
which does this.

Thomas.

The newly added check for valid stack pointer address breaks at least for
64bit kernels.  Use __get_user() for accessing stack content to avoid crashes,
when doing the backtrace.

Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
---

 arch/mips/kernel/traps.c |   16 ++++++++--------
 1 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/arch/mips/kernel/traps.c b/arch/mips/kernel/traps.c
index cb8b0e2..c9ce8d6 100644
--- a/arch/mips/kernel/traps.c
+++ b/arch/mips/kernel/traps.c
@@ -81,22 +81,22 @@ void (*board_bind_eic_interrupt)(int irq, int regset);
 
 static void show_raw_backtrace(unsigned long reg29)
 {
-	unsigned long *sp = (unsigned long *)(reg29 & ~3);
+	unsigned long __user *sp = (unsigned long __user *)(reg29 & ~3);
 	unsigned long addr;
 
 	printk("Call Trace:");
 #ifdef CONFIG_KALLSYMS
 	printk("\n");
 #endif
-#define IS_KVA01(a) ((((unsigned int)a) & 0xc0000000) == 0x80000000)
-	if (IS_KVA01(sp)) {
-		while (!kstack_end(sp)) {
-			addr = *sp++;
-			if (__kernel_text_address(addr))
-				print_ip_sym(addr);
+	while (!kstack_end(sp)) {
+		if (__get_user(addr, sp++)) {
+			printk(" (Bad stack address)");
+			break;
 		}
-		printk("\n");
+		if (__kernel_text_address(addr))
+			print_ip_sym(addr);
 	}
+	printk("\n");
 }
 
 #ifdef CONFIG_KALLSYMS

-- 
Crap can work. Given enough thrust pigs will fly, but it's not necessary a
good idea.                                                [ RFC1925, 2.3 ]

  parent reply	other threads:[~2008-05-03 22:49 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-05-01 16:33 Breakage in arch/mips/kernel/traps.c for 64bit Thomas Bogendoerfer
2008-05-01 21:01 ` Maciej W. Rozycki
2008-05-02 10:11 ` Ralf Baechle
2008-05-03 16:16   ` Atsushi Nemoto
2008-05-03 17:39     ` Ralf Baechle
2008-05-03 19:57       ` Maciej W. Rozycki
2008-05-03 22:48     ` Thomas Bogendoerfer [this message]
2008-05-04 13:39       ` Atsushi Nemoto
2008-05-04 22:08         ` Thomas Bogendoerfer
2008-05-05 14:58           ` Atsushi Nemoto

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=20080503224849.GA2314@alpha.franken.de \
    --to=tsbogend@alpha.franken.de \
    --cc=anemo@mba.ocn.ne.jp \
    --cc=linux-mips@linux-mips.org \
    --cc=ralf@linux-mips.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