From: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: lkml <linux-kernel@vger.kernel.org>,
Prasanna S Panchamukhi <prasanna@in.ibm.com>,
Jim Keniston <jkenisto@us.ibm.com>,
Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>,
Maneesh Soni <maneesh@in.ibm.com>
Subject: Re: [PATCH] Kprobes: Print details of kretprobe on assertion failure
Date: Tue, 3 Apr 2007 11:19:43 +0530 [thread overview]
Message-ID: <20070403054943.GA18459@in.ibm.com> (raw)
In-Reply-To: <20070402141732.166f50bd.akpm@linux-foundation.org>
On Mon, Apr 02, 2007 at 02:17:32PM -0700, Andrew Morton wrote:
> On Mon, 2 Apr 2007 14:56:36 +0530
> Ananth N Mavinakayanahalli <ananth@in.ibm.com> wrote:
>
> > From: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
> >
> > In certain cases like when the real return address can't be found or
> > when the number of tracked calls to a kretprobed function is less than
> > the number of returns, we may not be able to find the correct return
> > address after processing a kretprobe. Currently we just do a BUG_ON, but
> > no information is provided about the actual failing kretprobe.
> >
> > Print out details of the kretprobe before calling BUG().
> >
> > Signed-off-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
> >
> > ---
> > arch/i386/kernel/kprobes.c | 6 +++++-
> > arch/ia64/kernel/kprobes.c | 7 ++++++-
> > arch/powerpc/kernel/kprobes.c | 7 ++++++-
> > arch/x86_64/kernel/kprobes.c | 7 ++++++-
> > 4 files changed, 23 insertions(+), 4 deletions(-)
> >
> > Index: linux-2.6.21-rc5/arch/i386/kernel/kprobes.c
> > ===================================================================
> > --- linux-2.6.21-rc5.orig/arch/i386/kernel/kprobes.c
> > +++ linux-2.6.21-rc5/arch/i386/kernel/kprobes.c
> > @@ -440,7 +440,11 @@ fastcall void *__kprobes trampoline_hand
> > break;
> > }
> >
> > - BUG_ON(!orig_ret_address || (orig_ret_address == trampoline_address));
> > + if (!orig_ret_address || (orig_ret_address == trampoline_address)) {
> > + printk("kretprobe BUG!: Processing kretprobe %p @ %p\n",
> > + ri->rp, ri->rp->kp.addr);
> > + BUG();
> > + }
> >
> > spin_unlock_irqrestore(&kretprobe_lock, flags);
> >
> > Index: linux-2.6.21-rc5/arch/ia64/kernel/kprobes.c
> > ===================================================================
> > --- linux-2.6.21-rc5.orig/arch/ia64/kernel/kprobes.c
> > +++ linux-2.6.21-rc5/arch/ia64/kernel/kprobes.c
> > @@ -444,7 +444,12 @@ int __kprobes trampoline_probe_handler(s
> > break;
> > }
> >
> > - BUG_ON(!orig_ret_address || (orig_ret_address == trampoline_address));
> > + if (!orig_ret_address || (orig_ret_address == trampoline_address)) {
> > + printk("kretprobe BUG!: Processing kretprobe %p @ %p\n",
> > + ri->rp, ri->rp->kp.addr);
> > + BUG();
> > + }
> > +
> > regs->cr_iip = orig_ret_address;
> >
> > reset_current_kprobe();
> > Index: linux-2.6.21-rc5/arch/powerpc/kernel/kprobes.c
> > ===================================================================
> > --- linux-2.6.21-rc5.orig/arch/powerpc/kernel/kprobes.c
> > +++ linux-2.6.21-rc5/arch/powerpc/kernel/kprobes.c
> > @@ -293,7 +293,12 @@ int __kprobes trampoline_probe_handler(s
> > break;
> > }
> >
> > - BUG_ON(!orig_ret_address || (orig_ret_address == trampoline_address));
> > + if (!orig_ret_address || (orig_ret_address == trampoline_address)) {
> > + printk("kretprobe BUG!: Processing kretprobe %p @ %p\n",
> > + ri->rp, ri->rp->kp.addr);
> > + BUG();
> > + }
> > +
> > regs->nip = orig_ret_address;
> >
> > reset_current_kprobe();
> > Index: linux-2.6.21-rc5/arch/x86_64/kernel/kprobes.c
> > ===================================================================
> > --- linux-2.6.21-rc5.orig/arch/x86_64/kernel/kprobes.c
> > +++ linux-2.6.21-rc5/arch/x86_64/kernel/kprobes.c
> > @@ -438,7 +438,12 @@ int __kprobes trampoline_probe_handler(s
> > break;
> > }
> >
> > - BUG_ON(!orig_ret_address || (orig_ret_address == trampoline_address));
> > + if (!orig_ret_address || (orig_ret_address == trampoline_address)) {
> > + printk("kretprobe BUG!: Processing kretprobe %p @ %p\n",
> > + ri->rp, ri->rp->kp.addr);
> > + BUG();
> > + }
> > +
> > regs->rip = orig_ret_address;
> >
>
> A lot of copying-and-pasting there. Would it be better if this assertion
> was performed in a library function in kernel/kprobes.c?
Indeed. Here is the updated patch...
From: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
In certain cases like when the real return address can't be found or
when the number of tracked calls to a kretprobed function is less than
the number of returns, we may not be able to find the correct return
address after processing a kretprobe. Currently we just do a BUG_ON, but
no information is provided about the actual failing kretprobe.
Print out details of the kretprobe before calling BUG().
Signed-off-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
---
arch/i386/kernel/kprobes.c | 3 +--
arch/ia64/kernel/kprobes.c | 3 ++-
arch/powerpc/kernel/kprobes.c | 2 +-
arch/x86_64/kernel/kprobes.c | 2 +-
include/linux/kprobes.h | 10 ++++++++++
5 files changed, 15 insertions(+), 5 deletions(-)
Index: linux-2.6.21-rc5/arch/i386/kernel/kprobes.c
===================================================================
--- linux-2.6.21-rc5.orig/arch/i386/kernel/kprobes.c
+++ linux-2.6.21-rc5/arch/i386/kernel/kprobes.c
@@ -440,8 +440,7 @@ fastcall void *__kprobes trampoline_hand
break;
}
- BUG_ON(!orig_ret_address || (orig_ret_address == trampoline_address));
-
+ kretprobe_assert(ri, orig_ret_address, trampoline_address);
spin_unlock_irqrestore(&kretprobe_lock, flags);
hlist_for_each_entry_safe(ri, node, tmp, &empty_rp, hlist) {
Index: linux-2.6.21-rc5/arch/ia64/kernel/kprobes.c
===================================================================
--- linux-2.6.21-rc5.orig/arch/ia64/kernel/kprobes.c
+++ linux-2.6.21-rc5/arch/ia64/kernel/kprobes.c
@@ -444,7 +444,8 @@ int __kprobes trampoline_probe_handler(s
break;
}
- BUG_ON(!orig_ret_address || (orig_ret_address == trampoline_address));
+ kretprobe_assert(ri, orig_ret_address, trampoline_address);
+
regs->cr_iip = orig_ret_address;
reset_current_kprobe();
Index: linux-2.6.21-rc5/arch/powerpc/kernel/kprobes.c
===================================================================
--- linux-2.6.21-rc5.orig/arch/powerpc/kernel/kprobes.c
+++ linux-2.6.21-rc5/arch/powerpc/kernel/kprobes.c
@@ -293,7 +293,7 @@ int __kprobes trampoline_probe_handler(s
break;
}
- BUG_ON(!orig_ret_address || (orig_ret_address == trampoline_address));
+ kretprobe_assert(ri, orig_ret_address, trampoline_address);
regs->nip = orig_ret_address;
reset_current_kprobe();
Index: linux-2.6.21-rc5/arch/x86_64/kernel/kprobes.c
===================================================================
--- linux-2.6.21-rc5.orig/arch/x86_64/kernel/kprobes.c
+++ linux-2.6.21-rc5/arch/x86_64/kernel/kprobes.c
@@ -438,7 +438,7 @@ int __kprobes trampoline_probe_handler(s
break;
}
- BUG_ON(!orig_ret_address || (orig_ret_address == trampoline_address));
+ kretprobe_assert(ri, orig_ret_address, trampoline_address);
regs->rip = orig_ret_address;
reset_current_kprobe();
Index: linux-2.6.21-rc5/include/linux/kprobes.h
===================================================================
--- linux-2.6.21-rc5.orig/include/linux/kprobes.h
+++ linux-2.6.21-rc5/include/linux/kprobes.h
@@ -158,6 +158,16 @@ struct kretprobe_instance {
struct task_struct *task;
};
+static inline void kretprobe_assert(struct kretprobe_instance *ri,
+ unsigned long orig_ret_address, unsigned long trampoline_address)
+{
+ if (!orig_ret_address || (orig_ret_address == trampoline_address)) {
+ printk("kretprobe BUG!: Processing kretprobe %p @ %p\n",
+ ri->rp, ri->rp->kp.addr);
+ BUG();
+ }
+}
+
extern spinlock_t kretprobe_lock;
extern struct mutex kprobe_mutex;
extern int arch_prepare_kprobe(struct kprobe *p);
prev parent reply other threads:[~2007-04-03 5:46 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-04-02 9:26 [PATCH] Kprobes: Print details of kretprobe on assertion failure Ananth N Mavinakayanahalli
2007-04-02 21:17 ` Andrew Morton
2007-04-03 5:49 ` Ananth N Mavinakayanahalli [this message]
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=20070403054943.GA18459@in.ibm.com \
--to=ananth@in.ibm.com \
--cc=akpm@linux-foundation.org \
--cc=anil.s.keshavamurthy@intel.com \
--cc=jkenisto@us.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=maneesh@in.ibm.com \
--cc=prasanna@in.ibm.com \
/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