This is the mail archive of the
systemtap@sourceware.org
mailing list for the systemtap project.
Re: [RFC] [PATCH] To vunmap correct address in text_poke()(kprobes)
- From: "S. P. Prasanna" <prasanna at in dot ibm dot com>
- To: Srinivasa Ds <srinivasa at in dot ibm dot com>
- Cc: linux-kernel at vger dot kernel dot org, Andrew Morton <akpm at linux-foundation dot org>, ananth at in dot ibm dot com, anil dot s dot keshavamurthy at intel dot com, jkenisto at us dot ibm dot com, systemtap at sources dot redhat dot com, suzuki at in dot ibm dot com
- Date: Wed, 25 Jul 2007 19:41:17 +0530
- Subject: Re: [RFC] [PATCH] To vunmap correct address in text_poke()(kprobes)
- References: <200707251221.49827.srinivasa@in.ibm.com>
- Reply-to: prasanna at in dot ibm dot com
On Wed, Jul 25, 2007 at 12:21:49PM +0530, Srinivasa Ds wrote:
>
> When I was testing kprobes on x86_64 and I come across the below error message
> on latest 2.6.23-rc1 kernel.
> ==========================================
> Trying to vfree() bad address (ffffc20002233199)
> WARNING: at mm/vmalloc.c:330 __vunmap()
>
> Call Trace:
> [<ffffffff8023d199>] sys_gettimeofday+0x0/0x62
> [<ffffffff8047636c>] text_poke+0x119/0x124
> [<ffffffff80476b42>] arch_arm_kprobe+0x1c/0x21
> [<ffffffff80477cfd>] __register_kprobe+0x28a/0x2ed
> [<ffffffff8815a039>] :gettimeofday:kprobe_init+0x39/0x65
> [<ffffffff8025b5cf>] sys_init_module+0x1626/0x1788
> [<ffffffff802aaf12>] dput+0x3f/0xfa
> [<ffffffff8026bf91>] audit_syscall_entry+0x141/0x174
> [<ffffffff8020bee5>] tracesys+0xdc/0xe1
>
> ======================================================
>
> This indicates that vunmap() is not receving the page-aligned address in
> text_poke(). So the below attached patch will address this issue. Please let
> me know your comments.
>
> Signed-off-by: Srinivasa DS <srinivasa@in.ibm.com>
> Signed-off-by: Suzuki K P <suzuki@in.ibm.com>
>
>
> Index: linux-2.6.23-rc1/arch/i386/kernel/alternative.c
> ===================================================================
> --- linux-2.6.23-rc1.orig/arch/i386/kernel/alternative.c 2007-07-23
> 02:11:00.000000000 +0530
> +++ linux-2.6.23-rc1/arch/i386/kernel/alternative.c 2007-07-25
> 11:45:53.000000000 +0530
> @@ -447,5 +447,5 @@ void __kprobes text_poke(void *oaddr, un
> if (cpu_has_clflush)
> asm("clflush (%0) " :: "r" (oaddr) : "memory");
> if (addr != oaddr)
> - vunmap(addr);
> + vunmap(addr-(((unsigned long)oaddr) % PAGE_SIZE));
It is appropriate to use PAGE_MASK rather than doing all this,
as shown below.
vunmap((u8 *)((unsigned long)addr & PAGE_MASK));
Thanks
Prasanna
--
Prasanna S.P.
Linux Technology Center
India Software Labs, IBM Bangalore
Email: prasanna@in.ibm.com
Ph: 91-80-41776329