Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759747Ab3DBMQs (ORCPT ); Tue, 2 Apr 2013 08:16:48 -0400 Received: from seldrel01.sonyericsson.com ([212.209.106.2]:15286 "EHLO seldrel01.sonyericsson.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757255Ab3DBMQr (ORCPT ); Tue, 2 Apr 2013 08:16:47 -0400 From: Date: Tue, 2 Apr 2013 14:16:43 +0200 To: Masami Hiramatsu CC: Joonsoo Kim , "Collett, Toby" , "linux-kernel@vger.kernel.org" , Ananth N Mavinakayanahalli , Anil S Keshavamurthy , "David S. Miller" Subject: Re: [PATCH] kprobe: initialize kprobe_blacklist when it is used firstly Message-ID: <20130402121643.GA21724@caracas.corpusers.net> References: <1364799356-15768-1-git-send-email-iamjoonsoo.kim@lge.com> <515AB7C6.2030807@hitachi.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Disposition: inline In-Reply-To: <515AB7C6.2030807@hitachi.com> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 5701 Lines: 171 On 12:49 Tue 02 Apr , Masami Hiramatsu wrote: > Hi, > > (2013/04/01 15:55), Joonsoo Kim wrote: > > Currently, initializing kprobe_blacklist is done during boot process. > > It takes 230 ms on our android platform and this is significant amount > > for our use case. We can disable CONFIG_KPROBES for production kernel, > > but it is hassle. This kprobe functionality is not commonly used, > > so we don't need to pay this cost at all times. With this rationale, > > change code to initialize kprobe_blacklist when it is used firstly. > > I saw similar patch from Oskar and Toby, and I decided to wait > until his blacklist separating work. I'd like to ask Oskar > how the patch is going first. > I am currently preparing a series of 4 patches including the move of blacklist initialization (same as Joonsoo's patch) and separation of architecture specific blackpoints. The patches are dependent, so I appreciate if you wait a bit. You can expect the series at latest tomorrow. -Oskar > > > > Cc: Ananth N Mavinakayanahalli > > Cc: Anil S Keshavamurthy > > Cc: "David S. Miller" > > Cc: Masami Hiramatsu > > Signed-off-by: Joonsoo Kim > > --- > > I fotgot to add lkml. > > Sorry for noise. > > > > diff --git a/kernel/kprobes.c b/kernel/kprobes.c > > index e35be53..5e90092 100644 > > --- a/kernel/kprobes.c > > +++ b/kernel/kprobes.c > > @@ -101,6 +101,7 @@ static struct kprobe_blackpoint kprobe_blacklist[] = { > > {"mcount",}, /* mcount can be called from everywhere */ > > {NULL} /* Terminator */ > > }; > > +static bool kprobe_blacklist_initialized; > > > > #ifdef __ARCH_WANT_KPROBES_INSN_SLOT > > /* > > @@ -1324,6 +1325,49 @@ out: > > return ret; > > } > > > > +static void __kprobes init_kprobe_blacklist(void) > > +{ > > + unsigned long offset = 0, size = 0; > > + char *modname, namebuf[128]; > > + const char *symbol_name; > > + void *addr; > > + struct kprobe_blackpoint *kb; > > + > > + mutex_lock(&kprobe_mutex); > > + if (kprobe_blacklist_initialized) > > + goto out; > > + > > + /* > > + * Lookup and populate the kprobe_blacklist. > > + * > > + * Unlike the kretprobe blacklist, we'll need to determine > > + * the range of addresses that belong to the said functions, > > + * since a kprobe need not necessarily be at the beginning > > + * of a function. > > + */ > > + for (kb = kprobe_blacklist; kb->name != NULL; kb++) { > > + kprobe_lookup_name(kb->name, addr); > > + if (!addr) > > + continue; > > + > > + kb->start_addr = (unsigned long)addr; > > + symbol_name = kallsyms_lookup(kb->start_addr, > > + &size, &offset, &modname, namebuf); > > + if (!symbol_name) > > + kb->range = 0; > > + else > > + kb->range = size; > > + } > > + > > + /* This guarantee that who see initilized will > > + * get a updated data of kprobe_blacklist */ > > + smp_wmb(); > > + kprobe_blacklist_initialized = true; > > + > > +out: > > + mutex_unlock(&kprobe_mutex); > > +} > > + > > static int __kprobes in_kprobes_functions(unsigned long addr) > > { > > struct kprobe_blackpoint *kb; > > @@ -1331,6 +1375,7 @@ static int __kprobes in_kprobes_functions(unsigned long addr) > > if (addr >= (unsigned long)__kprobes_text_start && > > addr < (unsigned long)__kprobes_text_end) > > return -EINVAL; > > + > > /* > > * If there exists a kprobe_blacklist, verify and > > * fail any probe registration in the prohibited area > > @@ -1476,6 +1521,9 @@ int __kprobes register_kprobe(struct kprobe *p) > > struct module *probed_mod; > > kprobe_opcode_t *addr; > > > > + if (unlikely(!kprobe_blacklist_initialized)) > > + init_kprobe_blacklist(); > > + > > /* Adjust probe address from symbol */ > > addr = kprobe_addr(p); > > if (IS_ERR(addr)) > > @@ -2065,11 +2113,6 @@ static struct notifier_block kprobe_module_nb = { > > static int __init init_kprobes(void) > > { > > int i, err = 0; > > - unsigned long offset = 0, size = 0; > > - char *modname, namebuf[128]; > > - const char *symbol_name; > > - void *addr; > > - struct kprobe_blackpoint *kb; > > > > /* FIXME allocate the probe table, currently defined statically */ > > /* initialize all list heads */ > > @@ -2079,28 +2122,6 @@ static int __init init_kprobes(void) > > raw_spin_lock_init(&(kretprobe_table_locks[i].lock)); > > } > > > > - /* > > - * Lookup and populate the kprobe_blacklist. > > - * > > - * Unlike the kretprobe blacklist, we'll need to determine > > - * the range of addresses that belong to the said functions, > > - * since a kprobe need not necessarily be at the beginning > > - * of a function. > > - */ > > - for (kb = kprobe_blacklist; kb->name != NULL; kb++) { > > - kprobe_lookup_name(kb->name, addr); > > - if (!addr) > > - continue; > > - > > - kb->start_addr = (unsigned long)addr; > > - symbol_name = kallsyms_lookup(kb->start_addr, > > - &size, &offset, &modname, namebuf); > > - if (!symbol_name) > > - kb->range = 0; > > - else > > - kb->range = size; > > - } > > - > > if (kretprobe_blacklist_size) { > > /* lookup the function address from its name */ > > for (i = 0; kretprobe_blacklist[i].name != NULL; i++) { > > > > > -- > Masami HIRAMATSU > IT Management Research Dept. Linux Technology Center > Hitachi, Ltd., Yokohama Research Laboratory > E-mail: masami.hiramatsu.pt@hitachi.com > > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/