linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Anton Arapov <anton@redhat.com>
To: Anton Arapov <anton@redhat.com>, Oleg Nesterov <oleg@redhat.com>,
	Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: LKML <linux-kernel@vger.kernel.org>,
	Josh Stone <jistone@redhat.com>, Frank Eigler <fche@redhat.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@elte.hu>,
	Ananth N Mavinakayanahalli <ananth@in.ibm.com>,
	adrian.m.negreanu@intel.com, Torsten.Polle@gmx.de
Subject: [PATCH 5/7] uretprobes: return probe exit, invoke handlers
Date: Fri, 22 Mar 2013 14:09:02 +0100	[thread overview]
Message-ID: <1363957745-6657-6-git-send-email-anton@redhat.com> (raw)
In-Reply-To: <1363957745-6657-1-git-send-email-anton@redhat.com>

  Uretprobe handlers are invoked when the trampoline is hit, on completion the
trampoline is replaced with the saved return address and the uretprobe instance
deleted.

RFCv6 changes:
  - rework handle_uretprobe()

RFCv5 changes:
  - switch to simply linked list ->return_uprobes
  - rework handle_uretprobe()

RFCv4 changes:
  - check, whether utask is not NULL in handle_uretprobe()
  - get rid of area->rp_trampoline_vaddr
  - minor handle_uretprobe() fixups

RFCv3 changes:
  - protected uprobe with refcounter. See put_uprobe() in handle_uretprobe()
    that reflects increment in prepare_uretprobe()

RFCv2 changes:
  - get rid of ->return_consumers member from struct uprobe, introduce
    rp_handler() in consumer instead

Signed-off-by: Anton Arapov <anton@redhat.com>
---
 kernel/events/uprobes.c | 60 ++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 59 insertions(+), 1 deletion(-)

diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
index 4ea3e91..91edd2c 100644
--- a/kernel/events/uprobes.c
+++ b/kernel/events/uprobes.c
@@ -1623,6 +1623,56 @@ static void handler_chain(struct uprobe *uprobe, struct pt_regs *regs)
 	up_read(&uprobe->register_rwsem);
 }
 
+static void handler_uretprobe_chain(struct uprobe *uprobe, struct pt_regs *regs)
+{
+	struct uprobe_consumer *uc;
+
+	down_read(&uprobe->register_rwsem);
+	for (uc = uprobe->consumers; uc; uc = uc->next) {
+		if (uc->rp_handler)
+			uc->rp_handler(uc, regs);
+	}
+	up_read(&uprobe->register_rwsem);
+}
+
+static void handle_uretprobe(struct xol_area *area, struct pt_regs *regs)
+{
+	struct uprobe_task *utask;
+	struct return_instance *ri, *tmp;
+	unsigned long prev_ret_vaddr;
+
+	utask = get_utask();
+	if (!utask)
+		return;
+
+	ri = utask->return_instances;
+	if (!ri)
+		return;
+
+	instruction_pointer_set(regs, ri->orig_ret_vaddr);
+
+	while (ri) {
+		if (ri->uprobe->consumers)
+			handler_uretprobe_chain(ri->uprobe, regs);
+
+		put_uprobe(ri->uprobe);
+		tmp = ri;
+		prev_ret_vaddr = tmp->orig_ret_vaddr;
+		ri = ri->next;
+		kfree(tmp);
+
+		if (!ri || ri->dirty == false) {
+			/*
+			 * This is the first return uprobe (chronologically)
+			 * pushed for this particular instance of the probed
+			 * function.
+			 */
+			utask->return_instances = ri;
+			return;
+		}
+	}
+}
+
 /*
  * Run handler and ask thread to singlestep.
  * Ensure all non-fatal signals cannot interrupt thread while it singlesteps.
@@ -1631,11 +1681,19 @@ static void handle_swbp(struct pt_regs *regs)
 {
 	struct uprobe *uprobe;
 	unsigned long bp_vaddr;
+	struct xol_area *area;
 	int uninitialized_var(is_swbp);
 
 	bp_vaddr = uprobe_get_swbp_addr(regs);
-	uprobe = find_active_uprobe(bp_vaddr, &is_swbp);
+	area = get_xol_area();
+	if (area) {
+		if (bp_vaddr == get_trampoline_vaddr(area)) {
+			handle_uretprobe(area, regs);
+			return;
+		}
+	}
 
+	uprobe = find_active_uprobe(bp_vaddr, &is_swbp);
 	if (!uprobe) {
 		if (is_swbp > 0) {
 			/* No matching uprobe; signal SIGTRAP. */
-- 
1.8.1.4


  parent reply	other threads:[~2013-03-22 13:09 UTC|newest]

Thread overview: 50+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-03-22 13:08 [PATCH 0/7] uretprobes: return probes implementation Anton Arapov
2013-03-22 13:08 ` [PATCH 1/7] uretprobes: preparation patch Anton Arapov
2013-03-23 17:42   ` Oleg Nesterov
2013-03-22 13:08 ` [PATCH 2/7] uretprobes: extract fill_page() and trampoline implementation Anton Arapov
2013-03-24 14:41   ` Oleg Nesterov
2013-03-24 18:20     ` [PATCH 0/5] kmap cleanups for uretprobes (Was: extract fill_page() and trampoline implementation) Oleg Nesterov
2013-03-24 18:21       ` [PATCH 1/5] uprobes: Turn copy_opcode() into copy_from_page() Oleg Nesterov
2013-03-25 10:30         ` Anton Arapov
2013-03-26 11:59         ` Srikar Dronamraju
2013-03-24 18:21       ` [PATCH 2/5] uprobes: Change __copy_insn() to use copy_from_page() Oleg Nesterov
2013-03-25 10:31         ` Anton Arapov
2013-03-26 12:00         ` Srikar Dronamraju
2013-03-24 18:21       ` [PATCH 3/5] uprobes: Kill the unnecesary filp != NULL check in __copy_insn() Oleg Nesterov
2013-03-25 10:31         ` Anton Arapov
2013-03-26 12:00         ` Srikar Dronamraju
2013-03-24 18:21       ` [PATCH 4/5] uprobes: Introduce copy_to_page() Oleg Nesterov
2013-03-25 10:31         ` Anton Arapov
2013-03-26 12:02         ` Srikar Dronamraju
2013-03-24 18:21       ` [PATCH 5/5] uprobes: Change write_opcode() to use copy_*page() Oleg Nesterov
2013-03-25 10:31         ` Anton Arapov
2013-03-26 11:59         ` Srikar Dronamraju
2013-03-25 10:30       ` [PATCH 0/5] kmap cleanups for uretprobes (Was: extract fill_page() and trampoline implementation) Anton Arapov
2013-03-25 11:58     ` [PATCH 2/7] uretprobes: extract fill_page() and trampoline implementation Oleg Nesterov
2013-03-22 13:09 ` [PATCH 3/7] uretprobes/x86: hijack return address Anton Arapov
2013-03-24 14:59   ` Oleg Nesterov
2013-03-22 13:09 ` [PATCH 4/7] uretprobes: return probe entry, prepare_uretprobe() Anton Arapov
2013-03-22 15:02   ` Oleg Nesterov
2013-03-26 12:26     ` Anton Arapov
2013-03-26 14:34       ` Oleg Nesterov
2013-03-23 17:46   ` Oleg Nesterov
2013-03-24 15:26   ` Oleg Nesterov
2013-03-25 15:51     ` Anton Arapov
2013-03-26  8:45     ` Anton Arapov
2013-03-26  8:50       ` Anton Arapov
2013-03-22 13:09 ` Anton Arapov [this message]
2013-03-24 16:28   ` [PATCH 5/7] uretprobes: return probe exit, invoke handlers Oleg Nesterov
2013-03-25 12:31     ` Oleg Nesterov
2013-03-25 15:49     ` Anton Arapov
2013-03-25 16:38       ` Oleg Nesterov
2013-03-26  8:36         ` Anton Arapov
2013-03-22 13:09 ` [PATCH 6/7] uretprobes: limit the depth of return probe nestedness Anton Arapov
2013-03-24 16:54   ` Oleg Nesterov
2013-03-22 13:09 ` [PATCH 7/7] uretprobes: implemented, thus remove -ENOSYS Anton Arapov
2013-03-22 13:13   ` Anton Arapov
2013-03-22 13:09 ` [PATCH 7/7] uretprobes: remove -ENOSYS as return probes implemented Anton Arapov
2013-03-22 15:10 ` [PATCH 0/7] uretprobes: return probes implementation Oleg Nesterov
2013-03-22 21:40   ` Josh Stone
2013-03-23  6:43     ` Anton Arapov
2013-03-23 18:04       ` Oleg Nesterov
2013-03-23 17:56     ` Oleg Nesterov

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=1363957745-6657-6-git-send-email-anton@redhat.com \
    --to=anton@redhat.com \
    --cc=Torsten.Polle@gmx.de \
    --cc=adrian.m.negreanu@intel.com \
    --cc=ananth@in.ibm.com \
    --cc=fche@redhat.com \
    --cc=jistone@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=oleg@redhat.com \
    --cc=peterz@infradead.org \
    --cc=srikar@linux.vnet.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;
as well as URLs for NNTP newsgroup(s).