[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20111110184237.11361.87061.sendpatchset@srdronam.in.ibm.com>
Date: Fri, 11 Nov 2011 00:12:37 +0530
From: Srikar Dronamraju <srikar@...ux.vnet.ibm.com>
To: Peter Zijlstra <peterz@...radead.org>,
Linus Torvalds <torvalds@...ux-foundation.org>
Cc: Oleg Nesterov <oleg@...hat.com>,
Andrew Morton <akpm@...ux-foundation.org>,
LKML <linux-kernel@...r.kernel.org>,
Linux-mm <linux-mm@...ck.org>, Ingo Molnar <mingo@...e.hu>,
Andi Kleen <andi@...stfloor.org>,
Christoph Hellwig <hch@...radead.org>,
Steven Rostedt <rostedt@...dmis.org>,
Roland McGrath <roland@...k.frob.com>,
Thomas Gleixner <tglx@...utronix.de>,
Masami Hiramatsu <masami.hiramatsu.pt@...achi.com>,
Arnaldo Carvalho de Melo <acme@...radead.org>,
Anton Arapov <anton@...hat.com>,
Ananth N Mavinakayanahalli <ananth@...ibm.com>,
Jim Keniston <jkenisto@...ux.vnet.ibm.com>,
Stephen Wilson <wilsons@...rt.ca>
Subject: [PATCH v6 3.2-rc1 25/28] uprobes: call post_xol() unconditionally
Kill sstep_complete(), change uprobe_notify_resume() to use
post_xol() unconditionally.
It is wrong to assume that regs->ip always changes after the step.
rep or jmp/call to self for example. We know that this task has
already done the step, we can rely on DIE_DEBUG notification.
Original-patch-from: Oleg Nesterov <oleg@...hat.com>
Signed-off-by: Srikar Dronamraju <srikar@...ux.vnet.ibm.com>
---
include/linux/uprobes.h | 3 ++-
kernel/uprobes.c | 40 ++++++++++++----------------------------
2 files changed, 14 insertions(+), 29 deletions(-)
diff --git a/include/linux/uprobes.h b/include/linux/uprobes.h
index add5222..70d639c 100644
--- a/include/linux/uprobes.h
+++ b/include/linux/uprobes.h
@@ -74,7 +74,8 @@ struct uprobe {
enum uprobe_task_state {
UTASK_RUNNING,
UTASK_BP_HIT,
- UTASK_SSTEP
+ UTASK_SSTEP,
+ UTASK_SSTEP_ACK,
};
/*
diff --git a/kernel/uprobes.c b/kernel/uprobes.c
index 9e73cef..13b1d68 100644
--- a/kernel/uprobes.c
+++ b/kernel/uprobes.c
@@ -1316,24 +1316,6 @@ static int pre_ssout(struct uprobe *uprobe, struct pt_regs *regs,
}
/*
- * Verify from Instruction Pointer if singlestep has indeed occurred.
- * If Singlestep has occurred, then do post singlestep fix-ups.
- */
-static bool sstep_complete(struct uprobe *uprobe, struct pt_regs *regs)
-{
- unsigned long vaddr = instruction_pointer(regs);
-
- /*
- * If we have executed out of line, Instruction pointer
- * cannot be same as virtual address of XOL slot.
- */
- if (vaddr == current->utask->xol_vaddr)
- return false;
- post_xol(uprobe, regs);
- return true;
-}
-
-/*
* uprobe_notify_resume gets called in task context just before returning
* to userspace.
*
@@ -1381,17 +1363,18 @@ void uprobe_notify_resume(struct pt_regs *regs)
else
/* Cannot Singlestep; re-execute the instruction. */
goto cleanup_ret;
- } else if (utask->state == UTASK_SSTEP) {
+ } else {
u = utask->active_uprobe;
- if (sstep_complete(u, regs)) {
- put_uprobe(u);
- utask->active_uprobe = NULL;
- utask->state = UTASK_RUNNING;
- user_disable_single_step(current);
- xol_free_insn_slot(current);
-
- /* TODO Stop queueing signals. */
- }
+ if (utask->state == UTASK_SSTEP_ACK)
+ post_xol(u, regs);
+ else
+ WARN_ON_ONCE(1);
+
+ put_uprobe(u);
+ utask->active_uprobe = NULL;
+ utask->state = UTASK_RUNNING;
+ user_disable_single_step(current);
+ xol_free_insn_slot(current);
}
return;
@@ -1441,6 +1424,7 @@ int uprobe_post_notifier(struct pt_regs *regs)
/* task is currently not uprobed */
return 0;
+ utask->state = UTASK_SSTEP_ACK;
set_thread_flag(TIF_UPROBE);
return 1;
}
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists