From: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
To: Peter Zijlstra <peterz@infradead.org>, Ingo Molnar <mingo@elte.hu>
Cc: Steven Rostedt <rostedt@goodmis.org>,
Srikar Dronamraju <srikar@linux.vnet.ibm.com>,
Linux-mm <linux-mm@kvack.org>,
Arnaldo Carvalho de Melo <acme@infradead.org>,
Linus Torvalds <torvalds@linux-foundation.org>,
Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>,
Hugh Dickins <hughd@google.com>,
Christoph Hellwig <hch@infradead.org>,
Ananth N Mavinakayanahalli <ananth@in.ibm.com>,
Thomas Gleixner <tglx@linutronix.de>,
Jonathan Corbet <corbet@lwn.net>, Oleg Nesterov <oleg@redhat.com>,
LKML <linux-kernel@vger.kernel.org>,
Jim Keniston <jkenisto@linux.vnet.ibm.com>,
Roland McGrath <roland@hack.frob.com>,
Andi Kleen <andi@firstfloor.org>,
Andrew Morton <akpm@linux-foundation.org>
Subject: [PATCH v5 3.1.0-rc4-tip 17/26] x86: arch specific hooks for pre/post singlestep handling.
Date: Tue, 20 Sep 2011 17:33:25 +0530 [thread overview]
Message-ID: <20110920120325.25326.11641.sendpatchset@srdronam.in.ibm.com> (raw)
In-Reply-To: <20110920115938.25326.93059.sendpatchset@srdronam.in.ibm.com>
Hooks for handling pre singlestepping and post singlestepping.
Signed-off-by: Jim Keniston <jkenisto@us.ibm.com>
Signed-off-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
---
arch/x86/include/asm/uprobes.h | 2 +
arch/x86/kernel/uprobes.c | 138 ++++++++++++++++++++++++++++++++++++++++
2 files changed, 140 insertions(+), 0 deletions(-)
diff --git a/arch/x86/include/asm/uprobes.h b/arch/x86/include/asm/uprobes.h
index 2ad2c71..1c30cfd 100644
--- a/arch/x86/include/asm/uprobes.h
+++ b/arch/x86/include/asm/uprobes.h
@@ -47,6 +47,8 @@ struct uprobe_task_arch_info {};
struct uprobe;
extern int analyze_insn(struct task_struct *tsk, struct uprobe *uprobe);
extern void set_instruction_pointer(struct pt_regs *regs, unsigned long vaddr);
+extern int pre_xol(struct uprobe *uprobe, struct pt_regs *regs);
+extern int post_xol(struct uprobe *uprobe, struct pt_regs *regs);
extern int uprobe_exception_notify(struct notifier_block *self,
unsigned long val, void *data);
#endif /* _ASM_UPROBES_H */
diff --git a/arch/x86/kernel/uprobes.c b/arch/x86/kernel/uprobes.c
index 8ec759a..da1bc12 100644
--- a/arch/x86/kernel/uprobes.c
+++ b/arch/x86/kernel/uprobes.c
@@ -25,6 +25,7 @@
#include <linux/sched.h>
#include <linux/ptrace.h>
#include <linux/uprobes.h>
+#include <linux/uaccess.h>
#include <linux/kdebug.h>
#include <asm/insn.h>
@@ -395,6 +396,143 @@ void set_instruction_pointer(struct pt_regs *regs, unsigned long vaddr)
}
/*
+ * pre_xol - prepare to execute out of line.
+ * @uprobe: the probepoint information.
+ * @regs: reflects the saved user state of @tsk.
+ *
+ * If we're emulating a rip-relative instruction, save the contents
+ * of the scratch register and store the target address in that register.
+ *
+ * Returns true if @uprobe->opcode is @bkpt_insn.
+ */
+#ifdef CONFIG_X86_64
+int pre_xol(struct uprobe *uprobe, struct pt_regs *regs)
+{
+ struct uprobe_task_arch_info *tskinfo = ¤t->utask->tskinfo;
+
+ regs->ip = current->utask->xol_vaddr;
+ if (uprobe->fixups & UPROBES_FIX_RIP_AX) {
+ tskinfo->saved_scratch_register = regs->ax;
+ regs->ax = current->utask->vaddr;
+ regs->ax += uprobe->arch_info.rip_rela_target_address;
+ } else if (uprobe->fixups & UPROBES_FIX_RIP_CX) {
+ tskinfo->saved_scratch_register = regs->cx;
+ regs->cx = current->utask->vaddr;
+ regs->cx += uprobe->arch_info.rip_rela_target_address;
+ }
+ return 0;
+}
+#else
+int pre_xol(struct uprobe *uprobe, struct pt_regs *regs)
+{
+ regs->ip = current->utask->xol_vaddr;
+ return 0;
+}
+#endif
+
+/*
+ * Called by post_xol() to adjust the return address pushed by a call
+ * instruction executed out of line.
+ */
+static int adjust_ret_addr(unsigned long sp, long correction)
+{
+ int rasize, ncopied;
+ long ra = 0;
+
+ if (is_32bit_app(current))
+ rasize = 4;
+ else
+ rasize = 8;
+ ncopied = copy_from_user(&ra, (void __user *) sp, rasize);
+ if (unlikely(ncopied))
+ goto fail;
+ ra += correction;
+ ncopied = copy_to_user((void __user *) sp, &ra, rasize);
+ if (unlikely(ncopied))
+ goto fail;
+ return 0;
+
+fail:
+ pr_warn_once("uprobes: Failed to adjust return address after"
+ " single-stepping call instruction;"
+ " pid=%d, sp=%#lx\n", current->pid, sp);
+ return -EFAULT;
+}
+
+#ifdef CONFIG_X86_64
+static bool is_riprel_insn(struct uprobe *uprobe)
+{
+ return ((uprobe->fixups &
+ (UPROBES_FIX_RIP_AX | UPROBES_FIX_RIP_CX)) != 0);
+}
+
+static void handle_riprel_post_xol(struct uprobe *uprobe,
+ struct pt_regs *regs, long *correction)
+{
+ if (is_riprel_insn(uprobe)) {
+ struct uprobe_task_arch_info *tskinfo;
+ tskinfo = ¤t->utask->tskinfo;
+
+ if (uprobe->fixups & UPROBES_FIX_RIP_AX)
+ regs->ax = tskinfo->saved_scratch_register;
+ else
+ regs->cx = tskinfo->saved_scratch_register;
+ /*
+ * The original instruction includes a displacement, and so
+ * is 4 bytes longer than what we've just single-stepped.
+ * Fall through to handle stuff like "jmpq *...(%rip)" and
+ * "callq *...(%rip)".
+ */
+ *correction += 4;
+ }
+}
+#else
+static void handle_riprel_post_xol(struct uprobe *uprobe,
+ struct pt_regs *regs, long *correction)
+{
+}
+#endif
+
+/*
+ * Called after single-stepping. To avoid the SMP problems that can
+ * occur when we temporarily put back the original opcode to
+ * single-step, we single-stepped a copy of the instruction.
+ *
+ * This function prepares to resume execution after the single-step.
+ * We have to fix things up as follows:
+ *
+ * Typically, the new ip is relative to the copied instruction. We need
+ * to make it relative to the original instruction (FIX_IP). Exceptions
+ * are return instructions and absolute or indirect jump or call instructions.
+ *
+ * If the single-stepped instruction was a call, the return address that
+ * is atop the stack is the address following the copied instruction. We
+ * need to make it the address following the original instruction (FIX_CALL).
+ *
+ * If the original instruction was a rip-relative instruction such as
+ * "movl %edx,0xnnnn(%rip)", we have instead executed an equivalent
+ * instruction using a scratch register -- e.g., "movl %edx,(%rax)".
+ * We need to restore the contents of the scratch register and adjust
+ * the ip, keeping in mind that the instruction we executed is 4 bytes
+ * shorter than the original instruction (since we squeezed out the offset
+ * field). (FIX_RIP_AX or FIX_RIP_CX)
+ */
+int post_xol(struct uprobe *uprobe, struct pt_regs *regs)
+{
+ struct uprobe_task *utask = current->utask;
+ int result = 0;
+ long correction;
+
+ correction = (long)(utask->vaddr - utask->xol_vaddr);
+ handle_riprel_post_xol(uprobe, regs, &correction);
+ if (uprobe->fixups & UPROBES_FIX_IP)
+ regs->ip += correction;
+ if (uprobe->fixups & UPROBES_FIX_CALL)
+ result = adjust_ret_addr(regs->sp, correction);
+ return result;
+}
+
+/*
* Wrapper routine for handling exceptions.
*/
int uprobe_exception_notify(struct notifier_block *self,
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2011-09-20 12:17 UTC|newest]
Thread overview: 160+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-09-20 11:59 [PATCH v5 3.1.0-rc4-tip 0/26] Uprobes patchset with perf probe support Srikar Dronamraju
2011-09-20 11:59 ` [PATCH v5 3.1.0-rc4-tip 1/26] uprobes: Auxillary routines to insert, find, delete uprobes Srikar Dronamraju
2011-09-20 15:42 ` Stefan Hajnoczi
2011-09-26 11:18 ` Peter Zijlstra
2011-09-26 11:59 ` Srikar Dronamraju
2011-09-26 11:18 ` Peter Zijlstra
2011-09-26 12:02 ` Srikar Dronamraju
2011-09-26 13:35 ` Peter Zijlstra
2011-09-26 16:19 ` Srikar Dronamraju
2011-09-20 12:00 ` [PATCH v5 3.1.0-rc4-tip 2/26] Uprobes: Allow multiple consumers for an uprobe Srikar Dronamraju
2011-09-26 12:29 ` Peter Zijlstra
2011-09-20 12:00 ` [PATCH v5 3.1.0-rc4-tip 3/26] Uprobes: register/unregister probes Srikar Dronamraju
2011-09-20 16:50 ` Stefan Hajnoczi
2011-09-21 4:07 ` Srikar Dronamraju
2011-09-26 13:15 ` Peter Zijlstra
2011-09-26 13:23 ` Srikar Dronamraju
2011-10-03 12:46 ` Oleg Nesterov
2011-10-05 17:04 ` Srikar Dronamraju
2011-10-05 18:50 ` Oleg Nesterov
2011-10-06 6:51 ` Srikar Dronamraju
2011-10-07 17:03 ` Oleg Nesterov
2011-09-20 12:00 ` [PATCH v5 3.1.0-rc4-tip 4/26] uprobes: Define hooks for mmap/munmap Srikar Dronamraju
2011-09-20 17:03 ` Stefan Hajnoczi
2011-09-21 4:03 ` Srikar Dronamraju
2011-09-26 13:53 ` Peter Zijlstra
2011-09-26 15:44 ` Srikar Dronamraju
2011-09-27 11:37 ` Peter Zijlstra
2011-09-27 13:08 ` Srikar Dronamraju
2011-09-27 11:41 ` Peter Zijlstra
2011-09-27 12:59 ` Srikar Dronamraju
2011-09-27 11:42 ` Peter Zijlstra
2011-10-03 13:37 ` Oleg Nesterov
2011-10-06 11:05 ` Srikar Dronamraju
2011-10-07 17:36 ` Oleg Nesterov
2011-10-10 12:31 ` Srikar Dronamraju
2011-09-20 12:00 ` [PATCH v5 3.1.0-rc4-tip 5/26] Uprobes: copy of the original instruction Srikar Dronamraju
2011-10-03 16:29 ` Oleg Nesterov
2011-10-05 10:52 ` Srikar Dronamraju
2011-10-05 15:11 ` Oleg Nesterov
2011-10-05 16:09 ` Srikar Dronamraju
2011-10-05 17:53 ` Oleg Nesterov
2011-09-20 12:01 ` [PATCH v5 3.1.0-rc4-tip 6/26] Uprobes: define fixups Srikar Dronamraju
2011-09-20 12:01 ` [PATCH v5 3.1.0-rc4-tip 7/26] Uprobes: uprobes arch info Srikar Dronamraju
2011-09-20 12:01 ` [PATCH v5 3.1.0-rc4-tip 8/26] x86: analyze instruction and determine fixups Srikar Dronamraju
2011-09-20 17:13 ` Stefan Hajnoczi
2011-09-20 18:12 ` Christoph Hellwig
2011-09-20 20:53 ` Stefan Hajnoczi
2011-09-23 11:53 ` Masami Hiramatsu
2011-09-23 16:51 ` Stefan Hajnoczi
2011-09-26 19:59 ` Josh Stone
2011-09-27 1:32 ` Masami Hiramatsu
2011-09-27 2:59 ` Josh Stone
2011-09-27 7:08 ` Stefan Hajnoczi
2011-09-22 1:05 ` Josh Stone
2011-10-05 15:48 ` Oleg Nesterov
2011-10-05 16:12 ` Srikar Dronamraju
2011-09-20 12:01 ` [PATCH v5 3.1.0-rc4-tip 9/26] Uprobes: Background page replacement Srikar Dronamraju
2011-10-05 16:19 ` Oleg Nesterov
2011-10-06 6:53 ` Srikar Dronamraju
2011-09-20 12:01 ` [PATCH v5 3.1.0-rc4-tip 10/26] x86: Set instruction pointer Srikar Dronamraju
2011-10-05 16:29 ` Oleg Nesterov
2011-09-20 12:02 ` [PATCH v5 3.1.0-rc4-tip 11/26] x86: Introduce TIF_UPROBE FLAG Srikar Dronamraju
2011-09-20 12:02 ` [PATCH v5 3.1.0-rc4-tip 12/26] Uprobes: Handle breakpoint and Singlestep Srikar Dronamraju
2011-09-26 13:59 ` Peter Zijlstra
2011-09-26 16:01 ` Srikar Dronamraju
2011-09-26 16:25 ` Peter Zijlstra
2011-10-05 17:48 ` Oleg Nesterov
2011-09-26 14:02 ` Peter Zijlstra
2011-10-07 18:28 ` Oleg Nesterov
2011-10-09 13:31 ` Oleg Nesterov
2011-09-20 12:02 ` [PATCH v5 3.1.0-rc4-tip 13/26] x86: define a x86 specific exception notifier Srikar Dronamraju
2011-09-26 14:19 ` Peter Zijlstra
2011-09-26 15:52 ` Srikar Dronamraju
2011-09-27 11:46 ` Peter Zijlstra
2011-10-07 18:31 ` Oleg Nesterov
2011-09-20 12:02 ` [PATCH v5 3.1.0-rc4-tip 14/26] uprobe: register " Srikar Dronamraju
2011-09-20 12:03 ` [PATCH v5 3.1.0-rc4-tip 15/26] x86: Define x86_64 specific uprobe_task_arch_info structure Srikar Dronamraju
2011-09-20 12:03 ` [PATCH v5 3.1.0-rc4-tip 16/26] uprobes: Introduce " Srikar Dronamraju
2011-09-20 12:03 ` Srikar Dronamraju [this message]
2011-09-26 14:23 ` [PATCH v5 3.1.0-rc4-tip 17/26] x86: arch specific hooks for pre/post singlestep handling Peter Zijlstra
2011-09-26 16:34 ` Srikar Dronamraju
2011-09-27 11:44 ` Peter Zijlstra
2011-09-20 12:03 ` [PATCH v5 3.1.0-rc4-tip 18/26] uprobes: slot allocation Srikar Dronamraju
2011-09-27 11:49 ` Peter Zijlstra
2011-09-27 12:32 ` Srikar Dronamraju
2011-09-27 12:59 ` Peter Zijlstra
2011-09-27 12:18 ` Peter Zijlstra
2011-09-27 12:45 ` Srikar Dronamraju
2011-09-27 12:36 ` Peter Zijlstra
2011-09-27 12:37 ` Peter Zijlstra
2011-09-27 12:50 ` Srikar Dronamraju
2011-09-27 12:50 ` Peter Zijlstra
2011-09-27 12:55 ` Peter Zijlstra
2011-10-07 18:37 ` Oleg Nesterov
2011-10-09 11:47 ` Srikar Dronamraju
2011-09-20 12:03 ` [PATCH v5 3.1.0-rc4-tip 19/26] tracing: Extract out common code for kprobes/uprobes traceevents Srikar Dronamraju
2011-09-28 5:04 ` Masami Hiramatsu
2011-09-20 12:04 ` [PATCH v5 3.1.0-rc4-tip 20/26] tracing: uprobes trace_event interface Srikar Dronamraju
2011-09-20 12:04 ` [PATCH v5 3.1.0-rc4-tip 21/26] tracing: uprobes Documentation Srikar Dronamraju
2011-09-20 12:04 ` [PATCH v5 3.1.0-rc4-tip 22/26] perf: rename target_module to target Srikar Dronamraju
2011-09-20 12:04 ` [PATCH v5 3.1.0-rc4-tip 23/26] perf: perf interface for uprobes Srikar Dronamraju
2011-09-20 12:04 ` [PATCH v5 3.1.0-rc4-tip 24/26] perf: show possible probes in a given executable file or library Srikar Dronamraju
2011-09-20 12:05 ` [PATCH v5 3.1.0-rc4-tip 25/26] perf: Documentation for perf uprobes Srikar Dronamraju
2011-09-28 9:20 ` Masami Hiramatsu
2011-09-20 12:05 ` [PATCH v5 3.1.0-rc4-tip 26/26] uprobes: queue signals while thread is singlestepping Srikar Dronamraju
2011-09-27 13:03 ` Peter Zijlstra
2011-09-27 13:12 ` Srikar Dronamraju
2011-10-05 18:01 ` Oleg Nesterov
2011-10-06 5:47 ` Srikar Dronamraju
2011-10-07 16:58 ` Oleg Nesterov
2011-10-10 12:25 ` Srikar Dronamraju
2011-10-10 18:25 ` Oleg Nesterov
2011-10-11 17:24 ` Oleg Nesterov
2011-10-11 17:38 ` Srikar Dronamraju
2011-10-11 17:26 ` Srikar Dronamraju
2011-10-11 18:56 ` Oleg Nesterov
2011-10-12 12:01 ` Srikar Dronamraju
2011-10-12 19:34 ` Oleg Nesterov
2011-10-12 19:59 ` Oleg Nesterov
2011-09-20 13:34 ` [PATCH v5 3.1.0-rc4-tip 0/26] Uprobes patchset with perf probe support Christoph Hellwig
2011-09-20 14:12 ` Srikar Dronamraju
2011-09-20 14:28 ` Christoph Hellwig
2011-09-20 15:19 ` Srikar Dronamraju
2011-10-15 19:00 ` [PATCH 0/X] (Was: Uprobes patchset with perf probe support) Oleg Nesterov
2011-10-15 19:00 ` [PATCH 1/X] uprobes: write_opcode: the new page needs PG_uptodate Oleg Nesterov
2011-10-17 10:59 ` Srikar Dronamraju
2011-10-15 19:00 ` [PATCH 2/X] uprobes: write_opcode() needs put_page(new_page) unconditionally Oleg Nesterov
2011-10-18 16:47 ` Srikar Dronamraju
2011-10-15 19:01 ` [PATCH 3/X] uprobes: xol_add_vma: fix ->uprobes_xol_area initialization Oleg Nesterov
2011-10-15 19:01 ` [PATCH 4/X] uprobes: xol_add_vma: misc cleanups Oleg Nesterov
2011-10-15 19:01 ` [PATCH 5/X] uprobes: xol_alloc_area() needs memory barriers Oleg Nesterov
2011-10-16 16:13 ` [PATCH 6/X] uprobes: reimplement xol_add_vma() via install_special_mapping() Oleg Nesterov
2011-10-17 10:50 ` Srikar Dronamraju
2011-10-17 13:34 ` Stephen Smalley
2011-10-17 18:55 ` Oleg Nesterov
2011-10-16 16:14 ` [PATCH 7/X] uprobes: xol_add_vma: simply use TASK_SIZE as a hint Oleg Nesterov
2011-10-19 21:51 ` [PATCH 8-14/X] (Was: Uprobes patchset with perf probe support) Oleg Nesterov
2011-10-19 21:52 ` [PATCH 8/X] uprobes: kill sstep_complete() Oleg Nesterov
2011-10-19 21:52 ` [PATCH 9/X] uprobes: introduce UTASK_SSTEP_ACK state Oleg Nesterov
2011-10-19 21:52 ` [PATCH 10/X] uprobes: introduce uprobe_deny_signal() Oleg Nesterov
2011-10-19 21:53 ` [PATCH 11/X] uprobes: x86: introduce xol_was_trapped() Oleg Nesterov
2011-10-24 14:55 ` Srikar Dronamraju
2011-10-24 16:07 ` Oleg Nesterov
2011-10-19 21:53 ` [PATCH 12/X] uprobes: x86: introduce abort_xol() Oleg Nesterov
2011-10-21 14:42 ` Srikar Dronamraju
2011-10-21 16:22 ` Oleg Nesterov
2011-10-21 16:26 ` Ananth N Mavinakayanahalli
2011-10-21 16:42 ` Oleg Nesterov
2011-10-21 17:59 ` test-case (Was: [PATCH 12/X] uprobes: x86: introduce abort_xol()) Oleg Nesterov
2011-10-25 14:06 ` Srikar Dronamraju
2011-10-25 15:49 ` Oleg Nesterov
2011-10-22 7:09 ` [PATCH 12/X] uprobes: x86: introduce abort_xol() Ananth N Mavinakayanahalli
2011-10-19 21:53 ` [PATCH 13/X] uprobes: introduce UTASK_SSTEP_TRAPPED logic Oleg Nesterov
2011-10-22 7:20 ` Ananth N Mavinakayanahalli
2011-10-24 14:41 ` Oleg Nesterov
2011-10-24 15:16 ` Ananth N Mavinakayanahalli
2011-10-24 16:13 ` Oleg Nesterov
2011-10-25 6:01 ` Ananth N Mavinakayanahalli
2011-10-25 14:30 ` Oleg Nesterov
2011-10-19 21:54 ` [PATCH 14/X] uprobes: uprobe_deny_signal: check __fatal_signal_pending() 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=20110920120325.25326.11641.sendpatchset@srdronam.in.ibm.com \
--to=srikar@linux.vnet.ibm.com \
--cc=acme@infradead.org \
--cc=akpm@linux-foundation.org \
--cc=ananth@in.ibm.com \
--cc=andi@firstfloor.org \
--cc=corbet@lwn.net \
--cc=hch@infradead.org \
--cc=hughd@google.com \
--cc=jkenisto@linux.vnet.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=masami.hiramatsu.pt@hitachi.com \
--cc=mingo@elte.hu \
--cc=oleg@redhat.com \
--cc=peterz@infradead.org \
--cc=roland@hack.frob.com \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
/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).