From: Denys Vlasenko <vda.linux@googlemail.com>
To: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>,
Linus Torvalds <torvalds@linux-foundation.org>,
Oleg Nesterov <oleg@redhat.com>, Ingo Molnar <mingo@elte.hu>,
Andrew Morton <akpm@linux-foundation.org>,
LKML <linux-kernel@vger.kernel.org>,
Linux-mm <linux-mm@kvack.org>, Andi Kleen <andi@firstfloor.org>,
Christoph Hellwig <hch@infradead.org>,
Steven Rostedt <rostedt@goodmis.org>,
Roland McGrath <roland@hack.frob.com>,
Thomas Gleixner <tglx@linutronix.de>,
Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>,
Arnaldo Carvalho de Melo <acme@infradead.org>,
Anton Arapov <anton@redhat.com>,
Ananth N Mavinakayanahalli <ananth@in.ibm.com>,
Jim Keniston <jkenisto@linux.vnet.ibm.com>,
Stephen Rothwell <sfr@canb.auug.org.au>
Subject: Re: [PATCH v9 3.2 1/9] uprobes: Install and remove breakpoints.
Date: Wed, 25 Jan 2012 16:13:29 +0100 [thread overview]
Message-ID: <CAK1hOcO3pz+zBLQKfdty3UwQG8zxXwBWo9euFaE+zKawiqTE2g@mail.gmail.com> (raw)
In-Reply-To: <20120110114831.17610.88468.sendpatchset@srdronam.in.ibm.com>
On Tue, Jan 10, 2012 at 12:48 PM, Srikar Dronamraju
<srikar@linux.vnet.ibm.com> wrote:
> +/*
> + * If uprobe->insn doesn't use rip-relative addressing, return
> + * immediately. Otherwise, rewrite the instruction so that it accesses
> + * its memory operand indirectly through a scratch register. Set
> + * uprobe->arch_info.fixups and uprobe->arch_info.rip_rela_target_address
> + * accordingly. (The contents of the scratch register will be saved
> + * before we single-step the modified instruction, and restored
> + * afterward.)
> + *
> + * We do this because a rip-relative instruction can access only a
> + * relatively small area (+/- 2 GB from the instruction), and the XOL
> + * area typically lies beyond that area. At least for instructions
> + * that store to memory, we can't execute the original instruction
> + * and "fix things up" later, because the misdirected store could be
> + * disastrous.
> + *
> + * Some useful facts about rip-relative instructions:
> + * - There's always a modrm byte.
> + * - There's never a SIB byte.
> + * - The displacement is always 4 bytes.
> + */
> +static void handle_riprel_insn(struct mm_struct *mm, struct uprobe *uprobe,
> + struct insn *insn)
> +{
> + u8 *cursor;
> + u8 reg;
> +
> + if (mm->context.ia32_compat)
> + return;
> +
> + uprobe->arch_info.rip_rela_target_address = 0x0;
> + if (!insn_rip_relative(insn))
> + return;
> +
> + /*
> + * Point cursor at the modrm byte. The next 4 bytes are the
> + * displacement. Beyond the displacement, for some instructions,
> + * is the immediate operand.
> + */
> + cursor = uprobe->insn + insn->prefixes.nbytes
> + + insn->rex_prefix.nbytes + insn->opcode.nbytes;
> + insn_get_length(insn);
> +
> + /*
> + * Convert from rip-relative addressing to indirect addressing
> + * via a scratch register. Change the r/m field from 0x5 (%rip)
> + * to 0x0 (%rax) or 0x1 (%rcx), and squeeze out the offset field.
> + */
> + reg = MODRM_REG(insn);
> + if (reg == 0) {
> + /*
> + * The register operand (if any) is either the A register
> + * (%rax, %eax, etc.) or (if the 0x4 bit is set in the
> + * REX prefix) %r8. In any case, we know the C register
> + * is NOT the register operand, so we use %rcx (register
> + * #1) for the scratch register.
> + */
> + uprobe->arch_info.fixups = UPROBES_FIX_RIP_CX;
> + /* Change modrm from 00 000 101 to 00 000 001. */
> + *cursor = 0x1;
> + } else {
> + /* Use %rax (register #0) for the scratch register. */
> + uprobe->arch_info.fixups = UPROBES_FIX_RIP_AX;
> + /* Change modrm from 00 xxx 101 to 00 xxx 000 */
> + *cursor = (reg << 3);
> + }
> +
> + /* Target address = address of next instruction + (signed) offset */
> + uprobe->arch_info.rip_rela_target_address = (long)insn->length
> + + insn->displacement.value;
> + /* Displacement field is gone; slide immediate field (if any) over. */
> + if (insn->immediate.nbytes) {
> + cursor++;
> + memmove(cursor, cursor + insn->displacement.nbytes,
> + insn->immediate.nbytes);
> + }
> + return;
> +}
It seems to be possible to store RIP value *without displacement*
into AX/CX and convert rip-relative instruction into AX/CX *relative* one.
Example:
c7 05 78 56 34 12 2a 00 00 00 movl $0x2a,0x12345678(%rip)
converts to:
c7 81 78 56 34 12 2a 00 00 00 movl $0x2a,0x12345678(%rcx)
This way instruction size stays the same and you don't need
to memmove immediate value.
--
vda
next prev parent reply other threads:[~2012-01-25 15:13 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-01-10 11:48 [PATCH v9 3.2 0/9] Uprobes patchset with perf probe support Srikar Dronamraju
2012-01-10 11:48 ` [PATCH v9 3.2 1/9] uprobes: Install and remove breakpoints Srikar Dronamraju
2012-01-25 14:39 ` Denys Vlasenko
2012-01-25 16:31 ` Oleg Nesterov
2012-01-25 15:13 ` Denys Vlasenko [this message]
2012-01-25 15:32 ` Denys Vlasenko
2012-01-26 14:14 ` Masami Hiramatsu
2012-01-26 18:28 ` Denys Vlasenko
2012-01-30 9:51 ` Masami Hiramatsu
2012-01-26 13:38 ` Masami Hiramatsu
2012-01-26 13:45 ` Masami Hiramatsu
2012-01-10 11:48 ` [PATCH v9 3.2 2/9] uprobes: handle breakpoint and signal step exception Srikar Dronamraju
2012-01-18 8:39 ` Anton Arapov
2012-01-18 9:02 ` Srikar Dronamraju
2012-01-18 10:18 ` Mike Frysinger
2012-01-18 10:47 ` Srikar Dronamraju
2012-01-18 11:01 ` Mike Frysinger
2012-01-20 22:57 ` Mike Frysinger
2012-01-25 8:12 ` Srikar Dronamraju
2012-01-10 11:48 ` [PATCH v9 3.2 3/9] uprobes: slot allocation Srikar Dronamraju
2012-01-10 11:49 ` [PATCH v9 3.2 4/9] uprobes: counter to optimize probe hits Srikar Dronamraju
2012-01-10 11:49 ` [PATCH v9 3.2 5/9] tracing: modify is_delete, is_return from ints to bool Srikar Dronamraju
2012-01-10 11:49 ` [PATCH v9 3.2 6/9] tracing: Extract out common code for kprobes/uprobes traceevents Srikar Dronamraju
2012-01-10 11:49 ` [PATCH v9 3.2 7/9] tracing: uprobes trace_event interface Srikar Dronamraju
2012-01-16 13:11 ` Jiri Olsa
2012-01-16 14:45 ` Srikar Dronamraju
2012-01-16 15:33 ` Jiri Olsa
2012-01-16 16:41 ` Srikar Dronamraju
2012-01-17 9:28 ` Ingo Molnar
2012-01-17 10:22 ` Srikar Dronamraju
2012-01-17 10:59 ` Ingo Molnar
2012-01-17 11:03 ` Srikar Dronamraju
2012-01-17 11:22 ` Ingo Molnar
2012-01-17 11:57 ` Srikar Dronamraju
2012-01-17 12:23 ` Ingo Molnar
2012-01-17 12:27 ` Ingo Molnar
2012-01-17 12:11 ` Ingo Molnar
2012-01-17 12:21 ` Ingo Molnar
2012-01-18 10:07 ` Srikar Dronamraju
2012-01-10 11:49 ` [PATCH v9 3.2 8/9] perf: rename target_module to target Srikar Dronamraju
2012-01-16 8:34 ` [PATCH v9 3.2 0/9] Uprobes patchset with perf probe support Ingo Molnar
2012-01-16 15:17 ` Srikar Dronamraju
2012-01-17 9:39 ` Ingo Molnar
2012-01-25 14:11 ` Peter Zijlstra
2012-01-26 11:10 ` Ingo Molnar
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=CAK1hOcO3pz+zBLQKfdty3UwQG8zxXwBWo9euFaE+zKawiqTE2g@mail.gmail.com \
--to=vda.linux@googlemail.com \
--cc=acme@infradead.org \
--cc=akpm@linux-foundation.org \
--cc=ananth@in.ibm.com \
--cc=andi@firstfloor.org \
--cc=anton@redhat.com \
--cc=hch@infradead.org \
--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=sfr@canb.auug.org.au \
--cc=srikar@linux.vnet.ibm.com \
--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).