Re: [PATCH 2/2] [POWERPC] uprobes: powerpc port

On Tue, Jun 12, 2012 at 02:01:46PM +1000, Michael Ellerman wrote:

On Fri, 2012-06-08 at 14:51 +0530, Ananth N Mavinakayanahalli wrote:

quoted

On Fri, Jun 08, 2012 at 04:38:17PM +1000, Michael Ellerman wrote:

quoted

On Fri, 2012-06-08 at 11:49 +0530, Ananth N Mavinakayanahalli wrote:

quoted

On Fri, Jun 08, 2012 at 04:17:44PM +1000, Michael Ellerman wrote:

quoted

On Fri, 2012-06-08 at 11:31 +0530, Ananth N Mavinakayanahalli wrote:

quoted

On Fri, Jun 08, 2012 at 03:51:54PM +1000, Michael Ellerman wrote:

quoted

On Fri, 2012-06-08 at 10:06 +0530, Ananth N Mavinakayanahalli wrote:

But MSR_PR=1 and hence emulate_step() will return -1 and hence we will
end up single-stepping using user_enable_single_step(). Same with rfid.

Right. But that was exactly Jim's point, you may be asked to emulate
those instructions even though you wouldn't expect to see them in
userspace code, so you need to handle it.

Luckily it looks like emulate_step() will do the right thing for you.
It'd be good to test it to make 100% sure.

Sure. Will add that check and send v2.

Sorry I didn't mean add a test in the code, I meant construct a test
case to confirm that it works as expected.

Michael,

I just hand-coded the instr to emulate_step() and here are the results:

MSR_PR is set
insn = 7c600124, ret = 0 /* mtmsr */
insn = 7c600164, ret = 0 /* mtmsrd */
insn = 4c000024, ret = -1 /* rfid */
insn = 4c000064, ret = 0 /* rfi */

Also verified that standalone programs with those instructions in inline
asm will die with a SIGILL.

So, for mtmsr, mtmsrd and rfi, we have to single-step them which will
result in a SIGILL in turn.

What happens in the rfid case? You don't handle -1 from emulate_step()
any differently AFAICS, so don't we try to single step that too?

-1 is just emulate_step() flagging cases where instructions must not be
single-stepped (rfi[d], mtmsr that clears MSR_RI). But as with the other
OEA instructions in user space, we fail with a SIGILL.

As the application is hozed in any case if we encounter an OEA
instruction, I'd think there is no point in handling a -1 from
emulate_step() any differently.

Ananth