From: Josh Poimboeuf <jpoimboe@redhat.com>
To: Alexei Starovoitov <alexei.starovoitov@gmail.com>
Cc: x86@kernel.org, linux-kernel@vger.kernel.org,
Alexei Starovoitov <ast@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
netdev@vger.kernel.org, bpf@vger.kernel.org,
Peter Zijlstra <peterz@infradead.org>,
Song Liu <songliubraving@fb.com>, Kairui Song <kasong@redhat.com>
Subject: Re: [PATCH 7/9] x86/unwind/orc: Fall back to using frame pointers for generated code
Date: Thu, 13 Jun 2019 21:28:48 -0500 [thread overview]
Message-ID: <20190614022848.ly4vlgsz6fa4bcbl@treble> (raw)
In-Reply-To: <20190614015848.todgfogryjn573nd@treble>
On Thu, Jun 13, 2019 at 08:58:48PM -0500, Josh Poimboeuf wrote:
> On Thu, Jun 13, 2019 at 06:42:45PM -0700, Alexei Starovoitov wrote:
> > On Thu, Jun 13, 2019 at 08:30:51PM -0500, Josh Poimboeuf wrote:
> > > On Thu, Jun 13, 2019 at 03:00:55PM -0700, Alexei Starovoitov wrote:
> > > > > @@ -392,8 +402,16 @@ bool unwind_next_frame(struct unwind_state *state)
> > > > > * calls and calls to noreturn functions.
> > > > > */
> > > > > orc = orc_find(state->signal ? state->ip : state->ip - 1);
> > > > > - if (!orc)
> > > > > - goto err;
> > > > > + if (!orc) {
> > > > > + /*
> > > > > + * As a fallback, try to assume this code uses a frame pointer.
> > > > > + * This is useful for generated code, like BPF, which ORC
> > > > > + * doesn't know about. This is just a guess, so the rest of
> > > > > + * the unwind is no longer considered reliable.
> > > > > + */
> > > > > + orc = &orc_fp_entry;
> > > > > + state->error = true;
> > > >
> > > > That seems fragile.
> > >
> > > I don't think so. The unwinder has sanity checks to make sure it
> > > doesn't go off the rails. And it works just fine. The beauty is that
> > > it should work for all generated code (not just BPF).
> > >
> > > > Can't we populate orc_unwind tables after JIT ?
> > >
> > > As I mentioned it would introduce a lot more complexity. For each JIT
> > > function, BPF would have to tell ORC the following:
> > >
> > > - where the BPF function lives
> > > - how big the stack frame is
> > > - where RBP and other callee-saved regs are on the stack
> >
> > that sounds like straightforward addition that ORC should have anyway.
> > right now we're not using rbp in the jit-ed code,
> > but one day we definitely will.
> > Same goes for r12. It's reserved right now for 'strategic use'.
> > We've been thinking to add another register to bpf isa.
> > It will map to r12 on x86. arm64 and others have plenty of regs to use.
> > The programs are getting bigger and register spill/fill starting to
> > become a performance concern. Extra register will give us more room.
>
> With CONFIG_FRAME_POINTER, RBP isn't available. If you look at all the
> code in the entire kernel you'll notice that BPF JIT is pretty much the
> only one still clobbering it.
Hm. If you wanted to eventually use R12 for other purposes, there might
be a way to abstract BPF_REG_FP such that it doesn't actually need a
dedicated register. The BPF program's frame pointer will always be a
certain constant offset away from RBP (real frame pointer), so accesses
to BPF_REG_FP could still be based on RBP, but with an offset added to
it.
--
Josh
next prev parent reply other threads:[~2019-06-14 2:28 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-06-13 13:20 [PATCH 0/9] x86/bpf: unwinder fixes Josh Poimboeuf
2019-06-13 13:20 ` [PATCH 1/9] perf/x86: Always store regs->ip in perf_callchain_kernel() Josh Poimboeuf
2019-06-13 13:20 ` [PATCH 2/9] objtool: Fix ORC unwinding in non-JIT BPF generated code Josh Poimboeuf
2019-06-13 20:57 ` Alexei Starovoitov
2019-06-14 1:20 ` Josh Poimboeuf
2019-06-14 1:37 ` Alexei Starovoitov
2019-06-14 1:51 ` Josh Poimboeuf
2019-06-14 7:08 ` Peter Zijlstra
2019-06-14 7:35 ` Alexei Starovoitov
2019-06-14 8:11 ` Peter Zijlstra
2019-06-14 15:13 ` Alexei Starovoitov
2019-06-14 16:11 ` Josh Poimboeuf
2019-06-13 13:21 ` [PATCH 3/9] x86/bpf: Move epilogue generation to a dedicated function Josh Poimboeuf
2019-06-13 18:57 ` Song Liu
2019-06-13 19:12 ` Josh Poimboeuf
2019-06-13 13:21 ` [PATCH 4/9] x86/bpf: Simplify prologue generation Josh Poimboeuf
2019-06-13 13:21 ` [PATCH 5/9] x86/bpf: Support SIB byte generation Josh Poimboeuf
2019-06-13 13:21 ` [PATCH 6/9] x86/bpf: Fix JIT frame pointer usage Josh Poimboeuf
2019-06-13 21:58 ` Alexei Starovoitov
2019-06-14 1:22 ` Josh Poimboeuf
2019-06-14 1:39 ` Alexei Starovoitov
2019-06-14 1:52 ` Josh Poimboeuf
2019-06-14 10:50 ` David Laight
2019-06-14 13:44 ` Josh Poimboeuf
2019-06-14 13:58 ` David Laight
2019-06-14 17:07 ` Josh Poimboeuf
2019-06-17 15:07 ` David Laight
2019-06-13 13:21 ` [PATCH 7/9] x86/unwind/orc: Fall back to using frame pointers for generated code Josh Poimboeuf
2019-06-13 22:00 ` Alexei Starovoitov
2019-06-14 1:30 ` Josh Poimboeuf
2019-06-14 1:42 ` Alexei Starovoitov
2019-06-14 1:58 ` Josh Poimboeuf
2019-06-14 2:28 ` Josh Poimboeuf [this message]
2019-06-14 4:50 ` Josh Poimboeuf
2019-06-14 6:00 ` Alexei Starovoitov
2019-06-14 7:41 ` Peter Zijlstra
2019-06-14 13:31 ` Josh Poimboeuf
2019-06-14 15:29 ` Alexei Starovoitov
2019-06-14 13:34 ` Josh Poimboeuf
2019-06-14 15:31 ` Alexei Starovoitov
2019-06-14 15:56 ` Josh Poimboeuf
2019-06-13 13:21 ` [PATCH 8/9] x86/bpf: Convert asm comments to AT&T syntax Josh Poimboeuf
2019-06-13 18:52 ` Song Liu
2019-06-13 19:11 ` Josh Poimboeuf
2019-06-14 7:42 ` Peter Zijlstra
2019-06-14 15:13 ` Song Liu
2019-06-13 13:21 ` [PATCH 9/9] x86/bpf: Convert MOV function/macro argument ordering " Josh Poimboeuf
2019-06-13 19:00 ` [PATCH 0/9] x86/bpf: unwinder fixes Song Liu
2019-06-13 20:41 ` Peter Zijlstra
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=20190614022848.ly4vlgsz6fa4bcbl@treble \
--to=jpoimboe@redhat.com \
--cc=alexei.starovoitov@gmail.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=kasong@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=peterz@infradead.org \
--cc=songliubraving@fb.com \
--cc=x86@kernel.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).