From: Andy Lutomirski <luto@kernel.org>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>,
LKML <linux-kernel@vger.kernel.org>, X86 ML <x86@kernel.org>,
Andy Lutomirski <luto@kernel.org>,
Catalin Marinas <catalin.marinas@arm.com>,
Will Deacon <will@kernel.org>,
Mark Rutland <mark.rutland@arm.com>,
Marc Zyngier <maz@kernel.org>,
Paolo Bonzini <pbonzini@redhat.com>,
kvm list <kvm@vger.kernel.org>,
linux-arch <linux-arch@vger.kernel.org>,
Juergen Gross <jgross@suse.com>
Subject: Re: [RFC patch 10/15] x86/entry: Move irq tracing to C code
Date: Mon, 23 Sep 2019 10:24:01 -0700 [thread overview]
Message-ID: <CALCETrW7T5KPCFO6vWUkExCYKdXVkv6mkfA53A542uxukP+eew@mail.gmail.com> (raw)
In-Reply-To: <20190923121001.GG2332@hirez.programming.kicks-ass.net>
On Mon, Sep 23, 2019 at 5:10 AM Peter Zijlstra <peterz@infradead.org> wrote:
>
> On Mon, Sep 23, 2019 at 01:55:51PM +0200, Peter Zijlstra wrote:
> > On Mon, Sep 23, 2019 at 01:49:20PM +0200, Peter Zijlstra wrote:
> > > While walking the kids to school I wondered WTH we need to call
> > > TRACE_IRQS_OFF in the first place. If this is the return from exception
> > > path, interrupts had better be disabled already (in exception enter).
> > >
> > > For entry_64.S we have:
> > >
> > > - idtentry_part; which does TRACE_IRQS_OFF at the start and error_exit
> > > at the end.
> > >
> > > - xen_do_hypervisor_callback, xen_failsafe_callback -- which are
> > > confusing.
> > >
> > > So in the normal case, it appears we can simply do:
> > >
> > > diff --git a/arch/x86/entry/entry_64.S b/arch/x86/entry/entry_64.S
> > > index b7c3ea4cb19d..e9cf59ac554e 100644
> > > --- a/arch/x86/entry/entry_64.S
> > > +++ b/arch/x86/entry/entry_64.S
> > > @@ -1368,8 +1368,6 @@ END(error_entry)
> > >
> > > ENTRY(error_exit)
> > > UNWIND_HINT_REGS
> > > - DISABLE_INTERRUPTS(CLBR_ANY)
> > > - TRACE_IRQS_OFF
> > > testb $3, CS(%rsp)
> > > jz retint_kernel
> > > jmp retint_user
> > >
> > > and all should be well. This leaves Xen...
> > >
> > > For entry_32.S it looks like nothing uses 'resume_userspace' so that
> > > ENTRY can go away. Which then leaves:
> > >
> > > * ret_from_intr:
> > > - common_spurious: TRACE_IRQS_OFF
> > > - common_interrupt: TRACE_IRQS_OFF
> > > - BUILD_INTERRUPT3: TRACE_IRQS_OFF
> > > - xen_do_upcall: ...
> > >
> > > * ret_from_exception:
> > > - xen_failsafe_callback: ...
> > > - common_exception_read_cr2: TRACE_IRQS_OFF
> > > - common_exception: TRACE_IRQS_OFF
> > > - int3: TRACE_IRQS_OFF
> > >
> > > Which again suggests:
> > >
> > > diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S
> > > index f83ca5aa8b77..7a19e7413a8e 100644
> > > --- a/arch/x86/entry/entry_32.S
> > > +++ b/arch/x86/entry/entry_32.S
> > > @@ -825,9 +825,6 @@ END(ret_from_fork)
> > > cmpl $USER_RPL, %eax
> > > jb restore_all_kernel # not returning to v8086 or userspace
> > >
> > > -ENTRY(resume_userspace)
> > > - DISABLE_INTERRUPTS(CLBR_ANY)
> > > - TRACE_IRQS_OFF
> > > movl %esp, %eax
> > > call prepare_exit_to_usermode
> > > jmp restore_all
> > >
> > > with the notable exception (oh teh pun!) being Xen... _again_.
> > >
> > > With these patchlets on, we'd want prepare_exit_to_usermode() to
> > > validate the IRQ state, but AFAICT it _should_ all just 'work' (famous
> > > last words).
> > >
> > > Cc Juergen because Xen
> >
> > Arrgh.. faults!! they do local_irq_enable() but never disable them
> > again. Arguably those functions should be symmetric and restore IF when
> > they muck with it instead of relying on the exit path fixing things up.
>
> ---
> diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S
> index f83ca5aa8b77..7a19e7413a8e 100644
> --- a/arch/x86/entry/entry_32.S
> +++ b/arch/x86/entry/entry_32.S
> @@ -825,9 +825,6 @@ END(ret_from_fork)
> cmpl $USER_RPL, %eax
> jb restore_all_kernel # not returning to v8086 or userspace
...
Yes, please, but can you add an assertion under CONFIG_DEBUG_ENTRY
that IRQs are actually off?
next prev parent reply other threads:[~2019-09-23 17:24 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-09-19 15:03 [RFC patch 00/15] entry: Provide generic implementation for host and guest entry/exit work Thomas Gleixner
2019-09-19 15:03 ` [RFC patch 01/15] entry: Provide generic syscall entry functionality Thomas Gleixner
2019-09-20 23:38 ` Andy Lutomirski
2019-10-20 11:49 ` Thomas Gleixner
2019-09-23 9:05 ` Mike Rapoport
2019-09-19 15:03 ` [RFC patch 02/15] x86/entry: Remove _TIF_NOHZ from _TIF_WORK_SYSCALL_ENTRY Thomas Gleixner
2019-09-20 23:39 ` Andy Lutomirski
2019-09-23 20:43 ` Thomas Gleixner
2019-09-19 15:03 ` [RFC patch 03/15] x86/entry: Use generic syscall entry function Thomas Gleixner
2019-09-20 23:41 ` Andy Lutomirski
2019-09-23 8:31 ` Peter Zijlstra
2019-09-23 8:40 ` Thomas Gleixner
2019-09-19 15:03 ` [RFC patch 04/15] arm64/entry: " Thomas Gleixner
2019-09-20 12:21 ` Catalin Marinas
2019-09-19 15:03 ` [RFC patch 05/15] entry: Provide generic syscall exit function Thomas Gleixner
2019-09-19 15:03 ` [RFC patch 06/15] x86/entry: Use generic syscall exit functionality Thomas Gleixner
2019-09-19 15:03 ` [RFC patch 07/15] arm64/syscall: Remove obscure flag check Thomas Gleixner
2019-09-20 14:29 ` Catalin Marinas
2019-09-19 15:03 ` [RFC patch 08/15] arm64/syscall: Use generic syscall exit functionality Thomas Gleixner
2019-09-19 15:03 ` [RFC patch 09/15] entry: Provide generic exit to usermode functionality Thomas Gleixner
2019-09-23 8:30 ` Peter Zijlstra
2019-09-19 15:03 ` [RFC patch 10/15] x86/entry: Move irq tracing to C code Thomas Gleixner
2019-09-23 8:47 ` Peter Zijlstra
2019-09-23 10:27 ` Thomas Gleixner
2019-09-23 11:49 ` Peter Zijlstra
2019-09-23 11:55 ` Peter Zijlstra
2019-09-23 12:10 ` Peter Zijlstra
2019-09-23 17:24 ` Andy Lutomirski [this message]
2019-09-26 2:59 ` Josh Poimboeuf
2019-09-19 15:03 ` [RFC patch 11/15] x86/entry: Use generic exit to usermode Thomas Gleixner
2019-09-19 15:03 ` [RFC patch 12/15] arm64/entry: " Thomas Gleixner
2019-09-19 15:03 ` [RFC patch 13/15] arm64/entry: Move FPU restore out of exit_to_usermode() loop Thomas Gleixner
2019-09-19 15:03 ` [RFC patch 14/15] workpending: Provide infrastructure for work before entering a guest Thomas Gleixner
2019-09-19 15:40 ` Paolo Bonzini
2019-09-20 11:48 ` Thomas Gleixner
2019-09-23 18:17 ` Andy Lutomirski
2019-09-26 11:35 ` Miroslav Benes
2019-09-19 15:03 ` [RFC patch 15/15] x86/kvm: Use GENERIC_EXIT_WORKPENDING Thomas Gleixner
2019-09-19 15:40 ` Paolo Bonzini
2019-09-20 15:12 ` [RFC patch 00/15] entry: Provide generic implementation for host and guest entry/exit work Mark Rutland
2019-09-23 20:50 ` Thomas Gleixner
2019-09-23 18:18 ` Andy Lutomirski
2019-09-24 6:50 ` Christian Borntraeger
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=CALCETrW7T5KPCFO6vWUkExCYKdXVkv6mkfA53A542uxukP+eew@mail.gmail.com \
--to=luto@kernel.org \
--cc=catalin.marinas@arm.com \
--cc=jgross@suse.com \
--cc=kvm@vger.kernel.org \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mark.rutland@arm.com \
--cc=maz@kernel.org \
--cc=pbonzini@redhat.com \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--cc=will@kernel.org \
--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).