All of lore.kernel.org
 help / color / mirror / Atom feed
From: Josh Poimboeuf <jpoimboe@redhat.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: "Jürgen Groß" <jgross@suse.com>,
	xen-devel@lists.xenproject.org, x86@kernel.org,
	linux-kernel@vger.kernel.org,
	virtualization@lists.linux-foundation.org,
	linux-hyperv@vger.kernel.org, kvm@vger.kernel.org,
	luto@kernel.org, "Thomas Gleixner" <tglx@linutronix.de>,
	"Ingo Molnar" <mingo@redhat.com>,
	"Borislav Petkov" <bp@alien8.de>,
	"H. Peter Anvin" <hpa@zytor.com>,
	"Boris Ostrovsky" <boris.ostrovsky@oracle.com>,
	"Stefano Stabellini" <sstabellini@kernel.org>,
	"Deep Shah" <sdeep@vmware.com>,
	"VMware, Inc." <pv-drivers@vmware.com>,
	"K. Y. Srinivasan" <kys@microsoft.com>,
	"Haiyang Zhang" <haiyangz@microsoft.com>,
	"Stephen Hemminger" <sthemmin@microsoft.com>,
	"Wei Liu" <wei.liu@kernel.org>,
	"Paolo Bonzini" <pbonzini@redhat.com>,
	"Sean Christopherson" <sean.j.christopherson@intel.com>,
	"Vitaly Kuznetsov" <vkuznets@redhat.com>,
	"Wanpeng Li" <wanpengli@tencent.com>,
	"Jim Mattson" <jmattson@google.com>,
	"Joerg Roedel" <joro@8bytes.org>,
	"Daniel Lezcano" <daniel.lezcano@linaro.org>,
	"Juri Lelli" <juri.lelli@redhat.com>,
	"Vincent Guittot" <vincent.guittot@linaro.org>,
	"Dietmar Eggemann" <dietmar.eggemann@arm.com>,
	"Steven Rostedt" <rostedt@goodmis.org>,
	"Ben Segall" <bsegall@google.com>, "Mel Gorman" <mgorman@suse.de>,
	"Daniel Bristot de Oliveira" <bristot@redhat.com>
Subject: Re: [PATCH v2 00/12] x86: major paravirt cleanup
Date: Tue, 15 Dec 2020 18:38:02 -0600	[thread overview]
Message-ID: <20201216003802.5fpklvx37yuiufrt@treble> (raw)
In-Reply-To: <20201215145408.GR3092@hirez.programming.kicks-ass.net>

On Tue, Dec 15, 2020 at 03:54:08PM +0100, Peter Zijlstra wrote:
> The problem is that a single instance of unwind information (ORC) must
> capture and correctly unwind all alternatives. Since the trivially
> correct mandate is out, implement the straight forward brute-force
> approach:
> 
>  1) generate CFI information for each alternative
> 
>  2) unwind every alternative with the merge-sort of the previously
>     generated CFI information -- O(n^2)
> 
>  3) for any possible conflict: yell.
> 
>  4) Generate ORC with merge-sort
> 
> Specifically for 3 there are two possible classes of conflicts:
> 
>  - the merge-sort itself could find conflicting CFI for the same
>    offset.
> 
>  - the unwind can fail with the merged CFI.

So much algorithm.  Could we make it easier by caching the shared
per-alt-group CFI state somewhere along the way?

For example:

struct alt_group_info {

	/* first original insn in the group */
	struct instruction *orig_insn;

	/* max # of bytes in the group (cfi array size) */
	unsigned long nbytes;

	/* byte-offset-addressed array of CFI pointers */
	struct cfi_state **cfi;
};

We could change 'insn->alt_group' to be a pointer to a shared instance
of the above struct, so that all original and replacement instructions
in a group have a pointer to it.

Starting out, 'cfi' array is all NULLs.  Then when updating CFI, check
'insn->alt_group.cfi[offset]'.

[ 'offset' is a byte offset from the beginning of the group.  It could
  be calculated based on 'orig_insn' or 'orig_insn->alts', depending on
  whether 'insn' is an original or a replacement. ]

If the array entry is NULL, just update it with a pointer to the CFI.
If it's not NULL, make sure it matches the existing CFI, and WARN if it
doesn't.

Also, with this data structure, the ORC generation should also be a lot
more straightforward, just ignore the NULL entries.

Thoughts?  This is all theoretical of course, I could try to do a patch
tomorrow.

-- 
Josh


WARNING: multiple messages have this Message-ID (diff)
From: Josh Poimboeuf <jpoimboe@redhat.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: "Juri Lelli" <juri.lelli@redhat.com>,
	linux-hyperv@vger.kernel.org,
	"Daniel Lezcano" <daniel.lezcano@linaro.org>,
	"Wanpeng Li" <wanpengli@tencent.com>,
	kvm@vger.kernel.org, "VMware, Inc." <pv-drivers@vmware.com>,
	virtualization@lists.linux-foundation.org,
	"Ben Segall" <bsegall@google.com>,
	"H. Peter Anvin" <hpa@zytor.com>,
	"Boris Ostrovsky" <boris.ostrovsky@oracle.com>,
	"Wei Liu" <wei.liu@kernel.org>,
	"Stefano Stabellini" <sstabellini@kernel.org>,
	"Stephen Hemminger" <sthemmin@microsoft.com>,
	"Joerg Roedel" <joro@8bytes.org>,
	x86@kernel.org, "Ingo Molnar" <mingo@redhat.com>,
	"Mel Gorman" <mgorman@suse.de>,
	xen-devel@lists.xenproject.org,
	"Haiyang Zhang" <haiyangz@microsoft.com>,
	"Steven Rostedt" <rostedt@goodmis.org>,
	"Borislav Petkov" <bp@alien8.de>,
	luto@kernel.org, "Vincent Guittot" <vincent.guittot@linaro.org>,
	"Thomas Gleixner" <tglx@linutronix.de>,
	"Dietmar Eggemann" <dietmar.eggemann@arm.com>,
	"Jim Mattson" <jmattson@google.com>,
	"Jürgen Groß" <jgross@suse.com>,
	linux-kernel@vger.kernel.org,
	"Sean Christopherson" <sean.j.christopherson@intel.com>,
	"Paolo Bonzini" <pbonzini@redhat.com>,
	"Daniel Bristot de Oliveira" <bristot@redhat.com>
Subject: Re: [PATCH v2 00/12] x86: major paravirt cleanup
Date: Tue, 15 Dec 2020 18:38:02 -0600	[thread overview]
Message-ID: <20201216003802.5fpklvx37yuiufrt@treble> (raw)
In-Reply-To: <20201215145408.GR3092@hirez.programming.kicks-ass.net>

On Tue, Dec 15, 2020 at 03:54:08PM +0100, Peter Zijlstra wrote:
> The problem is that a single instance of unwind information (ORC) must
> capture and correctly unwind all alternatives. Since the trivially
> correct mandate is out, implement the straight forward brute-force
> approach:
> 
>  1) generate CFI information for each alternative
> 
>  2) unwind every alternative with the merge-sort of the previously
>     generated CFI information -- O(n^2)
> 
>  3) for any possible conflict: yell.
> 
>  4) Generate ORC with merge-sort
> 
> Specifically for 3 there are two possible classes of conflicts:
> 
>  - the merge-sort itself could find conflicting CFI for the same
>    offset.
> 
>  - the unwind can fail with the merged CFI.

So much algorithm.  Could we make it easier by caching the shared
per-alt-group CFI state somewhere along the way?

For example:

struct alt_group_info {

	/* first original insn in the group */
	struct instruction *orig_insn;

	/* max # of bytes in the group (cfi array size) */
	unsigned long nbytes;

	/* byte-offset-addressed array of CFI pointers */
	struct cfi_state **cfi;
};

We could change 'insn->alt_group' to be a pointer to a shared instance
of the above struct, so that all original and replacement instructions
in a group have a pointer to it.

Starting out, 'cfi' array is all NULLs.  Then when updating CFI, check
'insn->alt_group.cfi[offset]'.

[ 'offset' is a byte offset from the beginning of the group.  It could
  be calculated based on 'orig_insn' or 'orig_insn->alts', depending on
  whether 'insn' is an original or a replacement. ]

If the array entry is NULL, just update it with a pointer to the CFI.
If it's not NULL, make sure it matches the existing CFI, and WARN if it
doesn't.

Also, with this data structure, the ORC generation should also be a lot
more straightforward, just ignore the NULL entries.

Thoughts?  This is all theoretical of course, I could try to do a patch
tomorrow.

-- 
Josh

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

  parent reply	other threads:[~2020-12-16  0:40 UTC|newest]

Thread overview: 107+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-11-20 11:46 [PATCH v2 00/12] x86: major paravirt cleanup Juergen Gross
2020-11-20 11:46 ` Juergen Gross via Virtualization
2020-11-20 11:46 ` [PATCH v2 01/12] x86/xen: use specific Xen pv interrupt entry for MCE Juergen Gross
2020-12-09 21:03   ` Thomas Gleixner
2020-11-20 11:46 ` [PATCH v2 02/12] x86/xen: use specific Xen pv interrupt entry for DF Juergen Gross
2020-12-09 21:03   ` Thomas Gleixner
2020-11-20 11:46 ` [PATCH v2 03/12] x86/pv: switch SWAPGS to ALTERNATIVE Juergen Gross
2020-11-20 11:46   ` Juergen Gross via Virtualization
2020-11-27 11:31   ` Borislav Petkov
2020-11-27 11:31     ` Borislav Petkov
2020-12-09 21:07   ` Thomas Gleixner
2020-12-09 21:07     ` Thomas Gleixner
2020-11-20 11:46 ` [PATCH v2 04/12] x86/xen: drop USERGS_SYSRET64 paravirt call Juergen Gross
2020-11-20 11:46   ` Juergen Gross via Virtualization
2020-12-02 12:32   ` Borislav Petkov
2020-12-02 12:32     ` Borislav Petkov
2020-12-02 14:48     ` Jürgen Groß
2020-12-02 14:48       ` Jürgen Groß via Virtualization
2020-12-02 17:08       ` Borislav Petkov
2020-12-02 17:08         ` Borislav Petkov
2020-11-20 11:46 ` [PATCH v2 05/12] x86: rework arch_local_irq_restore() to not use popf Juergen Gross
2020-11-20 11:46   ` Juergen Gross via Virtualization
2020-11-20 11:59   ` Peter Zijlstra
2020-11-20 11:59     ` Peter Zijlstra
2020-11-20 12:05     ` Jürgen Groß
2020-11-20 12:05       ` Jürgen Groß via Virtualization
2020-11-22  6:55     ` Jürgen Groß
2020-11-22  6:55       ` Jürgen Groß via Virtualization
2020-11-22 21:44       ` Andy Lutomirski
2020-11-22 21:44         ` Andy Lutomirski
2020-11-22 21:44         ` Andy Lutomirski
2020-11-23  5:21         ` Jürgen Groß
2020-11-23  5:21           ` Jürgen Groß via Virtualization
2020-11-23 15:15           ` Andy Lutomirski
2020-11-23 15:15             ` Andy Lutomirski
2020-12-09 13:27         ` Mark Rutland
2020-12-09 13:27           ` Mark Rutland
2020-12-09 14:02           ` Mark Rutland
2020-12-09 14:02             ` Mark Rutland
2020-12-09 14:05             ` Jürgen Groß
2020-12-09 14:05               ` Jürgen Groß via Virtualization
2020-12-09 18:15     ` Mark Rutland
2020-12-09 18:15       ` Mark Rutland
2020-12-09 18:54       ` Thomas Gleixner
2020-12-09 18:54         ` Thomas Gleixner
2020-12-10 11:10         ` Mark Rutland
2020-12-10 11:10           ` Mark Rutland
2020-12-10 20:15           ` x86/ioapic: Cleanup the timer_works() irqflags mess Thomas Gleixner
2020-12-10 20:15             ` Thomas Gleixner
2020-12-10 22:04             ` [tip: x86/apic] " tip-bot2 for Thomas Gleixner
2020-12-11  5:10             ` Jürgen Groß
2020-12-11  5:10               ` Jürgen Groß via Virtualization
2020-11-27  2:20   ` [x86] 97e8f0134a: fio.write_iops 8.6% improvement kernel test robot
2020-11-20 11:46 ` [PATCH v2 06/12] x86/paravirt: switch time pvops functions to use static_call() Juergen Gross
2020-11-20 11:46   ` Juergen Gross via Virtualization
2020-11-20 12:01   ` Peter Zijlstra
2020-11-20 12:01     ` Peter Zijlstra
2020-11-20 12:07     ` Jürgen Groß
2020-11-20 12:07       ` Jürgen Groß via Virtualization
2020-11-20 11:46 ` [PATCH v2 07/12] x86: add new features for paravirt patching Juergen Gross
2020-12-08 18:43   ` Borislav Petkov
2020-12-09  7:30     ` Jürgen Groß
2020-12-09 12:03       ` Borislav Petkov
2020-12-09 12:22         ` Jürgen Groß
2020-12-10 17:58           ` Borislav Petkov
2020-11-20 11:46 ` [PATCH v2 08/12] x86/paravirt: remove no longer needed 32-bit pvops cruft Juergen Gross
2020-11-20 11:46   ` Juergen Gross via Virtualization
2020-11-20 12:08   ` Peter Zijlstra
2020-11-20 12:08     ` Peter Zijlstra
2020-11-20 12:16     ` Jürgen Groß
2020-11-20 12:16       ` Jürgen Groß via Virtualization
2020-11-20 16:52   ` Arvind Sankar
2020-11-20 11:46 ` [PATCH v2 09/12] x86/paravirt: switch iret pvops to ALTERNATIVE Juergen Gross
2020-11-20 11:46   ` Juergen Gross via Virtualization
2020-11-20 11:46 ` [PATCH v2 10/12] x86/paravirt: add new macros PVOP_ALT* supporting pvops in ALTERNATIVEs Juergen Gross
2020-11-20 11:46   ` Juergen Gross via Virtualization
2020-11-20 11:46 ` [PATCH v2 11/12] x86/paravirt: switch functions with custom code to ALTERNATIVE Juergen Gross
2020-11-20 11:46   ` Juergen Gross via Virtualization
2020-11-20 15:46   ` kernel test robot
2020-11-20 15:46     ` kernel test robot
2020-11-20 15:46     ` kernel test robot
2020-11-25 15:46   ` [x86/paravirt] fd8d46a7a2: kernel-selftests.livepatch.test-callbacks.sh.fail kernel test robot
2020-11-20 11:46 ` [PATCH v2 12/12] x86/paravirt: have only one paravirt patch function Juergen Gross
2020-11-20 11:46   ` Juergen Gross via Virtualization
2020-11-20 14:18   ` kernel test robot
2020-11-20 14:18     ` kernel test robot
2020-11-20 14:18     ` kernel test robot
2020-11-20 12:53 ` [PATCH v2 00/12] x86: major paravirt cleanup Peter Zijlstra
2020-11-20 12:53   ` Peter Zijlstra
2020-11-23 13:43   ` Peter Zijlstra
2020-11-23 13:43     ` Peter Zijlstra
2020-12-15 11:42     ` Jürgen Groß
2020-12-15 11:42       ` Jürgen Groß via Virtualization
2020-12-15 14:18       ` Peter Zijlstra
2020-12-15 14:18         ` Peter Zijlstra
2020-12-15 14:54         ` Peter Zijlstra
2020-12-15 14:54           ` Peter Zijlstra
2020-12-15 15:07           ` Jürgen Groß
2020-12-15 15:07             ` Jürgen Groß via Virtualization
2020-12-16  0:38           ` Josh Poimboeuf [this message]
2020-12-16  0:38             ` Josh Poimboeuf
2020-12-16  8:40             ` Peter Zijlstra
2020-12-16  8:40               ` Peter Zijlstra
2020-12-16 16:56               ` Josh Poimboeuf
2020-12-16 16:56                 ` Josh Poimboeuf
2020-12-16 17:58                 ` Peter Zijlstra
2020-12-16 17:58                   ` 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=20201216003802.5fpklvx37yuiufrt@treble \
    --to=jpoimboe@redhat.com \
    --cc=boris.ostrovsky@oracle.com \
    --cc=bp@alien8.de \
    --cc=bristot@redhat.com \
    --cc=bsegall@google.com \
    --cc=daniel.lezcano@linaro.org \
    --cc=dietmar.eggemann@arm.com \
    --cc=haiyangz@microsoft.com \
    --cc=hpa@zytor.com \
    --cc=jgross@suse.com \
    --cc=jmattson@google.com \
    --cc=joro@8bytes.org \
    --cc=juri.lelli@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=kys@microsoft.com \
    --cc=linux-hyperv@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=luto@kernel.org \
    --cc=mgorman@suse.de \
    --cc=mingo@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=peterz@infradead.org \
    --cc=pv-drivers@vmware.com \
    --cc=rostedt@goodmis.org \
    --cc=sdeep@vmware.com \
    --cc=sean.j.christopherson@intel.com \
    --cc=sstabellini@kernel.org \
    --cc=sthemmin@microsoft.com \
    --cc=tglx@linutronix.de \
    --cc=vincent.guittot@linaro.org \
    --cc=virtualization@lists.linux-foundation.org \
    --cc=vkuznets@redhat.com \
    --cc=wanpengli@tencent.com \
    --cc=wei.liu@kernel.org \
    --cc=x86@kernel.org \
    --cc=xen-devel@lists.xenproject.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.