public inbox for linux-arm-kernel@lists.infradead.org 
 help / color / mirror / Atom feed
From: marc.zyngier@arm•com (Marc Zyngier)
To: linux-arm-kernel@lists•infradead.org
Subject: [PATCH v2 00/21] arm64: KVM: world switch in C
Date: Tue, 01 Dec 2015 09:58:23 +0000	[thread overview]
Message-ID: <565D6F3F.5030700@arm.com> (raw)
In-Reply-To: <20151130203345.GI11704@cbox>

On 30/11/15 20:33, Christoffer Dall wrote:
> On Fri, Nov 27, 2015 at 06:49:54PM +0000, Marc Zyngier wrote:
>> Once upon a time, the KVM/arm64 world switch was a nice, clean, lean
>> and mean piece of hand-crafted assembly code. Over time, features have
>> crept in, the code has become harder to maintain, and the smallest
>> change is a pain to introduce. The VHE patches are a prime example of
>> why this doesn't work anymore.
>>
>> This series rewrites most of the existing assembly code in C, but keeps
>> the existing code structure in place (most function names will look
>> familiar to the reader). The biggest change is that we don't have to
>> deal with a static register allocation (the compiler does it for us),
>> we can easily follow structure and pointers, and only the lowest level
>> is still in assembly code. Oh, and a negative diffstat.
>>
>> There is still a healthy dose of inline assembly (system register
>> accessors, runtime code patching), but I've tried not to make it too
>> invasive. The generated code, while not exactly brilliant, doesn't
>> look too shaby. I do expect a small performance degradation, but I
>> believe this is something we can improve over time (my initial
>> measurements don't show any obvious regression though).
> 
> I ran this through my experimental setup on m400 and got this:

[...]

> What this tells me is that we do take a noticable hit on the
> world-switch path, which shows up in the TCP_RR and hackbench workloads,
> which have a high precision in their output.
> 
> Note that the memcached number is well within its variability between
> individual benchmark runs, where it varies to 12% of its average in over
> 80% of the executions.
> 
> I don't think this is a showstopper thought, but we could consider
> looking more closely at a breakdown of the world-switch path and verify
> if/where we are really taking a hit.

Thanks for doing so, very interesting. As a data point, what compiler
are you using? I'd expect some variability based on the compiler version...

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...

  parent reply	other threads:[~2015-12-01  9:58 UTC|newest]

Thread overview: 88+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-11-27 18:49 [PATCH v2 00/21] arm64: KVM: world switch in C Marc Zyngier
2015-11-27 18:49 ` [PATCH v2 01/21] arm64: Add macros to read/write system registers Marc Zyngier
2015-11-30 20:00   ` Christoffer Dall
2015-11-27 18:49 ` [PATCH v2 02/21] arm64: KVM: Add a HYP-specific header file Marc Zyngier
2015-11-30 20:00   ` Christoffer Dall
2015-12-01 11:41     ` Marc Zyngier
2015-12-01 11:47       ` Christoffer Dall
2015-11-27 18:49 ` [PATCH v2 03/21] arm64: KVM: Implement vgic-v2 save/restore Marc Zyngier
2015-11-30 20:00   ` Christoffer Dall
2015-12-01 11:39     ` Marc Zyngier
2015-11-27 18:49 ` [PATCH v2 04/21] arm64: KVM: Implement vgic-v3 save/restore Marc Zyngier
2015-11-30  9:59   ` Alex Bennée
2015-11-30 10:43     ` Marc Zyngier
2015-11-30 19:50   ` Christoffer Dall
2015-12-01 11:32     ` Marc Zyngier
2015-12-01 11:44       ` Christoffer Dall
2015-12-01 11:50         ` Christoffer Dall
2015-12-01 11:57           ` Marc Zyngier
2015-12-01 12:24             ` Christoffer Dall
2015-12-01 12:49               ` Marc Zyngier
2015-12-01 11:54         ` Marc Zyngier
2015-11-27 18:49 ` [PATCH v2 05/21] arm64: KVM: Implement timer save/restore Marc Zyngier
2015-11-30 19:59   ` Christoffer Dall
2015-12-01 11:34     ` Marc Zyngier
2015-11-27 18:50 ` [PATCH v2 06/21] arm64: KVM: Implement system register save/restore Marc Zyngier
2015-12-01 15:53   ` Christoffer Dall
2015-11-27 18:50 ` [PATCH v2 07/21] arm64: KVM: Implement 32bit " Marc Zyngier
2015-12-01 15:52   ` Christoffer Dall
2015-11-27 18:50 ` [PATCH v2 08/21] arm64: KVM: Implement debug save/restore Marc Zyngier
2015-11-30 12:00   ` Alex Bennée
2015-11-30 12:24     ` Marc Zyngier
2015-12-01 12:56   ` Christoffer Dall
2015-12-01 13:06     ` Marc Zyngier
2015-12-01 13:19       ` Alex Bennée
2015-12-01 13:34         ` Marc Zyngier
2015-12-01 14:47       ` Christoffer Dall
2015-12-01 14:56         ` Christoffer Dall
2015-12-01 15:01         ` Marc Zyngier
2015-12-01 15:41           ` Christoffer Dall
2015-12-01 18:34             ` Marc Zyngier
2015-11-27 18:50 ` [PATCH v2 09/21] arm64: KVM: Implement guest entry Marc Zyngier
2015-12-01 15:29   ` Christoffer Dall
2015-12-01 18:41     ` Marc Zyngier
2015-11-27 18:50 ` [PATCH v2 10/21] arm64: KVM: Add patchable function selector Marc Zyngier
2015-12-01 15:39   ` Christoffer Dall
2015-12-01 18:51     ` Marc Zyngier
2015-12-02  9:27       ` Christoffer Dall
2015-12-02  9:47         ` Marc Zyngier
2015-12-02 11:53           ` Christoffer Dall
2015-12-02 13:19             ` Marc Zyngier
2015-12-02 16:19               ` Christoffer Dall
2015-12-02 22:34   ` Andrew Jones
2015-12-03  8:18     ` Marc Zyngier
2015-11-27 18:50 ` [PATCH v2 11/21] arm64: KVM: Implement the core world switch Marc Zyngier
2015-12-01 15:55   ` Christoffer Dall
2015-11-27 18:50 ` [PATCH v2 12/21] arm64: KVM: Implement fpsimd save/restore Marc Zyngier
2015-12-02 11:53   ` Christoffer Dall
2015-12-02 15:29     ` Marc Zyngier
2015-12-02 16:19       ` Christoffer Dall
2015-11-27 18:50 ` [PATCH v2 13/21] arm64: KVM: Implement TLB handling Marc Zyngier
2015-12-02 11:53   ` Christoffer Dall
2015-11-27 18:50 ` [PATCH v2 14/21] arm64: KVM: HYP mode entry points Marc Zyngier
2015-12-02 11:53   ` Christoffer Dall
2015-11-27 18:50 ` [PATCH v2 15/21] arm64: KVM: Add panic handling Marc Zyngier
2015-12-02 11:53   ` Christoffer Dall
2015-11-27 18:50 ` [PATCH v2 16/21] arm64: KVM: Add compatibility aliases Marc Zyngier
2015-12-02 11:49   ` Christoffer Dall
2015-12-02 15:23     ` Marc Zyngier
2015-11-27 18:50 ` [PATCH v2 17/21] arm64: KVM: Map the kernel RO section into HYP Marc Zyngier
2015-12-02 11:49   ` Christoffer Dall
2015-11-27 18:50 ` [PATCH v2 18/21] arm64: KVM: Move away from the assembly version of the world switch Marc Zyngier
2015-12-02 11:49   ` Christoffer Dall
2015-11-27 18:50 ` [PATCH v2 19/21] arm64: KVM: Turn system register numbers to an enum Marc Zyngier
2015-12-02 11:51   ` Christoffer Dall
2015-12-02 15:26     ` Marc Zyngier
2015-11-27 18:50 ` [PATCH v2 20/21] arm64: KVM: Cleanup asm-offset.c Marc Zyngier
2015-12-02 11:51   ` Christoffer Dall
2015-11-27 18:50 ` [PATCH v2 21/21] arm64: KVM: Remove weak attributes Marc Zyngier
2015-12-02 11:47   ` Christoffer Dall
2015-12-02 15:21     ` Marc Zyngier
2015-12-02 16:21       ` Christoffer Dall
2015-12-02 17:52         ` Marc Zyngier
2015-11-30 20:33 ` [PATCH v2 00/21] arm64: KVM: world switch in C Christoffer Dall
2015-12-01  3:19   ` Mario Smarduch
2015-12-01  9:58   ` Marc Zyngier [this message]
2015-12-01 12:00     ` Christoffer Dall
2015-12-01 17:51       ` Marc Zyngier
2015-12-01 19:34         ` Christoffer Dall

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=565D6F3F.5030700@arm.com \
    --to=marc.zyngier@arm$(echo .)com \
    --cc=linux-arm-kernel@lists$(echo .)infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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