[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <01d2f33c-6e50-ae88-73ff-84042504c26e@gmail.com>
Date: Thu, 5 Aug 2021 23:51:43 +0800
From: Tianyu Lan <ltykernel@...il.com>
To: Dave Hansen <dave.hansen@...el.com>,
Peter Zijlstra <peterz@...radead.org>
Cc: kys@...rosoft.com, haiyangz@...rosoft.com, sthemmin@...rosoft.com,
wei.liu@...nel.org, decui@...rosoft.com, tglx@...utronix.de,
mingo@...hat.com, bp@...en8.de, x86@...nel.org, hpa@...or.com,
dave.hansen@...ux.intel.com, luto@...nel.org,
konrad.wilk@...cle.com, boris.ostrovsky@...cle.com,
jgross@...e.com, sstabellini@...nel.org, joro@...tes.org,
will@...nel.org, davem@...emloft.net, kuba@...nel.org,
jejb@...ux.ibm.com, martin.petersen@...cle.com, arnd@...db.de,
hch@....de, m.szyprowski@...sung.com, robin.murphy@....com,
Tianyu.Lan@...rosoft.com, rppt@...nel.org,
kirill.shutemov@...ux.intel.com, akpm@...ux-foundation.org,
brijesh.singh@....com, thomas.lendacky@....com, pgonda@...gle.com,
david@...hat.com, krish.sadhukhan@...cle.com, saravanand@...com,
aneesh.kumar@...ux.ibm.com, xen-devel@...ts.xenproject.org,
martin.b.radev@...il.com, ardb@...nel.org, rientjes@...gle.com,
tj@...nel.org, keescook@...omium.org,
michael.h.kelley@...rosoft.com, iommu@...ts.linux-foundation.org,
linux-arch@...r.kernel.org, linux-hyperv@...r.kernel.org,
linux-kernel@...r.kernel.org, linux-scsi@...r.kernel.org,
netdev@...r.kernel.org, vkuznets@...hat.com, parri.andrea@...il.com
Subject: Re: [PATCH V2 03/14] x86/set_memory: Add x86_set_memory_enc static
call support
On 8/5/2021 10:29 PM, Dave Hansen wrote:
> On 8/5/21 7:23 AM, Peter Zijlstra wrote:
>> This is assuming any of this is actually performance critical, based off
>> of this using static_call() to begin with.
>
> This code is not performance critical.
>
> I think I sent folks off on a wild goose chase when I asked that we make
> an effort to optimize code that does:
>
> if (some_hyperv_check())
> foo();
>
> if (some_amd_feature_check())
> bar();
>
> with checks that will actually compile away when Hyper-V or
> some_amd_feature() is disabled. That's less about performance and just
> about good hygiene. I *wanted* to see
> cpu_feature_enabled(X86_FEATURE...) checks.
>
> Someone suggested using static calls, and off we went...
>
> Could we please just use cpu_feature_enabled()?
>
Yes, cpu_feature_enabled() works. The target is just to run platform
code after platform check. I will update this in the next version.
Powered by blists - more mailing lists