[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <0e04d1be-be64-c5c3-b167-f28376b7bdc8@loongson.cn>
Date: Tue, 12 Sep 2023 09:57:59 +0800
From: zhaotianrui <zhaotianrui@...ngson.cn>
To: WANG Xuerui <kernel@...0n.name>, linux-kernel@...r.kernel.org,
kvm@...r.kernel.org
Cc: Paolo Bonzini <pbonzini@...hat.com>,
Huacai Chen <chenhuacai@...nel.org>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
loongarch@...ts.linux.dev, Jens Axboe <axboe@...nel.dk>,
Mark Brown <broonie@...nel.org>,
Alex Deucher <alexander.deucher@....com>,
Oliver Upton <oliver.upton@...ux.dev>, maobibo@...ngson.cn,
Xi Ruoyao <xry111@...111.site>,
kernel test robot <lkp@...el.com>
Subject: Re: [PATCH v20 28/30] LoongArch: KVM: Enable kvm config and add the
makefile
在 2023/9/11 下午3:30, WANG Xuerui 写道:
> On 8/31/23 16:30, Tianrui Zhao wrote:
>> [snip]
>> +
>> +config AS_HAS_LVZ_EXTENSION
>> + def_bool $(as-instr,hvcl 0)
>
> Upon closer look it looks like this piece could use some improvement
> as well: while presence of "hvcl" indeed always implies full support
> for LVZ instructions, "hvcl" however isn't used anywhere in the
> series, so it's not trivially making sense. It may be better to test
> for an instruction actually going to be used like "gcsrrd". What do
> you think?
Thanks for your advice, and I think both "hvcl" and "gcsrrd"
instructions which are supported by binutils mean that the cpu support
LVZ extension, so I think they have the same meaning there.
Thanks
Tianrui Zhao
Powered by blists - more mailing lists