On Nov 3, 2021, at 11:23 PM, gaoliming <gaoliming@...> wrote:

Lefi:

-----邮件原件-----
发件人: devel@edk2.groups.io <devel@edk2.groups.io> 代表 Leif Lindholm
发送时间: 2021年11月2日 17:51
收件人: brbarkel@... <bret@...>
抄送: devel@edk2.groups.io; Ard Biesheuvel <ardb+tianocore@...>;
Michael D Kinney <michael.d.kinney@...>; Liming Gao
<gaoliming@...>; Zhiguang Liu <zhiguang.liu@...>; Sami
Mujawar <sami.mujawar@...>; Jiewen Yao <jiewen.yao@...>;
Supreeth Venkatesh <supreeth.venkatesh@...>; Maciej Rabeda
<maciej.rabeda@...>; Jiaxin Wu <jiaxin.wu@...>; Siyuan
Fu <siyuan.fu@...>; Ray Ni <ray.ni@...>; Zhichao Gao
<zhichao.gao@...>; Sean Brogan <sean.brogan@...>
主题: Re: [edk2-devel] [PATCH v1 05/16] ArmPkg and MdePkg: Move the Arm
CompilerIntrinsicsLib to MdePkg

On Mon, Nov 01, 2021 at 12:56:37 -0700, brbarkel@... wrote:

From: Bret Barkelew <brbarkel@...>

This aligns better with Mu's philosophy around dependency structuring
and is one of the steps to enable Basecore to have zero CI dependencies
on other Mu repos.

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3649

Only one comment on this patch really.

ArmPkg/Library/CompilerIntrinsicsLib/CompilerIntrinsicsLib.inf =>

MdePkg/Library/CompilerIntrinsicsLib/ArmCompilerIntrinsicsLib.inf | 3 +--

Could we just keep the .inf name as is?

I agree your suggestion. CompilerIntrinsicsLib can support X86 in future.
So,
CompilerIntrinsicsLib.inf should be used.

Thanks
Liming

I think we're getting closer to the x86 folk accepting that they
need this too, rather than keep inventing new dialects of C in the
desperate hope that the compiler won't generate stdlib calls it's
fully permitted to generate whenever it feels like.

/
Leif