On Thu, Apr 4, 2019 at 1:00 PM Andrew Fish &= lt;afish@apple.com> wrote:

Gary,

The edk2 is freestanding UEFI = firmware and the runtime does not support floating point exceptions.=C2=A0<= /div>

I think a long time ago there was an optimization = bug and the -mno-sse fixed it.=C2=A0

The UEFI x64 = ABI is Windows like and not Sys V like.=C2=A0

Thanks= ,

Andrew Fish

On Apr 4, 2019, at 10:13 AM, Gary Zibrat via Groups.Io <gzibrat=3Dgoo= gle.com@groups.io> wrote:

Does anyone know = why -mno-sse is specified on Clang x64?
https://github.com/tianocore/edk2/commit/6f756db5ea057ff10dd16= 8c7627163cec046a3d9
There doesn't seem to be a corres= ponding flag for other compilers.

I ask becaus= e a commit in LLVM breaks compilation of EDK2:
https://review= s.llvm.org/rL357317#change-QSW9B7DqCGUI

--0000000000004bcde10585b9f3e9-- From mboxrd@z Thu Jan 1 00:00:00 1970 Authentication-Results: mx.groups.io; dkim=pass header.i=@apple.com header.s=20180706 header.b=wRnGAZzM; spf=pass (domain: apple.com, ip: 17.151.62.66, mailfrom: afish@apple.com) Received: from nwk-aaemail-lapp01.apple.com (nwk-aaemail-lapp01.apple.com [17.151.62.66]) by groups.io with SMTP; Thu, 04 Apr 2019 13:12:43 -0700 Received: from pps.filterd (nwk-aaemail-lapp01.apple.com [127.0.0.1]) by nwk-aaemail-lapp01.apple.com (8.16.0.27/8.16.0.27) with SMTP id x34K2J77033615; Thu, 4 Apr 2019 13:12:43 -0700 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=apple.com; h=mime-version : content-type : sender : from : subject : date : references : to : in-reply-to : message-id; s=20180706; bh=Y60N/Kki+AUufQAmOKzC+Z2eiUSoLl/R2eu5C9NLSa4=; b=wRnGAZzMebWOBb+0xe/TGcUcvsemF+mnQENkhTi+WLJP0xjlkxPsQtgZ7XQ1ZP/okH5e fr6ISH9IspKxKGorjqkwygpP0WwY/RxKFvZCDHCxcLhbVlGwBaO52h4rYqESabjjX36a C5y6FivZcfCst5r+gLCT1ZWZfDaFPzoSja/WcI3N4z1stZ2UPig6cPmbgzhNYxOVmrjA E/WIiSZ048G/NXpzNx7JApe3ieBXnpqMxO/z9b6fe6LttOQIs4XSUZVg0SHE4DCptCwS RH0MiSRIhPXOsS0fRPcbAyZ18hYNsuoNXm9+EH0SSIQmBRCRr6qI8nOiIpdYxe5eNbeQ Yw== Received: from ma1-mtap-s03.corp.apple.com (ma1-mtap-s03.corp.apple.com [17.40.76.7]) by nwk-aaemail-lapp01.apple.com with ESMTP id 2rmg25d38x-2 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NO); Thu, 04 Apr 2019 13:12:43 -0700 MIME-version: 1.0 Received: from nwk-mmpp-sz13.apple.com (nwk-mmpp-sz13.apple.com [17.128.115.216]) by ma1-mtap-s03.corp.apple.com (Oracle Communications Messaging Server 8.0.2.3.20181024 64bit (built Oct 24 2018)) with ESMTPS id <0PPG0075KDH51OF0@ma1-mtap-s03.corp.apple.com>; Thu, 04 Apr 2019 13:12:42 -0700 (PDT) Received: from process_milters-daemon.nwk-mmpp-sz13.apple.com by nwk-mmpp-sz13.apple.com (Oracle Communications Messaging Server 8.0.2.3.20181024 64bit (built Oct 24 2018)) id <0PPG00H00CGQTC00@nwk-mmpp-sz13.apple.com>; Thu, 04 Apr 2019 13:12:41 -0700 (PDT) X-Va-A: X-Va-T-CD: c01e0da5a1b2c363686fd10c98c9e006 X-Va-E-CD: 741062ba1fafd478fc0525dab04cad1c X-Va-R-CD: 875399ea7acc22cd0b728349e28566db X-Va-CD: 0 X-Va-ID: 70175b01-5652-4c63-8438-7d0090729f54 X-V-A: X-V-T-CD: adba116a933446bf0bfca850aefd636f X-V-E-CD: 741062ba1fafd478fc0525dab04cad1c X-V-R-CD: 875399ea7acc22cd0b728349e28566db X-V-CD: 0 X-V-ID: c6f01379-af94-4109-92b1-20ed792e6a2a X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:,, definitions=2019-04-04_11:,, signatures=0 Received: from [17.226.41.11] (unknown [17.226.41.11]) by nwk-mmpp-sz13.apple.com (Oracle Communications Messaging Server 8.0.2.3.20181024 64bit (built Oct 24 2018)) with ESMTPSA id <0PPG004KEDH50R10@nwk-mmpp-sz13.apple.com>; Thu, 04 Apr 2019 13:12:41 -0700 (PDT) Sender: afish@apple.com From: "Andrew Fish" Subject: Re: [edk2-devel] Why is specified -mno-sse on clang x64? Date: Thu, 04 Apr 2019 13:12:38 -0700 References: <2A7A0205-FADC-4E8F-B744-E0B5C476A2C6@apple.com> To: devel@edk2.groups.io, gzibrat@google.com In-reply-to: Message-id: X-Mailer: Apple Mail (2.3445.6.18) X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:,, definitions=2019-04-04_11:,, signatures=0 Content-type: multipart/alternative; boundary="Boundary_(ID_WAuoJQlwq7d6iq2pTWHg8g)" --Boundary_(ID_WAuoJQlwq7d6iq2pTWHg8g) Content-type: text/plain; CHARSET=US-ASCII Content-transfer-encoding: 7BIT > On Apr 4, 2019, at 1:09 PM, Gary Zibrat via Groups.Io wrote: > > It seems like the "-mno-mmx -msoft-float -mno-implicit-float" part of the flags should be enough to handle that now. > Is it ok to take -mno-sse out (in my local branch) for new versions of clang without having to worry about silent failures? > Since EFIAPI uses the windows ABI, it should be ok to use XMM registers for integer operations. > ' Gary, Conceptually it should be OK. But obviously no one has tested it :). Thanks, Andrew Fish > On Thu, Apr 4, 2019 at 1:00 PM Andrew Fish > wrote: > Gary, > > The edk2 is freestanding UEFI firmware and the runtime does not support floating point exceptions. > > I think a long time ago there was an optimization bug and the -mno-sse fixed it. > > The UEFI x64 ABI is Windows like and not Sys V like. > > Thanks, > > Andrew Fish > >> On Apr 4, 2019, at 10:13 AM, Gary Zibrat via Groups.Io > wrote: >> >> Does anyone know why -mno-sse is specified on Clang x64? >> https://github.com/tianocore/edk2/commit/6f756db5ea057ff10dd168c7627163cec046a3d9 >> There doesn't seem to be a corresponding flag for other compilers. >> >> I ask because a commit in LLVM breaks compilation of EDK2: >> https://reviews.llvm.org/rL357317#change-QSW9B7DqCGUI > > --Boundary_(ID_WAuoJQlwq7d6iq2pTWHg8g) Content-type: text/html; CHARSET=US-ASCII Content-transfer-encoding: quoted-printable

On Apr 4,= 2019, at 1:09 PM, Gary Zibrat via Groups.Io <gzibrat=3Dgoogle.com@groups.io>= wrote:

It seems like the "-mno-mmx -= msoft-float -mno-implicit-float" part of the flags should be enough to hand= le that now.
Is it ok to tak= e -mno-sse out (in my local branch) for new versions of clang without havin= g to worry about silent failures?
Since EF= IAPI uses the windows ABI, it should be ok to use XMM registers for integer= operations.
'

Gary,

Conceptually it should be OK. But obviously no one has tested it :).&= nbsp;

Thanks,

=
Andrew Fish

On Thu, Apr 4,= 2019 at 1:00 PM Andrew Fish <afish@apple.com> wrote:
Gary,
The edk2 is freestanding U= EFI firmware and the runtime does not support floating point exceptions.&nb= sp;

I think a lon= g time ago there was an optimization bug and the -mno-sse fixed it.

The UEFI x64 ABI i= s Windows like and not Sys V like.

Thanks,

Andrew Fish

On Apr 4, 2019,= at 10:13 AM, Gary Zibrat via Groups.Io <gzibrat=3Dgoogle.com@groups= .io> wrote:

Does anyone know why -mno-sse is specified on Clang x64?
https://github.com/t= ianocore/edk2/commit/6f756db5ea057ff10dd168c7627163cec046a3d9
There doesn't seem to be a corresponding flag = for other compilers.

I ask because a commit in LLVM breaks compilation of EDK= 2:
https://reviews.llvm.org/rL357317#change-QSW9B7= DqCGUI

--Boundary_(ID_WAuoJQlwq7d6iq2pTWHg8g)-- From mboxrd@z Thu Jan 1 00:00:00 1970 Authentication-Results: mx.groups.io; dkim=pass header.i=@google.com header.s=20161025 header.b=E7uYJvFe; spf=pass (domain: google.com, ip: 209.85.166.170, mailfrom: gzibrat@google.com) Received: from mail-it1-f170.google.com (mail-it1-f170.google.com [209.85.166.170]) by groups.io with SMTP; Mon, 08 Apr 2019 14:12:16 -0700 Received: by mail-it1-f170.google.com with SMTP id y134so1490222itc.5 for ; Mon, 08 Apr 2019 14:12:16 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20161025; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=trvA/UZGnsoLm8nIDLCHdLkYIPNSVT8pmuauEgCI+/M=; b=E7uYJvFe6EuyFwQmKhq/yrelA8enPCbJswoKgZiQqahIn3+SSdVwf1+8/16EEvOymt uBeYqkW1LoFYSKeqVlAGXqkEVAjnlj5hvRUuDEUptS+elWWMsaouybcgBpS6H09uMAzP 0HxJ0BPoPVM+PDrRMaFUimSX8mWnLNQDOBHZLH4jm3cjEWJGsAe1NpBWnmhm8iMICqcQ RCSVkN1YHQojEkbbLBiqyCqrvLyed6nzuPUegITkvM43tkK9RkYOD8zV+yHp15mbTv5+ oByWYsVOMO8CUPLAqh5JCyuAnZl5yWrmbeL50PTI0IyalrRns9syGv7dYH+J62dELnlK ZSKw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=trvA/UZGnsoLm8nIDLCHdLkYIPNSVT8pmuauEgCI+/M=; b=DYrQwfwvqdN+dRbpTfHQyrRs4IHV1RVT4tdakn+2fsyoBgRDabYBpD9ThJN4/igqMs A4cfW/ifj3ZMfqLLDM3UzAgVNXyZSvlsexl+tSA4zMHNCea2cZcfSWCc7TBE/k+IntM4 kuZN1yJeLwjwJnwYDQnKeRkTQZB9EufZhKAU3kLOd8bIGQqz+9JonMD23ZHSpQBBm4aW 49vMRz8zj5zS9qIvn8pazeY7wq3L6KfhyNhtbDRn8RRxhB4E9KKHTPJSqdM5Yyo/iIbu 6AVAAaAbEYbep+Vc0BqQ7j2ePm4wk1VEvkRsgiFi93Rit32Go0BzYyoci4BUaTZrR98D 4JKA== X-Gm-Message-State: APjAAAXUilYrTps3zWBlfqmYEvT1omWgW31XNvy8H3U/x5yi7k3HeRWI OIqSxseRrA/5KC59ePZskfkPxsmgUm8zfB3y/S4KPw== X-Google-Smtp-Source: APXvYqxqC0bFLj0RPHP90jiZuS8xjqUMY0a206cyT8vIAgNBq9FmBr3rWz1o8QfQbrUYA9jYotm/zXCmsi8FSNc7Zok= X-Received: by 2002:a24:c544:: with SMTP id f65mr22362289itg.90.1554757935259; Mon, 08 Apr 2019 14:12:15 -0700 (PDT) MIME-Version: 1.0 References: <2A7A0205-FADC-4E8F-B744-E0B5C476A2C6@apple.com>

In-Reply-To: From: "Gary Zibrat" Date: Mon, 8 Apr 2019 14:12:04 -0700 Message-ID: Subject: Re: [edk2-devel] Why is specified -mno-sse on clang x64? To: Andrew Fish Cc: devel@edk2.groups.io Content-Type: multipart/alternative; boundary="00000000000071a94a05860b4ad4" --00000000000071a94a05860b4ad4 Content-Type: text/plain; charset="UTF-8" I've messed around with it locally and nothing seemed to stop working. Would there be any interest in adding a newer clang configuration with -mno-sse removed? On Thu, Apr 4, 2019 at 1:12 PM Andrew Fish wrote: > > > On Apr 4, 2019, at 1:09 PM, Gary Zibrat via Groups.Io < > gzibrat=google.com@groups.io> wrote: > > It seems like the "-mno-mmx -msoft-float -mno-implicit-float" part of the > flags should be enough to handle that now. > Is it ok to take -mno-sse out (in my local branch) for new versions of > clang without having to worry about silent failures? > Since EFIAPI uses the windows ABI, it should be ok to use XMM registers > for integer operations. > ' > > > Gary, > > Conceptually it should be OK. But obviously no one has tested it :). > > Thanks, > > Andrew Fish > > > On Thu, Apr 4, 2019 at 1:00 PM Andrew Fish wrote: > >> Gary, >> >> The edk2 is freestanding UEFI firmware and the runtime does not support >> floating point exceptions. >> >> I think a long time ago there was an optimization bug and the -mno-sse >> fixed it. >> >> The UEFI x64 ABI is Windows like and not Sys V like. >> >> Thanks, >> >> Andrew Fish >> >> On Apr 4, 2019, at 10:13 AM, Gary Zibrat via Groups.Io < >> gzibrat=google.com@groups.io> wrote: >> >> Does anyone know why -mno-sse is specified on Clang x64? >> >> https://github.com/tianocore/edk2/commit/6f756db5ea057ff10dd168c7627163cec046a3d9 >> There doesn't seem to be a corresponding flag for other compilers. >> >> I ask because a commit in LLVM breaks compilation of EDK2: >> https://reviews.llvm.org/rL357317#change-QSW9B7DqCGUI >> >> >> > > > --00000000000071a94a05860b4ad4 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable

I've messed around with it locally and nothing seemed = to stop working.

Would there be any interest in adding a newer clang co= nfiguration with -mno-sse removed?

On Thu, Apr 4, 2019 at 1:12 PM Andr= ew Fish <afish@apple.com> wrot= e:

On= Apr 4, 2019, at 1:09 PM, Gary Zibrat via Groups.Io <gzibrat=3Dgoogle.com@group= s.io> wrote:

It seems like the &q= uot;-mno-mmx -msoft-float -mno-implicit-float" part of the flags shoul= d be enough to handle that now.
Is it ok to take -mno-sse out= (in my local branch) for new versions of clang without having to worry abo= ut silent failures?
Since EFIAPI uses the windows ABI= , it should be ok to use XMM registers for integer operations.
&= #39;

Gary,
<= div>
Conceptually it should be OK. But obviously no one has t= ested it :).=C2=A0

Thanks,

Andrew Fish

= On Thu, Apr 4, 2019 at 1:00 PM Andrew Fish <afish@apple.com> wrote:
Gary,

The e= dk2 is freestanding UEFI firmware and the runtime does not support floating= point exceptions.=C2=A0

I think a long time ago t= here was an optimization bug and the -mno-sse fixed it.=C2=A0
The UEFI x64 ABI is Windows like and not Sys V like.=C2=A0
<= div>
Thanks,

Andrew Fish
On Apr 4, 2019, at 10:13 AM, Gary Zibrat vi= a Groups.Io <gzibrat=3Dgoogle.com@groups.io> wrote:

Does anyone know why -mno-sse is specified = on Clang x64?
https://github= .com/tianocore/edk2/commit/6f756db5ea057ff10dd168c7627163cec046a3d9
=
There doesn't seem to be a corresponding flag for other comp= ilers.

I ask because a commit in LLVM breaks c= ompilation of EDK2:
https://reviews.llvm.org/rL357317#change-= QSW9B7DqCGUI