From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received-SPF: Pass (sender SPF authorized) identity=mailfrom; client-ip=210.71.195.43; helo=out03.hibox.biz; envelope-from=tim.lewis@insyde.com; receiver=edk2-devel@lists.01.org Received: from out03.hibox.biz (out03.hibox.biz [210.71.195.43]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ml01.01.org (Postfix) with ESMTPS id 79FFE207E5401 for ; Tue, 8 May 2018 18:28:00 -0700 (PDT) X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: =?us-ascii?q?A2D5AADlTfJa/ww0GKxCGhkBAQEBAQEBA?= =?us-ascii?q?QEBAQEHAQEBAQGDGCthF2Moi3FejhExAV2CFpESFIFkCxgLCAGEOgQCAoJkIjQ?= =?us-ascii?q?YAQIBAQEBAQECbBwMhSgBAQEBAwEBBgIeLggQFwEDAgYDDQQBAwEBKAcZDgERD?= =?us-ascii?q?QMBBQgCBAESCwWDDgQsgT4DFQ8wqGAxGgKIJYJIgiuEfYJRP4EPgwuCehcBAQE?= =?us-ascii?q?BAYEihg8CjB8qLoszCAKBZ4N8hQyDXIE1PIV0hH2HQYFON4RAgkaBJRw4DYFFc?= =?us-ascii?q?FCCQwk2gWEXegEIgkKFFIVeHzABgQwIDAGNdSuCGQEB?= X-IronPort-AV: E=Sophos;i="5.49,379,1520870400"; d="scan'208";a="7313985" Received: from unknown (HELO hb3-BKT202.hibox.biz) ([172.24.52.12]) by out03.hibox.biz with ESMTP; 09 May 2018 09:27:56 +0800 Received: from unknown (HELO hb3-BKT103.hibox.biz) ([172.24.51.13]) by hb3-BKT202.hibox.biz with ESMTP; 09 May 2018 09:27:55 +0800 Received: from unknown (HELO hb3-IN02.hibox.biz) ([172.24.12.12]) by hb3-BKT103.hibox.biz with ESMTP; 09 May 2018 09:27:56 +0800 X-Remote-IP: 73.116.1.175 X-Remote-Host: c-73-116-1-175.hsd1.ca.comcast.net X-SBRS: -10.0 X-MID: 11758225 X-Auth-ID: tim.lewis@insyde.com X-EnvelopeFrom: tim.lewis@insyde.com hiBox-Sender: 1 Received: from c-73-116-1-175.hsd1.ca.comcast.net (HELO DESKTOPAVHFBJF) ([73.116.1.175]) by hb3-IN02.hibox.biz with ESMTP/TLS/AES256-SHA; 09 May 2018 09:27:54 +0800 From: "Tim Lewis" To: "'Yao, Jiewen'" , , References: <74D8A39837DF1E4DA445A8C0B3885C503AB98291@shsmsx102.ccr.corp.intel.com> <74D8A39837DF1E4DA445A8C0B3885C503AB98303@shsmsx102.ccr.corp.intel.com> <74D8A39837DF1E4DA445A8C0B3885C503AB983CA@shsmsx102.ccr.corp.intel.com> In-Reply-To: <74D8A39837DF1E4DA445A8C0B3885C503AB983CA@shsmsx102.ccr.corp.intel.com> Date: Tue, 8 May 2018 18:27:49 -0700 Message-ID: <003701d3e734$f31a7770$d94f6650$@insyde.com> MIME-Version: 1.0 X-Mailer: Microsoft Outlook 16.0 Thread-Index: AQH+/tuOV9z+ScBXPyNfb+V7j60Y7AHO5MhdAghrJxgCPMX98AH2v1pvAUVkrF2jhen+YA== Subject: Re: [PATCH] MdePkg/BaseLib: Globally include ChkStk.c X-BeenThere: edk2-devel@lists.01.org X-Mailman-Version: 2.1.26 Precedence: list List-Id: EDK II Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 09 May 2018 01:28:05 -0000 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Content-Language: en-us I think that this is a fatal error in EDK2. It basically says, "we are = out of stack space." The alternative is: the system hangs in an unexpected = way since the stack overflows into other pages. Tim -----Original Message----- From: edk2-devel On Behalf Of Yao, = Jiewen Sent: Tuesday, May 8, 2018 5:25 PM To: Marvin.Haeuser@outlook.com; edk2-devel@lists.01.org Subject: Re: [edk2] [PATCH] MdePkg/BaseLib: Globally include ChkStk.c We discussed internally in Intel. The quick workaround is: use /Gs65536. :-) At the same time, our recommendation is to revisit the code which = triggers this error. Why this function need such a big stack? And try to reduce = the local stack usage. What is why we still use /Gs32768 as default, instead of /Gs65536. Thank you Yao Jiewen > -----Original Message----- > From: edk2-devel [mailto:edk2-devel-bounces@lists.01.org] On Behalf Of = > Marvin H?user > Sent: Tuesday, May 8, 2018 5:21 PM > To: edk2-devel@lists.01.org > Cc: Yao, Jiewen > Subject: Re: [edk2] [PATCH] MdePkg/BaseLib: Globally include ChkStk.c >=20 > Oh, if you are prefering a good implementation, I will be all for=20 > that. This was just a 'quick workaround', same as currently done for = GCC. > I'm actually unsure whether a good implementation is possible with a=20 > flat memory model. It will likely be mere luck whether there is enough = > space free below the stack, except for maybe when it's located very=20 > high (preferably past the 32-bit space). > Has there been any prior discussion on this topic? Would be interested = > to follow up if there was. >=20 > Thanks, > Marvin >=20 > > -----Original Message----- > > From: Yao, Jiewen > > Sent: Wednesday, May 9, 2018 2:13 AM > > To: Marvin.Haeuser@outlook.com; edk2-devel@lists.01.org > > Cc: Gao, Liming > > Subject: RE: [PATCH] MdePkg/BaseLib: Globally include ChkStk.c > > > > There are some open source implementation: > > MSVC: https://github.com/ispc/ispc/issues/542 > > GCC: > > https://android.googlesource.com/toolchain/gcc/+/b094d6c4bf572654a03 > > 1e cc4afe675154c886dc5/gcc-4.4.3/gcc/config/i386/chkstk.asm > > > > The compiler generated code assumes the stack is enlarged after we=20 > > can chkstk. > > > > I agree empty function can make build pass. > > But more important, we need make it work if there is a need to=20 > > increase the stack. > > The potential issue is that the later code (after chkstk) assumes=20 > > the stack is increased, but actually it is not. > > > > That is why I ask how this is validated. > > > > Thank you > > Yao Jiewen > > > > > > > -----Original Message----- > > > From: Marvin H=E4user [mailto:Marvin.Haeuser@outlook.com] > > > Sent: Tuesday, May 8, 2018 4:58 PM > > > To: edk2-devel@lists.01.org > > > Cc: Yao, Jiewen ; Gao, Liming=20 > > > > > > Subject: RE: [PATCH] MdePkg/BaseLib: Globally include ChkStk.c > > > > > > Hey Yao, > > > > > > As far as I am aware, all __chkstk does is access the stack in 4=20 > > > KB intervals from the current location to the maximum requested=20 > > > location to eventually hit the Windows Guard Page, which then=20 > > > triggers the stack > > increase. > > > Because I do not know of any equivalent concept in edk2 and the=20 > > > intrinsic was already disabled for GCC, I supposed it was a good=20 > > > idea to do so globally. Are the potential issues I am not aware = of? > > > > > > Thanks, > > > Marvin. > > > > > > > -----Original Message----- > > > > From: Yao, Jiewen > > > > Sent: Wednesday, May 9, 2018 1:52 AM > > > > To: Marvin.Haeuser@outlook.com; edk2-devel@lists.01.org > > > > Cc: Gao, Liming > > > > Subject: RE: [PATCH] MdePkg/BaseLib: Globally include ChkStk.c > > > > > > > > HI Marvin > > > > Would you mind to share the information on how you test this = update? > > > > > > > > > > > > Per my experience, chkstk not only does the check but also does=20 > > > > the real work to allocate more stack. > > > > > > > > /Gs can be used to indicate the size (/Gs[num] control stack=20 > > > > checking calls) > > > > > > > > We use /Gs32768 by default, see tools_def.txt. > > > > > > > > Usually, we just increase this value to override the default=20 > > > > one, if we hit this issue. > > > > > > > > Thank you > > > > Yao Jiewen > > > > > > > > > -----Original Message----- > > > > > From: edk2-devel [mailto:edk2-devel-bounces@lists.01.org] On=20 > > > > > Behalf Of Marvin H?user > > > > > Sent: Tuesday, May 8, 2018 4:37 PM > > > > > To: edk2-devel@lists.01.org > > > > > Cc: Gao, Liming > > > > > Subject: Re: [edk2] [PATCH] MdePkg/BaseLib: Globally include=20 > > > > > ChkStk.c > > > > > > > > > > Hey Liming, > > > > > > > > > > According to the MSDN documentation, the call will be inserted = > > > > > for a stack usage past a defined threshold - 4 KB for IA32, 8=20 > > > > > KB for X64 > > (source: > > > > > = https://msdn.microsoft.com/en-us/library/ms648426(v=3Dvs.85).aspx). > > > > > > > > > > Thanks, > > > > > Marvin > > > > > > > > > > > -----Urspr=FCngliche Nachricht----- > > > > > > Von: Gao, Liming > > > > > > Gesendet: Montag, 7. Mai 2018 14:16 > > > > > > An: Marvin.Haeuser@outlook.com; edk2-devel@lists.01.org > > > > > > Cc: Kinney, Michael D > > > > > > Betreff: RE: [PATCH] MdePkg/BaseLib: Globally include=20 > > > > > > ChkStk.c > > > > > > > > > > > > Marvin: > > > > > > In VS compiler, what case will trig this intrinsic = function? > > > > > > > > > > > > Thanks > > > > > > Liming > > > > > > > -----Original Message----- > > > > > > > From: edk2-devel [mailto:edk2-devel-bounces@lists.01.org]=20 > > > > > > > On Behalf Of Marvin H?user > > > > > > > Sent: Saturday, May 5, 2018 10:25 PM > > > > > > > To: edk2-devel@lists.01.org > > > > > > > Cc: Kinney, Michael D ; Gao,=20 > > > > > > > Liming > > > > > > > Subject: [edk2] [PATCH] MdePkg/BaseLib: Globally include=20 > > > > > > > ChkStk.c > > > > > > > > > > > > > > Initially added for GCC build support, this patch includes = > > > > > > > the function for all compilers and all architectures. This = > > > > > > > is done as huge variables on the stack may cause the=20 > > > > > > > generation of calls to this intrinsic function for=20 > > > > > > > Microsoft compilers, even for the IA32 architecture, too. > > > > > > > > > > > > > > Contributed-under: TianoCore Contribution Agreement 1.1 > > > > > > > Signed-off-by: Marvin Haeuser > > > > > > > --- > > > > > > > MdePkg/Library/BaseLib/{ChkStkGcc.c =3D> ChkStk.c} | 48 > > > > > > > ++++++++++------ > > > > > > ---- > > > > > > > MdePkg/Library/BaseLib/BaseLib.inf | 4 +- > > > > > > > 2 files changed, 26 insertions(+), 26 deletions(-) > > > > > > > > > > > > > > diff --git a/MdePkg/Library/BaseLib/ChkStkGcc.c > > > > > > > b/MdePkg/Library/BaseLib/ChkStk.c similarity index 74%=20 > > > > > > > rename from MdePkg/Library/BaseLib/ChkStkGcc.c > > > > > > > rename to MdePkg/Library/BaseLib/ChkStk.c index > > > > > > > ecba3853b159..59e6d73f9331 100644 > > > > > > > --- a/MdePkg/Library/BaseLib/ChkStkGcc.c > > > > > > > +++ b/MdePkg/Library/BaseLib/ChkStk.c > > > > > > > @@ -1,24 +1,24 @@ > > > > > > > -/** @file > > > > > > > - Provides hack function for passng GCC build. > > > > > > > - > > > > > > > - Copyright (c) 2006 - 2008, Intel Corporation. All=20 > > > > > > > rights reserved.
> > > > > > > - This program and the accompanying materials > > > > > > > - are licensed and made available under the terms and=20 > > > > > > > conditions of the BSD License > > > > > > > - which accompanies this distribution. The full text of=20 > > > > > > > the license may be found at > > > > > > > - http://opensource.org/licenses/bsd-license.php. > > > > > > > - > > > > > > > - THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN > > "AS > > > > IS" > > > > > > > BASIS, > > > > > > > - WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, > > EITHER > > > > > > EXPRESS OR IMPLIED. > > > > > > > - > > > > > > > -**/ > > > > > > > - > > > > > > > -#include "BaseLibInternals.h" > > > > > > > - > > > > > > > -/** > > > > > > > - Hack function for passing GCC build. > > > > > > > -**/ > > > > > > > -VOID > > > > > > > -__chkstk() > > > > > > > -{ > > > > > > > -} > > > > > > > - > > > > > > > +/** @file > > > > > > > + Provides hack function for passing build. > > > > > > > + > > > > > > > + Copyright (c) 2006 - 2018, Intel Corporation. All=20 > > > > > > > + rights reserved.
This program and the accompanying=20 > > > > > > > + materials are licensed and made available under the=20 > > > > > > > + terms and conditions of the BSD License which=20 > > > > > > > + accompanies this distribution. The full text of the=20 > > > > > > > + license may be found at > > > > > > > + http://opensource.org/licenses/bsd- > > > > > > license.php. > > > > > > > + > > > > > > > + THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN > > "AS > > > > IS" > > > > > > > + BASIS, WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY > > KIND, > > > > > > EITHER EXPRESS OR IMPLIED. > > > > > > > + > > > > > > > +**/ > > > > > > > + > > > > > > > +#include "BaseLibInternals.h" > > > > > > > + > > > > > > > +/** > > > > > > > + Hack function for passing build. > > > > > > > +**/ > > > > > > > +VOID > > > > > > > +__chkstk() > > > > > > > +{ > > > > > > > +} > > > > > > > + > > > > > > > diff --git a/MdePkg/Library/BaseLib/BaseLib.inf > > > > > > > b/MdePkg/Library/BaseLib/BaseLib.inf > > > > > > > index 5fbbd02a94b6..d23a6db2581a 100644 > > > > > > > --- a/MdePkg/Library/BaseLib/BaseLib.inf > > > > > > > +++ b/MdePkg/Library/BaseLib/BaseLib.inf > > > > > > > @@ -1,7 +1,7 @@ > > > > > > > ## @file > > > > > > > # Base Library implementation. > > > > > > > # > > > > > > > -# Copyright (c) 2007 - 2016, Intel Corporation. All=20 > > > > > > > rights reserved.
> > > > > > > +# Copyright (c) 2007 - 2018, Intel Corporation. All=20 > > > > > > > +rights reserved.
> > > > > > > # Portions copyright (c) 2008 - 2009, Apple Inc. All=20 > > > > > > > rights reserved.
# Portions copyright (c) 2011 - = 2013, ARM Ltd. > > > > > > > All rights reserved.
# @@ -64,6 +64,7 @@ [Sources] > > > > > > > SafeString.c > > > > > > > String.c > > > > > > > FilePaths.c > > > > > > > + ChkStk.c > > > > > > > BaseLibInternals.h > > > > > > > > > > > > > > [Sources.Ia32] > > > > > > > @@ -781,7 +782,6 @@ [Sources.X64] > > > > > > > X64/DisableCache.S | GCC > > > > > > > X64/RdRand.nasm| GCC > > > > > > > X64/RdRand.S | GCC > > > > > > > - ChkStkGcc.c | GCC > > > > > > > > > > > > > > [Sources.IPF] > > > > > > > Ipf/AccessGp.s > > > > > > > -- > > > > > > > 2.17.0.windows.1 > > > > > > > > > > > > > > _______________________________________________ > > > > > > > edk2-devel mailing list > > > > > > > edk2-devel@lists.01.org > > > > > > > https://lists.01.org/mailman/listinfo/edk2-devel > > > > > _______________________________________________ > > > > > edk2-devel mailing list > > > > > edk2-devel@lists.01.org > > > > > https://lists.01.org/mailman/listinfo/edk2-devel >=20 > _______________________________________________ > edk2-devel mailing list > edk2-devel@lists.01.org > https://lists.01.org/mailman/listinfo/edk2-devel _______________________________________________ edk2-devel mailing list edk2-devel@lists.01.org https://lists.01.org/mailman/listinfo/edk2-devel