From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received-SPF: Pass (sender SPF authorized) identity=mailfrom; client-ip=134.134.136.126; helo=mga18.intel.com; envelope-from=jordan.l.justen@intel.com; receiver=edk2-devel@lists.01.org Received: from mga18.intel.com (mga18.intel.com [134.134.136.126]) (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 CE7252194D3AE for ; Tue, 19 Feb 2019 11:27:05 -0800 (PST) X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga001.jf.intel.com ([10.7.209.18]) by orsmga106.jf.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 19 Feb 2019 11:27:04 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.58,388,1544515200"; d="scan'208";a="139910516" Received: from mmdandap-mobl1.amr.corp.intel.com (HELO localhost) ([10.254.8.66]) by orsmga001.jf.intel.com with ESMTP; 19 Feb 2019 11:27:03 -0800 MIME-Version: 1.0 In-Reply-To: <734D49CCEBEEF84792F5B80ED585239D5C023766@SHSMSX104.ccr.corp.intel.com> References: <20190218041141.21363-1-jordan.l.justen@intel.com> <734D49CCEBEEF84792F5B80ED585239D5C023766@SHSMSX104.ccr.corp.intel.com> Cc: Liu Yu , Andrew Fish , Anthony Perard , Ard Biesheuvel , "Wu, Hao A" , "Wang, Jian J" , Julien Grall , Laszlo Ersek , "Zeng, Star" , "Gao, Liming" From: Jordan Justen To: "Ni, Ray" , "edk2-devel@lists.01.org" Message-ID: <155060442351.4842.17716845099299751595@jljusten-skl> User-Agent: alot/0.8 Date: Tue, 19 Feb 2019 11:27:03 -0800 Subject: Re: [PATCH 00/10] Fix PEI Core issue during TemporaryRamMigration X-BeenThere: edk2-devel@lists.01.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: EDK II Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 19 Feb 2019 19:27:06 -0000 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable On 2019-02-18 18:46:24, Ni, Ray wrote: > Jordan, > I find many real platforms do not implement the temporary ram migration > PPI and rely on the PeiCore migration logic. > So perhaps TemporaryRamMigration PPI was added to help platform to > destroy the temporary RAM (CAR in x86 platform). I guess since it is in the PI spec, we can't be sure it is only used for this case, or that it might not be used in the future. > But with the introduction of TemporaryRamDone PPI, maybe the > TemporaryRamMigration PPI can be retired. > The logic in PeiCore to call TemporaryRamMigration is just for backward > compatibility. > If that's true, do you still need to enhance PeiCore? I checked the PI 1.4 spec, and I didn't see anything indicating that TemporaryRamSupport PPI is deprecated. Since it is not deprecated, should we ignore the known issue? > For the Emulator case, I already found without TemporaryRamMigration > the platform can still boot. >=20 > Does OVMF hard-depend on TemporaryRamMigration? Or it can reply on > the PeiCore migration logic + TemporaryDone PPI? This is a good question. If it is true that TemporaryRamSupport is rarely used, then maybe it is better to have the sample platforms use the more commonly used path. Personally, I think we should still address the issue with TemporaryRamSupport, and leave the question of whether to test TemporaryRamSupport code paths in the sample platforms as a separate task. At the least, I think we should still continue to use TemporaryRamDone to reset the temp-ram contents to help make sure nothing accidentally depends on a temp-ram pointer. Unfortunately, this would mean that the TemporaryRamSupport path is not really being tested, but it might be the better choise if TemporaryRamSupport is never used in real platforms. -Jordan >=20 > > -----Original Message----- > > From: Justen, Jordan L > > Sent: Monday, February 18, 2019 12:12 PM > > To: edk2-devel@lists.01.org > > Cc: Justen, Jordan L ; Liu Yu > > ; Andrew Fish ; Anthony > > Perard ; Ard Biesheuvel > > ; Wu, Hao A ; Wang, Jian > > J ; Julien Grall ; Lasz= lo Ersek > > ; Ni, Ray ; Zeng, Star > > > > Subject: [PATCH 00/10] Fix PEI Core issue during TemporaryRamMigration > >=20 > > https://github.com/jljusten/edk2.git temp-ram-support > >=20 > > This series fixes an issue that, while rare, is possible based on the w= ay the > > TemporaryRamSupport PPI is defined along with how it is used by the PEI > > Core. > >=20 > > Liu Yu reported a boot issue with EmulatorPkg, which I believe was caus= ed by > > this issue. > >=20 > > The details of the issue are described in the commit message of the > > "MdeModePkg/Core/Pei: Add code path to allow assembly temp-ram > > migration" patch. > >=20 > > Along with this, I added a few Temporary RAM patches for EmulatorPkg and > > OvmfPkg. > >=20 > > Cc: Liu Yu > > Cc: Andrew Fish > > Cc: Anthony Perard > > Cc: Ard Biesheuvel > > Cc: Hao Wu > > Cc: Jian J Wang > > Cc: Julien Grall > > Cc: Laszlo Ersek > > Cc: Ray Ni > > Cc: Star Zeng > >=20 > > Jordan Justen (10): > > EmulatorPkg/build.sh: Fix missing usage of -b BUILDTARGET parameter > > EmulatorPkg/Unix/Host: Use PcdInitValueInTempStack to init temp-ram > > EmulatorPkg/Sec: Replace assembly temp-ram support with C code > > EmulatorPkg/Sec: Disable optimizations for TemporaryRamMigration > > function > > OvmfPkg/Sec: Swap TemporaryRam Stack and Heap locations > > OvmfPkg/Sec: Disable optimizations for TemporaryRamMigration > > MdeModePkg/Core/Pei: Add code path to allow assembly temp-ram > > migration > > MdeModulePkg/Core/Pei: Use assembly for X64 TemporaryRamMigration > > MdeModulePkg/Core/Pei: Use assembly for IA32 TemporaryRamMigration > > OvmfPkg/Sec: Fill Temp Ram after TemporaryRamMigration > >=20 > > EmulatorPkg/Sec/Ia32/SwitchRam.S | 95 ------------------- > > EmulatorPkg/Sec/Ia32/SwitchRam.asm | 94 ------------------ > > EmulatorPkg/Sec/Ia32/TempRam.c | 65 ------------- > > EmulatorPkg/Sec/Sec.c | 76 ++++++++++++++- > > EmulatorPkg/Sec/Sec.inf | 13 +-- > > EmulatorPkg/Sec/X64/SwitchRam.S | 72 -------------- > > EmulatorPkg/Sec/X64/SwitchRam.asm | 76 --------------- > > EmulatorPkg/Unix/Host/Host.c | 2 +- > > EmulatorPkg/Unix/Host/Host.inf | 1 + > > EmulatorPkg/build.sh | 10 +- > > MdeModulePkg/Core/Pei/Dispatcher/Dispatcher.c | 59 ++++++++---- > > .../Dispatcher/Ia32/TemporaryRamMigration.S | 72 ++++++++++++++ > > .../Ia32/TemporaryRamMigration.nasm | 78 +++++++++++++++ > > .../Pei/Dispatcher/TemporaryRamMigration.c | 52 ++++++++++ > > .../Dispatcher/X64/TemporaryRamMigration.S | 69 ++++++++++++++ > > .../Dispatcher/X64/TemporaryRamMigration.nasm | 75 +++++++++++++++ > > MdeModulePkg/Core/Pei/PeiMain.h | 52 ++++++++++ > > MdeModulePkg/Core/Pei/PeiMain.inf | 14 +++ > > OvmfPkg/Sec/Ia32/SecEntry.nasm | 2 +- > > OvmfPkg/Sec/SecMain.c | 59 ++++++++---- > > OvmfPkg/Sec/X64/SecEntry.nasm | 2 +- > > 21 files changed, 577 insertions(+), 461 deletions(-) delete mode 100= 644 > > EmulatorPkg/Sec/Ia32/SwitchRam.S delete mode 100644 > > EmulatorPkg/Sec/Ia32/SwitchRam.asm > > delete mode 100644 EmulatorPkg/Sec/Ia32/TempRam.c delete mode > > 100644 EmulatorPkg/Sec/X64/SwitchRam.S delete mode 100644 > > EmulatorPkg/Sec/X64/SwitchRam.asm create mode 100644 > > MdeModulePkg/Core/Pei/Dispatcher/Ia32/TemporaryRamMigration.S > > create mode 100644 > > MdeModulePkg/Core/Pei/Dispatcher/Ia32/TemporaryRamMigration.nasm > > create mode 100644 > > MdeModulePkg/Core/Pei/Dispatcher/TemporaryRamMigration.c > > create mode 100644 > > MdeModulePkg/Core/Pei/Dispatcher/X64/TemporaryRamMigration.S > > create mode 100644 > > MdeModulePkg/Core/Pei/Dispatcher/X64/TemporaryRamMigration.nasm > >=20 > > -- > > 2.20.0.rc1 >=20