From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from us-smtp-delivery-1.mimecast.com (us-smtp-delivery-1.mimecast.com [207.211.31.81]) by mx.groups.io with SMTP id smtpd.web12.32841.1599202699113070477 for ; Thu, 03 Sep 2020 23:58:19 -0700 Authentication-Results: mx.groups.io; dkim=missing; spf=pass (domain: redhat.com, ip: 207.211.31.81, mailfrom: lersek@redhat.com) Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-355-7ld6ee5qM0GxgvvdWDtRNQ-1; Fri, 04 Sep 2020 02:58:17 -0400 X-MC-Unique: 7ld6ee5qM0GxgvvdWDtRNQ-1 Received: from smtp.corp.redhat.com (int-mx01.intmail.prod.int.phx2.redhat.com [10.5.11.11]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id A60AC192CC6B; Fri, 4 Sep 2020 06:58:12 +0000 (UTC) Received: from lacos-laptop-7.usersys.redhat.com (ovpn-112-161.ams2.redhat.com [10.36.112.161]) by smtp.corp.redhat.com (Postfix) with ESMTP id 3842B821E3; Fri, 4 Sep 2020 06:58:11 +0000 (UTC) Subject: =?UTF-8?B?UmU6IOWbnuWkjTogW2VkazItZGV2ZWxdIFtQQVRDSCB2Ml0gVWVmaUNwdVBrZy9NcEluaXRMaWI6IEFkZCBjaGVjayBmb3IgQ1IzL0dEVC9JRFQu?= To: Fan Jeff , "devel@edk2.groups.io" , "eric.dong@intel.com" , "Ni, Ray" Cc: "Lou, Yun" References: <20200903151147.1196-1-eric.dong@intel.com> <9c9d8289-4f8e-75d8-2816-750195a54842@redhat.com> From: "Laszlo Ersek" Message-ID: Date: Fri, 4 Sep 2020 08:58:10 +0200 MIME-Version: 1.0 In-Reply-To: X-Scanned-By: MIMEDefang 2.79 on 10.5.11.11 X-Mimecast-Spam-Score: 0.002 X-Mimecast-Originator: redhat.com Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Content-Language: en-US On 09/04/20 04:18, Fan Jeff wrote: > Laszlo & Eric, > > Introducing one new PCD to handle such rare test case is too heavy. I think We could do validating CR3/GDT/IDT space < 4GB address always in MpInitLib. I disagree. What the UEFI application does (interfere with GDT / IDT / CR3 placement) is invalid. It changes system properties under the feet of platform DXE drivers. UEFI applications are supposed to be written against public protocols and services in the UEFI spec (and maybe in the PI spec). If this application is a test application that purposely massages low-level system properties, that's fine; but then, if we change core edk2 components to be somewhat compatible with this application, then we should make sure that platforms that do not care about this specific use case do not suffer a performance hit or a code complexity hit. What I could accept, under your proposal, is the following: add three ASSERT()s to FillExchangeInfoData(), where we fetch the IDTR / GDTR / CR3 anyway. This would be fine because it only expresses existing assumptions / requirements. However, my understanding is that this would not solve Eric's problem. The system would hang -- in DEBUG / NOOPT builds -- or crash -- in a RELEASE build -- just the same as before. Now, *if* FillExchangeInfoData() is currently *wrong* to have these 32-bit assumptions, because edk2 modules themselves can break those assumptions (without the custom UEFI application), then we have a more serious problem. But for that problem, just "checking and rejecting" is not a sufficient solution, regardless of how and where we check and reject. Thanks Laszlo > > Jeff > > 发件人: Dong, Eric > 发送时间: 2020年9月4日 10:01 > 收件人: Ni, Ray; devel@edk2.groups.io; lersek@redhat.com > 抄送: Lou, Yun > 主题: Re: [edk2-devel] [PATCH v2] UefiCpuPkg/MpInitLib: Add check for CR3/GDT/IDT. > > I guess Laszlo think this check is not always needed, just used for this special shell application case. He wants to use default FALSE to always ignore this check and make code logic clear. > > Thanks, > Eric > > From: Ni, Ray > Sent: Friday, September 4, 2020 9:34 AM > To: devel@edk2.groups.io; lersek@redhat.com; Dong, Eric > Cc: Lou, Yun > Subject: Re: [edk2-devel] [PATCH v2] UefiCpuPkg/MpInitLib: Add check for CR3/GDT/IDT. > > Why do we need a new PCD to control such check? Under what circumstance the PCD is false? > We may need to move such check out of MpLib.c. Because when bps runs at 32bit mode, AP doesn’t need to switch to long mode, such check is not needed and also always passes. > > We should not assume PEI runs at 32 bit mode. > > > 发件人: devel@edk2.groups.io > 代表 Laszlo Ersek > > 发送时间: Friday, September 4, 2020 3:55:47 AM > 收件人: Dong, Eric >; devel@edk2.groups.io > > 抄送: Ni, Ray > > 主题: Re: [edk2-devel] [PATCH v2] UefiCpuPkg/MpInitLib: Add check for CR3/GDT/IDT. > > On 09/03/20 21:00, Laszlo Ersek wrote: > >> (10) More importantly, ValidCR3GdtIdtCheck() should not be called in the >> Worker functions for StartupAllAPs, StartupThisAP, SwitchBSP, and >> EnableDisableAP, in "UefiCpuPkg/Library/MpInitLib/MpLib.c". >> >> Instead, the calls should be made in the DXE instance of the library >> ("UefiCpuPkg/Library/MpInitLib/DxeMpLib.c"), at the very top of the >> functions: >> >> - MpInitLibStartupAllAPs >> - MpInitLibStartupThisAP >> - MpInitLibSwitchBSP >> - MpInitLibEnableDisableAP >> >> Here's why: >> >> (a) The symptom does not affect the PEI phase -- by the time the UEFI >> application is executed, the PEI phase has ended; there's no need to >> modify the PEI instance of the library. >> >> (b) The currently proposed failure exits are too late. For example, in >> the SwitchBSPWorker() function, by the time we exit, we have called >> DisableApicTimerInterrupt(), SaveAndDisableInterrupts(), and >> DisableLvtInterrupts() -- and the error path does not restore the >> original environment. >> >> (c) According to the PI spec (v1.7), the StartupAllAPs(), >> StartupThisAP(), SwitchBSP(), EnableDisableAP() member functions of >> EFI_MP_SERVICES_PROTOCOL may only be called on the (current) BSP. >> Because of this, it is OK to call ValidCR3GdtIdtCheck() as the very >> first action in the above-listed DxeMpLib functions. >> >> (Assuming the protocol members are called from an AP, and consequently >> we check CR3 / GDTR / IDTR on the AP (and not on the BSP), that's the >> *caller's* fault, per spec!) > > This means we can move the ValidCr3GdtIdtCheck() function to > "DxeMpLib.c", and it is not necessary to modify "MpLib.h". > > Thanks > Laszlo > > >