From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received-SPF: Pass (sender SPF authorized) identity=mailfrom; client-ip=192.55.52.88; helo=mga01.intel.com; envelope-from=ruiyu.ni@intel.com; receiver=edk2-devel@lists.01.org Received: from mga01.intel.com (mga01.intel.com [192.55.52.88]) (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 A519A2035624D for ; Tue, 5 Dec 2017 00:32:37 -0800 (PST) Received: from fmsmga008.fm.intel.com ([10.253.24.58]) by fmsmga101.fm.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 05 Dec 2017 00:37:08 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.45,363,1508828400"; d="scan'208";a="226307" Received: from ray-dev.ccr.corp.intel.com (HELO [10.239.9.14]) ([10.239.9.14]) by fmsmga008.fm.intel.com with ESMTP; 05 Dec 2017 00:37:06 -0800 To: "Zeng, Star" , Ard Biesheuvel , "edk2-devel@lists.01.org" Cc: "Wu, Hao A" , "Kinney, Michael D" , "Tian, Feng" , "leif.lindholm@linaro.org" References: <20171130101132.18317-1-ard.biesheuvel@linaro.org> <20171130101132.18317-2-ard.biesheuvel@linaro.org> <1c1b4095-3d21-552e-c54e-251a34191740@Intel.com> <0C09AFA07DD0434D9E2A0C6AEB0483103B9BFA78@shsmsx102.ccr.corp.intel.com> From: "Ni, Ruiyu" Message-ID: <9c301a27-024d-3ba8-65f5-260303d2c9ac@Intel.com> Date: Tue, 5 Dec 2017 16:37:06 +0800 User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:52.0) Gecko/20100101 Thunderbird/52.5.0 MIME-Version: 1.0 In-Reply-To: <0C09AFA07DD0434D9E2A0C6AEB0483103B9BFA78@shsmsx102.ccr.corp.intel.com> Subject: Re: [PATCH v2 1/2] MdeModulePkg: introduce SD/MMC override protocol X-BeenThere: edk2-devel@lists.01.org X-Mailman-Version: 2.1.22 Precedence: list List-Id: EDK II Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 05 Dec 2017 08:32:37 -0000 Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: 7bit On 12/5/2017 3:20 PM, Zeng, Star wrote: > If making this protocol a platform level singleton instance, is it so hard to define the interfaces and parameters since different controllers may need different hook points and parameters? > > Thanks, > Star > -----Original Message----- > From: Ni, Ruiyu > Sent: Tuesday, December 5, 2017 3:09 PM > To: Ard Biesheuvel ; edk2-devel@lists.01.org > Cc: "hao.a.wu@intel.com; Kinney.d.michael"@intel.com; Tian, Feng ; Zeng, Star ; leif.lindholm@linaro.org > Subject: Re: [edk2] [PATCH v2 1/2] MdeModulePkg: introduce SD/MMC override protocol > > On 11/30/2017 6:11 PM, Ard Biesheuvel wrote: >> Many ARM based SoCs have integrated SDHCI controllers, and often, >> these implementations deviate in subtle ways from the pertinent >> specifications. On the one hand, these deviations are quite easy to >> work around, but on the other hand, having a collection of SoC >> specific workarounds in the generic driver stack is undesirable. >> >> So let's introduce an optional SD/MMC override protocol that we can >> invoke at the appropriate moments in the device initialization. >> That way, the workaround itself remains platform specific, but we can >> still use the generic driver stack on such platforms. >> >> Contributed-under: TianoCore Contribution Agreement 1.1 >> Signed-off-by: Ard Biesheuvel >> --- >> MdeModulePkg/Include/Protocol/SdMmcOverride.h | 103 ++++++++++++++++++++ >> MdeModulePkg/MdeModulePkg.dec | 3 + >> 2 files changed, 106 insertions(+) >> >> diff --git a/MdeModulePkg/Include/Protocol/SdMmcOverride.h >> b/MdeModulePkg/Include/Protocol/SdMmcOverride.h >> new file mode 100644 >> index 000000000000..5a5c393896f4 >> --- /dev/null >> +++ b/MdeModulePkg/Include/Protocol/SdMmcOverride.h >> @@ -0,0 +1,103 @@ >> +/** @file >> + Protocol to describe overrides required to support non-standard >> +SDHCI >> + implementations >> + >> + Copyright (c) 2017, Linaro, Ltd. All rights reserved.
>> + >> + This program and the accompanying materials are licensed and made >> + available under the terms and conditions of the BSD License which >> + accompanies this distribution. The full text of 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. >> + >> +**/ >> + >> +#ifndef __SD_MMC_OVERRIDE_H__ >> +#define __SD_MMC_OVERRIDE_H__ >> + >> +#include >> + >> +#define EDKII_SD_MMC_OVERRIDE_PROTOCOL_GUID \ >> + { 0xeaf9e3c1, 0xc9cd, 0x46db, { 0xa5, 0xe5, 0x5a, 0x12, 0x4c, 0x83, >> +0x23, 0x23 } } >> + >> +#define EDKII_SD_MMC_OVERRIDE_PROTOCOL_VERSION 0x1 >> + >> +typedef struct _SD_MMC_OVERRIDE SD_MMC_OVERRIDE; >> + >> +typedef enum { >> + SD_MMC_OVERRIDE_RESET_PRE_HOOK, >> + SD_MMC_OVERRIDE_RESET_POST_HOOK, >> + SD_MMC_OVERRIDE_INIT_HOST_PRE_HOOK, >> + SD_MMC_OVERRIDE_INIT_HOST_POST_HOOK, >> +} SD_MMC_OVERRIDE_HOOK; >> + >> +/** >> + >> + Override function for SDHCI capability bits >> + >> + @param[in] PassThru A pointer to the >> + EFI_SD_MMC_PASS_THRU_PROTOCOL instance. >> + @param[in] ControllerHandle The EFI_HANDLE of the controller. >> + @param[in] Slot The 0 based slot index. >> + @param[in,out] SdMmcHcSlotCapability The SDHCI capability structure. >> + >> + @retval EFI_SUCCESS The override function completed successfully. >> + @retval EFI_NOT_FOUND The specified controller or slot does not exist. >> + @retval EFI_INVALID_PARAMETER SdMmcHcSlotCapability is NULL >> + >> +**/ >> +typedef >> +EFI_STATUS >> +(EFIAPI * SD_MMC_OVERRIDE_CAPABILITY) ( >> + IN EFI_SD_MMC_PASS_THRU_PROTOCOL *PassThru, >> + IN EFI_HANDLE ControllerHandle, >> + IN UINT8 Slot, >> + IN OUT UINT64 *SdMmcHcSlotCapability >> + ); >> + >> +/** >> + >> + Override function for SDHCI controller operations >> + >> + @param[in] PassThru A pointer to the >> + EFI_SD_MMC_PASS_THRU_PROTOCOL instance. >> + @param[in] ControllerHandle The EFI_HANDLE of the controller. >> + @param[in] Slot The 0 based slot index. >> + @param[in,out] HookType The type of operation and whether the >> + hook is invoked right before (pre) or >> + right after (post) >> + >> + @retval EFI_SUCCESS The override function completed successfully. >> + @retval EFI_NOT_FOUND The specified controller or slot does not exist. >> + @retval EFI_INVALID_PARAMETER HookType is invalid >> + >> +**/ >> +typedef >> +EFI_STATUS >> +(EFIAPI * SD_MMC_OVERRIDE_INVOKE_HOOK) ( >> + IN EFI_SD_MMC_PASS_THRU_PROTOCOL *PassThru, >> + IN EFI_HANDLE ControllerHandle, >> + IN UINT8 Slot, >> + IN SD_MMC_OVERRIDE_HOOK HookType >> + ); >> + >> +struct _SD_MMC_OVERRIDE { >> + // >> + // Protocol version of this implementation >> + // >> + UINTN Version; >> + // >> + // Callback to override SD/MMC host controller capability bits >> + // >> + SD_MMC_OVERRIDE_CAPABILITY OverrideCapability; >> + // >> + // Callback to invoke SD/MMC override hooks >> + // >> + SD_MMC_OVERRIDE_INVOKE_HOOK InvokeHook; >> +}; >> + >> +extern EFI_GUID gEdkiiSdMmcOverrideProtocolGuid; >> + >> +#endif >> diff --git a/MdeModulePkg/MdeModulePkg.dec >> b/MdeModulePkg/MdeModulePkg.dec index 856d67aceb21..64ceea029f94 >> 100644 >> --- a/MdeModulePkg/MdeModulePkg.dec >> +++ b/MdeModulePkg/MdeModulePkg.dec >> @@ -562,6 +562,9 @@ [Protocols] >> ## Include/Protocol/SmmMemoryAttribute.h >> gEdkiiSmmMemoryAttributeProtocolGuid = { 0x69b792ea, 0x39ce, >> 0x402d, { 0xa2, 0xa6, 0xf7, 0x21, 0xde, 0x35, 0x1d, 0xfe } } >> >> + ## Include/Protocol/SdMmcOverride.h >> + gEdkiiSdMmcOverrideProtocolGuid = { 0xeaf9e3c1, 0xc9cd, 0x46db, { >> + 0xa5, 0xe5, 0x5a, 0x12, 0x4c, 0x83, 0x23, 0x23 } } >> + >> # >> # [Error.gEfiMdeModulePkgTokenSpaceGuid] >> # 0x80000001 | Invalid value provided. >> > Based on your patch #2, I think the protocol is supposed to be installed by a platform driver (let's call it SdMmcOverridePlatform) on the controller handle. > > There are two ways to produce this Override protocol: > 1. SdMmcOverridePlatform registers a protocol notification on PciIo and find the SDMMC host controller PCI handle, then install SdMmcOverride on that handle. > 2. SdMMcOverridePlatform is written as a driver model driver. > DriverBindingStart() opens PciIo BY_DRIVER and produces SdMmcOverride. > Then SdMmcPciHc driver opens SdMmcOverride BY_DRIVER and produces SdMmcPassthru. > > But since the SdMmcOverride protocol is optional, that makes #2 not workable. > > But #1 cannot handle the "reconnect -r" case. Because there is no protocol notification on uninstall. > > So I am thinking if it's more proper to make this protocol a platform level singleton instance. I noticed that all APIs exposed by this protocol already carry the Controller Handle. > > -- > Thanks, > Ray > Star, I meant to create a SdMmcPlatform protocol. -- Thanks, Ray