Received: by 2002:ac0:a581:0:0:0:0:0 with SMTP id m1-v6csp1786545imm; Thu, 21 Jun 2018 02:18:53 -0700 (PDT) X-Google-Smtp-Source: ADUXVKK9VUJF5IxQEGgf/i2r2nYjRWTY8pksTdt5meVrOZJF3p5/z5eL0rPD1cfH3Yn8st9CCjGu X-Received: by 2002:a62:3994:: with SMTP id u20-v6mr26744166pfj.95.1529572733019; Thu, 21 Jun 2018 02:18:53 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1529572732; cv=none; d=google.com; s=arc-20160816; b=tCRxNW7cYwKNnape5t5yMSnMbVaYv0Hn0JU+un+twFSFjcaBZ3SFNGG7fwiXu+w4Xk J+C4/IWCQ+dE9mJiZ2FuoMOJmMSH30Du7J80THu2Oa5v8vpH08hoVMe2rAG8Mvwaj+bI QtWqopEmoVBe+Tz4m7bdlHm43lWKoMu/eIimXUcss5R0eb3yHLjwQd7Txu+kGlvIHxLR EB5iHjYhpoIbya0Qh3TXD9W8GMVvHsPWyTEj7EHP5bz66eVTPD+K72OvljYzF/2kO+zJ tDvcsbIML/lVFWKJFMc2paFbNZGLuUftWRRUQvp2w+4tmAAfcsDrhKASp0vvgxK0okAS 8FWw== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:sender:thread-index:content-language :content-transfer-encoding:mime-version:message-id:date:subject :in-reply-to:references:cc:to:from:dmarc-filter:dkim-signature :dkim-signature:arc-authentication-results; bh=foi1eXU6MY8SLh3I6o+2ScSrYxk5HR4c6GG7DdOUmDc=; b=Umy7s8h0IFgQDjKHWUV9D/lXsZUK4oDrA9CjuAIUvhKi1P8CEFkiqkM45F+OQUYJ47 1jfOfcZwrBQqF6gDB7wG1CbWLERP2SNTg1dc1W/oFQq4E49WHGYtgOYkneB24Ef5j/u+ DAWxj9soq46hZbcLQ5zK8SLXE+BHWdsyd8fTB2n7FlQprwG2EEJV/bL9haOEJ3hNUZPY GfAHc90d4xjxGh8uwyG+eo1CSXWfEQlNi7kWcqvqSVrVdXJtSL7T5L5JGxHA8qe/+bXj q1bGorgiIFX5beLLCJdyW2dlLoCGWt9ENJKs66hqoQeDK6eP8o/0lpcaqUoljotEsKid APWQ== ARC-Authentication-Results: i=1; mx.google.com; dkim=pass header.i=@codeaurora.org header.s=default header.b=cHRw81+b; dkim=pass header.i=@codeaurora.org header.s=default header.b=Qzexztfz; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org Return-Path: Received: from vger.kernel.org (vger.kernel.org. [209.132.180.67]) by mx.google.com with ESMTP id 97-v6si4690607pld.345.2018.06.21.02.18.39; Thu, 21 Jun 2018 02:18:52 -0700 (PDT) Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67; Authentication-Results: mx.google.com; dkim=pass header.i=@codeaurora.org header.s=default header.b=cHRw81+b; dkim=pass header.i=@codeaurora.org header.s=default header.b=Qzexztfz; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932874AbeFUJRu (ORCPT + 99 others); Thu, 21 Jun 2018 05:17:50 -0400 Received: from smtp.codeaurora.org ([198.145.29.96]:33344 "EHLO smtp.codeaurora.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932534AbeFUJRp (ORCPT ); Thu, 21 Jun 2018 05:17:45 -0400 Received: by smtp.codeaurora.org (Postfix, from userid 1000) id A69226060A; Thu, 21 Jun 2018 09:17:44 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=codeaurora.org; s=default; t=1529572664; bh=MQY4LXF2/6zKvzdzfpuQm+uPaMloU2dhFrsf7Ip96GE=; h=From:To:Cc:References:In-Reply-To:Subject:Date:From; b=cHRw81+bvdkpqnqCX39wSmXvTSOpvxKFkt+C2lG+MnNSehxpNQgEK1Gk8oOuIZtjP UsRFsy5wjogax0ouU4d8hvYLGprPUosV0UBf+VH2QbG3D/kw1E7leTZl/7+6LNb81c nF2wAip8EzOtUeTNJ1GU31jb/BOHeVU6HLlVFOrU= X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on pdx-caf-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-0.6 required=2.0 tests=ALL_TRUSTED,BAYES_00, DKIM_SIGNED,TVD_RCVD_SINGLE,T_DKIM_INVALID autolearn=no autolearn_force=no version=3.4.0 Received: from SAYALIL (blr-c-bdr-fw-01_globalnat_allzones-outside.qualcomm.com [103.229.19.19]) (using TLSv1 with cipher ECDHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) (Authenticated sender: sayalil@smtp.codeaurora.org) by smtp.codeaurora.org (Postfix) with ESMTPSA id 934DA6021A; Thu, 21 Jun 2018 09:17:39 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=codeaurora.org; s=default; t=1529572663; bh=MQY4LXF2/6zKvzdzfpuQm+uPaMloU2dhFrsf7Ip96GE=; h=From:To:Cc:References:In-Reply-To:Subject:Date:From; b=QzexztfzkIVSAXdmlqUPT7sPPRq2Upz1yf8mP9b4JXq4UZWAGRXh/T9/aTP92zziJ S7GrXQVVhDIZPomfkSalfCl4l152XXXf3fAJa3GIIBSdyWOWT8+0sj/zq1saxxsORT cnwj9Nz3W12HvuV5r4ihTRqXUDXZkXgf+IK8hUW8= DMARC-Filter: OpenDMARC Filter v1.3.2 smtp.codeaurora.org 934DA6021A Authentication-Results: pdx-caf-mail.web.codeaurora.org; dmarc=none (p=none dis=none) header.from=codeaurora.org Authentication-Results: pdx-caf-mail.web.codeaurora.org; spf=none smtp.mailfrom=sayalil@codeaurora.org From: "sayali" To: "'Kyuho Choi'" Cc: , , , , , , , , , , , "'open list'" References: <1528981432-23065-1-git-send-email-sayalil@codeaurora.org> <1528981432-23065-3-git-send-email-sayalil@codeaurora.org> In-Reply-To: Subject: RE: [PATCH V3 2/3] scsi: ufs: Add ufs provisioning support Date: Thu, 21 Jun 2018 14:47:36 +0530 Message-ID: <002c01d40940$b5a91120$20fb3360$@codeaurora.org> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable X-Mailer: Microsoft Outlook 16.0 Content-Language: en-us Thread-Index: AQHqrwfiTmcauPZKO5758rWQ5wPzzwH9kfIcAm+N5IukGbmhgA== Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Kyuho, Comment inline. Thanks, Sayali -----Original Message----- From: Kyuho Choi [mailto:chlrbgh0@gmail.com]=20 Sent: Friday, June 15, 2018 8:28 AM To: Sayali Lokhande Cc: subhashj@codeaurora.org; cang@codeaurora.org; = vivek.gautam@codeaurora.org; rnayak@codeaurora.org; = vinholikatti@gmail.com; jejb@linux.vnet.ibm.com; = martin.petersen@oracle.com; asutoshd@codeaurora.org; = evgreen@chromium.org; riteshh@codeaurora.org; = linux-scsi@vger.kernel.org; open list Subject: Re: [PATCH V3 2/3] scsi: ufs: Add ufs provisioning support Hi, On 6/14/18, Sayali Lokhande wrote: > A new api ufshcd_do_config_device() is added in driver to support UFS=20 > provisioning at runtime. Configfs support is added to trigger=20 > provisioning. > Device and Unit configurable parameters are parsed from vendor=20 > specific provisioning data or file and passed via configfs node at=20 > runtime to provision ufs device. > > Signed-off-by: Sayali Lokhande > --- > drivers/scsi/ufs/ufs.h | 28 ++++++++ > drivers/scsi/ufs/ufshcd.c | 180 > ++++++++++++++++++++++++++++++++++++++++++++++ > drivers/scsi/ufs/ufshcd.h | 2 + > 3 files changed, 210 insertions(+) > > diff --git a/drivers/scsi/ufs/ufs.h b/drivers/scsi/ufs/ufs.h index=20 > e15deb0..1f99904 100644 > --- a/drivers/scsi/ufs/ufs.h > +++ b/drivers/scsi/ufs/ufs.h > @@ -333,6 +333,7 @@ enum { > UFSHCD_AMP =3D 3, > }; > > +#define UFS_BLOCK_SIZE 4096 > #define POWER_DESC_MAX_SIZE 0x62 > #define POWER_DESC_MAX_ACTV_ICC_LVLS 16 > > @@ -425,6 +426,33 @@ enum { > MASK_TM_SERVICE_RESP =3D 0xFF, > }; > > +struct ufs_unit_desc { > + u8 bLUEnable; /* 1 for enabled LU */ > + u8 bBootLunID; /* 0 for using this LU for boot */ > + u8 bLUWriteProtect; /* 1 =3D power on WP, 2 =3D permanent = WP */ > + u8 bMemoryType; /* 0 for enhanced memory type */ > + u32 dNumAllocUnits; /* Number of alloc unit for this LU = */ > + u8 bDataReliability; /* 0 for reliable write support */ > + u8 bLogicalBlockSize; /* See section 13.2.3 of UFS standard = */ > + u8 bProvisioningType; /* 0 for thin provisioning */ > + u16 wContextCapabilities; /* refer Unit Descriptor Description = */ > +}; > + > +struct ufs_config_descr { > + u8 bNumberLU; /* Total number of active LUs */ > + u8 bBootEnable; /* enabling device for partial init = */ > + u8 bDescrAccessEn; /* desc access during partial init */ > + u8 bInitPowerMode; /* Initial device power mode */ > + u8 bHighPriorityLUN; /* LUN of the high priority LU */ > + u8 bSecureRemovalType; /* Erase config for data removal */ > + u8 bInitActiveICCLevel; /* ICC level after reset */ > + u16 wPeriodicRTCUpdate; /* 0 to set a priodic RTC update rate = */ > + u32 bConfigDescrLock; /* 1 to lock Configation Descriptor = */ > + u32 qVendorConfigCode; /* Vendor specific configuration code = */ > + struct ufs_unit_desc unit[8]; > + u8 lun_to_grow; > +}; > + > /* Task management service response */ enum { > UPIU_TASK_MANAGEMENT_FUNC_COMPL =3D 0x00, > diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c=20 > index 4abc7ae..c0235a4 100644 > --- a/drivers/scsi/ufs/ufshcd.c > +++ b/drivers/scsi/ufs/ufshcd.c > @@ -42,6 +42,7 @@ > #include > #include > #include > +#include > #include "ufshcd.h" > #include "ufs_quirks.h" > #include "unipro.h" > @@ -3063,6 +3064,14 @@ static inline int ufshcd_read_power_desc(struct = > ufs_hba *hba, > return ufshcd_read_desc(hba, QUERY_DESC_IDN_POWER, 0, buf, size); } > > +static inline int ufshcd_read_geometry_desc(struct ufs_hba *hba, > + u8 *buf, > + u32 size) > +{ > + return ufshcd_read_desc(hba, QUERY_DESC_IDN_GEOMETRY, 0, buf, size); = > +} > + > + > static int ufshcd_read_device_desc(struct ufs_hba *hba, u8 *buf, u32=20 > size) { > return ufshcd_read_desc(hba, QUERY_DESC_IDN_DEVICE, 0, buf, size);=20 > @@ -6344,6 +6353,177 @@ static int ufshcd_set_dev_ref_clk(struct=20 > ufs_hba > *hba) > } > > /** > + * ufshcd_do_config_device - API function for UFS provisioning > + * hba: per-adapter instance > + * Returns 0 for success, non-zero in case of failure. > + */ > +int ufshcd_do_config_device(struct ufs_hba *hba) { > + struct ufs_config_descr *cfg =3D &hba->cfgs; > + int buff_len =3D QUERY_DESC_CONFIGURATION_DEF_SIZE; > + u8 desc_buf[QUERY_DESC_CONFIGURATION_DEF_SIZE] =3D {0}; > + int i, ret =3D 0; > + int lun_to_grow =3D -1; > + u64 qTotalRawDeviceCapacity; > + u16 wEnhanced1CapAdjFac, wEnhanced2CapAdjFac; > + u32 dEnhanced1MaxNAllocU, dEnhanced2MaxNAllocU; > + size_t alloc_units, units_to_create =3D 0; > + size_t capacity_to_alloc_factor; > + size_t enhanced1_units =3D 0, enhanced2_units =3D 0; > + size_t conversion_ratio =3D 1; > + u8 *pt; > + u32 blocks_per_alloc_unit =3D 1024; > + int geo_len =3D hba->desc_size.geom_desc; > + u8 geo_buf[hba->desc_size.geom_desc]; > + unsigned int max_partitions =3D 8; > + > + WARN_ON(!hba || !cfg); > + > + ret =3D ufshcd_read_geometry_desc(hba, geo_buf, geo_len); > + if (ret) { > + dev_err(hba->dev, "%s: Failed getting geometry_desc %d\n", > + __func__, ret); > + return ret; > + } > + > + /* > + * Get Geomtric parameters like total configurable memory > + * quantity (Offset 0x04 to 0x0b), Capacity Adjustment > + * Factors (Offset 0x30, 0x31, 0x36, 0x37), Max allocation > + * units (Offset 0x2c to 0x2f, 0x32 to 0x35) used to configure > + * the device logical units. > + */ > + qTotalRawDeviceCapacity =3D get_unaligned_be64(&geo_buf[0x04]); > + wEnhanced1CapAdjFac =3D get_unaligned_be16(&geo_buf[0x30]); > + wEnhanced2CapAdjFac =3D get_unaligned_be16(&geo_buf[0x36]); > + dEnhanced1MaxNAllocU =3D get_unaligned_be32(&geo_buf[0x2c]); > + dEnhanced2MaxNAllocU =3D get_unaligned_be32(&geo_buf[0x32]); > + > + capacity_to_alloc_factor =3D > + (blocks_per_alloc_unit * UFS_BLOCK_SIZE) / 512; > + > + if (qTotalRawDeviceCapacity % capacity_to_alloc_factor !=3D 0) { > + dev_err(hba->dev, > + "%s: Raw capacity(%llu) not multiple of alloc factor(%zu)\n", > + __func__, qTotalRawDeviceCapacity, > + capacity_to_alloc_factor); > + return -EINVAL; > + } > + alloc_units =3D (qTotalRawDeviceCapacity / = capacity_to_alloc_factor); > + units_to_create =3D 0; > + enhanced1_units =3D enhanced2_units =3D 0; > + > + /* > + * Calculate number of allocation units to be assigned to a logical = unit > + * considering the capacity adjustment factor of respective memory = type. > + */ > + for (i =3D 0; i < (max_partitions - 1) && > + units_to_create <=3D alloc_units; i++) { > + if ((cfg->unit[i].dNumAllocUnits % blocks_per_alloc_unit) =3D=3D 0) > + cfg->unit[i].dNumAllocUnits /=3D blocks_per_alloc_unit; > + else > + cfg->unit[i].dNumAllocUnits =3D > + cfg->unit[i].dNumAllocUnits / blocks_per_alloc_unit + 1; > + > + if (cfg->unit[i].bMemoryType =3D=3D 0) > + units_to_create +=3D cfg->unit[i].dNumAllocUnits; > + else if (cfg->unit[i].bMemoryType =3D=3D 3) { > + enhanced1_units +=3D cfg->unit[i].dNumAllocUnits; > + cfg->unit[i].dNumAllocUnits *=3D > + (wEnhanced1CapAdjFac / 0x100); > + units_to_create +=3D cfg->unit[i].dNumAllocUnits; > + } else if (cfg->unit[i].bMemoryType =3D=3D 4) { > + enhanced2_units +=3D cfg->unit[i].dNumAllocUnits; > + cfg->unit[i].dNumAllocUnits *=3D > + (wEnhanced1CapAdjFac / 0x100); > + units_to_create +=3D cfg->unit[i].dNumAllocUnits; > + } else { > + dev_err(hba->dev, "%s: Unsupported memory type %d\n", > + __func__, cfg->unit[i].bMemoryType); > + return -EINVAL; > + } > + } > + if (enhanced1_units > dEnhanced1MaxNAllocU) { > + dev_err(hba->dev, "%s: size %zu exceeds max enhanced1 area size = %u\n", > + __func__, enhanced1_units, dEnhanced1MaxNAllocU); > + return -ERANGE; > + } > + if (enhanced2_units > dEnhanced2MaxNAllocU) { > + dev_err(hba->dev, "%s: size %zu exceeds max enhanced2 area size = %u\n", > + __func__, enhanced2_units, dEnhanced2MaxNAllocU); > + return -ERANGE; > + } > + if (units_to_create > alloc_units) { > + dev_err(hba->dev, "%s: Specified size %zu exceeds device size = %zu\n", > + __func__, units_to_create, alloc_units); > + return -ERANGE; > + } > + lun_to_grow =3D cfg->lun_to_grow; > + if (lun_to_grow !=3D -1) { > + if (cfg->unit[i].bMemoryType =3D=3D 0) > + conversion_ratio =3D 1; > + else if (cfg->unit[i].bMemoryType =3D=3D 3) > + conversion_ratio =3D (wEnhanced1CapAdjFac / 0x100); > + else if (cfg->unit[i].bMemoryType =3D=3D 4) > + conversion_ratio =3D (wEnhanced2CapAdjFac / 0x100); > + > + cfg->unit[lun_to_grow].dNumAllocUnits +=3D > + ((alloc_units - units_to_create) / conversion_ratio); > + dev_dbg(hba->dev, "%s: conversion_ratio %zu for lun %d\n", > + __func__, conversion_ratio, i); > + } > + > + /* Fill in the buffer with configuration data */ > + pt =3D desc_buf; > + *pt++ =3D 0x90; // bLength > + *pt++ =3D 0x01; // bDescriptorType > + *pt++ =3D 0; // Reserved in UFS2.0 and onward > + *pt++ =3D cfg->bBootEnable; > + *pt++ =3D cfg->bDescrAccessEn; > + *pt++ =3D cfg->bInitPowerMode; > + *pt++ =3D cfg->bHighPriorityLUN; > + *pt++ =3D cfg->bSecureRemovalType; > + *pt++ =3D cfg->bInitActiveICCLevel; > + put_unaligned_be16(cfg->wPeriodicRTCUpdate, pt); > + pt =3D pt + 7; // Reserved fields set to 0 > + > + /* Fill in the buffer with per logical unit data */ > + for (i =3D 0; i < UFS_UPIU_MAX_GENERAL_LUN; i++) { > + *pt++ =3D cfg->unit[i].bLUEnable; > + *pt++ =3D cfg->unit[i].bBootLunID; > + *pt++ =3D cfg->unit[i].bLUWriteProtect; > + *pt++ =3D cfg->unit[i].bMemoryType; > + put_unaligned_be32(cfg->unit[i].dNumAllocUnits, pt); > + pt =3D pt + 4; > + *pt++ =3D cfg->unit[i].bDataReliability; > + *pt++ =3D cfg->unit[i].bLogicalBlockSize; > + *pt++ =3D cfg->unit[i].bProvisioningType; > + put_unaligned_be16(cfg->unit[i].wContextCapabilities, pt); > + pt =3D pt + 5; // Reserved fields set to 0 > + } > + > + ret =3D ufshcd_query_descriptor_retry(hba, = UPIU_QUERY_OPCODE_WRITE_DESC, > + QUERY_DESC_IDN_CONFIGURATION, 0, 0, desc_buf, &buff_len); > + > + if (ret) { > + dev_err(hba->dev, "%s: Failed writing descriptor. desc_idn %d,=20 > +opcode %x > ret %d\n", > + __func__, QUERY_DESC_IDN_CONFIGURATION, > + UPIU_QUERY_OPCODE_WRITE_DESC, ret); > + return ret; > + } To support stable operation and prevent race condition, how about adding = some functions during in runtime provisioning?. pm_runtime_get/put_sync and scsi_block/unblock_requests. How you thought?. [Sayali] Agreed. Will check and update in next patch set. > + > + if (cfg->bConfigDescrLock =3D=3D 1) { > + ret =3D ufshcd_query_attr(hba, UPIU_QUERY_OPCODE_WRITE_ATTR, > + QUERY_ATTR_IDN_CONF_DESC_LOCK, 0, 0, &cfg->bConfigDescrLock); > + if (ret) > + dev_err(hba->dev, "%s: Failed writing bConfigDescrLock %d\n", > + __func__, ret); > + } > + > + return ret; > +} > + > +/** > * ufshcd_probe_hba - probe hba to detect device and initialize > * @hba: per-adapter instance > * > diff --git a/drivers/scsi/ufs/ufshcd.h b/drivers/scsi/ufs/ufshcd.h=20 > index b026ad8..816b674 100644 > --- a/drivers/scsi/ufs/ufshcd.h > +++ b/drivers/scsi/ufs/ufshcd.h > @@ -549,6 +549,7 @@ struct ufs_hba { > unsigned int irq; > bool is_irq_enabled; > u32 dev_ref_clk_freq; > + struct ufs_config_descr cfgs; > > /* Interrupt aggregation support is broken */ > #define UFSHCD_QUIRK_BROKEN_INTR_AGGR 0x1 > @@ -866,6 +867,7 @@ int ufshcd_read_string_desc(struct ufs_hba *hba,=20 > int desc_index, > > int ufshcd_hold(struct ufs_hba *hba, bool async); void=20 > ufshcd_release(struct ufs_hba *hba); > +int ufshcd_do_config_device(struct ufs_hba *hba); > > int ufshcd_map_desc_id_to_length(struct ufs_hba *hba, enum desc_idn=20 > desc_id, > int *desc_length); > -- > The Qualcomm Innovation Center, Inc. is a member of the Code Aurora=20 > Forum, a Linux Foundation Collaborative Project > > BR, Kyuho Choi