From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-8.2 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_HELO_NONE,SPF_PASS, UNPARSEABLE_RELAY,USER_AGENT_SANE_1 autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 320C5C54E8D for ; Mon, 11 May 2020 08:36:34 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 1BF4E20A8B for ; Mon, 11 May 2020 08:36:34 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729264AbgEKIgd (ORCPT ); Mon, 11 May 2020 04:36:33 -0400 Received: from smtp2207-205.mail.aliyun.com ([121.197.207.205]:57119 "EHLO smtp2207-205.mail.aliyun.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1728574AbgEKIgd (ORCPT ); Mon, 11 May 2020 04:36:33 -0400 X-Alimail-AntiSpam: AC=CONTINUE;BC=0.07436282|-1;CH=green;DM=|CONTINUE|false|;DS=CONTINUE|ham_system_inform|0.0537608-0.000255462-0.945984;FP=0|0|0|0|0|-1|-1|-1;HT=e02c03278;MF=liaoweixiong@allwinnertech.com;NM=1;PH=DS;RN=16;RT=16;SR=0;TI=SMTPD_---.HWnA6YQ_1589186181; Received: from 172.16.10.102(mailfrom:liaoweixiong@allwinnertech.com fp:SMTPD_---.HWnA6YQ_1589186181) by smtp.aliyun-inc.com(10.147.43.230); Mon, 11 May 2020 16:36:21 +0800 Subject: Re: [PATCH v7 10/18] pstore/zone,blk: Add support for pmsg frontend To: Kees Cook Cc: Anton Vorontsov , Colin Cross , Tony Luck , Jonathan Corbet , Miquel Raynal , Richard Weinberger , Vignesh Raghavendra , Petr Mladek , Sergey Senozhatsky , Steven Rostedt , Rob Herring , Pavel Tatashin , linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org, linux-mtd@lists.infradead.org References: <20200510202436.63222-1-keescook@chromium.org> <20200510202436.63222-11-keescook@chromium.org> From: WeiXiong Liao Message-ID: <3b14333e-9f9b-9ac5-e30e-79b42bbd6526@allwinnertech.com> Date: Mon, 11 May 2020 16:36:30 +0800 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.6.1 MIME-Version: 1.0 In-Reply-To: <20200510202436.63222-11-keescook@chromium.org> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org hi Kees Cook, On 2020/5/11 AM 4:24, Kees Cook wrote: > From: WeiXiong Liao > > Add pmsg support to pstore/blk (through pstore/zone). To enable, pmsg_size > must be greater than 0 and a multiple of 4096. > > Signed-off-by: WeiXiong Liao > Link: https://lore.kernel.org/r/1585126506-18635-5-git-send-email-liaoweixiong@allwinnertech.com > Signed-off-by: Kees Cook > --- > fs/pstore/Kconfig | 12 ++ > fs/pstore/blk.c | 9 ++ > fs/pstore/zone.c | 269 ++++++++++++++++++++++++++++++++++-- > include/linux/pstore_zone.h | 2 + > 4 files changed, 282 insertions(+), 10 deletions(-) > > diff --git a/fs/pstore/Kconfig b/fs/pstore/Kconfig > index 92ba73bd0b62..f18cd126d83f 100644 > --- a/fs/pstore/Kconfig > +++ b/fs/pstore/Kconfig > @@ -224,3 +224,15 @@ config PSTORE_BLK_MAX_REASON > > NOTE that, both Kconfig and module parameters can configure > pstore/blk, but module parameters have priority over Kconfig. > + > +config PSTORE_BLK_PMSG_SIZE > + int "Size in Kbytes of pmsg to store" > + depends on PSTORE_BLK > + depends on PSTORE_PMSG > + default 64 > + help > + This just sets size of pmsg (pmsg_size) for pstore/blk. The size is > + in KB and must be a multiple of 4. > + > + NOTE that, both Kconfig and module parameters can configure > + pstore/blk, but module parameters have priority over Kconfig. > diff --git a/fs/pstore/blk.c b/fs/pstore/blk.c > index cec1fa261d1b..516a5f80d627 100644 > --- a/fs/pstore/blk.c > +++ b/fs/pstore/blk.c > @@ -27,6 +27,14 @@ module_param(max_reason, int, 0400); > MODULE_PARM_DESC(max_reason, > "maximum reason for kmsg dump (default 2: Oops and Panic)"); > > +#if IS_ENABLED(CONFIG_PSTORE_PMSG) > +static long pmsg_size = CONFIG_PSTORE_BLK_PMSG_SIZE; > +#else > +static long pmsg_size = -1; > +#endif > +module_param(pmsg_size, long, 0400); > +MODULE_PARM_DESC(pmsg_size, "pmsg size in kbytes"); > + > /* > * blkdev - the block device to use for pstore storage > * > @@ -132,6 +140,7 @@ static int psblk_register_do(struct pstore_device_info *dev) > } > > verify_size(kmsg_size, 4096, dev->flags & PSTORE_FLAGS_DMESG); > + verify_size(pmsg_size, 4096, dev->flags & PSTORE_FLAGS_PMSG); > #undef verify_size > > pstore_zone_info->total_size = dev->total_size; > diff --git a/fs/pstore/zone.c b/fs/pstore/zone.c > index a388c299ebb1..03dc87791c24 100644 > --- a/fs/pstore/zone.c > +++ b/fs/pstore/zone.c > @@ -27,12 +27,14 @@ > * > * @sig: signature to indicate header (PSZ_SIG xor PSZONE-type value) > * @datalen: length of data in @data > + * @start: offset into @data where the beginning of the stored bytes begin > * @data: zone data. > */ > struct psz_buffer { > #define PSZ_SIG (0x43474244) /* DBGC */ > uint32_t sig; > atomic_t datalen; > + atomic_t start; > uint8_t data[]; > }; > > @@ -88,9 +90,11 @@ struct pstore_zone { > * struct psz_context - all about running state of pstore/zone > * > * @opszs: oops/panic storage zones > + * @ppsz: pmsg storage zone > * @kmsg_max_cnt: max count of @opszs > * @kmsg_read_cnt: counter to read oops zone > * @kmsg_write_cnt: counter to write > + * @pmsg_read_cnt: counter to read pmsg zone > * @oops_counter: counter to oops > * @panic_counter: counter to panic > * @recovered: whether finish recovering data from storage > @@ -101,9 +105,11 @@ struct pstore_zone { > */ > struct psz_context { > struct pstore_zone **opszs; > + struct pstore_zone *ppsz; > unsigned int kmsg_max_cnt; > unsigned int kmsg_read_cnt; > unsigned int kmsg_write_cnt; > + unsigned int pmsg_read_cnt; > /* > * the counter should be recovered when recover. > * It records the oops/panic times after burning rather than booting. > @@ -143,15 +149,20 @@ static inline int buffer_datalen(struct pstore_zone *zone) > return atomic_read(&zone->buffer->datalen); > } > > +static inline int buffer_start(struct pstore_zone *zone) > +{ > + return atomic_read(&zone->buffer->start); > +} > + > static inline bool is_on_panic(void) > { > return atomic_read(&pstore_zone_cxt.on_panic); > } > > -static ssize_t psz_zone_read(struct pstore_zone *zone, char *buf, > +static ssize_t psz_zone_read_buffer(struct pstore_zone *zone, char *buf, > size_t len, unsigned long off) > { > - if (!buf || !zone->buffer) > + if (!buf || !zone || !zone->buffer) > return -EINVAL; > if (off > zone->buffer_size) > return -EINVAL; > @@ -160,6 +171,18 @@ static ssize_t psz_zone_read(struct pstore_zone *zone, char *buf, > return len; > } > > +static int psz_zone_read_oldbuf(struct pstore_zone *zone, char *buf, > + size_t len, unsigned long off) > +{ > + if (!buf || !zone || !zone->oldbuf) > + return -EINVAL; > + if (off > zone->buffer_size) > + return -EINVAL; > + len = min_t(size_t, len, zone->buffer_size - off); > + memcpy(buf, zone->oldbuf->data + off, len); > + return 0; > +} > + > static int psz_zone_write(struct pstore_zone *zone, > enum psz_flush_mode flush_mode, const char *buf, > size_t len, unsigned long off) > @@ -415,6 +438,93 @@ static int psz_kmsg_recover(struct psz_context *cxt) > return ret; > } > > +static int psz_recover_zone(struct psz_context *cxt, struct pstore_zone *zone) > +{ > + struct pstore_zone_info *info = cxt->pstore_zone_info; > + struct psz_buffer *oldbuf, tmpbuf; > + int ret = 0; > + char *buf; > + ssize_t rcnt, len, start, off; > + > + if (!zone || zone->oldbuf) > + return 0; > + > + if (is_on_panic()) { > + /* save data as much as possible */ > + psz_flush_dirty_zone(zone); > + return 0; > + } > + > + if (unlikely(!info->read)) > + return -EINVAL; > + > + len = sizeof(struct psz_buffer); > + rcnt = info->read((char *)&tmpbuf, len, zone->off); > + if (rcnt != len) { > + pr_debug("read zone %s failed\n", zone->name); > + return (int)rcnt < 0 ? (int)rcnt : -EIO; > + } > + > + if (tmpbuf.sig != zone->buffer->sig) { > + pr_debug("no valid data in zone %s\n", zone->name); > + return 0; > + } > + > + if (zone->buffer_size < atomic_read(&tmpbuf.datalen) || > + zone->buffer_size < atomic_read(&tmpbuf.start)) { > + pr_info("found overtop zone: %s: off %lld, size %zu\n", > + zone->name, zone->off, zone->buffer_size); > + /* just keep going */ > + return 0; > + } > + > + if (!atomic_read(&tmpbuf.datalen)) { > + pr_debug("found erased zone: %s: off %lld, size %zu, datalen %d\n", > + zone->name, zone->off, zone->buffer_size, > + atomic_read(&tmpbuf.datalen)); > + return 0; > + } > + > + pr_debug("found nice zone: %s: off %lld, size %zu, datalen %d\n", > + zone->name, zone->off, zone->buffer_size, > + atomic_read(&tmpbuf.datalen)); > + > + len = atomic_read(&tmpbuf.datalen) + sizeof(*oldbuf); > + oldbuf = kzalloc(len, GFP_KERNEL); > + if (!oldbuf) > + return -ENOMEM; > + > + memcpy(oldbuf, &tmpbuf, sizeof(*oldbuf)); > + buf = (char *)oldbuf + sizeof(*oldbuf); > + len = atomic_read(&oldbuf->datalen); > + start = atomic_read(&oldbuf->start); > + off = zone->off + sizeof(*oldbuf); > + > + /* get part of data */ > + rcnt = info->read(buf, len - start, off + start); > + if (rcnt != len - start) { > + pr_err("read zone %s failed\n", zone->name); > + ret = (int)rcnt < 0 ? (int)rcnt : -EIO; > + goto free_oldbuf; > + } > + > + /* get the rest of data */ > + rcnt = info->read(buf + len - start, start, off); > + if (rcnt != start) { > + pr_err("read zone %s failed\n", zone->name); > + ret = (int)rcnt < 0 ? (int)rcnt : -EIO; > + goto free_oldbuf; > + } > + > + zone->oldbuf = oldbuf; > + psz_flush_dirty_zone(zone); > + return 0; > + > +free_oldbuf: > + kfree(oldbuf); > + return ret; > +} > + > /** > * psz_recovery() - recover data from storage > * @cxt: the context of pstore/zone > @@ -434,6 +544,10 @@ static inline int psz_recovery(struct psz_context *cxt) > if (ret) > goto recover_fail; > > + ret = psz_recover_zone(cxt, cxt->ppsz); > + if (ret) > + goto recover_fail; > + > pr_debug("recover end!\n"); > atomic_set(&cxt->recovered, 1); > return 0; > @@ -448,9 +562,17 @@ static int psz_pstore_open(struct pstore_info *psi) > struct psz_context *cxt = psi->data; > > cxt->kmsg_read_cnt = 0; > + cxt->pmsg_read_cnt = 0; > return 0; > } > > +static inline bool psz_old_ok(struct pstore_zone *zone) > +{ > + if (zone && zone->oldbuf && atomic_read(&zone->oldbuf->datalen)) > + return true; > + return false; > +} > + > static inline bool psz_ok(struct pstore_zone *zone) > { > if (zone && zone->buffer && buffer_datalen(zone)) > @@ -475,6 +597,25 @@ static inline int psz_kmsg_erase(struct psz_context *cxt, > return psz_zone_write(zone, FLUSH_META, NULL, 0, 0); > } > > +static inline int psz_record_erase(struct psz_context *cxt, > + struct pstore_zone *zone) > +{ > + if (unlikely(!psz_old_ok(zone))) > + return 0; > + > + kfree(zone->oldbuf); > + zone->oldbuf = NULL; > + /* > + * if there are new data in zone buffer, that means the old data > + * are already invalid. It is no need to flush 0 (erase) to > + * block device. > + */ > + if (!buffer_datalen(zone)) > + return psz_zone_write(zone, FLUSH_META, NULL, 0, 0); > + psz_flush_dirty_zone(zone); > + return 0; > +} > + > static int psz_pstore_erase(struct pstore_record *record) > { > struct psz_context *cxt = record->psi->data; > @@ -484,6 +625,8 @@ static int psz_pstore_erase(struct pstore_record *record) > if (record->id >= cxt->kmsg_max_cnt) > return -EINVAL; > return psz_kmsg_erase(cxt, cxt->opszs[record->id], record); > + case PSTORE_TYPE_PMSG: > + return psz_record_erase(cxt, cxt->ppsz); > default: > return -EINVAL; > } > @@ -504,8 +647,10 @@ static void psz_write_kmsg_hdr(struct pstore_zone *zone, > hdr->reason = record->reason; > if (hdr->reason == KMSG_DUMP_OOPS) > hdr->counter = ++cxt->oops_counter; > - else > + else if (hdr->reason == KMSG_DUMP_PANIC) > hdr->counter = ++cxt->panic_counter; > + else > + hdr->counter = 0; I think This difference should move to first patch to support > } > > static inline int notrace psz_kmsg_write_record(struct psz_context *cxt, > @@ -555,6 +700,53 @@ static int notrace psz_kmsg_write(struct psz_context *cxt, > return 0; > } > > +static int notrace psz_record_write(struct pstore_zone *zone, > + struct pstore_record *record) > +{ > + size_t start, rem; > + int cnt = record->size; > + bool is_full_data = false; > + char *buf = record->buf; > + > + if (!zone || !record) > + return -ENOSPC; > + > + if (atomic_read(&zone->buffer->datalen) >= zone->buffer_size) > + is_full_data = true; > + > + if (unlikely(cnt > zone->buffer_size)) { > + buf += cnt - zone->buffer_size; > + cnt = zone->buffer_size; > + } > + > + start = buffer_start(zone); > + rem = zone->buffer_size - start; > + if (unlikely(rem < cnt)) { > + psz_zone_write(zone, FLUSH_PART, buf, rem, start); > + buf += rem; > + cnt -= rem; > + start = 0; > + is_full_data = true; > + } > + > + atomic_set(&zone->buffer->start, cnt + start); > + psz_zone_write(zone, FLUSH_PART, buf, cnt, start); > + > + /** > + * psz_zone_write will set datalen as start + cnt. > + * It work if actual data length lesser than buffer size. > + * If data length greater than buffer size, pmsg will rewrite to > + * beginning of zone, which make buffer->datalen wrongly. > + * So we should reset datalen as buffer size once actual data length > + * greater than buffer size. > + */ > + if (is_full_data) { > + atomic_set(&zone->buffer->datalen, zone->buffer_size); > + psz_zone_write(zone, FLUSH_META, NULL, 0, 0); > + } > + return 0; > +} > + > static int notrace psz_pstore_write(struct pstore_record *record) > { > struct psz_context *cxt = record->psi->data; > @@ -566,6 +758,8 @@ static int notrace psz_pstore_write(struct pstore_record *record) > switch (record->type) { > case PSTORE_TYPE_DMESG: > return psz_kmsg_write(cxt, record); > + case PSTORE_TYPE_PMSG: > + return psz_record_write(cxt->ppsz, record); > default: > return -EINVAL; > } > @@ -581,6 +775,13 @@ static struct pstore_zone *psz_read_next_zone(struct psz_context *cxt) > return zone; > } > > + if (cxt->pmsg_read_cnt == 0) { > + cxt->pmsg_read_cnt++; > + zone = cxt->ppsz; > + if (psz_old_ok(zone)) > + return zone; > + } > + > return NULL; > } > > @@ -631,7 +832,7 @@ static ssize_t psz_kmsg_read(struct pstore_zone *zone, > return -ENOMEM; > } > > - size = psz_zone_read(zone, record->buf + hlen, size, > + size = psz_zone_read_buffer(zone, record->buf + hlen, size, > sizeof(struct psz_kmsg_header)); > if (unlikely(size < 0)) { > kfree(record->buf); > @@ -641,6 +842,32 @@ static ssize_t psz_kmsg_read(struct pstore_zone *zone, > return size + hlen; > } > > +static ssize_t psz_record_read(struct pstore_zone *zone, > + struct pstore_record *record) > +{ > + size_t len; > + struct psz_buffer *buf; > + > + if (!zone || !record) > + return -ENOSPC; > + > + buf = (struct psz_buffer *)zone->oldbuf; > + if (!buf) > + return -ENOMSG; > + > + len = atomic_read(&buf->datalen); > + record->buf = kmalloc(len, GFP_KERNEL); > + if (!record->buf) > + return -ENOMEM; > + > + if (unlikely(psz_zone_read_oldbuf(zone, record->buf, len, 0))) { > + kfree(record->buf); > + return -ENOMSG; > + } > + > + return len; > +} > + > static ssize_t psz_pstore_read(struct pstore_record *record) > { > struct psz_context *cxt = record->psi->data; > @@ -665,6 +892,9 @@ static ssize_t psz_pstore_read(struct pstore_record *record) > readop = psz_kmsg_read; > record->id = cxt->kmsg_read_cnt - 1; > break; > + case PSTORE_TYPE_PMSG: > + readop = psz_record_read; > + break; > default: > goto next_zone; > } > @@ -720,6 +950,8 @@ static void psz_free_all_zones(struct psz_context *cxt) > { > if (cxt->opszs) > psz_free_zones(&cxt->opszs, &cxt->kmsg_max_cnt); > + if (cxt->ppsz) > + psz_free_zone(&cxt->ppsz); > } > > static struct pstore_zone *psz_init_zone(enum pstore_type_id type, > @@ -753,8 +985,10 @@ static struct pstore_zone *psz_init_zone(enum pstore_type_id type, > zone->type = type; > zone->buffer_size = size - sizeof(struct psz_buffer); > zone->buffer->sig = type ^ PSZ_SIG; > + zone->oldbuf = NULL; > atomic_set(&zone->dirty, 0); > atomic_set(&zone->buffer->datalen, 0); > + atomic_set(&zone->buffer->start, 0); > > *off += size; > > @@ -809,19 +1043,28 @@ static int psz_alloc_zones(struct psz_context *cxt) > struct pstore_zone_info *info = cxt->pstore_zone_info; > loff_t off = 0; > int err; > - size_t size; > + size_t off_size = 0; > + > + off_size += info->pmsg_size; > + cxt->ppsz = psz_init_zone(PSTORE_TYPE_PMSG, &off, info->pmsg_size); > + if (IS_ERR(cxt->ppsz)) { > + err = PTR_ERR(cxt->ppsz); > + cxt->ppsz = NULL; > + goto free_out; > + } > > - size = info->total_size; > - cxt->opszs = psz_init_zones(PSTORE_TYPE_DMESG, &off, size, > + cxt->opszs = psz_init_zones(PSTORE_TYPE_DMESG, &off, > + info->total_size - off_size, > info->kmsg_size, &cxt->kmsg_max_cnt); > if (IS_ERR(cxt->opszs)) { > err = PTR_ERR(cxt->opszs); > cxt->opszs = NULL; > - goto fail_out; > + goto free_out; > } > > return 0; > -fail_out: > +free_out: > + psz_free_all_zones(cxt); > return err; > } > > @@ -844,7 +1087,7 @@ int register_pstore_zone(struct pstore_zone_info *info) > return -EINVAL; > } > > - if (!info->kmsg_size) { > + if (!info->kmsg_size && !info->pmsg_size) { > pr_warn("at least one record size must be non-zero\n"); > return -EINVAL; > } > @@ -866,6 +1109,7 @@ int register_pstore_zone(struct pstore_zone_info *info) > > check_size(total_size, 4096); > check_size(kmsg_size, SECTOR_SIZE); > + check_size(pmsg_size, SECTOR_SIZE); > > #undef check_size > > @@ -891,6 +1135,7 @@ int register_pstore_zone(struct pstore_zone_info *info) > pr_debug("register %s with properties:\n", info->name); > pr_debug("\ttotal size : %ld Bytes\n", info->total_size); > pr_debug("\tkmsg size : %ld Bytes\n", info->kmsg_size); > + pr_debug("\tpmsg size : %ld Bytes\n", info->pmsg_size); > > err = psz_alloc_zones(cxt); > if (err) { > @@ -920,6 +1165,10 @@ int register_pstore_zone(struct pstore_zone_info *info) > pr_cont(",panic_write"); > pr_cont(")"); > } > + if (info->pmsg_size) { > + cxt->pstore.flags |= PSTORE_FLAGS_PMSG; > + pr_cont(" pmsg"); > + } > pr_cont("\n"); > > err = pstore_register(&cxt->pstore); > diff --git a/include/linux/pstore_zone.h b/include/linux/pstore_zone.h > index eb005d9ae40c..29c367a3bd80 100644 > --- a/include/linux/pstore_zone.h > +++ b/include/linux/pstore_zone.h > @@ -17,6 +17,7 @@ typedef ssize_t (*pstore_zone_write_op)(const char *, size_t, loff_t); > * @kmsg_size: The size of oops/panic zone. Zero means disabled, otherwise, > * it must be multiple of SECTOR_SIZE(512 Bytes). > * @max_reason: Maximum kmsg dump reason to store. > + * @pmsg_size: The size of pmsg zone which is the same as @kmsg_size. > * @read: The general read operation. Both of the function parameters > * @size and @offset are relative value to storage. > * On success, the number of bytes should be returned, others > @@ -33,6 +34,7 @@ struct pstore_zone_info { > unsigned long total_size; > unsigned long kmsg_size; > int max_reason; > + unsigned long pmsg_size; > pstore_zone_read_op read; > pstore_zone_write_op write; > pstore_zone_write_op panic_write; > -- WeiXiong Liao