pstore/zone,blk: Add console frontend support

Support backend for console. To enable console backend, just make
console_size be greater than 0 and a multiple of 4096.

Signed-off-by: WeiXiong Liao <[email protected]>
Link: https://lore.kernel.org/lkml/[email protected]/
Signed-off-by: Kees Cook <[email protected]>

Author: WeiXiong Liao

fs/pstore/Kconfig

@@ -180,11 +180,11 @@ config PSTORE_BLK_BLKDEV
 	help
 	  Which block device should be used for pstore/blk.
 
-	  It accept the following variants:
+	  It accepts the following variants:
 	  1) <hex_major><hex_minor> device number in hexadecimal representation,
 	     with no leading 0x, for example b302.
-	  2) /dev/<disk_name> represents the device number of disk
-	  3) /dev/<disk_name><decimal> represents the device number
+	  2) /dev/<disk_name> represents the device name of disk
+	  3) /dev/<disk_name><decimal> represents the device name and number
 	     of partition - device number of disk plus the partition number
 	  4) /dev/<disk_name>p<decimal> - same as the above, this form is
 	     used when disk name of partitioned disk ends with a digit.
@@ -236,3 +236,15 @@ config PSTORE_BLK_PMSG_SIZE
 
 	  NOTE that, both Kconfig and module parameters can configure
 	  pstore/blk, but module parameters have priority over Kconfig.
+
+config PSTORE_BLK_CONSOLE_SIZE
+	int "Size in Kbytes of console log to store"
+	depends on PSTORE_BLK
+	depends on PSTORE_CONSOLE
+	default 64
+	help
+	  This just sets size of console log (console_size) to store via
+	  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.

fs/pstore/blk.c

@@ -35,6 +35,14 @@ static long pmsg_size = -1;
 module_param(pmsg_size, long, 0400);
 MODULE_PARM_DESC(pmsg_size, "pmsg size in kbytes");
 
+#if IS_ENABLED(CONFIG_PSTORE_CONSOLE)
+static long console_size = CONFIG_PSTORE_BLK_CONSOLE_SIZE;
+#else
+static long console_size = -1;
+#endif
+module_param(console_size, long, 0400);
+MODULE_PARM_DESC(console_size, "console size in kbytes");
+
 /*
  * blkdev - the block device to use for pstore storage
  *
@@ -91,7 +99,8 @@ struct bdev_info {
  *		whole disk).
  *		On success, the number of bytes should be returned, others
  *		means error.
- * @write:	The same as @read.
+ * @write:	The same as @read, but the following error number:
+ *		-EBUSY means try to write again later.
  * @panic_write:The write operation only used for panic case. It's optional
  *		if you do not care panic log. The parameters and return value
  *		are the same as @read.
@@ -142,6 +151,7 @@ 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);
+	verify_size(console_size, 4096, dev->flags & PSTORE_FLAGS_CONSOLE);
 #undef verify_size
 
 	pstore_zone_info->total_size = dev->total_size;

fs/pstore/zone.c

@@ -91,10 +91,12 @@ struct pstore_zone {
  *
  * @kpszs: kmsg dump storage zones
  * @ppsz: pmsg storage zone
+ * @cpsz: console storage zone
  * @kmsg_max_cnt: max count of @kpszs
  * @kmsg_read_cnt: counter of total read kmsg dumps
  * @kmsg_write_cnt: counter of total kmsg dump writes
  * @pmsg_read_cnt: counter of total read pmsg zone
+ * @console_read_cnt: counter of total read console zone
  * @oops_counter: counter of oops dumps
  * @panic_counter: counter of panic dumps
  * @recovered: whether finished recovering data from storage
@@ -106,10 +108,12 @@ struct pstore_zone {
 struct psz_context {
 	struct pstore_zone **kpszs;
 	struct pstore_zone *ppsz;
+	struct pstore_zone *cpsz;
 	unsigned int kmsg_max_cnt;
 	unsigned int kmsg_read_cnt;
 	unsigned int kmsg_write_cnt;
 	unsigned int pmsg_read_cnt;
+	unsigned int console_read_cnt;
 	/*
 	 * These counters should be calculated during recovery.
 	 * It records the oops/panic times after crashes rather than boots.
@@ -129,6 +133,9 @@ struct psz_context {
 };
 static struct psz_context pstore_zone_cxt;
 
+static void psz_flush_all_dirty_zones(struct work_struct *);
+static DECLARE_DELAYED_WORK(psz_cleaner, psz_flush_all_dirty_zones);
+
 /**
  * enum psz_flush_mode - flush mode for psz_zone_write()
  *
@@ -237,6 +244,9 @@ static int psz_zone_write(struct pstore_zone *zone,
 	return 0;
 dirty:
 	atomic_set(&zone->dirty, true);
+	/* flush dirty zones nicely */
+	if (wcnt == -EBUSY && !is_on_panic())
+		schedule_delayed_work(&psz_cleaner, msecs_to_jiffies(500));
 	return -EBUSY;
 }
 
@@ -293,6 +303,21 @@ static int psz_move_zone(struct pstore_zone *old, struct pstore_zone *new)
 	return 0;
 }
 
+static void psz_flush_all_dirty_zones(struct work_struct *work)
+{
+	struct psz_context *cxt = &pstore_zone_cxt;
+	int ret = 0;
+
+	if (cxt->ppsz)
+		ret |= psz_flush_dirty_zone(cxt->ppsz);
+	if (cxt->cpsz)
+		ret |= psz_flush_dirty_zone(cxt->cpsz);
+	if (cxt->kpszs)
+		ret |= psz_flush_dirty_zones(cxt->kpszs, cxt->kmsg_max_cnt);
+	if (ret && cxt->pstore_zone_info)
+		schedule_delayed_work(&psz_cleaner, msecs_to_jiffies(1000));
+}
+
 static int psz_kmsg_recover_data(struct psz_context *cxt)
 {
 	struct pstore_zone_info *info = cxt->pstore_zone_info;
@@ -545,6 +570,10 @@ static inline int psz_recovery(struct psz_context *cxt)
 		goto out;
 
 	ret = psz_recover_zone(cxt, cxt->ppsz);
+	if (ret)
+		goto out;
+
+	ret = psz_recover_zone(cxt, cxt->cpsz);
 
 out:
 	if (unlikely(ret))
@@ -562,6 +591,7 @@ static int psz_pstore_open(struct pstore_info *psi)
 
 	cxt->kmsg_read_cnt = 0;
 	cxt->pmsg_read_cnt = 0;
+	cxt->console_read_cnt = 0;
 	return 0;
 }
 
@@ -626,8 +656,9 @@ static int psz_pstore_erase(struct pstore_record *record)
 		return psz_kmsg_erase(cxt, cxt->kpszs[record->id], record);
 	case PSTORE_TYPE_PMSG:
 		return psz_record_erase(cxt, cxt->ppsz);
-	default:
-		return -EINVAL;
+	case PSTORE_TYPE_CONSOLE:
+		return psz_record_erase(cxt, cxt->cpsz);
+	default: return -EINVAL;
 	}
 }
 
@@ -690,9 +721,10 @@ static int notrace psz_kmsg_write(struct psz_context *cxt,
 		return -ENOSPC;
 
 	ret = psz_kmsg_write_record(cxt, record);
-	if (!ret) {
+	if (!ret && is_on_panic()) {
+		/* ensure all data are flushed to storage when panic */
 		pr_debug("try to flush other dirty zones\n");
-		psz_flush_dirty_zones(cxt->kpszs, cxt->kmsg_max_cnt);
+		psz_flush_all_dirty_zones(NULL);
 	}
 
 	/* always return 0 as we had handled it on buffer */
@@ -756,9 +788,18 @@ static int notrace psz_pstore_write(struct pstore_record *record)
 			record->reason == KMSG_DUMP_PANIC)
 		atomic_set(&cxt->on_panic, 1);
 
+	/*
+	 * if on panic, do not write except panic records
+	 * Fix case that panic_write prints log which wakes up console backend.
+	 */
+	if (is_on_panic() && record->type != PSTORE_TYPE_DMESG)
+		return -EBUSY;
+
 	switch (record->type) {
 	case PSTORE_TYPE_DMESG:
 		return psz_kmsg_write(cxt, record);
+	case PSTORE_TYPE_CONSOLE:
+		return psz_record_write(cxt->cpsz, record);
 	case PSTORE_TYPE_PMSG:
 		return psz_record_write(cxt->ppsz, record);
 	default:
@@ -783,6 +824,13 @@ static struct pstore_zone *psz_read_next_zone(struct psz_context *cxt)
 			return zone;
 	}
 
+	if (cxt->console_read_cnt == 0) {
+		cxt->console_read_cnt++;
+		zone = cxt->cpsz;
+		if (psz_old_ok(zone))
+			return zone;
+	}
+
 	return NULL;
 }
 
@@ -893,6 +941,8 @@ 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_CONSOLE:
+		fallthrough;
 	case PSTORE_TYPE_PMSG:
 		readop = psz_record_read;
 		break;
@@ -953,6 +1003,8 @@ static void psz_free_all_zones(struct psz_context *cxt)
 		psz_free_zones(&cxt->kpszs, &cxt->kmsg_max_cnt);
 	if (cxt->ppsz)
 		psz_free_zone(&cxt->ppsz);
+	if (cxt->cpsz)
+		psz_free_zone(&cxt->cpsz);
 }
 
 static struct pstore_zone *psz_init_zone(enum pstore_type_id type,
@@ -1054,6 +1106,15 @@ static int psz_alloc_zones(struct psz_context *cxt)
 		goto free_out;
 	}
 
+	off_size += info->console_size;
+	cxt->cpsz = psz_init_zone(PSTORE_TYPE_CONSOLE, &off,
+			info->console_size);
+	if (IS_ERR(cxt->cpsz)) {
+		err = PTR_ERR(cxt->cpsz);
+		cxt->cpsz = NULL;
+		goto free_out;
+	}
+
 	cxt->kpszs = psz_init_zones(PSTORE_TYPE_DMESG, &off,
 			info->total_size - off_size,
 			info->kmsg_size, &cxt->kmsg_max_cnt);
@@ -1088,7 +1149,7 @@ int register_pstore_zone(struct pstore_zone_info *info)
 		return -EINVAL;
 	}
 
-	if (!info->kmsg_size && !info->pmsg_size) {
+	if (!info->kmsg_size && !info->pmsg_size && !info->console_size) {
 		pr_warn("at least one record size must be non-zero\n");
 		return -EINVAL;
 	}
@@ -1111,6 +1172,7 @@ 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);
+	check_size(console_size, SECTOR_SIZE);
 
 #undef check_size
 
@@ -1137,6 +1199,7 @@ int register_pstore_zone(struct pstore_zone_info *info)
 	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);
+	pr_debug("\tconsole size : %ld Bytes\n", info->console_size);
 
 	err = psz_alloc_zones(cxt);
 	if (err) {
@@ -1170,6 +1233,10 @@ int register_pstore_zone(struct pstore_zone_info *info)
 		cxt->pstore.flags |= PSTORE_FLAGS_PMSG;
 		pr_cont(" pmsg");
 	}
+	if (info->console_size) {
+		cxt->pstore.flags |= PSTORE_FLAGS_CONSOLE;
+		pr_cont(" console");
+	}
 	pr_cont("\n");
 
 	err = pstore_register(&cxt->pstore);
@@ -1211,6 +1278,10 @@ void unregister_pstore_zone(struct pstore_zone_info *info)
 	/* Stop incoming writes from pstore. */
 	pstore_unregister(&cxt->pstore);
 
+	/* Flush any pending writes. */
+	psz_flush_all_dirty_zones(NULL);
+	flush_delayed_work(&psz_cleaner);
+
 	/* Clean up allocations. */
 	kfree(cxt->pstore.buf);
 	cxt->pstore.buf = NULL;

include/linux/pstore_zone.h

@@ -18,11 +18,12 @@ typedef ssize_t (*pstore_zone_write_op)(const char *, size_t, loff_t);
  *		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.
+ * @console_size:The size of console 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
  *		means error.
- * @write:	The same as @read.
+ * @write:	The same as @read, but -EBUSY means try to write again later.
  * @panic_write:The write operation only used for panic case. It's optional
  *		if you do not care panic log. The parameters and return value
  *		are the same as @read.
@@ -35,6 +36,7 @@ struct pstore_zone_info {
 	unsigned long kmsg_size;
 	int max_reason;
 	unsigned long pmsg_size;
+	unsigned long console_size;
 	pstore_zone_read_op read;
 	pstore_zone_write_op write;
 	pstore_zone_write_op panic_write;