From 78989f0a5592182a3d45d869ddaafc71c8f673af Mon Sep 17 00:00:00 2001 From: Hari Bathini Date: Fri, 6 Feb 2015 01:06:04 +0530 Subject: [PATCH] powerpc/nvram: Move generic code for nvram and pstore With minor checks, we can move most of the code for nvram under pseries to a common place to be re-used by other powerpc platforms like powernv. This patch moves such common code to arch/powerpc/kernel/nvram_64.c file. Signed-off-by: Hari Bathini [mpe: Move select of ZLIB_DEFLATE to PPC64 to fix the build] Signed-off-by: Michael Ellerman --- arch/powerpc/include/asm/nvram.h | 50 +- arch/powerpc/include/asm/rtas.h | 4 + arch/powerpc/kernel/nvram_64.c | 656 ++++++++++++++++++++++++ arch/powerpc/platforms/Kconfig.cputype | 1 + arch/powerpc/platforms/pseries/Kconfig | 1 - arch/powerpc/platforms/pseries/nvram.c | 666 +------------------------ 6 files changed, 715 insertions(+), 663 deletions(-) diff --git a/arch/powerpc/include/asm/nvram.h b/arch/powerpc/include/asm/nvram.h index b0fe0fe4e626..09a518bb7c03 100644 --- a/arch/powerpc/include/asm/nvram.h +++ b/arch/powerpc/include/asm/nvram.h @@ -9,12 +9,43 @@ #ifndef _ASM_POWERPC_NVRAM_H #define _ASM_POWERPC_NVRAM_H - +#include #include #include #include +/* + * Set oops header version to distinguish between old and new format header. + * lnx,oops-log partition max size is 4000, header version > 4000 will + * help in identifying new header. + */ +#define OOPS_HDR_VERSION 5000 + +struct err_log_info { + __be32 error_type; + __be32 seq_num; +}; + +struct nvram_os_partition { + const char *name; + int req_size; /* desired size, in bytes */ + int min_size; /* minimum acceptable size (0 means req_size) */ + long size; /* size of data portion (excluding err_log_info) */ + long index; /* offset of data portion of partition */ + bool os_partition; /* partition initialized by OS, not FW */ +}; + +struct oops_log_info { + __be16 version; + __be16 report_length; + __be64 timestamp; +} __attribute__((packed)); + +extern struct nvram_os_partition oops_log_partition; + #ifdef CONFIG_PPC_PSERIES +extern struct nvram_os_partition rtas_log_partition; + extern int nvram_write_error_log(char * buff, int length, unsigned int err_type, unsigned int err_seq); extern int nvram_read_error_log(char * buff, int length, @@ -50,6 +81,23 @@ extern void pmac_xpram_write(int xpaddr, u8 data); /* Synchronize NVRAM */ extern void nvram_sync(void); +/* Initialize NVRAM OS partition */ +extern int __init nvram_init_os_partition(struct nvram_os_partition *part); + +/* Initialize NVRAM oops partition */ +extern void __init nvram_init_oops_partition(int rtas_partition_exists); + +/* Read a NVRAM partition */ +extern int nvram_read_partition(struct nvram_os_partition *part, char *buff, + int length, unsigned int *err_type, + unsigned int *error_log_cnt); + +/* Write to NVRAM OS partition */ +extern int nvram_write_os_partition(struct nvram_os_partition *part, + char *buff, int length, + unsigned int err_type, + unsigned int error_log_cnt); + /* Determine NVRAM size */ extern ssize_t nvram_get_size(void); diff --git a/arch/powerpc/include/asm/rtas.h b/arch/powerpc/include/asm/rtas.h index 2e23e92a4372..d1bd001566df 100644 --- a/arch/powerpc/include/asm/rtas.h +++ b/arch/powerpc/include/asm/rtas.h @@ -343,8 +343,12 @@ extern int early_init_dt_scan_rtas(unsigned long node, extern void pSeries_log_error(char *buf, unsigned int err_type, int fatal); #ifdef CONFIG_PPC_PSERIES +extern unsigned long last_rtas_event; +extern int clobbering_unread_rtas_event(void); extern int pseries_devicetree_update(s32 scope); extern void post_mobility_fixup(void); +#else +static inline int clobbering_unread_rtas_event(void) { return 0; } #endif #ifdef CONFIG_PPC_RTAS_DAEMON diff --git a/arch/powerpc/kernel/nvram_64.c b/arch/powerpc/kernel/nvram_64.c index 34f7c9b7cd96..42e5c6a9c214 100644 --- a/arch/powerpc/kernel/nvram_64.c +++ b/arch/powerpc/kernel/nvram_64.c @@ -26,6 +26,9 @@ #include #include #include +#include +#include +#include #include #include #include @@ -54,6 +57,659 @@ struct nvram_partition { static LIST_HEAD(nvram_partitions); +#ifdef CONFIG_PPC_PSERIES +struct nvram_os_partition rtas_log_partition = { + .name = "ibm,rtas-log", + .req_size = 2079, + .min_size = 1055, + .index = -1, + .os_partition = true +}; +#endif + +struct nvram_os_partition oops_log_partition = { + .name = "lnx,oops-log", + .req_size = 4000, + .min_size = 2000, + .index = -1, + .os_partition = true +}; + +static const char *nvram_os_partitions[] = { +#ifdef CONFIG_PPC_PSERIES + "ibm,rtas-log", +#endif + "lnx,oops-log", + NULL +}; + +static void oops_to_nvram(struct kmsg_dumper *dumper, + enum kmsg_dump_reason reason); + +static struct kmsg_dumper nvram_kmsg_dumper = { + .dump = oops_to_nvram +}; + +/* + * For capturing and compressing an oops or panic report... + + * big_oops_buf[] holds the uncompressed text we're capturing. + * + * oops_buf[] holds the compressed text, preceded by a oops header. + * oops header has u16 holding the version of oops header (to differentiate + * between old and new format header) followed by u16 holding the length of + * the compressed* text (*Or uncompressed, if compression fails.) and u64 + * holding the timestamp. oops_buf[] gets written to NVRAM. + * + * oops_log_info points to the header. oops_data points to the compressed text. + * + * +- oops_buf + * | +- oops_data + * v v + * +-----------+-----------+-----------+------------------------+ + * | version | length | timestamp | text | + * | (2 bytes) | (2 bytes) | (8 bytes) | (oops_data_sz bytes) | + * +-----------+-----------+-----------+------------------------+ + * ^ + * +- oops_log_info + * + * We preallocate these buffers during init to avoid kmalloc during oops/panic. + */ +static size_t big_oops_buf_sz; +static char *big_oops_buf, *oops_buf; +static char *oops_data; +static size_t oops_data_sz; + +/* Compression parameters */ +#define COMPR_LEVEL 6 +#define WINDOW_BITS 12 +#define MEM_LEVEL 4 +static struct z_stream_s stream; + +#ifdef CONFIG_PSTORE +#ifdef CONFIG_PPC_PSERIES +static struct nvram_os_partition of_config_partition = { + .name = "of-config", + .index = -1, + .os_partition = false +}; +#endif + +static struct nvram_os_partition common_partition = { + .name = "common", + .index = -1, + .os_partition = false +}; + +static enum pstore_type_id nvram_type_ids[] = { + PSTORE_TYPE_DMESG, + PSTORE_TYPE_PPC_COMMON, + -1, + -1, + -1 +}; +static int read_type; +#endif + +/* nvram_write_os_partition + * + * We need to buffer the error logs into nvram to ensure that we have + * the failure information to decode. If we have a severe error there + * is no way to guarantee that the OS or the machine is in a state to + * get back to user land and write the error to disk. For example if + * the SCSI device driver causes a Machine Check by writing to a bad + * IO address, there is no way of guaranteeing that the device driver + * is in any state that is would also be able to write the error data + * captured to disk, thus we buffer it in NVRAM for analysis on the + * next boot. + * + * In NVRAM the partition containing the error log buffer will looks like: + * Header (in bytes): + * +-----------+----------+--------+------------+------------------+ + * | signature | checksum | length | name | data | + * |0 |1 |2 3|4 15|16 length-1| + * +-----------+----------+--------+------------+------------------+ + * + * The 'data' section would look like (in bytes): + * +--------------+------------+-----------------------------------+ + * | event_logged | sequence # | error log | + * |0 3|4 7|8 error_log_size-1| + * +--------------+------------+-----------------------------------+ + * + * event_logged: 0 if event has not been logged to syslog, 1 if it has + * sequence #: The unique sequence # for each event. (until it wraps) + * error log: The error log from event_scan + */ +int nvram_write_os_partition(struct nvram_os_partition *part, + char *buff, int length, + unsigned int err_type, + unsigned int error_log_cnt) +{ + int rc; + loff_t tmp_index; + struct err_log_info info; + + if (part->index == -1) + return -ESPIPE; + + if (length > part->size) + length = part->size; + + info.error_type = cpu_to_be32(err_type); + info.seq_num = cpu_to_be32(error_log_cnt); + + tmp_index = part->index; + + rc = ppc_md.nvram_write((char *)&info, sizeof(struct err_log_info), + &tmp_index); + if (rc <= 0) { + pr_err("%s: Failed nvram_write (%d)\n", __func__, rc); + return rc; + } + + rc = ppc_md.nvram_write(buff, length, &tmp_index); + if (rc <= 0) { + pr_err("%s: Failed nvram_write (%d)\n", __func__, rc); + return rc; + } + + return 0; +} + +/* nvram_read_partition + * + * Reads nvram partition for at most 'length' + */ +int nvram_read_partition(struct nvram_os_partition *part, char *buff, + int length, unsigned int *err_type, + unsigned int *error_log_cnt) +{ + int rc; + loff_t tmp_index; + struct err_log_info info; + + if (part->index == -1) + return -1; + + if (length > part->size) + length = part->size; + + tmp_index = part->index; + + if (part->os_partition) { + rc = ppc_md.nvram_read((char *)&info, + sizeof(struct err_log_info), + &tmp_index); + if (rc <= 0) { + pr_err("%s: Failed nvram_read (%d)\n", __func__, rc); + return rc; + } + } + + rc = ppc_md.nvram_read(buff, length, &tmp_index); + if (rc <= 0) { + pr_err("%s: Failed nvram_read (%d)\n", __func__, rc); + return rc; + } + + if (part->os_partition) { + *error_log_cnt = be32_to_cpu(info.seq_num); + *err_type = be32_to_cpu(info.error_type); + } + + return 0; +} + +/* nvram_init_os_partition + * + * This sets up a partition with an "OS" signature. + * + * The general strategy is the following: + * 1.) If a partition with the indicated name already exists... + * - If it's large enough, use it. + * - Otherwise, recycle it and keep going. + * 2.) Search for a free partition that is large enough. + * 3.) If there's not a free partition large enough, recycle any obsolete + * OS partitions and try again. + * 4.) Will first try getting a chunk that will satisfy the requested size. + * 5.) If a chunk of the requested size cannot be allocated, then try finding + * a chunk that will satisfy the minum needed. + * + * Returns 0 on success, else -1. + */ +int __init nvram_init_os_partition(struct nvram_os_partition *part) +{ + loff_t p; + int size; + + /* Look for ours */ + p = nvram_find_partition(part->name, NVRAM_SIG_OS, &size); + + /* Found one but too small, remove it */ + if (p && size < part->min_size) { + pr_info("nvram: Found too small %s partition," + " removing it...\n", part->name); + nvram_remove_partition(part->name, NVRAM_SIG_OS, NULL); + p = 0; + } + + /* Create one if we didn't find */ + if (!p) { + p = nvram_create_partition(part->name, NVRAM_SIG_OS, + part->req_size, part->min_size); + if (p == -ENOSPC) { + pr_info("nvram: No room to create %s partition, " + "deleting any obsolete OS partitions...\n", + part->name); + nvram_remove_partition(NULL, NVRAM_SIG_OS, + nvram_os_partitions); + p = nvram_create_partition(part->name, NVRAM_SIG_OS, + part->req_size, part->min_size); + } + } + + if (p <= 0) { + pr_err("nvram: Failed to find or create %s" + " partition, err %d\n", part->name, (int)p); + return -1; + } + + part->index = p; + part->size = nvram_get_partition_size(p) - sizeof(struct err_log_info); + + return 0; +} + +/* Derived from logfs_compress() */ +static int nvram_compress(const void *in, void *out, size_t inlen, + size_t outlen) +{ + int err, ret; + + ret = -EIO; + err = zlib_deflateInit2(&stream, COMPR_LEVEL, Z_DEFLATED, WINDOW_BITS, + MEM_LEVEL, Z_DEFAULT_STRATEGY); + if (err != Z_OK) + goto error; + + stream.next_in = in; + stream.avail_in = inlen; + stream.total_in = 0; + stream.next_out = out; + stream.avail_out = outlen; + stream.total_out = 0; + + err = zlib_deflate(&stream, Z_FINISH); + if (err != Z_STREAM_END) + goto error; + + err = zlib_deflateEnd(&stream); + if (err != Z_OK) + goto error; + + if (stream.total_out >= stream.total_in) + goto error; + + ret = stream.total_out; +error: + return ret; +} + +/* Compress the text from big_oops_buf into oops_buf. */ +static int zip_oops(size_t text_len) +{ + struct oops_log_info *oops_hdr = (struct oops_log_info *)oops_buf; + int zipped_len = nvram_compress(big_oops_buf, oops_data, text_len, + oops_data_sz); + if (zipped_len < 0) { + pr_err("nvram: compression failed; returned %d\n", zipped_len); + pr_err("nvram: logging uncompressed oops/panic report\n"); + return -1; + } + oops_hdr->version = cpu_to_be16(OOPS_HDR_VERSION); + oops_hdr->report_length = cpu_to_be16(zipped_len); + oops_hdr->timestamp = cpu_to_be64(get_seconds()); + return 0; +} + +#ifdef CONFIG_PSTORE +static int nvram_pstore_open(struct pstore_info *psi) +{ + /* Reset the iterator to start reading partitions again */ + read_type = -1; + return 0; +} + +/** + * nvram_pstore_write - pstore write callback for nvram + * @type: Type of message logged + * @reason: reason behind dump (oops/panic) + * @id: identifier to indicate the write performed + * @part: pstore writes data to registered buffer in parts, + * part number will indicate the same. + * @count: Indicates oops count + * @compressed: Flag to indicate the log is compressed + * @size: number of bytes written to the registered buffer + * @psi: registered pstore_info structure + * + * Called by pstore_dump() when an oops or panic report is logged in the + * printk buffer. + * Returns 0 on successful write. + */ +static int nvram_pstore_write(enum pstore_type_id type, + enum kmsg_dump_reason reason, + u64 *id, unsigned int part, int count, + bool compressed, size_t size, + struct pstore_info *psi) +{ + int rc; + unsigned int err_type = ERR_TYPE_KERNEL_PANIC; + struct oops_log_info *oops_hdr = (struct oops_log_info *) oops_buf; + + /* part 1 has the recent messages from printk buffer */ + if (part > 1 || (type != PSTORE_TYPE_DMESG)) + return -1; + + if (clobbering_unread_rtas_event()) + return -1; + + oops_hdr->version = cpu_to_be16(OOPS_HDR_VERSION); + oops_hdr->report_length = cpu_to_be16(size); + oops_hdr->timestamp = cpu_to_be64(get_seconds()); + + if (compressed) + err_type = ERR_TYPE_KERNEL_PANIC_GZ; + + rc = nvram_write_os_partition(&oops_log_partition, oops_buf, + (int) (sizeof(*oops_hdr) + size), err_type, count); + + if (rc != 0) + return rc; + + *id = part; + return 0; +} + +/* + * Reads the oops/panic report, rtas, of-config and common partition. + * Returns the length of the data we read from each partition. + * Returns 0 if we've been called before. + */ +static ssize_t nvram_pstore_read(u64 *id, enum pstore_type_id *type, + int *count, struct timespec *time, char **buf, + bool *compressed, struct pstore_info *psi) +{ + struct oops_log_info *oops_hdr; + unsigned int err_type, id_no, size = 0; + struct nvram_os_partition *part = NULL; + char *buff = NULL; + int sig = 0; + loff_t p; + + read_type++; + + switch (nvram_type_ids[read_type]) { + case PSTORE_TYPE_DMESG: + part = &oops_log_partition; + *type = PSTORE_TYPE_DMESG; + break; + case PSTORE_TYPE_PPC_COMMON: + sig = NVRAM_SIG_SYS; + part = &common_partition; + *type = PSTORE_TYPE_PPC_COMMON; + *id = PSTORE_TYPE_PPC_COMMON; + time->tv_sec = 0; + time->tv_nsec = 0; + break; +#ifdef CONFIG_PPC_PSERIES + case PSTORE_TYPE_PPC_RTAS: + part = &rtas_log_partition; + *type = PSTORE_TYPE_PPC_RTAS; + time->tv_sec = last_rtas_event; + time->tv_nsec = 0; + break; + case PSTORE_TYPE_PPC_OF: + sig = NVRAM_SIG_OF; + part = &of_config_partition; + *type = PSTORE_TYPE_PPC_OF; + *id = PSTORE_TYPE_PPC_OF; + time->tv_sec = 0; + time->tv_nsec = 0; + break; +#endif + default: + return 0; + } + + if (!part->os_partition) { + p = nvram_find_partition(part->name, sig, &size); + if (p <= 0) { + pr_err("nvram: Failed to find partition %s, " + "err %d\n", part->name, (int)p); + return 0; + } + part->index = p; + part->size = size; + } + + buff = kmalloc(part->size, GFP_KERNEL); + + if (!buff) + return -ENOMEM; + + if (nvram_read_partition(part, buff, part->size, &err_type, &id_no)) { + kfree(buff); + return 0; + } + + *count = 0; + + if (part->os_partition) + *id = id_no; + + if (nvram_type_ids[read_type] == PSTORE_TYPE_DMESG) { + size_t length, hdr_size; + + oops_hdr = (struct oops_log_info *)buff; + if (be16_to_cpu(oops_hdr->version) < OOPS_HDR_VERSION) { + /* Old format oops header had 2-byte record size */ + hdr_size = sizeof(u16); + length = be16_to_cpu(oops_hdr->version); + time->tv_sec = 0; + time->tv_nsec = 0; + } else { + hdr_size = sizeof(*oops_hdr); + length = be16_to_cpu(oops_hdr->report_length); + time->tv_sec = be64_to_cpu(oops_hdr->timestamp); + time->tv_nsec = 0; + } + *buf = kmalloc(length, GFP_KERNEL); + if (*buf == NULL) + return -ENOMEM; + memcpy(*buf, buff + hdr_size, length); + kfree(buff); + + if (err_type == ERR_TYPE_KERNEL_PANIC_GZ) + *compressed = true; + else + *compressed = false; + return length; + } + + *buf = buff; + return part->size; +} + +static struct pstore_info nvram_pstore_info = { + .owner = THIS_MODULE, + .name = "nvram", + .open = nvram_pstore_open, + .read = nvram_pstore_read, + .write = nvram_pstore_write, +}; + +static int nvram_pstore_init(void) +{ + int rc = 0; + + nvram_type_ids[2] = PSTORE_TYPE_PPC_RTAS; + nvram_type_ids[3] = PSTORE_TYPE_PPC_OF; + + nvram_pstore_info.buf = oops_data; + nvram_pstore_info.bufsize = oops_data_sz; + + spin_lock_init(&nvram_pstore_info.buf_lock); + + rc = pstore_register(&nvram_pstore_info); + if (rc != 0) + pr_err("nvram: pstore_register() failed, defaults to " + "kmsg_dump; returned %d\n", rc); + + return rc; +} +#else +static int nvram_pstore_init(void) +{ + return -1; +} +#endif + +void __init nvram_init_oops_partition(int rtas_partition_exists) +{ + int rc; + + rc = nvram_init_os_partition(&oops_log_partition); + if (rc != 0) { +#ifdef CONFIG_PPC_PSERIES + if (!rtas_partition_exists) { + pr_err("nvram: Failed to initialize oops partition!"); + return; + } + pr_notice("nvram: Using %s partition to log both" + " RTAS errors and oops/panic reports\n", + rtas_log_partition.name); + memcpy(&oops_log_partition, &rtas_log_partition, + sizeof(rtas_log_partition)); +#else + pr_err("nvram: Failed to initialize oops partition!"); + return; +#endif + } + oops_buf = kmalloc(oops_log_partition.size, GFP_KERNEL); + if (!oops_buf) { + pr_err("nvram: No memory for %s partition\n", + oops_log_partition.name); + return; + } + oops_data = oops_buf + sizeof(struct oops_log_info); + oops_data_sz = oops_log_partition.size - sizeof(struct oops_log_info); + + rc = nvram_pstore_init(); + + if (!rc) + return; + + /* + * Figure compression (preceded by elimination of each line's + * severity prefix) will reduce the oops/panic report to at most + * 45% of its original size. + */ + big_oops_buf_sz = (oops_data_sz * 100) / 45; + big_oops_buf = kmalloc(big_oops_buf_sz, GFP_KERNEL); + if (big_oops_buf) { + stream.workspace = kmalloc(zlib_deflate_workspacesize( + WINDOW_BITS, MEM_LEVEL), GFP_KERNEL); + if (!stream.workspace) { + pr_err("nvram: No memory for compression workspace; " + "skipping compression of %s partition data\n", + oops_log_partition.name); + kfree(big_oops_buf); + big_oops_buf = NULL; + } + } else { + pr_err("No memory for uncompressed %s data; " + "skipping compression\n", oops_log_partition.name); + stream.workspace = NULL; + } + + rc = kmsg_dump_register(&nvram_kmsg_dumper); + if (rc != 0) { + pr_err("nvram: kmsg_dump_register() failed; returned %d\n", rc); + kfree(oops_buf); + kfree(big_oops_buf); + kfree(stream.workspace); + } +} + +/* + * This is our kmsg_dump callback, called after an oops or panic report + * has been written to the printk buffer. We want to capture as much + * of the printk buffer as possible. First, capture as much as we can + * that we think will compress sufficiently to fit in the lnx,oops-log + * partition. If that's too much, go back and capture uncompressed text. + */ +static void oops_to_nvram(struct kmsg_dumper *dumper, + enum kmsg_dump_reason reason) +{ + struct oops_log_info *oops_hdr = (struct oops_log_info *)oops_buf; + static unsigned int oops_count = 0; + static bool panicking = false; + static DEFINE_SPINLOCK(lock); + unsigned long flags; + size_t text_len; + unsigned int err_type = ERR_TYPE_KERNEL_PANIC_GZ; + int rc = -1; + + switch (reason) { + case KMSG_DUMP_RESTART: + case KMSG_DUMP_HALT: + case KMSG_DUMP_POWEROFF: + /* These are almost always orderly shutdowns. */ + return; + case KMSG_DUMP_OOPS: + break; + case KMSG_DUMP_PANIC: + panicking = true; + break; + case KMSG_DUMP_EMERG: + if (panicking) + /* Panic report already captured. */ + return; + break; + default: + pr_err("%s: ignoring unrecognized KMSG_DUMP_* reason %d\n", + __func__, (int) reason); + return; + } + + if (clobbering_unread_rtas_event()) + return; + + if (!spin_trylock_irqsave(&lock, flags)) + return; + + if (big_oops_buf) { + kmsg_dump_get_buffer(dumper, false, + big_oops_buf, big_oops_buf_sz, &text_len); + rc = zip_oops(text_len); + } + if (rc != 0) { + kmsg_dump_rewind(dumper); + kmsg_dump_get_buffer(dumper, false, + oops_data, oops_data_sz, &text_len); + err_type = ERR_TYPE_KERNEL_PANIC; + oops_hdr->version = cpu_to_be16(OOPS_HDR_VERSION); + oops_hdr->report_length = cpu_to_be16(text_len); + oops_hdr->timestamp = cpu_to_be64(get_seconds()); + } + + (void) nvram_write_os_partition(&oops_log_partition, oops_buf, + (int) (sizeof(*oops_hdr) + text_len), err_type, + ++oops_count); + + spin_unlock_irqrestore(&lock, flags); +} + static loff_t dev_nvram_llseek(struct file *file, loff_t offset, int origin) { int size; diff --git a/arch/powerpc/platforms/Kconfig.cputype b/arch/powerpc/platforms/Kconfig.cputype index 76483e3acd60..80614c6088bc 100644 --- a/arch/powerpc/platforms/Kconfig.cputype +++ b/arch/powerpc/platforms/Kconfig.cputype @@ -2,6 +2,7 @@ config PPC64 bool "64-bit kernel" default n select HAVE_VIRT_CPU_ACCOUNTING + select ZLIB_DEFLATE help This option selects whether a 32-bit or a 64-bit kernel will be built. diff --git a/arch/powerpc/platforms/pseries/Kconfig b/arch/powerpc/platforms/pseries/Kconfig index a758a9c3bbba..54c87d5d349d 100644 --- a/arch/powerpc/platforms/pseries/Kconfig +++ b/arch/powerpc/platforms/pseries/Kconfig @@ -16,7 +16,6 @@ config PPC_PSERIES select PPC_UDBG_16550 select PPC_NATIVE select PPC_PCI_CHOICE if EXPERT - select ZLIB_DEFLATE select PPC_DOORBELL select HAVE_CONTEXT_TRACKING select HOTPLUG_CPU if SMP diff --git a/arch/powerpc/platforms/pseries/nvram.c b/arch/powerpc/platforms/pseries/nvram.c index 054a0ed5c7ee..533807efed61 100644 --- a/arch/powerpc/platforms/pseries/nvram.c +++ b/arch/powerpc/platforms/pseries/nvram.c @@ -20,7 +20,6 @@ #include #include #include -#include #include #include #include @@ -30,129 +29,17 @@ /* Max bytes to read/write in one go */ #define NVRW_CNT 0x20 -/* - * Set oops header version to distinguish between old and new format header. - * lnx,oops-log partition max size is 4000, header version > 4000 will - * help in identifying new header. - */ -#define OOPS_HDR_VERSION 5000 - static unsigned int nvram_size; static int nvram_fetch, nvram_store; static char nvram_buf[NVRW_CNT]; /* assume this is in the first 4GB */ static DEFINE_SPINLOCK(nvram_lock); -struct err_log_info { - __be32 error_type; - __be32 seq_num; -}; - -struct nvram_os_partition { - const char *name; - int req_size; /* desired size, in bytes */ - int min_size; /* minimum acceptable size (0 means req_size) */ - long size; /* size of data portion (excluding err_log_info) */ - long index; /* offset of data portion of partition */ - bool os_partition; /* partition initialized by OS, not FW */ -}; - -static struct nvram_os_partition rtas_log_partition = { - .name = "ibm,rtas-log", - .req_size = 2079, - .min_size = 1055, - .index = -1, - .os_partition = true -}; - -static struct nvram_os_partition oops_log_partition = { - .name = "lnx,oops-log", - .req_size = 4000, - .min_size = 2000, - .index = -1, - .os_partition = true -}; - -static const char *pseries_nvram_os_partitions[] = { - "ibm,rtas-log", - "lnx,oops-log", - NULL -}; - -struct oops_log_info { - __be16 version; - __be16 report_length; - __be64 timestamp; -} __attribute__((packed)); - -static void oops_to_nvram(struct kmsg_dumper *dumper, - enum kmsg_dump_reason reason); - -static struct kmsg_dumper nvram_kmsg_dumper = { - .dump = oops_to_nvram -}; - /* See clobbering_unread_rtas_event() */ #define NVRAM_RTAS_READ_TIMEOUT 5 /* seconds */ static unsigned long last_unread_rtas_event; /* timestamp */ -/* - * For capturing and compressing an oops or panic report... - - * big_oops_buf[] holds the uncompressed text we're capturing. - * - * oops_buf[] holds the compressed text, preceded by a oops header. - * oops header has u16 holding the version of oops header (to differentiate - * between old and new format header) followed by u16 holding the length of - * the compressed* text (*Or uncompressed, if compression fails.) and u64 - * holding the timestamp. oops_buf[] gets written to NVRAM. - * - * oops_log_info points to the header. oops_data points to the compressed text. - * - * +- oops_buf - * | +- oops_data - * v v - * +-----------+-----------+-----------+------------------------+ - * | version | length | timestamp | text | - * | (2 bytes) | (2 bytes) | (8 bytes) | (oops_data_sz bytes) | - * +-----------+-----------+-----------+------------------------+ - * ^ - * +- oops_log_info - * - * We preallocate these buffers during init to avoid kmalloc during oops/panic. - */ -static size_t big_oops_buf_sz; -static char *big_oops_buf, *oops_buf; -static char *oops_data; -static size_t oops_data_sz; - -/* Compression parameters */ -#define COMPR_LEVEL 6 -#define WINDOW_BITS 12 -#define MEM_LEVEL 4 -static struct z_stream_s stream; - #ifdef CONFIG_PSTORE -static struct nvram_os_partition of_config_partition = { - .name = "of-config", - .index = -1, - .os_partition = false -}; - -static struct nvram_os_partition common_partition = { - .name = "common", - .index = -1, - .os_partition = false -}; - -static enum pstore_type_id nvram_type_ids[] = { - PSTORE_TYPE_DMESG, - PSTORE_TYPE_PPC_RTAS, - PSTORE_TYPE_PPC_OF, - PSTORE_TYPE_PPC_COMMON, - -1 -}; -static int read_type; -static unsigned long last_rtas_event; +unsigned long last_rtas_event; #endif static ssize_t pSeries_nvram_read(char *buf, size_t count, loff_t *index) @@ -246,73 +133,11 @@ static ssize_t pSeries_nvram_get_size(void) return nvram_size ? nvram_size : -ENODEV; } - -/* nvram_write_os_partition, nvram_write_error_log +/* nvram_write_error_log * * We need to buffer the error logs into nvram to ensure that we have - * the failure information to decode. If we have a severe error there - * is no way to guarantee that the OS or the machine is in a state to - * get back to user land and write the error to disk. For example if - * the SCSI device driver causes a Machine Check by writing to a bad - * IO address, there is no way of guaranteeing that the device driver - * is in any state that is would also be able to write the error data - * captured to disk, thus we buffer it in NVRAM for analysis on the - * next boot. - * - * In NVRAM the partition containing the error log buffer will looks like: - * Header (in bytes): - * +-----------+----------+--------+------------+------------------+ - * | signature | checksum | length | name | data | - * |0 |1 |2 3|4 15|16 length-1| - * +-----------+----------+--------+------------+------------------+ - * - * The 'data' section would look like (in bytes): - * +--------------+------------+-----------------------------------+ - * | event_logged | sequence # | error log | - * |0 3|4 7|8 error_log_size-1| - * +--------------+------------+-----------------------------------+ - * - * event_logged: 0 if event has not been logged to syslog, 1 if it has - * sequence #: The unique sequence # for each event. (until it wraps) - * error log: The error log from event_scan + * the failure information to decode. */ -static int nvram_write_os_partition(struct nvram_os_partition *part, - char *buff, int length, - unsigned int err_type, - unsigned int error_log_cnt) -{ - int rc; - loff_t tmp_index; - struct err_log_info info; - - if (part->index == -1) { - return -ESPIPE; - } - - if (length > part->size) { - length = part->size; - } - - info.error_type = cpu_to_be32(err_type); - info.seq_num = cpu_to_be32(error_log_cnt); - - tmp_index = part->index; - - rc = ppc_md.nvram_write((char *)&info, sizeof(struct err_log_info), &tmp_index); - if (rc <= 0) { - pr_err("%s: Failed nvram_write (%d)\n", __func__, rc); - return rc; - } - - rc = ppc_md.nvram_write(buff, length, &tmp_index); - if (rc <= 0) { - pr_err("%s: Failed nvram_write (%d)\n", __func__, rc); - return rc; - } - - return 0; -} - int nvram_write_error_log(char * buff, int length, unsigned int err_type, unsigned int error_log_cnt) { @@ -328,50 +153,6 @@ int nvram_write_error_log(char * buff, int length, return rc; } -/* nvram_read_partition - * - * Reads nvram partition for at most 'length' - */ -static int nvram_read_partition(struct nvram_os_partition *part, char *buff, - int length, unsigned int *err_type, - unsigned int *error_log_cnt) -{ - int rc; - loff_t tmp_index; - struct err_log_info info; - - if (part->index == -1) - return -1; - - if (length > part->size) - length = part->size; - - tmp_index = part->index; - - if (part->os_partition) { - rc = ppc_md.nvram_read((char *)&info, - sizeof(struct err_log_info), - &tmp_index); - if (rc <= 0) { - pr_err("%s: Failed nvram_read (%d)\n", __func__, rc); - return rc; - } - } - - rc = ppc_md.nvram_read(buff, length, &tmp_index); - if (rc <= 0) { - pr_err("%s: Failed nvram_read (%d)\n", __func__, rc); - return rc; - } - - if (part->os_partition) { - *error_log_cnt = be32_to_cpu(info.seq_num); - *err_type = be32_to_cpu(info.error_type); - } - - return 0; -} - /* nvram_read_error_log * * Reads nvram for error log for at most 'length' @@ -407,67 +188,6 @@ int nvram_clear_error_log(void) return 0; } -/* pseries_nvram_init_os_partition - * - * This sets up a partition with an "OS" signature. - * - * The general strategy is the following: - * 1.) If a partition with the indicated name already exists... - * - If it's large enough, use it. - * - Otherwise, recycle it and keep going. - * 2.) Search for a free partition that is large enough. - * 3.) If there's not a free partition large enough, recycle any obsolete - * OS partitions and try again. - * 4.) Will first try getting a chunk that will satisfy the requested size. - * 5.) If a chunk of the requested size cannot be allocated, then try finding - * a chunk that will satisfy the minum needed. - * - * Returns 0 on success, else -1. - */ -static int __init pseries_nvram_init_os_partition(struct nvram_os_partition - *part) -{ - loff_t p; - int size; - - /* Look for ours */ - p = nvram_find_partition(part->name, NVRAM_SIG_OS, &size); - - /* Found one but too small, remove it */ - if (p && size < part->min_size) { - pr_info("nvram: Found too small %s partition," - " removing it...\n", part->name); - nvram_remove_partition(part->name, NVRAM_SIG_OS, NULL); - p = 0; - } - - /* Create one if we didn't find */ - if (!p) { - p = nvram_create_partition(part->name, NVRAM_SIG_OS, - part->req_size, part->min_size); - if (p == -ENOSPC) { - pr_info("nvram: No room to create %s partition, " - "deleting any obsolete OS partitions...\n", - part->name); - nvram_remove_partition(NULL, NVRAM_SIG_OS, - pseries_nvram_os_partitions); - p = nvram_create_partition(part->name, NVRAM_SIG_OS, - part->req_size, part->min_size); - } - } - - if (p <= 0) { - pr_err("nvram: Failed to find or create %s" - " partition, err %d\n", part->name, (int)p); - return -1; - } - - part->index = p; - part->size = nvram_get_partition_size(p) - sizeof(struct err_log_info); - - return 0; -} - /* * Are we using the ibm,rtas-log for oops/panic reports? And if so, * would logging this oops/panic overwrite an RTAS event that rtas_errd @@ -476,7 +196,7 @@ static int __init pseries_nvram_init_os_partition(struct nvram_os_partition * We assume that if rtas_errd hasn't read the RTAS event in * NVRAM_RTAS_READ_TIMEOUT seconds, it's probably not going to. */ -static int clobbering_unread_rtas_event(void) +int clobbering_unread_rtas_event(void) { return (oops_log_partition.index == rtas_log_partition.index && last_unread_rtas_event @@ -484,313 +204,6 @@ static int clobbering_unread_rtas_event(void) NVRAM_RTAS_READ_TIMEOUT); } -/* Derived from logfs_compress() */ -static int nvram_compress(const void *in, void *out, size_t inlen, - size_t outlen) -{ - int err, ret; - - ret = -EIO; - err = zlib_deflateInit2(&stream, COMPR_LEVEL, Z_DEFLATED, WINDOW_BITS, - MEM_LEVEL, Z_DEFAULT_STRATEGY); - if (err != Z_OK) - goto error; - - stream.next_in = in; - stream.avail_in = inlen; - stream.total_in = 0; - stream.next_out = out; - stream.avail_out = outlen; - stream.total_out = 0; - - err = zlib_deflate(&stream, Z_FINISH); - if (err != Z_STREAM_END) - goto error; - - err = zlib_deflateEnd(&stream); - if (err != Z_OK) - goto error; - - if (stream.total_out >= stream.total_in) - goto error; - - ret = stream.total_out; -error: - return ret; -} - -/* Compress the text from big_oops_buf into oops_buf. */ -static int zip_oops(size_t text_len) -{ - struct oops_log_info *oops_hdr = (struct oops_log_info *)oops_buf; - int zipped_len = nvram_compress(big_oops_buf, oops_data, text_len, - oops_data_sz); - if (zipped_len < 0) { - pr_err("nvram: compression failed; returned %d\n", zipped_len); - pr_err("nvram: logging uncompressed oops/panic report\n"); - return -1; - } - oops_hdr->version = cpu_to_be16(OOPS_HDR_VERSION); - oops_hdr->report_length = cpu_to_be16(zipped_len); - oops_hdr->timestamp = cpu_to_be64(get_seconds()); - return 0; -} - -#ifdef CONFIG_PSTORE -static int nvram_pstore_open(struct pstore_info *psi) -{ - /* Reset the iterator to start reading partitions again */ - read_type = -1; - return 0; -} - -/** - * nvram_pstore_write - pstore write callback for nvram - * @type: Type of message logged - * @reason: reason behind dump (oops/panic) - * @id: identifier to indicate the write performed - * @part: pstore writes data to registered buffer in parts, - * part number will indicate the same. - * @count: Indicates oops count - * @compressed: Flag to indicate the log is compressed - * @size: number of bytes written to the registered buffer - * @psi: registered pstore_info structure - * - * Called by pstore_dump() when an oops or panic report is logged in the - * printk buffer. - * Returns 0 on successful write. - */ -static int nvram_pstore_write(enum pstore_type_id type, - enum kmsg_dump_reason reason, - u64 *id, unsigned int part, int count, - bool compressed, size_t size, - struct pstore_info *psi) -{ - int rc; - unsigned int err_type = ERR_TYPE_KERNEL_PANIC; - struct oops_log_info *oops_hdr = (struct oops_log_info *) oops_buf; - - /* part 1 has the recent messages from printk buffer */ - if (part > 1 || type != PSTORE_TYPE_DMESG || - clobbering_unread_rtas_event()) - return -1; - - oops_hdr->version = cpu_to_be16(OOPS_HDR_VERSION); - oops_hdr->report_length = cpu_to_be16(size); - oops_hdr->timestamp = cpu_to_be64(get_seconds()); - - if (compressed) - err_type = ERR_TYPE_KERNEL_PANIC_GZ; - - rc = nvram_write_os_partition(&oops_log_partition, oops_buf, - (int) (sizeof(*oops_hdr) + size), err_type, count); - - if (rc != 0) - return rc; - - *id = part; - return 0; -} - -/* - * Reads the oops/panic report, rtas, of-config and common partition. - * Returns the length of the data we read from each partition. - * Returns 0 if we've been called before. - */ -static ssize_t nvram_pstore_read(u64 *id, enum pstore_type_id *type, - int *count, struct timespec *time, char **buf, - bool *compressed, struct pstore_info *psi) -{ - struct oops_log_info *oops_hdr; - unsigned int err_type, id_no, size = 0; - struct nvram_os_partition *part = NULL; - char *buff = NULL; - int sig = 0; - loff_t p; - - read_type++; - - switch (nvram_type_ids[read_type]) { - case PSTORE_TYPE_DMESG: - part = &oops_log_partition; - *type = PSTORE_TYPE_DMESG; - break; - case PSTORE_TYPE_PPC_RTAS: - part = &rtas_log_partition; - *type = PSTORE_TYPE_PPC_RTAS; - time->tv_sec = last_rtas_event; - time->tv_nsec = 0; - break; - case PSTORE_TYPE_PPC_OF: - sig = NVRAM_SIG_OF; - part = &of_config_partition; - *type = PSTORE_TYPE_PPC_OF; - *id = PSTORE_TYPE_PPC_OF; - time->tv_sec = 0; - time->tv_nsec = 0; - break; - case PSTORE_TYPE_PPC_COMMON: - sig = NVRAM_SIG_SYS; - part = &common_partition; - *type = PSTORE_TYPE_PPC_COMMON; - *id = PSTORE_TYPE_PPC_COMMON; - time->tv_sec = 0; - time->tv_nsec = 0; - break; - default: - return 0; - } - - if (!part->os_partition) { - p = nvram_find_partition(part->name, sig, &size); - if (p <= 0) { - pr_err("nvram: Failed to find partition %s, " - "err %d\n", part->name, (int)p); - return 0; - } - part->index = p; - part->size = size; - } - - buff = kmalloc(part->size, GFP_KERNEL); - - if (!buff) - return -ENOMEM; - - if (nvram_read_partition(part, buff, part->size, &err_type, &id_no)) { - kfree(buff); - return 0; - } - - *count = 0; - - if (part->os_partition) - *id = id_no; - - if (nvram_type_ids[read_type] == PSTORE_TYPE_DMESG) { - size_t length, hdr_size; - - oops_hdr = (struct oops_log_info *)buff; - if (be16_to_cpu(oops_hdr->version) < OOPS_HDR_VERSION) { - /* Old format oops header had 2-byte record size */ - hdr_size = sizeof(u16); - length = be16_to_cpu(oops_hdr->version); - time->tv_sec = 0; - time->tv_nsec = 0; - } else { - hdr_size = sizeof(*oops_hdr); - length = be16_to_cpu(oops_hdr->report_length); - time->tv_sec = be64_to_cpu(oops_hdr->timestamp); - time->tv_nsec = 0; - } - *buf = kmalloc(length, GFP_KERNEL); - if (*buf == NULL) - return -ENOMEM; - memcpy(*buf, buff + hdr_size, length); - kfree(buff); - - if (err_type == ERR_TYPE_KERNEL_PANIC_GZ) - *compressed = true; - else - *compressed = false; - return length; - } - - *buf = buff; - return part->size; -} - -static struct pstore_info nvram_pstore_info = { - .owner = THIS_MODULE, - .name = "nvram", - .open = nvram_pstore_open, - .read = nvram_pstore_read, - .write = nvram_pstore_write, -}; - -static int nvram_pstore_init(void) -{ - int rc = 0; - - nvram_pstore_info.buf = oops_data; - nvram_pstore_info.bufsize = oops_data_sz; - - spin_lock_init(&nvram_pstore_info.buf_lock); - - rc = pstore_register(&nvram_pstore_info); - if (rc != 0) - pr_err("nvram: pstore_register() failed, defaults to " - "kmsg_dump; returned %d\n", rc); - - return rc; -} -#else -static int nvram_pstore_init(void) -{ - return -1; -} -#endif - -static void __init nvram_init_oops_partition(int rtas_partition_exists) -{ - int rc; - - rc = pseries_nvram_init_os_partition(&oops_log_partition); - if (rc != 0) { - if (!rtas_partition_exists) - return; - pr_notice("nvram: Using %s partition to log both" - " RTAS errors and oops/panic reports\n", - rtas_log_partition.name); - memcpy(&oops_log_partition, &rtas_log_partition, - sizeof(rtas_log_partition)); - } - oops_buf = kmalloc(oops_log_partition.size, GFP_KERNEL); - if (!oops_buf) { - pr_err("nvram: No memory for %s partition\n", - oops_log_partition.name); - return; - } - oops_data = oops_buf + sizeof(struct oops_log_info); - oops_data_sz = oops_log_partition.size - sizeof(struct oops_log_info); - - rc = nvram_pstore_init(); - - if (!rc) - return; - - /* - * Figure compression (preceded by elimination of each line's - * severity prefix) will reduce the oops/panic report to at most - * 45% of its original size. - */ - big_oops_buf_sz = (oops_data_sz * 100) / 45; - big_oops_buf = kmalloc(big_oops_buf_sz, GFP_KERNEL); - if (big_oops_buf) { - stream.workspace = kmalloc(zlib_deflate_workspacesize( - WINDOW_BITS, MEM_LEVEL), GFP_KERNEL); - if (!stream.workspace) { - pr_err("nvram: No memory for compression workspace; " - "skipping compression of %s partition data\n", - oops_log_partition.name); - kfree(big_oops_buf); - big_oops_buf = NULL; - } - } else { - pr_err("No memory for uncompressed %s data; " - "skipping compression\n", oops_log_partition.name); - stream.workspace = NULL; - } - - rc = kmsg_dump_register(&nvram_kmsg_dumper); - if (rc != 0) { - pr_err("nvram: kmsg_dump_register() failed; returned %d\n", rc); - kfree(oops_buf); - kfree(big_oops_buf); - kfree(stream.workspace); - } -} - static int __init pseries_nvram_init_log_partitions(void) { int rc; @@ -798,7 +211,7 @@ static int __init pseries_nvram_init_log_partitions(void) /* Scan nvram for partitions */ nvram_scan_partitions(); - rc = pseries_nvram_init_os_partition(&rtas_log_partition); + rc = nvram_init_os_partition(&rtas_log_partition); nvram_init_oops_partition(rc == 0); return 0; } @@ -834,72 +247,3 @@ int __init pSeries_nvram_init(void) return 0; } - -/* - * This is our kmsg_dump callback, called after an oops or panic report - * has been written to the printk buffer. We want to capture as much - * of the printk buffer as possible. First, capture as much as we can - * that we think will compress sufficiently to fit in the lnx,oops-log - * partition. If that's too much, go back and capture uncompressed text. - */ -static void oops_to_nvram(struct kmsg_dumper *dumper, - enum kmsg_dump_reason reason) -{ - struct oops_log_info *oops_hdr = (struct oops_log_info *)oops_buf; - static unsigned int oops_count = 0; - static bool panicking = false; - static DEFINE_SPINLOCK(lock); - unsigned long flags; - size_t text_len; - unsigned int err_type = ERR_TYPE_KERNEL_PANIC_GZ; - int rc = -1; - - switch (reason) { - case KMSG_DUMP_RESTART: - case KMSG_DUMP_HALT: - case KMSG_DUMP_POWEROFF: - /* These are almost always orderly shutdowns. */ - return; - case KMSG_DUMP_OOPS: - break; - case KMSG_DUMP_PANIC: - panicking = true; - break; - case KMSG_DUMP_EMERG: - if (panicking) - /* Panic report already captured. */ - return; - break; - default: - pr_err("%s: ignoring unrecognized KMSG_DUMP_* reason %d\n", - __func__, (int) reason); - return; - } - - if (clobbering_unread_rtas_event()) - return; - - if (!spin_trylock_irqsave(&lock, flags)) - return; - - if (big_oops_buf) { - kmsg_dump_get_buffer(dumper, false, - big_oops_buf, big_oops_buf_sz, &text_len); - rc = zip_oops(text_len); - } - if (rc != 0) { - kmsg_dump_rewind(dumper); - kmsg_dump_get_buffer(dumper, false, - oops_data, oops_data_sz, &text_len); - err_type = ERR_TYPE_KERNEL_PANIC; - oops_hdr->version = cpu_to_be16(OOPS_HDR_VERSION); - oops_hdr->report_length = cpu_to_be16(text_len); - oops_hdr->timestamp = cpu_to_be64(get_seconds()); - } - - (void) nvram_write_os_partition(&oops_log_partition, oops_buf, - (int) (sizeof(*oops_hdr) + text_len), err_type, - ++oops_count); - - spin_unlock_irqrestore(&lock, flags); -}