linux/fs/pstore/ram_core.c
Rob Herring 7ae9cb8193 pstore-ram: Fix hangs by using write-combine mappings
Currently trying to use pstore on at least ARMs can hang as we're
mapping the peristent RAM with pgprot_noncached().

On ARMs, pgprot_noncached() will actually make the memory strongly
ordered, and as the atomic operations pstore uses are implementation
defined for strongly ordered memory, they may not work. So basically
atomic operations have undefined behavior on ARM for device or strongly
ordered memory types.

Let's fix the issue by using write-combine variants for mappings. This
corresponds to normal, non-cacheable memory on ARM. For many other
architectures, this change does not change the mapping type as by
default we have:

#define pgprot_writecombine pgprot_noncached

The reason why pgprot_noncached() was originaly used for pstore
is because Colin Cross <ccross@android.com> had observed lost
debug prints right before a device hanging write operation on some
systems. For the platforms supporting pgprot_noncached(), we can
add a an optional configuration option to support that. But let's
get pstore working first before adding new features.

Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Anton Vorontsov <cbouatmailru@gmail.com>
Cc: Colin Cross <ccross@android.com>
Cc: Olof Johansson <olof@lixom.net>
Cc: linux-kernel@vger.kernel.org
Cc: stable@vger.kernel.org
Acked-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Rob Herring <rob.herring@calxeda.com>
[tony@atomide.com: updated description]
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
2014-12-11 13:35:49 -08:00

527 lines
13 KiB
C

/*
* Copyright (C) 2012 Google, Inc.
*
* This software is licensed under the terms of the GNU General Public
* License version 2, as published by the Free Software Foundation, and
* may be copied, distributed, and modified under those terms.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
*/
#define pr_fmt(fmt) "persistent_ram: " fmt
#include <linux/device.h>
#include <linux/err.h>
#include <linux/errno.h>
#include <linux/kernel.h>
#include <linux/init.h>
#include <linux/io.h>
#include <linux/list.h>
#include <linux/memblock.h>
#include <linux/rslib.h>
#include <linux/slab.h>
#include <linux/vmalloc.h>
#include <linux/pstore_ram.h>
#include <asm/page.h>
struct persistent_ram_buffer {
uint32_t sig;
atomic_t start;
atomic_t size;
uint8_t data[0];
};
#define PERSISTENT_RAM_SIG (0x43474244) /* DBGC */
static inline size_t buffer_size(struct persistent_ram_zone *prz)
{
return atomic_read(&prz->buffer->size);
}
static inline size_t buffer_start(struct persistent_ram_zone *prz)
{
return atomic_read(&prz->buffer->start);
}
/* increase and wrap the start pointer, returning the old value */
static size_t buffer_start_add_atomic(struct persistent_ram_zone *prz, size_t a)
{
int old;
int new;
do {
old = atomic_read(&prz->buffer->start);
new = old + a;
while (unlikely(new >= prz->buffer_size))
new -= prz->buffer_size;
} while (atomic_cmpxchg(&prz->buffer->start, old, new) != old);
return old;
}
/* increase the size counter until it hits the max size */
static void buffer_size_add_atomic(struct persistent_ram_zone *prz, size_t a)
{
size_t old;
size_t new;
if (atomic_read(&prz->buffer->size) == prz->buffer_size)
return;
do {
old = atomic_read(&prz->buffer->size);
new = old + a;
if (new > prz->buffer_size)
new = prz->buffer_size;
} while (atomic_cmpxchg(&prz->buffer->size, old, new) != old);
}
static DEFINE_RAW_SPINLOCK(buffer_lock);
/* increase and wrap the start pointer, returning the old value */
static size_t buffer_start_add_locked(struct persistent_ram_zone *prz, size_t a)
{
int old;
int new;
unsigned long flags;
raw_spin_lock_irqsave(&buffer_lock, flags);
old = atomic_read(&prz->buffer->start);
new = old + a;
while (unlikely(new >= prz->buffer_size))
new -= prz->buffer_size;
atomic_set(&prz->buffer->start, new);
raw_spin_unlock_irqrestore(&buffer_lock, flags);
return old;
}
/* increase the size counter until it hits the max size */
static void buffer_size_add_locked(struct persistent_ram_zone *prz, size_t a)
{
size_t old;
size_t new;
unsigned long flags;
raw_spin_lock_irqsave(&buffer_lock, flags);
old = atomic_read(&prz->buffer->size);
if (old == prz->buffer_size)
goto exit;
new = old + a;
if (new > prz->buffer_size)
new = prz->buffer_size;
atomic_set(&prz->buffer->size, new);
exit:
raw_spin_unlock_irqrestore(&buffer_lock, flags);
}
static size_t (*buffer_start_add)(struct persistent_ram_zone *, size_t) = buffer_start_add_atomic;
static void (*buffer_size_add)(struct persistent_ram_zone *, size_t) = buffer_size_add_atomic;
static void notrace persistent_ram_encode_rs8(struct persistent_ram_zone *prz,
uint8_t *data, size_t len, uint8_t *ecc)
{
int i;
uint16_t par[prz->ecc_info.ecc_size];
/* Initialize the parity buffer */
memset(par, 0, sizeof(par));
encode_rs8(prz->rs_decoder, data, len, par, 0);
for (i = 0; i < prz->ecc_info.ecc_size; i++)
ecc[i] = par[i];
}
static int persistent_ram_decode_rs8(struct persistent_ram_zone *prz,
void *data, size_t len, uint8_t *ecc)
{
int i;
uint16_t par[prz->ecc_info.ecc_size];
for (i = 0; i < prz->ecc_info.ecc_size; i++)
par[i] = ecc[i];
return decode_rs8(prz->rs_decoder, data, par, len,
NULL, 0, NULL, 0, NULL);
}
static void notrace persistent_ram_update_ecc(struct persistent_ram_zone *prz,
unsigned int start, unsigned int count)
{
struct persistent_ram_buffer *buffer = prz->buffer;
uint8_t *buffer_end = buffer->data + prz->buffer_size;
uint8_t *block;
uint8_t *par;
int ecc_block_size = prz->ecc_info.block_size;
int ecc_size = prz->ecc_info.ecc_size;
int size = ecc_block_size;
if (!ecc_size)
return;
block = buffer->data + (start & ~(ecc_block_size - 1));
par = prz->par_buffer + (start / ecc_block_size) * ecc_size;
do {
if (block + ecc_block_size > buffer_end)
size = buffer_end - block;
persistent_ram_encode_rs8(prz, block, size, par);
block += ecc_block_size;
par += ecc_size;
} while (block < buffer->data + start + count);
}
static void persistent_ram_update_header_ecc(struct persistent_ram_zone *prz)
{
struct persistent_ram_buffer *buffer = prz->buffer;
if (!prz->ecc_info.ecc_size)
return;
persistent_ram_encode_rs8(prz, (uint8_t *)buffer, sizeof(*buffer),
prz->par_header);
}
static void persistent_ram_ecc_old(struct persistent_ram_zone *prz)
{
struct persistent_ram_buffer *buffer = prz->buffer;
uint8_t *block;
uint8_t *par;
if (!prz->ecc_info.ecc_size)
return;
block = buffer->data;
par = prz->par_buffer;
while (block < buffer->data + buffer_size(prz)) {
int numerr;
int size = prz->ecc_info.block_size;
if (block + size > buffer->data + prz->buffer_size)
size = buffer->data + prz->buffer_size - block;
numerr = persistent_ram_decode_rs8(prz, block, size, par);
if (numerr > 0) {
pr_devel("error in block %p, %d\n", block, numerr);
prz->corrected_bytes += numerr;
} else if (numerr < 0) {
pr_devel("uncorrectable error in block %p\n", block);
prz->bad_blocks++;
}
block += prz->ecc_info.block_size;
par += prz->ecc_info.ecc_size;
}
}
static int persistent_ram_init_ecc(struct persistent_ram_zone *prz,
struct persistent_ram_ecc_info *ecc_info)
{
int numerr;
struct persistent_ram_buffer *buffer = prz->buffer;
int ecc_blocks;
size_t ecc_total;
if (!ecc_info || !ecc_info->ecc_size)
return 0;
prz->ecc_info.block_size = ecc_info->block_size ?: 128;
prz->ecc_info.ecc_size = ecc_info->ecc_size ?: 16;
prz->ecc_info.symsize = ecc_info->symsize ?: 8;
prz->ecc_info.poly = ecc_info->poly ?: 0x11d;
ecc_blocks = DIV_ROUND_UP(prz->buffer_size - prz->ecc_info.ecc_size,
prz->ecc_info.block_size +
prz->ecc_info.ecc_size);
ecc_total = (ecc_blocks + 1) * prz->ecc_info.ecc_size;
if (ecc_total >= prz->buffer_size) {
pr_err("%s: invalid ecc_size %u (total %zu, buffer size %zu)\n",
__func__, prz->ecc_info.ecc_size,
ecc_total, prz->buffer_size);
return -EINVAL;
}
prz->buffer_size -= ecc_total;
prz->par_buffer = buffer->data + prz->buffer_size;
prz->par_header = prz->par_buffer +
ecc_blocks * prz->ecc_info.ecc_size;
/*
* first consecutive root is 0
* primitive element to generate roots = 1
*/
prz->rs_decoder = init_rs(prz->ecc_info.symsize, prz->ecc_info.poly,
0, 1, prz->ecc_info.ecc_size);
if (prz->rs_decoder == NULL) {
pr_info("init_rs failed\n");
return -EINVAL;
}
prz->corrected_bytes = 0;
prz->bad_blocks = 0;
numerr = persistent_ram_decode_rs8(prz, buffer, sizeof(*buffer),
prz->par_header);
if (numerr > 0) {
pr_info("error in header, %d\n", numerr);
prz->corrected_bytes += numerr;
} else if (numerr < 0) {
pr_info("uncorrectable error in header\n");
prz->bad_blocks++;
}
return 0;
}
ssize_t persistent_ram_ecc_string(struct persistent_ram_zone *prz,
char *str, size_t len)
{
ssize_t ret;
if (!prz->ecc_info.ecc_size)
return 0;
if (prz->corrected_bytes || prz->bad_blocks)
ret = snprintf(str, len, ""
"\n%d Corrected bytes, %d unrecoverable blocks\n",
prz->corrected_bytes, prz->bad_blocks);
else
ret = snprintf(str, len, "\nNo errors detected\n");
return ret;
}
static void notrace persistent_ram_update(struct persistent_ram_zone *prz,
const void *s, unsigned int start, unsigned int count)
{
struct persistent_ram_buffer *buffer = prz->buffer;
memcpy(buffer->data + start, s, count);
persistent_ram_update_ecc(prz, start, count);
}
void persistent_ram_save_old(struct persistent_ram_zone *prz)
{
struct persistent_ram_buffer *buffer = prz->buffer;
size_t size = buffer_size(prz);
size_t start = buffer_start(prz);
if (!size)
return;
if (!prz->old_log) {
persistent_ram_ecc_old(prz);
prz->old_log = kmalloc(size, GFP_KERNEL);
}
if (!prz->old_log) {
pr_err("failed to allocate buffer\n");
return;
}
prz->old_log_size = size;
memcpy(prz->old_log, &buffer->data[start], size - start);
memcpy(prz->old_log + size - start, &buffer->data[0], start);
}
int notrace persistent_ram_write(struct persistent_ram_zone *prz,
const void *s, unsigned int count)
{
int rem;
int c = count;
size_t start;
if (unlikely(c > prz->buffer_size)) {
s += c - prz->buffer_size;
c = prz->buffer_size;
}
buffer_size_add(prz, c);
start = buffer_start_add(prz, c);
rem = prz->buffer_size - start;
if (unlikely(rem < c)) {
persistent_ram_update(prz, s, start, rem);
s += rem;
c -= rem;
start = 0;
}
persistent_ram_update(prz, s, start, c);
persistent_ram_update_header_ecc(prz);
return count;
}
size_t persistent_ram_old_size(struct persistent_ram_zone *prz)
{
return prz->old_log_size;
}
void *persistent_ram_old(struct persistent_ram_zone *prz)
{
return prz->old_log;
}
void persistent_ram_free_old(struct persistent_ram_zone *prz)
{
kfree(prz->old_log);
prz->old_log = NULL;
prz->old_log_size = 0;
}
void persistent_ram_zap(struct persistent_ram_zone *prz)
{
atomic_set(&prz->buffer->start, 0);
atomic_set(&prz->buffer->size, 0);
persistent_ram_update_header_ecc(prz);
}
static void *persistent_ram_vmap(phys_addr_t start, size_t size)
{
struct page **pages;
phys_addr_t page_start;
unsigned int page_count;
pgprot_t prot;
unsigned int i;
void *vaddr;
page_start = start - offset_in_page(start);
page_count = DIV_ROUND_UP(size + offset_in_page(start), PAGE_SIZE);
prot = pgprot_writecombine(PAGE_KERNEL);
pages = kmalloc_array(page_count, sizeof(struct page *), GFP_KERNEL);
if (!pages) {
pr_err("%s: Failed to allocate array for %u pages\n",
__func__, page_count);
return NULL;
}
for (i = 0; i < page_count; i++) {
phys_addr_t addr = page_start + i * PAGE_SIZE;
pages[i] = pfn_to_page(addr >> PAGE_SHIFT);
}
vaddr = vmap(pages, page_count, VM_MAP, prot);
kfree(pages);
return vaddr;
}
static void *persistent_ram_iomap(phys_addr_t start, size_t size)
{
if (!request_mem_region(start, size, "persistent_ram")) {
pr_err("request mem region (0x%llx@0x%llx) failed\n",
(unsigned long long)size, (unsigned long long)start);
return NULL;
}
buffer_start_add = buffer_start_add_locked;
buffer_size_add = buffer_size_add_locked;
return ioremap_wc(start, size);
}
static int persistent_ram_buffer_map(phys_addr_t start, phys_addr_t size,
struct persistent_ram_zone *prz)
{
prz->paddr = start;
prz->size = size;
if (pfn_valid(start >> PAGE_SHIFT))
prz->vaddr = persistent_ram_vmap(start, size);
else
prz->vaddr = persistent_ram_iomap(start, size);
if (!prz->vaddr) {
pr_err("%s: Failed to map 0x%llx pages at 0x%llx\n", __func__,
(unsigned long long)size, (unsigned long long)start);
return -ENOMEM;
}
prz->buffer = prz->vaddr + offset_in_page(start);
prz->buffer_size = size - sizeof(struct persistent_ram_buffer);
return 0;
}
static int persistent_ram_post_init(struct persistent_ram_zone *prz, u32 sig,
struct persistent_ram_ecc_info *ecc_info)
{
int ret;
ret = persistent_ram_init_ecc(prz, ecc_info);
if (ret)
return ret;
sig ^= PERSISTENT_RAM_SIG;
if (prz->buffer->sig == sig) {
if (buffer_size(prz) > prz->buffer_size ||
buffer_start(prz) > buffer_size(prz))
pr_info("found existing invalid buffer, size %zu, start %zu\n",
buffer_size(prz), buffer_start(prz));
else {
pr_debug("found existing buffer, size %zu, start %zu\n",
buffer_size(prz), buffer_start(prz));
persistent_ram_save_old(prz);
return 0;
}
} else {
pr_debug("no valid data in buffer (sig = 0x%08x)\n",
prz->buffer->sig);
}
prz->buffer->sig = sig;
persistent_ram_zap(prz);
return 0;
}
void persistent_ram_free(struct persistent_ram_zone *prz)
{
if (!prz)
return;
if (prz->vaddr) {
if (pfn_valid(prz->paddr >> PAGE_SHIFT)) {
vunmap(prz->vaddr);
} else {
iounmap(prz->vaddr);
release_mem_region(prz->paddr, prz->size);
}
prz->vaddr = NULL;
}
persistent_ram_free_old(prz);
kfree(prz);
}
struct persistent_ram_zone *persistent_ram_new(phys_addr_t start, size_t size,
u32 sig, struct persistent_ram_ecc_info *ecc_info)
{
struct persistent_ram_zone *prz;
int ret = -ENOMEM;
prz = kzalloc(sizeof(struct persistent_ram_zone), GFP_KERNEL);
if (!prz) {
pr_err("failed to allocate persistent ram zone\n");
goto err;
}
ret = persistent_ram_buffer_map(start, size, prz);
if (ret)
goto err;
ret = persistent_ram_post_init(prz, sig, ecc_info);
if (ret)
goto err;
return prz;
err:
persistent_ram_free(prz);
return ERR_PTR(ret);
}