linux/fs/jffs2/read.c

217 lines
6.9 KiB
C
Raw Normal View History

/*
* JFFS2 -- Journalling Flash File System, Version 2.
*
* Copyright © 2001-2007 Red Hat, Inc.
*
* Created by David Woodhouse <dwmw2@infradead.org>
*
* For licensing information, see the file 'LICENCE' in this directory.
*
*/
#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/crc32.h>
#include <linux/pagemap.h>
#include <linux/mtd/mtd.h>
#include <linux/compiler.h>
#include "nodelist.h"
#include "compr.h"
int jffs2_read_dnode(struct jffs2_sb_info *c, struct jffs2_inode_info *f,
struct jffs2_full_dnode *fd, unsigned char *buf,
int ofs, int len)
{
struct jffs2_raw_inode *ri;
size_t readlen;
uint32_t crc;
unsigned char *decomprbuf = NULL;
unsigned char *readbuf = NULL;
int ret = 0;
ri = jffs2_alloc_raw_inode();
if (!ri)
return -ENOMEM;
ret = jffs2_flash_read(c, ref_offset(fd->raw), sizeof(*ri), &readlen, (char *)ri);
if (ret) {
jffs2_free_raw_inode(ri);
printk(KERN_WARNING "Error reading node from 0x%08x: %d\n", ref_offset(fd->raw), ret);
return ret;
}
if (readlen != sizeof(*ri)) {
jffs2_free_raw_inode(ri);
printk(KERN_WARNING "Short read from 0x%08x: wanted 0x%zx bytes, got 0x%zx\n",
ref_offset(fd->raw), sizeof(*ri), readlen);
return -EIO;
}
crc = crc32(0, ri, sizeof(*ri)-8);
D1(printk(KERN_DEBUG "Node read from %08x: node_crc %08x, calculated CRC %08x. dsize %x, csize %x, offset %x, buf %p\n",
ref_offset(fd->raw), je32_to_cpu(ri->node_crc),
crc, je32_to_cpu(ri->dsize), je32_to_cpu(ri->csize),
je32_to_cpu(ri->offset), buf));
if (crc != je32_to_cpu(ri->node_crc)) {
printk(KERN_WARNING "Node CRC %08x != calculated CRC %08x for node at %08x\n",
je32_to_cpu(ri->node_crc), crc, ref_offset(fd->raw));
ret = -EIO;
goto out_ri;
}
/* There was a bug where we wrote hole nodes out with csize/dsize
swapped. Deal with it */
if (ri->compr == JFFS2_COMPR_ZERO && !je32_to_cpu(ri->dsize) &&
je32_to_cpu(ri->csize)) {
ri->dsize = ri->csize;
ri->csize = cpu_to_je32(0);
}
D1(if(ofs + len > je32_to_cpu(ri->dsize)) {
printk(KERN_WARNING "jffs2_read_dnode() asked for %d bytes at %d from %d-byte node\n",
len, ofs, je32_to_cpu(ri->dsize));
ret = -EINVAL;
goto out_ri;
});
if (ri->compr == JFFS2_COMPR_ZERO) {
memset(buf, 0, len);
goto out_ri;
}
/* Cases:
Reading whole node and it's uncompressed - read directly to buffer provided, check CRC.
Reading whole node and it's compressed - read into comprbuf, check CRC and decompress to buffer provided
Reading partial node and it's uncompressed - read into readbuf, check CRC, and copy
Reading partial node and it's compressed - read into readbuf, check checksum, decompress to decomprbuf and copy
*/
if (ri->compr == JFFS2_COMPR_NONE && len == je32_to_cpu(ri->dsize)) {
readbuf = buf;
} else {
readbuf = kmalloc(je32_to_cpu(ri->csize), GFP_KERNEL);
if (!readbuf) {
ret = -ENOMEM;
goto out_ri;
}
}
if (ri->compr != JFFS2_COMPR_NONE) {
if (len < je32_to_cpu(ri->dsize)) {
decomprbuf = kmalloc(je32_to_cpu(ri->dsize), GFP_KERNEL);
if (!decomprbuf) {
ret = -ENOMEM;
goto out_readbuf;
}
} else {
decomprbuf = buf;
}
} else {
decomprbuf = readbuf;
}
D2(printk(KERN_DEBUG "Read %d bytes to %p\n", je32_to_cpu(ri->csize),
readbuf));
ret = jffs2_flash_read(c, (ref_offset(fd->raw)) + sizeof(*ri),
je32_to_cpu(ri->csize), &readlen, readbuf);
if (!ret && readlen != je32_to_cpu(ri->csize))
ret = -EIO;
if (ret)
goto out_decomprbuf;
crc = crc32(0, readbuf, je32_to_cpu(ri->csize));
if (crc != je32_to_cpu(ri->data_crc)) {
printk(KERN_WARNING "Data CRC %08x != calculated CRC %08x for node at %08x\n",
je32_to_cpu(ri->data_crc), crc, ref_offset(fd->raw));
ret = -EIO;
goto out_decomprbuf;
}
D2(printk(KERN_DEBUG "Data CRC matches calculated CRC %08x\n", crc));
if (ri->compr != JFFS2_COMPR_NONE) {
D2(printk(KERN_DEBUG "Decompress %d bytes from %p to %d bytes at %p\n",
je32_to_cpu(ri->csize), readbuf, je32_to_cpu(ri->dsize), decomprbuf));
ret = jffs2_decompress(c, f, ri->compr | (ri->usercompr << 8), readbuf, decomprbuf, je32_to_cpu(ri->csize), je32_to_cpu(ri->dsize));
if (ret) {
printk(KERN_WARNING "Error: jffs2_decompress returned %d\n", ret);
goto out_decomprbuf;
}
}
if (len < je32_to_cpu(ri->dsize)) {
memcpy(buf, decomprbuf+ofs, len);
}
out_decomprbuf:
if(decomprbuf != buf && decomprbuf != readbuf)
kfree(decomprbuf);
out_readbuf:
if(readbuf != buf)
kfree(readbuf);
out_ri:
jffs2_free_raw_inode(ri);
return ret;
}
int jffs2_read_inode_range(struct jffs2_sb_info *c, struct jffs2_inode_info *f,
unsigned char *buf, uint32_t offset, uint32_t len)
{
uint32_t end = offset + len;
struct jffs2_node_frag *frag;
int ret;
D1(printk(KERN_DEBUG "jffs2_read_inode_range: ino #%u, range 0x%08x-0x%08x\n",
f->inocache->ino, offset, offset+len));
frag = jffs2_lookup_node_frag(&f->fragtree, offset);
/* XXX FIXME: Where a single physical node actually shows up in two
frags, we read it twice. Don't do that. */
jffs2: Fix memory corruption in jffs2_read_inode_range() In 2.6.23 kernel, commit a32ea1e1f925399e0d81ca3f7394a44a6dafa12c ("Fix read/truncate race") fixed a race in the generic code, and as a side effect, now do_generic_file_read() can ask us to readpage() past the i_size. This seems to be correctly handled by the block routines (e.g. block_read_full_page() fills the page with zeroes in case if somebody is trying to read past the last inode's block). JFFS2 doesn't handle this; it assumes that it won't be asked to read pages which don't exist -- and thus that there will be at least _one_ valid 'frag' on the page it's being asked to read. It will fill any holes with the following memset: memset(buf, 0, min(end, frag->ofs + frag->size) - offset); When the 'closest smaller match' returned by jffs2_lookup_node_frag() is actually on a previous page and ends before 'offset', that results in: memset(buf, 0, <huge unsigned negative>); Hopefully, in most cases the corruption is fatal, and quickly causing random oopses, like this: root@10.0.0.4:~/ltp-fs-20090531# ./testcases/kernel/fs/ftest/ftest01 Unable to handle kernel paging request for data at address 0x00000008 Faulting instruction address: 0xc01cd980 Oops: Kernel access of bad area, sig: 11 [#1] [...] NIP [c01cd980] rb_insert_color+0x38/0x184 LR [c0043978] enqueue_hrtimer+0x88/0xc4 Call Trace: [c6c63b60] [c004f9a8] tick_sched_timer+0xa0/0xe4 (unreliable) [c6c63b80] [c0043978] enqueue_hrtimer+0x88/0xc4 [c6c63b90] [c0043a48] __run_hrtimer+0x94/0xbc [c6c63bb0] [c0044628] hrtimer_interrupt+0x140/0x2b8 [c6c63c10] [c000f8e8] timer_interrupt+0x13c/0x254 [c6c63c30] [c001352c] ret_from_except+0x0/0x14 --- Exception: 901 at memset+0x38/0x5c LR = jffs2_read_inode_range+0x144/0x17c [c6c63cf0] [00000000] (null) (unreliable) This patch fixes the issue, plus fixes all LTP tests on NAND/UBI with JFFS2 filesystem that were failing since 2.6.23 (seems like the bug above also broke the truncation). Reported-By: Anton Vorontsov <avorontsov@ru.mvista.com> Tested-By: Anton Vorontsov <avorontsov@ru.mvista.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com> Cc: stable@kernel.org Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-11-30 10:06:40 +01:00
/* Now we're pointing at the first frag which overlaps our page
* (or perhaps is before it, if we've been asked to read off the
* end of the file). */
while(offset < end) {
D2(printk(KERN_DEBUG "jffs2_read_inode_range: offset %d, end %d\n", offset, end));
jffs2: Fix memory corruption in jffs2_read_inode_range() In 2.6.23 kernel, commit a32ea1e1f925399e0d81ca3f7394a44a6dafa12c ("Fix read/truncate race") fixed a race in the generic code, and as a side effect, now do_generic_file_read() can ask us to readpage() past the i_size. This seems to be correctly handled by the block routines (e.g. block_read_full_page() fills the page with zeroes in case if somebody is trying to read past the last inode's block). JFFS2 doesn't handle this; it assumes that it won't be asked to read pages which don't exist -- and thus that there will be at least _one_ valid 'frag' on the page it's being asked to read. It will fill any holes with the following memset: memset(buf, 0, min(end, frag->ofs + frag->size) - offset); When the 'closest smaller match' returned by jffs2_lookup_node_frag() is actually on a previous page and ends before 'offset', that results in: memset(buf, 0, <huge unsigned negative>); Hopefully, in most cases the corruption is fatal, and quickly causing random oopses, like this: root@10.0.0.4:~/ltp-fs-20090531# ./testcases/kernel/fs/ftest/ftest01 Unable to handle kernel paging request for data at address 0x00000008 Faulting instruction address: 0xc01cd980 Oops: Kernel access of bad area, sig: 11 [#1] [...] NIP [c01cd980] rb_insert_color+0x38/0x184 LR [c0043978] enqueue_hrtimer+0x88/0xc4 Call Trace: [c6c63b60] [c004f9a8] tick_sched_timer+0xa0/0xe4 (unreliable) [c6c63b80] [c0043978] enqueue_hrtimer+0x88/0xc4 [c6c63b90] [c0043a48] __run_hrtimer+0x94/0xbc [c6c63bb0] [c0044628] hrtimer_interrupt+0x140/0x2b8 [c6c63c10] [c000f8e8] timer_interrupt+0x13c/0x254 [c6c63c30] [c001352c] ret_from_except+0x0/0x14 --- Exception: 901 at memset+0x38/0x5c LR = jffs2_read_inode_range+0x144/0x17c [c6c63cf0] [00000000] (null) (unreliable) This patch fixes the issue, plus fixes all LTP tests on NAND/UBI with JFFS2 filesystem that were failing since 2.6.23 (seems like the bug above also broke the truncation). Reported-By: Anton Vorontsov <avorontsov@ru.mvista.com> Tested-By: Anton Vorontsov <avorontsov@ru.mvista.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com> Cc: stable@kernel.org Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-11-30 10:06:40 +01:00
if (unlikely(!frag || frag->ofs > offset ||
frag->ofs + frag->size <= offset)) {
uint32_t holesize = end - offset;
jffs2: Fix memory corruption in jffs2_read_inode_range() In 2.6.23 kernel, commit a32ea1e1f925399e0d81ca3f7394a44a6dafa12c ("Fix read/truncate race") fixed a race in the generic code, and as a side effect, now do_generic_file_read() can ask us to readpage() past the i_size. This seems to be correctly handled by the block routines (e.g. block_read_full_page() fills the page with zeroes in case if somebody is trying to read past the last inode's block). JFFS2 doesn't handle this; it assumes that it won't be asked to read pages which don't exist -- and thus that there will be at least _one_ valid 'frag' on the page it's being asked to read. It will fill any holes with the following memset: memset(buf, 0, min(end, frag->ofs + frag->size) - offset); When the 'closest smaller match' returned by jffs2_lookup_node_frag() is actually on a previous page and ends before 'offset', that results in: memset(buf, 0, <huge unsigned negative>); Hopefully, in most cases the corruption is fatal, and quickly causing random oopses, like this: root@10.0.0.4:~/ltp-fs-20090531# ./testcases/kernel/fs/ftest/ftest01 Unable to handle kernel paging request for data at address 0x00000008 Faulting instruction address: 0xc01cd980 Oops: Kernel access of bad area, sig: 11 [#1] [...] NIP [c01cd980] rb_insert_color+0x38/0x184 LR [c0043978] enqueue_hrtimer+0x88/0xc4 Call Trace: [c6c63b60] [c004f9a8] tick_sched_timer+0xa0/0xe4 (unreliable) [c6c63b80] [c0043978] enqueue_hrtimer+0x88/0xc4 [c6c63b90] [c0043a48] __run_hrtimer+0x94/0xbc [c6c63bb0] [c0044628] hrtimer_interrupt+0x140/0x2b8 [c6c63c10] [c000f8e8] timer_interrupt+0x13c/0x254 [c6c63c30] [c001352c] ret_from_except+0x0/0x14 --- Exception: 901 at memset+0x38/0x5c LR = jffs2_read_inode_range+0x144/0x17c [c6c63cf0] [00000000] (null) (unreliable) This patch fixes the issue, plus fixes all LTP tests on NAND/UBI with JFFS2 filesystem that were failing since 2.6.23 (seems like the bug above also broke the truncation). Reported-By: Anton Vorontsov <avorontsov@ru.mvista.com> Tested-By: Anton Vorontsov <avorontsov@ru.mvista.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com> Cc: stable@kernel.org Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-11-30 10:06:40 +01:00
if (frag && frag->ofs > offset) {
D1(printk(KERN_NOTICE "Eep. Hole in ino #%u fraglist. frag->ofs = 0x%08x, offset = 0x%08x\n", f->inocache->ino, frag->ofs, offset));
holesize = min(holesize, frag->ofs - offset);
}
D1(printk(KERN_DEBUG "Filling non-frag hole from %d-%d\n", offset, offset+holesize));
memset(buf, 0, holesize);
buf += holesize;
offset += holesize;
continue;
} else if (unlikely(!frag->node)) {
uint32_t holeend = min(end, frag->ofs + frag->size);
D1(printk(KERN_DEBUG "Filling frag hole from %d-%d (frag 0x%x 0x%x)\n", offset, holeend, frag->ofs, frag->ofs + frag->size));
memset(buf, 0, holeend - offset);
buf += holeend - offset;
offset = holeend;
frag = frag_next(frag);
continue;
} else {
uint32_t readlen;
uint32_t fragofs; /* offset within the frag to start reading */
fragofs = offset - frag->ofs;
readlen = min(frag->size - fragofs, end - offset);
D1(printk(KERN_DEBUG "Reading %d-%d from node at 0x%08x (%d)\n",
frag->ofs+fragofs, frag->ofs+fragofs+readlen,
ref_offset(frag->node->raw), ref_flags(frag->node->raw)));
ret = jffs2_read_dnode(c, f, frag->node, buf, fragofs + frag->ofs - frag->node->ofs, readlen);
D2(printk(KERN_DEBUG "node read done\n"));
if (ret) {
D1(printk(KERN_DEBUG"jffs2_read_inode_range error %d\n",ret));
memset(buf, 0, readlen);
return ret;
}
buf += readlen;
offset += readlen;
frag = frag_next(frag);
D2(printk(KERN_DEBUG "node read was OK. Looping\n"));
}
}
return 0;
}