2006-01-16 17:50:04 +01:00
|
|
|
/*
|
|
|
|
* Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
|
2008-01-28 18:13:02 +01:00
|
|
|
* Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved.
|
2006-01-16 17:50:04 +01:00
|
|
|
*
|
|
|
|
* This copyrighted material is made available to anyone wishing to use,
|
|
|
|
* modify, copy, or redistribute it subject to the terms and conditions
|
2006-09-01 17:05:15 +02:00
|
|
|
* of the GNU General Public License version 2.
|
2006-01-16 17:50:04 +01:00
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/slab.h>
|
|
|
|
#include <linux/spinlock.h>
|
|
|
|
#include <linux/completion.h>
|
|
|
|
#include <linux/buffer_head.h>
|
2006-01-30 19:34:10 +01:00
|
|
|
#include <linux/fs.h>
|
2006-02-27 23:23:27 +01:00
|
|
|
#include <linux/gfs2_ondisk.h>
|
2008-03-11 00:17:47 +01:00
|
|
|
#include <linux/prefetch.h>
|
2009-02-09 10:25:01 +01:00
|
|
|
#include <linux/blkdev.h>
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
#include "gfs2.h"
|
2006-02-27 23:23:27 +01:00
|
|
|
#include "incore.h"
|
2006-01-16 17:50:04 +01:00
|
|
|
#include "glock.h"
|
|
|
|
#include "glops.h"
|
|
|
|
#include "lops.h"
|
|
|
|
#include "meta_io.h"
|
|
|
|
#include "quota.h"
|
|
|
|
#include "rgrp.h"
|
|
|
|
#include "super.h"
|
|
|
|
#include "trans.h"
|
2006-02-27 23:23:27 +01:00
|
|
|
#include "util.h"
|
[GFS2] flush the log if a transaction can't allocate space
This is a fix for bz #208514. When GFS2 frees up space, the freed blocks
aren't available for reuse until the resource group is successfully written
to the ondisk journal. So in rare cases, GFS2 operations will fail, saying
that the filesystem is out of space, when in reality, you are just waiting for
a log flush. For instance, on a 1Gig filesystem, if I continually write 10 Mb
to a file, and then truncate it, after a hundred interations, the write will
fail with -ENOSPC, even though the filesystem is just 1% full.
The attached patch calls a log flush in these cases. I tested this patch
fairly heavily to check if there were any locking issues that I missed, and
it seems to work just fine. Also, this patch only does the log flush if
get_local_rgrp makes a complete loop of resource groups without skipping
any do to locking issues. The code would be slightly simpler if it just always
did the log flush after the first failed pass, and you could only ever have
to go through the loop twice, instead of up to three times. However, I guessed
that failing to find a rg simply do to locking issues would be common enough
to skip the log flush in that case, but I'm not certain that this is the right
way to go. Either way, I don't suppose this code will be hit all that often.
Signed-off-by: Benjamin E. Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-03-23 21:51:56 +01:00
|
|
|
#include "log.h"
|
2007-06-08 11:05:33 +02:00
|
|
|
#include "inode.h"
|
2007-10-15 15:42:35 +02:00
|
|
|
#include "ops_address.h"
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2006-09-05 21:41:57 +02:00
|
|
|
#define BFITNOENT ((u32)~0)
|
2007-07-24 21:09:32 +02:00
|
|
|
#define NO_BLOCK ((u64)~0)
|
2006-05-18 19:52:39 +02:00
|
|
|
|
2008-03-11 00:17:47 +01:00
|
|
|
#if BITS_PER_LONG == 32
|
|
|
|
#define LBITMASK (0x55555555UL)
|
|
|
|
#define LBITSKIP55 (0x55555555UL)
|
|
|
|
#define LBITSKIP00 (0x00000000UL)
|
|
|
|
#else
|
|
|
|
#define LBITMASK (0x5555555555555555UL)
|
|
|
|
#define LBITSKIP55 (0x5555555555555555UL)
|
|
|
|
#define LBITSKIP00 (0x0000000000000000UL)
|
|
|
|
#endif
|
|
|
|
|
2006-05-18 19:52:39 +02:00
|
|
|
/*
|
|
|
|
* These routines are used by the resource group routines (rgrp.c)
|
|
|
|
* to keep track of block allocation. Each block is represented by two
|
2006-06-14 21:32:57 +02:00
|
|
|
* bits. So, each byte represents GFS2_NBBY (i.e. 4) blocks.
|
|
|
|
*
|
|
|
|
* 0 = Free
|
|
|
|
* 1 = Used (not metadata)
|
|
|
|
* 2 = Unlinked (still in use) inode
|
|
|
|
* 3 = Used (metadata)
|
2006-05-18 19:52:39 +02:00
|
|
|
*/
|
|
|
|
|
|
|
|
static const char valid_change[16] = {
|
|
|
|
/* current */
|
2006-06-14 21:32:57 +02:00
|
|
|
/* n */ 0, 1, 1, 1,
|
2006-05-18 19:52:39 +02:00
|
|
|
/* e */ 1, 0, 0, 0,
|
2006-06-14 21:32:57 +02:00
|
|
|
/* w */ 0, 0, 0, 1,
|
2006-05-18 19:52:39 +02:00
|
|
|
1, 0, 0, 0
|
|
|
|
};
|
|
|
|
|
2007-06-08 11:05:33 +02:00
|
|
|
static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal,
|
2008-02-06 11:11:15 +01:00
|
|
|
unsigned char old_state, unsigned char new_state,
|
|
|
|
unsigned int *n);
|
2007-06-08 11:05:33 +02:00
|
|
|
|
2006-05-18 19:52:39 +02:00
|
|
|
/**
|
|
|
|
* gfs2_setbit - Set a bit in the bitmaps
|
|
|
|
* @buffer: the buffer that holds the bitmaps
|
|
|
|
* @buflen: the length (in bytes) of the buffer
|
|
|
|
* @block: the block to set
|
|
|
|
* @new_state: the new state of the block
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2008-02-06 11:11:15 +01:00
|
|
|
static inline void gfs2_setbit(struct gfs2_rgrpd *rgd, unsigned char *buf1,
|
|
|
|
unsigned char *buf2, unsigned int offset,
|
|
|
|
unsigned int buflen, u32 block,
|
|
|
|
unsigned char new_state)
|
2006-05-18 19:52:39 +02:00
|
|
|
{
|
2008-02-06 11:11:15 +01:00
|
|
|
unsigned char *byte1, *byte2, *end, cur_state;
|
|
|
|
const unsigned int bit = (block % GFS2_NBBY) * GFS2_BIT_SIZE;
|
2006-05-18 19:52:39 +02:00
|
|
|
|
2008-02-06 11:11:15 +01:00
|
|
|
byte1 = buf1 + offset + (block / GFS2_NBBY);
|
|
|
|
end = buf1 + offset + buflen;
|
2006-05-18 19:52:39 +02:00
|
|
|
|
2008-02-06 11:11:15 +01:00
|
|
|
BUG_ON(byte1 >= end);
|
2006-05-18 19:52:39 +02:00
|
|
|
|
2008-02-06 11:11:15 +01:00
|
|
|
cur_state = (*byte1 >> bit) & GFS2_BIT_MASK;
|
2006-05-18 19:52:39 +02:00
|
|
|
|
2008-02-06 11:11:15 +01:00
|
|
|
if (unlikely(!valid_change[new_state * 4 + cur_state])) {
|
2006-05-18 19:52:39 +02:00
|
|
|
gfs2_consist_rgrpd(rgd);
|
2008-02-06 11:11:15 +01:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
*byte1 ^= (cur_state ^ new_state) << bit;
|
|
|
|
|
|
|
|
if (buf2) {
|
|
|
|
byte2 = buf2 + offset + (block / GFS2_NBBY);
|
|
|
|
cur_state = (*byte2 >> bit) & GFS2_BIT_MASK;
|
|
|
|
*byte2 ^= (cur_state ^ new_state) << bit;
|
|
|
|
}
|
2006-05-18 19:52:39 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_testbit - test a bit in the bitmaps
|
|
|
|
* @buffer: the buffer that holds the bitmaps
|
|
|
|
* @buflen: the length (in bytes) of the buffer
|
|
|
|
* @block: the block to read
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2008-02-06 11:11:15 +01:00
|
|
|
static inline unsigned char gfs2_testbit(struct gfs2_rgrpd *rgd,
|
|
|
|
const unsigned char *buffer,
|
|
|
|
unsigned int buflen, u32 block)
|
2006-05-18 19:52:39 +02:00
|
|
|
{
|
2008-02-06 11:11:15 +01:00
|
|
|
const unsigned char *byte, *end;
|
|
|
|
unsigned char cur_state;
|
2006-05-18 19:52:39 +02:00
|
|
|
unsigned int bit;
|
|
|
|
|
|
|
|
byte = buffer + (block / GFS2_NBBY);
|
|
|
|
bit = (block % GFS2_NBBY) * GFS2_BIT_SIZE;
|
|
|
|
end = buffer + buflen;
|
|
|
|
|
|
|
|
gfs2_assert(rgd->rd_sbd, byte < end);
|
|
|
|
|
|
|
|
cur_state = (*byte >> bit) & GFS2_BIT_MASK;
|
|
|
|
|
|
|
|
return cur_state;
|
|
|
|
}
|
|
|
|
|
2009-02-17 15:13:35 +01:00
|
|
|
/**
|
|
|
|
* gfs2_bit_search
|
|
|
|
* @ptr: Pointer to bitmap data
|
|
|
|
* @mask: Mask to use (normally 0x55555.... but adjusted for search start)
|
|
|
|
* @state: The state we are searching for
|
|
|
|
*
|
|
|
|
* We xor the bitmap data with a patter which is the bitwise opposite
|
|
|
|
* of what we are looking for, this gives rise to a pattern of ones
|
|
|
|
* wherever there is a match. Since we have two bits per entry, we
|
|
|
|
* take this pattern, shift it down by one place and then and it with
|
|
|
|
* the original. All the even bit positions (0,2,4, etc) then represent
|
|
|
|
* successful matches, so we mask with 0x55555..... to remove the unwanted
|
|
|
|
* odd bit positions.
|
|
|
|
*
|
|
|
|
* This allows searching of a whole u64 at once (32 blocks) with a
|
|
|
|
* single test (on 64 bit arches).
|
|
|
|
*/
|
|
|
|
|
|
|
|
static inline u64 gfs2_bit_search(const __le64 *ptr, u64 mask, u8 state)
|
|
|
|
{
|
|
|
|
u64 tmp;
|
|
|
|
static const u64 search[] = {
|
2009-02-21 02:11:42 +01:00
|
|
|
[0] = 0xffffffffffffffffULL,
|
|
|
|
[1] = 0xaaaaaaaaaaaaaaaaULL,
|
|
|
|
[2] = 0x5555555555555555ULL,
|
|
|
|
[3] = 0x0000000000000000ULL,
|
2009-02-17 15:13:35 +01:00
|
|
|
};
|
|
|
|
tmp = le64_to_cpu(*ptr) ^ search[state];
|
|
|
|
tmp &= (tmp >> 1);
|
|
|
|
tmp &= mask;
|
|
|
|
return tmp;
|
|
|
|
}
|
|
|
|
|
2006-05-18 19:52:39 +02:00
|
|
|
/**
|
|
|
|
* gfs2_bitfit - Search an rgrp's bitmap buffer to find a bit-pair representing
|
|
|
|
* a block in a given allocation state.
|
|
|
|
* @buffer: the buffer that holds the bitmaps
|
2009-02-17 15:13:35 +01:00
|
|
|
* @len: the length (in bytes) of the buffer
|
2006-05-18 19:52:39 +02:00
|
|
|
* @goal: start search at this block's bit-pair (within @buffer)
|
2009-02-17 15:13:35 +01:00
|
|
|
* @state: GFS2_BLKST_XXX the state of the block we're looking for.
|
2006-05-18 19:52:39 +02:00
|
|
|
*
|
|
|
|
* Scope of @goal and returned block number is only within this bitmap buffer,
|
|
|
|
* not entire rgrp or filesystem. @buffer will be offset from the actual
|
2009-02-17 15:13:35 +01:00
|
|
|
* beginning of a bitmap block buffer, skipping any header structures, but
|
|
|
|
* headers are always a multiple of 64 bits long so that the buffer is
|
|
|
|
* always aligned to a 64 bit boundary.
|
|
|
|
*
|
|
|
|
* The size of the buffer is in bytes, but is it assumed that it is
|
|
|
|
* always ok to to read a complete multiple of 64 bits at the end
|
|
|
|
* of the block in case the end is no aligned to a natural boundary.
|
2006-05-18 19:52:39 +02:00
|
|
|
*
|
|
|
|
* Return: the block number (bitmap buffer scope) that was found
|
|
|
|
*/
|
|
|
|
|
2009-02-21 02:12:05 +01:00
|
|
|
static u32 gfs2_bitfit(const u8 *buf, const unsigned int len,
|
|
|
|
u32 goal, u8 state)
|
2006-05-18 19:52:39 +02:00
|
|
|
{
|
2009-02-17 15:13:35 +01:00
|
|
|
u32 spoint = (goal << 1) & ((8*sizeof(u64)) - 1);
|
|
|
|
const __le64 *ptr = ((__le64 *)buf) + (goal >> 5);
|
|
|
|
const __le64 *end = (__le64 *)(buf + ALIGN(len, sizeof(u64)));
|
|
|
|
u64 tmp;
|
2009-02-21 02:11:42 +01:00
|
|
|
u64 mask = 0x5555555555555555ULL;
|
2009-02-17 15:13:35 +01:00
|
|
|
u32 bit;
|
|
|
|
|
|
|
|
BUG_ON(state > 3);
|
|
|
|
|
|
|
|
/* Mask off bits we don't care about at the start of the search */
|
|
|
|
mask <<= spoint;
|
|
|
|
tmp = gfs2_bit_search(ptr, mask, state);
|
|
|
|
ptr++;
|
|
|
|
while(tmp == 0 && ptr < end) {
|
2009-02-21 02:11:42 +01:00
|
|
|
tmp = gfs2_bit_search(ptr, 0x5555555555555555ULL, state);
|
2009-02-17 15:13:35 +01:00
|
|
|
ptr++;
|
2008-03-11 00:17:47 +01:00
|
|
|
}
|
2009-02-17 15:13:35 +01:00
|
|
|
/* Mask off any bits which are more than len bytes from the start */
|
|
|
|
if (ptr == end && (len & (sizeof(u64) - 1)))
|
|
|
|
tmp &= (((u64)~0) >> (64 - 8*(len & (sizeof(u64) - 1))));
|
|
|
|
/* Didn't find anything, so return */
|
|
|
|
if (tmp == 0)
|
|
|
|
return BFITNOENT;
|
|
|
|
ptr--;
|
2009-04-23 09:54:02 +02:00
|
|
|
bit = __ffs64(tmp);
|
2009-02-17 15:13:35 +01:00
|
|
|
bit /= 2; /* two bits per entry in the bitmap */
|
|
|
|
return (((const unsigned char *)ptr - buf) * GFS2_NBBY) + bit;
|
2006-05-18 19:52:39 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_bitcount - count the number of bits in a certain state
|
|
|
|
* @buffer: the buffer that holds the bitmaps
|
|
|
|
* @buflen: the length (in bytes) of the buffer
|
|
|
|
* @state: the state of the block we're looking for
|
|
|
|
*
|
|
|
|
* Returns: The number of bits
|
|
|
|
*/
|
|
|
|
|
2008-01-29 14:30:20 +01:00
|
|
|
static u32 gfs2_bitcount(struct gfs2_rgrpd *rgd, const u8 *buffer,
|
|
|
|
unsigned int buflen, u8 state)
|
2006-05-18 19:52:39 +02:00
|
|
|
{
|
2008-01-29 14:30:20 +01:00
|
|
|
const u8 *byte = buffer;
|
|
|
|
const u8 *end = buffer + buflen;
|
|
|
|
const u8 state1 = state << 2;
|
|
|
|
const u8 state2 = state << 4;
|
|
|
|
const u8 state3 = state << 6;
|
2006-09-04 18:49:07 +02:00
|
|
|
u32 count = 0;
|
2006-05-18 19:52:39 +02:00
|
|
|
|
|
|
|
for (; byte < end; byte++) {
|
|
|
|
if (((*byte) & 0x03) == state)
|
|
|
|
count++;
|
|
|
|
if (((*byte) & 0x0C) == state1)
|
|
|
|
count++;
|
|
|
|
if (((*byte) & 0x30) == state2)
|
|
|
|
count++;
|
|
|
|
if (((*byte) & 0xC0) == state3)
|
|
|
|
count++;
|
|
|
|
}
|
|
|
|
|
|
|
|
return count;
|
|
|
|
}
|
|
|
|
|
2006-01-16 17:50:04 +01:00
|
|
|
/**
|
|
|
|
* gfs2_rgrp_verify - Verify that a resource group is consistent
|
|
|
|
* @sdp: the filesystem
|
|
|
|
* @rgd: the rgrp
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
|
|
|
void gfs2_rgrp_verify(struct gfs2_rgrpd *rgd)
|
|
|
|
{
|
|
|
|
struct gfs2_sbd *sdp = rgd->rd_sbd;
|
|
|
|
struct gfs2_bitmap *bi = NULL;
|
2007-06-01 15:11:58 +02:00
|
|
|
u32 length = rgd->rd_length;
|
2006-09-04 18:49:07 +02:00
|
|
|
u32 count[4], tmp;
|
2006-01-16 17:50:04 +01:00
|
|
|
int buf, x;
|
|
|
|
|
2006-09-04 18:49:07 +02:00
|
|
|
memset(count, 0, 4 * sizeof(u32));
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
/* Count # blocks in each of 4 possible allocation states */
|
|
|
|
for (buf = 0; buf < length; buf++) {
|
|
|
|
bi = rgd->rd_bits + buf;
|
|
|
|
for (x = 0; x < 4; x++)
|
|
|
|
count[x] += gfs2_bitcount(rgd,
|
|
|
|
bi->bi_bh->b_data +
|
|
|
|
bi->bi_offset,
|
|
|
|
bi->bi_len, x);
|
|
|
|
}
|
|
|
|
|
2008-11-04 11:25:13 +01:00
|
|
|
if (count[0] != rgd->rd_free) {
|
2006-01-16 17:50:04 +01:00
|
|
|
if (gfs2_consist_rgrpd(rgd))
|
|
|
|
fs_err(sdp, "free data mismatch: %u != %u\n",
|
2008-11-04 11:25:13 +01:00
|
|
|
count[0], rgd->rd_free);
|
2006-01-16 17:50:04 +01:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2008-11-04 11:32:57 +01:00
|
|
|
tmp = rgd->rd_data - rgd->rd_free - rgd->rd_dinodes;
|
2006-06-14 21:32:57 +02:00
|
|
|
if (count[1] + count[2] != tmp) {
|
2006-01-16 17:50:04 +01:00
|
|
|
if (gfs2_consist_rgrpd(rgd))
|
|
|
|
fs_err(sdp, "used data mismatch: %u != %u\n",
|
|
|
|
count[1], tmp);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2008-11-04 11:32:57 +01:00
|
|
|
if (count[3] != rgd->rd_dinodes) {
|
2006-01-16 17:50:04 +01:00
|
|
|
if (gfs2_consist_rgrpd(rgd))
|
2006-06-14 21:32:57 +02:00
|
|
|
fs_err(sdp, "used metadata mismatch: %u != %u\n",
|
2008-11-04 11:32:57 +01:00
|
|
|
count[3], rgd->rd_dinodes);
|
2006-01-16 17:50:04 +01:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2006-06-14 21:32:57 +02:00
|
|
|
if (count[2] > count[3]) {
|
2006-01-16 17:50:04 +01:00
|
|
|
if (gfs2_consist_rgrpd(rgd))
|
2006-06-14 21:32:57 +02:00
|
|
|
fs_err(sdp, "unlinked inodes > inodes: %u\n",
|
|
|
|
count[2]);
|
2006-01-16 17:50:04 +01:00
|
|
|
return;
|
|
|
|
}
|
2006-06-14 21:32:57 +02:00
|
|
|
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
2007-06-01 15:11:58 +02:00
|
|
|
static inline int rgrp_contains_block(struct gfs2_rgrpd *rgd, u64 block)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
2007-06-01 15:11:58 +02:00
|
|
|
u64 first = rgd->rd_data0;
|
|
|
|
u64 last = first + rgd->rd_data;
|
2006-09-05 17:15:45 +02:00
|
|
|
return first <= block && block < last;
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_blk2rgrpd - Find resource group for a given data/meta block number
|
|
|
|
* @sdp: The GFS2 superblock
|
|
|
|
* @n: The data block number
|
|
|
|
*
|
|
|
|
* Returns: The resource group, or NULL if not found
|
|
|
|
*/
|
|
|
|
|
2006-09-04 18:49:07 +02:00
|
|
|
struct gfs2_rgrpd *gfs2_blk2rgrpd(struct gfs2_sbd *sdp, u64 blk)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
|
|
|
struct gfs2_rgrpd *rgd;
|
|
|
|
|
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
|
|
|
|
|
|
|
list_for_each_entry(rgd, &sdp->sd_rindex_mru_list, rd_list_mru) {
|
2007-06-01 15:11:58 +02:00
|
|
|
if (rgrp_contains_block(rgd, blk)) {
|
2006-01-16 17:50:04 +01:00
|
|
|
list_move(&rgd->rd_list_mru, &sdp->sd_rindex_mru_list);
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
return rgd;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_rgrpd_get_first - get the first Resource Group in the filesystem
|
|
|
|
* @sdp: The GFS2 superblock
|
|
|
|
*
|
|
|
|
* Returns: The first rgrp in the filesystem
|
|
|
|
*/
|
|
|
|
|
|
|
|
struct gfs2_rgrpd *gfs2_rgrpd_get_first(struct gfs2_sbd *sdp)
|
|
|
|
{
|
|
|
|
gfs2_assert(sdp, !list_empty(&sdp->sd_rindex_list));
|
|
|
|
return list_entry(sdp->sd_rindex_list.next, struct gfs2_rgrpd, rd_list);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_rgrpd_get_next - get the next RG
|
|
|
|
* @rgd: A RG
|
|
|
|
*
|
|
|
|
* Returns: The next rgrp
|
|
|
|
*/
|
|
|
|
|
|
|
|
struct gfs2_rgrpd *gfs2_rgrpd_get_next(struct gfs2_rgrpd *rgd)
|
|
|
|
{
|
|
|
|
if (rgd->rd_list.next == &rgd->rd_sbd->sd_rindex_list)
|
|
|
|
return NULL;
|
|
|
|
return list_entry(rgd->rd_list.next, struct gfs2_rgrpd, rd_list);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void clear_rgrpdi(struct gfs2_sbd *sdp)
|
|
|
|
{
|
|
|
|
struct list_head *head;
|
|
|
|
struct gfs2_rgrpd *rgd;
|
|
|
|
struct gfs2_glock *gl;
|
|
|
|
|
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
|
|
|
sdp->sd_rindex_forward = NULL;
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
|
|
|
|
head = &sdp->sd_rindex_list;
|
|
|
|
while (!list_empty(head)) {
|
|
|
|
rgd = list_entry(head->next, struct gfs2_rgrpd, rd_list);
|
|
|
|
gl = rgd->rd_gl;
|
|
|
|
|
|
|
|
list_del(&rgd->rd_list);
|
|
|
|
list_del(&rgd->rd_list_mru);
|
|
|
|
|
|
|
|
if (gl) {
|
2006-02-27 23:23:27 +01:00
|
|
|
gl->gl_object = NULL;
|
2006-01-16 17:50:04 +01:00
|
|
|
gfs2_glock_put(gl);
|
|
|
|
}
|
|
|
|
|
|
|
|
kfree(rgd->rd_bits);
|
2008-01-29 00:20:26 +01:00
|
|
|
kmem_cache_free(gfs2_rgrpd_cachep, rgd);
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void gfs2_clear_rgrpd(struct gfs2_sbd *sdp)
|
|
|
|
{
|
2006-02-21 13:51:39 +01:00
|
|
|
mutex_lock(&sdp->sd_rindex_mutex);
|
2006-01-16 17:50:04 +01:00
|
|
|
clear_rgrpdi(sdp);
|
2006-02-21 13:51:39 +01:00
|
|
|
mutex_unlock(&sdp->sd_rindex_mutex);
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
2007-06-01 15:11:58 +02:00
|
|
|
static void gfs2_rindex_print(const struct gfs2_rgrpd *rgd)
|
|
|
|
{
|
|
|
|
printk(KERN_INFO " ri_addr = %llu\n", (unsigned long long)rgd->rd_addr);
|
|
|
|
printk(KERN_INFO " ri_length = %u\n", rgd->rd_length);
|
|
|
|
printk(KERN_INFO " ri_data0 = %llu\n", (unsigned long long)rgd->rd_data0);
|
|
|
|
printk(KERN_INFO " ri_data = %u\n", rgd->rd_data);
|
|
|
|
printk(KERN_INFO " ri_bitbytes = %u\n", rgd->rd_bitbytes);
|
|
|
|
}
|
|
|
|
|
2006-01-16 17:50:04 +01:00
|
|
|
/**
|
|
|
|
* gfs2_compute_bitstructs - Compute the bitmap sizes
|
|
|
|
* @rgd: The resource group descriptor
|
|
|
|
*
|
|
|
|
* Calculates bitmap descriptors, one for each block that contains bitmap data
|
|
|
|
*
|
|
|
|
* Returns: errno
|
|
|
|
*/
|
|
|
|
|
|
|
|
static int compute_bitstructs(struct gfs2_rgrpd *rgd)
|
|
|
|
{
|
|
|
|
struct gfs2_sbd *sdp = rgd->rd_sbd;
|
|
|
|
struct gfs2_bitmap *bi;
|
2007-06-01 15:11:58 +02:00
|
|
|
u32 length = rgd->rd_length; /* # blocks in hdr & bitmap */
|
2006-09-04 18:49:07 +02:00
|
|
|
u32 bytes_left, bytes;
|
2006-01-16 17:50:04 +01:00
|
|
|
int x;
|
|
|
|
|
2006-06-14 21:32:57 +02:00
|
|
|
if (!length)
|
|
|
|
return -EINVAL;
|
|
|
|
|
2006-07-27 20:29:00 +02:00
|
|
|
rgd->rd_bits = kcalloc(length, sizeof(struct gfs2_bitmap), GFP_NOFS);
|
2006-01-16 17:50:04 +01:00
|
|
|
if (!rgd->rd_bits)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
2007-06-01 15:11:58 +02:00
|
|
|
bytes_left = rgd->rd_bitbytes;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
for (x = 0; x < length; x++) {
|
|
|
|
bi = rgd->rd_bits + x;
|
|
|
|
|
|
|
|
/* small rgrp; bitmap stored completely in header block */
|
|
|
|
if (length == 1) {
|
|
|
|
bytes = bytes_left;
|
|
|
|
bi->bi_offset = sizeof(struct gfs2_rgrp);
|
|
|
|
bi->bi_start = 0;
|
|
|
|
bi->bi_len = bytes;
|
|
|
|
/* header block */
|
|
|
|
} else if (x == 0) {
|
|
|
|
bytes = sdp->sd_sb.sb_bsize - sizeof(struct gfs2_rgrp);
|
|
|
|
bi->bi_offset = sizeof(struct gfs2_rgrp);
|
|
|
|
bi->bi_start = 0;
|
|
|
|
bi->bi_len = bytes;
|
|
|
|
/* last block */
|
|
|
|
} else if (x + 1 == length) {
|
|
|
|
bytes = bytes_left;
|
|
|
|
bi->bi_offset = sizeof(struct gfs2_meta_header);
|
2007-06-01 15:11:58 +02:00
|
|
|
bi->bi_start = rgd->rd_bitbytes - bytes_left;
|
2006-01-16 17:50:04 +01:00
|
|
|
bi->bi_len = bytes;
|
|
|
|
/* other blocks */
|
|
|
|
} else {
|
2006-02-27 18:00:42 +01:00
|
|
|
bytes = sdp->sd_sb.sb_bsize -
|
|
|
|
sizeof(struct gfs2_meta_header);
|
2006-01-16 17:50:04 +01:00
|
|
|
bi->bi_offset = sizeof(struct gfs2_meta_header);
|
2007-06-01 15:11:58 +02:00
|
|
|
bi->bi_start = rgd->rd_bitbytes - bytes_left;
|
2006-01-16 17:50:04 +01:00
|
|
|
bi->bi_len = bytes;
|
|
|
|
}
|
|
|
|
|
|
|
|
bytes_left -= bytes;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (bytes_left) {
|
|
|
|
gfs2_consist_rgrpd(rgd);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
bi = rgd->rd_bits + (length - 1);
|
2007-06-01 15:11:58 +02:00
|
|
|
if ((bi->bi_start + bi->bi_len) * GFS2_NBBY != rgd->rd_data) {
|
2006-01-16 17:50:04 +01:00
|
|
|
if (gfs2_consist_rgrpd(rgd)) {
|
2007-06-01 15:11:58 +02:00
|
|
|
gfs2_rindex_print(rgd);
|
2006-01-16 17:50:04 +01:00
|
|
|
fs_err(sdp, "start=%u len=%u offset=%u\n",
|
|
|
|
bi->bi_start, bi->bi_len, bi->bi_offset);
|
|
|
|
}
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
[GFS2] kernel changes to support new gfs2_grow command
This is another revision of my gfs2 kernel patch that allows
gfs2_grow to function properly.
Steve Whitehouse expressed some concerns about the previous
patch and I restructured it based on his comments.
The previous patch was doing the statfs_change at file close time,
under its own transaction. The current patch does the statfs_change
inside the gfs2_commit_write function, which keeps it under the
umbrella of the inode transaction.
I can't call ri_update to re-read the rindex file during the
transaction because the transaction may have outstanding unwritten
buffers attached to the rgrps that would be otherwise blown away.
So instead, I created a new function, gfs2_ri_total, that will
re-read the rindex file just to total the file system space
for the sake of the statfs_change. The ri_update will happen
later, when gfs2 realizes the version number has changed, as it
happened before my patch.
Since the statfs_change is happening at write_commit time and there
may be multiple writes to the rindex file for one grow operation.
So one consequence of this restructuring is that instead of getting
one kernel message to indicate the change, you may see several.
For example, before when you did a gfs2_grow, you'd get a single
message like:
GFS2: File system extended by 247876 blocks (968MB)
Now you get something like:
GFS2: File system extended by 207896 blocks (812MB)
GFS2: File system extended by 39980 blocks (156MB)
This version has also been successfully run against the hours-long
"gfs2_fsck_hellfire" test that does several gfs2_grow and gfs2_fsck
while interjecting file system damage. It does this repeatedly
under a variety Resource Group conditions.
Signed-off-By: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-05-09 16:37:57 +02:00
|
|
|
/**
|
|
|
|
* gfs2_ri_total - Total up the file system space, according to the rindex.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
u64 gfs2_ri_total(struct gfs2_sbd *sdp)
|
|
|
|
{
|
|
|
|
u64 total_data = 0;
|
|
|
|
struct inode *inode = sdp->sd_rindex;
|
|
|
|
struct gfs2_inode *ip = GFS2_I(inode);
|
|
|
|
char buf[sizeof(struct gfs2_rindex)];
|
|
|
|
struct file_ra_state ra_state;
|
|
|
|
int error, rgrps;
|
|
|
|
|
|
|
|
mutex_lock(&sdp->sd_rindex_mutex);
|
|
|
|
file_ra_state_init(&ra_state, inode->i_mapping);
|
|
|
|
for (rgrps = 0;; rgrps++) {
|
|
|
|
loff_t pos = rgrps * sizeof(struct gfs2_rindex);
|
|
|
|
|
2008-11-04 10:47:33 +01:00
|
|
|
if (pos + sizeof(struct gfs2_rindex) >= ip->i_disksize)
|
[GFS2] kernel changes to support new gfs2_grow command
This is another revision of my gfs2 kernel patch that allows
gfs2_grow to function properly.
Steve Whitehouse expressed some concerns about the previous
patch and I restructured it based on his comments.
The previous patch was doing the statfs_change at file close time,
under its own transaction. The current patch does the statfs_change
inside the gfs2_commit_write function, which keeps it under the
umbrella of the inode transaction.
I can't call ri_update to re-read the rindex file during the
transaction because the transaction may have outstanding unwritten
buffers attached to the rgrps that would be otherwise blown away.
So instead, I created a new function, gfs2_ri_total, that will
re-read the rindex file just to total the file system space
for the sake of the statfs_change. The ri_update will happen
later, when gfs2 realizes the version number has changed, as it
happened before my patch.
Since the statfs_change is happening at write_commit time and there
may be multiple writes to the rindex file for one grow operation.
So one consequence of this restructuring is that instead of getting
one kernel message to indicate the change, you may see several.
For example, before when you did a gfs2_grow, you'd get a single
message like:
GFS2: File system extended by 247876 blocks (968MB)
Now you get something like:
GFS2: File system extended by 207896 blocks (812MB)
GFS2: File system extended by 39980 blocks (156MB)
This version has also been successfully run against the hours-long
"gfs2_fsck_hellfire" test that does several gfs2_grow and gfs2_fsck
while interjecting file system damage. It does this repeatedly
under a variety Resource Group conditions.
Signed-off-By: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-05-09 16:37:57 +02:00
|
|
|
break;
|
|
|
|
error = gfs2_internal_read(ip, &ra_state, buf, &pos,
|
|
|
|
sizeof(struct gfs2_rindex));
|
|
|
|
if (error != sizeof(struct gfs2_rindex))
|
|
|
|
break;
|
2007-06-01 15:11:58 +02:00
|
|
|
total_data += be32_to_cpu(((struct gfs2_rindex *)buf)->ri_data);
|
[GFS2] kernel changes to support new gfs2_grow command
This is another revision of my gfs2 kernel patch that allows
gfs2_grow to function properly.
Steve Whitehouse expressed some concerns about the previous
patch and I restructured it based on his comments.
The previous patch was doing the statfs_change at file close time,
under its own transaction. The current patch does the statfs_change
inside the gfs2_commit_write function, which keeps it under the
umbrella of the inode transaction.
I can't call ri_update to re-read the rindex file during the
transaction because the transaction may have outstanding unwritten
buffers attached to the rgrps that would be otherwise blown away.
So instead, I created a new function, gfs2_ri_total, that will
re-read the rindex file just to total the file system space
for the sake of the statfs_change. The ri_update will happen
later, when gfs2 realizes the version number has changed, as it
happened before my patch.
Since the statfs_change is happening at write_commit time and there
may be multiple writes to the rindex file for one grow operation.
So one consequence of this restructuring is that instead of getting
one kernel message to indicate the change, you may see several.
For example, before when you did a gfs2_grow, you'd get a single
message like:
GFS2: File system extended by 247876 blocks (968MB)
Now you get something like:
GFS2: File system extended by 207896 blocks (812MB)
GFS2: File system extended by 39980 blocks (156MB)
This version has also been successfully run against the hours-long
"gfs2_fsck_hellfire" test that does several gfs2_grow and gfs2_fsck
while interjecting file system damage. It does this repeatedly
under a variety Resource Group conditions.
Signed-off-By: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-05-09 16:37:57 +02:00
|
|
|
}
|
|
|
|
mutex_unlock(&sdp->sd_rindex_mutex);
|
|
|
|
return total_data;
|
|
|
|
}
|
|
|
|
|
2007-06-01 15:11:58 +02:00
|
|
|
static void gfs2_rindex_in(struct gfs2_rgrpd *rgd, const void *buf)
|
|
|
|
{
|
|
|
|
const struct gfs2_rindex *str = buf;
|
|
|
|
|
|
|
|
rgd->rd_addr = be64_to_cpu(str->ri_addr);
|
|
|
|
rgd->rd_length = be32_to_cpu(str->ri_length);
|
|
|
|
rgd->rd_data0 = be64_to_cpu(str->ri_data0);
|
|
|
|
rgd->rd_data = be32_to_cpu(str->ri_data);
|
|
|
|
rgd->rd_bitbytes = be32_to_cpu(str->ri_bitbytes);
|
|
|
|
}
|
|
|
|
|
2006-01-16 17:50:04 +01:00
|
|
|
/**
|
2007-05-10 23:54:38 +02:00
|
|
|
* read_rindex_entry - Pull in a new resource index entry from the disk
|
2006-01-16 17:50:04 +01:00
|
|
|
* @gl: The glock covering the rindex inode
|
|
|
|
*
|
2007-05-10 23:54:38 +02:00
|
|
|
* Returns: 0 on success, error code otherwise
|
|
|
|
*/
|
|
|
|
|
|
|
|
static int read_rindex_entry(struct gfs2_inode *ip,
|
|
|
|
struct file_ra_state *ra_state)
|
|
|
|
{
|
|
|
|
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
|
|
|
|
loff_t pos = sdp->sd_rgrps * sizeof(struct gfs2_rindex);
|
|
|
|
char buf[sizeof(struct gfs2_rindex)];
|
|
|
|
int error;
|
|
|
|
struct gfs2_rgrpd *rgd;
|
|
|
|
|
|
|
|
error = gfs2_internal_read(ip, ra_state, buf, &pos,
|
|
|
|
sizeof(struct gfs2_rindex));
|
|
|
|
if (!error)
|
|
|
|
return 0;
|
|
|
|
if (error != sizeof(struct gfs2_rindex)) {
|
|
|
|
if (error > 0)
|
|
|
|
error = -EIO;
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2008-01-29 00:20:26 +01:00
|
|
|
rgd = kmem_cache_zalloc(gfs2_rgrpd_cachep, GFP_NOFS);
|
2007-05-10 23:54:38 +02:00
|
|
|
error = -ENOMEM;
|
|
|
|
if (!rgd)
|
|
|
|
return error;
|
|
|
|
|
|
|
|
mutex_init(&rgd->rd_mutex);
|
|
|
|
lops_init_le(&rgd->rd_le, &gfs2_rg_lops);
|
|
|
|
rgd->rd_sbd = sdp;
|
|
|
|
|
|
|
|
list_add_tail(&rgd->rd_list, &sdp->sd_rindex_list);
|
|
|
|
list_add_tail(&rgd->rd_list_mru, &sdp->sd_rindex_mru_list);
|
|
|
|
|
2007-06-01 15:11:58 +02:00
|
|
|
gfs2_rindex_in(rgd, buf);
|
2007-05-10 23:54:38 +02:00
|
|
|
error = compute_bitstructs(rgd);
|
|
|
|
if (error)
|
|
|
|
return error;
|
|
|
|
|
2007-06-01 15:11:58 +02:00
|
|
|
error = gfs2_glock_get(sdp, rgd->rd_addr,
|
2007-05-10 23:54:38 +02:00
|
|
|
&gfs2_rgrp_glops, CREATE, &rgd->rd_gl);
|
|
|
|
if (error)
|
|
|
|
return error;
|
|
|
|
|
|
|
|
rgd->rd_gl->gl_object = rgd;
|
2008-01-31 17:31:39 +01:00
|
|
|
rgd->rd_flags &= ~GFS2_RDF_UPTODATE;
|
2007-06-08 11:05:33 +02:00
|
|
|
rgd->rd_flags |= GFS2_RDF_CHECK;
|
2007-05-10 23:54:38 +02:00
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_ri_update - Pull in a new resource index from the disk
|
|
|
|
* @ip: pointer to the rindex inode
|
|
|
|
*
|
2006-01-16 17:50:04 +01:00
|
|
|
* Returns: 0 on successful update, error code otherwise
|
|
|
|
*/
|
|
|
|
|
|
|
|
static int gfs2_ri_update(struct gfs2_inode *ip)
|
|
|
|
{
|
2006-06-14 21:32:57 +02:00
|
|
|
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
|
|
|
|
struct inode *inode = &ip->i_inode;
|
2006-01-30 19:34:10 +01:00
|
|
|
struct file_ra_state ra_state;
|
2008-11-04 10:47:33 +01:00
|
|
|
u64 rgrp_count = ip->i_disksize;
|
2006-01-16 17:50:04 +01:00
|
|
|
int error;
|
|
|
|
|
2007-05-14 19:42:18 +02:00
|
|
|
if (do_div(rgrp_count, sizeof(struct gfs2_rindex))) {
|
2006-01-16 17:50:04 +01:00
|
|
|
gfs2_consist_inode(ip);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
clear_rgrpdi(sdp);
|
|
|
|
|
2006-01-30 19:34:10 +01:00
|
|
|
file_ra_state_init(&ra_state, inode->i_mapping);
|
2007-05-14 19:42:18 +02:00
|
|
|
for (sdp->sd_rgrps = 0; sdp->sd_rgrps < rgrp_count; sdp->sd_rgrps++) {
|
2007-05-10 23:54:38 +02:00
|
|
|
error = read_rindex_entry(ip, &ra_state);
|
|
|
|
if (error) {
|
|
|
|
clear_rgrpdi(sdp);
|
|
|
|
return error;
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
2007-05-10 23:54:38 +02:00
|
|
|
}
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2008-01-31 17:31:39 +01:00
|
|
|
sdp->sd_rindex_uptodate = 1;
|
2007-05-10 23:54:38 +02:00
|
|
|
return 0;
|
|
|
|
}
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2007-05-10 23:54:38 +02:00
|
|
|
/**
|
|
|
|
* gfs2_ri_update_special - Pull in a new resource index from the disk
|
|
|
|
*
|
|
|
|
* This is a special version that's safe to call from gfs2_inplace_reserve_i.
|
|
|
|
* In this case we know that we don't have any resource groups in memory yet.
|
|
|
|
*
|
|
|
|
* @ip: pointer to the rindex inode
|
|
|
|
*
|
|
|
|
* Returns: 0 on successful update, error code otherwise
|
|
|
|
*/
|
|
|
|
static int gfs2_ri_update_special(struct gfs2_inode *ip)
|
|
|
|
{
|
|
|
|
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
|
|
|
|
struct inode *inode = &ip->i_inode;
|
|
|
|
struct file_ra_state ra_state;
|
|
|
|
int error;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2007-05-10 23:54:38 +02:00
|
|
|
file_ra_state_init(&ra_state, inode->i_mapping);
|
|
|
|
for (sdp->sd_rgrps = 0;; sdp->sd_rgrps++) {
|
|
|
|
/* Ignore partials */
|
|
|
|
if ((sdp->sd_rgrps + 1) * sizeof(struct gfs2_rindex) >
|
2008-11-04 10:47:33 +01:00
|
|
|
ip->i_disksize)
|
2007-05-10 23:54:38 +02:00
|
|
|
break;
|
|
|
|
error = read_rindex_entry(ip, &ra_state);
|
|
|
|
if (error) {
|
|
|
|
clear_rgrpdi(sdp);
|
|
|
|
return error;
|
|
|
|
}
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
2008-01-31 17:31:39 +01:00
|
|
|
sdp->sd_rindex_uptodate = 1;
|
2006-01-16 17:50:04 +01:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_rindex_hold - Grab a lock on the rindex
|
|
|
|
* @sdp: The GFS2 superblock
|
|
|
|
* @ri_gh: the glock holder
|
|
|
|
*
|
|
|
|
* We grab a lock on the rindex inode to make sure that it doesn't
|
|
|
|
* change whilst we are performing an operation. We keep this lock
|
|
|
|
* for quite long periods of time compared to other locks. This
|
|
|
|
* doesn't matter, since it is shared and it is very, very rarely
|
|
|
|
* accessed in the exclusive mode (i.e. only when expanding the filesystem).
|
|
|
|
*
|
|
|
|
* This makes sure that we're using the latest copy of the resource index
|
|
|
|
* special file, which might have been updated if someone expanded the
|
|
|
|
* filesystem (via gfs2_grow utility), which adds new resource groups.
|
|
|
|
*
|
|
|
|
* Returns: 0 on success, error code otherwise
|
|
|
|
*/
|
|
|
|
|
|
|
|
int gfs2_rindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ri_gh)
|
|
|
|
{
|
2006-06-14 21:32:57 +02:00
|
|
|
struct gfs2_inode *ip = GFS2_I(sdp->sd_rindex);
|
2006-01-16 17:50:04 +01:00
|
|
|
struct gfs2_glock *gl = ip->i_gl;
|
|
|
|
int error;
|
|
|
|
|
|
|
|
error = gfs2_glock_nq_init(gl, LM_ST_SHARED, 0, ri_gh);
|
|
|
|
if (error)
|
|
|
|
return error;
|
|
|
|
|
|
|
|
/* Read new copy from disk if we don't have the latest */
|
2008-01-31 17:31:39 +01:00
|
|
|
if (!sdp->sd_rindex_uptodate) {
|
2006-02-21 13:51:39 +01:00
|
|
|
mutex_lock(&sdp->sd_rindex_mutex);
|
2008-01-31 17:31:39 +01:00
|
|
|
if (!sdp->sd_rindex_uptodate) {
|
2006-01-16 17:50:04 +01:00
|
|
|
error = gfs2_ri_update(ip);
|
|
|
|
if (error)
|
|
|
|
gfs2_glock_dq_uninit(ri_gh);
|
|
|
|
}
|
2006-02-21 13:51:39 +01:00
|
|
|
mutex_unlock(&sdp->sd_rindex_mutex);
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2008-01-29 01:38:07 +01:00
|
|
|
static void gfs2_rgrp_in(struct gfs2_rgrpd *rgd, const void *buf)
|
2007-06-01 15:11:58 +02:00
|
|
|
{
|
|
|
|
const struct gfs2_rgrp *str = buf;
|
2008-01-29 01:38:07 +01:00
|
|
|
u32 rg_flags;
|
2007-06-01 15:11:58 +02:00
|
|
|
|
2008-01-29 01:38:07 +01:00
|
|
|
rg_flags = be32_to_cpu(str->rg_flags);
|
|
|
|
if (rg_flags & GFS2_RGF_NOALLOC)
|
|
|
|
rgd->rd_flags |= GFS2_RDF_NOALLOC;
|
|
|
|
else
|
|
|
|
rgd->rd_flags &= ~GFS2_RDF_NOALLOC;
|
2008-11-04 11:25:13 +01:00
|
|
|
rgd->rd_free = be32_to_cpu(str->rg_free);
|
2008-11-04 11:32:57 +01:00
|
|
|
rgd->rd_dinodes = be32_to_cpu(str->rg_dinodes);
|
2008-11-04 11:19:03 +01:00
|
|
|
rgd->rd_igeneration = be64_to_cpu(str->rg_igeneration);
|
2007-06-01 15:11:58 +02:00
|
|
|
}
|
|
|
|
|
2008-01-29 01:38:07 +01:00
|
|
|
static void gfs2_rgrp_out(struct gfs2_rgrpd *rgd, void *buf)
|
2007-06-01 15:11:58 +02:00
|
|
|
{
|
|
|
|
struct gfs2_rgrp *str = buf;
|
2008-01-29 01:38:07 +01:00
|
|
|
u32 rg_flags = 0;
|
2007-06-01 15:11:58 +02:00
|
|
|
|
2008-01-29 01:38:07 +01:00
|
|
|
if (rgd->rd_flags & GFS2_RDF_NOALLOC)
|
|
|
|
rg_flags |= GFS2_RGF_NOALLOC;
|
|
|
|
str->rg_flags = cpu_to_be32(rg_flags);
|
2008-11-04 11:25:13 +01:00
|
|
|
str->rg_free = cpu_to_be32(rgd->rd_free);
|
2008-11-04 11:32:57 +01:00
|
|
|
str->rg_dinodes = cpu_to_be32(rgd->rd_dinodes);
|
2007-06-01 15:11:58 +02:00
|
|
|
str->__pad = cpu_to_be32(0);
|
2008-11-04 11:19:03 +01:00
|
|
|
str->rg_igeneration = cpu_to_be64(rgd->rd_igeneration);
|
2007-06-01 15:11:58 +02:00
|
|
|
memset(&str->rg_reserved, 0, sizeof(str->rg_reserved));
|
|
|
|
}
|
|
|
|
|
2006-01-16 17:50:04 +01:00
|
|
|
/**
|
|
|
|
* gfs2_rgrp_bh_get - Read in a RG's header and bitmaps
|
|
|
|
* @rgd: the struct gfs2_rgrpd describing the RG to read in
|
|
|
|
*
|
|
|
|
* Read in all of a Resource Group's header and bitmap blocks.
|
|
|
|
* Caller must eventually call gfs2_rgrp_relse() to free the bitmaps.
|
|
|
|
*
|
|
|
|
* Returns: errno
|
|
|
|
*/
|
|
|
|
|
|
|
|
int gfs2_rgrp_bh_get(struct gfs2_rgrpd *rgd)
|
|
|
|
{
|
|
|
|
struct gfs2_sbd *sdp = rgd->rd_sbd;
|
|
|
|
struct gfs2_glock *gl = rgd->rd_gl;
|
2007-06-01 15:11:58 +02:00
|
|
|
unsigned int length = rgd->rd_length;
|
2006-01-16 17:50:04 +01:00
|
|
|
struct gfs2_bitmap *bi;
|
|
|
|
unsigned int x, y;
|
|
|
|
int error;
|
|
|
|
|
2006-02-21 13:51:39 +01:00
|
|
|
mutex_lock(&rgd->rd_mutex);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
|
|
|
if (rgd->rd_bh_count) {
|
|
|
|
rgd->rd_bh_count++;
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
2006-02-21 13:51:39 +01:00
|
|
|
mutex_unlock(&rgd->rd_mutex);
|
2006-01-16 17:50:04 +01:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
|
|
|
|
for (x = 0; x < length; x++) {
|
|
|
|
bi = rgd->rd_bits + x;
|
2007-06-01 15:11:58 +02:00
|
|
|
error = gfs2_meta_read(gl, rgd->rd_addr + x, 0, &bi->bi_bh);
|
2006-01-16 17:50:04 +01:00
|
|
|
if (error)
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (y = length; y--;) {
|
|
|
|
bi = rgd->rd_bits + y;
|
2006-09-21 23:05:23 +02:00
|
|
|
error = gfs2_meta_wait(sdp, bi->bi_bh);
|
2006-01-16 17:50:04 +01:00
|
|
|
if (error)
|
|
|
|
goto fail;
|
2006-06-14 21:32:57 +02:00
|
|
|
if (gfs2_metatype_check(sdp, bi->bi_bh, y ? GFS2_METATYPE_RB :
|
2006-01-16 17:50:04 +01:00
|
|
|
GFS2_METATYPE_RG)) {
|
|
|
|
error = -EIO;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-01-31 17:31:39 +01:00
|
|
|
if (!(rgd->rd_flags & GFS2_RDF_UPTODATE)) {
|
2008-01-29 01:38:07 +01:00
|
|
|
gfs2_rgrp_in(rgd, (rgd->rd_bits[0].bi_bh)->b_data);
|
2008-01-31 17:31:39 +01:00
|
|
|
rgd->rd_flags |= GFS2_RDF_UPTODATE;
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
2008-11-04 11:25:13 +01:00
|
|
|
rgd->rd_free_clone = rgd->rd_free;
|
2006-01-16 17:50:04 +01:00
|
|
|
rgd->rd_bh_count++;
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
|
2006-02-21 13:51:39 +01:00
|
|
|
mutex_unlock(&rgd->rd_mutex);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
2006-06-14 21:32:57 +02:00
|
|
|
fail:
|
2006-01-16 17:50:04 +01:00
|
|
|
while (x--) {
|
|
|
|
bi = rgd->rd_bits + x;
|
|
|
|
brelse(bi->bi_bh);
|
|
|
|
bi->bi_bh = NULL;
|
|
|
|
gfs2_assert_warn(sdp, !bi->bi_clone);
|
|
|
|
}
|
2006-02-21 13:51:39 +01:00
|
|
|
mutex_unlock(&rgd->rd_mutex);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
|
|
|
void gfs2_rgrp_bh_hold(struct gfs2_rgrpd *rgd)
|
|
|
|
{
|
|
|
|
struct gfs2_sbd *sdp = rgd->rd_sbd;
|
|
|
|
|
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
|
|
|
gfs2_assert_warn(rgd->rd_sbd, rgd->rd_bh_count);
|
|
|
|
rgd->rd_bh_count++;
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_rgrp_bh_put - Release RG bitmaps read in with gfs2_rgrp_bh_get()
|
|
|
|
* @rgd: the struct gfs2_rgrpd describing the RG to read in
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
|
|
|
void gfs2_rgrp_bh_put(struct gfs2_rgrpd *rgd)
|
|
|
|
{
|
|
|
|
struct gfs2_sbd *sdp = rgd->rd_sbd;
|
2007-06-01 15:11:58 +02:00
|
|
|
int x, length = rgd->rd_length;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
|
|
|
gfs2_assert_warn(rgd->rd_sbd, rgd->rd_bh_count);
|
|
|
|
if (--rgd->rd_bh_count) {
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (x = 0; x < length; x++) {
|
|
|
|
struct gfs2_bitmap *bi = rgd->rd_bits + x;
|
|
|
|
kfree(bi->bi_clone);
|
|
|
|
bi->bi_clone = NULL;
|
|
|
|
brelse(bi->bi_bh);
|
|
|
|
bi->bi_bh = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
}
|
|
|
|
|
2009-02-09 10:25:01 +01:00
|
|
|
static void gfs2_rgrp_send_discards(struct gfs2_sbd *sdp, u64 offset,
|
|
|
|
const struct gfs2_bitmap *bi)
|
|
|
|
{
|
|
|
|
struct super_block *sb = sdp->sd_vfs;
|
|
|
|
struct block_device *bdev = sb->s_bdev;
|
|
|
|
const unsigned int sects_per_blk = sdp->sd_sb.sb_bsize /
|
|
|
|
bdev_hardsect_size(sb->s_bdev);
|
|
|
|
u64 blk;
|
GFS2: Add a "demote a glock" interface to sysfs
This adds a sysfs file called demote_rq to GFS2's
per filesystem directory. Its possible to use this
file to demote arbitrary glocks in exactly the same
way as if a request had come in from a remote node.
This is intended for testing issues relating to caching
of data under glocks. Despite that, the interface is
generic enough to send requests to any type of glock,
but be careful as its not always safe to send an
arbitrary message to an arbitrary glock. For that reason
and to prevent DoS, this interface is restricted to root
only.
The messages look like this:
<type>:<glocknumber> <mode>
Example:
echo -n "2:13324 EX" >/sys/fs/gfs2/unity:myfs/demote_rq
Which means "please demote inode glock (type 2) number 13324 so that
I can get an EX (exclusive) lock". The lock modes are those which
would normally be sent by a remote node in its callback so if you
want to unlock a glock, you use EX, to demote to shared, use SH or PR
(depending on whether you like GFS2 or DLM lock modes better!).
If the glock doesn't exist, you'll get -ENOENT returned. If the
arguments don't make sense, you'll get -EINVAL returned.
The plan is that this interface will be used in combination with
the blktrace patch which I recently posted for comments although
it is, of course, still useful in its own right.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2009-02-12 14:31:58 +01:00
|
|
|
sector_t start = 0;
|
2009-02-09 10:25:01 +01:00
|
|
|
sector_t nr_sects = 0;
|
|
|
|
int rv;
|
|
|
|
unsigned int x;
|
|
|
|
|
|
|
|
for (x = 0; x < bi->bi_len; x++) {
|
|
|
|
const u8 *orig = bi->bi_bh->b_data + bi->bi_offset + x;
|
|
|
|
const u8 *clone = bi->bi_clone + bi->bi_offset + x;
|
|
|
|
u8 diff = ~(*orig | (*orig >> 1)) & (*clone | (*clone >> 1));
|
|
|
|
diff &= 0x55;
|
|
|
|
if (diff == 0)
|
|
|
|
continue;
|
|
|
|
blk = offset + ((bi->bi_start + x) * GFS2_NBBY);
|
|
|
|
blk *= sects_per_blk; /* convert to sectors */
|
|
|
|
while(diff) {
|
|
|
|
if (diff & 1) {
|
|
|
|
if (nr_sects == 0)
|
|
|
|
goto start_new_extent;
|
|
|
|
if ((start + nr_sects) != blk) {
|
|
|
|
rv = blkdev_issue_discard(bdev, start,
|
|
|
|
nr_sects, GFP_NOFS);
|
|
|
|
if (rv)
|
|
|
|
goto fail;
|
|
|
|
nr_sects = 0;
|
|
|
|
start_new_extent:
|
|
|
|
start = blk;
|
|
|
|
}
|
|
|
|
nr_sects += sects_per_blk;
|
|
|
|
}
|
|
|
|
diff >>= 2;
|
|
|
|
blk += sects_per_blk;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (nr_sects) {
|
|
|
|
rv = blkdev_issue_discard(bdev, start, nr_sects, GFP_NOFS);
|
|
|
|
if (rv)
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
return;
|
|
|
|
fail:
|
|
|
|
fs_warn(sdp, "error %d on discard request, turning discards off for this filesystem", rv);
|
|
|
|
sdp->sd_args.ar_discard = 0;
|
|
|
|
}
|
|
|
|
|
2006-01-16 17:50:04 +01:00
|
|
|
void gfs2_rgrp_repolish_clones(struct gfs2_rgrpd *rgd)
|
|
|
|
{
|
|
|
|
struct gfs2_sbd *sdp = rgd->rd_sbd;
|
2007-06-01 15:11:58 +02:00
|
|
|
unsigned int length = rgd->rd_length;
|
2006-01-16 17:50:04 +01:00
|
|
|
unsigned int x;
|
|
|
|
|
|
|
|
for (x = 0; x < length; x++) {
|
|
|
|
struct gfs2_bitmap *bi = rgd->rd_bits + x;
|
|
|
|
if (!bi->bi_clone)
|
|
|
|
continue;
|
2009-02-09 10:25:01 +01:00
|
|
|
if (sdp->sd_args.ar_discard)
|
|
|
|
gfs2_rgrp_send_discards(sdp, rgd->rd_data0, bi);
|
2006-01-16 17:50:04 +01:00
|
|
|
memcpy(bi->bi_clone + bi->bi_offset,
|
2006-06-14 21:32:57 +02:00
|
|
|
bi->bi_bh->b_data + bi->bi_offset, bi->bi_len);
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
2008-11-04 11:25:13 +01:00
|
|
|
rgd->rd_free_clone = rgd->rd_free;
|
2006-01-16 17:50:04 +01:00
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_alloc_get - get the struct gfs2_alloc structure for an inode
|
|
|
|
* @ip: the incore GFS2 inode structure
|
|
|
|
*
|
|
|
|
* Returns: the struct gfs2_alloc
|
|
|
|
*/
|
|
|
|
|
|
|
|
struct gfs2_alloc *gfs2_alloc_get(struct gfs2_inode *ip)
|
|
|
|
{
|
2008-01-10 16:18:55 +01:00
|
|
|
BUG_ON(ip->i_alloc != NULL);
|
|
|
|
ip->i_alloc = kzalloc(sizeof(struct gfs2_alloc), GFP_KERNEL);
|
|
|
|
return ip->i_alloc;
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* try_rgrp_fit - See if a given reservation will fit in a given RG
|
|
|
|
* @rgd: the RG data
|
|
|
|
* @al: the struct gfs2_alloc structure describing the reservation
|
|
|
|
*
|
|
|
|
* If there's room for the requested blocks to be allocated from the RG:
|
|
|
|
* Sets the $al_rgd field in @al.
|
|
|
|
*
|
|
|
|
* Returns: 1 on success (it fits), 0 on failure (it doesn't fit)
|
|
|
|
*/
|
|
|
|
|
|
|
|
static int try_rgrp_fit(struct gfs2_rgrpd *rgd, struct gfs2_alloc *al)
|
|
|
|
{
|
|
|
|
struct gfs2_sbd *sdp = rgd->rd_sbd;
|
|
|
|
int ret = 0;
|
|
|
|
|
2008-01-29 01:38:07 +01:00
|
|
|
if (rgd->rd_flags & GFS2_RDF_NOALLOC)
|
2007-04-02 11:48:17 +02:00
|
|
|
return 0;
|
|
|
|
|
2006-01-16 17:50:04 +01:00
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
|
|
|
if (rgd->rd_free_clone >= al->al_requested) {
|
|
|
|
al->al_rgd = rgd;
|
|
|
|
ret = 1;
|
|
|
|
}
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2007-06-08 11:05:33 +02:00
|
|
|
/**
|
|
|
|
* try_rgrp_unlink - Look for any unlinked, allocated, but unused inodes
|
|
|
|
* @rgd: The rgrp
|
|
|
|
*
|
|
|
|
* Returns: The inode, if one has been found
|
|
|
|
*/
|
|
|
|
|
|
|
|
static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked)
|
|
|
|
{
|
|
|
|
struct inode *inode;
|
2007-07-24 21:09:32 +02:00
|
|
|
u32 goal = 0, block;
|
2007-06-27 23:07:08 +02:00
|
|
|
u64 no_addr;
|
2007-08-08 23:52:09 +02:00
|
|
|
struct gfs2_sbd *sdp = rgd->rd_sbd;
|
2008-02-06 11:11:15 +01:00
|
|
|
unsigned int n;
|
2007-06-08 11:05:33 +02:00
|
|
|
|
|
|
|
for(;;) {
|
2007-07-12 23:58:50 +02:00
|
|
|
if (goal >= rgd->rd_data)
|
|
|
|
break;
|
2007-08-08 23:52:09 +02:00
|
|
|
down_write(&sdp->sd_log_flush_lock);
|
2008-02-06 11:11:15 +01:00
|
|
|
n = 1;
|
2007-07-24 21:09:32 +02:00
|
|
|
block = rgblk_search(rgd, goal, GFS2_BLKST_UNLINKED,
|
2008-02-06 11:11:15 +01:00
|
|
|
GFS2_BLKST_UNLINKED, &n);
|
2007-08-08 23:52:09 +02:00
|
|
|
up_write(&sdp->sd_log_flush_lock);
|
2007-07-24 21:09:32 +02:00
|
|
|
if (block == BFITNOENT)
|
2007-07-12 23:58:50 +02:00
|
|
|
break;
|
2007-07-24 21:09:32 +02:00
|
|
|
/* rgblk_search can return a block < goal, so we need to
|
|
|
|
keep it marching forward. */
|
|
|
|
no_addr = block + rgd->rd_data0;
|
2007-07-12 23:58:50 +02:00
|
|
|
goal++;
|
2007-07-24 21:09:32 +02:00
|
|
|
if (*last_unlinked != NO_BLOCK && no_addr <= *last_unlinked)
|
2007-06-08 11:05:33 +02:00
|
|
|
continue;
|
2007-06-27 23:07:08 +02:00
|
|
|
*last_unlinked = no_addr;
|
|
|
|
inode = gfs2_inode_lookup(rgd->rd_sbd->sd_vfs, DT_UNKNOWN,
|
[GFS2] Alternate gfs2_iget to avoid looking up inodes being freed
There is a possible deadlock between two processes on the same node, where one
process is deleting an inode, and another process is looking for allocated but
unused inodes to delete in order to create more space.
process A does an iput() on inode X, and it's i_count drops to 0. This causes
iput_final() to be called, which puts an inode into state I_FREEING at
generic_delete_inode(). There no point between when iput_final() is called, and
when I_FREEING is set where GFS2 could acquire any glocks. Once I_FREEING is
set, no other process on that node can successfully look up that inode until
the delete finishes.
process B locks the the resource group for the same inode in get_local_rgrp(),
which is called by gfs2_inplace_reserve_i()
process A tries to lock the resource group for the inode in
gfs2_dinode_dealloc(), but it's already locked by process B
process B waits in find_inode for the inode to have the I_FREEING state cleared.
Deadlock.
This patch solves the problem by adding an alternative to gfs2_iget(),
gfs2_iget_skip(), that simply skips any inodes that are in the I_FREEING
state.o The alternate test function is just like the original one, except that
it fails if the inode is being freed, and sets a skipped flag. The alternate
set function is just like the original, except that it fails if the skipped
flag is set. Only try_rgrp_unlink() calls gfs2_iget_skip() instead of
gfs2_iget().
Signed-off-by: Benjamin E. Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-09-18 20:33:18 +02:00
|
|
|
no_addr, -1, 1);
|
2007-06-08 11:05:33 +02:00
|
|
|
if (!IS_ERR(inode))
|
|
|
|
return inode;
|
|
|
|
}
|
|
|
|
|
|
|
|
rgd->rd_flags &= ~GFS2_RDF_CHECK;
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2006-01-16 17:50:04 +01:00
|
|
|
/**
|
|
|
|
* recent_rgrp_next - get next RG from "recent" list
|
|
|
|
* @cur_rgd: current rgrp
|
|
|
|
*
|
|
|
|
* Returns: The next rgrp in the recent list
|
|
|
|
*/
|
|
|
|
|
2008-07-10 16:54:12 +02:00
|
|
|
static struct gfs2_rgrpd *recent_rgrp_next(struct gfs2_rgrpd *cur_rgd)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
|
|
|
struct gfs2_sbd *sdp = cur_rgd->rd_sbd;
|
|
|
|
struct list_head *head;
|
|
|
|
struct gfs2_rgrpd *rgd;
|
|
|
|
|
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
2008-07-10 16:54:12 +02:00
|
|
|
head = &sdp->sd_rindex_mru_list;
|
|
|
|
if (unlikely(cur_rgd->rd_list_mru.next == head)) {
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
return NULL;
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
2008-07-10 16:54:12 +02:00
|
|
|
rgd = list_entry(cur_rgd->rd_list_mru.next, struct gfs2_rgrpd, rd_list_mru);
|
2006-01-16 17:50:04 +01:00
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
return rgd;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* forward_rgrp_get - get an rgrp to try next from full list
|
|
|
|
* @sdp: The GFS2 superblock
|
|
|
|
*
|
|
|
|
* Returns: The rgrp to try next
|
|
|
|
*/
|
|
|
|
|
|
|
|
static struct gfs2_rgrpd *forward_rgrp_get(struct gfs2_sbd *sdp)
|
|
|
|
{
|
|
|
|
struct gfs2_rgrpd *rgd;
|
|
|
|
unsigned int journals = gfs2_jindex_size(sdp);
|
|
|
|
unsigned int rg = 0, x;
|
|
|
|
|
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
|
|
|
|
|
|
|
rgd = sdp->sd_rindex_forward;
|
|
|
|
if (!rgd) {
|
|
|
|
if (sdp->sd_rgrps >= journals)
|
|
|
|
rg = sdp->sd_rgrps * sdp->sd_jdesc->jd_jid / journals;
|
|
|
|
|
2006-08-22 22:25:50 +02:00
|
|
|
for (x = 0, rgd = gfs2_rgrpd_get_first(sdp); x < rg;
|
2006-01-16 17:50:04 +01:00
|
|
|
x++, rgd = gfs2_rgrpd_get_next(rgd))
|
|
|
|
/* Do Nothing */;
|
|
|
|
|
|
|
|
sdp->sd_rindex_forward = rgd;
|
|
|
|
}
|
|
|
|
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
|
|
|
|
return rgd;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* forward_rgrp_set - set the forward rgrp pointer
|
|
|
|
* @sdp: the filesystem
|
|
|
|
* @rgd: The new forward rgrp
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
|
|
|
static void forward_rgrp_set(struct gfs2_sbd *sdp, struct gfs2_rgrpd *rgd)
|
|
|
|
{
|
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
|
|
|
sdp->sd_rindex_forward = rgd;
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* get_local_rgrp - Choose and lock a rgrp for allocation
|
|
|
|
* @ip: the inode to reserve space for
|
|
|
|
* @rgp: the chosen and locked rgrp
|
|
|
|
*
|
|
|
|
* Try to acquire rgrp in way which avoids contending with others.
|
|
|
|
*
|
|
|
|
* Returns: errno
|
|
|
|
*/
|
|
|
|
|
2007-06-08 11:05:33 +02:00
|
|
|
static struct inode *get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
2007-06-08 11:05:33 +02:00
|
|
|
struct inode *inode = NULL;
|
2006-06-14 21:32:57 +02:00
|
|
|
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
|
2006-01-16 17:50:04 +01:00
|
|
|
struct gfs2_rgrpd *rgd, *begin = NULL;
|
2008-01-10 16:18:55 +01:00
|
|
|
struct gfs2_alloc *al = ip->i_alloc;
|
2006-01-16 17:50:04 +01:00
|
|
|
int flags = LM_FLAG_TRY;
|
|
|
|
int skipped = 0;
|
|
|
|
int loops = 0;
|
2007-11-29 21:13:54 +01:00
|
|
|
int error, rg_locked;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2008-07-10 16:54:12 +02:00
|
|
|
rgd = gfs2_blk2rgrpd(sdp, ip->i_goal);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
while (rgd) {
|
2007-11-29 21:13:54 +01:00
|
|
|
rg_locked = 0;
|
|
|
|
|
|
|
|
if (gfs2_glock_is_locked_by_me(rgd->rd_gl)) {
|
|
|
|
rg_locked = 1;
|
|
|
|
error = 0;
|
|
|
|
} else {
|
|
|
|
error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE,
|
|
|
|
LM_FLAG_TRY, &al->al_rgd_gh);
|
|
|
|
}
|
2006-01-16 17:50:04 +01:00
|
|
|
switch (error) {
|
|
|
|
case 0:
|
|
|
|
if (try_rgrp_fit(rgd, al))
|
|
|
|
goto out;
|
2007-06-08 11:05:33 +02:00
|
|
|
if (rgd->rd_flags & GFS2_RDF_CHECK)
|
|
|
|
inode = try_rgrp_unlink(rgd, last_unlinked);
|
2007-11-29 21:13:54 +01:00
|
|
|
if (!rg_locked)
|
|
|
|
gfs2_glock_dq_uninit(&al->al_rgd_gh);
|
2007-06-08 11:05:33 +02:00
|
|
|
if (inode)
|
|
|
|
return inode;
|
2008-07-10 16:54:12 +02:00
|
|
|
/* fall through */
|
2006-01-16 17:50:04 +01:00
|
|
|
case GLR_TRYFAILED:
|
2008-07-10 16:54:12 +02:00
|
|
|
rgd = recent_rgrp_next(rgd);
|
2006-01-16 17:50:04 +01:00
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
2007-06-08 11:05:33 +02:00
|
|
|
return ERR_PTR(error);
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Go through full list of rgrps */
|
|
|
|
|
|
|
|
begin = rgd = forward_rgrp_get(sdp);
|
|
|
|
|
|
|
|
for (;;) {
|
2007-11-29 21:13:54 +01:00
|
|
|
rg_locked = 0;
|
|
|
|
|
|
|
|
if (gfs2_glock_is_locked_by_me(rgd->rd_gl)) {
|
|
|
|
rg_locked = 1;
|
|
|
|
error = 0;
|
|
|
|
} else {
|
|
|
|
error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, flags,
|
|
|
|
&al->al_rgd_gh);
|
|
|
|
}
|
2006-01-16 17:50:04 +01:00
|
|
|
switch (error) {
|
|
|
|
case 0:
|
|
|
|
if (try_rgrp_fit(rgd, al))
|
|
|
|
goto out;
|
2007-06-08 11:05:33 +02:00
|
|
|
if (rgd->rd_flags & GFS2_RDF_CHECK)
|
|
|
|
inode = try_rgrp_unlink(rgd, last_unlinked);
|
2007-11-29 21:13:54 +01:00
|
|
|
if (!rg_locked)
|
|
|
|
gfs2_glock_dq_uninit(&al->al_rgd_gh);
|
2007-06-08 11:05:33 +02:00
|
|
|
if (inode)
|
|
|
|
return inode;
|
2006-01-16 17:50:04 +01:00
|
|
|
break;
|
|
|
|
|
|
|
|
case GLR_TRYFAILED:
|
|
|
|
skipped++;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
2007-06-08 11:05:33 +02:00
|
|
|
return ERR_PTR(error);
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
rgd = gfs2_rgrpd_get_next(rgd);
|
|
|
|
if (!rgd)
|
|
|
|
rgd = gfs2_rgrpd_get_first(sdp);
|
|
|
|
|
|
|
|
if (rgd == begin) {
|
[GFS2] flush the log if a transaction can't allocate space
This is a fix for bz #208514. When GFS2 frees up space, the freed blocks
aren't available for reuse until the resource group is successfully written
to the ondisk journal. So in rare cases, GFS2 operations will fail, saying
that the filesystem is out of space, when in reality, you are just waiting for
a log flush. For instance, on a 1Gig filesystem, if I continually write 10 Mb
to a file, and then truncate it, after a hundred interations, the write will
fail with -ENOSPC, even though the filesystem is just 1% full.
The attached patch calls a log flush in these cases. I tested this patch
fairly heavily to check if there were any locking issues that I missed, and
it seems to work just fine. Also, this patch only does the log flush if
get_local_rgrp makes a complete loop of resource groups without skipping
any do to locking issues. The code would be slightly simpler if it just always
did the log flush after the first failed pass, and you could only ever have
to go through the loop twice, instead of up to three times. However, I guessed
that failing to find a rg simply do to locking issues would be common enough
to skip the log flush in that case, but I'm not certain that this is the right
way to go. Either way, I don't suppose this code will be hit all that often.
Signed-off-by: Benjamin E. Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-03-23 21:51:56 +01:00
|
|
|
if (++loops >= 3)
|
2007-06-08 11:05:33 +02:00
|
|
|
return ERR_PTR(-ENOSPC);
|
[GFS2] flush the log if a transaction can't allocate space
This is a fix for bz #208514. When GFS2 frees up space, the freed blocks
aren't available for reuse until the resource group is successfully written
to the ondisk journal. So in rare cases, GFS2 operations will fail, saying
that the filesystem is out of space, when in reality, you are just waiting for
a log flush. For instance, on a 1Gig filesystem, if I continually write 10 Mb
to a file, and then truncate it, after a hundred interations, the write will
fail with -ENOSPC, even though the filesystem is just 1% full.
The attached patch calls a log flush in these cases. I tested this patch
fairly heavily to check if there were any locking issues that I missed, and
it seems to work just fine. Also, this patch only does the log flush if
get_local_rgrp makes a complete loop of resource groups without skipping
any do to locking issues. The code would be slightly simpler if it just always
did the log flush after the first failed pass, and you could only ever have
to go through the loop twice, instead of up to three times. However, I guessed
that failing to find a rg simply do to locking issues would be common enough
to skip the log flush in that case, but I'm not certain that this is the right
way to go. Either way, I don't suppose this code will be hit all that often.
Signed-off-by: Benjamin E. Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-03-23 21:51:56 +01:00
|
|
|
if (!skipped)
|
|
|
|
loops++;
|
2006-01-16 17:50:04 +01:00
|
|
|
flags = 0;
|
[GFS2] flush the log if a transaction can't allocate space
This is a fix for bz #208514. When GFS2 frees up space, the freed blocks
aren't available for reuse until the resource group is successfully written
to the ondisk journal. So in rare cases, GFS2 operations will fail, saying
that the filesystem is out of space, when in reality, you are just waiting for
a log flush. For instance, on a 1Gig filesystem, if I continually write 10 Mb
to a file, and then truncate it, after a hundred interations, the write will
fail with -ENOSPC, even though the filesystem is just 1% full.
The attached patch calls a log flush in these cases. I tested this patch
fairly heavily to check if there were any locking issues that I missed, and
it seems to work just fine. Also, this patch only does the log flush if
get_local_rgrp makes a complete loop of resource groups without skipping
any do to locking issues. The code would be slightly simpler if it just always
did the log flush after the first failed pass, and you could only ever have
to go through the loop twice, instead of up to three times. However, I guessed
that failing to find a rg simply do to locking issues would be common enough
to skip the log flush in that case, but I'm not certain that this is the right
way to go. Either way, I don't suppose this code will be hit all that often.
Signed-off-by: Benjamin E. Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-03-23 21:51:56 +01:00
|
|
|
if (loops == 2)
|
|
|
|
gfs2_log_flush(sdp, NULL);
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2006-06-14 21:32:57 +02:00
|
|
|
out:
|
2006-01-16 17:50:04 +01:00
|
|
|
if (begin) {
|
2008-07-10 16:54:12 +02:00
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
|
|
|
list_move(&rgd->rd_list_mru, &sdp->sd_rindex_mru_list);
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
2006-01-16 17:50:04 +01:00
|
|
|
rgd = gfs2_rgrpd_get_next(rgd);
|
|
|
|
if (!rgd)
|
|
|
|
rgd = gfs2_rgrpd_get_first(sdp);
|
|
|
|
forward_rgrp_set(sdp, rgd);
|
|
|
|
}
|
|
|
|
|
2007-06-08 11:05:33 +02:00
|
|
|
return NULL;
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_inplace_reserve_i - Reserve space in the filesystem
|
|
|
|
* @ip: the inode to reserve space for
|
|
|
|
*
|
|
|
|
* Returns: errno
|
|
|
|
*/
|
|
|
|
|
|
|
|
int gfs2_inplace_reserve_i(struct gfs2_inode *ip, char *file, unsigned int line)
|
|
|
|
{
|
2006-06-14 21:32:57 +02:00
|
|
|
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
|
2008-01-10 16:18:55 +01:00
|
|
|
struct gfs2_alloc *al = ip->i_alloc;
|
2007-06-08 11:05:33 +02:00
|
|
|
struct inode *inode;
|
[GFS2] kernel changes to support new gfs2_grow command
This is another revision of my gfs2 kernel patch that allows
gfs2_grow to function properly.
Steve Whitehouse expressed some concerns about the previous
patch and I restructured it based on his comments.
The previous patch was doing the statfs_change at file close time,
under its own transaction. The current patch does the statfs_change
inside the gfs2_commit_write function, which keeps it under the
umbrella of the inode transaction.
I can't call ri_update to re-read the rindex file during the
transaction because the transaction may have outstanding unwritten
buffers attached to the rgrps that would be otherwise blown away.
So instead, I created a new function, gfs2_ri_total, that will
re-read the rindex file just to total the file system space
for the sake of the statfs_change. The ri_update will happen
later, when gfs2 realizes the version number has changed, as it
happened before my patch.
Since the statfs_change is happening at write_commit time and there
may be multiple writes to the rindex file for one grow operation.
So one consequence of this restructuring is that instead of getting
one kernel message to indicate the change, you may see several.
For example, before when you did a gfs2_grow, you'd get a single
message like:
GFS2: File system extended by 247876 blocks (968MB)
Now you get something like:
GFS2: File system extended by 207896 blocks (812MB)
GFS2: File system extended by 39980 blocks (156MB)
This version has also been successfully run against the hours-long
"gfs2_fsck_hellfire" test that does several gfs2_grow and gfs2_fsck
while interjecting file system damage. It does this repeatedly
under a variety Resource Group conditions.
Signed-off-By: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-05-09 16:37:57 +02:00
|
|
|
int error = 0;
|
2007-07-24 21:09:32 +02:00
|
|
|
u64 last_unlinked = NO_BLOCK;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
if (gfs2_assert_warn(sdp, al->al_requested))
|
|
|
|
return -EINVAL;
|
|
|
|
|
2007-06-08 11:05:33 +02:00
|
|
|
try_again:
|
[GFS2] kernel changes to support new gfs2_grow command
This is another revision of my gfs2 kernel patch that allows
gfs2_grow to function properly.
Steve Whitehouse expressed some concerns about the previous
patch and I restructured it based on his comments.
The previous patch was doing the statfs_change at file close time,
under its own transaction. The current patch does the statfs_change
inside the gfs2_commit_write function, which keeps it under the
umbrella of the inode transaction.
I can't call ri_update to re-read the rindex file during the
transaction because the transaction may have outstanding unwritten
buffers attached to the rgrps that would be otherwise blown away.
So instead, I created a new function, gfs2_ri_total, that will
re-read the rindex file just to total the file system space
for the sake of the statfs_change. The ri_update will happen
later, when gfs2 realizes the version number has changed, as it
happened before my patch.
Since the statfs_change is happening at write_commit time and there
may be multiple writes to the rindex file for one grow operation.
So one consequence of this restructuring is that instead of getting
one kernel message to indicate the change, you may see several.
For example, before when you did a gfs2_grow, you'd get a single
message like:
GFS2: File system extended by 247876 blocks (968MB)
Now you get something like:
GFS2: File system extended by 207896 blocks (812MB)
GFS2: File system extended by 39980 blocks (156MB)
This version has also been successfully run against the hours-long
"gfs2_fsck_hellfire" test that does several gfs2_grow and gfs2_fsck
while interjecting file system damage. It does this repeatedly
under a variety Resource Group conditions.
Signed-off-By: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-05-09 16:37:57 +02:00
|
|
|
/* We need to hold the rindex unless the inode we're using is
|
|
|
|
the rindex itself, in which case it's already held. */
|
|
|
|
if (ip != GFS2_I(sdp->sd_rindex))
|
|
|
|
error = gfs2_rindex_hold(sdp, &al->al_ri_gh);
|
|
|
|
else if (!sdp->sd_rgrps) /* We may not have the rindex read in, so: */
|
2007-05-10 23:54:38 +02:00
|
|
|
error = gfs2_ri_update_special(ip);
|
[GFS2] kernel changes to support new gfs2_grow command
This is another revision of my gfs2 kernel patch that allows
gfs2_grow to function properly.
Steve Whitehouse expressed some concerns about the previous
patch and I restructured it based on his comments.
The previous patch was doing the statfs_change at file close time,
under its own transaction. The current patch does the statfs_change
inside the gfs2_commit_write function, which keeps it under the
umbrella of the inode transaction.
I can't call ri_update to re-read the rindex file during the
transaction because the transaction may have outstanding unwritten
buffers attached to the rgrps that would be otherwise blown away.
So instead, I created a new function, gfs2_ri_total, that will
re-read the rindex file just to total the file system space
for the sake of the statfs_change. The ri_update will happen
later, when gfs2 realizes the version number has changed, as it
happened before my patch.
Since the statfs_change is happening at write_commit time and there
may be multiple writes to the rindex file for one grow operation.
So one consequence of this restructuring is that instead of getting
one kernel message to indicate the change, you may see several.
For example, before when you did a gfs2_grow, you'd get a single
message like:
GFS2: File system extended by 247876 blocks (968MB)
Now you get something like:
GFS2: File system extended by 207896 blocks (812MB)
GFS2: File system extended by 39980 blocks (156MB)
This version has also been successfully run against the hours-long
"gfs2_fsck_hellfire" test that does several gfs2_grow and gfs2_fsck
while interjecting file system damage. It does this repeatedly
under a variety Resource Group conditions.
Signed-off-By: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-05-09 16:37:57 +02:00
|
|
|
|
2006-01-16 17:50:04 +01:00
|
|
|
if (error)
|
|
|
|
return error;
|
|
|
|
|
2007-06-08 11:05:33 +02:00
|
|
|
inode = get_local_rgrp(ip, &last_unlinked);
|
|
|
|
if (inode) {
|
[GFS2] kernel changes to support new gfs2_grow command
This is another revision of my gfs2 kernel patch that allows
gfs2_grow to function properly.
Steve Whitehouse expressed some concerns about the previous
patch and I restructured it based on his comments.
The previous patch was doing the statfs_change at file close time,
under its own transaction. The current patch does the statfs_change
inside the gfs2_commit_write function, which keeps it under the
umbrella of the inode transaction.
I can't call ri_update to re-read the rindex file during the
transaction because the transaction may have outstanding unwritten
buffers attached to the rgrps that would be otherwise blown away.
So instead, I created a new function, gfs2_ri_total, that will
re-read the rindex file just to total the file system space
for the sake of the statfs_change. The ri_update will happen
later, when gfs2 realizes the version number has changed, as it
happened before my patch.
Since the statfs_change is happening at write_commit time and there
may be multiple writes to the rindex file for one grow operation.
So one consequence of this restructuring is that instead of getting
one kernel message to indicate the change, you may see several.
For example, before when you did a gfs2_grow, you'd get a single
message like:
GFS2: File system extended by 247876 blocks (968MB)
Now you get something like:
GFS2: File system extended by 207896 blocks (812MB)
GFS2: File system extended by 39980 blocks (156MB)
This version has also been successfully run against the hours-long
"gfs2_fsck_hellfire" test that does several gfs2_grow and gfs2_fsck
while interjecting file system damage. It does this repeatedly
under a variety Resource Group conditions.
Signed-off-By: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-05-09 16:37:57 +02:00
|
|
|
if (ip != GFS2_I(sdp->sd_rindex))
|
|
|
|
gfs2_glock_dq_uninit(&al->al_ri_gh);
|
2007-06-08 11:05:33 +02:00
|
|
|
if (IS_ERR(inode))
|
|
|
|
return PTR_ERR(inode);
|
|
|
|
iput(inode);
|
|
|
|
gfs2_log_flush(sdp, NULL);
|
|
|
|
goto try_again;
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
al->al_file = file;
|
|
|
|
al->al_line = line;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_inplace_release - release an inplace reservation
|
|
|
|
* @ip: the inode the reservation was taken out on
|
|
|
|
*
|
|
|
|
* Release a reservation made by gfs2_inplace_reserve().
|
|
|
|
*/
|
|
|
|
|
|
|
|
void gfs2_inplace_release(struct gfs2_inode *ip)
|
|
|
|
{
|
2006-06-14 21:32:57 +02:00
|
|
|
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
|
2008-01-10 16:18:55 +01:00
|
|
|
struct gfs2_alloc *al = ip->i_alloc;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
if (gfs2_assert_warn(sdp, al->al_alloced <= al->al_requested) == -1)
|
|
|
|
fs_warn(sdp, "al_alloced = %u, al_requested = %u "
|
|
|
|
"al_file = %s, al_line = %u\n",
|
|
|
|
al->al_alloced, al->al_requested, al->al_file,
|
|
|
|
al->al_line);
|
|
|
|
|
|
|
|
al->al_rgd = NULL;
|
2007-11-29 21:13:54 +01:00
|
|
|
if (al->al_rgd_gh.gh_gl)
|
|
|
|
gfs2_glock_dq_uninit(&al->al_rgd_gh);
|
[GFS2] kernel changes to support new gfs2_grow command
This is another revision of my gfs2 kernel patch that allows
gfs2_grow to function properly.
Steve Whitehouse expressed some concerns about the previous
patch and I restructured it based on his comments.
The previous patch was doing the statfs_change at file close time,
under its own transaction. The current patch does the statfs_change
inside the gfs2_commit_write function, which keeps it under the
umbrella of the inode transaction.
I can't call ri_update to re-read the rindex file during the
transaction because the transaction may have outstanding unwritten
buffers attached to the rgrps that would be otherwise blown away.
So instead, I created a new function, gfs2_ri_total, that will
re-read the rindex file just to total the file system space
for the sake of the statfs_change. The ri_update will happen
later, when gfs2 realizes the version number has changed, as it
happened before my patch.
Since the statfs_change is happening at write_commit time and there
may be multiple writes to the rindex file for one grow operation.
So one consequence of this restructuring is that instead of getting
one kernel message to indicate the change, you may see several.
For example, before when you did a gfs2_grow, you'd get a single
message like:
GFS2: File system extended by 247876 blocks (968MB)
Now you get something like:
GFS2: File system extended by 207896 blocks (812MB)
GFS2: File system extended by 39980 blocks (156MB)
This version has also been successfully run against the hours-long
"gfs2_fsck_hellfire" test that does several gfs2_grow and gfs2_fsck
while interjecting file system damage. It does this repeatedly
under a variety Resource Group conditions.
Signed-off-By: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2007-05-09 16:37:57 +02:00
|
|
|
if (ip != GFS2_I(sdp->sd_rindex))
|
|
|
|
gfs2_glock_dq_uninit(&al->al_ri_gh);
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_get_block_type - Check a block in a RG is of given type
|
|
|
|
* @rgd: the resource group holding the block
|
|
|
|
* @block: the block number
|
|
|
|
*
|
|
|
|
* Returns: The block type (GFS2_BLKST_*)
|
|
|
|
*/
|
|
|
|
|
2006-09-04 18:49:07 +02:00
|
|
|
unsigned char gfs2_get_block_type(struct gfs2_rgrpd *rgd, u64 block)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
|
|
|
struct gfs2_bitmap *bi = NULL;
|
2006-09-04 18:49:07 +02:00
|
|
|
u32 length, rgrp_block, buf_block;
|
2006-01-16 17:50:04 +01:00
|
|
|
unsigned int buf;
|
|
|
|
unsigned char type;
|
|
|
|
|
2007-06-01 15:11:58 +02:00
|
|
|
length = rgd->rd_length;
|
|
|
|
rgrp_block = block - rgd->rd_data0;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
for (buf = 0; buf < length; buf++) {
|
|
|
|
bi = rgd->rd_bits + buf;
|
|
|
|
if (rgrp_block < (bi->bi_start + bi->bi_len) * GFS2_NBBY)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
gfs2_assert(rgd->rd_sbd, buf < length);
|
|
|
|
buf_block = rgrp_block - bi->bi_start * GFS2_NBBY;
|
|
|
|
|
2006-06-14 21:32:57 +02:00
|
|
|
type = gfs2_testbit(rgd, bi->bi_bh->b_data + bi->bi_offset,
|
2006-01-16 17:50:04 +01:00
|
|
|
bi->bi_len, buf_block);
|
|
|
|
|
|
|
|
return type;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* rgblk_search - find a block in @old_state, change allocation
|
|
|
|
* state to @new_state
|
|
|
|
* @rgd: the resource group descriptor
|
|
|
|
* @goal: the goal block within the RG (start here to search for avail block)
|
|
|
|
* @old_state: GFS2_BLKST_XXX the before-allocation state to find
|
|
|
|
* @new_state: GFS2_BLKST_XXX the after-allocation block state
|
2008-02-06 11:11:15 +01:00
|
|
|
* @n: The extent length
|
2006-01-16 17:50:04 +01:00
|
|
|
*
|
|
|
|
* Walk rgrp's bitmap to find bits that represent a block in @old_state.
|
|
|
|
* Add the found bitmap buffer to the transaction.
|
|
|
|
* Set the found bits to @new_state to change block's allocation state.
|
|
|
|
*
|
|
|
|
* This function never fails, because we wouldn't call it unless we
|
|
|
|
* know (from reservation results, etc.) that a block is available.
|
|
|
|
*
|
|
|
|
* Scope of @goal and returned block is just within rgrp, not the whole
|
|
|
|
* filesystem.
|
|
|
|
*
|
|
|
|
* Returns: the block number allocated
|
|
|
|
*/
|
|
|
|
|
2006-09-04 18:49:07 +02:00
|
|
|
static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal,
|
2008-02-06 11:11:15 +01:00
|
|
|
unsigned char old_state, unsigned char new_state,
|
|
|
|
unsigned int *n)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
|
|
|
struct gfs2_bitmap *bi = NULL;
|
2008-02-06 11:11:15 +01:00
|
|
|
const u32 length = rgd->rd_length;
|
2006-09-04 18:49:07 +02:00
|
|
|
u32 blk = 0;
|
2006-01-16 17:50:04 +01:00
|
|
|
unsigned int buf, x;
|
2008-02-06 11:11:15 +01:00
|
|
|
const unsigned int elen = *n;
|
|
|
|
const u8 *buffer;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2008-02-06 11:11:15 +01:00
|
|
|
*n = 0;
|
2006-01-16 17:50:04 +01:00
|
|
|
/* Find bitmap block that contains bits for goal block */
|
|
|
|
for (buf = 0; buf < length; buf++) {
|
|
|
|
bi = rgd->rd_bits + buf;
|
|
|
|
if (goal < (bi->bi_start + bi->bi_len) * GFS2_NBBY)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
gfs2_assert(rgd->rd_sbd, buf < length);
|
|
|
|
|
|
|
|
/* Convert scope of "goal" from rgrp-wide to within found bit block */
|
|
|
|
goal -= bi->bi_start * GFS2_NBBY;
|
|
|
|
|
|
|
|
/* Search (up to entire) bitmap in this rgrp for allocatable block.
|
|
|
|
"x <= length", instead of "x < length", because we typically start
|
|
|
|
the search in the middle of a bit block, but if we can't find an
|
|
|
|
allocatable block anywhere else, we want to be able wrap around and
|
|
|
|
search in the first part of our first-searched bit block. */
|
|
|
|
for (x = 0; x <= length; x++) {
|
2007-08-08 23:52:09 +02:00
|
|
|
/* The GFS2_BLKST_UNLINKED state doesn't apply to the clone
|
|
|
|
bitmaps, so we must search the originals for that. */
|
2008-02-06 11:11:15 +01:00
|
|
|
buffer = bi->bi_bh->b_data + bi->bi_offset;
|
2007-08-08 23:52:09 +02:00
|
|
|
if (old_state != GFS2_BLKST_UNLINKED && bi->bi_clone)
|
2008-01-29 14:30:20 +01:00
|
|
|
buffer = bi->bi_clone + bi->bi_offset;
|
|
|
|
|
|
|
|
blk = gfs2_bitfit(buffer, bi->bi_len, goal, old_state);
|
2006-01-16 17:50:04 +01:00
|
|
|
if (blk != BFITNOENT)
|
|
|
|
break;
|
|
|
|
|
|
|
|
/* Try next bitmap block (wrap back to rgrp header if at end) */
|
|
|
|
buf = (buf + 1) % length;
|
|
|
|
bi = rgd->rd_bits + buf;
|
|
|
|
goal = 0;
|
|
|
|
}
|
|
|
|
|
2007-07-24 21:09:32 +02:00
|
|
|
if (blk != BFITNOENT && old_state != new_state) {
|
2008-02-06 11:11:15 +01:00
|
|
|
*n = 1;
|
2007-06-08 11:05:33 +02:00
|
|
|
gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1);
|
2008-02-06 11:11:15 +01:00
|
|
|
gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone, bi->bi_offset,
|
2006-01-16 17:50:04 +01:00
|
|
|
bi->bi_len, blk, new_state);
|
2008-02-22 17:09:31 +01:00
|
|
|
goal = blk;
|
|
|
|
while (*n < elen) {
|
2008-02-06 11:11:15 +01:00
|
|
|
goal++;
|
2008-02-22 17:09:31 +01:00
|
|
|
if (goal >= (bi->bi_len * GFS2_NBBY))
|
2008-02-06 11:11:15 +01:00
|
|
|
break;
|
|
|
|
if (gfs2_testbit(rgd, buffer, bi->bi_len, goal) !=
|
|
|
|
GFS2_BLKST_FREE)
|
|
|
|
break;
|
|
|
|
gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone,
|
2008-02-22 17:09:31 +01:00
|
|
|
bi->bi_offset, bi->bi_len, goal,
|
|
|
|
new_state);
|
|
|
|
(*n)++;
|
2008-02-06 11:11:15 +01:00
|
|
|
}
|
2007-06-08 11:05:33 +02:00
|
|
|
}
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2007-07-17 11:26:56 +02:00
|
|
|
return (blk == BFITNOENT) ? blk : (bi->bi_start * GFS2_NBBY) + blk;
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* rgblk_free - Change alloc state of given block(s)
|
|
|
|
* @sdp: the filesystem
|
|
|
|
* @bstart: the start of a run of blocks to free
|
|
|
|
* @blen: the length of the block run (all must lie within ONE RG!)
|
|
|
|
* @new_state: GFS2_BLKST_XXX the after-allocation block state
|
|
|
|
*
|
|
|
|
* Returns: Resource group containing the block(s)
|
|
|
|
*/
|
|
|
|
|
2006-09-04 18:49:07 +02:00
|
|
|
static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart,
|
|
|
|
u32 blen, unsigned char new_state)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
|
|
|
struct gfs2_rgrpd *rgd;
|
|
|
|
struct gfs2_bitmap *bi = NULL;
|
2006-09-04 18:49:07 +02:00
|
|
|
u32 length, rgrp_blk, buf_blk;
|
2006-01-16 17:50:04 +01:00
|
|
|
unsigned int buf;
|
|
|
|
|
|
|
|
rgd = gfs2_blk2rgrpd(sdp, bstart);
|
|
|
|
if (!rgd) {
|
|
|
|
if (gfs2_consist(sdp))
|
2006-05-24 16:22:09 +02:00
|
|
|
fs_err(sdp, "block = %llu\n", (unsigned long long)bstart);
|
2006-01-16 17:50:04 +01:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2007-06-01 15:11:58 +02:00
|
|
|
length = rgd->rd_length;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2007-06-01 15:11:58 +02:00
|
|
|
rgrp_blk = bstart - rgd->rd_data0;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
while (blen--) {
|
|
|
|
for (buf = 0; buf < length; buf++) {
|
|
|
|
bi = rgd->rd_bits + buf;
|
|
|
|
if (rgrp_blk < (bi->bi_start + bi->bi_len) * GFS2_NBBY)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
gfs2_assert(rgd->rd_sbd, buf < length);
|
|
|
|
|
|
|
|
buf_blk = rgrp_blk - bi->bi_start * GFS2_NBBY;
|
|
|
|
rgrp_blk++;
|
|
|
|
|
|
|
|
if (!bi->bi_clone) {
|
|
|
|
bi->bi_clone = kmalloc(bi->bi_bh->b_size,
|
2006-07-27 20:29:00 +02:00
|
|
|
GFP_NOFS | __GFP_NOFAIL);
|
2006-01-16 17:50:04 +01:00
|
|
|
memcpy(bi->bi_clone + bi->bi_offset,
|
|
|
|
bi->bi_bh->b_data + bi->bi_offset,
|
|
|
|
bi->bi_len);
|
|
|
|
}
|
2006-01-18 12:19:28 +01:00
|
|
|
gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1);
|
2008-02-06 11:11:15 +01:00
|
|
|
gfs2_setbit(rgd, bi->bi_bh->b_data, NULL, bi->bi_offset,
|
2006-01-16 17:50:04 +01:00
|
|
|
bi->bi_len, buf_blk, new_state);
|
|
|
|
}
|
|
|
|
|
|
|
|
return rgd;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2008-02-01 15:52:30 +01:00
|
|
|
* gfs2_alloc_block - Allocate a block
|
|
|
|
* @ip: the inode to allocate the block for
|
2006-01-16 17:50:04 +01:00
|
|
|
*
|
|
|
|
* Returns: the allocated block
|
|
|
|
*/
|
|
|
|
|
2008-02-06 11:11:15 +01:00
|
|
|
u64 gfs2_alloc_block(struct gfs2_inode *ip, unsigned int *n)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
2006-06-14 21:32:57 +02:00
|
|
|
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
|
2009-04-23 09:59:41 +02:00
|
|
|
struct buffer_head *dibh;
|
2008-01-10 16:18:55 +01:00
|
|
|
struct gfs2_alloc *al = ip->i_alloc;
|
2006-01-16 17:50:04 +01:00
|
|
|
struct gfs2_rgrpd *rgd = al->al_rgd;
|
2006-09-04 18:49:07 +02:00
|
|
|
u32 goal, blk;
|
|
|
|
u64 block;
|
2009-04-23 09:59:41 +02:00
|
|
|
int error;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2008-02-06 10:25:45 +01:00
|
|
|
if (rgrp_contains_block(rgd, ip->i_goal))
|
|
|
|
goal = ip->i_goal - rgd->rd_data0;
|
2006-01-16 17:50:04 +01:00
|
|
|
else
|
2008-02-01 11:34:15 +01:00
|
|
|
goal = rgd->rd_last_alloc;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2008-02-06 11:11:15 +01:00
|
|
|
blk = rgblk_search(rgd, goal, GFS2_BLKST_FREE, GFS2_BLKST_USED, n);
|
2007-07-17 11:26:56 +02:00
|
|
|
BUG_ON(blk == BFITNOENT);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2008-02-06 11:11:15 +01:00
|
|
|
rgd->rd_last_alloc = blk;
|
2007-06-01 15:11:58 +02:00
|
|
|
block = rgd->rd_data0 + blk;
|
2008-02-06 10:25:45 +01:00
|
|
|
ip->i_goal = block;
|
2009-04-23 09:59:41 +02:00
|
|
|
error = gfs2_meta_inode_buffer(ip, &dibh);
|
|
|
|
if (error == 0) {
|
|
|
|
struct gfs2_dinode *di = (struct gfs2_dinode *)dibh->b_data;
|
|
|
|
gfs2_trans_add_bh(ip->i_gl, dibh, 1);
|
|
|
|
di->di_goal_meta = di->di_goal_data = cpu_to_be64(ip->i_goal);
|
|
|
|
brelse(dibh);
|
|
|
|
}
|
2008-11-04 11:25:13 +01:00
|
|
|
gfs2_assert_withdraw(sdp, rgd->rd_free >= *n);
|
|
|
|
rgd->rd_free -= *n;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2006-01-18 12:19:28 +01:00
|
|
|
gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
|
2008-01-29 01:38:07 +01:00
|
|
|
gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2008-02-06 11:11:15 +01:00
|
|
|
al->al_alloced += *n;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2008-05-01 12:55:38 +02:00
|
|
|
gfs2_statfs_change(sdp, 0, -(s64)*n, 0);
|
2008-02-06 11:11:15 +01:00
|
|
|
gfs2_quota_change(ip, *n, ip->i_inode.i_uid, ip->i_inode.i_gid);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
2008-02-06 11:11:15 +01:00
|
|
|
rgd->rd_free_clone -= *n;
|
2006-01-16 17:50:04 +01:00
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
|
|
|
|
return block;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_alloc_di - Allocate a dinode
|
|
|
|
* @dip: the directory that the inode is going in
|
|
|
|
*
|
|
|
|
* Returns: the block allocated
|
|
|
|
*/
|
|
|
|
|
2006-07-11 15:46:33 +02:00
|
|
|
u64 gfs2_alloc_di(struct gfs2_inode *dip, u64 *generation)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
2006-06-14 21:32:57 +02:00
|
|
|
struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
|
2008-01-10 16:18:55 +01:00
|
|
|
struct gfs2_alloc *al = dip->i_alloc;
|
2006-01-16 17:50:04 +01:00
|
|
|
struct gfs2_rgrpd *rgd = al->al_rgd;
|
2006-07-11 15:46:33 +02:00
|
|
|
u32 blk;
|
|
|
|
u64 block;
|
2008-02-06 11:11:15 +01:00
|
|
|
unsigned int n = 1;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2008-02-01 11:34:15 +01:00
|
|
|
blk = rgblk_search(rgd, rgd->rd_last_alloc,
|
2008-02-06 11:11:15 +01:00
|
|
|
GFS2_BLKST_FREE, GFS2_BLKST_DINODE, &n);
|
2007-07-17 11:26:56 +02:00
|
|
|
BUG_ON(blk == BFITNOENT);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2008-02-01 11:34:15 +01:00
|
|
|
rgd->rd_last_alloc = blk;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2007-06-01 15:11:58 +02:00
|
|
|
block = rgd->rd_data0 + blk;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2008-11-04 11:25:13 +01:00
|
|
|
gfs2_assert_withdraw(sdp, rgd->rd_free);
|
|
|
|
rgd->rd_free--;
|
2008-11-04 11:32:57 +01:00
|
|
|
rgd->rd_dinodes++;
|
2008-11-04 11:19:03 +01:00
|
|
|
*generation = rgd->rd_igeneration++;
|
2006-01-18 12:19:28 +01:00
|
|
|
gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
|
2008-01-29 01:38:07 +01:00
|
|
|
gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
al->al_alloced++;
|
|
|
|
|
|
|
|
gfs2_statfs_change(sdp, 0, -1, +1);
|
2008-02-01 14:16:55 +01:00
|
|
|
gfs2_trans_add_unrevoke(sdp, block, 1);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
spin_lock(&sdp->sd_rindex_spin);
|
|
|
|
rgd->rd_free_clone--;
|
|
|
|
spin_unlock(&sdp->sd_rindex_spin);
|
|
|
|
|
|
|
|
return block;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_free_data - free a contiguous run of data block(s)
|
|
|
|
* @ip: the inode these blocks are being freed from
|
|
|
|
* @bstart: first block of a run of contiguous blocks
|
|
|
|
* @blen: the length of the block run
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2006-09-04 18:49:07 +02:00
|
|
|
void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
2006-06-14 21:32:57 +02:00
|
|
|
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
|
2006-01-16 17:50:04 +01:00
|
|
|
struct gfs2_rgrpd *rgd;
|
|
|
|
|
|
|
|
rgd = rgblk_free(sdp, bstart, blen, GFS2_BLKST_FREE);
|
|
|
|
if (!rgd)
|
|
|
|
return;
|
|
|
|
|
2008-11-04 11:25:13 +01:00
|
|
|
rgd->rd_free += blen;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2006-01-18 12:19:28 +01:00
|
|
|
gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
|
2008-01-29 01:38:07 +01:00
|
|
|
gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
gfs2_trans_add_rg(rgd);
|
|
|
|
|
|
|
|
gfs2_statfs_change(sdp, 0, +blen, 0);
|
2006-11-01 19:23:29 +01:00
|
|
|
gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid);
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_free_meta - free a contiguous run of data block(s)
|
|
|
|
* @ip: the inode these blocks are being freed from
|
|
|
|
* @bstart: first block of a run of contiguous blocks
|
|
|
|
* @blen: the length of the block run
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2006-09-04 18:49:07 +02:00
|
|
|
void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
2006-06-14 21:32:57 +02:00
|
|
|
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
|
2006-01-16 17:50:04 +01:00
|
|
|
struct gfs2_rgrpd *rgd;
|
|
|
|
|
|
|
|
rgd = rgblk_free(sdp, bstart, blen, GFS2_BLKST_FREE);
|
|
|
|
if (!rgd)
|
|
|
|
return;
|
|
|
|
|
2008-11-04 11:25:13 +01:00
|
|
|
rgd->rd_free += blen;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2006-01-18 12:19:28 +01:00
|
|
|
gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
|
2008-01-29 01:38:07 +01:00
|
|
|
gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
gfs2_trans_add_rg(rgd);
|
|
|
|
|
|
|
|
gfs2_statfs_change(sdp, 0, +blen, 0);
|
2006-11-01 19:23:29 +01:00
|
|
|
gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid);
|
2006-01-16 17:50:04 +01:00
|
|
|
gfs2_meta_wipe(ip, bstart, blen);
|
|
|
|
}
|
|
|
|
|
2006-06-14 21:32:57 +02:00
|
|
|
void gfs2_unlink_di(struct inode *inode)
|
|
|
|
{
|
|
|
|
struct gfs2_inode *ip = GFS2_I(inode);
|
|
|
|
struct gfs2_sbd *sdp = GFS2_SB(inode);
|
|
|
|
struct gfs2_rgrpd *rgd;
|
2007-05-15 16:37:50 +02:00
|
|
|
u64 blkno = ip->i_no_addr;
|
2006-06-14 21:32:57 +02:00
|
|
|
|
|
|
|
rgd = rgblk_free(sdp, blkno, 1, GFS2_BLKST_UNLINKED);
|
|
|
|
if (!rgd)
|
|
|
|
return;
|
|
|
|
gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
|
2008-01-29 01:38:07 +01:00
|
|
|
gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
|
2006-06-14 21:32:57 +02:00
|
|
|
gfs2_trans_add_rg(rgd);
|
|
|
|
}
|
|
|
|
|
2006-09-04 18:49:07 +02:00
|
|
|
static void gfs2_free_uninit_di(struct gfs2_rgrpd *rgd, u64 blkno)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
|
|
|
struct gfs2_sbd *sdp = rgd->rd_sbd;
|
|
|
|
struct gfs2_rgrpd *tmp_rgd;
|
|
|
|
|
|
|
|
tmp_rgd = rgblk_free(sdp, blkno, 1, GFS2_BLKST_FREE);
|
|
|
|
if (!tmp_rgd)
|
|
|
|
return;
|
|
|
|
gfs2_assert_withdraw(sdp, rgd == tmp_rgd);
|
|
|
|
|
2008-11-04 11:32:57 +01:00
|
|
|
if (!rgd->rd_dinodes)
|
2006-01-16 17:50:04 +01:00
|
|
|
gfs2_consist_rgrpd(rgd);
|
2008-11-04 11:32:57 +01:00
|
|
|
rgd->rd_dinodes--;
|
2008-11-04 11:25:13 +01:00
|
|
|
rgd->rd_free++;
|
2006-01-16 17:50:04 +01:00
|
|
|
|
2006-01-18 12:19:28 +01:00
|
|
|
gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
|
2008-01-29 01:38:07 +01:00
|
|
|
gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
gfs2_statfs_change(sdp, 0, +1, -1);
|
|
|
|
gfs2_trans_add_rg(rgd);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
void gfs2_free_di(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip)
|
|
|
|
{
|
2007-05-15 16:37:50 +02:00
|
|
|
gfs2_free_uninit_di(rgd, ip->i_no_addr);
|
2006-11-01 19:23:29 +01:00
|
|
|
gfs2_quota_change(ip, -1, ip->i_inode.i_uid, ip->i_inode.i_gid);
|
2007-05-15 16:37:50 +02:00
|
|
|
gfs2_meta_wipe(ip, ip->i_no_addr, 1);
|
2006-01-16 17:50:04 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_rlist_add - add a RG to a list of RGs
|
|
|
|
* @sdp: the filesystem
|
|
|
|
* @rlist: the list of resource groups
|
|
|
|
* @block: the block
|
|
|
|
*
|
|
|
|
* Figure out what RG a block belongs to and add that RG to the list
|
|
|
|
*
|
|
|
|
* FIXME: Don't use NOFAIL
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
|
|
|
void gfs2_rlist_add(struct gfs2_sbd *sdp, struct gfs2_rgrp_list *rlist,
|
2006-09-04 18:49:07 +02:00
|
|
|
u64 block)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
|
|
|
struct gfs2_rgrpd *rgd;
|
|
|
|
struct gfs2_rgrpd **tmp;
|
|
|
|
unsigned int new_space;
|
|
|
|
unsigned int x;
|
|
|
|
|
|
|
|
if (gfs2_assert_warn(sdp, !rlist->rl_ghs))
|
|
|
|
return;
|
|
|
|
|
|
|
|
rgd = gfs2_blk2rgrpd(sdp, block);
|
|
|
|
if (!rgd) {
|
|
|
|
if (gfs2_consist(sdp))
|
2006-05-24 16:22:09 +02:00
|
|
|
fs_err(sdp, "block = %llu\n", (unsigned long long)block);
|
2006-01-16 17:50:04 +01:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (x = 0; x < rlist->rl_rgrps; x++)
|
|
|
|
if (rlist->rl_rgd[x] == rgd)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (rlist->rl_rgrps == rlist->rl_space) {
|
|
|
|
new_space = rlist->rl_space + 10;
|
|
|
|
|
|
|
|
tmp = kcalloc(new_space, sizeof(struct gfs2_rgrpd *),
|
2006-07-27 20:29:00 +02:00
|
|
|
GFP_NOFS | __GFP_NOFAIL);
|
2006-01-16 17:50:04 +01:00
|
|
|
|
|
|
|
if (rlist->rl_rgd) {
|
|
|
|
memcpy(tmp, rlist->rl_rgd,
|
|
|
|
rlist->rl_space * sizeof(struct gfs2_rgrpd *));
|
|
|
|
kfree(rlist->rl_rgd);
|
|
|
|
}
|
|
|
|
|
|
|
|
rlist->rl_space = new_space;
|
|
|
|
rlist->rl_rgd = tmp;
|
|
|
|
}
|
|
|
|
|
|
|
|
rlist->rl_rgd[rlist->rl_rgrps++] = rgd;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_rlist_alloc - all RGs have been added to the rlist, now allocate
|
|
|
|
* and initialize an array of glock holders for them
|
|
|
|
* @rlist: the list of resource groups
|
|
|
|
* @state: the lock state to acquire the RG lock in
|
|
|
|
* @flags: the modifier flags for the holder structures
|
|
|
|
*
|
|
|
|
* FIXME: Don't use NOFAIL
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2008-01-28 18:13:02 +01:00
|
|
|
void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist, unsigned int state)
|
2006-01-16 17:50:04 +01:00
|
|
|
{
|
|
|
|
unsigned int x;
|
|
|
|
|
|
|
|
rlist->rl_ghs = kcalloc(rlist->rl_rgrps, sizeof(struct gfs2_holder),
|
2006-07-27 20:29:00 +02:00
|
|
|
GFP_NOFS | __GFP_NOFAIL);
|
2006-01-16 17:50:04 +01:00
|
|
|
for (x = 0; x < rlist->rl_rgrps; x++)
|
|
|
|
gfs2_holder_init(rlist->rl_rgd[x]->rd_gl,
|
2008-01-28 18:13:02 +01:00
|
|
|
state, 0,
|
2006-01-16 17:50:04 +01:00
|
|
|
&rlist->rl_ghs[x]);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* gfs2_rlist_free - free a resource group list
|
|
|
|
* @list: the list of resource groups
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
|
|
|
void gfs2_rlist_free(struct gfs2_rgrp_list *rlist)
|
|
|
|
{
|
|
|
|
unsigned int x;
|
|
|
|
|
|
|
|
kfree(rlist->rl_rgd);
|
|
|
|
|
|
|
|
if (rlist->rl_ghs) {
|
|
|
|
for (x = 0; x < rlist->rl_rgrps; x++)
|
|
|
|
gfs2_holder_uninit(&rlist->rl_ghs[x]);
|
|
|
|
kfree(rlist->rl_ghs);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|