X-Git-Url: http://ftp.safe.ca/?a=blobdiff_plain;f=fs%2Fgfs2%2Frgrp.c;h=daa4ae341a292fe03175c378d441e767706fb9c8;hb=fc00127fb67b2a7d2b66f0f4096a5367b581f045;hp=3f10b1fafd663a6186b1102ce9f36b4db5e05ea8;hpb=cf45b752c9f23939e40d823b0600bf876e97b0e0;p=safe%2Fjmp%2Flinux-2.6 diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index 3f10b1f..daa4ae3 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -13,7 +13,8 @@ #include #include #include -#include +#include +#include #include "gfs2.h" #include "incore.h" @@ -28,11 +29,21 @@ #include "util.h" #include "log.h" #include "inode.h" -#include "ops_address.h" +#include "trace_gfs2.h" #define BFITNOENT ((u32)~0) #define NO_BLOCK ((u64)~0) +#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 + /* * These routines are used by the resource group routines (rgrp.c) * to keep track of block allocation. Each block is represented by two @@ -53,7 +64,8 @@ static const char valid_change[16] = { }; static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, - unsigned char old_state, unsigned char new_state); + unsigned char old_state, unsigned char new_state, + unsigned int *n); /** * gfs2_setbit - Set a bit in the bitmaps @@ -64,26 +76,32 @@ static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, * */ -static void gfs2_setbit(struct gfs2_rgrpd *rgd, unsigned char *buffer, - unsigned int buflen, u32 block, - unsigned char new_state) +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) { - unsigned char *byte, *end, cur_state; - unsigned int bit; + unsigned char *byte1, *byte2, *end, cur_state; + const unsigned int bit = (block % GFS2_NBBY) * GFS2_BIT_SIZE; - byte = buffer + (block / GFS2_NBBY); - bit = (block % GFS2_NBBY) * GFS2_BIT_SIZE; - end = buffer + buflen; + byte1 = buf1 + offset + (block / GFS2_NBBY); + end = buf1 + offset + buflen; - gfs2_assert(rgd->rd_sbd, byte < end); + BUG_ON(byte1 >= end); - cur_state = (*byte >> bit) & GFS2_BIT_MASK; + cur_state = (*byte1 >> bit) & GFS2_BIT_MASK; - if (valid_change[new_state * 4 + cur_state]) { - *byte ^= cur_state << bit; - *byte |= new_state << bit; - } else + if (unlikely(!valid_change[new_state * 4 + cur_state])) { gfs2_consist_rgrpd(rgd); + 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; + } } /** @@ -94,10 +112,12 @@ static void gfs2_setbit(struct gfs2_rgrpd *rgd, unsigned char *buffer, * */ -static unsigned char gfs2_testbit(struct gfs2_rgrpd *rgd, unsigned char *buffer, - unsigned int buflen, u32 block) +static inline unsigned char gfs2_testbit(struct gfs2_rgrpd *rgd, + const unsigned char *buffer, + unsigned int buflen, u32 block) { - unsigned char *byte, *end, cur_state; + const unsigned char *byte, *end; + unsigned char cur_state; unsigned int bit; byte = buffer + (block / GFS2_NBBY); @@ -112,63 +132,89 @@ static unsigned char gfs2_testbit(struct gfs2_rgrpd *rgd, unsigned char *buffer, } /** + * 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[] = { + [0] = 0xffffffffffffffffULL, + [1] = 0xaaaaaaaaaaaaaaaaULL, + [2] = 0x5555555555555555ULL, + [3] = 0x0000000000000000ULL, + }; + tmp = le64_to_cpu(*ptr) ^ search[state]; + tmp &= (tmp >> 1); + tmp &= mask; + return tmp; +} + +/** * 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 - * @buflen: the length (in bytes) of the buffer + * @len: the length (in bytes) of the buffer * @goal: start search at this block's bit-pair (within @buffer) - * @old_state: GFS2_BLKST_XXX the state of the block we're looking for. + * @state: GFS2_BLKST_XXX the state of the block we're looking for. * * 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 - * beginning of a bitmap block buffer, skipping any header structures. + * 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. * * Return: the block number (bitmap buffer scope) that was found */ -static u32 gfs2_bitfit(const u8 *buffer, unsigned int buflen, u32 goal, - u8 old_state) +static u32 gfs2_bitfit(const u8 *buf, const unsigned int len, + u32 goal, u8 state) { - const u8 *byte; - u32 blk = goal; - unsigned int bit, bitlong; - const unsigned long *plong; -#if BITS_PER_LONG == 32 - const unsigned long plong55 = 0x55555555; -#else - const unsigned long plong55 = 0x5555555555555555; -#endif - - byte = buffer + (goal / GFS2_NBBY); - plong = (const unsigned long *)(buffer + (goal / GFS2_NBBY)); - bit = (goal % GFS2_NBBY) * GFS2_BIT_SIZE; - bitlong = bit; - - while (byte < buffer + buflen) { - - if (bitlong == 0 && old_state == 0 && *plong == plong55) { - plong++; - byte += sizeof(unsigned long); - blk += sizeof(unsigned long) * GFS2_NBBY; - continue; - } - if (((*byte >> bit) & GFS2_BIT_MASK) == old_state) - return blk; - bit += GFS2_BIT_SIZE; - if (bit >= 8) { - bit = 0; - byte++; - } - bitlong += GFS2_BIT_SIZE; - if (bitlong >= sizeof(unsigned long) * 8) { - bitlong = 0; - plong++; - } - - blk++; + 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; + u64 mask = 0x5555555555555555ULL; + 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) { + tmp = gfs2_bit_search(ptr, 0x5555555555555555ULL, state); + ptr++; } - - return BFITNOENT; + /* 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--; + bit = __ffs64(tmp); + bit /= 2; /* two bits per entry in the bitmap */ + return (((const unsigned char *)ptr - buf) * GFS2_NBBY) + bit; } /** @@ -231,16 +277,14 @@ void gfs2_rgrp_verify(struct gfs2_rgrpd *rgd) bi->bi_len, x); } - if (count[0] != rgd->rd_rg.rg_free) { + if (count[0] != rgd->rd_free) { if (gfs2_consist_rgrpd(rgd)) fs_err(sdp, "free data mismatch: %u != %u\n", - count[0], rgd->rd_rg.rg_free); + count[0], rgd->rd_free); return; } - tmp = rgd->rd_data - - rgd->rd_rg.rg_free - - rgd->rd_rg.rg_dinodes; + tmp = rgd->rd_data - rgd->rd_free - rgd->rd_dinodes; if (count[1] + count[2] != tmp) { if (gfs2_consist_rgrpd(rgd)) fs_err(sdp, "used data mismatch: %u != %u\n", @@ -248,10 +292,10 @@ void gfs2_rgrp_verify(struct gfs2_rgrpd *rgd) return; } - if (count[3] != rgd->rd_rg.rg_dinodes) { + if (count[3] != rgd->rd_dinodes) { if (gfs2_consist_rgrpd(rgd)) fs_err(sdp, "used metadata mismatch: %u != %u\n", - count[3], rgd->rd_rg.rg_dinodes); + count[3], rgd->rd_dinodes); return; } @@ -333,11 +377,6 @@ static void clear_rgrpdi(struct gfs2_sbd *sdp) spin_lock(&sdp->sd_rindex_spin); sdp->sd_rindex_forward = NULL; - head = &sdp->sd_rindex_recent_list; - while (!list_empty(head)) { - rgd = list_entry(head->next, struct gfs2_rgrpd, rd_recent); - list_del(&rgd->rd_recent); - } spin_unlock(&sdp->sd_rindex_spin); head = &sdp->sd_rindex_list; @@ -403,6 +442,7 @@ static int compute_bitstructs(struct gfs2_rgrpd *rgd) for (x = 0; x < length; x++) { bi = rgd->rd_bits + x; + bi->bi_flags = 0; /* small rgrp; bitmap stored completely in header block */ if (length == 1) { bytes = bytes_left; @@ -468,7 +508,7 @@ u64 gfs2_ri_total(struct gfs2_sbd *sdp) for (rgrps = 0;; rgrps++) { loff_t pos = rgrps * sizeof(struct gfs2_rindex); - if (pos + sizeof(struct gfs2_rindex) >= ip->i_di.di_size) + if (pos + sizeof(struct gfs2_rindex) >= ip->i_disksize) break; error = gfs2_internal_read(ip, &ra_state, buf, &pos, sizeof(struct gfs2_rindex)); @@ -541,7 +581,6 @@ static int read_rindex_entry(struct gfs2_inode *ip, rgd->rd_gl->gl_object = rgd; rgd->rd_flags &= ~GFS2_RDF_UPTODATE; - rgd->rd_flags |= GFS2_RDF_CHECK; return error; } @@ -557,7 +596,7 @@ static int gfs2_ri_update(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; - u64 rgrp_count = ip->i_di.di_size; + u64 rgrp_count = ip->i_disksize; int error; if (do_div(rgrp_count, sizeof(struct gfs2_rindex))) { @@ -601,7 +640,7 @@ static int gfs2_ri_update_special(struct gfs2_inode *ip) for (sdp->sd_rgrps = 0;; sdp->sd_rgrps++) { /* Ignore partials */ if ((sdp->sd_rgrps + 1) * sizeof(struct gfs2_rindex) > - ip->i_di.di_size) + ip->i_disksize) break; error = read_rindex_entry(ip, &ra_state); if (error) { @@ -659,32 +698,26 @@ int gfs2_rindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ri_gh) static void gfs2_rgrp_in(struct gfs2_rgrpd *rgd, const void *buf) { const struct gfs2_rgrp *str = buf; - struct gfs2_rgrp_host *rg = &rgd->rd_rg; u32 rg_flags; 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; - rg->rg_free = be32_to_cpu(str->rg_free); - rg->rg_dinodes = be32_to_cpu(str->rg_dinodes); - rg->rg_igeneration = be64_to_cpu(str->rg_igeneration); + rg_flags &= ~GFS2_RDF_MASK; + rgd->rd_flags &= GFS2_RDF_MASK; + rgd->rd_flags |= rg_flags; + rgd->rd_free = be32_to_cpu(str->rg_free); + rgd->rd_dinodes = be32_to_cpu(str->rg_dinodes); + rgd->rd_igeneration = be64_to_cpu(str->rg_igeneration); } static void gfs2_rgrp_out(struct gfs2_rgrpd *rgd, void *buf) { struct gfs2_rgrp *str = buf; - struct gfs2_rgrp_host *rg = &rgd->rd_rg; - u32 rg_flags = 0; - - if (rgd->rd_flags & GFS2_RDF_NOALLOC) - rg_flags |= GFS2_RGF_NOALLOC; - str->rg_flags = cpu_to_be32(rg_flags); - str->rg_free = cpu_to_be32(rg->rg_free); - str->rg_dinodes = cpu_to_be32(rg->rg_dinodes); + + str->rg_flags = cpu_to_be32(rgd->rd_flags & ~GFS2_RDF_MASK); + str->rg_free = cpu_to_be32(rgd->rd_free); + str->rg_dinodes = cpu_to_be32(rgd->rd_dinodes); str->__pad = cpu_to_be32(0); - str->rg_igeneration = cpu_to_be64(rg->rg_igeneration); + str->rg_igeneration = cpu_to_be64(rgd->rd_igeneration); memset(&str->rg_reserved, 0, sizeof(str->rg_reserved)); } @@ -738,12 +771,14 @@ int gfs2_rgrp_bh_get(struct gfs2_rgrpd *rgd) } if (!(rgd->rd_flags & GFS2_RDF_UPTODATE)) { + for (x = 0; x < length; x++) + clear_bit(GBF_FULL, &rgd->rd_bits[x].bi_flags); gfs2_rgrp_in(rgd, (rgd->rd_bits[0].bi_bh)->b_data); - rgd->rd_flags |= GFS2_RDF_UPTODATE; + rgd->rd_flags |= (GFS2_RDF_UPTODATE | GFS2_RDF_CHECK); } spin_lock(&sdp->sd_rindex_spin); - rgd->rd_free_clone = rgd->rd_rg.rg_free; + rgd->rd_free_clone = rgd->rd_free; rgd->rd_bh_count++; spin_unlock(&sdp->sd_rindex_spin); @@ -802,6 +837,58 @@ void gfs2_rgrp_bh_put(struct gfs2_rgrpd *rgd) spin_unlock(&sdp->sd_rindex_spin); } +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_logical_block_size(sb->s_bdev); + u64 blk; + sector_t start = 0; + 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; +} + void gfs2_rgrp_repolish_clones(struct gfs2_rgrpd *rgd) { struct gfs2_sbd *sdp = rgd->rd_sbd; @@ -812,12 +899,15 @@ void gfs2_rgrp_repolish_clones(struct gfs2_rgrpd *rgd) struct gfs2_bitmap *bi = rgd->rd_bits + x; if (!bi->bi_clone) continue; + if (sdp->sd_args.ar_discard) + gfs2_rgrp_send_discards(sdp, rgd->rd_data0, bi); + clear_bit(GBF_FULL, &bi->bi_flags); memcpy(bi->bi_clone + bi->bi_offset, bi->bi_bh->b_data + bi->bi_offset, bi->bi_len); } spin_lock(&sdp->sd_rindex_spin); - rgd->rd_free_clone = rgd->rd_rg.rg_free; + rgd->rd_free_clone = rgd->rd_free; spin_unlock(&sdp->sd_rindex_spin); } @@ -851,7 +941,7 @@ static int try_rgrp_fit(struct gfs2_rgrpd *rgd, struct gfs2_alloc *al) struct gfs2_sbd *sdp = rgd->rd_sbd; int ret = 0; - if (rgd->rd_flags & GFS2_RDF_NOALLOC) + if (rgd->rd_flags & (GFS2_RGF_NOALLOC | GFS2_RDF_ERROR)) return 0; spin_lock(&sdp->sd_rindex_spin); @@ -877,13 +967,15 @@ static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked) u32 goal = 0, block; u64 no_addr; struct gfs2_sbd *sdp = rgd->rd_sbd; + unsigned int n; for(;;) { if (goal >= rgd->rd_data) break; down_write(&sdp->sd_log_flush_lock); + n = 1; block = rgblk_search(rgd, goal, GFS2_BLKST_UNLINKED, - GFS2_BLKST_UNLINKED); + GFS2_BLKST_UNLINKED, &n); up_write(&sdp->sd_log_flush_lock); if (block == BFITNOENT) break; @@ -905,111 +997,30 @@ static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked) } /** - * recent_rgrp_first - get first RG from "recent" list - * @sdp: The GFS2 superblock - * @rglast: address of the rgrp used last - * - * Returns: The first rgrp in the recent list - */ - -static struct gfs2_rgrpd *recent_rgrp_first(struct gfs2_sbd *sdp, - u64 rglast) -{ - struct gfs2_rgrpd *rgd = NULL; - - spin_lock(&sdp->sd_rindex_spin); - - if (list_empty(&sdp->sd_rindex_recent_list)) - goto out; - - if (!rglast) - goto first; - - list_for_each_entry(rgd, &sdp->sd_rindex_recent_list, rd_recent) { - if (rgd->rd_addr == rglast) - goto out; - } - -first: - rgd = list_entry(sdp->sd_rindex_recent_list.next, struct gfs2_rgrpd, - rd_recent); -out: - spin_unlock(&sdp->sd_rindex_spin); - return rgd; -} - -/** * recent_rgrp_next - get next RG from "recent" list * @cur_rgd: current rgrp - * @remove: * * Returns: The next rgrp in the recent list */ -static struct gfs2_rgrpd *recent_rgrp_next(struct gfs2_rgrpd *cur_rgd, - int remove) +static struct gfs2_rgrpd *recent_rgrp_next(struct gfs2_rgrpd *cur_rgd) { struct gfs2_sbd *sdp = cur_rgd->rd_sbd; struct list_head *head; struct gfs2_rgrpd *rgd; spin_lock(&sdp->sd_rindex_spin); - - head = &sdp->sd_rindex_recent_list; - - list_for_each_entry(rgd, head, rd_recent) { - if (rgd == cur_rgd) { - if (cur_rgd->rd_recent.next != head) - rgd = list_entry(cur_rgd->rd_recent.next, - struct gfs2_rgrpd, rd_recent); - else - rgd = NULL; - - if (remove) - list_del(&cur_rgd->rd_recent); - - goto out; - } + head = &sdp->sd_rindex_mru_list; + if (unlikely(cur_rgd->rd_list_mru.next == head)) { + spin_unlock(&sdp->sd_rindex_spin); + return NULL; } - - rgd = NULL; - if (!list_empty(head)) - rgd = list_entry(head->next, struct gfs2_rgrpd, rd_recent); - -out: + rgd = list_entry(cur_rgd->rd_list_mru.next, struct gfs2_rgrpd, rd_list_mru); spin_unlock(&sdp->sd_rindex_spin); return rgd; } /** - * recent_rgrp_add - add an RG to tail of "recent" list - * @new_rgd: The rgrp to add - * - */ - -static void recent_rgrp_add(struct gfs2_rgrpd *new_rgd) -{ - struct gfs2_sbd *sdp = new_rgd->rd_sbd; - struct gfs2_rgrpd *rgd; - unsigned int count = 0; - unsigned int max = sdp->sd_rgrps / gfs2_jindex_size(sdp); - - spin_lock(&sdp->sd_rindex_spin); - - list_for_each_entry(rgd, &sdp->sd_rindex_recent_list, rd_recent) { - if (rgd == new_rgd) - goto out; - - if (++count >= max) - goto out; - } - list_add_tail(&new_rgd->rd_recent, &sdp->sd_rindex_recent_list); - -out: - spin_unlock(&sdp->sd_rindex_spin); -} - -/** * forward_rgrp_get - get an rgrp to try next from full list * @sdp: The GFS2 superblock * @@ -1076,9 +1087,7 @@ static struct inode *get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked) int loops = 0; int error, rg_locked; - /* Try recently successful rgrps */ - - rgd = recent_rgrp_first(sdp, ip->i_last_rg_alloc); + rgd = gfs2_blk2rgrpd(sdp, ip->i_goal); while (rgd) { rg_locked = 0; @@ -1100,11 +1109,9 @@ static struct inode *get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked) gfs2_glock_dq_uninit(&al->al_rgd_gh); if (inode) return inode; - rgd = recent_rgrp_next(rgd, 1); - break; - + /* fall through */ case GLR_TRYFAILED: - rgd = recent_rgrp_next(rgd, 0); + rgd = recent_rgrp_next(rgd); break; default: @@ -1162,10 +1169,10 @@ static struct inode *get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked) } out: - ip->i_last_rg_alloc = rgd->rd_addr; - if (begin) { - recent_rgrp_add(rgd); + spin_lock(&sdp->sd_rindex_spin); + list_move(&rgd->rd_list_mru, &sdp->sd_rindex_mru_list); + spin_unlock(&sdp->sd_rindex_spin); rgd = gfs2_rgrpd_get_next(rgd); if (!rgd) rgd = gfs2_rgrpd_get_first(sdp); @@ -1286,6 +1293,7 @@ unsigned char gfs2_get_block_type(struct gfs2_rgrpd *rgd, u64 block) * @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 + * @n: The extent length * * Walk rgrp's bitmap to find bits that represent a block in @old_state. * Add the found bitmap buffer to the transaction. @@ -1301,34 +1309,45 @@ unsigned char gfs2_get_block_type(struct gfs2_rgrpd *rgd, u64 block) */ static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, - unsigned char old_state, unsigned char new_state) + unsigned char old_state, unsigned char new_state, + unsigned int *n) { struct gfs2_bitmap *bi = NULL; - u32 length = rgd->rd_length; - u32 blk = 0; + const u32 length = rgd->rd_length; + u32 blk = BFITNOENT; unsigned int buf, x; + const unsigned int elen = *n; + const u8 *buffer = NULL; + *n = 0; /* 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; + /* Convert scope of "goal" from rgrp-wide to within found bit block */ + if (goal < (bi->bi_start + bi->bi_len) * GFS2_NBBY) { + goal -= bi->bi_start * GFS2_NBBY; + goto do_search; + } } + buf = 0; + goal = 0; - 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; - +do_search: /* 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++) { + bi = rgd->rd_bits + buf; + + if (test_bit(GBF_FULL, &bi->bi_flags) && + (old_state == GFS2_BLKST_FREE)) + goto skip; + /* The GFS2_BLKST_UNLINKED state doesn't apply to the clone bitmaps, so we must search the originals for that. */ - const u8 *buffer = bi->bi_bh->b_data + bi->bi_offset; + buffer = bi->bi_bh->b_data + bi->bi_offset; if (old_state != GFS2_BLKST_UNLINKED && bi->bi_clone) buffer = bi->bi_clone + bi->bi_offset; @@ -1336,22 +1355,39 @@ static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, if (blk != BFITNOENT) break; + if ((goal == 0) && (old_state == GFS2_BLKST_FREE)) + set_bit(GBF_FULL, &bi->bi_flags); + /* Try next bitmap block (wrap back to rgrp header if at end) */ - buf = (buf + 1) % length; - bi = rgd->rd_bits + buf; +skip: + buf++; + buf %= length; goal = 0; } - if (blk != BFITNOENT && old_state != new_state) { - gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1); - gfs2_setbit(rgd, bi->bi_bh->b_data + bi->bi_offset, - bi->bi_len, blk, new_state); - if (bi->bi_clone) - gfs2_setbit(rgd, bi->bi_clone + bi->bi_offset, - bi->bi_len, blk, new_state); - } + if (blk == BFITNOENT) + return blk; + *n = 1; + if (old_state == new_state) + goto out; - return (blk == BFITNOENT) ? blk : (bi->bi_start * GFS2_NBBY) + blk; + gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1); + gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone, bi->bi_offset, + bi->bi_len, blk, new_state); + goal = blk; + while (*n < elen) { + goal++; + if (goal >= (bi->bi_len * GFS2_NBBY)) + 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, bi->bi_offset, + bi->bi_len, goal, new_state); + (*n)++; + } +out: + return (bi->bi_start * GFS2_NBBY) + blk; } /** @@ -1403,7 +1439,7 @@ static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart, bi->bi_len); } gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1); - gfs2_setbit(rgd, bi->bi_bh->b_data + bi->bi_offset, + gfs2_setbit(rgd, bi->bi_bh->b_data, NULL, bi->bi_offset, bi->bi_len, buf_blk, new_state); } @@ -1411,94 +1447,90 @@ static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart, } /** - * gfs2_alloc_data - Allocate a data block - * @ip: the inode to allocate the data block for + * gfs2_rgrp_dump - print out an rgrp + * @seq: The iterator + * @gl: The glock in question * - * Returns: the allocated block */ -u64 gfs2_alloc_data(struct gfs2_inode *ip) +int gfs2_rgrp_dump(struct seq_file *seq, const struct gfs2_glock *gl) { - struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_alloc *al = ip->i_alloc; - struct gfs2_rgrpd *rgd = al->al_rgd; - u32 goal, blk; - u64 block; - - if (rgrp_contains_block(rgd, ip->i_di.di_goal_data)) - goal = ip->i_di.di_goal_data - rgd->rd_data0; - else - goal = rgd->rd_last_alloc_data; - - blk = rgblk_search(rgd, goal, GFS2_BLKST_FREE, GFS2_BLKST_USED); - BUG_ON(blk == BFITNOENT); - rgd->rd_last_alloc_data = blk; - - block = rgd->rd_data0 + blk; - ip->i_di.di_goal_data = block; - - gfs2_assert_withdraw(sdp, rgd->rd_rg.rg_free); - rgd->rd_rg.rg_free--; - - gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); - gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); - - al->al_alloced++; - - gfs2_statfs_change(sdp, 0, -1, 0); - gfs2_quota_change(ip, +1, ip->i_inode.i_uid, ip->i_inode.i_gid); - - spin_lock(&sdp->sd_rindex_spin); - rgd->rd_free_clone--; - spin_unlock(&sdp->sd_rindex_spin); - - return block; + const struct gfs2_rgrpd *rgd = gl->gl_object; + if (rgd == NULL) + return 0; + gfs2_print_dbg(seq, " R: n:%llu f:%02x b:%u/%u i:%u\n", + (unsigned long long)rgd->rd_addr, rgd->rd_flags, + rgd->rd_free, rgd->rd_free_clone, rgd->rd_dinodes); + return 0; } /** - * gfs2_alloc_meta - Allocate a metadata block - * @ip: the inode to allocate the metadata block for + * gfs2_alloc_block - Allocate one or more blocks + * @ip: the inode to allocate the block for + * @bn: Used to return the starting block number + * @n: requested number of blocks/extent length (value/result) * - * Returns: the allocated block + * Returns: 0 or error */ -u64 gfs2_alloc_meta(struct gfs2_inode *ip) +int gfs2_alloc_block(struct gfs2_inode *ip, u64 *bn, unsigned int *n) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + struct buffer_head *dibh; struct gfs2_alloc *al = ip->i_alloc; struct gfs2_rgrpd *rgd = al->al_rgd; u32 goal, blk; u64 block; + int error; - if (rgrp_contains_block(rgd, ip->i_di.di_goal_meta)) - goal = ip->i_di.di_goal_meta - rgd->rd_data0; + if (rgrp_contains_block(rgd, ip->i_goal)) + goal = ip->i_goal - rgd->rd_data0; else - goal = rgd->rd_last_alloc_meta; + goal = rgd->rd_last_alloc; - blk = rgblk_search(rgd, goal, GFS2_BLKST_FREE, GFS2_BLKST_USED); - BUG_ON(blk == BFITNOENT); - rgd->rd_last_alloc_meta = blk; + blk = rgblk_search(rgd, goal, GFS2_BLKST_FREE, GFS2_BLKST_USED, n); + /* Since all blocks are reserved in advance, this shouldn't happen */ + if (blk == BFITNOENT) + goto rgrp_error; + + rgd->rd_last_alloc = blk; block = rgd->rd_data0 + blk; - ip->i_di.di_goal_meta = block; + ip->i_goal = block; + 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); + } + if (rgd->rd_free < *n) + goto rgrp_error; - gfs2_assert_withdraw(sdp, rgd->rd_rg.rg_free); - rgd->rd_rg.rg_free--; + rgd->rd_free -= *n; gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); - al->al_alloced++; + al->al_alloced += *n; - gfs2_statfs_change(sdp, 0, -1, 0); - gfs2_quota_change(ip, +1, ip->i_inode.i_uid, ip->i_inode.i_gid); - gfs2_trans_add_unrevoke(sdp, block); + gfs2_statfs_change(sdp, 0, -(s64)*n, 0); + gfs2_quota_change(ip, *n, ip->i_inode.i_uid, ip->i_inode.i_gid); spin_lock(&sdp->sd_rindex_spin); - rgd->rd_free_clone--; + rgd->rd_free_clone -= *n; spin_unlock(&sdp->sd_rindex_spin); + trace_gfs2_block_alloc(ip, block, *n, GFS2_BLKST_USED); + *bn = block; + return 0; - return block; +rgrp_error: + fs_warn(sdp, "rgrp %llu has an error, marking it readonly until umount\n", + (unsigned long long)rgd->rd_addr); + fs_warn(sdp, "umount on all nodes and run fsck.gfs2 to fix the error\n"); + gfs2_rgrp_dump(NULL, rgd->rd_gl); + rgd->rd_flags |= GFS2_RDF_ERROR; + return -EIO; } /** @@ -1515,31 +1547,32 @@ u64 gfs2_alloc_di(struct gfs2_inode *dip, u64 *generation) struct gfs2_rgrpd *rgd = al->al_rgd; u32 blk; u64 block; + unsigned int n = 1; - blk = rgblk_search(rgd, rgd->rd_last_alloc_meta, - GFS2_BLKST_FREE, GFS2_BLKST_DINODE); + blk = rgblk_search(rgd, rgd->rd_last_alloc, + GFS2_BLKST_FREE, GFS2_BLKST_DINODE, &n); BUG_ON(blk == BFITNOENT); - rgd->rd_last_alloc_meta = blk; + rgd->rd_last_alloc = blk; block = rgd->rd_data0 + blk; - gfs2_assert_withdraw(sdp, rgd->rd_rg.rg_free); - rgd->rd_rg.rg_free--; - rgd->rd_rg.rg_dinodes++; - *generation = rgd->rd_rg.rg_igeneration++; + gfs2_assert_withdraw(sdp, rgd->rd_free); + rgd->rd_free--; + rgd->rd_dinodes++; + *generation = rgd->rd_igeneration++; gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); al->al_alloced++; gfs2_statfs_change(sdp, 0, -1, +1); - gfs2_trans_add_unrevoke(sdp, block); + gfs2_trans_add_unrevoke(sdp, block, 1); spin_lock(&sdp->sd_rindex_spin); rgd->rd_free_clone--; spin_unlock(&sdp->sd_rindex_spin); - + trace_gfs2_block_alloc(dip, block, 1, GFS2_BLKST_DINODE); return block; } @@ -1559,8 +1592,8 @@ void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen) rgd = rgblk_free(sdp, bstart, blen, GFS2_BLKST_FREE); if (!rgd) return; - - rgd->rd_rg.rg_free += blen; + trace_gfs2_block_alloc(ip, bstart, blen, GFS2_BLKST_FREE); + rgd->rd_free += blen; gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); @@ -1587,8 +1620,8 @@ void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen) rgd = rgblk_free(sdp, bstart, blen, GFS2_BLKST_FREE); if (!rgd) return; - - rgd->rd_rg.rg_free += blen; + trace_gfs2_block_alloc(ip, bstart, blen, GFS2_BLKST_FREE); + rgd->rd_free += blen; gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); @@ -1610,6 +1643,7 @@ void gfs2_unlink_di(struct inode *inode) rgd = rgblk_free(sdp, blkno, 1, GFS2_BLKST_UNLINKED); if (!rgd) return; + trace_gfs2_block_alloc(ip, blkno, 1, GFS2_BLKST_UNLINKED); gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); gfs2_trans_add_rg(rgd); @@ -1625,10 +1659,10 @@ static void gfs2_free_uninit_di(struct gfs2_rgrpd *rgd, u64 blkno) return; gfs2_assert_withdraw(sdp, rgd == tmp_rgd); - if (!rgd->rd_rg.rg_dinodes) + if (!rgd->rd_dinodes) gfs2_consist_rgrpd(rgd); - rgd->rd_rg.rg_dinodes--; - rgd->rd_rg.rg_free++; + rgd->rd_dinodes--; + rgd->rd_free++; gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); @@ -1641,6 +1675,7 @@ static void gfs2_free_uninit_di(struct gfs2_rgrpd *rgd, u64 blkno) void gfs2_free_di(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip) { gfs2_free_uninit_di(rgd, ip->i_no_addr); + trace_gfs2_block_alloc(ip, ip->i_no_addr, 1, GFS2_BLKST_FREE); gfs2_quota_change(ip, -1, ip->i_inode.i_uid, ip->i_inode.i_gid); gfs2_meta_wipe(ip, ip->i_no_addr, 1); }