X-Git-Url: https://err.no/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=fs%2Fgfs2%2Frgrp.c;h=3401628d742b6cbfd5b53c9a4ed11f42d3372413;hb=6924d1ab8b7bbe5ab416713f5701b3316b2df85b;hp=274a2df13f020e914ee19587daf47c985a2b30d7;hpb=1639431a3f57b43da1e15e9268a1d691ac01ba26;p=linux-2.6 diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index 274a2df13f..3401628d74 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -14,6 +14,7 @@ #include #include #include +#include #include "gfs2.h" #include "incore.h" @@ -33,6 +34,16 @@ #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); @@ -129,45 +149,63 @@ static unsigned char gfs2_testbit(struct gfs2_rgrpd *rgd, unsigned char *buffer, static u32 gfs2_bitfit(const u8 *buffer, unsigned int buflen, u32 goal, u8 old_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; + const u8 *byte, *start, *end; + int bit, startbit; + u32 g1, g2, misaligned; + unsigned long *plong; + unsigned long lskipval; + + lskipval = (old_state & GFS2_BLKST_USED) ? LBITSKIP00 : LBITSKIP55; + g1 = (goal / GFS2_NBBY); + start = buffer + g1; + byte = start; + end = buffer + buflen; + g2 = ALIGN(g1, sizeof(unsigned long)); + plong = (unsigned long *)(buffer + g2); + startbit = bit = (goal % GFS2_NBBY) * GFS2_BIT_SIZE; + misaligned = g2 - g1; + if (!misaligned) + goto ulong_aligned; +/* parse the bitmap a byte at a time */ +misaligned: + while (byte < end) { + if (((*byte >> bit) & GFS2_BIT_MASK) == old_state) { + return goal + + (((byte - start) * GFS2_NBBY) + + ((bit - startbit) >> 1)); } - if (((*byte >> bit) & GFS2_BIT_MASK) == old_state) - return blk; bit += GFS2_BIT_SIZE; - if (bit >= 8) { + if (bit >= GFS2_NBBY * GFS2_BIT_SIZE) { bit = 0; byte++; + misaligned--; + if (!misaligned) { + plong = (unsigned long *)byte; + goto ulong_aligned; + } } - bitlong += GFS2_BIT_SIZE; - if (bitlong >= sizeof(unsigned long) * 8) { - bitlong = 0; - plong++; - } - - blk++; } + return BFITNOENT; +/* parse the bitmap a unsigned long at a time */ +ulong_aligned: + /* Stop at "end - 1" or else prefetch can go past the end and segfault. + We could "if" it but we'd lose some of the performance gained. + This way will only slow down searching the very last 4/8 bytes + depending on architecture. I've experimented with several ways + of writing this section such as using an else before the goto + but this one seems to be the fastest. */ + while ((unsigned char *)plong < end - sizeof(unsigned long)) { + prefetch(plong + 1); + if (((*plong) & LBITMASK) != lskipval) + break; + plong++; + } + if ((unsigned char *)plong < end) { + byte = (const u8 *)plong; + misaligned += sizeof(unsigned long) - 1; + goto misaligned; + } return BFITNOENT; } @@ -877,13 +915,15 @@ 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; @@ -1280,6 +1320,7 @@ 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. @@ -1295,13 +1336,17 @@ 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; + const u32 length = rgd->rd_length; u32 blk = 0; unsigned int buf, x; + const unsigned int elen = *n; + const u8 *buffer; + *n = 0; /* Find bitmap block that contains bits for goal block */ for (buf = 0; buf < length; buf++) { bi = rgd->rd_bits + buf; @@ -1322,7 +1367,7 @@ static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, for (x = 0; x <= length; x++) { /* 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; @@ -1337,12 +1382,23 @@ static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, } if (blk != BFITNOENT && old_state != new_state) { + *n = 1; 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, bi->bi_clone, 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); + 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)++; + } } return (blk == BFITNOENT) ? blk : (bi->bi_start * GFS2_NBBY) + blk; @@ -1397,7 +1453,7 @@ 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); } @@ -1411,7 +1467,7 @@ static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart, * Returns: the allocated block */ -u64 gfs2_alloc_block(struct gfs2_inode *ip) +u64 gfs2_alloc_block(struct gfs2_inode *ip, unsigned int *n) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_alloc *al = ip->i_alloc; @@ -1424,26 +1480,26 @@ u64 gfs2_alloc_block(struct gfs2_inode *ip) else goal = rgd->rd_last_alloc; - blk = rgblk_search(rgd, goal, GFS2_BLKST_FREE, GFS2_BLKST_USED); + blk = rgblk_search(rgd, goal, GFS2_BLKST_FREE, GFS2_BLKST_USED, n); BUG_ON(blk == BFITNOENT); - rgd->rd_last_alloc = blk; + rgd->rd_last_alloc = blk; block = rgd->rd_data0 + blk; ip->i_goal = block; - gfs2_assert_withdraw(sdp, rgd->rd_rg.rg_free); - rgd->rd_rg.rg_free--; + gfs2_assert_withdraw(sdp, rgd->rd_rg.rg_free >= *n); + rgd->rd_rg.rg_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_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); return block; @@ -1463,9 +1519,10 @@ 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, - GFS2_BLKST_FREE, GFS2_BLKST_DINODE); + GFS2_BLKST_FREE, GFS2_BLKST_DINODE, &n); BUG_ON(blk == BFITNOENT); rgd->rd_last_alloc = blk;