[PATCH] md: merge raid5 and raid6 code
authorNeilBrown <neilb@suse.de>
Mon, 26 Jun 2006 07:27:38 +0000 (00:27 -0700)
committerLinus Torvalds <torvalds@g5.osdl.org>
Mon, 26 Jun 2006 16:58:37 +0000 (09:58 -0700)
There is a lot of commonality between raid5.c and raid6main.c.  This patches
merges both into one module called raid456.  This saves a lot of code, and
paves the way for online raid5->raid6 migrations.

There is still duplication, e.g.  between handle_stripe5 and handle_stripe6.
This will probably be cleaned up later.

Cc: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
drivers/md/Kconfig
drivers/md/Makefile
drivers/md/raid5.c
drivers/md/raid6main.c [deleted file]
include/linux/raid/raid5.h

index ac25a48362acab47ea103fa176b39590ed5c2e42..f657aa7ec78c2be6e698bc69ad32a50ac880c68c 100644 (file)
@@ -104,8 +104,8 @@ config MD_RAID10
 
          If unsure, say Y.
 
-config MD_RAID5
-       tristate "RAID-4/RAID-5 mode"
+config MD_RAID456
+       tristate "RAID-4/RAID-5/RAID-6 mode"
        depends on BLK_DEV_MD
        ---help---
          A RAID-5 set of N drives with a capacity of C MB per drive provides
@@ -116,14 +116,22 @@ config MD_RAID5
          while a RAID-5 set distributes the parity across the drives in one
          of the available parity distribution methods.
 
+         A RAID-6 set of N drives with a capacity of C MB per drive
+         provides the capacity of C * (N - 2) MB, and protects
+         against a failure of any two drives. For a given sector
+         (row) number, (N - 2) drives contain data sectors, and two
+         drives contains two independent redundancy syndromes.  Like
+         RAID-5, RAID-6 distributes the syndromes across the drives
+         in one of the available parity distribution methods.
+
          Information about Software RAID on Linux is contained in the
          Software-RAID mini-HOWTO, available from
          <http://www.tldp.org/docs.html#howto>. There you will also
          learn where to get the supporting user space utilities raidtools.
 
-         If you want to use such a RAID-4/RAID-5 set, say Y.  To
+         If you want to use such a RAID-4/RAID-5/RAID-6 set, say Y.  To
          compile this code as a module, choose M here: the module
-         will be called raid5.
+         will be called raid456.
 
          If unsure, say Y.
 
@@ -154,28 +162,6 @@ config MD_RAID5_RESHAPE
          There should be enough spares already present to make the new
          array workable.
 
-config MD_RAID6
-       tristate "RAID-6 mode"
-       depends on BLK_DEV_MD
-       ---help---
-         A RAID-6 set of N drives with a capacity of C MB per drive
-         provides the capacity of C * (N - 2) MB, and protects
-         against a failure of any two drives. For a given sector
-         (row) number, (N - 2) drives contain data sectors, and two
-         drives contains two independent redundancy syndromes.  Like
-         RAID-5, RAID-6 distributes the syndromes across the drives
-         in one of the available parity distribution methods.
-
-         RAID-6 requires mdadm-1.5.0 or later, available at:
-
-         ftp://ftp.kernel.org/pub/linux/utils/raid/mdadm/
-
-         If you want to use such a RAID-6 set, say Y.  To compile
-         this code as a module, choose M here: the module will be
-         called raid6.
-
-         If unsure, say Y.
-
 config MD_MULTIPATH
        tristate "Multipath I/O support"
        depends on BLK_DEV_MD
index d3efedf6a6ad8292d77bc3d5b0e70adff12ad3de..34957a68d9215ac08061706c58b0fd66dc499d9d 100644 (file)
@@ -8,7 +8,7 @@ dm-multipath-objs := dm-hw-handler.o dm-path-selector.o dm-mpath.o
 dm-snapshot-objs := dm-snap.o dm-exception-store.o
 dm-mirror-objs := dm-log.o dm-raid1.o
 md-mod-objs     := md.o bitmap.o
-raid6-objs     := raid6main.o raid6algos.o raid6recov.o raid6tables.o \
+raid456-objs   := raid5.o raid6algos.o raid6recov.o raid6tables.o \
                   raid6int1.o raid6int2.o raid6int4.o \
                   raid6int8.o raid6int16.o raid6int32.o \
                   raid6altivec1.o raid6altivec2.o raid6altivec4.o \
@@ -25,8 +25,7 @@ obj-$(CONFIG_MD_LINEAR)               += linear.o
 obj-$(CONFIG_MD_RAID0)         += raid0.o
 obj-$(CONFIG_MD_RAID1)         += raid1.o
 obj-$(CONFIG_MD_RAID10)                += raid10.o
-obj-$(CONFIG_MD_RAID5)         += raid5.o xor.o
-obj-$(CONFIG_MD_RAID6)         += raid6.o xor.o
+obj-$(CONFIG_MD_RAID456)       += raid456.o xor.o
 obj-$(CONFIG_MD_MULTIPATH)     += multipath.o
 obj-$(CONFIG_MD_FAULTY)                += faulty.o
 obj-$(CONFIG_BLK_DEV_MD)       += md-mod.o
index 122e64e557b1c13b14211b385fcb4f7c49143848..9ba73074df0475e795c71de4d783e9090df4da8b 100644 (file)
@@ -2,8 +2,11 @@
  * raid5.c : Multiple Devices driver for Linux
  *        Copyright (C) 1996, 1997 Ingo Molnar, Miguel de Icaza, Gadi Oxman
  *        Copyright (C) 1999, 2000 Ingo Molnar
+ *        Copyright (C) 2002, 2003 H. Peter Anvin
  *
- * RAID-5 management functions.
+ * RAID-4/5/6 management functions.
+ * Thanks to Penguin Computing for making the RAID-6 development possible
+ * by donating a test server!
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License as published by
 #include <linux/config.h>
 #include <linux/module.h>
 #include <linux/slab.h>
-#include <linux/raid/raid5.h>
 #include <linux/highmem.h>
 #include <linux/bitops.h>
 #include <linux/kthread.h>
 #include <asm/atomic.h>
+#include "raid6.h"
 
 #include <linux/raid/bitmap.h>
 
 #define __inline__
 #endif
 
+#if !RAID6_USE_EMPTY_ZERO_PAGE
+/* In .bss so it's zeroed */
+const char raid6_empty_zero_page[PAGE_SIZE] __attribute__((aligned(256)));
+#endif
+
+static inline int raid6_next_disk(int disk, int raid_disks)
+{
+       disk++;
+       return (disk < raid_disks) ? disk : 0;
+}
 static void print_raid5_conf (raid5_conf_t *conf);
 
 static void __release_stripe(raid5_conf_t *conf, struct stripe_head *sh)
@@ -104,7 +117,7 @@ static void release_stripe(struct stripe_head *sh)
 {
        raid5_conf_t *conf = sh->raid_conf;
        unsigned long flags;
-       
+
        spin_lock_irqsave(&conf->device_lock, flags);
        __release_stripe(conf, sh);
        spin_unlock_irqrestore(&conf->device_lock, flags);
@@ -117,7 +130,7 @@ static inline void remove_hash(struct stripe_head *sh)
        hlist_del_init(&sh->hash);
 }
 
-static void insert_hash(raid5_conf_t *conf, struct stripe_head *sh)
+static inline void insert_hash(raid5_conf_t *conf, struct stripe_head *sh)
 {
        struct hlist_head *hp = stripe_hash(conf, sh->sector);
 
@@ -190,7 +203,7 @@ static void init_stripe(struct stripe_head *sh, sector_t sector, int pd_idx, int
                (unsigned long long)sh->sector);
 
        remove_hash(sh);
-       
+
        sh->sector = sector;
        sh->pd_idx = pd_idx;
        sh->state = 0;
@@ -269,8 +282,9 @@ static struct stripe_head *get_active_stripe(raid5_conf_t *conf, sector_t sector
                        } else {
                                if (!test_bit(STRIPE_HANDLE, &sh->state))
                                        atomic_inc(&conf->active_stripes);
-                               if (!list_empty(&sh->lru))
-                                       list_del_init(&sh->lru);
+                               if (list_empty(&sh->lru))
+                                       BUG();
+                               list_del_init(&sh->lru);
                        }
                }
        } while (sh == NULL);
@@ -321,10 +335,9 @@ static int grow_stripes(raid5_conf_t *conf, int num)
                return 1;
        conf->slab_cache = sc;
        conf->pool_size = devs;
-       while (num--) {
+       while (num--)
                if (!grow_one_stripe(conf))
                        return 1;
-       }
        return 0;
 }
 
@@ -631,8 +644,7 @@ static void raid5_build_block (struct stripe_head *sh, int i)
        dev->req.bi_private = sh;
 
        dev->flags = 0;
-       if (i != sh->pd_idx)
-               dev->sector = compute_blocknr(sh, i);
+       dev->sector = compute_blocknr(sh, i);
 }
 
 static void error(mddev_t *mddev, mdk_rdev_t *rdev)
@@ -659,7 +671,7 @@ static void error(mddev_t *mddev, mdk_rdev_t *rdev)
                        " Operation continuing on %d devices\n",
                        bdevname(rdev->bdev,b), conf->working_disks);
        }
-}      
+}
 
 /*
  * Input: a 'big' sector number,
@@ -697,9 +709,12 @@ static sector_t raid5_compute_sector(sector_t r_sector, unsigned int raid_disks,
        /*
         * Select the parity disk based on the user selected algorithm.
         */
-       if (conf->level == 4)
+       switch(conf->level) {
+       case 4:
                *pd_idx = data_disks;
-       else switch (conf->algorithm) {
+               break;
+       case 5:
+               switch (conf->algorithm) {
                case ALGORITHM_LEFT_ASYMMETRIC:
                        *pd_idx = data_disks - stripe % raid_disks;
                        if (*dd_idx >= *pd_idx)
@@ -721,6 +736,39 @@ static sector_t raid5_compute_sector(sector_t r_sector, unsigned int raid_disks,
                default:
                        printk(KERN_ERR "raid5: unsupported algorithm %d\n",
                                conf->algorithm);
+               }
+               break;
+       case 6:
+
+               /**** FIX THIS ****/
+               switch (conf->algorithm) {
+               case ALGORITHM_LEFT_ASYMMETRIC:
+                       *pd_idx = raid_disks - 1 - (stripe % raid_disks);
+                       if (*pd_idx == raid_disks-1)
+                               (*dd_idx)++;    /* Q D D D P */
+                       else if (*dd_idx >= *pd_idx)
+                               (*dd_idx) += 2; /* D D P Q D */
+                       break;
+               case ALGORITHM_RIGHT_ASYMMETRIC:
+                       *pd_idx = stripe % raid_disks;
+                       if (*pd_idx == raid_disks-1)
+                               (*dd_idx)++;    /* Q D D D P */
+                       else if (*dd_idx >= *pd_idx)
+                               (*dd_idx) += 2; /* D D P Q D */
+                       break;
+               case ALGORITHM_LEFT_SYMMETRIC:
+                       *pd_idx = raid_disks - 1 - (stripe % raid_disks);
+                       *dd_idx = (*pd_idx + 2 + *dd_idx) % raid_disks;
+                       break;
+               case ALGORITHM_RIGHT_SYMMETRIC:
+                       *pd_idx = stripe % raid_disks;
+                       *dd_idx = (*pd_idx + 2 + *dd_idx) % raid_disks;
+                       break;
+               default:
+                       printk (KERN_CRIT "raid6: unsupported algorithm %d\n",
+                               conf->algorithm);
+               }
+               break;
        }
 
        /*
@@ -742,12 +790,17 @@ static sector_t compute_blocknr(struct stripe_head *sh, int i)
        int chunk_number, dummy1, dummy2, dd_idx = i;
        sector_t r_sector;
 
+
        chunk_offset = sector_div(new_sector, sectors_per_chunk);
        stripe = new_sector;
        BUG_ON(new_sector != stripe);
 
-       
-       switch (conf->algorithm) {
+       if (i == sh->pd_idx)
+               return 0;
+       switch(conf->level) {
+       case 4: break;
+       case 5:
+               switch (conf->algorithm) {
                case ALGORITHM_LEFT_ASYMMETRIC:
                case ALGORITHM_RIGHT_ASYMMETRIC:
                        if (i > sh->pd_idx)
@@ -761,7 +814,37 @@ static sector_t compute_blocknr(struct stripe_head *sh, int i)
                        break;
                default:
                        printk(KERN_ERR "raid5: unsupported algorithm %d\n",
+                              conf->algorithm);
+               }
+               break;
+       case 6:
+               data_disks = raid_disks - 2;
+               if (i == raid6_next_disk(sh->pd_idx, raid_disks))
+                       return 0; /* It is the Q disk */
+               switch (conf->algorithm) {
+               case ALGORITHM_LEFT_ASYMMETRIC:
+               case ALGORITHM_RIGHT_ASYMMETRIC:
+                       if (sh->pd_idx == raid_disks-1)
+                               i--;    /* Q D D D P */
+                       else if (i > sh->pd_idx)
+                               i -= 2; /* D D P Q D */
+                       break;
+               case ALGORITHM_LEFT_SYMMETRIC:
+               case ALGORITHM_RIGHT_SYMMETRIC:
+                       if (sh->pd_idx == raid_disks-1)
+                               i--; /* Q D D D P */
+                       else {
+                               /* D D P Q D */
+                               if (i < sh->pd_idx)
+                                       i += raid_disks;
+                               i -= (sh->pd_idx + 2);
+                       }
+                       break;
+               default:
+                       printk (KERN_CRIT "raid6: unsupported algorithm %d\n",
                                conf->algorithm);
+               }
+               break;
        }
 
        chunk_number = stripe * data_disks + i;
@@ -778,10 +861,11 @@ static sector_t compute_blocknr(struct stripe_head *sh, int i)
 
 
 /*
- * Copy data between a page in the stripe cache, and a bio.
- * There are no alignment or size guarantees between the page or the
- * bio except that there is some overlap.
- * All iovecs in the bio must be considered.
+ * Copy data between a page in the stripe cache, and one or more bion
+ * The page could align with the middle of the bio, or there could be
+ * several bion, each with several bio_vecs, which cover part of the page
+ * Multiple bion are linked together on bi_next.  There may be extras
+ * at the end of this list.  We ignore them.
  */
 static void copy_data(int frombio, struct bio *bio,
                     struct page *page,
@@ -810,7 +894,7 @@ static void copy_data(int frombio, struct bio *bio,
                if (len > 0 && page_offset + len > STRIPE_SIZE)
                        clen = STRIPE_SIZE - page_offset;
                else clen = len;
-                       
+
                if (clen > 0) {
                        char *ba = __bio_kmap_atomic(bio, i, KM_USER0);
                        if (frombio)
@@ -862,14 +946,14 @@ static void compute_block(struct stripe_head *sh, int dd_idx)
        set_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
 }
 
-static void compute_parity(struct stripe_head *sh, int method)
+static void compute_parity5(struct stripe_head *sh, int method)
 {
        raid5_conf_t *conf = sh->raid_conf;
        int i, pd_idx = sh->pd_idx, disks = sh->disks, count;
        void *ptr[MAX_XOR_BLOCKS];
        struct bio *chosen;
 
-       PRINTK("compute_parity, stripe %llu, method %d\n",
+       PRINTK("compute_parity5, stripe %llu, method %d\n",
                (unsigned long long)sh->sector, method);
 
        count = 1;
@@ -956,9 +1040,195 @@ static void compute_parity(struct stripe_head *sh, int method)
                clear_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
 }
 
+static void compute_parity6(struct stripe_head *sh, int method)
+{
+       raid6_conf_t *conf = sh->raid_conf;
+       int i, pd_idx = sh->pd_idx, qd_idx, d0_idx, disks = conf->raid_disks, count;
+       struct bio *chosen;
+       /**** FIX THIS: This could be very bad if disks is close to 256 ****/
+       void *ptrs[disks];
+
+       qd_idx = raid6_next_disk(pd_idx, disks);
+       d0_idx = raid6_next_disk(qd_idx, disks);
+
+       PRINTK("compute_parity, stripe %llu, method %d\n",
+               (unsigned long long)sh->sector, method);
+
+       switch(method) {
+       case READ_MODIFY_WRITE:
+               BUG();          /* READ_MODIFY_WRITE N/A for RAID-6 */
+       case RECONSTRUCT_WRITE:
+               for (i= disks; i-- ;)
+                       if ( i != pd_idx && i != qd_idx && sh->dev[i].towrite ) {
+                               chosen = sh->dev[i].towrite;
+                               sh->dev[i].towrite = NULL;
+
+                               if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
+                                       wake_up(&conf->wait_for_overlap);
+
+                               if (sh->dev[i].written) BUG();
+                               sh->dev[i].written = chosen;
+                       }
+               break;
+       case CHECK_PARITY:
+               BUG();          /* Not implemented yet */
+       }
+
+       for (i = disks; i--;)
+               if (sh->dev[i].written) {
+                       sector_t sector = sh->dev[i].sector;
+                       struct bio *wbi = sh->dev[i].written;
+                       while (wbi && wbi->bi_sector < sector + STRIPE_SECTORS) {
+                               copy_data(1, wbi, sh->dev[i].page, sector);
+                               wbi = r5_next_bio(wbi, sector);
+                       }
+
+                       set_bit(R5_LOCKED, &sh->dev[i].flags);
+                       set_bit(R5_UPTODATE, &sh->dev[i].flags);
+               }
+
+//     switch(method) {
+//     case RECONSTRUCT_WRITE:
+//     case CHECK_PARITY:
+//     case UPDATE_PARITY:
+               /* Note that unlike RAID-5, the ordering of the disks matters greatly. */
+               /* FIX: Is this ordering of drives even remotely optimal? */
+               count = 0;
+               i = d0_idx;
+               do {
+                       ptrs[count++] = page_address(sh->dev[i].page);
+                       if (count <= disks-2 && !test_bit(R5_UPTODATE, &sh->dev[i].flags))
+                               printk("block %d/%d not uptodate on parity calc\n", i,count);
+                       i = raid6_next_disk(i, disks);
+               } while ( i != d0_idx );
+//             break;
+//     }
+
+       raid6_call.gen_syndrome(disks, STRIPE_SIZE, ptrs);
+
+       switch(method) {
+       case RECONSTRUCT_WRITE:
+               set_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
+               set_bit(R5_UPTODATE, &sh->dev[qd_idx].flags);
+               set_bit(R5_LOCKED,   &sh->dev[pd_idx].flags);
+               set_bit(R5_LOCKED,   &sh->dev[qd_idx].flags);
+               break;
+       case UPDATE_PARITY:
+               set_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
+               set_bit(R5_UPTODATE, &sh->dev[qd_idx].flags);
+               break;
+       }
+}
+
+
+/* Compute one missing block */
+static void compute_block_1(struct stripe_head *sh, int dd_idx, int nozero)
+{
+       raid6_conf_t *conf = sh->raid_conf;
+       int i, count, disks = conf->raid_disks;
+       void *ptr[MAX_XOR_BLOCKS], *p;
+       int pd_idx = sh->pd_idx;
+       int qd_idx = raid6_next_disk(pd_idx, disks);
+
+       PRINTK("compute_block_1, stripe %llu, idx %d\n",
+               (unsigned long long)sh->sector, dd_idx);
+
+       if ( dd_idx == qd_idx ) {
+               /* We're actually computing the Q drive */
+               compute_parity6(sh, UPDATE_PARITY);
+       } else {
+               ptr[0] = page_address(sh->dev[dd_idx].page);
+               if (!nozero) memset(ptr[0], 0, STRIPE_SIZE);
+               count = 1;
+               for (i = disks ; i--; ) {
+                       if (i == dd_idx || i == qd_idx)
+                               continue;
+                       p = page_address(sh->dev[i].page);
+                       if (test_bit(R5_UPTODATE, &sh->dev[i].flags))
+                               ptr[count++] = p;
+                       else
+                               printk("compute_block() %d, stripe %llu, %d"
+                                      " not present\n", dd_idx,
+                                      (unsigned long long)sh->sector, i);
+
+                       check_xor();
+               }
+               if (count != 1)
+                       xor_block(count, STRIPE_SIZE, ptr);
+               if (!nozero) set_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
+               else clear_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
+       }
+}
+
+/* Compute two missing blocks */
+static void compute_block_2(struct stripe_head *sh, int dd_idx1, int dd_idx2)
+{
+       raid6_conf_t *conf = sh->raid_conf;
+       int i, count, disks = conf->raid_disks;
+       int pd_idx = sh->pd_idx;
+       int qd_idx = raid6_next_disk(pd_idx, disks);
+       int d0_idx = raid6_next_disk(qd_idx, disks);
+       int faila, failb;
+
+       /* faila and failb are disk numbers relative to d0_idx */
+       /* pd_idx become disks-2 and qd_idx become disks-1 */
+       faila = (dd_idx1 < d0_idx) ? dd_idx1+(disks-d0_idx) : dd_idx1-d0_idx;
+       failb = (dd_idx2 < d0_idx) ? dd_idx2+(disks-d0_idx) : dd_idx2-d0_idx;
+
+       BUG_ON(faila == failb);
+       if ( failb < faila ) { int tmp = faila; faila = failb; failb = tmp; }
+
+       PRINTK("compute_block_2, stripe %llu, idx %d,%d (%d,%d)\n",
+              (unsigned long long)sh->sector, dd_idx1, dd_idx2, faila, failb);
+
+       if ( failb == disks-1 ) {
+               /* Q disk is one of the missing disks */
+               if ( faila == disks-2 ) {
+                       /* Missing P+Q, just recompute */
+                       compute_parity6(sh, UPDATE_PARITY);
+                       return;
+               } else {
+                       /* We're missing D+Q; recompute D from P */
+                       compute_block_1(sh, (dd_idx1 == qd_idx) ? dd_idx2 : dd_idx1, 0);
+                       compute_parity6(sh, UPDATE_PARITY); /* Is this necessary? */
+                       return;
+               }
+       }
+
+       /* We're missing D+P or D+D; build pointer table */
+       {
+               /**** FIX THIS: This could be very bad if disks is close to 256 ****/
+               void *ptrs[disks];
+
+               count = 0;
+               i = d0_idx;
+               do {
+                       ptrs[count++] = page_address(sh->dev[i].page);
+                       i = raid6_next_disk(i, disks);
+                       if (i != dd_idx1 && i != dd_idx2 &&
+                           !test_bit(R5_UPTODATE, &sh->dev[i].flags))
+                               printk("compute_2 with missing block %d/%d\n", count, i);
+               } while ( i != d0_idx );
+
+               if ( failb == disks-2 ) {
+                       /* We're missing D+P. */
+                       raid6_datap_recov(disks, STRIPE_SIZE, faila, ptrs);
+               } else {
+                       /* We're missing D+D. */
+                       raid6_2data_recov(disks, STRIPE_SIZE, faila, failb, ptrs);
+               }
+
+               /* Both the above update both missing blocks */
+               set_bit(R5_UPTODATE, &sh->dev[dd_idx1].flags);
+               set_bit(R5_UPTODATE, &sh->dev[dd_idx2].flags);
+       }
+}
+
+
+
 /*
  * Each stripe/dev can have one or more bion attached.
- * toread/towrite point to the first in a chain. 
+ * toread/towrite point to the first in a chain.
  * The bi_next chain must be in order.
  */
 static int add_stripe_bio(struct stripe_head *sh, struct bio *bi, int dd_idx, int forwrite)
@@ -1031,6 +1301,13 @@ static int add_stripe_bio(struct stripe_head *sh, struct bio *bi, int dd_idx, in
 
 static void end_reshape(raid5_conf_t *conf);
 
+static int page_is_zero(struct page *p)
+{
+       char *a = page_address(p);
+       return ((*(u32*)a) == 0 &&
+               memcmp(a, a+4, STRIPE_SIZE-4)==0);
+}
+
 static int stripe_to_pdidx(sector_t stripe, raid5_conf_t *conf, int disks)
 {
        int sectors_per_chunk = conf->chunk_size >> 9;
@@ -1062,7 +1339,7 @@ static int stripe_to_pdidx(sector_t stripe, raid5_conf_t *conf, int disks)
  *
  */
  
-static void handle_stripe(struct stripe_head *sh)
+static void handle_stripe5(struct stripe_head *sh)
 {
        raid5_conf_t *conf = sh->raid_conf;
        int disks = sh->disks;
@@ -1394,7 +1671,7 @@ static void handle_stripe(struct stripe_head *sh)
                if (locked == 0 && (rcw == 0 ||rmw == 0) &&
                    !test_bit(STRIPE_BIT_DELAY, &sh->state)) {
                        PRINTK("Computing parity...\n");
-                       compute_parity(sh, rcw==0 ? RECONSTRUCT_WRITE : READ_MODIFY_WRITE);
+                       compute_parity5(sh, rcw==0 ? RECONSTRUCT_WRITE : READ_MODIFY_WRITE);
                        /* now every locked buffer is ready to be written */
                        for (i=disks; i--;)
                                if (test_bit(R5_LOCKED, &sh->dev[i].flags)) {
@@ -1421,13 +1698,10 @@ static void handle_stripe(struct stripe_head *sh)
            !test_bit(STRIPE_INSYNC, &sh->state)) {
                set_bit(STRIPE_HANDLE, &sh->state);
                if (failed == 0) {
-                       char *pagea;
                        BUG_ON(uptodate != disks);
-                       compute_parity(sh, CHECK_PARITY);
+                       compute_parity5(sh, CHECK_PARITY);
                        uptodate--;
-                       pagea = page_address(sh->dev[sh->pd_idx].page);
-                       if ((*(u32*)pagea) == 0 &&
-                           !memcmp(pagea, pagea+4, STRIPE_SIZE-4)) {
+                       if (page_is_zero(sh->dev[sh->pd_idx].page)) {
                                /* parity is correct (on disc, not in buffer any more) */
                                set_bit(STRIPE_INSYNC, &sh->state);
                        } else {
@@ -1487,7 +1761,7 @@ static void handle_stripe(struct stripe_head *sh)
                /* Need to write out all blocks after computing parity */
                sh->disks = conf->raid_disks;
                sh->pd_idx = stripe_to_pdidx(sh->sector, conf, conf->raid_disks);
-               compute_parity(sh, RECONSTRUCT_WRITE);
+               compute_parity5(sh, RECONSTRUCT_WRITE);
                for (i= conf->raid_disks; i--;) {
                        set_bit(R5_LOCKED, &sh->dev[i].flags);
                        locked++;
@@ -1615,71 +1889,634 @@ static void handle_stripe(struct stripe_head *sh)
        }
 }
 
-static void raid5_activate_delayed(raid5_conf_t *conf)
+static void handle_stripe6(struct stripe_head *sh, struct page *tmp_page)
 {
-       if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD) {
-               while (!list_empty(&conf->delayed_list)) {
-                       struct list_head *l = conf->delayed_list.next;
-                       struct stripe_head *sh;
-                       sh = list_entry(l, struct stripe_head, lru);
-                       list_del_init(l);
-                       clear_bit(STRIPE_DELAYED, &sh->state);
-                       if (!test_and_set_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
-                               atomic_inc(&conf->preread_active_stripes);
-                       list_add_tail(&sh->lru, &conf->handle_list);
-               }
-       }
-}
+       raid6_conf_t *conf = sh->raid_conf;
+       int disks = conf->raid_disks;
+       struct bio *return_bi= NULL;
+       struct bio *bi;
+       int i;
+       int syncing;
+       int locked=0, uptodate=0, to_read=0, to_write=0, failed=0, written=0;
+       int non_overwrite = 0;
+       int failed_num[2] = {0, 0};
+       struct r5dev *dev, *pdev, *qdev;
+       int pd_idx = sh->pd_idx;
+       int qd_idx = raid6_next_disk(pd_idx, disks);
+       int p_failed, q_failed;
 
-static void activate_bit_delay(raid5_conf_t *conf)
-{
-       /* device_lock is held */
-       struct list_head head;
-       list_add(&head, &conf->bitmap_list);
-       list_del_init(&conf->bitmap_list);
-       while (!list_empty(&head)) {
-               struct stripe_head *sh = list_entry(head.next, struct stripe_head, lru);
-               list_del_init(&sh->lru);
-               atomic_inc(&sh->count);
-               __release_stripe(conf, sh);
-       }
-}
+       PRINTK("handling stripe %llu, state=%#lx cnt=%d, pd_idx=%d, qd_idx=%d\n",
+              (unsigned long long)sh->sector, sh->state, atomic_read(&sh->count),
+              pd_idx, qd_idx);
 
-static void unplug_slaves(mddev_t *mddev)
-{
-       raid5_conf_t *conf = mddev_to_conf(mddev);
-       int i;
+       spin_lock(&sh->lock);
+       clear_bit(STRIPE_HANDLE, &sh->state);
+       clear_bit(STRIPE_DELAYED, &sh->state);
+
+       syncing = test_bit(STRIPE_SYNCING, &sh->state);
+       /* Now to look around and see what can be done */
 
        rcu_read_lock();
-       for (i=0; i<mddev->raid_disks; i++) {
-               mdk_rdev_t *rdev = rcu_dereference(conf->disks[i].rdev);
-               if (rdev && !test_bit(Faulty, &rdev->flags) && atomic_read(&rdev->nr_pending)) {
-                       request_queue_t *r_queue = bdev_get_queue(rdev->bdev);
+       for (i=disks; i--; ) {
+               mdk_rdev_t *rdev;
+               dev = &sh->dev[i];
+               clear_bit(R5_Insync, &dev->flags);
 
-                       atomic_inc(&rdev->nr_pending);
-                       rcu_read_unlock();
+               PRINTK("check %d: state 0x%lx read %p write %p written %p\n",
+                       i, dev->flags, dev->toread, dev->towrite, dev->written);
+               /* maybe we can reply to a read */
+               if (test_bit(R5_UPTODATE, &dev->flags) && dev->toread) {
+                       struct bio *rbi, *rbi2;
+                       PRINTK("Return read for disc %d\n", i);
+                       spin_lock_irq(&conf->device_lock);
+                       rbi = dev->toread;
+                       dev->toread = NULL;
+                       if (test_and_clear_bit(R5_Overlap, &dev->flags))
+                               wake_up(&conf->wait_for_overlap);
+                       spin_unlock_irq(&conf->device_lock);
+                       while (rbi && rbi->bi_sector < dev->sector + STRIPE_SECTORS) {
+                               copy_data(0, rbi, dev->page, dev->sector);
+                               rbi2 = r5_next_bio(rbi, dev->sector);
+                               spin_lock_irq(&conf->device_lock);
+                               if (--rbi->bi_phys_segments == 0) {
+                                       rbi->bi_next = return_bi;
+                                       return_bi = rbi;
+                               }
+                               spin_unlock_irq(&conf->device_lock);
+                               rbi = rbi2;
+                       }
+               }
 
-                       if (r_queue->unplug_fn)
-                               r_queue->unplug_fn(r_queue);
+               /* now count some things */
+               if (test_bit(R5_LOCKED, &dev->flags)) locked++;
+               if (test_bit(R5_UPTODATE, &dev->flags)) uptodate++;
 
-                       rdev_dec_pending(rdev, mddev);
-                       rcu_read_lock();
+
+               if (dev->toread) to_read++;
+               if (dev->towrite) {
+                       to_write++;
+                       if (!test_bit(R5_OVERWRITE, &dev->flags))
+                               non_overwrite++;
+               }
+               if (dev->written) written++;
+               rdev = rcu_dereference(conf->disks[i].rdev);
+               if (!rdev || !test_bit(In_sync, &rdev->flags)) {
+                       /* The ReadError flag will just be confusing now */
+                       clear_bit(R5_ReadError, &dev->flags);
+                       clear_bit(R5_ReWrite, &dev->flags);
                }
+               if (!rdev || !test_bit(In_sync, &rdev->flags)
+                   || test_bit(R5_ReadError, &dev->flags)) {
+                       if ( failed < 2 )
+                               failed_num[failed] = i;
+                       failed++;
+               } else
+                       set_bit(R5_Insync, &dev->flags);
        }
        rcu_read_unlock();
-}
+       PRINTK("locked=%d uptodate=%d to_read=%d"
+              " to_write=%d failed=%d failed_num=%d,%d\n",
+              locked, uptodate, to_read, to_write, failed,
+              failed_num[0], failed_num[1]);
+       /* check if the array has lost >2 devices and, if so, some requests might
+        * need to be failed
+        */
+       if (failed > 2 && to_read+to_write+written) {
+               for (i=disks; i--; ) {
+                       int bitmap_end = 0;
 
-static void raid5_unplug_device(request_queue_t *q)
-{
-       mddev_t *mddev = q->queuedata;
-       raid5_conf_t *conf = mddev_to_conf(mddev);
-       unsigned long flags;
+                       if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
+                               mdk_rdev_t *rdev;
+                               rcu_read_lock();
+                               rdev = rcu_dereference(conf->disks[i].rdev);
+                               if (rdev && test_bit(In_sync, &rdev->flags))
+                                       /* multiple read failures in one stripe */
+                                       md_error(conf->mddev, rdev);
+                               rcu_read_unlock();
+                       }
 
-       spin_lock_irqsave(&conf->device_lock, flags);
+                       spin_lock_irq(&conf->device_lock);
+                       /* fail all writes first */
+                       bi = sh->dev[i].towrite;
+                       sh->dev[i].towrite = NULL;
+                       if (bi) { to_write--; bitmap_end = 1; }
 
-       if (blk_remove_plug(q)) {
-               conf->seq_flush++;
-               raid5_activate_delayed(conf);
+                       if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
+                               wake_up(&conf->wait_for_overlap);
+
+                       while (bi && bi->bi_sector < sh->dev[i].sector + STRIPE_SECTORS){
+                               struct bio *nextbi = r5_next_bio(bi, sh->dev[i].sector);
+                               clear_bit(BIO_UPTODATE, &bi->bi_flags);
+                               if (--bi->bi_phys_segments == 0) {
+                                       md_write_end(conf->mddev);
+                                       bi->bi_next = return_bi;
+                                       return_bi = bi;
+                               }
+                               bi = nextbi;
+                       }
+                       /* and fail all 'written' */
+                       bi = sh->dev[i].written;
+                       sh->dev[i].written = NULL;
+                       if (bi) bitmap_end = 1;
+                       while (bi && bi->bi_sector < sh->dev[i].sector + STRIPE_SECTORS) {
+                               struct bio *bi2 = r5_next_bio(bi, sh->dev[i].sector);
+                               clear_bit(BIO_UPTODATE, &bi->bi_flags);
+                               if (--bi->bi_phys_segments == 0) {
+                                       md_write_end(conf->mddev);
+                                       bi->bi_next = return_bi;
+                                       return_bi = bi;
+                               }
+                               bi = bi2;
+                       }
+
+                       /* fail any reads if this device is non-operational */
+                       if (!test_bit(R5_Insync, &sh->dev[i].flags) ||
+                           test_bit(R5_ReadError, &sh->dev[i].flags)) {
+                               bi = sh->dev[i].toread;
+                               sh->dev[i].toread = NULL;
+                               if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
+                                       wake_up(&conf->wait_for_overlap);
+                               if (bi) to_read--;
+                               while (bi && bi->bi_sector < sh->dev[i].sector + STRIPE_SECTORS){
+                                       struct bio *nextbi = r5_next_bio(bi, sh->dev[i].sector);
+                                       clear_bit(BIO_UPTODATE, &bi->bi_flags);
+                                       if (--bi->bi_phys_segments == 0) {
+                                               bi->bi_next = return_bi;
+                                               return_bi = bi;
+                                       }
+                                       bi = nextbi;
+                               }
+                       }
+                       spin_unlock_irq(&conf->device_lock);
+                       if (bitmap_end)
+                               bitmap_endwrite(conf->mddev->bitmap, sh->sector,
+                                               STRIPE_SECTORS, 0, 0);
+               }
+       }
+       if (failed > 2 && syncing) {
+               md_done_sync(conf->mddev, STRIPE_SECTORS,0);
+               clear_bit(STRIPE_SYNCING, &sh->state);
+               syncing = 0;
+       }
+
+       /*
+        * might be able to return some write requests if the parity blocks
+        * are safe, or on a failed drive
+        */
+       pdev = &sh->dev[pd_idx];
+       p_failed = (failed >= 1 && failed_num[0] == pd_idx)
+               || (failed >= 2 && failed_num[1] == pd_idx);
+       qdev = &sh->dev[qd_idx];
+       q_failed = (failed >= 1 && failed_num[0] == qd_idx)
+               || (failed >= 2 && failed_num[1] == qd_idx);
+
+       if ( written &&
+            ( p_failed || ((test_bit(R5_Insync, &pdev->flags)
+                            && !test_bit(R5_LOCKED, &pdev->flags)
+                            && test_bit(R5_UPTODATE, &pdev->flags))) ) &&
+            ( q_failed || ((test_bit(R5_Insync, &qdev->flags)
+                            && !test_bit(R5_LOCKED, &qdev->flags)
+                            && test_bit(R5_UPTODATE, &qdev->flags))) ) ) {
+               /* any written block on an uptodate or failed drive can be
+                * returned.  Note that if we 'wrote' to a failed drive,
+                * it will be UPTODATE, but never LOCKED, so we don't need
+                * to test 'failed' directly.
+                */
+               for (i=disks; i--; )
+                       if (sh->dev[i].written) {
+                               dev = &sh->dev[i];
+                               if (!test_bit(R5_LOCKED, &dev->flags) &&
+                                   test_bit(R5_UPTODATE, &dev->flags) ) {
+                                       /* We can return any write requests */
+                                       int bitmap_end = 0;
+                                       struct bio *wbi, *wbi2;
+                                       PRINTK("Return write for stripe %llu disc %d\n",
+                                              (unsigned long long)sh->sector, i);
+                                       spin_lock_irq(&conf->device_lock);
+                                       wbi = dev->written;
+                                       dev->written = NULL;
+                                       while (wbi && wbi->bi_sector < dev->sector + STRIPE_SECTORS) {
+                                               wbi2 = r5_next_bio(wbi, dev->sector);
+                                               if (--wbi->bi_phys_segments == 0) {
+                                                       md_write_end(conf->mddev);
+                                                       wbi->bi_next = return_bi;
+                                                       return_bi = wbi;
+                                               }
+                                               wbi = wbi2;
+                                       }
+                                       if (dev->towrite == NULL)
+                                               bitmap_end = 1;
+                                       spin_unlock_irq(&conf->device_lock);
+                                       if (bitmap_end)
+                                               bitmap_endwrite(conf->mddev->bitmap, sh->sector,
+                                                               STRIPE_SECTORS,
+                                                               !test_bit(STRIPE_DEGRADED, &sh->state), 0);
+                               }
+                       }
+       }
+
+       /* Now we might consider reading some blocks, either to check/generate
+        * parity, or to satisfy requests
+        * or to load a block that is being partially written.
+        */
+       if (to_read || non_overwrite || (to_write && failed) || (syncing && (uptodate < disks))) {
+               for (i=disks; i--;) {
+                       dev = &sh->dev[i];
+                       if (!test_bit(R5_LOCKED, &dev->flags) && !test_bit(R5_UPTODATE, &dev->flags) &&
+                           (dev->toread ||
+                            (dev->towrite && !test_bit(R5_OVERWRITE, &dev->flags)) ||
+                            syncing ||
+                            (failed >= 1 && (sh->dev[failed_num[0]].toread || to_write)) ||
+                            (failed >= 2 && (sh->dev[failed_num[1]].toread || to_write))
+                                   )
+                               ) {
+                               /* we would like to get this block, possibly
+                                * by computing it, but we might not be able to
+                                */
+                               if (uptodate == disks-1) {
+                                       PRINTK("Computing stripe %llu block %d\n",
+                                              (unsigned long long)sh->sector, i);
+                                       compute_block_1(sh, i, 0);
+                                       uptodate++;
+                               } else if ( uptodate == disks-2 && failed >= 2 ) {
+                                       /* Computing 2-failure is *very* expensive; only do it if failed >= 2 */
+                                       int other;
+                                       for (other=disks; other--;) {
+                                               if ( other == i )
+                                                       continue;
+                                               if ( !test_bit(R5_UPTODATE, &sh->dev[other].flags) )
+                                                       break;
+                                       }
+                                       BUG_ON(other < 0);
+                                       PRINTK("Computing stripe %llu blocks %d,%d\n",
+                                              (unsigned long long)sh->sector, i, other);
+                                       compute_block_2(sh, i, other);
+                                       uptodate += 2;
+                               } else if (test_bit(R5_Insync, &dev->flags)) {
+                                       set_bit(R5_LOCKED, &dev->flags);
+                                       set_bit(R5_Wantread, &dev->flags);
+#if 0
+                                       /* if I am just reading this block and we don't have
+                                          a failed drive, or any pending writes then sidestep the cache */
+                                       if (sh->bh_read[i] && !sh->bh_read[i]->b_reqnext &&
+                                           ! syncing && !failed && !to_write) {
+                                               sh->bh_cache[i]->b_page =  sh->bh_read[i]->b_page;
+                                               sh->bh_cache[i]->b_data =  sh->bh_read[i]->b_data;
+                                       }
+#endif
+                                       locked++;
+                                       PRINTK("Reading block %d (sync=%d)\n",
+                                               i, syncing);
+                               }
+                       }
+               }
+               set_bit(STRIPE_HANDLE, &sh->state);
+       }
+
+       /* now to consider writing and what else, if anything should be read */
+       if (to_write) {
+               int rcw=0, must_compute=0;
+               for (i=disks ; i--;) {
+                       dev = &sh->dev[i];
+                       /* Would I have to read this buffer for reconstruct_write */
+                       if (!test_bit(R5_OVERWRITE, &dev->flags)
+                           && i != pd_idx && i != qd_idx
+                           && (!test_bit(R5_LOCKED, &dev->flags)
+#if 0
+                               || sh->bh_page[i] != bh->b_page
+#endif
+                                   ) &&
+                           !test_bit(R5_UPTODATE, &dev->flags)) {
+                               if (test_bit(R5_Insync, &dev->flags)) rcw++;
+                               else {
+                                       PRINTK("raid6: must_compute: disk %d flags=%#lx\n", i, dev->flags);
+                                       must_compute++;
+                               }
+                       }
+               }
+               PRINTK("for sector %llu, rcw=%d, must_compute=%d\n",
+                      (unsigned long long)sh->sector, rcw, must_compute);
+               set_bit(STRIPE_HANDLE, &sh->state);
+
+               if (rcw > 0)
+                       /* want reconstruct write, but need to get some data */
+                       for (i=disks; i--;) {
+                               dev = &sh->dev[i];
+                               if (!test_bit(R5_OVERWRITE, &dev->flags)
+                                   && !(failed == 0 && (i == pd_idx || i == qd_idx))
+                                   && !test_bit(R5_LOCKED, &dev->flags) && !test_bit(R5_UPTODATE, &dev->flags) &&
+                                   test_bit(R5_Insync, &dev->flags)) {
+                                       if (test_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
+                                       {
+                                               PRINTK("Read_old stripe %llu block %d for Reconstruct\n",
+                                                      (unsigned long long)sh->sector, i);
+                                               set_bit(R5_LOCKED, &dev->flags);
+                                               set_bit(R5_Wantread, &dev->flags);
+                                               locked++;
+                                       } else {
+                                               PRINTK("Request delayed stripe %llu block %d for Reconstruct\n",
+                                                      (unsigned long long)sh->sector, i);
+                                               set_bit(STRIPE_DELAYED, &sh->state);
+                                               set_bit(STRIPE_HANDLE, &sh->state);
+                                       }
+                               }
+                       }
+               /* now if nothing is locked, and if we have enough data, we can start a write request */
+               if (locked == 0 && rcw == 0 &&
+                   !test_bit(STRIPE_BIT_DELAY, &sh->state)) {
+                       if ( must_compute > 0 ) {
+                               /* We have failed blocks and need to compute them */
+                               switch ( failed ) {
+                               case 0: BUG();
+                               case 1: compute_block_1(sh, failed_num[0], 0); break;
+                               case 2: compute_block_2(sh, failed_num[0], failed_num[1]); break;
+                               default: BUG(); /* This request should have been failed? */
+                               }
+                       }
+
+                       PRINTK("Computing parity for stripe %llu\n", (unsigned long long)sh->sector);
+                       compute_parity6(sh, RECONSTRUCT_WRITE);
+                       /* now every locked buffer is ready to be written */
+                       for (i=disks; i--;)
+                               if (test_bit(R5_LOCKED, &sh->dev[i].flags)) {
+                                       PRINTK("Writing stripe %llu block %d\n",
+                                              (unsigned long long)sh->sector, i);
+                                       locked++;
+                                       set_bit(R5_Wantwrite, &sh->dev[i].flags);
+                               }
+                       /* after a RECONSTRUCT_WRITE, the stripe MUST be in-sync */
+                       set_bit(STRIPE_INSYNC, &sh->state);
+
+                       if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
+                               atomic_dec(&conf->preread_active_stripes);
+                               if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD)
+                                       md_wakeup_thread(conf->mddev->thread);
+                       }
+               }
+       }
+
+       /* maybe we need to check and possibly fix the parity for this stripe
+        * Any reads will already have been scheduled, so we just see if enough data
+        * is available
+        */
+       if (syncing && locked == 0 && !test_bit(STRIPE_INSYNC, &sh->state)) {
+               int update_p = 0, update_q = 0;
+               struct r5dev *dev;
+
+               set_bit(STRIPE_HANDLE, &sh->state);
+
+               BUG_ON(failed>2);
+               BUG_ON(uptodate < disks);
+               /* Want to check and possibly repair P and Q.
+                * However there could be one 'failed' device, in which
+                * case we can only check one of them, possibly using the
+                * other to generate missing data
+                */
+
+               /* If !tmp_page, we cannot do the calculations,
+                * but as we have set STRIPE_HANDLE, we will soon be called
+                * by stripe_handle with a tmp_page - just wait until then.
+                */
+               if (tmp_page) {
+                       if (failed == q_failed) {
+                               /* The only possible failed device holds 'Q', so it makes
+                                * sense to check P (If anything else were failed, we would
+                                * have used P to recreate it).
+                                */
+                               compute_block_1(sh, pd_idx, 1);
+                               if (!page_is_zero(sh->dev[pd_idx].page)) {
+                                       compute_block_1(sh,pd_idx,0);
+                                       update_p = 1;
+                               }
+                       }
+                       if (!q_failed && failed < 2) {
+                               /* q is not failed, and we didn't use it to generate
+                                * anything, so it makes sense to check it
+                                */
+                               memcpy(page_address(tmp_page),
+                                      page_address(sh->dev[qd_idx].page),
+                                      STRIPE_SIZE);
+                               compute_parity6(sh, UPDATE_PARITY);
+                               if (memcmp(page_address(tmp_page),
+                                          page_address(sh->dev[qd_idx].page),
+                                          STRIPE_SIZE)!= 0) {
+                                       clear_bit(STRIPE_INSYNC, &sh->state);
+                                       update_q = 1;
+                               }
+                       }
+                       if (update_p || update_q) {
+                               conf->mddev->resync_mismatches += STRIPE_SECTORS;
+                               if (test_bit(MD_RECOVERY_CHECK, &conf->mddev->recovery))
+                                       /* don't try to repair!! */
+                                       update_p = update_q = 0;
+                       }
+
+                       /* now write out any block on a failed drive,
+                        * or P or Q if they need it
+                        */
+
+                       if (failed == 2) {
+                               dev = &sh->dev[failed_num[1]];
+                               locked++;
+                               set_bit(R5_LOCKED, &dev->flags);
+                               set_bit(R5_Wantwrite, &dev->flags);
+                       }
+                       if (failed >= 1) {
+                               dev = &sh->dev[failed_num[0]];
+                               locked++;
+                               set_bit(R5_LOCKED, &dev->flags);
+                               set_bit(R5_Wantwrite, &dev->flags);
+                       }
+
+                       if (update_p) {
+                               dev = &sh->dev[pd_idx];
+                               locked ++;
+                               set_bit(R5_LOCKED, &dev->flags);
+                               set_bit(R5_Wantwrite, &dev->flags);
+                       }
+                       if (update_q) {
+                               dev = &sh->dev[qd_idx];
+                               locked++;
+                               set_bit(R5_LOCKED, &dev->flags);
+                               set_bit(R5_Wantwrite, &dev->flags);
+                       }
+                       clear_bit(STRIPE_DEGRADED, &sh->state);
+
+                       set_bit(STRIPE_INSYNC, &sh->state);
+               }
+       }
+
+       if (syncing && locked == 0 && test_bit(STRIPE_INSYNC, &sh->state)) {
+               md_done_sync(conf->mddev, STRIPE_SECTORS,1);
+               clear_bit(STRIPE_SYNCING, &sh->state);
+       }
+
+       /* If the failed drives are just a ReadError, then we might need
+        * to progress the repair/check process
+        */
+       if (failed <= 2 && ! conf->mddev->ro)
+               for (i=0; i<failed;i++) {
+                       dev = &sh->dev[failed_num[i]];
+                       if (test_bit(R5_ReadError, &dev->flags)
+                           && !test_bit(R5_LOCKED, &dev->flags)
+                           && test_bit(R5_UPTODATE, &dev->flags)
+                               ) {
+                               if (!test_bit(R5_ReWrite, &dev->flags)) {
+                                       set_bit(R5_Wantwrite, &dev->flags);
+                                       set_bit(R5_ReWrite, &dev->flags);
+                                       set_bit(R5_LOCKED, &dev->flags);
+                               } else {
+                                       /* let's read it back */
+                                       set_bit(R5_Wantread, &dev->flags);
+                                       set_bit(R5_LOCKED, &dev->flags);
+                               }
+                       }
+               }
+       spin_unlock(&sh->lock);
+
+       while ((bi=return_bi)) {
+               int bytes = bi->bi_size;
+
+               return_bi = bi->bi_next;
+               bi->bi_next = NULL;
+               bi->bi_size = 0;
+               bi->bi_end_io(bi, bytes, 0);
+       }
+       for (i=disks; i-- ;) {
+               int rw;
+               struct bio *bi;
+               mdk_rdev_t *rdev;
+               if (test_and_clear_bit(R5_Wantwrite, &sh->dev[i].flags))
+                       rw = 1;
+               else if (test_and_clear_bit(R5_Wantread, &sh->dev[i].flags))
+                       rw = 0;
+               else
+                       continue;
+
+               bi = &sh->dev[i].req;
+
+               bi->bi_rw = rw;
+               if (rw)
+                       bi->bi_end_io = raid5_end_write_request;
+               else
+                       bi->bi_end_io = raid5_end_read_request;
+
+               rcu_read_lock();
+               rdev = rcu_dereference(conf->disks[i].rdev);
+               if (rdev && test_bit(Faulty, &rdev->flags))
+                       rdev = NULL;
+               if (rdev)
+                       atomic_inc(&rdev->nr_pending);
+               rcu_read_unlock();
+
+               if (rdev) {
+                       if (syncing)
+                               md_sync_acct(rdev->bdev, STRIPE_SECTORS);
+
+                       bi->bi_bdev = rdev->bdev;
+                       PRINTK("for %llu schedule op %ld on disc %d\n",
+                               (unsigned long long)sh->sector, bi->bi_rw, i);
+                       atomic_inc(&sh->count);
+                       bi->bi_sector = sh->sector + rdev->data_offset;
+                       bi->bi_flags = 1 << BIO_UPTODATE;
+                       bi->bi_vcnt = 1;
+                       bi->bi_max_vecs = 1;
+                       bi->bi_idx = 0;
+                       bi->bi_io_vec = &sh->dev[i].vec;
+                       bi->bi_io_vec[0].bv_len = STRIPE_SIZE;
+                       bi->bi_io_vec[0].bv_offset = 0;
+                       bi->bi_size = STRIPE_SIZE;
+                       bi->bi_next = NULL;
+                       if (rw == WRITE &&
+                           test_bit(R5_ReWrite, &sh->dev[i].flags))
+                               atomic_add(STRIPE_SECTORS, &rdev->corrected_errors);
+                       generic_make_request(bi);
+               } else {
+                       if (rw == 1)
+                               set_bit(STRIPE_DEGRADED, &sh->state);
+                       PRINTK("skip op %ld on disc %d for sector %llu\n",
+                               bi->bi_rw, i, (unsigned long long)sh->sector);
+                       clear_bit(R5_LOCKED, &sh->dev[i].flags);
+                       set_bit(STRIPE_HANDLE, &sh->state);
+               }
+       }
+}
+
+static void handle_stripe(struct stripe_head *sh, struct page *tmp_page)
+{
+       if (sh->raid_conf->level == 6)
+               handle_stripe6(sh, tmp_page);
+       else
+               handle_stripe5(sh);
+}
+
+
+
+static void raid5_activate_delayed(raid5_conf_t *conf)
+{
+       if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD) {
+               while (!list_empty(&conf->delayed_list)) {
+                       struct list_head *l = conf->delayed_list.next;
+                       struct stripe_head *sh;
+                       sh = list_entry(l, struct stripe_head, lru);
+                       list_del_init(l);
+                       clear_bit(STRIPE_DELAYED, &sh->state);
+                       if (!test_and_set_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
+                               atomic_inc(&conf->preread_active_stripes);
+                       list_add_tail(&sh->lru, &conf->handle_list);
+               }
+       }
+}
+
+static void activate_bit_delay(raid5_conf_t *conf)
+{
+       /* device_lock is held */
+       struct list_head head;
+       list_add(&head, &conf->bitmap_list);
+       list_del_init(&conf->bitmap_list);
+       while (!list_empty(&head)) {
+               struct stripe_head *sh = list_entry(head.next, struct stripe_head, lru);
+               list_del_init(&sh->lru);
+               atomic_inc(&sh->count);
+               __release_stripe(conf, sh);
+       }
+}
+
+static void unplug_slaves(mddev_t *mddev)
+{
+       raid5_conf_t *conf = mddev_to_conf(mddev);
+       int i;
+
+       rcu_read_lock();
+       for (i=0; i<mddev->raid_disks; i++) {
+               mdk_rdev_t *rdev = rcu_dereference(conf->disks[i].rdev);
+               if (rdev && !test_bit(Faulty, &rdev->flags) && atomic_read(&rdev->nr_pending)) {
+                       request_queue_t *r_queue = bdev_get_queue(rdev->bdev);
+
+                       atomic_inc(&rdev->nr_pending);
+                       rcu_read_unlock();
+
+                       if (r_queue->unplug_fn)
+                               r_queue->unplug_fn(r_queue);
+
+                       rdev_dec_pending(rdev, mddev);
+                       rcu_read_lock();
+               }
+       }
+       rcu_read_unlock();
+}
+
+static void raid5_unplug_device(request_queue_t *q)
+{
+       mddev_t *mddev = q->queuedata;
+       raid5_conf_t *conf = mddev_to_conf(mddev);
+       unsigned long flags;
+
+       spin_lock_irqsave(&conf->device_lock, flags);
+
+       if (blk_remove_plug(q)) {
+               conf->seq_flush++;
+               raid5_activate_delayed(conf);
        }
        md_wakeup_thread(mddev->thread);
 
@@ -1753,7 +2590,7 @@ static int make_request(request_queue_t *q, struct bio * bi)
 
        for (;logical_sector < last_sector; logical_sector += STRIPE_SECTORS) {
                DEFINE_WAIT(w);
-               int disks;
+               int disks, data_disks;
 
        retry:
                prepare_to_wait(&conf->wait_for_overlap, &w, TASK_UNINTERRUPTIBLE);
@@ -1781,7 +2618,9 @@ static int make_request(request_queue_t *q, struct bio * bi)
                        }
                        spin_unlock_irq(&conf->device_lock);
                }
-               new_sector = raid5_compute_sector(logical_sector, disks, disks - 1,
+               data_disks = disks - conf->max_degraded;
+
+               new_sector = raid5_compute_sector(logical_sector, disks, data_disks,
                                                  &dd_idx, &pd_idx, conf);
                PRINTK("raid5: make_request, sector %llu logical %llu\n",
                        (unsigned long long)new_sector, 
@@ -1833,7 +2672,7 @@ static int make_request(request_queue_t *q, struct bio * bi)
                        }
                        finish_wait(&conf->wait_for_overlap, &w);
                        raid5_plug_device(conf);
-                       handle_stripe(sh);
+                       handle_stripe(sh, NULL);
                        release_stripe(sh);
                } else {
                        /* cannot get stripe for read-ahead, just give-up */
@@ -1849,7 +2688,7 @@ static int make_request(request_queue_t *q, struct bio * bi)
        if (remaining == 0) {
                int bytes = bi->bi_size;
 
-               if ( bio_data_dir(bi) == WRITE )
+               if ( rw == WRITE )
                        md_write_end(mddev);
                bi->bi_size = 0;
                bi->bi_end_io(bi, bytes, 0);
@@ -1865,9 +2704,11 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
        int pd_idx;
        sector_t first_sector, last_sector;
        int raid_disks = conf->raid_disks;
-       int data_disks = raid_disks-1;
+       int data_disks = raid_disks - conf->max_degraded;
        sector_t max_sector = mddev->size << 1;
        int sync_blocks;
+       int still_degraded = 0;
+       int i;
 
        if (sector_nr >= max_sector) {
                /* just being told to finish up .. nothing much to do */
@@ -1880,7 +2721,7 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
                if (mddev->curr_resync < max_sector) /* aborted */
                        bitmap_end_sync(mddev->bitmap, mddev->curr_resync,
                                        &sync_blocks, 1);
-               else /* compelted sync */
+               else /* completed sync */
                        conf->fullsync = 0;
                bitmap_close_sync(mddev->bitmap);
 
@@ -2003,11 +2844,12 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
                }
                return conf->chunk_size>>9;
        }
-       /* if there is 1 or more failed drives and we are trying
+       /* if there is too many failed drives and we are trying
         * to resync, then assert that we are finished, because there is
         * nothing we can do.
         */
-       if (mddev->degraded >= 1 && test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
+       if (mddev->degraded >= (data_disks - raid_disks) &&
+           test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
                sector_t rv = (mddev->size << 1) - sector_nr;
                *skipped = 1;
                return rv;
@@ -2026,17 +2868,26 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
        if (sh == NULL) {
                sh = get_active_stripe(conf, sector_nr, raid_disks, pd_idx, 0);
                /* make sure we don't swamp the stripe cache if someone else
-                * is trying to get access 
+                * is trying to get access
                 */
                schedule_timeout_uninterruptible(1);
        }
-       bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, 0);
-       spin_lock(&sh->lock);   
+       /* Need to check if array will still be degraded after recovery/resync
+        * We don't need to check the 'failed' flag as when that gets set,
+        * recovery aborts.
+        */
+       for (i=0; i<mddev->raid_disks; i++)
+               if (conf->disks[i].rdev == NULL)
+                       still_degraded = 1;
+
+       bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, still_degraded);
+
+       spin_lock(&sh->lock);
        set_bit(STRIPE_SYNCING, &sh->state);
        clear_bit(STRIPE_INSYNC, &sh->state);
        spin_unlock(&sh->lock);
 
-       handle_stripe(sh);
+       handle_stripe(sh, NULL);
        release_stripe(sh);
 
        return STRIPE_SECTORS;
@@ -2091,7 +2942,7 @@ static void raid5d (mddev_t *mddev)
                spin_unlock_irq(&conf->device_lock);
                
                handled++;
-               handle_stripe(sh);
+               handle_stripe(sh, conf->spare_page);
                release_stripe(sh);
 
                spin_lock_irq(&conf->device_lock);
@@ -2181,8 +3032,8 @@ static int run(mddev_t *mddev)
        struct disk_info *disk;
        struct list_head *tmp;
 
-       if (mddev->level != 5 && mddev->level != 4) {
-               printk(KERN_ERR "raid5: %s: raid level not set to 4/5 (%d)\n",
+       if (mddev->level != 5 && mddev->level != 4 && mddev->level != 6) {
+               printk(KERN_ERR "raid5: %s: raid level not set to 4/5/6 (%d)\n",
                       mdname(mddev), mddev->level);
                return -EIO;
        }
@@ -2251,6 +3102,11 @@ static int run(mddev_t *mddev)
        if ((conf->stripe_hashtbl = kzalloc(PAGE_SIZE, GFP_KERNEL)) == NULL)
                goto abort;
 
+       if (mddev->level == 6) {
+               conf->spare_page = alloc_page(GFP_KERNEL);
+               if (!conf->spare_page)
+                       goto abort;
+       }
        spin_lock_init(&conf->device_lock);
        init_waitqueue_head(&conf->wait_for_stripe);
        init_waitqueue_head(&conf->wait_for_overlap);
@@ -2282,12 +3138,16 @@ static int run(mddev_t *mddev)
        }
 
        /*
-        * 0 for a fully functional array, 1 for a degraded array.
+        * 0 for a fully functional array, 1 or 2 for a degraded array.
         */
        mddev->degraded = conf->failed_disks = conf->raid_disks - conf->working_disks;
        conf->mddev = mddev;
        conf->chunk_size = mddev->chunk_size;
        conf->level = mddev->level;
+       if (conf->level == 6)
+               conf->max_degraded = 2;
+       else
+               conf->max_degraded = 1;
        conf->algorithm = mddev->layout;
        conf->max_nr_stripes = NR_STRIPES;
        conf->expand_progress = mddev->reshape_position;
@@ -2296,6 +3156,11 @@ static int run(mddev_t *mddev)
        mddev->size &= ~(mddev->chunk_size/1024 -1);
        mddev->resync_max_sectors = mddev->size << 1;
 
+       if (conf->level == 6 && conf->raid_disks < 4) {
+               printk(KERN_ERR "raid6: not enough configured devices for %s (%d, minimum 4)\n",
+                      mdname(mddev), conf->raid_disks);
+               goto abort;
+       }
        if (!conf->chunk_size || conf->chunk_size % 4) {
                printk(KERN_ERR "raid5: invalid chunk size %d for %s\n",
                        conf->chunk_size, mdname(mddev));
@@ -2307,14 +3172,14 @@ static int run(mddev_t *mddev)
                        conf->algorithm, mdname(mddev));
                goto abort;
        }
-       if (mddev->degraded > 1) {
+       if (mddev->degraded > conf->max_degraded) {
                printk(KERN_ERR "raid5: not enough operational devices for %s"
                        " (%d/%d failed)\n",
                        mdname(mddev), conf->failed_disks, conf->raid_disks);
                goto abort;
        }
 
-       if (mddev->degraded == 1 &&
+       if (mddev->degraded > 0 &&
            mddev->recovery_cp != MaxSector) {
                if (mddev->ok_start_degraded)
                        printk(KERN_WARNING
@@ -2379,10 +3244,11 @@ static int run(mddev_t *mddev)
        }
 
        /* read-ahead size must cover two whole stripes, which is
-        * 2 * (n-1) * chunksize where 'n' is the number of raid devices
+        * 2 * (datadisks) * chunksize where 'n' is the number of raid devices
         */
        {
-               int stripe = (mddev->raid_disks-1) *
+               int data_disks = conf->previous_raid_disks - conf->max_degraded;
+               int stripe = data_disks *
                        (mddev->chunk_size / PAGE_SIZE);
                if (mddev->queue->backing_dev_info.ra_pages < 2 * stripe)
                        mddev->queue->backing_dev_info.ra_pages = 2 * stripe;
@@ -2393,12 +3259,14 @@ static int run(mddev_t *mddev)
 
        mddev->queue->unplug_fn = raid5_unplug_device;
        mddev->queue->issue_flush_fn = raid5_issue_flush;
-       mddev->array_size =  mddev->size * (conf->previous_raid_disks - 1);
+       mddev->array_size =  mddev->size * (conf->previous_raid_disks -
+                                           conf->max_degraded);
 
        return 0;
 abort:
        if (conf) {
                print_raid5_conf(conf);
+               safe_put_page(conf->spare_page);
                kfree(conf->disks);
                kfree(conf->stripe_hashtbl);
                kfree(conf);
@@ -2427,23 +3295,23 @@ static int stop(mddev_t *mddev)
 }
 
 #if RAID5_DEBUG
-static void print_sh (struct stripe_head *sh)
+static void print_sh (struct seq_file *seq, struct stripe_head *sh)
 {
        int i;
 
-       printk("sh %llu, pd_idx %d, state %ld.\n",
-               (unsigned long long)sh->sector, sh->pd_idx, sh->state);
-       printk("sh %llu,  count %d.\n",
-               (unsigned long long)sh->sector, atomic_read(&sh->count));
-       printk("sh %llu, ", (unsigned long long)sh->sector);
+       seq_printf(seq, "sh %llu, pd_idx %d, state %ld.\n",
+                  (unsigned long long)sh->sector, sh->pd_idx, sh->state);
+       seq_printf(seq, "sh %llu,  count %d.\n",
+                  (unsigned long long)sh->sector, atomic_read(&sh->count));
+       seq_printf(seq, "sh %llu, ", (unsigned long long)sh->sector);
        for (i = 0; i < sh->disks; i++) {
-               printk("(cache%d: %p %ld) ", 
-                       i, sh->dev[i].page, sh->dev[i].flags);
+               seq_printf(seq, "(cache%d: %p %ld) ",
+                          i, sh->dev[i].page, sh->dev[i].flags);
        }
-       printk("\n");
+       seq_printf(seq, "\n");
 }
 
-static void printall (raid5_conf_t *conf)
+static void printall (struct seq_file *seq, raid5_conf_t *conf)
 {
        struct stripe_head *sh;
        struct hlist_node *hn;
@@ -2454,7 +3322,7 @@ static void printall (raid5_conf_t *conf)
                hlist_for_each_entry(sh, hn, &conf->stripe_hashtbl[i], hash) {
                        if (sh->raid_conf != conf)
                                continue;
-                       print_sh(sh);
+                       print_sh(seq, sh);
                }
        }
        spin_unlock_irq(&conf->device_lock);
@@ -2474,9 +3342,8 @@ static void status (struct seq_file *seq, mddev_t *mddev)
                               test_bit(In_sync, &conf->disks[i].rdev->flags) ? "U" : "_");
        seq_printf (seq, "]");
 #if RAID5_DEBUG
-#define D(x) \
-       seq_printf (seq, "<"#x":%d>", atomic_read(&conf->x))
-       printall(conf);
+       seq_printf (seq, "\n");
+       printall(seq, conf);
 #endif
 }
 
@@ -2560,14 +3427,20 @@ static int raid5_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
        int disk;
        struct disk_info *p;
 
-       if (mddev->degraded > 1)
+       if (mddev->degraded > conf->max_degraded)
                /* no point adding a device */
                return 0;
 
        /*
-        * find the disk ...
+        * find the disk ... but prefer rdev->saved_raid_disk
+        * if possible.
         */
-       for (disk=0; disk < conf->raid_disks; disk++)
+       if (rdev->saved_raid_disk >= 0 &&
+           conf->disks[rdev->saved_raid_disk].rdev == NULL)
+               disk = rdev->saved_raid_disk;
+       else
+               disk = 0;
+       for ( ; disk < conf->raid_disks; disk++)
                if ((p=conf->disks + disk)->rdev == NULL) {
                        clear_bit(In_sync, &rdev->flags);
                        rdev->raid_disk = disk;
@@ -2590,8 +3463,10 @@ static int raid5_resize(mddev_t *mddev, sector_t sectors)
         * any io in the removed space completes, but it hardly seems
         * worth it.
         */
+       raid5_conf_t *conf = mddev_to_conf(mddev);
+
        sectors &= ~((sector_t)mddev->chunk_size/512 - 1);
-       mddev->array_size = (sectors * (mddev->raid_disks-1))>>1;
+       mddev->array_size = (sectors * (mddev->raid_disks-conf->max_degraded))>>1;
        set_capacity(mddev->gendisk, mddev->array_size << 1);
        mddev->changed = 1;
        if (sectors/2  > mddev->size && mddev->recovery_cp == MaxSector) {
@@ -2731,6 +3606,17 @@ static void end_reshape(raid5_conf_t *conf)
                conf->expand_progress = MaxSector;
                spin_unlock_irq(&conf->device_lock);
                conf->mddev->reshape_position = MaxSector;
+
+               /* read-ahead size must cover two whole stripes, which is
+                * 2 * (datadisks) * chunksize where 'n' is the number of raid devices
+                */
+               {
+                       int data_disks = conf->previous_raid_disks - conf->max_degraded;
+                       int stripe = data_disks *
+                               (conf->mddev->chunk_size / PAGE_SIZE);
+                       if (conf->mddev->queue->backing_dev_info.ra_pages < 2 * stripe)
+                               conf->mddev->queue->backing_dev_info.ra_pages = 2 * stripe;
+               }
        }
 }
 
@@ -2762,6 +3648,23 @@ static void raid5_quiesce(mddev_t *mddev, int state)
        }
 }
 
+static struct mdk_personality raid6_personality =
+{
+       .name           = "raid6",
+       .level          = 6,
+       .owner          = THIS_MODULE,
+       .make_request   = make_request,
+       .run            = run,
+       .stop           = stop,
+       .status         = status,
+       .error_handler  = error,
+       .hot_add_disk   = raid5_add_disk,
+       .hot_remove_disk= raid5_remove_disk,
+       .spare_active   = raid5_spare_active,
+       .sync_request   = sync_request,
+       .resize         = raid5_resize,
+       .quiesce        = raid5_quiesce,
+};
 static struct mdk_personality raid5_personality =
 {
        .name           = "raid5",
@@ -2804,6 +3707,12 @@ static struct mdk_personality raid4_personality =
 
 static int __init raid5_init(void)
 {
+       int e;
+
+       e = raid6_select_algo();
+       if ( e )
+               return e;
+       register_md_personality(&raid6_personality);
        register_md_personality(&raid5_personality);
        register_md_personality(&raid4_personality);
        return 0;
@@ -2811,6 +3720,7 @@ static int __init raid5_init(void)
 
 static void raid5_exit(void)
 {
+       unregister_md_personality(&raid6_personality);
        unregister_md_personality(&raid5_personality);
        unregister_md_personality(&raid4_personality);
 }
@@ -2823,3 +3733,10 @@ MODULE_ALIAS("md-raid5");
 MODULE_ALIAS("md-raid4");
 MODULE_ALIAS("md-level-5");
 MODULE_ALIAS("md-level-4");
+MODULE_ALIAS("md-personality-8"); /* RAID6 */
+MODULE_ALIAS("md-raid6");
+MODULE_ALIAS("md-level-6");
+
+/* This used to be two separate modules, they were: */
+MODULE_ALIAS("raid5");
+MODULE_ALIAS("raid6");
diff --git a/drivers/md/raid6main.c b/drivers/md/raid6main.c
deleted file mode 100644 (file)
index e53d2d9..0000000
+++ /dev/null
@@ -1,2427 +0,0 @@
-/*
- * raid6main.c : Multiple Devices driver for Linux
- *        Copyright (C) 1996, 1997 Ingo Molnar, Miguel de Icaza, Gadi Oxman
- *        Copyright (C) 1999, 2000 Ingo Molnar
- *        Copyright (C) 2002, 2003 H. Peter Anvin
- *
- * RAID-6 management functions.  This code is derived from raid5.c.
- * Last merge from raid5.c bkcvs version 1.79 (kernel 2.6.1).
- *
- * Thanks to Penguin Computing for making the RAID-6 development possible
- * by donating a test server!
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2, or (at your option)
- * any later version.
- *
- * You should have received a copy of the GNU General Public License
- * (for example /usr/src/linux/COPYING); if not, write to the Free
- * Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
- */
-
-
-#include <linux/config.h>
-#include <linux/module.h>
-#include <linux/slab.h>
-#include <linux/highmem.h>
-#include <linux/bitops.h>
-#include <asm/atomic.h>
-#include "raid6.h"
-
-#include <linux/raid/bitmap.h>
-
-/*
- * Stripe cache
- */
-
-#define NR_STRIPES             256
-#define STRIPE_SIZE            PAGE_SIZE
-#define STRIPE_SHIFT           (PAGE_SHIFT - 9)
-#define STRIPE_SECTORS         (STRIPE_SIZE>>9)
-#define        IO_THRESHOLD            1
-#define NR_HASH                        (PAGE_SIZE / sizeof(struct hlist_head))
-#define HASH_MASK              (NR_HASH - 1)
-
-#define stripe_hash(conf, sect)        (&((conf)->stripe_hashtbl[((sect) >> STRIPE_SHIFT) & HASH_MASK]))
-
-/* bio's attached to a stripe+device for I/O are linked together in bi_sector
- * order without overlap.  There may be several bio's per stripe+device, and
- * a bio could span several devices.
- * When walking this list for a particular stripe+device, we must never proceed
- * beyond a bio that extends past this device, as the next bio might no longer
- * be valid.
- * This macro is used to determine the 'next' bio in the list, given the sector
- * of the current stripe+device
- */
-#define r5_next_bio(bio, sect) ( ( (bio)->bi_sector + ((bio)->bi_size>>9) < sect + STRIPE_SECTORS) ? (bio)->bi_next : NULL)
-/*
- * The following can be used to debug the driver
- */
-#define RAID6_DEBUG    0       /* Extremely verbose printk */
-#define RAID6_PARANOIA 1       /* Check spinlocks */
-#define RAID6_DUMPSTATE 0      /* Include stripe cache state in /proc/mdstat */
-#if RAID6_PARANOIA && defined(CONFIG_SMP)
-# define CHECK_DEVLOCK() assert_spin_locked(&conf->device_lock)
-#else
-# define CHECK_DEVLOCK()
-#endif
-
-#define PRINTK(x...) ((void)(RAID6_DEBUG && printk(KERN_DEBUG x)))
-#if RAID6_DEBUG
-#undef inline
-#undef __inline__
-#define inline
-#define __inline__
-#endif
-
-#if !RAID6_USE_EMPTY_ZERO_PAGE
-/* In .bss so it's zeroed */
-const char raid6_empty_zero_page[PAGE_SIZE] __attribute__((aligned(256)));
-#endif
-
-static inline int raid6_next_disk(int disk, int raid_disks)
-{
-       disk++;
-       return (disk < raid_disks) ? disk : 0;
-}
-
-static void print_raid6_conf (raid6_conf_t *conf);
-
-static void __release_stripe(raid6_conf_t *conf, struct stripe_head *sh)
-{
-       if (atomic_dec_and_test(&sh->count)) {
-               BUG_ON(!list_empty(&sh->lru));
-               BUG_ON(atomic_read(&conf->active_stripes)==0);
-               if (test_bit(STRIPE_HANDLE, &sh->state)) {
-                       if (test_bit(STRIPE_DELAYED, &sh->state))
-                               list_add_tail(&sh->lru, &conf->delayed_list);
-                       else if (test_bit(STRIPE_BIT_DELAY, &sh->state) &&
-                                conf->seq_write == sh->bm_seq)
-                               list_add_tail(&sh->lru, &conf->bitmap_list);
-                       else {
-                               clear_bit(STRIPE_BIT_DELAY, &sh->state);
-                               list_add_tail(&sh->lru, &conf->handle_list);
-                       }
-                       md_wakeup_thread(conf->mddev->thread);
-               } else {
-                       if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
-                               atomic_dec(&conf->preread_active_stripes);
-                               if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD)
-                                       md_wakeup_thread(conf->mddev->thread);
-                       }
-                       list_add_tail(&sh->lru, &conf->inactive_list);
-                       atomic_dec(&conf->active_stripes);
-                       if (!conf->inactive_blocked ||
-                           atomic_read(&conf->active_stripes) < (conf->max_nr_stripes*3/4))
-                               wake_up(&conf->wait_for_stripe);
-               }
-       }
-}
-static void release_stripe(struct stripe_head *sh)
-{
-       raid6_conf_t *conf = sh->raid_conf;
-       unsigned long flags;
-
-       spin_lock_irqsave(&conf->device_lock, flags);
-       __release_stripe(conf, sh);
-       spin_unlock_irqrestore(&conf->device_lock, flags);
-}
-
-static inline void remove_hash(struct stripe_head *sh)
-{
-       PRINTK("remove_hash(), stripe %llu\n", (unsigned long long)sh->sector);
-
-       hlist_del_init(&sh->hash);
-}
-
-static inline void insert_hash(raid6_conf_t *conf, struct stripe_head *sh)
-{
-       struct hlist_head *hp = stripe_hash(conf, sh->sector);
-
-       PRINTK("insert_hash(), stripe %llu\n", (unsigned long long)sh->sector);
-
-       CHECK_DEVLOCK();
-       hlist_add_head(&sh->hash, hp);
-}
-
-
-/* find an idle stripe, make sure it is unhashed, and return it. */
-static struct stripe_head *get_free_stripe(raid6_conf_t *conf)
-{
-       struct stripe_head *sh = NULL;
-       struct list_head *first;
-
-       CHECK_DEVLOCK();
-       if (list_empty(&conf->inactive_list))
-               goto out;
-       first = conf->inactive_list.next;
-       sh = list_entry(first, struct stripe_head, lru);
-       list_del_init(first);
-       remove_hash(sh);
-       atomic_inc(&conf->active_stripes);
-out:
-       return sh;
-}
-
-static void shrink_buffers(struct stripe_head *sh, int num)
-{
-       struct page *p;
-       int i;
-
-       for (i=0; i<num ; i++) {
-               p = sh->dev[i].page;
-               if (!p)
-                       continue;
-               sh->dev[i].page = NULL;
-               put_page(p);
-       }
-}
-
-static int grow_buffers(struct stripe_head *sh, int num)
-{
-       int i;
-
-       for (i=0; i<num; i++) {
-               struct page *page;
-
-               if (!(page = alloc_page(GFP_KERNEL))) {
-                       return 1;
-               }
-               sh->dev[i].page = page;
-       }
-       return 0;
-}
-
-static void raid6_build_block (struct stripe_head *sh, int i);
-
-static void init_stripe(struct stripe_head *sh, sector_t sector, int pd_idx)
-{
-       raid6_conf_t *conf = sh->raid_conf;
-       int disks = conf->raid_disks, i;
-
-       BUG_ON(atomic_read(&sh->count) != 0);
-       BUG_ON(test_bit(STRIPE_HANDLE, &sh->state));
-
-       CHECK_DEVLOCK();
-       PRINTK("init_stripe called, stripe %llu\n",
-               (unsigned long long)sh->sector);
-
-       remove_hash(sh);
-
-       sh->sector = sector;
-       sh->pd_idx = pd_idx;
-       sh->state = 0;
-
-       for (i=disks; i--; ) {
-               struct r5dev *dev = &sh->dev[i];
-
-               if (dev->toread || dev->towrite || dev->written ||
-                   test_bit(R5_LOCKED, &dev->flags)) {
-                       PRINTK("sector=%llx i=%d %p %p %p %d\n",
-                              (unsigned long long)sh->sector, i, dev->toread,
-                              dev->towrite, dev->written,
-                              test_bit(R5_LOCKED, &dev->flags));
-                       BUG();
-               }
-               dev->flags = 0;
-               raid6_build_block(sh, i);
-       }
-       insert_hash(conf, sh);
-}
-
-static struct stripe_head *__find_stripe(raid6_conf_t *conf, sector_t sector)
-{
-       struct stripe_head *sh;
-       struct hlist_node *hn;
-
-       CHECK_DEVLOCK();
-       PRINTK("__find_stripe, sector %llu\n", (unsigned long long)sector);
-       hlist_for_each_entry (sh, hn,  stripe_hash(conf, sector), hash)
-               if (sh->sector == sector)
-                       return sh;
-       PRINTK("__stripe %llu not in cache\n", (unsigned long long)sector);
-       return NULL;
-}
-
-static void unplug_slaves(mddev_t *mddev);
-
-static struct stripe_head *get_active_stripe(raid6_conf_t *conf, sector_t sector,
-                                            int pd_idx, int noblock)
-{
-       struct stripe_head *sh;
-
-       PRINTK("get_stripe, sector %llu\n", (unsigned long long)sector);
-
-       spin_lock_irq(&conf->device_lock);
-
-       do {
-               wait_event_lock_irq(conf->wait_for_stripe,
-                                   conf->quiesce == 0,
-                                   conf->device_lock, /* nothing */);
-               sh = __find_stripe(conf, sector);
-               if (!sh) {
-                       if (!conf->inactive_blocked)
-                               sh = get_free_stripe(conf);
-                       if (noblock && sh == NULL)
-                               break;
-                       if (!sh) {
-                               conf->inactive_blocked = 1;
-                               wait_event_lock_irq(conf->wait_for_stripe,
-                                                   !list_empty(&conf->inactive_list) &&
-                                                   (atomic_read(&conf->active_stripes)
-                                                    < (conf->max_nr_stripes *3/4)
-                                                    || !conf->inactive_blocked),
-                                                   conf->device_lock,
-                                                   unplug_slaves(conf->mddev);
-                                       );
-                               conf->inactive_blocked = 0;
-                       } else
-                               init_stripe(sh, sector, pd_idx);
-               } else {
-                       if (atomic_read(&sh->count)) {
-                               BUG_ON(!list_empty(&sh->lru));
-                       } else {
-                               if (!test_bit(STRIPE_HANDLE, &sh->state))
-                                       atomic_inc(&conf->active_stripes);
-                               BUG_ON(list_empty(&sh->lru));
-                               list_del_init(&sh->lru);
-                       }
-               }
-       } while (sh == NULL);
-
-       if (sh)
-               atomic_inc(&sh->count);
-
-       spin_unlock_irq(&conf->device_lock);
-       return sh;
-}
-
-static int grow_one_stripe(raid6_conf_t *conf)
-{
-       struct stripe_head *sh;
-       sh = kmem_cache_alloc(conf->slab_cache, GFP_KERNEL);
-       if (!sh)
-               return 0;
-       memset(sh, 0, sizeof(*sh) + (conf->raid_disks-1)*sizeof(struct r5dev));
-       sh->raid_conf = conf;
-       spin_lock_init(&sh->lock);
-
-       if (grow_buffers(sh, conf->raid_disks)) {
-               shrink_buffers(sh, conf->raid_disks);
-               kmem_cache_free(conf->slab_cache, sh);
-               return 0;
-       }
-       /* we just created an active stripe so... */
-       atomic_set(&sh->count, 1);
-       atomic_inc(&conf->active_stripes);
-       INIT_LIST_HEAD(&sh->lru);
-       release_stripe(sh);
-       return 1;
-}
-
-static int grow_stripes(raid6_conf_t *conf, int num)
-{
-       kmem_cache_t *sc;
-       int devs = conf->raid_disks;
-
-       sprintf(conf->cache_name[0], "raid6/%s", mdname(conf->mddev));
-
-       sc = kmem_cache_create(conf->cache_name[0],
-                              sizeof(struct stripe_head)+(devs-1)*sizeof(struct r5dev),
-                              0, 0, NULL, NULL);
-       if (!sc)
-               return 1;
-       conf->slab_cache = sc;
-       while (num--)
-               if (!grow_one_stripe(conf))
-                       return 1;
-       return 0;
-}
-
-static int drop_one_stripe(raid6_conf_t *conf)
-{
-       struct stripe_head *sh;
-       spin_lock_irq(&conf->device_lock);
-       sh = get_free_stripe(conf);
-       spin_unlock_irq(&conf->device_lock);
-       if (!sh)
-               return 0;
-       BUG_ON(atomic_read(&sh->count));
-       shrink_buffers(sh, conf->raid_disks);
-       kmem_cache_free(conf->slab_cache, sh);
-       atomic_dec(&conf->active_stripes);
-       return 1;
-}
-
-static void shrink_stripes(raid6_conf_t *conf)
-{
-       while (drop_one_stripe(conf))
-               ;
-
-       if (conf->slab_cache)
-               kmem_cache_destroy(conf->slab_cache);
-       conf->slab_cache = NULL;
-}
-
-static int raid6_end_read_request(struct bio * bi, unsigned int bytes_done,
-                                 int error)
-{
-       struct stripe_head *sh = bi->bi_private;
-       raid6_conf_t *conf = sh->raid_conf;
-       int disks = conf->raid_disks, i;
-       int uptodate = test_bit(BIO_UPTODATE, &bi->bi_flags);
-
-       if (bi->bi_size)
-               return 1;
-
-       for (i=0 ; i<disks; i++)
-               if (bi == &sh->dev[i].req)
-                       break;
-
-       PRINTK("end_read_request %llu/%d, count: %d, uptodate %d.\n",
-               (unsigned long long)sh->sector, i, atomic_read(&sh->count),
-               uptodate);
-       if (i == disks) {
-               BUG();
-               return 0;
-       }
-
-       if (uptodate) {
-#if 0
-               struct bio *bio;
-               unsigned long flags;
-               spin_lock_irqsave(&conf->device_lock, flags);
-               /* we can return a buffer if we bypassed the cache or
-                * if the top buffer is not in highmem.  If there are
-                * multiple buffers, leave the extra work to
-                * handle_stripe
-                */
-               buffer = sh->bh_read[i];
-               if (buffer &&
-                   (!PageHighMem(buffer->b_page)
-                    || buffer->b_page == bh->b_page )
-                       ) {
-                       sh->bh_read[i] = buffer->b_reqnext;
-                       buffer->b_reqnext = NULL;
-               } else
-                       buffer = NULL;
-               spin_unlock_irqrestore(&conf->device_lock, flags);
-               if (sh->bh_page[i]==bh->b_page)
-                       set_buffer_uptodate(bh);
-               if (buffer) {
-                       if (buffer->b_page != bh->b_page)
-                               memcpy(buffer->b_data, bh->b_data, bh->b_size);
-                       buffer->b_end_io(buffer, 1);
-               }
-#else
-               set_bit(R5_UPTODATE, &sh->dev[i].flags);
-#endif
-               if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
-                       printk(KERN_INFO "raid6: read error corrected!!\n");
-                       clear_bit(R5_ReadError, &sh->dev[i].flags);
-                       clear_bit(R5_ReWrite, &sh->dev[i].flags);
-               }
-               if (atomic_read(&conf->disks[i].rdev->read_errors))
-                       atomic_set(&conf->disks[i].rdev->read_errors, 0);
-       } else {
-               int retry = 0;
-               clear_bit(R5_UPTODATE, &sh->dev[i].flags);
-               atomic_inc(&conf->disks[i].rdev->read_errors);
-               if (conf->mddev->degraded)
-                       printk(KERN_WARNING "raid6: read error not correctable.\n");
-               else if (test_bit(R5_ReWrite, &sh->dev[i].flags))
-                       /* Oh, no!!! */
-                       printk(KERN_WARNING "raid6: read error NOT corrected!!\n");
-               else if (atomic_read(&conf->disks[i].rdev->read_errors)
-                        > conf->max_nr_stripes)
-                       printk(KERN_WARNING
-                              "raid6: Too many read errors, failing device.\n");
-               else
-                       retry = 1;
-               if (retry)
-                       set_bit(R5_ReadError, &sh->dev[i].flags);
-               else {
-                       clear_bit(R5_ReadError, &sh->dev[i].flags);
-                       clear_bit(R5_ReWrite, &sh->dev[i].flags);
-                       md_error(conf->mddev, conf->disks[i].rdev);
-               }
-       }
-       rdev_dec_pending(conf->disks[i].rdev, conf->mddev);
-#if 0
-       /* must restore b_page before unlocking buffer... */
-       if (sh->bh_page[i] != bh->b_page) {
-               bh->b_page = sh->bh_page[i];
-               bh->b_data = page_address(bh->b_page);
-               clear_buffer_uptodate(bh);
-       }
-#endif
-       clear_bit(R5_LOCKED, &sh->dev[i].flags);
-       set_bit(STRIPE_HANDLE, &sh->state);
-       release_stripe(sh);
-       return 0;
-}
-
-static int raid6_end_write_request (struct bio *bi, unsigned int bytes_done,
-                                   int error)
-{
-       struct stripe_head *sh = bi->bi_private;
-       raid6_conf_t *conf = sh->raid_conf;
-       int disks = conf->raid_disks, i;
-       unsigned long flags;
-       int uptodate = test_bit(BIO_UPTODATE, &bi->bi_flags);
-
-       if (bi->bi_size)
-               return 1;
-
-       for (i=0 ; i<disks; i++)
-               if (bi == &sh->dev[i].req)
-                       break;
-
-       PRINTK("end_write_request %llu/%d, count %d, uptodate: %d.\n",
-               (unsigned long long)sh->sector, i, atomic_read(&sh->count),
-               uptodate);
-       if (i == disks) {
-               BUG();
-               return 0;
-       }
-
-       spin_lock_irqsave(&conf->device_lock, flags);
-       if (!uptodate)
-               md_error(conf->mddev, conf->disks[i].rdev);
-
-       rdev_dec_pending(conf->disks[i].rdev, conf->mddev);
-
-       clear_bit(R5_LOCKED, &sh->dev[i].flags);
-       set_bit(STRIPE_HANDLE, &sh->state);
-       __release_stripe(conf, sh);
-       spin_unlock_irqrestore(&conf->device_lock, flags);
-       return 0;
-}
-
-
-static sector_t compute_blocknr(struct stripe_head *sh, int i);
-
-static void raid6_build_block (struct stripe_head *sh, int i)
-{
-       struct r5dev *dev = &sh->dev[i];
-       int pd_idx = sh->pd_idx;
-       int qd_idx = raid6_next_disk(pd_idx, sh->raid_conf->raid_disks);
-
-       bio_init(&dev->req);
-       dev->req.bi_io_vec = &dev->vec;
-       dev->req.bi_vcnt++;
-       dev->req.bi_max_vecs++;
-       dev->vec.bv_page = dev->page;
-       dev->vec.bv_len = STRIPE_SIZE;
-       dev->vec.bv_offset = 0;
-
-       dev->req.bi_sector = sh->sector;
-       dev->req.bi_private = sh;
-
-       dev->flags = 0;
-       if (i != pd_idx && i != qd_idx)
-               dev->sector = compute_blocknr(sh, i);
-}
-
-static void error(mddev_t *mddev, mdk_rdev_t *rdev)
-{
-       char b[BDEVNAME_SIZE];
-       raid6_conf_t *conf = (raid6_conf_t *) mddev->private;
-       PRINTK("raid6: error called\n");
-
-       if (!test_bit(Faulty, &rdev->flags)) {
-               mddev->sb_dirty = 1;
-               if (test_bit(In_sync, &rdev->flags)) {
-                       conf->working_disks--;
-                       mddev->degraded++;
-                       conf->failed_disks++;
-                       clear_bit(In_sync, &rdev->flags);
-                       /*
-                        * if recovery was running, make sure it aborts.
-                        */
-                       set_bit(MD_RECOVERY_ERR, &mddev->recovery);
-               }
-               set_bit(Faulty, &rdev->flags);
-               printk (KERN_ALERT
-                       "raid6: Disk failure on %s, disabling device."
-                       " Operation continuing on %d devices\n",
-                       bdevname(rdev->bdev,b), conf->working_disks);
-       }
-}
-
-/*
- * Input: a 'big' sector number,
- * Output: index of the data and parity disk, and the sector # in them.
- */
-static sector_t raid6_compute_sector(sector_t r_sector, unsigned int raid_disks,
-                       unsigned int data_disks, unsigned int * dd_idx,
-                       unsigned int * pd_idx, raid6_conf_t *conf)
-{
-       long stripe;
-       unsigned long chunk_number;
-       unsigned int chunk_offset;
-       sector_t new_sector;
-       int sectors_per_chunk = conf->chunk_size >> 9;
-
-       /* First compute the information on this sector */
-
-       /*
-        * Compute the chunk number and the sector offset inside the chunk
-        */
-       chunk_offset = sector_div(r_sector, sectors_per_chunk);
-       chunk_number = r_sector;
-       if ( r_sector != chunk_number ) {
-               printk(KERN_CRIT "raid6: ERROR: r_sector = %llu, chunk_number = %lu\n",
-                      (unsigned long long)r_sector, (unsigned long)chunk_number);
-               BUG();
-       }
-
-       /*
-        * Compute the stripe number
-        */
-       stripe = chunk_number / data_disks;
-
-       /*
-        * Compute the data disk and parity disk indexes inside the stripe
-        */
-       *dd_idx = chunk_number % data_disks;
-
-       /*
-        * Select the parity disk based on the user selected algorithm.
-        */
-
-       /**** FIX THIS ****/
-       switch (conf->algorithm) {
-       case ALGORITHM_LEFT_ASYMMETRIC:
-               *pd_idx = raid_disks - 1 - (stripe % raid_disks);
-               if (*pd_idx == raid_disks-1)
-                       (*dd_idx)++;    /* Q D D D P */
-               else if (*dd_idx >= *pd_idx)
-                       (*dd_idx) += 2; /* D D P Q D */
-               break;
-       case ALGORITHM_RIGHT_ASYMMETRIC:
-               *pd_idx = stripe % raid_disks;
-               if (*pd_idx == raid_disks-1)
-                       (*dd_idx)++;    /* Q D D D P */
-               else if (*dd_idx >= *pd_idx)
-                       (*dd_idx) += 2; /* D D P Q D */
-               break;
-       case ALGORITHM_LEFT_SYMMETRIC:
-               *pd_idx = raid_disks - 1 - (stripe % raid_disks);
-               *dd_idx = (*pd_idx + 2 + *dd_idx) % raid_disks;
-               break;
-       case ALGORITHM_RIGHT_SYMMETRIC:
-               *pd_idx = stripe % raid_disks;
-               *dd_idx = (*pd_idx + 2 + *dd_idx) % raid_disks;
-               break;
-       default:
-               printk (KERN_CRIT "raid6: unsupported algorithm %d\n",
-                       conf->algorithm);
-       }
-
-       PRINTK("raid6: chunk_number = %lu, pd_idx = %u, dd_idx = %u\n",
-              chunk_number, *pd_idx, *dd_idx);
-
-       /*
-        * Finally, compute the new sector number
-        */
-       new_sector = (sector_t) stripe * sectors_per_chunk + chunk_offset;
-       return new_sector;
-}
-
-
-static sector_t compute_blocknr(struct stripe_head *sh, int i)
-{
-       raid6_conf_t *conf = sh->raid_conf;
-       int raid_disks = conf->raid_disks, data_disks = raid_disks - 2;
-       sector_t new_sector = sh->sector, check;
-       int sectors_per_chunk = conf->chunk_size >> 9;
-       sector_t stripe;
-       int chunk_offset;
-       int chunk_number, dummy1, dummy2, dd_idx = i;
-       sector_t r_sector;
-       int i0 = i;
-
-       chunk_offset = sector_div(new_sector, sectors_per_chunk);
-       stripe = new_sector;
-       if ( new_sector != stripe ) {
-               printk(KERN_CRIT "raid6: ERROR: new_sector = %llu, stripe = %lu\n",
-                      (unsigned long long)new_sector, (unsigned long)stripe);
-               BUG();
-       }
-
-       switch (conf->algorithm) {
-               case ALGORITHM_LEFT_ASYMMETRIC:
-               case ALGORITHM_RIGHT_ASYMMETRIC:
-                       if (sh->pd_idx == raid_disks-1)
-                               i--;    /* Q D D D P */
-                       else if (i > sh->pd_idx)
-                               i -= 2; /* D D P Q D */
-                       break;
-               case ALGORITHM_LEFT_SYMMETRIC:
-               case ALGORITHM_RIGHT_SYMMETRIC:
-                       if (sh->pd_idx == raid_disks-1)
-                               i--; /* Q D D D P */
-                       else {
-                               /* D D P Q D */
-                               if (i < sh->pd_idx)
-                                       i += raid_disks;
-                               i -= (sh->pd_idx + 2);
-                       }
-                       break;
-               default:
-                       printk (KERN_CRIT "raid6: unsupported algorithm %d\n",
-                               conf->algorithm);
-       }
-
-       PRINTK("raid6: compute_blocknr: pd_idx = %u, i0 = %u, i = %u\n", sh->pd_idx, i0, i);
-
-       chunk_number = stripe * data_disks + i;
-       r_sector = (sector_t)chunk_number * sectors_per_chunk + chunk_offset;
-
-       check = raid6_compute_sector (r_sector, raid_disks, data_disks, &dummy1, &dummy2, conf);
-       if (check != sh->sector || dummy1 != dd_idx || dummy2 != sh->pd_idx) {
-               printk(KERN_CRIT "raid6: compute_blocknr: map not correct\n");
-               return 0;
-       }
-       return r_sector;
-}
-
-
-
-/*
- * Copy data between a page in the stripe cache, and one or more bion
- * The page could align with the middle of the bio, or there could be
- * several bion, each with several bio_vecs, which cover part of the page
- * Multiple bion are linked together on bi_next.  There may be extras
- * at the end of this list.  We ignore them.
- */
-static void copy_data(int frombio, struct bio *bio,
-                    struct page *page,
-                    sector_t sector)
-{
-       char *pa = page_address(page);
-       struct bio_vec *bvl;
-       int i;
-       int page_offset;
-
-       if (bio->bi_sector >= sector)
-               page_offset = (signed)(bio->bi_sector - sector) * 512;
-       else
-               page_offset = (signed)(sector - bio->bi_sector) * -512;
-       bio_for_each_segment(bvl, bio, i) {
-               int len = bio_iovec_idx(bio,i)->bv_len;
-               int clen;
-               int b_offset = 0;
-
-               if (page_offset < 0) {
-                       b_offset = -page_offset;
-                       page_offset += b_offset;
-                       len -= b_offset;
-               }
-
-               if (len > 0 && page_offset + len > STRIPE_SIZE)
-                       clen = STRIPE_SIZE - page_offset;
-               else clen = len;
-
-               if (clen > 0) {
-                       char *ba = __bio_kmap_atomic(bio, i, KM_USER0);
-                       if (frombio)
-                               memcpy(pa+page_offset, ba+b_offset, clen);
-                       else
-                               memcpy(ba+b_offset, pa+page_offset, clen);
-                       __bio_kunmap_atomic(ba, KM_USER0);
-               }
-               if (clen < len) /* hit end of page */
-                       break;
-               page_offset +=  len;
-       }
-}
-
-#define check_xor()    do {                                            \
-                          if (count == MAX_XOR_BLOCKS) {               \
-                               xor_block(count, STRIPE_SIZE, ptr);     \
-                               count = 1;                              \
-                          }                                            \
-                       } while(0)
-
-/* Compute P and Q syndromes */
-static void compute_parity(struct stripe_head *sh, int method)
-{
-       raid6_conf_t *conf = sh->raid_conf;
-       int i, pd_idx = sh->pd_idx, qd_idx, d0_idx, disks = conf->raid_disks, count;
-       struct bio *chosen;
-       /**** FIX THIS: This could be very bad if disks is close to 256 ****/
-       void *ptrs[disks];
-
-       qd_idx = raid6_next_disk(pd_idx, disks);
-       d0_idx = raid6_next_disk(qd_idx, disks);
-
-       PRINTK("compute_parity, stripe %llu, method %d\n",
-               (unsigned long long)sh->sector, method);
-
-       switch(method) {
-       case READ_MODIFY_WRITE:
-               BUG();          /* READ_MODIFY_WRITE N/A for RAID-6 */
-       case RECONSTRUCT_WRITE:
-               for (i= disks; i-- ;)
-                       if ( i != pd_idx && i != qd_idx && sh->dev[i].towrite ) {
-                               chosen = sh->dev[i].towrite;
-                               sh->dev[i].towrite = NULL;
-
-                               if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
-                                       wake_up(&conf->wait_for_overlap);
-
-                               BUG_ON(sh->dev[i].written);
-                               sh->dev[i].written = chosen;
-                       }
-               break;
-       case CHECK_PARITY:
-               BUG();          /* Not implemented yet */
-       }
-
-       for (i = disks; i--;)
-               if (sh->dev[i].written) {
-                       sector_t sector = sh->dev[i].sector;
-                       struct bio *wbi = sh->dev[i].written;
-                       while (wbi && wbi->bi_sector < sector + STRIPE_SECTORS) {
-                               copy_data(1, wbi, sh->dev[i].page, sector);
-                               wbi = r5_next_bio(wbi, sector);
-                       }
-
-                       set_bit(R5_LOCKED, &sh->dev[i].flags);
-                       set_bit(R5_UPTODATE, &sh->dev[i].flags);
-               }
-
-//     switch(method) {
-//     case RECONSTRUCT_WRITE:
-//     case CHECK_PARITY:
-//     case UPDATE_PARITY:
-               /* Note that unlike RAID-5, the ordering of the disks matters greatly. */
-               /* FIX: Is this ordering of drives even remotely optimal? */
-               count = 0;
-               i = d0_idx;
-               do {
-                       ptrs[count++] = page_address(sh->dev[i].page);
-                       if (count <= disks-2 && !test_bit(R5_UPTODATE, &sh->dev[i].flags))
-                               printk("block %d/%d not uptodate on parity calc\n", i,count);
-                       i = raid6_next_disk(i, disks);
-               } while ( i != d0_idx );
-//             break;
-//     }
-
-       raid6_call.gen_syndrome(disks, STRIPE_SIZE, ptrs);
-
-       switch(method) {
-       case RECONSTRUCT_WRITE:
-               set_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
-               set_bit(R5_UPTODATE, &sh->dev[qd_idx].flags);
-               set_bit(R5_LOCKED,   &sh->dev[pd_idx].flags);
-               set_bit(R5_LOCKED,   &sh->dev[qd_idx].flags);
-               break;
-       case UPDATE_PARITY:
-               set_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
-               set_bit(R5_UPTODATE, &sh->dev[qd_idx].flags);
-               break;
-       }
-}
-
-/* Compute one missing block */
-static void compute_block_1(struct stripe_head *sh, int dd_idx, int nozero)
-{
-       raid6_conf_t *conf = sh->raid_conf;
-       int i, count, disks = conf->raid_disks;
-       void *ptr[MAX_XOR_BLOCKS], *p;
-       int pd_idx = sh->pd_idx;
-       int qd_idx = raid6_next_disk(pd_idx, disks);
-
-       PRINTK("compute_block_1, stripe %llu, idx %d\n",
-               (unsigned long long)sh->sector, dd_idx);
-
-       if ( dd_idx == qd_idx ) {
-               /* We're actually computing the Q drive */
-               compute_parity(sh, UPDATE_PARITY);
-       } else {
-               ptr[0] = page_address(sh->dev[dd_idx].page);
-               if (!nozero) memset(ptr[0], 0, STRIPE_SIZE);
-               count = 1;
-               for (i = disks ; i--; ) {
-                       if (i == dd_idx || i == qd_idx)
-                               continue;
-                       p = page_address(sh->dev[i].page);
-                       if (test_bit(R5_UPTODATE, &sh->dev[i].flags))
-                               ptr[count++] = p;
-                       else
-                               printk("compute_block() %d, stripe %llu, %d"
-                                      " not present\n", dd_idx,
-                                      (unsigned long long)sh->sector, i);
-
-                       check_xor();
-               }
-               if (count != 1)
-                       xor_block(count, STRIPE_SIZE, ptr);
-               if (!nozero) set_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
-               else clear_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
-       }
-}
-
-/* Compute two missing blocks */
-static void compute_block_2(struct stripe_head *sh, int dd_idx1, int dd_idx2)
-{
-       raid6_conf_t *conf = sh->raid_conf;
-       int i, count, disks = conf->raid_disks;
-       int pd_idx = sh->pd_idx;
-       int qd_idx = raid6_next_disk(pd_idx, disks);
-       int d0_idx = raid6_next_disk(qd_idx, disks);
-       int faila, failb;
-
-       /* faila and failb are disk numbers relative to d0_idx */
-       /* pd_idx become disks-2 and qd_idx become disks-1 */
-       faila = (dd_idx1 < d0_idx) ? dd_idx1+(disks-d0_idx) : dd_idx1-d0_idx;
-       failb = (dd_idx2 < d0_idx) ? dd_idx2+(disks-d0_idx) : dd_idx2-d0_idx;
-
-       BUG_ON(faila == failb);
-       if ( failb < faila ) { int tmp = faila; faila = failb; failb = tmp; }
-
-       PRINTK("compute_block_2, stripe %llu, idx %d,%d (%d,%d)\n",
-              (unsigned long long)sh->sector, dd_idx1, dd_idx2, faila, failb);
-
-       if ( failb == disks-1 ) {
-               /* Q disk is one of the missing disks */
-               if ( faila == disks-2 ) {
-                       /* Missing P+Q, just recompute */
-                       compute_parity(sh, UPDATE_PARITY);
-                       return;
-               } else {
-                       /* We're missing D+Q; recompute D from P */
-                       compute_block_1(sh, (dd_idx1 == qd_idx) ? dd_idx2 : dd_idx1, 0);
-                       compute_parity(sh, UPDATE_PARITY); /* Is this necessary? */
-                       return;
-               }
-       }
-
-       /* We're missing D+P or D+D; build pointer table */
-       {
-               /**** FIX THIS: This could be very bad if disks is close to 256 ****/
-               void *ptrs[disks];
-
-               count = 0;
-               i = d0_idx;
-               do {
-                       ptrs[count++] = page_address(sh->dev[i].page);
-                       i = raid6_next_disk(i, disks);
-                       if (i != dd_idx1 && i != dd_idx2 &&
-                           !test_bit(R5_UPTODATE, &sh->dev[i].flags))
-                               printk("compute_2 with missing block %d/%d\n", count, i);
-               } while ( i != d0_idx );
-
-               if ( failb == disks-2 ) {
-                       /* We're missing D+P. */
-                       raid6_datap_recov(disks, STRIPE_SIZE, faila, ptrs);
-               } else {
-                       /* We're missing D+D. */
-                       raid6_2data_recov(disks, STRIPE_SIZE, faila, failb, ptrs);
-               }
-
-               /* Both the above update both missing blocks */
-               set_bit(R5_UPTODATE, &sh->dev[dd_idx1].flags);
-               set_bit(R5_UPTODATE, &sh->dev[dd_idx2].flags);
-       }
-}
-
-
-/*
- * Each stripe/dev can have one or more bion attached.
- * toread/towrite point to the first in a chain.
- * The bi_next chain must be in order.
- */
-static int add_stripe_bio(struct stripe_head *sh, struct bio *bi, int dd_idx, int forwrite)
-{
-       struct bio **bip;
-       raid6_conf_t *conf = sh->raid_conf;
-       int firstwrite=0;
-
-       PRINTK("adding bh b#%llu to stripe s#%llu\n",
-               (unsigned long long)bi->bi_sector,
-               (unsigned long long)sh->sector);
-
-
-       spin_lock(&sh->lock);
-       spin_lock_irq(&conf->device_lock);
-       if (forwrite) {
-               bip = &sh->dev[dd_idx].towrite;
-               if (*bip == NULL && sh->dev[dd_idx].written == NULL)
-                       firstwrite = 1;
-       } else
-               bip = &sh->dev[dd_idx].toread;
-       while (*bip && (*bip)->bi_sector < bi->bi_sector) {
-               if ((*bip)->bi_sector + ((*bip)->bi_size >> 9) > bi->bi_sector)
-                       goto overlap;
-               bip = &(*bip)->bi_next;
-       }
-       if (*bip && (*bip)->bi_sector < bi->bi_sector + ((bi->bi_size)>>9))
-               goto overlap;
-
-       BUG_ON(*bip && bi->bi_next && (*bip) != bi->bi_next);
-       if (*bip)
-               bi->bi_next = *bip;
-       *bip = bi;
-       bi->bi_phys_segments ++;
-       spin_unlock_irq(&conf->device_lock);
-       spin_unlock(&sh->lock);
-
-       PRINTK("added bi b#%llu to stripe s#%llu, disk %d.\n",
-               (unsigned long long)bi->bi_sector,
-               (unsigned long long)sh->sector, dd_idx);
-
-       if (conf->mddev->bitmap && firstwrite) {
-               sh->bm_seq = conf->seq_write;
-               bitmap_startwrite(conf->mddev->bitmap, sh->sector,
-                                 STRIPE_SECTORS, 0);
-               set_bit(STRIPE_BIT_DELAY, &sh->state);
-       }
-
-       if (forwrite) {
-               /* check if page is covered */
-               sector_t sector = sh->dev[dd_idx].sector;
-               for (bi=sh->dev[dd_idx].towrite;
-                    sector < sh->dev[dd_idx].sector + STRIPE_SECTORS &&
-                            bi && bi->bi_sector <= sector;
-                    bi = r5_next_bio(bi, sh->dev[dd_idx].sector)) {
-                       if (bi->bi_sector + (bi->bi_size>>9) >= sector)
-                               sector = bi->bi_sector + (bi->bi_size>>9);
-               }
-               if (sector >= sh->dev[dd_idx].sector + STRIPE_SECTORS)
-                       set_bit(R5_OVERWRITE, &sh->dev[dd_idx].flags);
-       }
-       return 1;
-
- overlap:
-       set_bit(R5_Overlap, &sh->dev[dd_idx].flags);
-       spin_unlock_irq(&conf->device_lock);
-       spin_unlock(&sh->lock);
-       return 0;
-}
-
-
-static int page_is_zero(struct page *p)
-{
-       char *a = page_address(p);
-       return ((*(u32*)a) == 0 &&
-               memcmp(a, a+4, STRIPE_SIZE-4)==0);
-}
-/*
- * handle_stripe - do things to a stripe.
- *
- * We lock the stripe and then examine the state of various bits
- * to see what needs to be done.
- * Possible results:
- *    return some read request which now have data
- *    return some write requests which are safely on disc
- *    schedule a read on some buffers
- *    schedule a write of some buffers
- *    return confirmation of parity correctness
- *
- * Parity calculations are done inside the stripe lock
- * buffers are taken off read_list or write_list, and bh_cache buffers
- * get BH_Lock set before the stripe lock is released.
- *
- */
-
-static void handle_stripe(struct stripe_head *sh, struct page *tmp_page)
-{
-       raid6_conf_t *conf = sh->raid_conf;
-       int disks = conf->raid_disks;
-       struct bio *return_bi= NULL;
-       struct bio *bi;
-       int i;
-       int syncing;
-       int locked=0, uptodate=0, to_read=0, to_write=0, failed=0, written=0;
-       int non_overwrite = 0;
-       int failed_num[2] = {0, 0};
-       struct r5dev *dev, *pdev, *qdev;
-       int pd_idx = sh->pd_idx;
-       int qd_idx = raid6_next_disk(pd_idx, disks);
-       int p_failed, q_failed;
-
-       PRINTK("handling stripe %llu, state=%#lx cnt=%d, pd_idx=%d, qd_idx=%d\n",
-              (unsigned long long)sh->sector, sh->state, atomic_read(&sh->count),
-              pd_idx, qd_idx);
-
-       spin_lock(&sh->lock);
-       clear_bit(STRIPE_HANDLE, &sh->state);
-       clear_bit(STRIPE_DELAYED, &sh->state);
-
-       syncing = test_bit(STRIPE_SYNCING, &sh->state);
-       /* Now to look around and see what can be done */
-
-       rcu_read_lock();
-       for (i=disks; i--; ) {
-               mdk_rdev_t *rdev;
-               dev = &sh->dev[i];
-               clear_bit(R5_Insync, &dev->flags);
-
-               PRINTK("check %d: state 0x%lx read %p write %p written %p\n",
-                       i, dev->flags, dev->toread, dev->towrite, dev->written);
-               /* maybe we can reply to a read */
-               if (test_bit(R5_UPTODATE, &dev->flags) && dev->toread) {
-                       struct bio *rbi, *rbi2;
-                       PRINTK("Return read for disc %d\n", i);
-                       spin_lock_irq(&conf->device_lock);
-                       rbi = dev->toread;
-                       dev->toread = NULL;
-                       if (test_and_clear_bit(R5_Overlap, &dev->flags))
-                               wake_up(&conf->wait_for_overlap);
-                       spin_unlock_irq(&conf->device_lock);
-                       while (rbi && rbi->bi_sector < dev->sector + STRIPE_SECTORS) {
-                               copy_data(0, rbi, dev->page, dev->sector);
-                               rbi2 = r5_next_bio(rbi, dev->sector);
-                               spin_lock_irq(&conf->device_lock);
-                               if (--rbi->bi_phys_segments == 0) {
-                                       rbi->bi_next = return_bi;
-                                       return_bi = rbi;
-                               }
-                               spin_unlock_irq(&conf->device_lock);
-                               rbi = rbi2;
-                       }
-               }
-
-               /* now count some things */
-               if (test_bit(R5_LOCKED, &dev->flags)) locked++;
-               if (test_bit(R5_UPTODATE, &dev->flags)) uptodate++;
-
-
-               if (dev->toread) to_read++;
-               if (dev->towrite) {
-                       to_write++;
-                       if (!test_bit(R5_OVERWRITE, &dev->flags))
-                               non_overwrite++;
-               }
-               if (dev->written) written++;
-               rdev = rcu_dereference(conf->disks[i].rdev);
-               if (!rdev || !test_bit(In_sync, &rdev->flags)) {
-                       /* The ReadError flag will just be confusing now */
-                       clear_bit(R5_ReadError, &dev->flags);
-                       clear_bit(R5_ReWrite, &dev->flags);
-               }
-               if (!rdev || !test_bit(In_sync, &rdev->flags)
-                   || test_bit(R5_ReadError, &dev->flags)) {
-                       if ( failed < 2 )
-                               failed_num[failed] = i;
-                       failed++;
-               } else
-                       set_bit(R5_Insync, &dev->flags);
-       }
-       rcu_read_unlock();
-       PRINTK("locked=%d uptodate=%d to_read=%d"
-              " to_write=%d failed=%d failed_num=%d,%d\n",
-              locked, uptodate, to_read, to_write, failed,
-              failed_num[0], failed_num[1]);
-       /* check if the array has lost >2 devices and, if so, some requests might
-        * need to be failed
-        */
-       if (failed > 2 && to_read+to_write+written) {
-               for (i=disks; i--; ) {
-                       int bitmap_end = 0;
-
-                       if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
-                               mdk_rdev_t *rdev;
-                               rcu_read_lock();
-                               rdev = rcu_dereference(conf->disks[i].rdev);
-                               if (rdev && test_bit(In_sync, &rdev->flags))
-                                       /* multiple read failures in one stripe */
-                                       md_error(conf->mddev, rdev);
-                               rcu_read_unlock();
-                       }
-
-                       spin_lock_irq(&conf->device_lock);
-                       /* fail all writes first */
-                       bi = sh->dev[i].towrite;
-                       sh->dev[i].towrite = NULL;
-                       if (bi) { to_write--; bitmap_end = 1; }
-
-                       if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
-                               wake_up(&conf->wait_for_overlap);
-
-                       while (bi && bi->bi_sector < sh->dev[i].sector + STRIPE_SECTORS){
-                               struct bio *nextbi = r5_next_bio(bi, sh->dev[i].sector);
-                               clear_bit(BIO_UPTODATE, &bi->bi_flags);
-                               if (--bi->bi_phys_segments == 0) {
-                                       md_write_end(conf->mddev);
-                                       bi->bi_next = return_bi;
-                                       return_bi = bi;
-                               }
-                               bi = nextbi;
-                       }
-                       /* and fail all 'written' */
-                       bi = sh->dev[i].written;
-                       sh->dev[i].written = NULL;
-                       if (bi) bitmap_end = 1;
-                       while (bi && bi->bi_sector < sh->dev[i].sector + STRIPE_SECTORS) {
-                               struct bio *bi2 = r5_next_bio(bi, sh->dev[i].sector);
-                               clear_bit(BIO_UPTODATE, &bi->bi_flags);
-                               if (--bi->bi_phys_segments == 0) {
-                                       md_write_end(conf->mddev);
-                                       bi->bi_next = return_bi;
-                                       return_bi = bi;
-                               }
-                               bi = bi2;
-                       }
-
-                       /* fail any reads if this device is non-operational */
-                       if (!test_bit(R5_Insync, &sh->dev[i].flags) ||
-                           test_bit(R5_ReadError, &sh->dev[i].flags)) {
-                               bi = sh->dev[i].toread;
-                               sh->dev[i].toread = NULL;
-                               if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
-                                       wake_up(&conf->wait_for_overlap);
-                               if (bi) to_read--;
-                               while (bi && bi->bi_sector < sh->dev[i].sector + STRIPE_SECTORS){
-                                       struct bio *nextbi = r5_next_bio(bi, sh->dev[i].sector);
-                                       clear_bit(BIO_UPTODATE, &bi->bi_flags);
-                                       if (--bi->bi_phys_segments == 0) {
-                                               bi->bi_next = return_bi;
-                                               return_bi = bi;
-                                       }
-                                       bi = nextbi;
-                               }
-                       }
-                       spin_unlock_irq(&conf->device_lock);
-                       if (bitmap_end)
-                               bitmap_endwrite(conf->mddev->bitmap, sh->sector,
-                                               STRIPE_SECTORS, 0, 0);
-               }
-       }
-       if (failed > 2 && syncing) {
-               md_done_sync(conf->mddev, STRIPE_SECTORS,0);
-               clear_bit(STRIPE_SYNCING, &sh->state);
-               syncing = 0;
-       }
-
-       /*
-        * might be able to return some write requests if the parity blocks
-        * are safe, or on a failed drive
-        */
-       pdev = &sh->dev[pd_idx];
-       p_failed = (failed >= 1 && failed_num[0] == pd_idx)
-               || (failed >= 2 && failed_num[1] == pd_idx);
-       qdev = &sh->dev[qd_idx];
-       q_failed = (failed >= 1 && failed_num[0] == qd_idx)
-               || (failed >= 2 && failed_num[1] == qd_idx);
-
-       if ( written &&
-            ( p_failed || ((test_bit(R5_Insync, &pdev->flags)
-                            && !test_bit(R5_LOCKED, &pdev->flags)
-                            && test_bit(R5_UPTODATE, &pdev->flags))) ) &&
-            ( q_failed || ((test_bit(R5_Insync, &qdev->flags)
-                            && !test_bit(R5_LOCKED, &qdev->flags)
-                            && test_bit(R5_UPTODATE, &qdev->flags))) ) ) {
-               /* any written block on an uptodate or failed drive can be
-                * returned.  Note that if we 'wrote' to a failed drive,
-                * it will be UPTODATE, but never LOCKED, so we don't need
-                * to test 'failed' directly.
-                */
-               for (i=disks; i--; )
-                       if (sh->dev[i].written) {
-                               dev = &sh->dev[i];
-                               if (!test_bit(R5_LOCKED, &dev->flags) &&
-                                   test_bit(R5_UPTODATE, &dev->flags) ) {
-                                       /* We can return any write requests */
-                                       int bitmap_end = 0;
-                                       struct bio *wbi, *wbi2;
-                                       PRINTK("Return write for stripe %llu disc %d\n",
-                                              (unsigned long long)sh->sector, i);
-                                       spin_lock_irq(&conf->device_lock);
-                                       wbi = dev->written;
-                                       dev->written = NULL;
-                                       while (wbi && wbi->bi_sector < dev->sector + STRIPE_SECTORS) {
-                                               wbi2 = r5_next_bio(wbi, dev->sector);
-                                               if (--wbi->bi_phys_segments == 0) {
-                                                       md_write_end(conf->mddev);
-                                                       wbi->bi_next = return_bi;
-                                                       return_bi = wbi;
-                                               }
-                                               wbi = wbi2;
-                                       }
-                                       if (dev->towrite == NULL)
-                                               bitmap_end = 1;
-                                       spin_unlock_irq(&conf->device_lock);
-                                       if (bitmap_end)
-                                               bitmap_endwrite(conf->mddev->bitmap, sh->sector,
-                                                               STRIPE_SECTORS,
-                                                               !test_bit(STRIPE_DEGRADED, &sh->state), 0);
-                               }
-                       }
-       }
-
-       /* Now we might consider reading some blocks, either to check/generate
-        * parity, or to satisfy requests
-        * or to load a block that is being partially written.
-        */
-       if (to_read || non_overwrite || (to_write && failed) || (syncing && (uptodate < disks))) {
-               for (i=disks; i--;) {
-                       dev = &sh->dev[i];
-                       if (!test_bit(R5_LOCKED, &dev->flags) && !test_bit(R5_UPTODATE, &dev->flags) &&
-                           (dev->toread ||
-                            (dev->towrite && !test_bit(R5_OVERWRITE, &dev->flags)) ||
-                            syncing ||
-                            (failed >= 1 && (sh->dev[failed_num[0]].toread || to_write)) ||
-                            (failed >= 2 && (sh->dev[failed_num[1]].toread || to_write))
-                                   )
-                               ) {
-                               /* we would like to get this block, possibly
-                                * by computing it, but we might not be able to
-                                */
-                               if (uptodate == disks-1) {
-                                       PRINTK("Computing stripe %llu block %d\n",
-                                              (unsigned long long)sh->sector, i);
-                                       compute_block_1(sh, i, 0);
-                                       uptodate++;
-                               } else if ( uptodate == disks-2 && failed >= 2 ) {
-                                       /* Computing 2-failure is *very* expensive; only do it if failed >= 2 */
-                                       int other;
-                                       for (other=disks; other--;) {
-                                               if ( other == i )
-                                                       continue;
-                                               if ( !test_bit(R5_UPTODATE, &sh->dev[other].flags) )
-                                                       break;
-                                       }
-                                       BUG_ON(other < 0);
-                                       PRINTK("Computing stripe %llu blocks %d,%d\n",
-                                              (unsigned long long)sh->sector, i, other);
-                                       compute_block_2(sh, i, other);
-                                       uptodate += 2;
-                               } else if (test_bit(R5_Insync, &dev->flags)) {
-                                       set_bit(R5_LOCKED, &dev->flags);
-                                       set_bit(R5_Wantread, &dev->flags);
-#if 0
-                                       /* if I am just reading this block and we don't have
-                                          a failed drive, or any pending writes then sidestep the cache */
-                                       if (sh->bh_read[i] && !sh->bh_read[i]->b_reqnext &&
-                                           ! syncing && !failed && !to_write) {
-                                               sh->bh_cache[i]->b_page =  sh->bh_read[i]->b_page;
-                                               sh->bh_cache[i]->b_data =  sh->bh_read[i]->b_data;
-                                       }
-#endif
-                                       locked++;
-                                       PRINTK("Reading block %d (sync=%d)\n",
-                                               i, syncing);
-                               }
-                       }
-               }
-               set_bit(STRIPE_HANDLE, &sh->state);
-       }
-
-       /* now to consider writing and what else, if anything should be read */
-       if (to_write) {
-               int rcw=0, must_compute=0;
-               for (i=disks ; i--;) {
-                       dev = &sh->dev[i];
-                       /* Would I have to read this buffer for reconstruct_write */
-                       if (!test_bit(R5_OVERWRITE, &dev->flags)
-                           && i != pd_idx && i != qd_idx
-                           && (!test_bit(R5_LOCKED, &dev->flags)
-#if 0
-                               || sh->bh_page[i] != bh->b_page
-#endif
-                                   ) &&
-                           !test_bit(R5_UPTODATE, &dev->flags)) {
-                               if (test_bit(R5_Insync, &dev->flags)) rcw++;
-                               else {
-                                       PRINTK("raid6: must_compute: disk %d flags=%#lx\n", i, dev->flags);
-                                       must_compute++;
-                               }
-                       }
-               }
-               PRINTK("for sector %llu, rcw=%d, must_compute=%d\n",
-                      (unsigned long long)sh->sector, rcw, must_compute);
-               set_bit(STRIPE_HANDLE, &sh->state);
-
-               if (rcw > 0)
-                       /* want reconstruct write, but need to get some data */
-                       for (i=disks; i--;) {
-                               dev = &sh->dev[i];
-                               if (!test_bit(R5_OVERWRITE, &dev->flags)
-                                   && !(failed == 0 && (i == pd_idx || i == qd_idx))
-                                   && !test_bit(R5_LOCKED, &dev->flags) && !test_bit(R5_UPTODATE, &dev->flags) &&
-                                   test_bit(R5_Insync, &dev->flags)) {
-                                       if (test_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
-                                       {
-                                               PRINTK("Read_old stripe %llu block %d for Reconstruct\n",
-                                                      (unsigned long long)sh->sector, i);
-                                               set_bit(R5_LOCKED, &dev->flags);
-                                               set_bit(R5_Wantread, &dev->flags);
-                                               locked++;
-                                       } else {
-                                               PRINTK("Request delayed stripe %llu block %d for Reconstruct\n",
-                                                      (unsigned long long)sh->sector, i);
-                                               set_bit(STRIPE_DELAYED, &sh->state);
-                                               set_bit(STRIPE_HANDLE, &sh->state);
-                                       }
-                               }
-                       }
-               /* now if nothing is locked, and if we have enough data, we can start a write request */
-               if (locked == 0 && rcw == 0 &&
-                   !test_bit(STRIPE_BIT_DELAY, &sh->state)) {
-                       if ( must_compute > 0 ) {
-                               /* We have failed blocks and need to compute them */
-                               switch ( failed ) {
-                               case 0: BUG();
-                               case 1: compute_block_1(sh, failed_num[0], 0); break;
-                               case 2: compute_block_2(sh, failed_num[0], failed_num[1]); break;
-                               default: BUG(); /* This request should have been failed? */
-                               }
-                       }
-
-                       PRINTK("Computing parity for stripe %llu\n", (unsigned long long)sh->sector);
-                       compute_parity(sh, RECONSTRUCT_WRITE);
-                       /* now every locked buffer is ready to be written */
-                       for (i=disks; i--;)
-                               if (test_bit(R5_LOCKED, &sh->dev[i].flags)) {
-                                       PRINTK("Writing stripe %llu block %d\n",
-                                              (unsigned long long)sh->sector, i);
-                                       locked++;
-                                       set_bit(R5_Wantwrite, &sh->dev[i].flags);
-                               }
-                       /* after a RECONSTRUCT_WRITE, the stripe MUST be in-sync */
-                       set_bit(STRIPE_INSYNC, &sh->state);
-
-                       if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
-                               atomic_dec(&conf->preread_active_stripes);
-                               if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD)
-                                       md_wakeup_thread(conf->mddev->thread);
-                       }
-               }
-       }
-
-       /* maybe we need to check and possibly fix the parity for this stripe
-        * Any reads will already have been scheduled, so we just see if enough data
-        * is available
-        */
-       if (syncing && locked == 0 && !test_bit(STRIPE_INSYNC, &sh->state)) {
-               int update_p = 0, update_q = 0;
-               struct r5dev *dev;
-
-               set_bit(STRIPE_HANDLE, &sh->state);
-
-               BUG_ON(failed>2);
-               BUG_ON(uptodate < disks);
-               /* Want to check and possibly repair P and Q.
-                * However there could be one 'failed' device, in which
-                * case we can only check one of them, possibly using the
-                * other to generate missing data
-                */
-
-               /* If !tmp_page, we cannot do the calculations,
-                * but as we have set STRIPE_HANDLE, we will soon be called
-                * by stripe_handle with a tmp_page - just wait until then.
-                */
-               if (tmp_page) {
-                       if (failed == q_failed) {
-                               /* The only possible failed device holds 'Q', so it makes
-                                * sense to check P (If anything else were failed, we would
-                                * have used P to recreate it).
-                                */
-                               compute_block_1(sh, pd_idx, 1);
-                               if (!page_is_zero(sh->dev[pd_idx].page)) {
-                                       compute_block_1(sh,pd_idx,0);
-                                       update_p = 1;
-                               }
-                       }
-                       if (!q_failed && failed < 2) {
-                               /* q is not failed, and we didn't use it to generate
-                                * anything, so it makes sense to check it
-                                */
-                               memcpy(page_address(tmp_page),
-                                      page_address(sh->dev[qd_idx].page),
-                                      STRIPE_SIZE);
-                               compute_parity(sh, UPDATE_PARITY);
-                               if (memcmp(page_address(tmp_page),
-                                          page_address(sh->dev[qd_idx].page),
-                                          STRIPE_SIZE)!= 0) {
-                                       clear_bit(STRIPE_INSYNC, &sh->state);
-                                       update_q = 1;
-                               }
-                       }
-                       if (update_p || update_q) {
-                               conf->mddev->resync_mismatches += STRIPE_SECTORS;
-                               if (test_bit(MD_RECOVERY_CHECK, &conf->mddev->recovery))
-                                       /* don't try to repair!! */
-                                       update_p = update_q = 0;
-                       }
-
-                       /* now write out any block on a failed drive,
-                        * or P or Q if they need it
-                        */
-
-                       if (failed == 2) {
-                               dev = &sh->dev[failed_num[1]];
-                               locked++;
-                               set_bit(R5_LOCKED, &dev->flags);
-                               set_bit(R5_Wantwrite, &dev->flags);
-                       }
-                       if (failed >= 1) {
-                               dev = &sh->dev[failed_num[0]];
-                               locked++;
-                               set_bit(R5_LOCKED, &dev->flags);
-                               set_bit(R5_Wantwrite, &dev->flags);
-                       }
-
-                       if (update_p) {
-                               dev = &sh->dev[pd_idx];
-                               locked ++;
-                               set_bit(R5_LOCKED, &dev->flags);
-                               set_bit(R5_Wantwrite, &dev->flags);
-                       }
-                       if (update_q) {
-                               dev = &sh->dev[qd_idx];
-                               locked++;
-                               set_bit(R5_LOCKED, &dev->flags);
-                               set_bit(R5_Wantwrite, &dev->flags);
-                       }
-                       clear_bit(STRIPE_DEGRADED, &sh->state);
-
-                       set_bit(STRIPE_INSYNC, &sh->state);
-               }
-       }
-
-       if (syncing && locked == 0 && test_bit(STRIPE_INSYNC, &sh->state)) {
-               md_done_sync(conf->mddev, STRIPE_SECTORS,1);
-               clear_bit(STRIPE_SYNCING, &sh->state);
-       }
-
-       /* If the failed drives are just a ReadError, then we might need
-        * to progress the repair/check process
-        */
-       if (failed <= 2 && ! conf->mddev->ro)
-               for (i=0; i<failed;i++) {
-                       dev = &sh->dev[failed_num[i]];
-                       if (test_bit(R5_ReadError, &dev->flags)
-                           && !test_bit(R5_LOCKED, &dev->flags)
-                           && test_bit(R5_UPTODATE, &dev->flags)
-                               ) {
-                               if (!test_bit(R5_ReWrite, &dev->flags)) {
-                                       set_bit(R5_Wantwrite, &dev->flags);
-                                       set_bit(R5_ReWrite, &dev->flags);
-                                       set_bit(R5_LOCKED, &dev->flags);
-                               } else {
-                                       /* let's read it back */
-                                       set_bit(R5_Wantread, &dev->flags);
-                                       set_bit(R5_LOCKED, &dev->flags);
-                               }
-                       }
-               }
-       spin_unlock(&sh->lock);
-
-       while ((bi=return_bi)) {
-               int bytes = bi->bi_size;
-
-               return_bi = bi->bi_next;
-               bi->bi_next = NULL;
-               bi->bi_size = 0;
-               bi->bi_end_io(bi, bytes, 0);
-       }
-       for (i=disks; i-- ;) {
-               int rw;
-               struct bio *bi;
-               mdk_rdev_t *rdev;
-               if (test_and_clear_bit(R5_Wantwrite, &sh->dev[i].flags))
-                       rw = 1;
-               else if (test_and_clear_bit(R5_Wantread, &sh->dev[i].flags))
-                       rw = 0;
-               else
-                       continue;
-
-               bi = &sh->dev[i].req;
-
-               bi->bi_rw = rw;
-               if (rw)
-                       bi->bi_end_io = raid6_end_write_request;
-               else
-                       bi->bi_end_io = raid6_end_read_request;
-
-               rcu_read_lock();
-               rdev = rcu_dereference(conf->disks[i].rdev);
-               if (rdev && test_bit(Faulty, &rdev->flags))
-                       rdev = NULL;
-               if (rdev)
-                       atomic_inc(&rdev->nr_pending);
-               rcu_read_unlock();
-
-               if (rdev) {
-                       if (syncing)
-                               md_sync_acct(rdev->bdev, STRIPE_SECTORS);
-
-                       bi->bi_bdev = rdev->bdev;
-                       PRINTK("for %llu schedule op %ld on disc %d\n",
-                               (unsigned long long)sh->sector, bi->bi_rw, i);
-                       atomic_inc(&sh->count);
-                       bi->bi_sector = sh->sector + rdev->data_offset;
-                       bi->bi_flags = 1 << BIO_UPTODATE;
-                       bi->bi_vcnt = 1;
-                       bi->bi_max_vecs = 1;
-                       bi->bi_idx = 0;
-                       bi->bi_io_vec = &sh->dev[i].vec;
-                       bi->bi_io_vec[0].bv_len = STRIPE_SIZE;
-                       bi->bi_io_vec[0].bv_offset = 0;
-                       bi->bi_size = STRIPE_SIZE;
-                       bi->bi_next = NULL;
-                       if (rw == WRITE &&
-                           test_bit(R5_ReWrite, &sh->dev[i].flags))
-                               atomic_add(STRIPE_SECTORS, &rdev->corrected_errors);
-                       generic_make_request(bi);
-               } else {
-                       if (rw == 1)
-                               set_bit(STRIPE_DEGRADED, &sh->state);
-                       PRINTK("skip op %ld on disc %d for sector %llu\n",
-                               bi->bi_rw, i, (unsigned long long)sh->sector);
-                       clear_bit(R5_LOCKED, &sh->dev[i].flags);
-                       set_bit(STRIPE_HANDLE, &sh->state);
-               }
-       }
-}
-
-static void raid6_activate_delayed(raid6_conf_t *conf)
-{
-       if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD) {
-               while (!list_empty(&conf->delayed_list)) {
-                       struct list_head *l = conf->delayed_list.next;
-                       struct stripe_head *sh;
-                       sh = list_entry(l, struct stripe_head, lru);
-                       list_del_init(l);
-                       clear_bit(STRIPE_DELAYED, &sh->state);
-                       if (!test_and_set_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
-                               atomic_inc(&conf->preread_active_stripes);
-                       list_add_tail(&sh->lru, &conf->handle_list);
-               }
-       }
-}
-
-static void activate_bit_delay(raid6_conf_t *conf)
-{
-       /* device_lock is held */
-       struct list_head head;
-       list_add(&head, &conf->bitmap_list);
-       list_del_init(&conf->bitmap_list);
-       while (!list_empty(&head)) {
-               struct stripe_head *sh = list_entry(head.next, struct stripe_head, lru);
-               list_del_init(&sh->lru);
-               atomic_inc(&sh->count);
-               __release_stripe(conf, sh);
-       }
-}
-
-static void unplug_slaves(mddev_t *mddev)
-{
-       raid6_conf_t *conf = mddev_to_conf(mddev);
-       int i;
-
-       rcu_read_lock();
-       for (i=0; i<mddev->raid_disks; i++) {
-               mdk_rdev_t *rdev = rcu_dereference(conf->disks[i].rdev);
-               if (rdev && !test_bit(Faulty, &rdev->flags) && atomic_read(&rdev->nr_pending)) {
-                       request_queue_t *r_queue = bdev_get_queue(rdev->bdev);
-
-                       atomic_inc(&rdev->nr_pending);
-                       rcu_read_unlock();
-
-                       if (r_queue->unplug_fn)
-                               r_queue->unplug_fn(r_queue);
-
-                       rdev_dec_pending(rdev, mddev);
-                       rcu_read_lock();
-               }
-       }
-       rcu_read_unlock();
-}
-
-static void raid6_unplug_device(request_queue_t *q)
-{
-       mddev_t *mddev = q->queuedata;
-       raid6_conf_t *conf = mddev_to_conf(mddev);
-       unsigned long flags;
-
-       spin_lock_irqsave(&conf->device_lock, flags);
-
-       if (blk_remove_plug(q)) {
-               conf->seq_flush++;
-               raid6_activate_delayed(conf);
-       }
-       md_wakeup_thread(mddev->thread);
-
-       spin_unlock_irqrestore(&conf->device_lock, flags);
-
-       unplug_slaves(mddev);
-}
-
-static int raid6_issue_flush(request_queue_t *q, struct gendisk *disk,
-                            sector_t *error_sector)
-{
-       mddev_t *mddev = q->queuedata;
-       raid6_conf_t *conf = mddev_to_conf(mddev);
-       int i, ret = 0;
-
-       rcu_read_lock();
-       for (i=0; i<mddev->raid_disks && ret == 0; i++) {
-               mdk_rdev_t *rdev = rcu_dereference(conf->disks[i].rdev);
-               if (rdev && !test_bit(Faulty, &rdev->flags)) {
-                       struct block_device *bdev = rdev->bdev;
-                       request_queue_t *r_queue = bdev_get_queue(bdev);
-
-                       if (!r_queue->issue_flush_fn)
-                               ret = -EOPNOTSUPP;
-                       else {
-                               atomic_inc(&rdev->nr_pending);
-                               rcu_read_unlock();
-                               ret = r_queue->issue_flush_fn(r_queue, bdev->bd_disk,
-                                                             error_sector);
-                               rdev_dec_pending(rdev, mddev);
-                               rcu_read_lock();
-                       }
-               }
-       }
-       rcu_read_unlock();
-       return ret;
-}
-
-static inline void raid6_plug_device(raid6_conf_t *conf)
-{
-       spin_lock_irq(&conf->device_lock);
-       blk_plug_device(conf->mddev->queue);
-       spin_unlock_irq(&conf->device_lock);
-}
-
-static int make_request (request_queue_t *q, struct bio * bi)
-{
-       mddev_t *mddev = q->queuedata;
-       raid6_conf_t *conf = mddev_to_conf(mddev);
-       const unsigned int raid_disks = conf->raid_disks;
-       const unsigned int data_disks = raid_disks - 2;
-       unsigned int dd_idx, pd_idx;
-       sector_t new_sector;
-       sector_t logical_sector, last_sector;
-       struct stripe_head *sh;
-       const int rw = bio_data_dir(bi);
-
-       if (unlikely(bio_barrier(bi))) {
-               bio_endio(bi, bi->bi_size, -EOPNOTSUPP);
-               return 0;
-       }
-
-       md_write_start(mddev, bi);
-
-       disk_stat_inc(mddev->gendisk, ios[rw]);
-       disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bi));
-
-       logical_sector = bi->bi_sector & ~((sector_t)STRIPE_SECTORS-1);
-       last_sector = bi->bi_sector + (bi->bi_size>>9);
-
-       bi->bi_next = NULL;
-       bi->bi_phys_segments = 1;       /* over-loaded to count active stripes */
-
-       for (;logical_sector < last_sector; logical_sector += STRIPE_SECTORS) {
-               DEFINE_WAIT(w);
-
-               new_sector = raid6_compute_sector(logical_sector,
-                                                 raid_disks, data_disks, &dd_idx, &pd_idx, conf);
-
-               PRINTK("raid6: make_request, sector %llu logical %llu\n",
-                      (unsigned long long)new_sector,
-                      (unsigned long long)logical_sector);
-
-       retry:
-               prepare_to_wait(&conf->wait_for_overlap, &w, TASK_UNINTERRUPTIBLE);
-               sh = get_active_stripe(conf, new_sector, pd_idx, (bi->bi_rw&RWA_MASK));
-               if (sh) {
-                       if (!add_stripe_bio(sh, bi, dd_idx, (bi->bi_rw&RW_MASK))) {
-                               /* Add failed due to overlap.  Flush everything
-                                * and wait a while
-                                */
-                               raid6_unplug_device(mddev->queue);
-                               release_stripe(sh);
-                               schedule();
-                               goto retry;
-                       }
-                       finish_wait(&conf->wait_for_overlap, &w);
-                       raid6_plug_device(conf);
-                       handle_stripe(sh, NULL);
-                       release_stripe(sh);
-               } else {
-                       /* cannot get stripe for read-ahead, just give-up */
-                       clear_bit(BIO_UPTODATE, &bi->bi_flags);
-                       finish_wait(&conf->wait_for_overlap, &w);
-                       break;
-               }
-
-       }
-       spin_lock_irq(&conf->device_lock);
-       if (--bi->bi_phys_segments == 0) {
-               int bytes = bi->bi_size;
-
-               if (rw == WRITE )
-                       md_write_end(mddev);
-               bi->bi_size = 0;
-               bi->bi_end_io(bi, bytes, 0);
-       }
-       spin_unlock_irq(&conf->device_lock);
-       return 0;
-}
-
-/* FIXME go_faster isn't used */
-static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, int go_faster)
-{
-       raid6_conf_t *conf = (raid6_conf_t *) mddev->private;
-       struct stripe_head *sh;
-       int sectors_per_chunk = conf->chunk_size >> 9;
-       sector_t x;
-       unsigned long stripe;
-       int chunk_offset;
-       int dd_idx, pd_idx;
-       sector_t first_sector;
-       int raid_disks = conf->raid_disks;
-       int data_disks = raid_disks - 2;
-       sector_t max_sector = mddev->size << 1;
-       int sync_blocks;
-       int still_degraded = 0;
-       int i;
-
-       if (sector_nr >= max_sector) {
-               /* just being told to finish up .. nothing much to do */
-               unplug_slaves(mddev);
-
-               if (mddev->curr_resync < max_sector) /* aborted */
-                       bitmap_end_sync(mddev->bitmap, mddev->curr_resync,
-                                       &sync_blocks, 1);
-               else /* completed sync */
-                       conf->fullsync = 0;
-               bitmap_close_sync(mddev->bitmap);
-
-               return 0;
-       }
-       /* if there are 2 or more failed drives and we are trying
-        * to resync, then assert that we are finished, because there is
-        * nothing we can do.
-        */
-       if (mddev->degraded >= 2 && test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
-               sector_t rv = (mddev->size << 1) - sector_nr;
-               *skipped = 1;
-               return rv;
-       }
-       if (!bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, 1) &&
-           !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery) &&
-           !conf->fullsync && sync_blocks >= STRIPE_SECTORS) {
-               /* we can skip this block, and probably more */
-               sync_blocks /= STRIPE_SECTORS;
-               *skipped = 1;
-               return sync_blocks * STRIPE_SECTORS; /* keep things rounded to whole stripes */
-       }
-
-       x = sector_nr;
-       chunk_offset = sector_div(x, sectors_per_chunk);
-       stripe = x;
-       BUG_ON(x != stripe);
-
-       first_sector = raid6_compute_sector((sector_t)stripe*data_disks*sectors_per_chunk
-               + chunk_offset, raid_disks, data_disks, &dd_idx, &pd_idx, conf);
-       sh = get_active_stripe(conf, sector_nr, pd_idx, 1);
-       if (sh == NULL) {
-               sh = get_active_stripe(conf, sector_nr, pd_idx, 0);
-               /* make sure we don't swamp the stripe cache if someone else
-                * is trying to get access
-                */
-               schedule_timeout_uninterruptible(1);
-       }
-       /* Need to check if array will still be degraded after recovery/resync
-        * We don't need to check the 'failed' flag as when that gets set,
-        * recovery aborts.
-        */
-       for (i=0; i<mddev->raid_disks; i++)
-               if (conf->disks[i].rdev == NULL)
-                       still_degraded = 1;
-
-       bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, still_degraded);
-
-       spin_lock(&sh->lock);
-       set_bit(STRIPE_SYNCING, &sh->state);
-       clear_bit(STRIPE_INSYNC, &sh->state);
-       spin_unlock(&sh->lock);
-
-       handle_stripe(sh, NULL);
-       release_stripe(sh);
-
-       return STRIPE_SECTORS;
-}
-
-/*
- * This is our raid6 kernel thread.
- *
- * We scan the hash table for stripes which can be handled now.
- * During the scan, completed stripes are saved for us by the interrupt
- * handler, so that they will not have to wait for our next wakeup.
- */
-static void raid6d (mddev_t *mddev)
-{
-       struct stripe_head *sh;
-       raid6_conf_t *conf = mddev_to_conf(mddev);
-       int handled;
-
-       PRINTK("+++ raid6d active\n");
-
-       md_check_recovery(mddev);
-
-       handled = 0;
-       spin_lock_irq(&conf->device_lock);
-       while (1) {
-               struct list_head *first;
-
-               if (conf->seq_flush - conf->seq_write > 0) {
-                       int seq = conf->seq_flush;
-                       spin_unlock_irq(&conf->device_lock);
-                       bitmap_unplug(mddev->bitmap);
-                       spin_lock_irq(&conf->device_lock);
-                       conf->seq_write = seq;
-                       activate_bit_delay(conf);
-               }
-
-               if (list_empty(&conf->handle_list) &&
-                   atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD &&
-                   !blk_queue_plugged(mddev->queue) &&
-                   !list_empty(&conf->delayed_list))
-                       raid6_activate_delayed(conf);
-
-               if (list_empty(&conf->handle_list))
-                       break;
-
-               first = conf->handle_list.next;
-               sh = list_entry(first, struct stripe_head, lru);
-
-               list_del_init(first);
-               atomic_inc(&sh->count);
-               BUG_ON(atomic_read(&sh->count)!= 1);
-               spin_unlock_irq(&conf->device_lock);
-
-               handled++;
-               handle_stripe(sh, conf->spare_page);
-               release_stripe(sh);
-
-               spin_lock_irq(&conf->device_lock);
-       }
-       PRINTK("%d stripes handled\n", handled);
-
-       spin_unlock_irq(&conf->device_lock);
-
-       unplug_slaves(mddev);
-
-       PRINTK("--- raid6d inactive\n");
-}
-
-static ssize_t
-raid6_show_stripe_cache_size(mddev_t *mddev, char *page)
-{
-       raid6_conf_t *conf = mddev_to_conf(mddev);
-       if (conf)
-               return sprintf(page, "%d\n", conf->max_nr_stripes);
-       else
-               return 0;
-}
-
-static ssize_t
-raid6_store_stripe_cache_size(mddev_t *mddev, const char *page, size_t len)
-{
-       raid6_conf_t *conf = mddev_to_conf(mddev);
-       char *end;
-       int new;
-       if (len >= PAGE_SIZE)
-               return -EINVAL;
-       if (!conf)
-               return -ENODEV;
-
-       new = simple_strtoul(page, &end, 10);
-       if (!*page || (*end && *end != '\n') )
-               return -EINVAL;
-       if (new <= 16 || new > 32768)
-               return -EINVAL;
-       while (new < conf->max_nr_stripes) {
-               if (drop_one_stripe(conf))
-                       conf->max_nr_stripes--;
-               else
-                       break;
-       }
-       while (new > conf->max_nr_stripes) {
-               if (grow_one_stripe(conf))
-                       conf->max_nr_stripes++;
-               else break;
-       }
-       return len;
-}
-
-static struct md_sysfs_entry
-raid6_stripecache_size = __ATTR(stripe_cache_size, S_IRUGO | S_IWUSR,
-                               raid6_show_stripe_cache_size,
-                               raid6_store_stripe_cache_size);
-
-static ssize_t
-stripe_cache_active_show(mddev_t *mddev, char *page)
-{
-       raid6_conf_t *conf = mddev_to_conf(mddev);
-       if (conf)
-               return sprintf(page, "%d\n", atomic_read(&conf->active_stripes));
-       else
-               return 0;
-}
-
-static struct md_sysfs_entry
-raid6_stripecache_active = __ATTR_RO(stripe_cache_active);
-
-static struct attribute *raid6_attrs[] =  {
-       &raid6_stripecache_size.attr,
-       &raid6_stripecache_active.attr,
-       NULL,
-};
-static struct attribute_group raid6_attrs_group = {
-       .name = NULL,
-       .attrs = raid6_attrs,
-};
-
-static int run(mddev_t *mddev)
-{
-       raid6_conf_t *conf;
-       int raid_disk, memory;
-       mdk_rdev_t *rdev;
-       struct disk_info *disk;
-       struct list_head *tmp;
-
-       if (mddev->level != 6) {
-               PRINTK("raid6: %s: raid level not set to 6 (%d)\n", mdname(mddev), mddev->level);
-               return -EIO;
-       }
-
-       mddev->private = kzalloc(sizeof (raid6_conf_t), GFP_KERNEL);
-       if ((conf = mddev->private) == NULL)
-               goto abort;
-       conf->disks = kzalloc(mddev->raid_disks * sizeof(struct disk_info),
-                                GFP_KERNEL);
-       if (!conf->disks)
-               goto abort;
-
-       conf->mddev = mddev;
-
-       if ((conf->stripe_hashtbl = kzalloc(PAGE_SIZE, GFP_KERNEL)) == NULL)
-               goto abort;
-
-       conf->spare_page = alloc_page(GFP_KERNEL);
-       if (!conf->spare_page)
-               goto abort;
-
-       spin_lock_init(&conf->device_lock);
-       init_waitqueue_head(&conf->wait_for_stripe);
-       init_waitqueue_head(&conf->wait_for_overlap);
-       INIT_LIST_HEAD(&conf->handle_list);
-       INIT_LIST_HEAD(&conf->delayed_list);
-       INIT_LIST_HEAD(&conf->bitmap_list);
-       INIT_LIST_HEAD(&conf->inactive_list);
-       atomic_set(&conf->active_stripes, 0);
-       atomic_set(&conf->preread_active_stripes, 0);
-
-       PRINTK("raid6: run(%s) called.\n", mdname(mddev));
-
-       ITERATE_RDEV(mddev,rdev,tmp) {
-               raid_disk = rdev->raid_disk;
-               if (raid_disk >= mddev->raid_disks
-                   || raid_disk < 0)
-                       continue;
-               disk = conf->disks + raid_disk;
-
-               disk->rdev = rdev;
-
-               if (test_bit(In_sync, &rdev->flags)) {
-                       char b[BDEVNAME_SIZE];
-                       printk(KERN_INFO "raid6: device %s operational as raid"
-                              " disk %d\n", bdevname(rdev->bdev,b),
-                              raid_disk);
-                       conf->working_disks++;
-               }
-       }
-
-       conf->raid_disks = mddev->raid_disks;
-
-       /*
-        * 0 for a fully functional array, 1 or 2 for a degraded array.
-        */
-       mddev->degraded = conf->failed_disks = conf->raid_disks - conf->working_disks;
-       conf->mddev = mddev;
-       conf->chunk_size = mddev->chunk_size;
-       conf->level = mddev->level;
-       conf->algorithm = mddev->layout;
-       conf->max_nr_stripes = NR_STRIPES;
-
-       /* device size must be a multiple of chunk size */
-       mddev->size &= ~(mddev->chunk_size/1024 -1);
-       mddev->resync_max_sectors = mddev->size << 1;
-
-       if (conf->raid_disks < 4) {
-               printk(KERN_ERR "raid6: not enough configured devices for %s (%d, minimum 4)\n",
-                      mdname(mddev), conf->raid_disks);
-               goto abort;
-       }
-       if (!conf->chunk_size || conf->chunk_size % 4) {
-               printk(KERN_ERR "raid6: invalid chunk size %d for %s\n",
-                      conf->chunk_size, mdname(mddev));
-               goto abort;
-       }
-       if (conf->algorithm > ALGORITHM_RIGHT_SYMMETRIC) {
-               printk(KERN_ERR
-                      "raid6: unsupported parity algorithm %d for %s\n",
-                      conf->algorithm, mdname(mddev));
-               goto abort;
-       }
-       if (mddev->degraded > 2) {
-               printk(KERN_ERR "raid6: not enough operational devices for %s"
-                      " (%d/%d failed)\n",
-                      mdname(mddev), conf->failed_disks, conf->raid_disks);
-               goto abort;
-       }
-
-       if (mddev->degraded > 0 &&
-           mddev->recovery_cp != MaxSector) {
-               if (mddev->ok_start_degraded)
-                       printk(KERN_WARNING "raid6: starting dirty degraded array:%s"
-                              "- data corruption possible.\n",
-                              mdname(mddev));
-               else {
-                       printk(KERN_ERR "raid6: cannot start dirty degraded array"
-                              " for %s\n", mdname(mddev));
-                       goto abort;
-               }
-       }
-
-       {
-               mddev->thread = md_register_thread(raid6d, mddev, "%s_raid6");
-               if (!mddev->thread) {
-                       printk(KERN_ERR
-                              "raid6: couldn't allocate thread for %s\n",
-                              mdname(mddev));
-                       goto abort;
-               }
-       }
-
-       memory = conf->max_nr_stripes * (sizeof(struct stripe_head) +
-                conf->raid_disks * ((sizeof(struct bio) + PAGE_SIZE))) / 1024;
-       if (grow_stripes(conf, conf->max_nr_stripes)) {
-               printk(KERN_ERR
-                      "raid6: couldn't allocate %dkB for buffers\n", memory);
-               shrink_stripes(conf);
-               md_unregister_thread(mddev->thread);
-               goto abort;
-       } else
-               printk(KERN_INFO "raid6: allocated %dkB for %s\n",
-                      memory, mdname(mddev));
-
-       if (mddev->degraded == 0)
-               printk(KERN_INFO "raid6: raid level %d set %s active with %d out of %d"
-                      " devices, algorithm %d\n", conf->level, mdname(mddev),
-                      mddev->raid_disks-mddev->degraded, mddev->raid_disks,
-                      conf->algorithm);
-       else
-               printk(KERN_ALERT "raid6: raid level %d set %s active with %d"
-                      " out of %d devices, algorithm %d\n", conf->level,
-                      mdname(mddev), mddev->raid_disks - mddev->degraded,
-                      mddev->raid_disks, conf->algorithm);
-
-       print_raid6_conf(conf);
-
-       /* read-ahead size must cover two whole stripes, which is
-        * 2 * (n-2) * chunksize where 'n' is the number of raid devices
-        */
-       {
-               int stripe = (mddev->raid_disks-2) *
-                       (mddev->chunk_size / PAGE_SIZE);
-               if (mddev->queue->backing_dev_info.ra_pages < 2 * stripe)
-                       mddev->queue->backing_dev_info.ra_pages = 2 * stripe;
-       }
-
-       /* Ok, everything is just fine now */
-       sysfs_create_group(&mddev->kobj, &raid6_attrs_group);
-
-       mddev->array_size =  mddev->size * (mddev->raid_disks - 2);
-
-       mddev->queue->unplug_fn = raid6_unplug_device;
-       mddev->queue->issue_flush_fn = raid6_issue_flush;
-       return 0;
-abort:
-       if (conf) {
-               print_raid6_conf(conf);
-               safe_put_page(conf->spare_page);
-               kfree(conf->stripe_hashtbl);
-               kfree(conf->disks);
-               kfree(conf);
-       }
-       mddev->private = NULL;
-       printk(KERN_ALERT "raid6: failed to run raid set %s\n", mdname(mddev));
-       return -EIO;
-}
-
-
-
-static int stop (mddev_t *mddev)
-{
-       raid6_conf_t *conf = (raid6_conf_t *) mddev->private;
-
-       md_unregister_thread(mddev->thread);
-       mddev->thread = NULL;
-       shrink_stripes(conf);
-       kfree(conf->stripe_hashtbl);
-       blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
-       sysfs_remove_group(&mddev->kobj, &raid6_attrs_group);
-       kfree(conf);
-       mddev->private = NULL;
-       return 0;
-}
-
-#if RAID6_DUMPSTATE
-static void print_sh (struct seq_file *seq, struct stripe_head *sh)
-{
-       int i;
-
-       seq_printf(seq, "sh %llu, pd_idx %d, state %ld.\n",
-                  (unsigned long long)sh->sector, sh->pd_idx, sh->state);
-       seq_printf(seq, "sh %llu,  count %d.\n",
-                  (unsigned long long)sh->sector, atomic_read(&sh->count));
-       seq_printf(seq, "sh %llu, ", (unsigned long long)sh->sector);
-       for (i = 0; i < sh->raid_conf->raid_disks; i++) {
-               seq_printf(seq, "(cache%d: %p %ld) ",
-                          i, sh->dev[i].page, sh->dev[i].flags);
-       }
-       seq_printf(seq, "\n");
-}
-
-static void printall (struct seq_file *seq, raid6_conf_t *conf)
-{
-       struct stripe_head *sh;
-       struct hlist_node *hn;
-       int i;
-
-       spin_lock_irq(&conf->device_lock);
-       for (i = 0; i < NR_HASH; i++) {
-               sh = conf->stripe_hashtbl[i];
-               hlist_for_each_entry(sh, hn, &conf->stripe_hashtbl[i], hash) {
-                       if (sh->raid_conf != conf)
-                               continue;
-                       print_sh(seq, sh);
-               }
-       }
-       spin_unlock_irq(&conf->device_lock);
-}
-#endif
-
-static void status (struct seq_file *seq, mddev_t *mddev)
-{
-       raid6_conf_t *conf = (raid6_conf_t *) mddev->private;
-       int i;
-
-       seq_printf (seq, " level %d, %dk chunk, algorithm %d", mddev->level, mddev->chunk_size >> 10, mddev->layout);
-       seq_printf (seq, " [%d/%d] [", conf->raid_disks, conf->working_disks);
-       for (i = 0; i < conf->raid_disks; i++)
-               seq_printf (seq, "%s",
-                           conf->disks[i].rdev &&
-                           test_bit(In_sync, &conf->disks[i].rdev->flags) ? "U" : "_");
-       seq_printf (seq, "]");
-#if RAID6_DUMPSTATE
-       seq_printf (seq, "\n");
-       printall(seq, conf);
-#endif
-}
-
-static void print_raid6_conf (raid6_conf_t *conf)
-{
-       int i;
-       struct disk_info *tmp;
-
-       printk("RAID6 conf printout:\n");
-       if (!conf) {
-               printk("(conf==NULL)\n");
-               return;
-       }
-       printk(" --- rd:%d wd:%d fd:%d\n", conf->raid_disks,
-                conf->working_disks, conf->failed_disks);
-
-       for (i = 0; i < conf->raid_disks; i++) {
-               char b[BDEVNAME_SIZE];
-               tmp = conf->disks + i;
-               if (tmp->rdev)
-               printk(" disk %d, o:%d, dev:%s\n",
-                       i, !test_bit(Faulty, &tmp->rdev->flags),
-                       bdevname(tmp->rdev->bdev,b));
-       }
-}
-
-static int raid6_spare_active(mddev_t *mddev)
-{
-       int i;
-       raid6_conf_t *conf = mddev->private;
-       struct disk_info *tmp;
-
-       for (i = 0; i < conf->raid_disks; i++) {
-               tmp = conf->disks + i;
-               if (tmp->rdev
-                   && !test_bit(Faulty, &tmp->rdev->flags)
-                   && !test_bit(In_sync, &tmp->rdev->flags)) {
-                       mddev->degraded--;
-                       conf->failed_disks--;
-                       conf->working_disks++;
-                       set_bit(In_sync, &tmp->rdev->flags);
-               }
-       }
-       print_raid6_conf(conf);
-       return 0;
-}
-
-static int raid6_remove_disk(mddev_t *mddev, int number)
-{
-       raid6_conf_t *conf = mddev->private;
-       int err = 0;
-       mdk_rdev_t *rdev;
-       struct disk_info *p = conf->disks + number;
-
-       print_raid6_conf(conf);
-       rdev = p->rdev;
-       if (rdev) {
-               if (test_bit(In_sync, &rdev->flags) ||
-                   atomic_read(&rdev->nr_pending)) {
-                       err = -EBUSY;
-                       goto abort;
-               }
-               p->rdev = NULL;
-               synchronize_rcu();
-               if (atomic_read(&rdev->nr_pending)) {
-                       /* lost the race, try later */
-                       err = -EBUSY;
-                       p->rdev = rdev;
-               }
-       }
-
-abort:
-
-       print_raid6_conf(conf);
-       return err;
-}
-
-static int raid6_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
-{
-       raid6_conf_t *conf = mddev->private;
-       int found = 0;
-       int disk;
-       struct disk_info *p;
-
-       if (mddev->degraded > 2)
-               /* no point adding a device */
-               return 0;
-       /*
-        * find the disk ... but prefer rdev->saved_raid_disk
-        * if possible.
-        */
-       if (rdev->saved_raid_disk >= 0 &&
-           conf->disks[rdev->saved_raid_disk].rdev == NULL)
-               disk = rdev->saved_raid_disk;
-       else
-               disk = 0;
-       for ( ; disk < mddev->raid_disks; disk++)
-               if ((p=conf->disks + disk)->rdev == NULL) {
-                       clear_bit(In_sync, &rdev->flags);
-                       rdev->raid_disk = disk;
-                       found = 1;
-                       if (rdev->saved_raid_disk != disk)
-                               conf->fullsync = 1;
-                       rcu_assign_pointer(p->rdev, rdev);
-                       break;
-               }
-       print_raid6_conf(conf);
-       return found;
-}
-
-static int raid6_resize(mddev_t *mddev, sector_t sectors)
-{
-       /* no resync is happening, and there is enough space
-        * on all devices, so we can resize.
-        * We need to make sure resync covers any new space.
-        * If the array is shrinking we should possibly wait until
-        * any io in the removed space completes, but it hardly seems
-        * worth it.
-        */
-       sectors &= ~((sector_t)mddev->chunk_size/512 - 1);
-       mddev->array_size = (sectors * (mddev->raid_disks-2))>>1;
-       set_capacity(mddev->gendisk, mddev->array_size << 1);
-       mddev->changed = 1;
-       if (sectors/2  > mddev->size && mddev->recovery_cp == MaxSector) {
-               mddev->recovery_cp = mddev->size << 1;
-               set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
-       }
-       mddev->size = sectors /2;
-       mddev->resync_max_sectors = sectors;
-       return 0;
-}
-
-static void raid6_quiesce(mddev_t *mddev, int state)
-{
-       raid6_conf_t *conf = mddev_to_conf(mddev);
-
-       switch(state) {
-       case 1: /* stop all writes */
-               spin_lock_irq(&conf->device_lock);
-               conf->quiesce = 1;
-               wait_event_lock_irq(conf->wait_for_stripe,
-                                   atomic_read(&conf->active_stripes) == 0,
-                                   conf->device_lock, /* nothing */);
-               spin_unlock_irq(&conf->device_lock);
-               break;
-
-       case 0: /* re-enable writes */
-               spin_lock_irq(&conf->device_lock);
-               conf->quiesce = 0;
-               wake_up(&conf->wait_for_stripe);
-               spin_unlock_irq(&conf->device_lock);
-               break;
-       }
-}
-
-static struct mdk_personality raid6_personality =
-{
-       .name           = "raid6",
-       .level          = 6,
-       .owner          = THIS_MODULE,
-       .make_request   = make_request,
-       .run            = run,
-       .stop           = stop,
-       .status         = status,
-       .error_handler  = error,
-       .hot_add_disk   = raid6_add_disk,
-       .hot_remove_disk= raid6_remove_disk,
-       .spare_active   = raid6_spare_active,
-       .sync_request   = sync_request,
-       .resize         = raid6_resize,
-       .quiesce        = raid6_quiesce,
-};
-
-static int __init raid6_init(void)
-{
-       int e;
-
-       e = raid6_select_algo();
-       if ( e )
-               return e;
-
-       return register_md_personality(&raid6_personality);
-}
-
-static void raid6_exit (void)
-{
-       unregister_md_personality(&raid6_personality);
-}
-
-module_init(raid6_init);
-module_exit(raid6_exit);
-MODULE_LICENSE("GPL");
-MODULE_ALIAS("md-personality-8"); /* RAID6 */
-MODULE_ALIAS("md-raid6");
-MODULE_ALIAS("md-level-6");
index 914af667044f6f39eaf0a9ef51e6fa1966b44175..20ed4c997636fac9b92aa003a982c5c641d1a45d 100644 (file)
@@ -212,6 +212,7 @@ struct raid5_private_data {
        mddev_t                 *mddev;
        struct disk_info        *spare;
        int                     chunk_size, level, algorithm;
+       int                     max_degraded;
        int                     raid_disks, working_disks, failed_disks;
        int                     max_nr_stripes;