diff options
Diffstat (limited to 'drivers/block/aoe')
-rw-r--r-- | drivers/block/aoe/aoe.h | 93 | ||||
-rw-r--r-- | drivers/block/aoe/aoeblk.c | 91 | ||||
-rw-r--r-- | drivers/block/aoe/aoechr.c | 13 | ||||
-rw-r--r-- | drivers/block/aoe/aoecmd.c | 1232 | ||||
-rw-r--r-- | drivers/block/aoe/aoedev.c | 265 | ||||
-rw-r--r-- | drivers/block/aoe/aoemain.c | 10 | ||||
-rw-r--r-- | drivers/block/aoe/aoenet.c | 61 |
7 files changed, 1168 insertions, 597 deletions
diff --git a/drivers/block/aoe/aoe.h b/drivers/block/aoe/aoe.h index db195abad69..d2ed7f18d1a 100644 --- a/drivers/block/aoe/aoe.h +++ b/drivers/block/aoe/aoe.h @@ -1,5 +1,5 @@ -/* Copyright (c) 2007 Coraid, Inc. See COPYING for GPL terms. */ -#define VERSION "47" +/* Copyright (c) 2012 Coraid, Inc. See COPYING for GPL terms. */ +#define VERSION "50" #define AOE_MAJOR 152 #define DEVICE_NAME "aoe" @@ -10,9 +10,6 @@ #define AOE_PARTITIONS (16) #endif -#define SYSMINOR(aoemajor, aoeminor) ((aoemajor) * NPERSHELF + (aoeminor)) -#define AOEMAJOR(sysminor) ((sysminor) / NPERSHELF) -#define AOEMINOR(sysminor) ((sysminor) % NPERSHELF) #define WHITESPACE " \t\v\f\n" enum { @@ -75,72 +72,67 @@ enum { DEVFL_UP = 1, /* device is installed in system and ready for AoE->ATA commands */ DEVFL_TKILL = (1<<1), /* flag for timer to know when to kill self */ DEVFL_EXT = (1<<2), /* device accepts lba48 commands */ - DEVFL_CLOSEWAIT = (1<<3), /* device is waiting for all closes to revalidate */ - DEVFL_GDALLOC = (1<<4), /* need to alloc gendisk */ - DEVFL_KICKME = (1<<5), /* slow polling network card catch */ - DEVFL_NEWSIZE = (1<<6), /* need to update dev size in block layer */ - - BUFFL_FAIL = 1, + DEVFL_GDALLOC = (1<<3), /* need to alloc gendisk */ + DEVFL_KICKME = (1<<4), /* slow polling network card catch */ + DEVFL_NEWSIZE = (1<<5), /* need to update dev size in block layer */ }; enum { DEFAULTBCNT = 2 * 512, /* 2 sectors */ - NPERSHELF = 16, /* number of slots per shelf address */ - FREETAG = -1, MIN_BUFS = 16, NTARGETS = 8, NAOEIFS = 8, - NSKBPOOLMAX = 128, + NSKBPOOLMAX = 256, + NFACTIVE = 61, TIMERTICK = HZ / 10, MINTIMER = HZ >> 2, MAXTIMER = HZ << 1, - HELPWAIT = 20, }; struct buf { - struct list_head bufs; - ulong stime; /* for disk stats */ - ulong flags; ulong nframesout; ulong resid; ulong bv_resid; - ulong bv_off; sector_t sector; struct bio *bio; struct bio_vec *bv; + struct request *rq; }; struct frame { - int tag; + struct list_head head; + u32 tag; ulong waited; + struct aoetgt *t; /* parent target I belong to */ + sector_t lba; + struct sk_buff *skb; /* command skb freed on module exit */ + struct sk_buff *r_skb; /* response skb for async processing */ struct buf *buf; - char *bufaddr; + struct bio_vec *bv; ulong bcnt; - sector_t lba; - struct sk_buff *skb; + ulong bv_off; }; struct aoeif { struct net_device *nd; - unsigned char lost; - unsigned char lostjumbo; - ushort maxbcnt; + ulong lost; + int bcnt; }; struct aoetgt { unsigned char addr[6]; ushort nframes; - struct frame *frames; + struct aoedev *d; /* parent device I belong to */ + struct list_head ffree; /* list of free frames */ struct aoeif ifs[NAOEIFS]; struct aoeif *ifp; /* current aoeif in use */ ushort nout; ushort maxout; - u16 lasttag; /* last tag sent */ - u16 useme; + ulong falloc; ulong lastwadj; /* last window adjustment */ + int minbcnt; int wpkts, rpkts; - int dataref; }; struct aoedev { @@ -153,6 +145,9 @@ struct aoedev { u16 rttavg; /* round trip average of requests/responses */ u16 mintimer; u16 fw_ver; /* version of blade's firmware */ + u16 lasttag; /* last tag sent */ + u16 useme; + ulong ref; struct work_struct work;/* disk create work struct */ struct gendisk *gd; struct request_queue *blkq; @@ -160,16 +155,31 @@ struct aoedev { sector_t ssize; struct timer_list timer; spinlock_t lock; - struct sk_buff_head sendq; struct sk_buff_head skbpool; mempool_t *bufpool; /* for deadlock-free Buf allocation */ - struct list_head bufq; /* queue of bios to work on */ - struct buf *inprocess; /* the one we're currently working on */ + struct { /* pointers to work in progress */ + struct buf *buf; + struct bio *nxbio; + struct request *rq; + } ip; + ulong maxbcnt; + struct list_head factive[NFACTIVE]; /* hash of active frames */ struct aoetgt *targets[NTARGETS]; struct aoetgt **tgt; /* target in use when working */ - struct aoetgt **htgt; /* target needing rexmit assistance */ + struct aoetgt *htgt; /* target needing rexmit assistance */ + ulong ntargets; + ulong kicked; }; +/* kthread tracking */ +struct ktstate { + struct completion rendez; + struct task_struct *task; + wait_queue_head_t *waitq; + int (*fn) (void); + char *name; + spinlock_t *lock; +}; int aoeblk_init(void); void aoeblk_exit(void); @@ -182,22 +192,29 @@ void aoechr_error(char *); void aoecmd_work(struct aoedev *d); void aoecmd_cfg(ushort aoemajor, unsigned char aoeminor); -void aoecmd_ata_rsp(struct sk_buff *); +struct sk_buff *aoecmd_ata_rsp(struct sk_buff *); void aoecmd_cfg_rsp(struct sk_buff *); void aoecmd_sleepwork(struct work_struct *); void aoecmd_cleanslate(struct aoedev *); +void aoecmd_exit(void); +int aoecmd_init(void); struct sk_buff *aoecmd_ata_id(struct aoedev *); +void aoe_freetframe(struct frame *); +void aoe_flush_iocq(void); +void aoe_end_request(struct aoedev *, struct request *, int); +int aoe_ktstart(struct ktstate *k); +void aoe_ktstop(struct ktstate *k); int aoedev_init(void); void aoedev_exit(void); -struct aoedev *aoedev_by_aoeaddr(int maj, int min); -struct aoedev *aoedev_by_sysminor_m(ulong sysminor); +struct aoedev *aoedev_by_aoeaddr(ulong maj, int min, int do_alloc); void aoedev_downdev(struct aoedev *d); int aoedev_flush(const char __user *str, size_t size); +void aoe_failbuf(struct aoedev *, struct buf *); +void aoedev_put(struct aoedev *); int aoenet_init(void); void aoenet_exit(void); void aoenet_xmit(struct sk_buff_head *); int is_aoe_netif(struct net_device *ifp); int set_aoe_iflist(const char __user *str, size_t size); - diff --git a/drivers/block/aoe/aoeblk.c b/drivers/block/aoe/aoeblk.c index 321de7b6c44..00dfc5008ad 100644 --- a/drivers/block/aoe/aoeblk.c +++ b/drivers/block/aoe/aoeblk.c @@ -1,4 +1,4 @@ -/* Copyright (c) 2007 Coraid, Inc. See COPYING for GPL terms. */ +/* Copyright (c) 2012 Coraid, Inc. See COPYING for GPL terms. */ /* * aoeblk.c * block device routines @@ -161,68 +161,22 @@ aoeblk_release(struct gendisk *disk, fmode_t mode) } static void -aoeblk_make_request(struct request_queue *q, struct bio *bio) +aoeblk_request(struct request_queue *q) { - struct sk_buff_head queue; struct aoedev *d; - struct buf *buf; - ulong flags; - - blk_queue_bounce(q, &bio); - - if (bio == NULL) { - printk(KERN_ERR "aoe: bio is NULL\n"); - BUG(); - return; - } - d = bio->bi_bdev->bd_disk->private_data; - if (d == NULL) { - printk(KERN_ERR "aoe: bd_disk->private_data is NULL\n"); - BUG(); - bio_endio(bio, -ENXIO); - return; - } else if (bio->bi_io_vec == NULL) { - printk(KERN_ERR "aoe: bi_io_vec is NULL\n"); - BUG(); - bio_endio(bio, -ENXIO); - return; - } - buf = mempool_alloc(d->bufpool, GFP_NOIO); - if (buf == NULL) { - printk(KERN_INFO "aoe: buf allocation failure\n"); - bio_endio(bio, -ENOMEM); - return; - } - memset(buf, 0, sizeof(*buf)); - INIT_LIST_HEAD(&buf->bufs); - buf->stime = jiffies; - buf->bio = bio; - buf->resid = bio->bi_size; - buf->sector = bio->bi_sector; - buf->bv = &bio->bi_io_vec[bio->bi_idx]; - buf->bv_resid = buf->bv->bv_len; - WARN_ON(buf->bv_resid == 0); - buf->bv_off = buf->bv->bv_offset; - - spin_lock_irqsave(&d->lock, flags); + struct request *rq; + d = q->queuedata; if ((d->flags & DEVFL_UP) == 0) { pr_info_ratelimited("aoe: device %ld.%d is not up\n", d->aoemajor, d->aoeminor); - spin_unlock_irqrestore(&d->lock, flags); - mempool_free(buf, d->bufpool); - bio_endio(bio, -ENXIO); + while ((rq = blk_peek_request(q))) { + blk_start_request(rq); + aoe_end_request(d, rq, 1); + } return; } - - list_add_tail(&buf->bufs, &d->bufq); - aoecmd_work(d); - __skb_queue_head_init(&queue); - skb_queue_splice_init(&d->sendq, &queue); - - spin_unlock_irqrestore(&d->lock, flags); - aoenet_xmit(&queue); } static int @@ -254,41 +208,54 @@ aoeblk_gdalloc(void *vp) { struct aoedev *d = vp; struct gendisk *gd; + mempool_t *mp; + struct request_queue *q; + enum { KB = 1024, MB = KB * KB, READ_AHEAD = 2 * MB, }; ulong flags; gd = alloc_disk(AOE_PARTITIONS); if (gd == NULL) { - printk(KERN_ERR - "aoe: cannot allocate disk structure for %ld.%d\n", + pr_err("aoe: cannot allocate disk structure for %ld.%d\n", d->aoemajor, d->aoeminor); goto err; } - d->bufpool = mempool_create_slab_pool(MIN_BUFS, buf_pool_cache); - if (d->bufpool == NULL) { + mp = mempool_create(MIN_BUFS, mempool_alloc_slab, mempool_free_slab, + buf_pool_cache); + if (mp == NULL) { printk(KERN_ERR "aoe: cannot allocate bufpool for %ld.%d\n", d->aoemajor, d->aoeminor); goto err_disk; } + q = blk_init_queue(aoeblk_request, &d->lock); + if (q == NULL) { + pr_err("aoe: cannot allocate block queue for %ld.%d\n", + d->aoemajor, d->aoeminor); + mempool_destroy(mp); + goto err_disk; + } d->blkq = blk_alloc_queue(GFP_KERNEL); if (!d->blkq) goto err_mempool; - blk_queue_make_request(d->blkq, aoeblk_make_request); d->blkq->backing_dev_info.name = "aoe"; if (bdi_init(&d->blkq->backing_dev_info)) goto err_blkq; spin_lock_irqsave(&d->lock, flags); + blk_queue_max_hw_sectors(d->blkq, BLK_DEF_MAX_SECTORS); + q->backing_dev_info.ra_pages = READ_AHEAD / PAGE_CACHE_SIZE; + d->bufpool = mp; + d->blkq = gd->queue = q; + q->queuedata = d; + d->gd = gd; gd->major = AOE_MAJOR; - gd->first_minor = d->sysminor * AOE_PARTITIONS; + gd->first_minor = d->sysminor; gd->fops = &aoe_bdops; gd->private_data = d; set_capacity(gd, d->ssize); snprintf(gd->disk_name, sizeof gd->disk_name, "etherd/e%ld.%d", d->aoemajor, d->aoeminor); - gd->queue = d->blkq; - d->gd = gd; d->flags &= ~DEVFL_GDALLOC; d->flags |= DEVFL_UP; diff --git a/drivers/block/aoe/aoechr.c b/drivers/block/aoe/aoechr.c index e86d2062a16..ed57a890c64 100644 --- a/drivers/block/aoe/aoechr.c +++ b/drivers/block/aoe/aoechr.c @@ -1,4 +1,4 @@ -/* Copyright (c) 2007 Coraid, Inc. See COPYING for GPL terms. */ +/* Copyright (c) 2012 Coraid, Inc. See COPYING for GPL terms. */ /* * aoechr.c * AoE character device driver @@ -86,34 +86,34 @@ revalidate(const char __user *str, size_t size) if (copy_from_user(buf, str, size)) return -EFAULT; - /* should be e%d.%d format */ n = sscanf(buf, "e%d.%d", &major, &minor); if (n != 2) { - printk(KERN_ERR "aoe: invalid device specification\n"); + pr_err("aoe: invalid device specification %s\n", buf); return -EINVAL; } - d = aoedev_by_aoeaddr(major, minor); + d = aoedev_by_aoeaddr(major, minor, 0); if (!d) return -EINVAL; spin_lock_irqsave(&d->lock, flags); aoecmd_cleanslate(d); + aoecmd_cfg(major, minor); loop: skb = aoecmd_ata_id(d); spin_unlock_irqrestore(&d->lock, flags); /* try again if we are able to sleep a bit, * otherwise give up this revalidation */ - if (!skb && !msleep_interruptible(200)) { + if (!skb && !msleep_interruptible(250)) { spin_lock_irqsave(&d->lock, flags); goto loop; } + aoedev_put(d); if (skb) { struct sk_buff_head queue; __skb_queue_head_init(&queue); __skb_queue_tail(&queue, skb); aoenet_xmit(&queue); } - aoecmd_cfg(major, minor); return 0; } @@ -174,6 +174,7 @@ aoechr_write(struct file *filp, const char __user *buf, size_t cnt, loff_t *offp break; case MINOR_FLUSH: ret = aoedev_flush(buf, cnt); + break; } if (ret == 0) ret = cnt; diff --git a/drivers/block/aoe/aoecmd.c b/drivers/block/aoe/aoecmd.c index 887f68f6d79..3804a0af3ef 100644 --- a/drivers/block/aoe/aoecmd.c +++ b/drivers/block/aoe/aoecmd.c @@ -1,4 +1,4 @@ -/* Copyright (c) 2007 Coraid, Inc. See COPYING for GPL terms. */ +/* Copyright (c) 2012 Coraid, Inc. See COPYING for GPL terms. */ /* * aoecmd.c * Filesystem request handling methods @@ -12,10 +12,19 @@ #include <linux/netdevice.h> #include <linux/genhd.h> #include <linux/moduleparam.h> +#include <linux/workqueue.h> +#include <linux/kthread.h> #include <net/net_namespace.h> #include <asm/unaligned.h> +#include <linux/uio.h> #include "aoe.h" +#define MAXIOC (8192) /* default meant to avoid most soft lockups */ + +static void ktcomplete(struct frame *, struct sk_buff *); + +static struct buf *nextbuf(struct aoedev *); + static int aoe_deadsecs = 60 * 3; module_param(aoe_deadsecs, int, 0644); MODULE_PARM_DESC(aoe_deadsecs, "After aoe_deadsecs seconds, give up and fail dev."); @@ -25,6 +34,15 @@ module_param(aoe_maxout, int, 0644); MODULE_PARM_DESC(aoe_maxout, "Only aoe_maxout outstanding packets for every MAC on eX.Y."); +static wait_queue_head_t ktiowq; +static struct ktstate kts; + +/* io completion queue */ +static struct { + struct list_head head; + spinlock_t lock; +} iocq; + static struct sk_buff * new_skb(ulong len) { @@ -41,15 +59,21 @@ new_skb(ulong len) } static struct frame * -getframe(struct aoetgt *t, int tag) +getframe(struct aoedev *d, u32 tag) { - struct frame *f, *e; + struct frame *f; + struct list_head *head, *pos, *nx; + u32 n; - f = t->frames; - e = f + t->nframes; - for (; f<e; f++) - if (f->tag == tag) + n = tag % NFACTIVE; + head = &d->factive[n]; + list_for_each_safe(pos, nx, head) { + f = list_entry(pos, struct frame, head); + if (f->tag == tag) { + list_del(pos); return f; + } + } return NULL; } @@ -59,18 +83,18 @@ getframe(struct aoetgt *t, int tag) * This driver reserves tag -1 to mean "unused frame." */ static int -newtag(struct aoetgt *t) +newtag(struct aoedev *d) { register ulong n; n = jiffies & 0xffff; - return n |= (++t->lasttag & 0x7fff) << 16; + return n |= (++d->lasttag & 0x7fff) << 16; } -static int +static u32 aoehdr_atainit(struct aoedev *d, struct aoetgt *t, struct aoe_hdr *h) { - u32 host_tag = newtag(t); + u32 host_tag = newtag(d); memcpy(h->src, t->ifp->nd->dev_addr, sizeof h->src); memcpy(h->dst, t->addr, sizeof h->dst); @@ -95,16 +119,18 @@ put_lba(struct aoe_atahdr *ah, sector_t lba) ah->lba5 = lba >>= 8; } -static void +static struct aoeif * ifrotate(struct aoetgt *t) { - t->ifp++; - if (t->ifp >= &t->ifs[NAOEIFS] || t->ifp->nd == NULL) - t->ifp = t->ifs; - if (t->ifp->nd == NULL) { - printk(KERN_INFO "aoe: no interface to rotate to\n"); - BUG(); - } + struct aoeif *ifp; + + ifp = t->ifp; + ifp++; + if (ifp >= &t->ifs[NAOEIFS] || ifp->nd == NULL) + ifp = t->ifs; + if (ifp->nd == NULL) + return NULL; + return t->ifp = ifp; } static void @@ -129,78 +155,128 @@ skb_pool_get(struct aoedev *d) return NULL; } -/* freeframe is where we do our load balancing so it's a little hairy. */ +void +aoe_freetframe(struct frame *f) +{ + struct aoetgt *t; + + t = f->t; + f->buf = NULL; + f->bv = NULL; + f->r_skb = NULL; + list_add(&f->head, &t->ffree); +} + static struct frame * -freeframe(struct aoedev *d) +newtframe(struct aoedev *d, struct aoetgt *t) { - struct frame *f, *e, *rf; - struct aoetgt **t; + struct frame *f; struct sk_buff *skb; + struct list_head *pos; + + if (list_empty(&t->ffree)) { + if (t->falloc >= NSKBPOOLMAX*2) + return NULL; + f = kcalloc(1, sizeof(*f), GFP_ATOMIC); + if (f == NULL) + return NULL; + t->falloc++; + f->t = t; + } else { + pos = t->ffree.next; + list_del(pos); + f = list_entry(pos, struct frame, head); + } + + skb = f->skb; + if (skb == NULL) { + f->skb = skb = new_skb(ETH_ZLEN); + if (!skb) { +bail: aoe_freetframe(f); + return NULL; + } + } + + if (atomic_read(&skb_shinfo(skb)->dataref) != 1) { + skb = skb_pool_get(d); + if (skb == NULL) + goto bail; + skb_pool_put(d, f->skb); + f->skb = skb; + } + + skb->truesize -= skb->data_len; + skb_shinfo(skb)->nr_frags = skb->data_len = 0; + skb_trim(skb, 0); + return f; +} + +static struct frame * +newframe(struct aoedev *d) +{ + struct frame *f; + struct aoetgt *t, **tt; + int totout = 0; if (d->targets[0] == NULL) { /* shouldn't happen, but I'm paranoid */ printk(KERN_ERR "aoe: NULL TARGETS!\n"); return NULL; } - t = d->tgt; - t++; - if (t >= &d->targets[NTARGETS] || !*t) - t = d->targets; + tt = d->tgt; /* last used target */ for (;;) { - if ((*t)->nout < (*t)->maxout + tt++; + if (tt >= &d->targets[NTARGETS] || !*tt) + tt = d->targets; + t = *tt; + totout += t->nout; + if (t->nout < t->maxout && t != d->htgt - && (*t)->ifp->nd) { - rf = NULL; - f = (*t)->frames; - e = f + (*t)->nframes; - for (; f < e; f++) { - if (f->tag != FREETAG) - continue; - skb = f->skb; - if (!skb - && !(f->skb = skb = new_skb(ETH_ZLEN))) - continue; - if (atomic_read(&skb_shinfo(skb)->dataref) - != 1) { - if (!rf) - rf = f; - continue; - } -gotone: skb_shinfo(skb)->nr_frags = skb->data_len = 0; - skb_trim(skb, 0); - d->tgt = t; - ifrotate(*t); + && t->ifp->nd) { + f = newtframe(d, t); + if (f) { + ifrotate(t); + d->tgt = tt; return f; } - /* Work can be done, but the network layer is - holding our precious packets. Try to grab - one from the pool. */ - f = rf; - if (f == NULL) { /* more paranoia */ - printk(KERN_ERR - "aoe: freeframe: %s.\n", - "unexpected null rf"); - d->flags |= DEVFL_KICKME; - return NULL; - } - skb = skb_pool_get(d); - if (skb) { - skb_pool_put(d, f->skb); - f->skb = skb; - goto gotone; - } - (*t)->dataref++; - if ((*t)->nout == 0) - d->flags |= DEVFL_KICKME; } - if (t == d->tgt) /* we've looped and found nada */ + if (tt == d->tgt) /* we've looped and found nada */ break; - t++; - if (t >= &d->targets[NTARGETS] || !*t) - t = d->targets; + } + if (totout == 0) { + d->kicked++; + d->flags |= DEVFL_KICKME; } return NULL; } +static void +skb_fillup(struct sk_buff *skb, struct bio_vec *bv, ulong off, ulong cnt) +{ + int frag = 0; + ulong fcnt; +loop: + fcnt = bv->bv_len - (off - bv->bv_offset); + if (fcnt > cnt) + fcnt = cnt; + skb_fill_page_desc(skb, frag++, bv->bv_page, off, fcnt); + cnt -= fcnt; + if (cnt <= 0) + return; + bv++; + off = bv->bv_offset; + goto loop; +} + +static void +fhash(struct frame *f) +{ + struct aoedev *d = f->t->d; + u32 n; + + n = f->tag % NFACTIVE; + list_add_tail(&f->head, &d->factive[n]); +} + static int aoecmd_ata_rw(struct aoedev *d) { @@ -208,26 +284,47 @@ aoecmd_ata_rw(struct aoedev *d) struct aoe_hdr *h; struct aoe_atahdr *ah; struct buf *buf; - struct bio_vec *bv; struct aoetgt *t; struct sk_buff *skb; - ulong bcnt; + struct sk_buff_head queue; + ulong bcnt, fbcnt; char writebit, extbit; writebit = 0x10; extbit = 0x4; - f = freeframe(d); + buf = nextbuf(d); + if (buf == NULL) + return 0; + f = newframe(d); if (f == NULL) return 0; t = *d->tgt; - buf = d->inprocess; - bv = buf->bv; - bcnt = t->ifp->maxbcnt; + bcnt = d->maxbcnt; if (bcnt == 0) bcnt = DEFAULTBCNT; - if (bcnt > buf->bv_resid) - bcnt = buf->bv_resid; + if (bcnt > buf->resid) + bcnt = buf->resid; + fbcnt = bcnt; + f->bv = buf->bv; + f->bv_off = f->bv->bv_offset + (f->bv->bv_len - buf->bv_resid); + do { + if (fbcnt < buf->bv_resid) { + buf->bv_resid -= fbcnt; + buf->resid -= fbcnt; + break; + } + fbcnt -= buf->bv_resid; + buf->resid -= buf->bv_resid; + if (buf->resid == 0) { + d->ip.buf = NULL; + break; + } + buf->bv++; + buf->bv_resid = buf->bv->bv_len; + WARN_ON(buf->bv_resid == 0); + } while (fbcnt); + /* initialize the headers & frame */ skb = f->skb; h = (struct aoe_hdr *) skb_mac_header(skb); @@ -235,10 +332,10 @@ aoecmd_ata_rw(struct aoedev *d) skb_put(skb, sizeof *h + sizeof *ah); memset(h, 0, skb->len); f->tag = aoehdr_atainit(d, t, h); + fhash(f); t->nout++; f->waited = 0; f->buf = buf; - f->bufaddr = page_address(bv->bv_page) + buf->bv_off; f->bcnt = bcnt; f->lba = buf->sector; @@ -253,10 +350,11 @@ aoecmd_ata_rw(struct aoedev *d) ah->lba3 |= 0xe0; /* LBA bit + obsolete 0xa0 */ } if (bio_data_dir(buf->bio) == WRITE) { - skb_fill_page_desc(skb, 0, bv->bv_page, buf->bv_off, bcnt); + skb_fillup(skb, f->bv, f->bv_off, bcnt); ah->aflags |= AOEAFL_WRITE; skb->len += bcnt; skb->data_len = bcnt; + skb->truesize += bcnt; t->wpkts++; } else { t->rpkts++; @@ -267,23 +365,15 @@ aoecmd_ata_rw(struct aoedev *d) /* mark all tracking fields and load out */ buf->nframesout += 1; - buf->bv_off += bcnt; - buf->bv_resid -= bcnt; - buf->resid -= bcnt; buf->sector += bcnt >> 9; - if (buf->resid == 0) { - d->inprocess = NULL; - } else if (buf->bv_resid == 0) { - buf->bv = ++bv; - buf->bv_resid = bv->bv_len; - WARN_ON(buf->bv_resid == 0); - buf->bv_off = bv->bv_offset; - } skb->dev = t->ifp->nd; skb = skb_clone(skb, GFP_ATOMIC); - if (skb) - __skb_queue_tail(&d->sendq, skb); + if (skb) { + __skb_queue_head_init(&queue); + __skb_queue_tail(&queue, skb); + aoenet_xmit(&queue); + } return 1; } @@ -330,17 +420,25 @@ cont: } static void -resend(struct aoedev *d, struct aoetgt *t, struct frame *f) +resend(struct aoedev *d, struct frame *f) { struct sk_buff *skb; + struct sk_buff_head queue; struct aoe_hdr *h; struct aoe_atahdr *ah; + struct aoetgt *t; char buf[128]; u32 n; - ifrotate(t); - n = newtag(t); + t = f->t; + n = newtag(d); skb = f->skb; + if (ifrotate(t) == NULL) { + /* probably can't happen, but set it up to fail anyway */ + pr_info("aoe: resend: no interfaces to rotate to.\n"); + ktcomplete(f, NULL); + return; + } h = (struct aoe_hdr *) skb_mac_header(skb); ah = (struct aoe_atahdr *) (h+1); @@ -351,39 +449,22 @@ resend(struct aoedev *d, struct aoetgt *t, struct frame *f) aoechr_error(buf); f->tag = n; + fhash(f); h->tag = cpu_to_be32(n); memcpy(h->dst, t->addr, sizeof h->dst); memcpy(h->src, t->ifp->nd->dev_addr, sizeof h->src); - switch (ah->cmdstat) { - default: - break; - case ATA_CMD_PIO_READ: - case ATA_CMD_PIO_READ_EXT: - case ATA_CMD_PIO_WRITE: - case ATA_CMD_PIO_WRITE_EXT: - put_lba(ah, f->lba); - - n = f->bcnt; - if (n > DEFAULTBCNT) - n = DEFAULTBCNT; - ah->scnt = n >> 9; - if (ah->aflags & AOEAFL_WRITE) { - skb_fill_page_desc(skb, 0, virt_to_page(f->bufaddr), - offset_in_page(f->bufaddr), n); - skb->len = sizeof *h + sizeof *ah + n; - skb->data_len = n; - } - } skb->dev = t->ifp->nd; skb = skb_clone(skb, GFP_ATOMIC); if (skb == NULL) return; - __skb_queue_tail(&d->sendq, skb); + __skb_queue_head_init(&queue); + __skb_queue_tail(&queue, skb); + aoenet_xmit(&queue); } static int -tsince(int tag) +tsince(u32 tag) { int n; @@ -407,58 +488,65 @@ getif(struct aoetgt *t, struct net_device *nd) return NULL; } -static struct aoeif * -addif(struct aoetgt *t, struct net_device *nd) -{ - struct aoeif *p; - - p = getif(t, NULL); - if (!p) - return NULL; - p->nd = nd; - p->maxbcnt = DEFAULTBCNT; - p->lost = 0; - p->lostjumbo = 0; - return p; -} - static void ejectif(struct aoetgt *t, struct aoeif *ifp) { struct aoeif *e; + struct net_device *nd; ulong n; + nd = ifp->nd; e = t->ifs + NAOEIFS - 1; n = (e - ifp) * sizeof *ifp; memmove(ifp, ifp+1, n); e->nd = NULL; + dev_put(nd); } static int sthtith(struct aoedev *d) { - struct frame *f, *e, *nf; + struct frame *f, *nf; + struct list_head *nx, *pos, *head; struct sk_buff *skb; - struct aoetgt *ht = *d->htgt; - - f = ht->frames; - e = f + ht->nframes; - for (; f < e; f++) { - if (f->tag == FREETAG) - continue; - nf = freeframe(d); - if (!nf) - return 0; - skb = nf->skb; - *nf = *f; - f->skb = skb; - f->tag = FREETAG; - nf->waited = 0; - ht->nout--; - (*d->tgt)->nout++; - resend(d, *d->tgt, nf); + struct aoetgt *ht = d->htgt; + int i; + + for (i = 0; i < NFACTIVE; i++) { + head = &d->factive[i]; + list_for_each_safe(pos, nx, head) { + f = list_entry(pos, struct frame, head); + if (f->t != ht) + continue; + + nf = newframe(d); + if (!nf) + return 0; + + /* remove frame from active list */ + list_del(pos); + + /* reassign all pertinent bits to new outbound frame */ + skb = nf->skb; + nf->skb = f->skb; + nf->buf = f->buf; + nf->bcnt = f->bcnt; + nf->lba = f->lba; + nf->bv = f->bv; + nf->bv_off = f->bv_off; + nf->waited = 0; + f->skb = skb; + aoe_freetframe(f); + ht->nout--; + nf->t->nout++; + resend(d, nf); + } } - /* he's clean, he's useless. take away his interfaces */ + /* We've cleaned up the outstanding so take away his + * interfaces so he won't be used. We should remove him from + * the target array here, but cleaning up a target is + * involved. PUNT! + */ memset(ht->ifs, 0, sizeof ht->ifs); d->htgt = NULL; return 1; @@ -477,13 +565,15 @@ ata_scnt(unsigned char *packet) { static void rexmit_timer(ulong vp) { - struct sk_buff_head queue; struct aoedev *d; struct aoetgt *t, **tt, **te; struct aoeif *ifp; - struct frame *f, *e; + struct frame *f; + struct list_head *head, *pos, *nx; + LIST_HEAD(flist); register long timeout; ulong flags, n; + int i; d = (struct aoedev *) vp; @@ -497,58 +587,22 @@ rexmit_timer(ulong vp) spin_unlock_irqrestore(&d->lock, flags); return; } - tt = d->targets; - te = tt + NTARGETS; - for (; tt < te && *tt; tt++) { - t = *tt; - f = t->frames; - e = f + t->nframes; - for (; f < e; f++) { - if (f->tag == FREETAG - || tsince(f->tag) < timeout) - continue; - n = f->waited += timeout; - n /= HZ; - if (n > aoe_deadsecs) { - /* waited too long. device failure. */ - aoedev_downdev(d); - break; - } - - if (n > HELPWAIT /* see if another target can help */ - && (tt != d->targets || d->targets[1])) - d->htgt = tt; - - if (t->nout == t->maxout) { - if (t->maxout > 1) - t->maxout--; - t->lastwadj = jiffies; - } - - ifp = getif(t, f->skb->dev); - if (ifp && ++ifp->lost > (t->nframes << 1) - && (ifp != t->ifs || t->ifs[1].nd)) { - ejectif(t, ifp); - ifp = NULL; - } - if (ata_scnt(skb_mac_header(f->skb)) > DEFAULTBCNT / 512 - && ifp && ++ifp->lostjumbo > (t->nframes << 1) - && ifp->maxbcnt != DEFAULTBCNT) { - printk(KERN_INFO - "aoe: e%ld.%d: " - "too many lost jumbo on " - "%s:%pm - " - "falling back to %d frames.\n", - d->aoemajor, d->aoeminor, - ifp->nd->name, t->addr, - DEFAULTBCNT); - ifp->maxbcnt = 0; - } - resend(d, t, f); + /* collect all frames to rexmit into flist */ + for (i = 0; i < NFACTIVE; i++) { + head = &d->factive[i]; + list_for_each_safe(pos, nx, head) { + f = list_entry(pos, struct frame, head); + if (tsince(f->tag) < timeout) + break; /* end of expired frames */ + /* move to flist for later processing */ + list_move_tail(pos, &flist); } - - /* window check */ + } + /* window check */ + tt = d->targets; + te = tt + d->ntargets; + for (; tt < te && (t = *tt); tt++) { if (t->nout == t->maxout && t->maxout < t->nframes && (jiffies - t->lastwadj)/HZ > 10) { @@ -557,45 +611,173 @@ rexmit_timer(ulong vp) } } - if (!skb_queue_empty(&d->sendq)) { + if (!list_empty(&flist)) { /* retransmissions necessary */ n = d->rttavg <<= 1; if (n > MAXTIMER) d->rttavg = MAXTIMER; } - if (d->flags & DEVFL_KICKME || d->htgt) { - d->flags &= ~DEVFL_KICKME; - aoecmd_work(d); + /* process expired frames */ + while (!list_empty(&flist)) { + pos = flist.next; + f = list_entry(pos, struct frame, head); + n = f->waited += timeout; + n /= HZ; + if (n > aoe_deadsecs) { + /* Waited too long. Device failure. + * Hang all frames on first hash bucket for downdev + * to clean up. + */ + list_splice(&flist, &d->factive[0]); + aoedev_downdev(d); + break; + } + list_del(pos); + + t = f->t; + if (n > aoe_deadsecs/2) + d->htgt = t; /* see if another target can help */ + + if (t->nout == t->maxout) { + if (t->maxout > 1) + t->maxout--; + t->lastwadj = jiffies; + } + + ifp = getif(t, f->skb->dev); + if (ifp && ++ifp->lost > (t->nframes << 1) + && (ifp != t->ifs || t->ifs[1].nd)) { + ejectif(t, ifp); + ifp = NULL; + } + resend(d, f); } - __skb_queue_head_init(&queue); - skb_queue_splice_init(&d->sendq, &queue); + if ((d->flags & DEVFL_KICKME || d->htgt) && d->blkq) { + d->flags &= ~DEVFL_KICKME; + d->blkq->request_fn(d->blkq); + } d->timer.expires = jiffies + TIMERTICK; add_timer(&d->timer); spin_unlock_irqrestore(&d->lock, flags); +} - aoenet_xmit(&queue); +static unsigned long +rqbiocnt(struct request *r) +{ + struct bio *bio; + unsigned long n = 0; + + __rq_for_each_bio(bio, r) + n++; + return n; +} + +/* This can be removed if we are certain that no users of the block + * layer will ever use zero-count pages in bios. Otherwise we have to + * protect against the put_page sometimes done by the network layer. + * + * See http://oss.sgi.com/archives/xfs/2007-01/msg00594.html for + * discussion. + * + * We cannot use get_page in the workaround, because it insists on a + * positive page count as a precondition. So we use _count directly. + */ +static void +bio_pageinc(struct bio *bio) +{ + struct bio_vec *bv; + struct page *page; + int i; + + bio_for_each_segment(bv, bio, i) { + page = bv->bv_page; + /* Non-zero page count for non-head members of + * compound pages is no longer allowed by the kernel, + * but this has never been seen here. + */ + if (unlikely(PageCompound(page))) + if (compound_trans_head(page) != page) { + pr_crit("page tail used for block I/O\n"); + BUG(); + } + atomic_inc(&page->_count); + } +} + +static void +bio_pagedec(struct bio *bio) +{ + struct bio_vec *bv; + int i; + + bio_for_each_segment(bv, bio, i) + atomic_dec(&bv->bv_page->_count); +} + +static void +bufinit(struct buf *buf, struct request *rq, struct bio *bio) +{ + struct bio_vec *bv; + + memset(buf, 0, sizeof(*buf)); + buf->rq = rq; + buf->bio = bio; + buf->resid = bio->bi_size; + buf->sector = bio->bi_sector; + bio_pageinc(bio); + buf->bv = bv = &bio->bi_io_vec[bio->bi_idx]; + buf->bv_resid = bv->bv_len; + WARN_ON(buf->bv_resid == 0); +} + +static struct buf * +nextbuf(struct aoedev *d) +{ + struct request *rq; + struct request_queue *q; + struct buf *buf; + struct bio *bio; + + q = d->blkq; + if (q == NULL) + return NULL; /* initializing */ + if (d->ip.buf) + return d->ip.buf; + rq = d->ip.rq; + if (rq == NULL) { + rq = blk_peek_request(q); + if (rq == NULL) + return NULL; + blk_start_request(rq); + d->ip.rq = rq; + d->ip.nxbio = rq->bio; + rq->special = (void *) rqbiocnt(rq); + } + buf = mempool_alloc(d->bufpool, GFP_ATOMIC); + if (buf == NULL) { + pr_err("aoe: nextbuf: unable to mempool_alloc!\n"); + return NULL; + } + bio = d->ip.nxbio; + bufinit(buf, rq, bio); + bio = bio->bi_next; + d->ip.nxbio = bio; + if (bio == NULL) + d->ip.rq = NULL; + return d->ip.buf = buf; } /* enters with d->lock held */ void aoecmd_work(struct aoedev *d) { - struct buf *buf; -loop: if (d->htgt && !sthtith(d)) return; - if (d->inprocess == NULL) { - if (list_empty(&d->bufq)) - return; - buf = container_of(d->bufq.next, struct buf, bufs); - list_del(d->bufq.next); - d->inprocess = buf; - } - if (aoecmd_ata_rw(d)) - goto loop; + while (aoecmd_ata_rw(d)) + ; } /* this function performs work that has been deferred until sleeping is OK @@ -604,28 +786,25 @@ void aoecmd_sleepwork(struct work_struct *work) { struct aoedev *d = container_of(work, struct aoedev, work); + struct block_device *bd; + u64 ssize; if (d->flags & DEVFL_GDALLOC) aoeblk_gdalloc(d); if (d->flags & DEVFL_NEWSIZE) { - struct block_device *bd; - unsigned long flags; - u64 ssize; - ssize = get_capacity(d->gd); bd = bdget_disk(d->gd, 0); - if (bd) { mutex_lock(&bd->bd_inode->i_mutex); i_size_write(bd->bd_inode, (loff_t)ssize<<9); mutex_unlock(&bd->bd_inode->i_mutex); bdput(bd); } - spin_lock_irqsave(&d->lock, flags); + spin_lock_irq(&d->lock); d->flags |= DEVFL_UP; d->flags &= ~DEVFL_NEWSIZE; - spin_unlock_irqrestore(&d->lock, flags); + spin_unlock_irq(&d->lock); } } @@ -718,163 +897,299 @@ gettgt(struct aoedev *d, char *addr) return NULL; } -static inline void -diskstats(struct gendisk *disk, struct bio *bio, ulong duration, sector_t sector) +static void +bvcpy(struct bio_vec *bv, ulong off, struct sk_buff *skb, long cnt) +{ + ulong fcnt; + char *p; + int soff = 0; +loop: + fcnt = bv->bv_len - (off - bv->bv_offset); + if (fcnt > cnt) + fcnt = cnt; + p = page_address(bv->bv_page) + off; + skb_copy_bits(skb, soff, p, fcnt); + soff += fcnt; + cnt -= fcnt; + if (cnt <= 0) + return; + bv++; + off = bv->bv_offset; + goto loop; +} + +void +aoe_end_request(struct aoedev *d, struct request *rq, int fastfail) +{ + struct bio *bio; + int bok; + struct request_queue *q; + + q = d->blkq; + if (rq == d->ip.rq) + d->ip.rq = NULL; + do { + bio = rq->bio; + bok = !fastfail && test_bit(BIO_UPTODATE, &bio->bi_flags); + } while (__blk_end_request(rq, bok ? 0 : -EIO, bio->bi_size)); + + /* cf. http://lkml.org/lkml/2006/10/31/28 */ + if (!fastfail) + q->request_fn(q); +} + +static void +aoe_end_buf(struct aoedev *d, struct buf *buf) +{ + struct request *rq; + unsigned long n; + + if (buf == d->ip.buf) + d->ip.buf = NULL; + rq = buf->rq; + bio_pagedec(buf->bio); + mempool_free(buf, d->bufpool); + n = (unsigned long) rq->special; + rq->special = (void *) --n; + if (n == 0) + aoe_end_request(d, rq, 0); +} + +static void +ktiocomplete(struct frame *f) { - unsigned long n_sect = bio->bi_size >> 9; - const int rw = bio_data_dir(bio); - struct hd_struct *part; - int cpu; + struct aoe_hdr *hin, *hout; + struct aoe_atahdr *ahin, *ahout; + struct buf *buf; + struct sk_buff *skb; + struct aoetgt *t; + struct aoeif *ifp; + struct aoedev *d; + long n; + + if (f == NULL) + return; + + t = f->t; + d = t->d; + + hout = (struct aoe_hdr *) skb_mac_header(f->skb); + ahout = (struct aoe_atahdr *) (hout+1); + buf = f->buf; + skb = f->r_skb; + if (skb == NULL) + goto noskb; /* just fail the buf. */ + + hin = (struct aoe_hdr *) skb->data; + skb_pull(skb, sizeof(*hin)); + ahin = (struct aoe_atahdr *) skb->data; + skb_pull(skb, sizeof(*ahin)); + if (ahin->cmdstat & 0xa9) { /* these bits cleared on success */ + pr_err("aoe: ata error cmd=%2.2Xh stat=%2.2Xh from e%ld.%d\n", + ahout->cmdstat, ahin->cmdstat, + d->aoemajor, d->aoeminor); +noskb: if (buf) + clear_bit(BIO_UPTODATE, &buf->bio->bi_flags); + goto badrsp; + } - cpu = part_stat_lock(); - part = disk_map_sector_rcu(disk, sector); + n = ahout->scnt << 9; + switch (ahout->cmdstat) { + case ATA_CMD_PIO_READ: + case ATA_CMD_PIO_READ_EXT: + if (skb->len < n) { + pr_err("aoe: runt data size in read. skb->len=%d need=%ld\n", + skb->len, n); + clear_bit(BIO_UPTODATE, &buf->bio->bi_flags); + break; + } + bvcpy(f->bv, f->bv_off, skb, n); + case ATA_CMD_PIO_WRITE: + case ATA_CMD_PIO_WRITE_EXT: + spin_lock_irq(&d->lock); + ifp = getif(t, skb->dev); + if (ifp) + ifp->lost = 0; + if (d->htgt == t) /* I'll help myself, thank you. */ + d->htgt = NULL; + spin_unlock_irq(&d->lock); + break; + case ATA_CMD_ID_ATA: + if (skb->len < 512) { + pr_info("aoe: runt data size in ataid. skb->len=%d\n", + skb->len); + break; + } + if (skb_linearize(skb)) + break; + spin_lock_irq(&d->lock); + ataid_complete(d, t, skb->data); + spin_unlock_irq(&d->lock); + break; + default: + pr_info("aoe: unrecognized ata command %2.2Xh for %d.%d\n", + ahout->cmdstat, + be16_to_cpu(get_unaligned(&hin->major)), + hin->minor); + } +badrsp: + spin_lock_irq(&d->lock); + + aoe_freetframe(f); + + if (buf && --buf->nframesout == 0 && buf->resid == 0) + aoe_end_buf(d, buf); + + aoecmd_work(d); + + spin_unlock_irq(&d->lock); + aoedev_put(d); + dev_kfree_skb(skb); +} + +/* Enters with iocq.lock held. + * Returns true iff responses needing processing remain. + */ +static int +ktio(void) +{ + struct frame *f; + struct list_head *pos; + int i; - part_stat_inc(cpu, part, ios[rw]); - part_stat_add(cpu, part, ticks[rw], duration); - part_stat_add(cpu, part, sectors[rw], n_sect); - part_stat_add(cpu, part, io_ticks, duration); + for (i = 0; ; ++i) { + if (i == MAXIOC) + return 1; + if (list_empty(&iocq.head)) + return 0; + pos = iocq.head.next; + list_del(pos); + spin_unlock_irq(&iocq.lock); + f = list_entry(pos, struct frame, head); + ktiocomplete(f); + spin_lock_irq(&iocq.lock); + } +} - part_stat_unlock(); +static int +kthread(void *vp) +{ + struct ktstate *k; + DECLARE_WAITQUEUE(wait, current); + int more; + + k = vp; + current->flags |= PF_NOFREEZE; + set_user_nice(current, -10); + complete(&k->rendez); /* tell spawner we're running */ + do { + spin_lock_irq(k->lock); + more = k->fn(); + if (!more) { + add_wait_queue(k->waitq, &wait); + __set_current_state(TASK_INTERRUPTIBLE); + } + spin_unlock_irq(k->lock); + if (!more) { + schedule(); + remove_wait_queue(k->waitq, &wait); + } else + cond_resched(); + } while (!kthread_should_stop()); + complete(&k->rendez); /* tell spawner we're stopping */ + return 0; } void +aoe_ktstop(struct ktstate *k) +{ + kthread_stop(k->task); + wait_for_completion(&k->rendez); +} + +int +aoe_ktstart(struct ktstate *k) +{ + struct task_struct *task; + + init_completion(&k->rendez); + task = kthread_run(kthread, k, k->name); + if (task == NULL || IS_ERR(task)) + return -ENOMEM; + k->task = task; + wait_for_completion(&k->rendez); /* allow kthread to start */ + init_completion(&k->rendez); /* for waiting for exit later */ + return 0; +} + +/* pass it off to kthreads for processing */ +static void +ktcomplete(struct frame *f, struct sk_buff *skb) +{ + ulong flags; + + f->r_skb = skb; + spin_lock_irqsave(&iocq.lock, flags); + list_add_tail(&f->head, &iocq.head); + spin_unlock_irqrestore(&iocq.lock, flags); + wake_up(&ktiowq); +} + +struct sk_buff * aoecmd_ata_rsp(struct sk_buff *skb) { - struct sk_buff_head queue; struct aoedev *d; - struct aoe_hdr *hin, *hout; - struct aoe_atahdr *ahin, *ahout; + struct aoe_hdr *h; struct frame *f; - struct buf *buf; struct aoetgt *t; - struct aoeif *ifp; - register long n; + u32 n; ulong flags; char ebuf[128]; u16 aoemajor; - hin = (struct aoe_hdr *) skb_mac_header(skb); - aoemajor = get_unaligned_be16(&hin->major); - d = aoedev_by_aoeaddr(aoemajor, hin->minor); + h = (struct aoe_hdr *) skb->data; + aoemajor = be16_to_cpu(get_unaligned(&h->major)); + d = aoedev_by_aoeaddr(aoemajor, h->minor, 0); if (d == NULL) { snprintf(ebuf, sizeof ebuf, "aoecmd_ata_rsp: ata response " "for unknown device %d.%d\n", - aoemajor, hin->minor); + aoemajor, h->minor); aoechr_error(ebuf); - return; + return skb; } spin_lock_irqsave(&d->lock, flags); - n = get_unaligned_be32(&hin->tag); - t = gettgt(d, hin->src); - if (t == NULL) { - printk(KERN_INFO "aoe: can't find target e%ld.%d:%pm\n", - d->aoemajor, d->aoeminor, hin->src); - spin_unlock_irqrestore(&d->lock, flags); - return; - } - f = getframe(t, n); + n = be32_to_cpu(get_unaligned(&h->tag)); + f = getframe(d, n); if (f == NULL) { calc_rttavg(d, -tsince(n)); spin_unlock_irqrestore(&d->lock, flags); + aoedev_put(d); snprintf(ebuf, sizeof ebuf, "%15s e%d.%d tag=%08x@%08lx\n", "unexpected rsp", - get_unaligned_be16(&hin->major), - hin->minor, - get_unaligned_be32(&hin->tag), + get_unaligned_be16(&h->major), + h->minor, + get_unaligned_be32(&h->tag), jiffies); aoechr_error(ebuf); - return; + return skb; } - + t = f->t; calc_rttavg(d, tsince(f->tag)); - - ahin = (struct aoe_atahdr *) (hin+1); - hout = (struct aoe_hdr *) skb_mac_header(f->skb); - ahout = (struct aoe_atahdr *) (hout+1); - buf = f->buf; - - if (ahin->cmdstat & 0xa9) { /* these bits cleared on success */ - printk(KERN_ERR - "aoe: ata error cmd=%2.2Xh stat=%2.2Xh from e%ld.%d\n", - ahout->cmdstat, ahin->cmdstat, - d->aoemajor, d->aoeminor); - if (buf) - buf->flags |= BUFFL_FAIL; - } else { - if (d->htgt && t == *d->htgt) /* I'll help myself, thank you. */ - d->htgt = NULL; - n = ahout->scnt << 9; - switch (ahout->cmdstat) { - case ATA_CMD_PIO_READ: - case ATA_CMD_PIO_READ_EXT: - if (skb->len - sizeof *hin - sizeof *ahin < n) { - printk(KERN_ERR - "aoe: %s. skb->len=%d need=%ld\n", - "runt data size in read", skb->len, n); - /* fail frame f? just returning will rexmit. */ - spin_unlock_irqrestore(&d->lock, flags); - return; - } - memcpy(f->bufaddr, ahin+1, n); - case ATA_CMD_PIO_WRITE: - case ATA_CMD_PIO_WRITE_EXT: - ifp = getif(t, skb->dev); - if (ifp) { - ifp->lost = 0; - if (n > DEFAULTBCNT) - ifp->lostjumbo = 0; - } - if (f->bcnt -= n) { - f->lba += n >> 9; - f->bufaddr += n; - resend(d, t, f); - goto xmit; - } - break; - case ATA_CMD_ID_ATA: - if (skb->len - sizeof *hin - sizeof *ahin < 512) { - printk(KERN_INFO - "aoe: runt data size in ataid. skb->len=%d\n", - skb->len); - spin_unlock_irqrestore(&d->lock, flags); - return; - } - ataid_complete(d, t, (char *) (ahin+1)); - break; - default: - printk(KERN_INFO - "aoe: unrecognized ata command %2.2Xh for %d.%d\n", - ahout->cmdstat, - get_unaligned_be16(&hin->major), - hin->minor); - } - } - - if (buf && --buf->nframesout == 0 && buf->resid == 0) { - diskstats(d->gd, buf->bio, jiffies - buf->stime, buf->sector); - if (buf->flags & BUFFL_FAIL) - bio_endio(buf->bio, -EIO); - else { - bio_flush_dcache_pages(buf->bio); - bio_endio(buf->bio, 0); - } - mempool_free(buf, d->bufpool); - } - - f->buf = NULL; - f->tag = FREETAG; t->nout--; - aoecmd_work(d); -xmit: - __skb_queue_head_init(&queue); - skb_queue_splice_init(&d->sendq, &queue); spin_unlock_irqrestore(&d->lock, flags); - aoenet_xmit(&queue); + + ktcomplete(f, skb); + + /* + * Note here that we do not perform an aoedev_put, as we are + * leaving this reference for the ktio to release. + */ + return NULL; } void @@ -896,7 +1211,7 @@ aoecmd_ata_id(struct aoedev *d) struct sk_buff *skb; struct aoetgt *t; - f = freeframe(d); + f = newframe(d); if (f == NULL) return NULL; @@ -909,6 +1224,7 @@ aoecmd_ata_id(struct aoedev *d) skb_put(skb, sizeof *h + sizeof *ah); memset(h, 0, skb->len); f->tag = aoehdr_atainit(d, t, h); + fhash(f); t->nout++; f->waited = 0; @@ -929,7 +1245,6 @@ static struct aoetgt * addtgt(struct aoedev *d, char *addr, ulong nframes) { struct aoetgt *t, **tt, **te; - struct frame *f, *e; tt = d->targets; te = tt + NTARGETS; @@ -941,26 +1256,73 @@ addtgt(struct aoedev *d, char *addr, ulong nframes) "aoe: device addtgt failure; too many targets\n"); return NULL; } - t = kcalloc(1, sizeof *t, GFP_ATOMIC); - f = kcalloc(nframes, sizeof *f, GFP_ATOMIC); - if (!t || !f) { - kfree(f); - kfree(t); + t = kzalloc(sizeof(*t), GFP_ATOMIC); + if (!t) { printk(KERN_INFO "aoe: cannot allocate memory to add target\n"); return NULL; } + d->ntargets++; t->nframes = nframes; - t->frames = f; - e = f + nframes; - for (; f < e; f++) - f->tag = FREETAG; + t->d = d; memcpy(t->addr, addr, sizeof t->addr); t->ifp = t->ifs; t->maxout = t->nframes; + INIT_LIST_HEAD(&t->ffree); return *tt = t; } +static void +setdbcnt(struct aoedev *d) +{ + struct aoetgt **t, **e; + int bcnt = 0; + + t = d->targets; + e = t + NTARGETS; + for (; t < e && *t; t++) + if (bcnt == 0 || bcnt > (*t)->minbcnt) + bcnt = (*t)->minbcnt; + if (bcnt != d->maxbcnt) { + d->maxbcnt = bcnt; + pr_info("aoe: e%ld.%d: setting %d byte data frames\n", + d->aoemajor, d->aoeminor, bcnt); + } +} + +static void +setifbcnt(struct aoetgt *t, struct net_device *nd, int bcnt) +{ + struct aoedev *d; + struct aoeif *p, *e; + int minbcnt; + + d = t->d; + minbcnt = bcnt; + p = t->ifs; + e = p + NAOEIFS; + for (; p < e; p++) { + if (p->nd == NULL) + break; /* end of the valid interfaces */ + if (p->nd == nd) { + p->bcnt = bcnt; /* we're updating */ + nd = NULL; + } else if (minbcnt > p->bcnt) + minbcnt = p->bcnt; /* find the min interface */ + } + if (nd) { + if (p == e) { + pr_err("aoe: device setifbcnt failure; too many interfaces.\n"); + return; + } + dev_hold(nd); + p->nd = nd; + p->bcnt = bcnt; + } + t->minbcnt = minbcnt; + setdbcnt(d); +} + void aoecmd_cfg_rsp(struct sk_buff *skb) { @@ -968,11 +1330,12 @@ aoecmd_cfg_rsp(struct sk_buff *skb) struct aoe_hdr *h; struct aoe_cfghdr *ch; struct aoetgt *t; - struct aoeif *ifp; - ulong flags, sysminor, aoemajor; + ulong flags, aoemajor; struct sk_buff *sl; + struct sk_buff_head queue; u16 n; + sl = NULL; h = (struct aoe_hdr *) skb_mac_header(skb); ch = (struct aoe_cfghdr *) (h+1); @@ -986,10 +1349,13 @@ aoecmd_cfg_rsp(struct sk_buff *skb) "Check shelf dip switches.\n"); return; } - - sysminor = SYSMINOR(aoemajor, h->minor); - if (sysminor * AOE_PARTITIONS + AOE_PARTITIONS > MINORMASK) { - printk(KERN_INFO "aoe: e%ld.%d: minor number too large\n", + if (aoemajor == 0xffff) { + pr_info("aoe: e%ld.%d: broadcast shelf number invalid\n", + aoemajor, (int) h->minor); + return; + } + if (h->minor == 0xff) { + pr_info("aoe: e%ld.%d: broadcast slot number invalid\n", aoemajor, (int) h->minor); return; } @@ -998,9 +1364,9 @@ aoecmd_cfg_rsp(struct sk_buff *skb) if (n > aoe_maxout) /* keep it reasonable */ n = aoe_maxout; - d = aoedev_by_sysminor_m(sysminor); + d = aoedev_by_aoeaddr(aoemajor, h->minor, 1); if (d == NULL) { - printk(KERN_INFO "aoe: device sysminor_m failure\n"); + pr_info("aoe: device allocation failure\n"); return; } @@ -1009,52 +1375,26 @@ aoecmd_cfg_rsp(struct sk_buff *skb) t = gettgt(d, h->src); if (!t) { t = addtgt(d, h->src, n); - if (!t) { - spin_unlock_irqrestore(&d->lock, flags); - return; - } - } - ifp = getif(t, skb->dev); - if (!ifp) { - ifp = addif(t, skb->dev); - if (!ifp) { - printk(KERN_INFO - "aoe: device addif failure; " - "too many interfaces?\n"); - spin_unlock_irqrestore(&d->lock, flags); - return; - } - } - if (ifp->maxbcnt) { - n = ifp->nd->mtu; - n -= sizeof (struct aoe_hdr) + sizeof (struct aoe_atahdr); - n /= 512; - if (n > ch->scnt) - n = ch->scnt; - n = n ? n * 512 : DEFAULTBCNT; - if (n != ifp->maxbcnt) { - printk(KERN_INFO - "aoe: e%ld.%d: setting %d%s%s:%pm\n", - d->aoemajor, d->aoeminor, n, - " byte data frames on ", ifp->nd->name, - t->addr); - ifp->maxbcnt = n; - } + if (!t) + goto bail; } + n = skb->dev->mtu; + n -= sizeof(struct aoe_hdr) + sizeof(struct aoe_atahdr); + n /= 512; + if (n > ch->scnt) + n = ch->scnt; + n = n ? n * 512 : DEFAULTBCNT; + setifbcnt(t, skb->dev, n); /* don't change users' perspective */ - if (d->nopen) { - spin_unlock_irqrestore(&d->lock, flags); - return; + if (d->nopen == 0) { + d->fw_ver = be16_to_cpu(ch->fwver); + sl = aoecmd_ata_id(d); } - d->fw_ver = be16_to_cpu(ch->fwver); - - sl = aoecmd_ata_id(d); - +bail: spin_unlock_irqrestore(&d->lock, flags); - + aoedev_put(d); if (sl) { - struct sk_buff_head queue; __skb_queue_head_init(&queue); __skb_queue_tail(&queue, sl); aoenet_xmit(&queue); @@ -1065,20 +1405,74 @@ void aoecmd_cleanslate(struct aoedev *d) { struct aoetgt **t, **te; - struct aoeif *p, *e; d->mintimer = MINTIMER; + d->maxbcnt = 0; t = d->targets; te = t + NTARGETS; - for (; t < te && *t; t++) { + for (; t < te && *t; t++) (*t)->maxout = (*t)->nframes; - p = (*t)->ifs; - e = p + NAOEIFS; - for (; p < e; p++) { - p->lostjumbo = 0; - p->lost = 0; - p->maxbcnt = DEFAULTBCNT; +} + +void +aoe_failbuf(struct aoedev *d, struct buf *buf) +{ + if (buf == NULL) + return; + buf->resid = 0; + clear_bit(BIO_UPTODATE, &buf->bio->bi_flags); + if (buf->nframesout == 0) + aoe_end_buf(d, buf); +} + +void +aoe_flush_iocq(void) +{ + struct frame *f; + struct aoedev *d; + LIST_HEAD(flist); + struct list_head *pos; + struct sk_buff *skb; + ulong flags; + + spin_lock_irqsave(&iocq.lock, flags); + list_splice_init(&iocq.head, &flist); + spin_unlock_irqrestore(&iocq.lock, flags); + while (!list_empty(&flist)) { + pos = flist.next; + list_del(pos); + f = list_entry(pos, struct frame, head); + d = f->t->d; + skb = f->r_skb; + spin_lock_irqsave(&d->lock, flags); + if (f->buf) { + f->buf->nframesout--; + aoe_failbuf(d, f->buf); } + aoe_freetframe(f); + spin_unlock_irqrestore(&d->lock, flags); + dev_kfree_skb(skb); + aoedev_put(d); } } + +int __init +aoecmd_init(void) +{ + INIT_LIST_HEAD(&iocq.head); + spin_lock_init(&iocq.lock); + init_waitqueue_head(&ktiowq); + kts.name = "aoe_ktio"; + kts.fn = ktio; + kts.waitq = &ktiowq; + kts.lock = &iocq.lock; + return aoe_ktstart(&kts); +} + +void +aoecmd_exit(void) +{ + aoe_ktstop(&kts); + aoe_flush_iocq(); +} diff --git a/drivers/block/aoe/aoedev.c b/drivers/block/aoe/aoedev.c index 6b5110a4745..90e5b537f94 100644 --- a/drivers/block/aoe/aoedev.c +++ b/drivers/block/aoe/aoedev.c @@ -1,4 +1,4 @@ -/* Copyright (c) 2007 Coraid, Inc. See COPYING for GPL terms. */ +/* Copyright (c) 2012 Coraid, Inc. See COPYING for GPL terms. */ /* * aoedev.c * AoE device utility functions; maintains device list. @@ -9,6 +9,9 @@ #include <linux/netdevice.h> #include <linux/delay.h> #include <linux/slab.h> +#include <linux/bitmap.h> +#include <linux/kdev_t.h> +#include <linux/moduleparam.h> #include "aoe.h" static void dummy_timer(ulong); @@ -16,23 +19,121 @@ static void aoedev_freedev(struct aoedev *); static void freetgt(struct aoedev *d, struct aoetgt *t); static void skbpoolfree(struct aoedev *d); +static int aoe_dyndevs = 1; +module_param(aoe_dyndevs, int, 0644); +MODULE_PARM_DESC(aoe_dyndevs, "Use dynamic minor numbers for devices."); + static struct aoedev *devlist; static DEFINE_SPINLOCK(devlist_lock); -struct aoedev * -aoedev_by_aoeaddr(int maj, int min) +/* Because some systems will have one, many, or no + * - partitions, + * - slots per shelf, + * - or shelves, + * we need some flexibility in the way the minor numbers + * are allocated. So they are dynamic. + */ +#define N_DEVS ((1U<<MINORBITS)/AOE_PARTITIONS) + +static DEFINE_SPINLOCK(used_minors_lock); +static DECLARE_BITMAP(used_minors, N_DEVS); + +static int +minor_get_dyn(ulong *sysminor) { - struct aoedev *d; ulong flags; + ulong n; + int error = 0; + + spin_lock_irqsave(&used_minors_lock, flags); + n = find_first_zero_bit(used_minors, N_DEVS); + if (n < N_DEVS) + set_bit(n, used_minors); + else + error = -1; + spin_unlock_irqrestore(&used_minors_lock, flags); + + *sysminor = n * AOE_PARTITIONS; + return error; +} - spin_lock_irqsave(&devlist_lock, flags); +static int +minor_get_static(ulong *sysminor, ulong aoemaj, int aoemin) +{ + ulong flags; + ulong n; + int error = 0; + enum { + /* for backwards compatibility when !aoe_dyndevs, + * a static number of supported slots per shelf */ + NPERSHELF = 16, + }; + + n = aoemaj * NPERSHELF + aoemin; + if (aoemin >= NPERSHELF || n >= N_DEVS) { + pr_err("aoe: %s with e%ld.%d\n", + "cannot use static minor device numbers", + aoemaj, aoemin); + error = -1; + } else { + spin_lock_irqsave(&used_minors_lock, flags); + if (test_bit(n, used_minors)) { + pr_err("aoe: %s %lu\n", + "existing device already has static minor number", + n); + error = -1; + } else + set_bit(n, used_minors); + spin_unlock_irqrestore(&used_minors_lock, flags); + } - for (d=devlist; d; d=d->next) - if (d->aoemajor == maj && d->aoeminor == min) - break; + *sysminor = n; + return error; +} + +static int +minor_get(ulong *sysminor, ulong aoemaj, int aoemin) +{ + if (aoe_dyndevs) + return minor_get_dyn(sysminor); + else + return minor_get_static(sysminor, aoemaj, aoemin); +} + +static void +minor_free(ulong minor) +{ + ulong flags; + + minor /= AOE_PARTITIONS; + BUG_ON(minor >= N_DEVS); + + spin_lock_irqsave(&used_minors_lock, flags); + BUG_ON(!test_bit(minor, used_minors)); + clear_bit(minor, used_minors); + spin_unlock_irqrestore(&used_minors_lock, flags); +} + +/* + * Users who grab a pointer to the device with aoedev_by_aoeaddr + * automatically get a reference count and must be responsible + * for performing a aoedev_put. With the addition of async + * kthread processing I'm no longer confident that we can + * guarantee consistency in the face of device flushes. + * + * For the time being, we only bother to add extra references for + * frames sitting on the iocq. When the kthreads finish processing + * these frames, they will aoedev_put the device. + */ + +void +aoedev_put(struct aoedev *d) +{ + ulong flags; + spin_lock_irqsave(&devlist_lock, flags); + d->ref--; spin_unlock_irqrestore(&devlist_lock, flags); - return d; } static void @@ -47,54 +148,74 @@ dummy_timer(ulong vp) add_timer(&d->timer); } +static void +aoe_failip(struct aoedev *d) +{ + struct request *rq; + struct bio *bio; + unsigned long n; + + aoe_failbuf(d, d->ip.buf); + + rq = d->ip.rq; + if (rq == NULL) + return; + while ((bio = d->ip.nxbio)) { + clear_bit(BIO_UPTODATE, &bio->bi_flags); + d->ip.nxbio = bio->bi_next; + n = (unsigned long) rq->special; + rq->special = (void *) --n; + } + if ((unsigned long) rq->special == 0) + aoe_end_request(d, rq, 0); +} + void aoedev_downdev(struct aoedev *d) { - struct aoetgt **t, **te; - struct frame *f, *e; - struct buf *buf; - struct bio *bio; + struct aoetgt *t, **tt, **te; + struct frame *f; + struct list_head *head, *pos, *nx; + struct request *rq; + int i; - t = d->targets; - te = t + NTARGETS; - for (; t < te && *t; t++) { - f = (*t)->frames; - e = f + (*t)->nframes; - for (; f < e; f->tag = FREETAG, f->buf = NULL, f++) { - if (f->tag == FREETAG || f->buf == NULL) - continue; - buf = f->buf; - bio = buf->bio; - if (--buf->nframesout == 0 - && buf != d->inprocess) { - mempool_free(buf, d->bufpool); - bio_endio(bio, -EIO); + d->flags &= ~DEVFL_UP; + + /* clean out active buffers */ + for (i = 0; i < NFACTIVE; i++) { + head = &d->factive[i]; + list_for_each_safe(pos, nx, head) { + f = list_entry(pos, struct frame, head); + list_del(pos); + if (f->buf) { + f->buf->nframesout--; + aoe_failbuf(d, f->buf); } + aoe_freetframe(f); } - (*t)->maxout = (*t)->nframes; - (*t)->nout = 0; } - buf = d->inprocess; - if (buf) { - bio = buf->bio; - mempool_free(buf, d->bufpool); - bio_endio(bio, -EIO); + /* reset window dressings */ + tt = d->targets; + te = tt + NTARGETS; + for (; tt < te && (t = *tt); tt++) { + t->maxout = t->nframes; + t->nout = 0; } - d->inprocess = NULL; + + /* clean out the in-process request (if any) */ + aoe_failip(d); d->htgt = NULL; - while (!list_empty(&d->bufq)) { - buf = container_of(d->bufq.next, struct buf, bufs); - list_del(d->bufq.next); - bio = buf->bio; - mempool_free(buf, d->bufpool); - bio_endio(bio, -EIO); + /* fast fail all pending I/O */ + if (d->blkq) { + while ((rq = blk_peek_request(d->blkq))) { + blk_start_request(rq); + aoe_end_request(d, rq, 1); + } } if (d->gd) set_capacity(d->gd, 0); - - d->flags &= ~DEVFL_UP; } static void @@ -107,6 +228,7 @@ aoedev_freedev(struct aoedev *d) aoedisk_rm_sysfs(d); del_gendisk(d->gd); put_disk(d->gd); + blk_cleanup_queue(d->blkq); } t = d->targets; e = t + NTARGETS; @@ -115,7 +237,7 @@ aoedev_freedev(struct aoedev *d) if (d->bufpool) mempool_destroy(d->bufpool); skbpoolfree(d); - blk_cleanup_queue(d->blkq); + minor_free(d->sysminor); kfree(d); } @@ -142,7 +264,8 @@ aoedev_flush(const char __user *str, size_t cnt) spin_lock(&d->lock); if ((!all && (d->flags & DEVFL_UP)) || (d->flags & (DEVFL_GDALLOC|DEVFL_NEWSIZE)) - || d->nopen) { + || d->nopen + || d->ref) { spin_unlock(&d->lock); dd = &d->next; continue; @@ -163,12 +286,15 @@ aoedev_flush(const char __user *str, size_t cnt) return 0; } -/* I'm not really sure that this is a realistic problem, but if the -network driver goes gonzo let's just leak memory after complaining. */ +/* This has been confirmed to occur once with Tms=3*1000 due to the + * driver changing link and not processing its transmit ring. The + * problem is hard enough to solve by returning an error that I'm + * still punting on "solving" this. + */ static void skbfree(struct sk_buff *skb) { - enum { Sms = 100, Tms = 3*1000}; + enum { Sms = 250, Tms = 30 * 1000}; int i = Tms / Sms; if (skb == NULL) @@ -182,6 +308,7 @@ skbfree(struct sk_buff *skb) "cannot free skb -- memory leaked."); return; } + skb->truesize -= skb->data_len; skb_shinfo(skb)->nr_frags = skb->data_len = 0; skb_trim(skb, 0); dev_kfree_skb(skb); @@ -198,26 +325,29 @@ skbpoolfree(struct aoedev *d) __skb_queue_head_init(&d->skbpool); } -/* find it or malloc it */ +/* find it or allocate it */ struct aoedev * -aoedev_by_sysminor_m(ulong sysminor) +aoedev_by_aoeaddr(ulong maj, int min, int do_alloc) { struct aoedev *d; + int i; ulong flags; + ulong sysminor; spin_lock_irqsave(&devlist_lock, flags); for (d=devlist; d; d=d->next) - if (d->sysminor == sysminor) + if (d->aoemajor == maj && d->aoeminor == min) { + d->ref++; break; - if (d) + } + if (d || !do_alloc || minor_get(&sysminor, maj, min) < 0) goto out; d = kcalloc(1, sizeof *d, GFP_ATOMIC); if (!d) goto out; INIT_WORK(&d->work, aoecmd_sleepwork); spin_lock_init(&d->lock); - skb_queue_head_init(&d->sendq); skb_queue_head_init(&d->skbpool); init_timer(&d->timer); d->timer.data = (ulong) d; @@ -226,10 +356,12 @@ aoedev_by_sysminor_m(ulong sysminor) add_timer(&d->timer); d->bufpool = NULL; /* defer to aoeblk_gdalloc */ d->tgt = d->targets; - INIT_LIST_HEAD(&d->bufq); + d->ref = 1; + for (i = 0; i < NFACTIVE; i++) + INIT_LIST_HEAD(&d->factive[i]); d->sysminor = sysminor; - d->aoemajor = AOEMAJOR(sysminor); - d->aoeminor = AOEMINOR(sysminor); + d->aoemajor = maj; + d->aoeminor = min; d->mintimer = MINTIMER; d->next = devlist; devlist = d; @@ -241,13 +373,23 @@ aoedev_by_sysminor_m(ulong sysminor) static void freetgt(struct aoedev *d, struct aoetgt *t) { - struct frame *f, *e; + struct frame *f; + struct list_head *pos, *nx, *head; + struct aoeif *ifp; - f = t->frames; - e = f + t->nframes; - for (; f < e; f++) + for (ifp = t->ifs; ifp < &t->ifs[NAOEIFS]; ++ifp) { + if (!ifp->nd) + break; + dev_put(ifp->nd); + } + + head = &t->ffree; + list_for_each_safe(pos, nx, head) { + list_del(pos); + f = list_entry(pos, struct frame, head); skbfree(f->skb); - kfree(t->frames); + kfree(f); + } kfree(t); } @@ -257,6 +399,7 @@ aoedev_exit(void) struct aoedev *d; ulong flags; + aoe_flush_iocq(); while ((d = devlist)) { devlist = d->next; diff --git a/drivers/block/aoe/aoemain.c b/drivers/block/aoe/aoemain.c index 7f83ad90e76..04793c2c701 100644 --- a/drivers/block/aoe/aoemain.c +++ b/drivers/block/aoe/aoemain.c @@ -1,4 +1,4 @@ -/* Copyright (c) 2007 Coraid, Inc. See COPYING for GPL terms. */ +/* Copyright (c) 2012 Coraid, Inc. See COPYING for GPL terms. */ /* * aoemain.c * Module initialization routines, discover timer @@ -61,6 +61,7 @@ aoe_exit(void) aoenet_exit(); unregister_blkdev(AOE_MAJOR, DEVICE_NAME); + aoecmd_exit(); aoechr_exit(); aoedev_exit(); aoeblk_exit(); /* free cache after de-allocating bufs */ @@ -83,17 +84,20 @@ aoe_init(void) ret = aoenet_init(); if (ret) goto net_fail; + ret = aoecmd_init(); + if (ret) + goto cmd_fail; ret = register_blkdev(AOE_MAJOR, DEVICE_NAME); if (ret < 0) { printk(KERN_ERR "aoe: can't register major\n"); goto blkreg_fail; } - printk(KERN_INFO "aoe: AoE v%s initialised.\n", VERSION); discover_timer(TINIT); return 0; - blkreg_fail: + aoecmd_exit(); + cmd_fail: aoenet_exit(); net_fail: aoeblk_exit(); diff --git a/drivers/block/aoe/aoenet.c b/drivers/block/aoe/aoenet.c index 4d3bc0d49df..162c6471275 100644 --- a/drivers/block/aoe/aoenet.c +++ b/drivers/block/aoe/aoenet.c @@ -1,4 +1,4 @@ -/* Copyright (c) 2007 Coraid, Inc. See COPYING for GPL terms. */ +/* Copyright (c) 2012 Coraid, Inc. See COPYING for GPL terms. */ /* * aoenet.c * Ethernet portion of AoE driver @@ -33,6 +33,9 @@ static char aoe_iflist[IFLISTSZ]; module_param_string(aoe_iflist, aoe_iflist, IFLISTSZ, 0600); MODULE_PARM_DESC(aoe_iflist, "aoe_iflist=\"dev1 [dev2 ...]\""); +static wait_queue_head_t txwq; +static struct ktstate kts; + #ifndef MODULE static int __init aoe_iflist_setup(char *str) { @@ -44,6 +47,23 @@ static int __init aoe_iflist_setup(char *str) __setup("aoe_iflist=", aoe_iflist_setup); #endif +static spinlock_t txlock; +static struct sk_buff_head skbtxq; + +/* enters with txlock held */ +static int +tx(void) +{ + struct sk_buff *skb; + + while ((skb = skb_dequeue(&skbtxq))) { + spin_unlock_irq(&txlock); + dev_queue_xmit(skb); + spin_lock_irq(&txlock); + } + return 0; +} + int is_aoe_netif(struct net_device *ifp) { @@ -88,10 +108,14 @@ void aoenet_xmit(struct sk_buff_head *queue) { struct sk_buff *skb, *tmp; + ulong flags; skb_queue_walk_safe(queue, skb, tmp) { __skb_unlink(skb, queue); - dev_queue_xmit(skb); + spin_lock_irqsave(&txlock, flags); + skb_queue_tail(&skbtxq, skb); + spin_unlock_irqrestore(&txlock, flags); + wake_up(&txwq); } } @@ -102,7 +126,9 @@ static int aoenet_rcv(struct sk_buff *skb, struct net_device *ifp, struct packet_type *pt, struct net_device *orig_dev) { struct aoe_hdr *h; + struct aoe_atahdr *ah; u32 n; + int sn; if (dev_net(ifp) != &init_net) goto exit; @@ -110,13 +136,16 @@ aoenet_rcv(struct sk_buff *skb, struct net_device *ifp, struct packet_type *pt, skb = skb_share_check(skb, GFP_ATOMIC); if (skb == NULL) return 0; - if (skb_linearize(skb)) - goto exit; if (!is_aoe_netif(ifp)) goto exit; skb_push(skb, ETH_HLEN); /* (1) */ - - h = (struct aoe_hdr *) skb_mac_header(skb); + sn = sizeof(*h) + sizeof(*ah); + if (skb->len >= sn) { + sn -= skb_headlen(skb); + if (sn > 0 && !__pskb_pull_tail(skb, sn)) + goto exit; + } + h = (struct aoe_hdr *) skb->data; n = get_unaligned_be32(&h->tag); if ((h->verfl & AOEFL_RSP) == 0 || (n & 1<<31)) goto exit; @@ -137,7 +166,8 @@ aoenet_rcv(struct sk_buff *skb, struct net_device *ifp, struct packet_type *pt, switch (h->cmd) { case AOECMD_ATA: - aoecmd_ata_rsp(skb); + /* ata_rsp may keep skb for later processing or give it back */ + skb = aoecmd_ata_rsp(skb); break; case AOECMD_CFG: aoecmd_cfg_rsp(skb); @@ -145,8 +175,12 @@ aoenet_rcv(struct sk_buff *skb, struct net_device *ifp, struct packet_type *pt, default: if (h->cmd >= AOECMD_VEND_MIN) break; /* don't complain about vendor commands */ - printk(KERN_INFO "aoe: unknown cmd %d\n", h->cmd); + pr_info("aoe: unknown AoE command type 0x%02x\n", h->cmd); + break; } + + if (!skb) + return 0; exit: dev_kfree_skb(skb); return 0; @@ -160,6 +194,15 @@ static struct packet_type aoe_pt __read_mostly = { int __init aoenet_init(void) { + skb_queue_head_init(&skbtxq); + init_waitqueue_head(&txwq); + spin_lock_init(&txlock); + kts.lock = &txlock; + kts.fn = tx; + kts.waitq = &txwq; + kts.name = "aoe_tx"; + if (aoe_ktstart(&kts)) + return -EAGAIN; dev_add_pack(&aoe_pt); return 0; } @@ -167,6 +210,8 @@ aoenet_init(void) void aoenet_exit(void) { + aoe_ktstop(&kts); + skb_queue_purge(&skbtxq); dev_remove_pack(&aoe_pt); } |