[PATCH 2.6.18-rc4] aoe [09/14]: zero copy write 2 of 2

From: Ed L. Cashin
Date: Wed Sep 20 2006 - 14:58:12 EST


Avoid memory copy on writes.
(This patch follows patch 4.)

Signed-off-by: "Ed L. Cashin" <ecashin@xxxxxxxxxx>
---

diff -upr 2.6.18-rc4-orig/drivers/block/aoe/aoe.h 2.6.18-rc4-aoe/drivers/block/aoe/aoe.h
--- 2.6.18-rc4-orig/drivers/block/aoe/aoe.h 2006-09-20 14:29:36.000000000 -0400
+++ 2.6.18-rc4-aoe/drivers/block/aoe/aoe.h 2006-09-20 14:29:36.000000000 -0400
@@ -84,6 +84,7 @@ enum {
DEVFL_PAUSE = (1<<5),
DEVFL_NEWSIZE = (1<<6), /* need to update dev size in block layer */
DEVFL_MAXBCNT = (1<<7), /* d->maxbcnt is not changeable */
+ DEVFL_KICKME = (1<<8),

BUFFL_FAIL = 1,
};
diff -upr 2.6.18-rc4-orig/drivers/block/aoe/aoecmd.c 2.6.18-rc4-aoe/drivers/block/aoe/aoecmd.c
--- 2.6.18-rc4-orig/drivers/block/aoe/aoecmd.c 2006-09-20 14:29:36.000000000 -0400
+++ 2.6.18-rc4-aoe/drivers/block/aoe/aoecmd.c 2006-09-20 14:29:36.000000000 -0400
@@ -120,7 +120,7 @@ aoecmd_ata_rw(struct aoedev *d, struct f
h = (struct aoe_hdr *) skb->mac.raw;
ah = (struct aoe_atahdr *) (h+1);
skb->len = sizeof *h + sizeof *ah;
- memset(h, 0, skb->len);
+ memset(h, 0, ETH_ZLEN);
f->tag = aoehdr_atainit(d, h);
f->waited = 0;
f->buf = buf;
@@ -143,8 +143,9 @@ aoecmd_ata_rw(struct aoedev *d, struct f
skb_fill_page_desc(skb, 0, virt_to_page(f->bufaddr),
offset_in_page(f->bufaddr), bcnt);
ah->aflags |= AOEAFL_WRITE;
+ skb->len += bcnt;
+ skb->data_len = bcnt;
} else {
- skb_shinfo(skb)->nr_frags = 0;
skb->len = ETH_ZLEN;
writebit = 0;
}
@@ -167,8 +168,9 @@ aoecmd_ata_rw(struct aoedev *d, struct f
}

skb->dev = d->ifp;
- skb_get(skb);
- skb->next = NULL;
+ skb = skb_clone(skb, GFP_ATOMIC);
+ if (skb == NULL)
+ return;
if (d->sendq_hd)
d->sendq_tl->next = skb;
else
@@ -224,6 +226,29 @@ aoecmd_cfg_pkts(ushort aoemajor, unsigne
return sl;
}

+static struct frame *
+freeframe(struct aoedev *d)
+{
+ struct frame *f, *e;
+ int n = 0;
+
+ f = d->frames;
+ e = f + d->nframes;
+ for (; f<e; f++) {
+ if (f->tag != FREETAG)
+ continue;
+ if (atomic_read(&skb_shinfo(f->skb)->dataref) == 1) {
+ skb_shinfo(f->skb)->nr_frags = f->skb->data_len = 0;
+ return f;
+ }
+ n++;
+ }
+ if (n == d->nframes) /* wait for network layer */
+ d->flags |= DEVFL_KICKME;
+
+ return NULL;
+}
+
/* enters with d->lock held */
void
aoecmd_work(struct aoedev *d)
@@ -239,7 +264,7 @@ aoecmd_work(struct aoedev *d)
}

loop:
- f = getframe(d, FREETAG);
+ f = freeframe(d);
if (f == NULL)
return;
if (d->inprocess == NULL) {
@@ -282,20 +307,25 @@ rexmit(struct aoedev *d, struct frame *f
n = DEFAULTBCNT / 512;
if (ah->scnt > n) {
ah->scnt = n;
- if (ah->aflags & AOEAFL_WRITE)
+ if (ah->aflags & AOEAFL_WRITE) {
skb_fill_page_desc(skb, 0, virt_to_page(f->bufaddr),
offset_in_page(f->bufaddr), DEFAULTBCNT);
+ skb->len = sizeof *h + sizeof *ah + DEFAULTBCNT;
+ skb->data_len = DEFAULTBCNT;
+ }
if (++d->lostjumbo > (d->nframes << 1))
if (d->maxbcnt != DEFAULTBCNT) {
- iprintk("too many lost jumbo - using 1KB frames.\n");
+ iprintk("e%ld.%ld: too many lost jumbo on %s - using 1KB frames.\n",
+ d->aoemajor, d->aoeminor, d->ifp->name);
d->maxbcnt = DEFAULTBCNT;
d->flags |= DEVFL_MAXBCNT;
}
}

skb->dev = d->ifp;
- skb_get(skb);
- skb->next = NULL;
+ skb = skb_clone(skb, GFP_ATOMIC);
+ if (skb == NULL)
+ return;
if (d->sendq_hd)
d->sendq_tl->next = skb;
else
@@ -350,6 +380,10 @@ rexmit_timer(ulong vp)
rexmit(d, f);
}
}
+ if (d->flags & DEVFL_KICKME) {
+ d->flags &= ~DEVFL_KICKME;
+ aoecmd_work(d);
+ }

sl = d->sendq_hd;
d->sendq_hd = d->sendq_tl = NULL;
@@ -552,23 +586,27 @@ aoecmd_ata_rsp(struct sk_buff *skb)
case WIN_WRITE:
case WIN_WRITE_EXT:
if (f->bcnt -= n) {
+ skb = f->skb;
f->bufaddr += n;
put_lba(ahout, f->lba += ahout->scnt);
n = f->bcnt;
if (n > DEFAULTBCNT)
n = DEFAULTBCNT;
ahout->scnt = n >> 9;
- if (ahout->aflags & AOEAFL_WRITE)
- skb_fill_page_desc(f->skb, 0,
+ if (ahout->aflags & AOEAFL_WRITE) {
+ skb_fill_page_desc(skb, 0,
virt_to_page(f->bufaddr),
offset_in_page(f->bufaddr), n);
+ skb->len = sizeof *hout + sizeof *ahout + n;
+ skb->data_len = n;
+ }
f->tag = newtag(d);
hout->tag = cpu_to_be32(f->tag);
skb->dev = d->ifp;
- skb_get(f->skb);
- f->skb->next = NULL;
+ skb = skb_clone(skb, GFP_ATOMIC);
spin_unlock_irqrestore(&d->lock, flags);
- aoenet_xmit(f->skb);
+ if (skb)
+ aoenet_xmit(skb);
return;
}
if (n > DEFAULTBCNT)
@@ -642,7 +680,7 @@ aoecmd_ata_id(struct aoedev *d)
struct frame *f;
struct sk_buff *skb;

- f = getframe(d, FREETAG);
+ f = freeframe(d);
if (f == NULL) {
eprintk("can't get a frame. This shouldn't happen.\n");
return NULL;
@@ -652,8 +690,8 @@ aoecmd_ata_id(struct aoedev *d)
skb = f->skb;
h = (struct aoe_hdr *) skb->mac.raw;
ah = (struct aoe_atahdr *) (h+1);
- skb->len = sizeof *h + sizeof *ah;
- memset(h, 0, skb->len);
+ skb->len = ETH_ZLEN;
+ memset(h, 0, ETH_ZLEN);
f->tag = aoehdr_atainit(d, h);
f->waited = 0;

@@ -663,12 +701,11 @@ aoecmd_ata_id(struct aoedev *d)
ah->lba3 = 0xa0;

skb->dev = d->ifp;
- skb_get(skb);

d->rttavg = MAXTIMER;
d->timer.function = rexmit_timer;

- return skb;
+ return skb_clone(skb, GFP_ATOMIC);
}

void
@@ -724,7 +761,12 @@ aoecmd_cfg_rsp(struct sk_buff *skb)
n /= 512;
if (n > ch->scnt)
n = ch->scnt;
- d->maxbcnt = n ? n * 512 : DEFAULTBCNT;
+ n = n ? n * 512 : DEFAULTBCNT;
+ if (n != d->maxbcnt) {
+ iprintk("e%ld.%ld: setting %d byte data frames on %s\n",
+ d->aoemajor, d->aoeminor, n, d->ifp->name);
+ d->maxbcnt = n;
+ }
}

/* don't change users' perspective */
diff -upr 2.6.18-rc4-orig/drivers/block/aoe/aoedev.c 2.6.18-rc4-aoe/drivers/block/aoe/aoedev.c
--- 2.6.18-rc4-orig/drivers/block/aoe/aoedev.c 2006-09-20 14:29:35.000000000 -0400
+++ 2.6.18-rc4-aoe/drivers/block/aoe/aoedev.c 2006-09-20 14:29:36.000000000 -0400
@@ -121,6 +121,7 @@ aoedev_downdev(struct aoedev *d)
mempool_free(buf, d->bufpool);
bio_endio(bio, bio->bi_size, -EIO);
}
+ skb_shinfo(f->skb)->nr_frags = f->skb->data_len = 0;
}
d->inprocess = NULL;



--
"Ed L. Cashin" <ecashin@xxxxxxxxxx>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/