Konrad Rzeszutek Wilk
2011-Nov-30 18:15 UTC
[PATCH] Xen block improvements for 3.3 [secure-discard flag](v1)
Way back in 3.2 time-frame we worked out some of these patches and how
to improve the code. I don''t think I had posted them after we chatted
about what to do or how to do it...
Either way, posting them here to kick-start the conversation and
see if we can improve these patches some more. I''ve been running
with these patches for some time now so I am quite confident that I got
the 32-bit to 64-bit structure padding worked out.
(Ran it with old guests/new guests/32-bit/64-bit/Linux/Windows
combination. Around 18 different scenarios).
Konrad Rzeszutek Wilk (3):
xen/blk[front|back]: Squash blkif_request_rw and blkif_request_discard
together
xen/blk[front|back]: Enhance discard support with secure erasing support.
xen/blkback: Move processing of BLKIF_OP_DISCARD from dispatch_rw_block_io
Li Dongyang (1):
xen-blkback: convert hole punching to discard request on loop devices
drivers/block/xen-blkback/blkback.c | 84 ++++++++++++++--------------------
drivers/block/xen-blkback/common.h | 67 ++++++++++++++++------------
drivers/block/xen-blkback/xenbus.c | 12 +++++
drivers/block/xen-blkfront.c | 77 +++++++++++++++++++++-----------
include/xen/interface/io/blkif.h | 40 ++++++++++++++---
5 files changed, 169 insertions(+), 111 deletions(-)
Konrad Rzeszutek Wilk
2011-Nov-30 18:15 UTC
[PATCH 1/4] xen/blk[front|back]: Squash blkif_request_rw and blkif_request_discard together
In a union type structure to deal with the overlapping
attributes in a easier manner.
Suggested-by: Ian Campbell <Ian.Campbell@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
drivers/block/xen-blkback/blkback.c | 13 ++++---
drivers/block/xen-blkback/common.h | 64 +++++++++++++++++++---------------
drivers/block/xen-blkfront.c | 44 ++++++++++++-----------
include/xen/interface/io/blkif.h | 24 +++++++++----
4 files changed, 83 insertions(+), 62 deletions(-)
diff --git a/drivers/block/xen-blkback/blkback.c
b/drivers/block/xen-blkback/blkback.c
index 15ec4db..d7104ab 100644
--- a/drivers/block/xen-blkback/blkback.c
+++ b/drivers/block/xen-blkback/blkback.c
@@ -362,7 +362,7 @@ static int xen_blkbk_map(struct blkif_request *req,
{
struct gnttab_map_grant_ref map[BLKIF_MAX_SEGMENTS_PER_REQUEST];
int i;
- int nseg = req->nr_segments;
+ int nseg = req->u.rw.nr_segments;
int ret = 0;
/*
@@ -449,7 +449,7 @@ static void xen_blk_discard(struct xen_blkif *blkif, struct
blkif_request *req)
} else if (err)
status = BLKIF_RSP_ERROR;
- make_response(blkif, req->id, req->operation, status);
+ make_response(blkif, req->u.discard.id, req->operation, status);
}
static void xen_blk_drain_io(struct xen_blkif *blkif)
@@ -644,7 +644,8 @@ static int dispatch_rw_block_io(struct xen_blkif *blkif,
}
/* Check that the number of segments is sane. */
- nseg = req->nr_segments;
+ nseg = req->u.rw.nr_segments;
+
if (unlikely(nseg == 0 && operation != WRITE_FLUSH &&
operation != REQ_DISCARD) ||
unlikely(nseg > BLKIF_MAX_SEGMENTS_PER_REQUEST)) {
@@ -654,12 +655,12 @@ static int dispatch_rw_block_io(struct xen_blkif *blkif,
goto fail_response;
}
- preq.dev = req->handle;
+ preq.dev = req->u.rw.handle;
preq.sector_number = req->u.rw.sector_number;
preq.nr_sects = 0;
pending_req->blkif = blkif;
- pending_req->id = req->id;
+ pending_req->id = req->u.rw.id;
pending_req->operation = req->operation;
pending_req->status = BLKIF_RSP_OKAY;
pending_req->nr_pages = nseg;
@@ -784,7 +785,7 @@ static int dispatch_rw_block_io(struct xen_blkif *blkif,
xen_blkbk_unmap(pending_req);
fail_response:
/* Haven''t submitted any bio''s yet. */
- make_response(blkif, req->id, req->operation, BLKIF_RSP_ERROR);
+ make_response(blkif, req->u.rw.id, req->operation, BLKIF_RSP_ERROR);
free_req(pending_req);
msleep(1); /* back off a bit */
return -EIO;
diff --git a/drivers/block/xen-blkback/common.h
b/drivers/block/xen-blkback/common.h
index dfb1b3a..dbfe7b3 100644
--- a/drivers/block/xen-blkback/common.h
+++ b/drivers/block/xen-blkback/common.h
@@ -60,58 +60,66 @@ struct blkif_common_response {
char dummy;
};
-/* i386 protocol version */
-#pragma pack(push, 4)
-
struct blkif_x86_32_request_rw {
+ uint8_t nr_segments; /* number of segments */
+ blkif_vdev_t handle; /* only for read/write requests */
+ uint64_t id; /* private guest value, echoed in resp */
blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
-};
+} __attribute__((__packed__));
struct blkif_x86_32_request_discard {
+ uint8_t nr_segments; /* number of segments */
+ blkif_vdev_t _pad1; /* was "handle" for read/write requests
*/
+ uint64_t id; /* private guest value, echoed in resp */
blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
- uint64_t nr_sectors;
-};
+ uint64_t nr_sectors;
+} __attribute__((__packed__));
struct blkif_x86_32_request {
uint8_t operation; /* BLKIF_OP_??? */
- uint8_t nr_segments; /* number of segments */
- blkif_vdev_t handle; /* only for read/write requests */
- uint64_t id; /* private guest value, echoed in resp */
union {
struct blkif_x86_32_request_rw rw;
struct blkif_x86_32_request_discard discard;
} u;
-};
+} __attribute__((__packed__));
+
+/* i386 protocol version */
+#pragma pack(push, 4)
struct blkif_x86_32_response {
uint64_t id; /* copied from request */
uint8_t operation; /* copied from request */
int16_t status; /* BLKIF_RSP_??? */
};
#pragma pack(pop)
-
/* x86_64 protocol version */
struct blkif_x86_64_request_rw {
+ uint8_t nr_segments; /* number of segments */
+ blkif_vdev_t handle; /* only for read/write requests */
+ uint32_t _pad1; /* offsetof(blkif_reqest..,u.rw.id)==8 */
+ uint64_t id;
blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
-};
+} __attribute__((__packed__));
struct blkif_x86_64_request_discard {
+ uint8_t nr_segments; /* number of segments */
+ blkif_vdev_t _pad1; /* was "handle" for read/write requests
*/
+ uint32_t _pad2; /* offsetof(blkif_..,u.discard.id)==8 */
+ uint64_t id;
blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
- uint64_t nr_sectors;
-};
+ uint64_t nr_sectors;
+} __attribute__((__packed__));
struct blkif_x86_64_request {
uint8_t operation; /* BLKIF_OP_??? */
- uint8_t nr_segments; /* number of segments */
- blkif_vdev_t handle; /* only for read/write requests */
- uint64_t __attribute__((__aligned__(8))) id;
union {
struct blkif_x86_64_request_rw rw;
struct blkif_x86_64_request_discard discard;
} u;
-};
+} __attribute__((__packed__));
+
struct blkif_x86_64_response {
uint64_t __attribute__((__aligned__(8))) id;
uint8_t operation; /* copied from request */
@@ -237,18 +245,18 @@ static inline void blkif_get_x86_32_req(struct
blkif_request *dst,
{
int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST;
dst->operation = src->operation;
- dst->nr_segments = src->nr_segments;
- dst->handle = src->handle;
- dst->id = src->id;
switch (src->operation) {
case BLKIF_OP_READ:
case BLKIF_OP_WRITE:
case BLKIF_OP_WRITE_BARRIER:
case BLKIF_OP_FLUSH_DISKCACHE:
+ dst->u.rw.nr_segments = src->u.rw.nr_segments;
+ dst->u.rw.handle = src->u.rw.handle;
+ dst->u.rw.id = src->u.rw.id;
dst->u.rw.sector_number = src->u.rw.sector_number;
barrier();
- if (n > dst->nr_segments)
- n = dst->nr_segments;
+ if (n > dst->u.rw.nr_segments)
+ n = dst->u.rw.nr_segments;
for (i = 0; i < n; i++)
dst->u.rw.seg[i] = src->u.rw.seg[i];
break;
@@ -266,18 +274,18 @@ static inline void blkif_get_x86_64_req(struct
blkif_request *dst,
{
int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST;
dst->operation = src->operation;
- dst->nr_segments = src->nr_segments;
- dst->handle = src->handle;
- dst->id = src->id;
switch (src->operation) {
case BLKIF_OP_READ:
case BLKIF_OP_WRITE:
case BLKIF_OP_WRITE_BARRIER:
case BLKIF_OP_FLUSH_DISKCACHE:
+ dst->u.rw.nr_segments = src->u.rw.nr_segments;
+ dst->u.rw.handle = src->u.rw.handle;
+ dst->u.rw.id = src->u.rw.id;
dst->u.rw.sector_number = src->u.rw.sector_number;
barrier();
- if (n > dst->nr_segments)
- n = dst->nr_segments;
+ if (n > dst->u.rw.nr_segments)
+ n = dst->u.rw.nr_segments;
for (i = 0; i < n; i++)
dst->u.rw.seg[i] = src->u.rw.seg[i];
break;
diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
index 7b2ec59..2c2c4be 100644
--- a/drivers/block/xen-blkfront.c
+++ b/drivers/block/xen-blkfront.c
@@ -135,15 +135,15 @@ static int get_id_from_freelist(struct blkfront_info
*info)
{
unsigned long free = info->shadow_free;
BUG_ON(free >= BLK_RING_SIZE);
- info->shadow_free = info->shadow[free].req.id;
- info->shadow[free].req.id = 0x0fffffee; /* debug */
+ info->shadow_free = info->shadow[free].req.u.rw.id;
+ info->shadow[free].req.u.rw.id = 0x0fffffee; /* debug */
return free;
}
static void add_id_to_freelist(struct blkfront_info *info,
unsigned long id)
{
- info->shadow[id].req.id = info->shadow_free;
+ info->shadow[id].req.u.rw.id = info->shadow_free;
info->shadow[id].request = NULL;
info->shadow_free = id;
}
@@ -287,9 +287,9 @@ static int blkif_queue_request(struct request *req)
id = get_id_from_freelist(info);
info->shadow[id].request = req;
- ring_req->id = id;
+ ring_req->u.rw.id = id;
ring_req->u.rw.sector_number = (blkif_sector_t)blk_rq_pos(req);
- ring_req->handle = info->handle;
+ ring_req->u.rw.handle = info->handle;
ring_req->operation = rq_data_dir(req) ?
BLKIF_OP_WRITE : BLKIF_OP_READ;
@@ -308,13 +308,15 @@ static int blkif_queue_request(struct request *req)
if (unlikely(req->cmd_flags & REQ_DISCARD)) {
/* id, sector_number and handle are set above. */
ring_req->operation = BLKIF_OP_DISCARD;
- ring_req->nr_segments = 0;
+ ring_req->u.discard.nr_segments = 0;
ring_req->u.discard.nr_sectors = blk_rq_sectors(req);
} else {
- ring_req->nr_segments = blk_rq_map_sg(req->q, req, info->sg);
- BUG_ON(ring_req->nr_segments > BLKIF_MAX_SEGMENTS_PER_REQUEST);
+ ring_req->u.rw.nr_segments = blk_rq_map_sg(req->q, req,
+ info->sg);
+ BUG_ON(ring_req->u.rw.nr_segments >
+ BLKIF_MAX_SEGMENTS_PER_REQUEST);
- for_each_sg(info->sg, sg, ring_req->nr_segments, i) {
+ for_each_sg(info->sg, sg, ring_req->u.rw.nr_segments, i) {
buffer_mfn = pfn_to_mfn(page_to_pfn(sg_page(sg)));
fsect = sg->offset >> 9;
lsect = fsect + (sg->length >> 9) - 1;
@@ -705,7 +707,7 @@ static void blkif_free(struct blkfront_info *info, int
suspend)
static void blkif_completion(struct blk_shadow *s)
{
int i;
- for (i = 0; i < s->req.nr_segments; i++)
+ for (i = 0; i < s->req.u.rw.nr_segments; i++)
gnttab_end_foreign_access(s->req.u.rw.seg[i].gref, 0, 0UL);
}
@@ -763,7 +765,7 @@ static irqreturn_t blkif_interrupt(int irq, void *dev_id)
error = -EOPNOTSUPP;
}
if (unlikely(bret->status == BLKIF_RSP_ERROR &&
- info->shadow[id].req.nr_segments == 0)) {
+ info->shadow[id].req.u.rw.nr_segments == 0)) {
printk(KERN_WARNING "blkfront: %s: empty write %s op failed\n",
info->flush_op == BLKIF_OP_WRITE_BARRIER ?
"barrier" : "flush disk cache",
@@ -984,8 +986,8 @@ static int blkfront_probe(struct xenbus_device *dev,
INIT_WORK(&info->work, blkif_restart_queue);
for (i = 0; i < BLK_RING_SIZE; i++)
- info->shadow[i].req.id = i+1;
- info->shadow[BLK_RING_SIZE-1].req.id = 0x0fffffff;
+ info->shadow[i].req.u.rw.id = i+1;
+ info->shadow[BLK_RING_SIZE-1].req.u.rw.id = 0x0fffffff;
/* Front end dir is a number, which is used as the id. */
info->handle = simple_strtoul(strrchr(dev->nodename,
''/'')+1, NULL, 0);
@@ -1019,9 +1021,9 @@ static int blkif_recover(struct blkfront_info *info)
/* Stage 2: Set up free list. */
memset(&info->shadow, 0, sizeof(info->shadow));
for (i = 0; i < BLK_RING_SIZE; i++)
- info->shadow[i].req.id = i+1;
+ info->shadow[i].req.u.rw.id = i+1;
info->shadow_free = info->ring.req_prod_pvt;
- info->shadow[BLK_RING_SIZE-1].req.id = 0x0fffffff;
+ info->shadow[BLK_RING_SIZE-1].req.u.rw.id = 0x0fffffff;
/* Stage 3: Find pending requests and requeue them. */
for (i = 0; i < BLK_RING_SIZE; i++) {
@@ -1034,17 +1036,17 @@ static int blkif_recover(struct blkfront_info *info)
*req = copy[i].req;
/* We get a new request id, and must reset the shadow state. */
- req->id = get_id_from_freelist(info);
- memcpy(&info->shadow[req->id], ©[i], sizeof(copy[i]));
+ req->u.rw.id = get_id_from_freelist(info);
+ memcpy(&info->shadow[req->u.rw.id], ©[i], sizeof(copy[i]));
/* Rewrite any grant references invalidated by susp/resume. */
- for (j = 0; j < req->nr_segments; j++)
+ for (j = 0; j < req->u.rw.nr_segments; j++)
gnttab_grant_foreign_access_ref(
req->u.rw.seg[j].gref,
info->xbdev->otherend_id,
- pfn_to_mfn(info->shadow[req->id].frame[j]),
- rq_data_dir(info->shadow[req->id].request));
- info->shadow[req->id].req = *req;
+ pfn_to_mfn(info->shadow[req->u.rw.id].frame[j]),
+ rq_data_dir(info->shadow[req->u.rw.id].request));
+ info->shadow[req->u.rw.id].req = *req;
info->ring.req_prod_pvt++;
}
diff --git a/include/xen/interface/io/blkif.h b/include/xen/interface/io/blkif.h
index 9324488..f88e28b 100644
--- a/include/xen/interface/io/blkif.h
+++ b/include/xen/interface/io/blkif.h
@@ -95,6 +95,12 @@ typedef uint64_t blkif_sector_t;
#define BLKIF_MAX_SEGMENTS_PER_REQUEST 11
struct blkif_request_rw {
+ uint8_t nr_segments; /* number of segments */
+ blkif_vdev_t handle; /* only for read/write requests */
+#ifdef CONFIG_X86_64
+ uint32_t _pad1; /* offsetof(blkif_request,u.rw.id) == 8 */
+#endif
+ uint64_t id; /* private guest value, echoed in resp */
blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
struct blkif_request_segment {
grant_ref_t gref; /* reference to I/O buffer frame */
@@ -102,23 +108,27 @@ struct blkif_request_rw {
/* @last_sect: last sector in frame to transfer (inclusive). */
uint8_t first_sect, last_sect;
} seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
-};
+} __attribute__((__packed__));
struct blkif_request_discard {
+ uint8_t nr_segments; /* number of segments */
+ blkif_vdev_t _pad1; /* only for read/write requests */
+#ifdef CONFIG_X86_64
+ uint32_t _pad2; /* offsetof(blkif_req..,u.discard.id)==8*/
+#endif
+ uint64_t id; /* private guest value, echoed in resp */
blkif_sector_t sector_number;
- uint64_t nr_sectors;
-};
+ uint64_t nr_sectors;
+ uint8_t _pad3;
+} __attribute__((__packed__));
struct blkif_request {
uint8_t operation; /* BLKIF_OP_??? */
- uint8_t nr_segments; /* number of segments */
- blkif_vdev_t handle; /* only for read/write requests */
- uint64_t id; /* private guest value, echoed in resp */
union {
struct blkif_request_rw rw;
struct blkif_request_discard discard;
} u;
-};
+} __attribute__((__packed__));
struct blkif_response {
uint64_t id; /* copied from request */
--
1.7.7.3
Konrad Rzeszutek Wilk
2011-Nov-30 18:15 UTC
[PATCH 2/4] xen/blk[front|back]: Enhance discard support with secure erasing support.
Part of the blkdev_issue_discard(xx) operation is that it can also
issue a secure discard operation that will permanantly remove the
sectors in question. We advertise that we can support that via the
''discard-secure'' attribute and on the request, if the
''secure'' bit
is set, we will attempt to pass in REQ_DISCARD | REQ_SECURE.
CC: Li Dongyang <lidongyang@novell.com>
[v1: Used ''flag'' instead of ''secure:1'' bit]
[v2: Use ''reserved'' uint8_t instead of adding a new value]
[v3: Check for nseg when mapping instead of operation]
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
drivers/block/xen-blkback/blkback.c | 18 ++++++++++-----
drivers/block/xen-blkback/common.h | 7 ++++-
drivers/block/xen-blkback/xenbus.c | 12 ++++++++++
drivers/block/xen-blkfront.c | 41 ++++++++++++++++++++++++++--------
include/xen/interface/io/blkif.h | 18 ++++++++++++++-
5 files changed, 77 insertions(+), 19 deletions(-)
diff --git a/drivers/block/xen-blkback/blkback.c
b/drivers/block/xen-blkback/blkback.c
index d7104ab..9d2261b 100644
--- a/drivers/block/xen-blkback/blkback.c
+++ b/drivers/block/xen-blkback/blkback.c
@@ -422,13 +422,16 @@ static void xen_blk_discard(struct xen_blkif *blkif,
struct blkif_request *req)
int status = BLKIF_RSP_OKAY;
struct block_device *bdev = blkif->vbd.bdev;
- if (blkif->blk_backend_type == BLKIF_BACKEND_PHY)
+ if (blkif->blk_backend_type == BLKIF_BACKEND_PHY) {
+ unsigned long secure = (blkif->vbd.discard_secure &&
+ (req->u.discard.flag & BLKIF_DISCARD_SECURE)) ?
+ BLKDEV_DISCARD_SECURE : 0;
/* just forward the discard request */
err = blkdev_issue_discard(bdev,
req->u.discard.sector_number,
req->u.discard.nr_sectors,
- GFP_KERNEL, 0);
- else if (blkif->blk_backend_type == BLKIF_BACKEND_FILE) {
+ GFP_KERNEL, secure);
+ } else if (blkif->blk_backend_type == BLKIF_BACKEND_FILE) {
/* punch a hole in the backing file */
struct loop_device *lo = bdev->bd_disk->private_data;
struct file *file = lo->lo_backing_file;
@@ -643,8 +646,11 @@ static int dispatch_rw_block_io(struct xen_blkif *blkif,
break;
}
- /* Check that the number of segments is sane. */
- nseg = req->u.rw.nr_segments;
+ if (unlikely(operation == REQ_DISCARD))
+ nseg = 0;
+ else
+ /* Check that the number of segments is sane. */
+ nseg = req->u.rw.nr_segments;
if (unlikely(nseg == 0 && operation != WRITE_FLUSH &&
operation != REQ_DISCARD) ||
@@ -708,7 +714,7 @@ static int dispatch_rw_block_io(struct xen_blkif *blkif,
* the hypercall to unmap the grants - that is all done in
* xen_blkbk_unmap.
*/
- if (operation != REQ_DISCARD && xen_blkbk_map(req, pending_req, seg))
+ if (nseg && xen_blkbk_map(req, pending_req, seg))
goto fail_flush;
/*
diff --git a/drivers/block/xen-blkback/common.h
b/drivers/block/xen-blkback/common.h
index dbfe7b3..d0ee7ed 100644
--- a/drivers/block/xen-blkback/common.h
+++ b/drivers/block/xen-blkback/common.h
@@ -69,7 +69,7 @@ struct blkif_x86_32_request_rw {
} __attribute__((__packed__));
struct blkif_x86_32_request_discard {
- uint8_t nr_segments; /* number of segments */
+ uint8_t flag; /* BLKIF_DISCARD_SECURE or zero */
blkif_vdev_t _pad1; /* was "handle" for read/write requests
*/
uint64_t id; /* private guest value, echoed in resp */
blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
@@ -104,7 +104,7 @@ struct blkif_x86_64_request_rw {
} __attribute__((__packed__));
struct blkif_x86_64_request_discard {
- uint8_t nr_segments; /* number of segments */
+ uint8_t flag; /* BLKIF_DISCARD_SECURE or zero */
blkif_vdev_t _pad1; /* was "handle" for read/write requests
*/
uint32_t _pad2; /* offsetof(blkif_..,u.discard.id)==8 */
uint64_t id;
@@ -164,6 +164,7 @@ struct xen_vbd {
/* Cached size parameter. */
sector_t size;
bool flush_support;
+ bool discard_secure;
};
struct backend_info;
@@ -261,6 +262,7 @@ static inline void blkif_get_x86_32_req(struct blkif_request
*dst,
dst->u.rw.seg[i] = src->u.rw.seg[i];
break;
case BLKIF_OP_DISCARD:
+ dst->u.discard.flag = src->u.discard.flag;
dst->u.discard.sector_number = src->u.discard.sector_number;
dst->u.discard.nr_sectors = src->u.discard.nr_sectors;
break;
@@ -290,6 +292,7 @@ static inline void blkif_get_x86_64_req(struct blkif_request
*dst,
dst->u.rw.seg[i] = src->u.rw.seg[i];
break;
case BLKIF_OP_DISCARD:
+ dst->u.discard.flag = src->u.discard.flag;
dst->u.discard.sector_number = src->u.discard.sector_number;
dst->u.discard.nr_sectors = src->u.discard.nr_sectors;
break;
diff --git a/drivers/block/xen-blkback/xenbus.c
b/drivers/block/xen-blkback/xenbus.c
index f759ad4..187fd2c 100644
--- a/drivers/block/xen-blkback/xenbus.c
+++ b/drivers/block/xen-blkback/xenbus.c
@@ -338,6 +338,9 @@ static int xen_vbd_create(struct xen_blkif *blkif,
blkif_vdev_t handle,
if (q && q->flush_flags)
vbd->flush_support = true;
+ if (q && blk_queue_secdiscard(q))
+ vbd->discard_secure = true;
+
DPRINTK("Successful creation of handle=%04x (dom=%u)\n",
handle, blkif->domid);
return 0;
@@ -420,6 +423,15 @@ int xen_blkbk_discard(struct xenbus_transaction xbt, struct
backend_info *be)
state = 1;
blkif->blk_backend_type = BLKIF_BACKEND_PHY;
}
+ /* Optional. */
+ err = xenbus_printf(xbt, dev->nodename,
+ "discard-secure", "%d",
+ blkif->vbd.discard_secure);
+ if (err) {
+ xenbus_dev_fatal(dev, err,
+ "writting discard-secure");
+ goto kfree;
+ }
}
} else {
err = PTR_ERR(type);
diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
index 2c2c4be..351ddef 100644
--- a/drivers/block/xen-blkfront.c
+++ b/drivers/block/xen-blkfront.c
@@ -98,7 +98,8 @@ struct blkfront_info
unsigned long shadow_free;
unsigned int feature_flush;
unsigned int flush_op;
- unsigned int feature_discard;
+ unsigned int feature_discard:1;
+ unsigned int feature_secdiscard:1;
unsigned int discard_granularity;
unsigned int discard_alignment;
int is_ready;
@@ -305,11 +306,14 @@ static int blkif_queue_request(struct request *req)
ring_req->operation = info->flush_op;
}
- if (unlikely(req->cmd_flags & REQ_DISCARD)) {
+ if (unlikely(req->cmd_flags & (REQ_DISCARD | REQ_SECURE))) {
/* id, sector_number and handle are set above. */
ring_req->operation = BLKIF_OP_DISCARD;
- ring_req->u.discard.nr_segments = 0;
ring_req->u.discard.nr_sectors = blk_rq_sectors(req);
+ if ((req->cmd_flags & REQ_SECURE) &&
info->feature_secdiscard)
+ ring_req->u.discard.flag = BLKIF_DISCARD_SECURE;
+ else
+ ring_req->u.discard.flag = 0;
} else {
ring_req->u.rw.nr_segments = blk_rq_map_sg(req->q, req,
info->sg);
@@ -426,6 +430,8 @@ static int xlvbd_init_blk_queue(struct gendisk *gd, u16
sector_size)
blk_queue_max_discard_sectors(rq, get_capacity(gd));
rq->limits.discard_granularity = info->discard_granularity;
rq->limits.discard_alignment = info->discard_alignment;
+ if (info->feature_secdiscard)
+ queue_flag_set_unlocked(QUEUE_FLAG_SECDISCARD, rq);
}
/* Hard sector size and max sectors impersonate the equiv. hardware. */
@@ -707,6 +713,8 @@ static void blkif_free(struct blkfront_info *info, int
suspend)
static void blkif_completion(struct blk_shadow *s)
{
int i;
+ /* Do not let BLKIF_OP_DISCARD as nr_segment is in the same place
+ * flag. */
for (i = 0; i < s->req.u.rw.nr_segments; i++)
gnttab_end_foreign_access(s->req.u.rw.seg[i].gref, 0, 0UL);
}
@@ -738,7 +746,8 @@ static irqreturn_t blkif_interrupt(int irq, void *dev_id)
id = bret->id;
req = info->shadow[id].request;
- blkif_completion(&info->shadow[id]);
+ if (bret->operation != BLKIF_OP_DISCARD)
+ blkif_completion(&info->shadow[id]);
add_id_to_freelist(info, id);
@@ -751,7 +760,9 @@ static irqreturn_t blkif_interrupt(int irq, void *dev_id)
info->gd->disk_name);
error = -EOPNOTSUPP;
info->feature_discard = 0;
+ info->feature_secdiscard = 0;
queue_flag_clear(QUEUE_FLAG_DISCARD, rq);
+ queue_flag_clear(QUEUE_FLAG_SECDISCARD, rq);
}
__blk_end_request_all(req, error);
break;
@@ -1039,13 +1050,15 @@ static int blkif_recover(struct blkfront_info *info)
req->u.rw.id = get_id_from_freelist(info);
memcpy(&info->shadow[req->u.rw.id], ©[i], sizeof(copy[i]));
+ if (req->operation != BLKIF_OP_DISCARD) {
/* Rewrite any grant references invalidated by susp/resume. */
- for (j = 0; j < req->u.rw.nr_segments; j++)
- gnttab_grant_foreign_access_ref(
- req->u.rw.seg[j].gref,
- info->xbdev->otherend_id,
- pfn_to_mfn(info->shadow[req->u.rw.id].frame[j]),
- rq_data_dir(info->shadow[req->u.rw.id].request));
+ for (j = 0; j < req->u.rw.nr_segments; j++)
+ gnttab_grant_foreign_access_ref(
+ req->u.rw.seg[j].gref,
+ info->xbdev->otherend_id,
+ pfn_to_mfn(info->shadow[req->u.rw.id].frame[j]),
+ rq_data_dir(info->shadow[req->u.rw.id].request));
+ }
info->shadow[req->u.rw.id].req = *req;
info->ring.req_prod_pvt++;
@@ -1137,11 +1150,13 @@ static void blkfront_setup_discard(struct blkfront_info
*info)
char *type;
unsigned int discard_granularity;
unsigned int discard_alignment;
+ unsigned int discard_secure;
type = xenbus_read(XBT_NIL, info->xbdev->otherend, "type",
NULL);
if (IS_ERR(type))
return;
+ info->feature_secdiscard = 0;
if (strncmp(type, "phy", 3) == 0) {
err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
"discard-granularity", "%u", &discard_granularity,
@@ -1152,6 +1167,12 @@ static void blkfront_setup_discard(struct blkfront_info
*info)
info->discard_granularity = discard_granularity;
info->discard_alignment = discard_alignment;
}
+ err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
+ "discard-secure", "%d", &discard_secure,
+ NULL);
+ if (!err)
+ info->feature_secdiscard = discard_secure;
+
} else if (strncmp(type, "file", 4) == 0)
info->feature_discard = 1;
diff --git a/include/xen/interface/io/blkif.h b/include/xen/interface/io/blkif.h
index f88e28b..ee338bf 100644
--- a/include/xen/interface/io/blkif.h
+++ b/include/xen/interface/io/blkif.h
@@ -84,6 +84,21 @@ typedef uint64_t blkif_sector_t;
* e07154r6-Data_Set_Management_Proposal_for_ATA-ACS2.doc
* http://www.seagate.com/staticfiles/support/disc/manuals/
* Interface%20manuals/100293068c.pdf
+ * The backend can optionally provide three extra XenBus attributes to
+ * further optimize the discard functionality:
+ * ''discard-aligment'' - Devices that support discard
functionality may
+ * internally allocate space in units that are bigger than the exported
+ * logical block size. The discard-alignment parameter indicates how many bytes
+ * the beginning of the partition is offset from the internal allocation
unit''s
+ * natural alignment.
+ * ''discard-granularity'' - Devices that support discard
functionality may
+ * internally allocate space using units that are bigger than the logical block
+ * size. The discard-granularity parameter indicates the size of the internal
+ * allocation unit in bytes if reported by the device. Otherwise the
+ * discard-granularity will be set to match the device''s physical
block size.
+ * ''discard-secure'' - All copies of the discarded sectors
(potentially created
+ * by garbage collection) must also be erased. To use this feature, the flag
+ * BLKIF_DISCARD_SECURE must be set in the blkif_request_trim.
*/
#define BLKIF_OP_DISCARD 5
@@ -111,7 +126,8 @@ struct blkif_request_rw {
} __attribute__((__packed__));
struct blkif_request_discard {
- uint8_t nr_segments; /* number of segments */
+ uint8_t flag; /* BLKIF_DISCARD_SECURE or zero. */
+#define BLKIF_DISCARD_SECURE (1<<0) /* ignored if discard-secure=0
*/
blkif_vdev_t _pad1; /* only for read/write requests */
#ifdef CONFIG_X86_64
uint32_t _pad2; /* offsetof(blkif_req..,u.discard.id)==8*/
--
1.7.7.3
Konrad Rzeszutek Wilk
2011-Nov-30 18:15 UTC
[PATCH 3/4] xen/blkback: Move processing of BLKIF_OP_DISCARD from dispatch_rw_block_io
. and move it to its own function that will deal with the
discard operation.
Suggested-by: Jan Beulich <JBeulich@novell.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
drivers/block/xen-blkback/blkback.c | 54 +++++++++++++++-------------------
1 files changed, 24 insertions(+), 30 deletions(-)
diff --git a/drivers/block/xen-blkback/blkback.c
b/drivers/block/xen-blkback/blkback.c
index 9d2261b..b058de7 100644
--- a/drivers/block/xen-blkback/blkback.c
+++ b/drivers/block/xen-blkback/blkback.c
@@ -416,12 +416,16 @@ static int xen_blkbk_map(struct blkif_request *req,
return ret;
}
-static void xen_blk_discard(struct xen_blkif *blkif, struct blkif_request *req)
+static int dispatch_discard_io(struct xen_blkif *blkif,
+ struct blkif_request *req)
{
int err = 0;
int status = BLKIF_RSP_OKAY;
struct block_device *bdev = blkif->vbd.bdev;
+ blkif->st_ds_req++;
+
+ xen_blkif_get(blkif);
if (blkif->blk_backend_type == BLKIF_BACKEND_PHY) {
unsigned long secure = (blkif->vbd.discard_secure &&
(req->u.discard.flag & BLKIF_DISCARD_SECURE)) ?
@@ -453,6 +457,8 @@ static void xen_blk_discard(struct xen_blkif *blkif, struct
blkif_request *req)
status = BLKIF_RSP_ERROR;
make_response(blkif, req->u.discard.id, req->operation, status);
+ xen_blkif_put(blkif);
+ return err;
}
static void xen_blk_drain_io(struct xen_blkif *blkif)
@@ -576,8 +582,11 @@ __do_block_io_op(struct xen_blkif *blkif)
/* Apply all sanity checks to /private copy/ of request. */
barrier();
-
- if (dispatch_rw_block_io(blkif, &req, pending_req))
+ if (unlikely(req.operation == BLKIF_OP_DISCARD)) {
+ free_req(pending_req);
+ if (dispatch_discard_io(blkif, &req))
+ break;
+ } else if (dispatch_rw_block_io(blkif, &req, pending_req))
break;
/* Yield point for this unbounded loop. */
@@ -636,24 +645,16 @@ static int dispatch_rw_block_io(struct xen_blkif *blkif,
blkif->st_f_req++;
operation = WRITE_FLUSH;
break;
- case BLKIF_OP_DISCARD:
- blkif->st_ds_req++;
- operation = REQ_DISCARD;
- break;
default:
operation = 0; /* make gcc happy */
goto fail_response;
break;
}
- if (unlikely(operation == REQ_DISCARD))
- nseg = 0;
- else
- /* Check that the number of segments is sane. */
- nseg = req->u.rw.nr_segments;
+ /* Check that the number of segments is sane. */
+ nseg = req->u.rw.nr_segments;
- if (unlikely(nseg == 0 && operation != WRITE_FLUSH &&
- operation != REQ_DISCARD) ||
+ if (unlikely(nseg == 0 && operation != WRITE_FLUSH) ||
unlikely(nseg > BLKIF_MAX_SEGMENTS_PER_REQUEST)) {
pr_debug(DRV_PFX "Bad number of segments in request (%d)\n",
nseg);
@@ -714,7 +715,7 @@ static int dispatch_rw_block_io(struct xen_blkif *blkif,
* the hypercall to unmap the grants - that is all done in
* xen_blkbk_unmap.
*/
- if (nseg && xen_blkbk_map(req, pending_req, seg))
+ if (xen_blkbk_map(req, pending_req, seg))
goto fail_flush;
/*
@@ -746,23 +747,16 @@ static int dispatch_rw_block_io(struct xen_blkif *blkif,
/* This will be hit if the operation was a flush or discard. */
if (!bio) {
- BUG_ON(operation != WRITE_FLUSH && operation != REQ_DISCARD);
+ BUG_ON(operation != WRITE_FLUSH);
- if (operation == WRITE_FLUSH) {
- bio = bio_alloc(GFP_KERNEL, 0);
- if (unlikely(bio == NULL))
- goto fail_put_bio;
+ bio = bio_alloc(GFP_KERNEL, 0);
+ if (unlikely(bio == NULL))
+ goto fail_put_bio;
- biolist[nbio++] = bio;
- bio->bi_bdev = preq.bdev;
- bio->bi_private = pending_req;
- bio->bi_end_io = end_block_io_op;
- } else if (operation == REQ_DISCARD) {
- xen_blk_discard(blkif, req);
- xen_blkif_put(blkif);
- free_req(pending_req);
- return 0;
- }
+ biolist[nbio++] = bio;
+ bio->bi_bdev = preq.bdev;
+ bio->bi_private = pending_req;
+ bio->bi_end_io = end_block_io_op;
}
/*
--
1.7.7.3
Konrad Rzeszutek Wilk
2011-Nov-30 18:15 UTC
[PATCH 4/4] xen-blkback: convert hole punching to discard request on loop devices
From: Li Dongyang <lidongyang@novell.com>
As of dfaa2ef68e80c378e610e3c8c536f1c239e8d3ef, loop devices support
discard request now. We could just issue a discard request, and
the loop driver will punch the hole for us, so we don''t need to touch
the internals of loop device and punch the hole ourselves, Thanks.
V0->V1: rebased on devel/for-jens-3.3
Signed-off-by: Li Dongyang <lidongyang@novell.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
drivers/block/xen-blkback/blkback.c | 19 ++-----------------
1 files changed, 2 insertions(+), 17 deletions(-)
diff --git a/drivers/block/xen-blkback/blkback.c
b/drivers/block/xen-blkback/blkback.c
index b058de7..0088bf6 100644
--- a/drivers/block/xen-blkback/blkback.c
+++ b/drivers/block/xen-blkback/blkback.c
@@ -39,9 +39,6 @@
#include <linux/list.h>
#include <linux/delay.h>
#include <linux/freezer.h>
-#include <linux/loop.h>
-#include <linux/falloc.h>
-#include <linux/fs.h>
#include <xen/events.h>
#include <xen/page.h>
@@ -426,27 +423,15 @@ static int dispatch_discard_io(struct xen_blkif *blkif,
blkif->st_ds_req++;
xen_blkif_get(blkif);
- if (blkif->blk_backend_type == BLKIF_BACKEND_PHY) {
+ if (blkif->blk_backend_type == BLKIF_BACKEND_PHY ||
+ blkif->blk_backend_type == BLKIF_BACKEND_FILE) {
unsigned long secure = (blkif->vbd.discard_secure &&
(req->u.discard.flag & BLKIF_DISCARD_SECURE)) ?
BLKDEV_DISCARD_SECURE : 0;
- /* just forward the discard request */
err = blkdev_issue_discard(bdev,
req->u.discard.sector_number,
req->u.discard.nr_sectors,
GFP_KERNEL, secure);
- } else if (blkif->blk_backend_type == BLKIF_BACKEND_FILE) {
- /* punch a hole in the backing file */
- struct loop_device *lo = bdev->bd_disk->private_data;
- struct file *file = lo->lo_backing_file;
-
- if (file->f_op->fallocate)
- err = file->f_op->fallocate(file,
- FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE,
- req->u.discard.sector_number << 9,
- req->u.discard.nr_sectors << 9);
- else
- err = -EOPNOTSUPP;
} else
err = -EOPNOTSUPP;
--
1.7.7.3