From: Avi Kivity <avi@redhat.com>
To: Anthony Liguori <anthony@codemonkey.ws>
Cc: qemu-devel@nongnu.org
Subject: [Qemu-devel] [PATCH 3/5] Implement cancellation method for dma async I/O
Date: Mon, 16 Mar 2009 15:11:09 +0200 [thread overview]
Message-ID: <1237209071-26942-4-git-send-email-avi@redhat.com> (raw)
In-Reply-To: <1237209071-26942-1-git-send-email-avi@redhat.com>
Move the dma helpers to a private aio pool, and implement a cancellation
method for them. Should prevent issues when cancelling I/O while dma is
in progress.
Signed-off-by: Avi Kivity <avi@redhat.com>
---
dma-helpers.c | 15 ++++++++++++++-
dma.h | 1 +
vl.c | 1 +
3 files changed, 16 insertions(+), 1 deletions(-)
diff --git a/dma-helpers.c b/dma-helpers.c
index b2ade19..19fa4f0 100644
--- a/dma-helpers.c
+++ b/dma-helpers.c
@@ -10,6 +10,8 @@
#include "dma.h"
#include "block_int.h"
+static AIOPool dma_aio_pool;
+
void qemu_sglist_init(QEMUSGList *qsg, int alloc_hint)
{
qsg->sg = qemu_malloc(alloc_hint * sizeof(ScatterGatherEntry));
@@ -126,7 +128,7 @@ static BlockDriverAIOCB *dma_bdrv_io(
DMABlockState *dbs = qemu_malloc(sizeof(*dbs));
dbs->bs = bs;
- dbs->acb = qemu_aio_get(bs, cb, opaque);
+ dbs->acb = qemu_aio_get_pool(&dma_aio_pool, bs, cb, opaque);
dbs->sg = sg;
dbs->sector_num = sector_num;
dbs->sg_cur_index = 0;
@@ -153,3 +155,14 @@ BlockDriverAIOCB *dma_bdrv_write(BlockDriverState *bs,
return dma_bdrv_io(bs, sg, sector, cb, opaque, 1);
}
+static void dma_aio_cancel(BlockDriverAIOCB *acb)
+{
+ DMABlockState *dbs = (DMABlockState *)acb->opaque;
+
+ bdrv_aio_cancel(dbs->acb);
+}
+
+void dma_helper_init(void)
+{
+ aio_pool_init(&dma_aio_pool, sizeof(BlockDriverAIOCB), dma_aio_cancel);
+}
diff --git a/dma.h b/dma.h
index d596717..7d2ab45 100644
--- a/dma.h
+++ b/dma.h
@@ -37,5 +37,6 @@ BlockDriverAIOCB *dma_bdrv_read(BlockDriverState *bs,
BlockDriverAIOCB *dma_bdrv_write(BlockDriverState *bs,
QEMUSGList *sg, uint64_t sector,
BlockDriverCompletionFunc *cb, void *opaque);
+void dma_helper_init(void);
#endif
diff --git a/vl.c b/vl.c
index 2adfebf..4016c37 100644
--- a/vl.c
+++ b/vl.c
@@ -5517,6 +5517,7 @@ int main(int argc, char **argv, char **envp)
cpu_exec_init_all(tb_size * 1024 * 1024);
bdrv_init();
+ dma_helper_init();
/* we always create the cdrom drive, even if no disk is there */
--
1.6.1.1
next prev parent reply other threads:[~2009-03-16 13:11 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-03-16 13:11 [Qemu-devel] [PATCH 0/5] Fix aio cancellation with synthetic aiocbs Avi Kivity
2009-03-16 13:11 ` [Qemu-devel] [PATCH 1/5] Refactor aio callback allocation to use an aiocb pool Avi Kivity
2009-03-16 13:11 ` [Qemu-devel] [PATCH 2/5] Convert vectored aio emulation to use a dedicated pool Avi Kivity
2009-03-16 13:32 ` Christoph Hellwig
2009-03-16 13:39 ` Avi Kivity
2009-03-16 13:44 ` Christoph Hellwig
2009-03-16 13:11 ` Avi Kivity [this message]
2009-03-16 13:11 ` [Qemu-devel] [PATCH 4/5] Use vectored aiocb storage to store vector translation state Avi Kivity
2009-03-16 13:11 ` [Qemu-devel] [PATCH 5/5] Move block dma helpers aiocb to store dma state Avi Kivity
2009-03-20 18:26 ` [Qemu-devel] Re: [PATCH 0/5] Fix aio cancellation with synthetic aiocbs Anthony Liguori
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1237209071-26942-4-git-send-email-avi@redhat.com \
--to=avi@redhat.com \
--cc=anthony@codemonkey.ws \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).