From patchwork Tue Jun 14 18:18:22 2011 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Subject: [04/13] qed: extract qed_start_allocating_write() Date: Tue, 14 Jun 2011 08:18:22 -0000 From: Stefan Hajnoczi X-Patchwork-Id: 100416 Message-Id: <1308075511-4745-5-git-send-email-stefanha@linux.vnet.ibm.com> To: Cc: Kevin Wolf , Anthony Liguori , Stefan Hajnoczi , Adam Litke Copy-on-read requests are a form of allocating write and will need to be queued like other allocating writes. This patch extracts the request queuing code for allocating writes so that it can be reused for copy-on-read in a later patch. Signed-off-by: Stefan Hajnoczi --- block/qed.c | 32 ++++++++++++++++++++++++++------ 1 files changed, 26 insertions(+), 6 deletions(-) diff --git a/block/qed.c b/block/qed.c index 565bbc1..cc193ad 100644 --- a/block/qed.c +++ b/block/qed.c @@ -1097,14 +1097,15 @@ static bool qed_should_set_need_check(BDRVQEDState *s) } /** - * Write new data cluster + * Start an allocating write request or queue it * - * @acb: Write request - * @len: Length in bytes + * @ret: true if request can proceed, false if queued * - * This path is taken when writing to previously unallocated clusters. + * If a request is queued this function returns false and the caller should + * return. When it becomes time for the request to proceed the qed_aio_next() + * function will be called. */ -static void qed_aio_write_alloc(QEDAIOCB *acb, size_t len) +static bool qed_start_allocating_write(QEDAIOCB *acb) { BDRVQEDState *s = acb_to_s(acb); @@ -1119,7 +1120,26 @@ static void qed_aio_write_alloc(QEDAIOCB *acb, size_t len) } if (acb != QSIMPLEQ_FIRST(&s->allocating_write_reqs) || s->allocating_write_reqs_plugged) { - return; /* wait for existing request to finish */ + return false; + } + return true; +} + +/** + * Write new data cluster + * + * @acb: Write request + * @len: Length in bytes + * + * This path is taken when writing to previously unallocated clusters. + */ +static void qed_aio_write_alloc(QEDAIOCB *acb, size_t len) +{ + BDRVQEDState *s = acb_to_s(acb); + BlockDriverCompletionFunc *cb; + + if (!qed_start_allocating_write(acb)) { + return; } acb->cur_nclusters = qed_bytes_to_clusters(s,