From: "Nicholas A. Bellinger" <nab@linux-iscsi.org>
To: Andy Grover <agrover@redhat.com>, Christoph Hellwig <hch@lst.de>
Cc: target-devel <target-devel@vger.kernel.org>,
linux-scsi <linux-scsi@vger.kernel.org>,
Nicholas Bellinger <nab@linux-iscsi.org>
Subject: [PATCH 1/7] iscsi-target: Fix iscsit_alloc_buffs() breakage with offset_in_page
Date: Fri, 3 Jun 2011 18:18:12 -0700 [thread overview]
Message-ID: <1307150298-23921-2-git-send-email-nab@linux-iscsi.org> (raw)
In-Reply-To: <1307150298-23921-1-git-send-email-nab@linux-iscsi.org>
From: Nicholas Bellinger <nab@linux-iscsi.org>
This patch changes iscsit_alloc_buffs() w/ SCF_SCSI_CONTROL_NONSG_IO_CDB
to take into account the offset_in_page(), and also handles the case
where length and PAGE_SIZE are identical, and require nents to be incremented.
It also moves iscsit_allocate_iovecs() for the SCSI payload case until
after transport_generic_map_mem_to_cmd() has been called because the
iovec allocation depends upon cmd->se_cmd.t_tasks_se_num having been
set by the return of transport_map_sg_to_mem() done in RX thread context.
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
---
drivers/target/iscsi/iscsi_target.c | 78 +++++++++++++++++++++++------------
1 files changed, 52 insertions(+), 26 deletions(-)
diff --git a/drivers/target/iscsi/iscsi_target.c b/drivers/target/iscsi/iscsi_target.c
index 2503ef3..959f616 100644
--- a/drivers/target/iscsi/iscsi_target.c
+++ b/drivers/target/iscsi/iscsi_target.c
@@ -779,27 +779,17 @@ static int iscsit_allocate_iovecs(struct iscsi_cmd *cmd)
static int iscsit_alloc_buffs(struct iscsi_cmd *cmd)
{
- u32 length = cmd->se_cmd.data_length;
struct scatterlist *sgl;
+ void *buf, *cur;
+ u32 length = cmd->se_cmd.data_length;
int nents = DIV_ROUND_UP(length, PAGE_SIZE);
- int i;
- void *buf;
- void *cur;
- int ret;
-
- /* Even no-length cmds need some iovecs, apparently? */
- ret = iscsit_allocate_iovecs(cmd);
- if (ret < 0)
- return ret;
-
+ int i = 0, ret;
+ /*
+ * If no SCSI payload is present, allocate the default iovecs used for
+ * iSCSI PDU Header
+ */
if (!length)
- return 0;
-
- sgl = kzalloc(sizeof(*sgl) * nents, GFP_KERNEL);
- if (!sgl)
- return -ENOMEM;
- sg_init_table(sgl, nents);
-
+ return iscsit_allocate_iovecs(cmd);
/*
* Allocate from slab if nonsg, but sgl should point
* to the malloced mem.
@@ -807,17 +797,38 @@ static int iscsit_alloc_buffs(struct iscsi_cmd *cmd)
* Alloc pages if sg.
*/
if (cmd->se_cmd.se_cmd_flags & SCF_SCSI_CONTROL_NONSG_IO_CDB) {
+ int pg_off = 0, buf_size;
buf = kmalloc(length, GFP_KERNEL);
if (!buf)
return -ENOMEM;
+ /*
+ * Allocate extra SGL for offset_in_page exceeding DIV_ROUND_UP
+ */
+ pg_off = offset_in_page(buf);
+ if ((pg_off + length) > (PAGE_SIZE * nents))
+ nents++;
+
+ sgl = kzalloc(sizeof(*sgl) * nents, GFP_KERNEL);
+ if (!sgl)
+ return -ENOMEM;
+ sg_init_table(sgl, nents);
cur = buf;
while (length) {
- int buf_size = min_t(int, length, PAGE_SIZE);
-
+ if (pg_off != 0) {
+ buf_size = min_t(int, PAGE_SIZE - pg_off, length);
+ pg_off = 0;
+ } else
+ buf_size = min_t(int, length, PAGE_SIZE);
+
sg_set_buf(&sgl[i], cur, buf_size);
+ if (sgl[i].length < 0) {
+ printk("sg_set_buf: page: %p, len: %d, offset: %d\n",
+ sg_page(&sgl[i]), sgl[i].length, sgl[i].offset);
+ BUG();
+ }
length -= buf_size;
cur += buf_size;
i++;
@@ -826,7 +837,12 @@ static int iscsit_alloc_buffs(struct iscsi_cmd *cmd)
cmd->t_mem = buf;
} else {
- i = 0;
+ sgl = kzalloc(sizeof(*sgl) * nents, GFP_KERNEL);
+ if (!sgl)
+ return -ENOMEM;
+
+ sg_init_table(sgl, nents);
+
while (length) {
int buf_size = min_t(int, length, PAGE_SIZE);
struct page *page;
@@ -849,6 +865,13 @@ static int iscsit_alloc_buffs(struct iscsi_cmd *cmd)
/* make se_mem list from the memory */
transport_generic_map_mem_to_cmd(&cmd->se_cmd, sgl, nents, NULL, 0);
+ /*
+ * Allocate iovecs for SCSI payload after transport_generic_map_mem_to_cmd
+ * so that cmd->se_cmd.t_tasks_se_num has been set.
+ */
+ ret = iscsit_allocate_iovecs(cmd);
+ if (ret < 0)
+ goto page_alloc_failed;
return 0;
@@ -1081,12 +1104,15 @@ attach_cmd:
* Active/NonOptimized primary access state..
*/
core_alua_check_nonop_delay(SE_CMD(cmd));
-
+ /*
+ * Allocate and setup SGL used with transport_generic_map_mem_to_cmd().
+ * also call iscsit_allocate_iovecs()
+ */
ret = iscsit_alloc_buffs(cmd);
- if (ret < 0) {
- return ret;
- }
-
+ if (ret < 0)
+ return iscsit_add_reject_from_cmd(
+ ISCSI_REASON_BOOKMARK_NO_RESOURCES,
+ 1, 1, buf, cmd);
/*
* Check the CmdSN against ExpCmdSN/MaxCmdSN here if
* the Immediate Bit is not set, and no Immediate
--
1.7.2.5
next prev parent reply other threads:[~2011-06-04 1:18 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-06-04 1:18 [PATCH 0/7] iscsi-target: Bugfixes for t_mem_list -> t_mem_sg[] conversion Nicholas A. Bellinger
2011-06-04 1:18 ` Nicholas A. Bellinger [this message]
2011-06-04 14:21 ` [PATCH 1/7] iscsi-target: Fix iscsit_alloc_buffs() breakage with offset_in_page Christoph Hellwig
2011-06-07 21:43 ` Andy Grover
2011-06-04 1:18 ` [PATCH 2/7] iscsi-target: Fix padding breakage in iscsit_send_data_in Nicholas A. Bellinger
2011-06-04 1:18 ` [PATCH 3/7] iscsi-target: Clear map_sg usage for non ISTATE_SEND_DATAIN ops Nicholas A. Bellinger
2011-06-04 1:18 ` [PATCH 4/7] iscsi-target: Handle transport_generic_new_cmd failure Nicholas A. Bellinger
2011-06-04 14:00 ` Christoph Hellwig
2011-06-04 1:18 ` [PATCH 5/7] iscsi-target: Fix iscsit_map_iovec cur_len + data_offset breakage Nicholas A. Bellinger
2011-06-04 1:18 ` [PATCH 6/7] iscsi-target: Fix iscsit_fe_sendpage_sg breakage Nicholas A. Bellinger
2011-06-04 1:18 ` [PATCH 7/7] iscsi-target: Fix iscsit_do_crypto_hash_sg() bug Nicholas A. Bellinger
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=1307150298-23921-2-git-send-email-nab@linux-iscsi.org \
--to=nab@linux-iscsi.org \
--cc=agrover@redhat.com \
--cc=hch@lst.de \
--cc=linux-scsi@vger.kernel.org \
--cc=target-devel@vger.kernel.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