From: Mike Christie <michael.christie@oracle.com>
To: mingzhe.zou@easystack.cn, torvalds@linux-foundation.org,
zgrieee@gmail.com, martin.petersen@oracle.com
Cc: linux-kernel@vger.kernel.org, linux-scsi@vger.kernel.org,
target-devel@vger.kernel.org, dongsheng.yang@easystack.cn,
zoumingzhe@qq.com
Subject: Re: [PATCH] scsi: target: fixup incorrect use of 'cpumask_t'
Date: Sat, 14 May 2022 11:30:06 -0500 [thread overview]
Message-ID: <64f16117-beb1-def1-5017-c5006c362070@oracle.com> (raw)
In-Reply-To: <20220513062745.3245-1-mingzhe.zou@easystack.cn>
On 5/13/22 1:27 AM, mingzhe.zou@easystack.cn wrote:
> From: mingzhe <mingzhe.zou@easystack.cn>
>
> In commit d72d827f2f26, I used 'cpumask_t' incorrectly.
> ```
> void iscsit_thread_get_cpumask(struct iscsi_conn *conn)
> {
> int ord, cpu;
> cpumask_t conn_allowed_cpumask;
> ......
> }
>
> static ssize_t lio_target_wwn_cpus_allowed_list_store(
> struct config_item *item, const char *page, size_t count)
> {
> int ret;
> char *orig;
> cpumask_t new_allowed_cpumask;
> ......
> }
> ```
>
> So, that the correct pattern should be as follows:
> ```
> cpumask_var_t mask;
>
> if (!alloc_cpumask_var(&mask, GFP_KERNEL))
> return -ENOMEM;
> ... use 'mask' here as a ...
> free_cpumask_var(mask);
> ```
>
Add a:
Fixes: d72d827f2f26 ("scsi: target: Add iscsi/cpus_allowed_list in configfs")
to make it easier for people to pick up.
> Reported-by: Test Bot <zgrieee@gmail.com>
> Signed-off-by: mingzhe <mingzhe.zou@easystack.cn>
> ---
> drivers/target/iscsi/iscsi_target.c | 32 ++++++++++++++------
> drivers/target/iscsi/iscsi_target_configfs.c | 20 ++++++++----
> 2 files changed, 36 insertions(+), 16 deletions(-)
>
> diff --git a/drivers/target/iscsi/iscsi_target.c b/drivers/target/iscsi/iscsi_target.c
> index 6fe6a6bab3f4..3f9f5b8879fe 100644
> --- a/drivers/target/iscsi/iscsi_target.c
> +++ b/drivers/target/iscsi/iscsi_target.c
> @@ -3596,10 +3596,7 @@ static int iscsit_send_reject(
> void iscsit_thread_get_cpumask(struct iscsi_conn *conn)
> {
> int ord, cpu;
> - cpumask_t conn_allowed_cpumask;
> -
> - cpumask_and(&conn_allowed_cpumask, iscsit_global->allowed_cpumask,
> - cpu_online_mask);
> + cpumask_var_t conn_allowed_cpumask;
>
> /*
> * bitmap_id is assigned from iscsit_global->ts_bitmap from
> @@ -3609,13 +3606,28 @@ void iscsit_thread_get_cpumask(struct iscsi_conn *conn)
> * iSCSI connection's RX/TX threads will be scheduled to
> * execute upon.
> */
> - cpumask_clear(conn->conn_cpumask);
> - ord = conn->bitmap_id % cpumask_weight(&conn_allowed_cpumask);
> - for_each_cpu(cpu, &conn_allowed_cpumask) {
> - if (ord-- == 0) {
> - cpumask_set_cpu(cpu, conn->conn_cpumask);
> - return;
> + if (!alloc_cpumask_var(&conn_allowed_cpumask, GFP_KERNEL)) {
Do a zalloc_cpumask_var to make sure it's all initialized.
> + ord = conn->bitmap_id % cpumask_weight(cpu_online_mask);
> + for_each_online_cpu(cpu) {
> + if (ord-- == 0) {
> + cpumask_set_cpu(cpu, conn->conn_cpumask);
> + return;
> + }
> + }
> + } else {
> + cpumask_and(conn_allowed_cpumask, iscsit_global->allowed_cpumask,
> + cpu_online_mask);
> +
> + cpumask_clear(conn->conn_cpumask);
> + ord = conn->bitmap_id % cpumask_weight(conn_allowed_cpumask);
> + for_each_cpu(cpu, conn_allowed_cpumask) {
> + if (ord-- == 0) {
> + cpumask_set_cpu(cpu, conn->conn_cpumask);
> + free_cpumask_var(conn_allowed_cpumask);
> + return;
> + }
> }
> + free_cpumask_var(conn_allowed_cpumask);
> }
> /*
> * This should never be reached..
> diff --git a/drivers/target/iscsi/iscsi_target_configfs.c b/drivers/target/iscsi/iscsi_target_configfs.c
> index 0cedcfe207b5..ae9319934cd0 100644
> --- a/drivers/target/iscsi/iscsi_target_configfs.c
> +++ b/drivers/target/iscsi/iscsi_target_configfs.c
> @@ -1139,20 +1139,28 @@ static ssize_t lio_target_wwn_cpus_allowed_list_store(
> {
> int ret;
> char *orig;
> - cpumask_t new_allowed_cpumask;
> + cpumask_var_t new_allowed_cpumask;
> +
> + if (!alloc_cpumask_var(&new_allowed_cpumask, GFP_KERNEL))
> + return -ENOMEM;
You can do zalloc_cpumask_var and then drop the clear call below.
>
> orig = kstrdup(page, GFP_KERNEL);
> - if (!orig)
> + if (!orig) {
> + free_cpumask_var(new_allowed_cpumask);
> return -ENOMEM;
Just add a goto at the bottom before free_cpumask_var then have this
and the other failure path use it. Make sure to also fix up the
return code (set count to -ENOMEM/ret or use ret at the bottom and
set count to ret).
> + }
>
> - cpumask_clear(&new_allowed_cpumask);
> - ret = cpulist_parse(orig, &new_allowed_cpumask);
> + cpumask_clear(new_allowed_cpumask);
> + ret = cpulist_parse(orig, new_allowed_cpumask);
>
> kfree(orig);
> - if (ret != 0)
> + if (ret != 0) {
> + free_cpumask_var(new_allowed_cpumask);
> return ret;
> + }
>
> - cpumask_copy(iscsit_global->allowed_cpumask, &new_allowed_cpumask);
> + cpumask_copy(iscsit_global->allowed_cpumask, new_allowed_cpumask);
> + free_cpumask_var(new_allowed_cpumask);
> return count;
> }
>
prev parent reply other threads:[~2022-05-14 16:30 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-05-13 6:27 mingzhe.zou
2022-05-14 16:30 ` Mike Christie [this message]
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=64f16117-beb1-def1-5017-c5006c362070@oracle.com \
--to=michael.christie@oracle.com \
--cc=dongsheng.yang@easystack.cn \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=martin.petersen@oracle.com \
--cc=mingzhe.zou@easystack.cn \
--cc=target-devel@vger.kernel.org \
--cc=torvalds@linux-foundation.org \
--cc=zgrieee@gmail.com \
--cc=zoumingzhe@qq.com \
/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
all inboxes | Powered by JetHome®