From: Leon Romanovsky <leon@kernel.org>
To: "Wei Hu (Xavier)" <xavier.huwei@huawei.com>
Cc: dledford@redhat.com, linux-rdma@vger.kernel.org,
lijun_nudt@163.com, oulijun@huawei.com,
charles.chenxin@huawei.com, liuyixian@huawei.com,
xushaobo2@huawei.com, zhangxiping3@huawei.com,
xavier.huwei@tom.com, linuxarm@huawei.com,
linux-kernel@vger.kernel.org, shaobohsu@163.com
Subject: Re: [PATCH for-next 3/9] RDMA/hns: Add return statement when kzalloc return NULL in hns_roce_v1_recreate_lp_qp
Date: Fri, 29 Sep 2017 13:23:37 +0300 [thread overview]
Message-ID: <20170929102337.GK2297@mtr-leonro.local> (raw)
In-Reply-To: <59CDE31A.5090707@huawei.com>
[-- Attachment #1: Type: text/plain, Size: 3883 bytes --]
On Fri, Sep 29, 2017 at 02:07:22PM +0800, Wei Hu (Xavier) wrote:
>
>
> On 2017/9/28 20:59, Leon Romanovsky wrote:
> > On Thu, Sep 28, 2017 at 07:56:59PM +0800, Wei Hu (Xavier) wrote:
> > >
> > > On 2017/9/28 17:13, Leon Romanovsky wrote:
> > > > On Thu, Sep 28, 2017 at 12:57:28PM +0800, Wei Hu (Xavier) wrote:
> > > > > From: Lijun Ou <oulijun@huawei.com>
> > > > >
> > > > > When lp_qp_work is NULL, it should be returned ENOMEM. This patch
> > > > > mainly fixes it.
> > > > >
> > > > > Ihis patch fixes the smatch error as below:
> > > > > drivers/infiniband/hw/hns/hns_roce_hw_v1.c:918 hns_roce_v1_recreate_lp_qp()
> > > > > error: potential null dereference 'lp_qp_work'. (kzalloc returns null)
> > > > >
> > > > > Signed-off-by: Lijun Ou <oulijun@huawei.com>
> > > > > Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
> > > > > Signed-off-by: Shaobo Xu <xushaobo2@huawei.com>
> > > > > ---
> > > > > drivers/infiniband/hw/hns/hns_roce_hw_v1.c | 2 ++
> > > > > 1 file changed, 2 insertions(+)
> > > > >
> > > > > diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
> > > > > index 95f5c88..1071fa2 100644
> > > > > --- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
> > > > > +++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
> > > > > @@ -912,6 +912,8 @@ static int hns_roce_v1_recreate_lp_qp(struct hns_roce_dev *hr_dev)
> > > > >
> > > > > lp_qp_work = kzalloc(sizeof(struct hns_roce_recreate_lp_qp_work),
> > > > > GFP_KERNEL);
> > > > > + if (!lp_qp_work)
> > > > > + return -ENOMEM;
> > > > >
> > > > You will treat this error in the same was as you will treat timeout,
> > > > which is wrong.
> > > Thanks, Leon
> > > We will send v2 to fix the compatible warn info.
> > No, you missed the point.
> > From the code flow below the behavior of hns_roce_v1_recreate_lp_qp
> > for ENOMEM and ETIMEOUT returns will be the same and it is wrong.
> >
> > For the ETIMEOUT, you can continue, for ENOMEM, you should properly
> > unfold the whole flow.
> >
> > Thanks
> >
> Hi, Leon
> We prepare to modify the warn info as bleow:
>
> if (hr_dev->hw->dereg_mr && hns_roce_v1_recreate_lp_qp(hr_dev))
> dev_warn(&hr_dev->pdev->dev, "recreate lp qp failed!\n");
>
> for -ETIMEDOUT, there is a warn info as blow, but there isn't this one
> for -ENOMEM.
> dev_warn(dev, "recreate lp qp failed 20s timeout and return
> failed!\n");
>
> static int hns_roce_v1_recreate_lp_qp(struct hns_roce_dev *hr_dev)
> {
> <snip>
> lp_qp_work = kzalloc(sizeof(struct
> hns_roce_recreate_lp_qp_work),
> GFP_KERNEL);
> if (!lp_qp_work)
> return -ENOMEM;
>
> <snip>
> dev_warn(dev, "recreate lp qp failed 20s timeout and return
> failed!\n");
> return -ETIMEDOUT;
> }
>
> Regards
> Wei Hu
Hi Wei,
It will be helpful, if you post your suggestions in git diff format.
My expectation is to see the following code:
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index 747efd1ae5a6..0b9ec7c24f2d 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -1648,14 +1648,18 @@ void hns_roce_v1_set_mac(struct hns_roce_dev *hr_dev, u8 phy_port, u8 *addr)
u16 *p_h;
u32 *p;
u32 val;
/*
* When mac changed, loopback may fail
* because of smac not equal to dmac.
* We Need to release and create reserved qp again.
*/
- if (hr_dev->hw->dereg_mr && hns_roce_v1_recreate_lp_qp(hr_dev))
- dev_warn(&hr_dev->pdev->dev, "recreate lp qp timeout!\n");
+ if (hr_dev->hw->dereg_mr) {
+ int ret;
+ ret = hns_roce_v1_recreate_lp_qp(hr_dev);
+ if (ret && ret != -ETIMEDOUT)
+ return ret;
+ }
p = (u32 *)(&addr[0]);
reg_smac_l = *p;
Thanks
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]
next prev parent reply other threads:[~2017-09-29 10:23 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-09-28 4:57 [PATCH for-next 0/9] Bug fixes & Code improvements in hip06 and hip08 RoCE driver Wei Hu (Xavier)
2017-09-28 4:57 ` [PATCH for-next 1/9] RDMA/hns: Modify the value with rd&dest_rd of qp_attr Wei Hu (Xavier)
2017-09-28 4:57 ` [PATCH for-next 2/9] RDMA/hns: Factor out the code for checking sdb status into a new function Wei Hu (Xavier)
2017-09-28 13:50 ` Leon Romanovsky
2017-09-29 2:05 ` Wei Hu (Xavier)
2017-09-28 4:57 ` [PATCH for-next 3/9] RDMA/hns: Add return statement when kzalloc return NULL in hns_roce_v1_recreate_lp_qp Wei Hu (Xavier)
2017-09-28 9:13 ` Leon Romanovsky
2017-09-28 11:56 ` Wei Hu (Xavier)
2017-09-28 12:59 ` Leon Romanovsky
2017-09-29 6:07 ` Wei Hu (Xavier)
2017-09-29 10:23 ` Leon Romanovsky [this message]
2017-09-29 13:15 ` Wei Hu (Xavier)
2017-09-28 4:57 ` [PATCH for-next 4/9] RDMA/hns: Set mask for destination qp field of qp context assignment Wei Hu (Xavier)
2017-09-28 4:57 ` [PATCH for-next 5/9] RDMA/hns: Set rdma_ah_attr type for querying qp Wei Hu (Xavier)
2017-09-28 4:57 ` [PATCH for-next 6/9] RDMA/hns: Add return statement when checking error in hns_roce_v1_mr_free_work_fn Wei Hu (Xavier)
2017-09-28 4:57 ` [PATCH for-next 7/9] RDMA/hns: Remove unnecessarily calling unregister_inetaddr_notifier function Wei Hu (Xavier)
2017-09-28 4:57 ` [PATCH for-next 8/9] RDMA/hns: Remove unused struct members in hns-abi.h Wei Hu (Xavier)
2017-09-28 9:02 ` Leon Romanovsky
2017-09-28 11:56 ` Wei Hu (Xavier)
2017-09-28 13:04 ` Leon Romanovsky
2017-09-28 16:12 ` Wei Hu (Xavier)
2017-09-28 4:57 ` [PATCH for-next 9/9] RDMA/hns: Replace usleep_range with udelay when checking command status Wei Hu (Xavier)
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=20170929102337.GK2297@mtr-leonro.local \
--to=leon@kernel.org \
--cc=charles.chenxin@huawei.com \
--cc=dledford@redhat.com \
--cc=lijun_nudt@163.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rdma@vger.kernel.org \
--cc=linuxarm@huawei.com \
--cc=liuyixian@huawei.com \
--cc=oulijun@huawei.com \
--cc=shaobohsu@163.com \
--cc=xavier.huwei@huawei.com \
--cc=xavier.huwei@tom.com \
--cc=xushaobo2@huawei.com \
--cc=zhangxiping3@huawei.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;
as well as URLs for NNTP newsgroup(s).