netdev - Re: [PATCH 1/3] infiniband: IB/hns: add Hisilicon RoCE support

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20160311104239.GD3931@nanopsycho.orion>
Date:	Fri, 11 Mar 2016 11:42:40 +0100
From:	Jiri Pirko <jiri@...nulli.us>
To:	Lijun Ou <oulijun@...wei.com>
Cc:	dledford@...hat.com, sean.hefty@...el.com,
	hal.rosenstock@...il.com, davem@...emloft.net,
	jeffrey.t.kirsher@...el.com, jiri@...lanox.com,
	ogerlitz@...lanox.com, linux-rdma@...r.kernel.org,
	linux-kernel@...r.kernel.org, netdev@...r.kernel.org,
	gongyangming@...wei.com, xiaokun@...wei.com,
	tangchaofei@...wei.com, haifeng.wei@...wei.com,
	yisen.zhuang@...wei.com, yankejian@...wei.com,
	lisheng011@...wei.com, charles.chenxin@...wei.com,
	linuxarm@...wei.com
Subject: Re: [PATCH 1/3] infiniband: IB/hns: add Hisilicon RoCE support

Fri, Mar 11, 2016 at 11:37:09AM CET, oulijun@...wei.com wrote:
>The driver for Hisilicon RoCE is a platform driver.
>The driver will support mulitple versions of hardware. Currently only "v1"
>for hip06 SOC is supported.
>The driver includes two parts: common driver and hardware-specific
>operations. hns_roce_v1_hw.c and hns_roce_v1_hw.h are files for
>hardware-specific operations only for v1 engine, and other files(.c and .h)
>for common algorithm and common hardware operations
>
>Signed-off-by: Lijun Ou <oulijun@...wei.com>
>Signed-off-by: Wei Hu(Xavier) <xavier.huwei@...wei.com>
>Signed-off-by: Znlong <zhaonenglong@...wei.com>

<snip>

I'm sorry to be nitpicking, but you still have style issues in your
code. I believe that for newly submitted code, this should be avoided. I
already pointed that out as a comment to your last version, but you
ignored it. So again, couple of examples:

>+struct ib_ah *hns_roce_create_ah(struct ib_pd *ibpd,
>+					struct ib_ah_attr *ah_attr)

<snip>	
	
>+	ret = ib_get_cached_gid(ibpd->device, ah_attr->port_num,
>+			ah_attr->grh.sgid_index, &sgid, &gid_attr);

<snip>

>+int hns_roce_bitmap_alloc_range(
>+					struct hns_roce_bitmap *bitmap,
>+					int cnt, int align, u32 *obj)

<snip>

>+			pages =
>+				kmalloc(sizeof(*pages) * buf->nbufs,
>+					GFP_KERNEL);

<snip>

>+		dev_err(dev,
>+			"CQ alloc.Failed to find cq buf addr.\n");

<snip>

>+	resp.qp_tab_size      = hr_dev->caps.num_qps;

<snip>

>+		buddy->bits[i] =
>+			kmalloc(s * sizeof(long), GFP_KERNEL);


and many, many others similar to this. Please fix this.


Also, I don't understand why you have "_" prefix for labels:

>+
>+_error_failed_register_device:
>+	hns_roce_engine_uninit(hr_dev);
>+
>+_error_failed_engine_init:
>+	hns_roce_cleanup_bitmap(hr_dev);
>+
>+_error_failed_setup_hca:
>+	hns_roce_cleanup_icm(hr_dev);
>+
>+_error_failed_init_icm:
>+	if (hr_dev->cmd_mod)
>+		hns_roce_cmd_use_polling(hr_dev);
>+
>+_error_failed_use_event:
>+	hns_roce_cleanup_eq_table(hr_dev);
>+
>+_error_failed_eq_tabel:
>+	hns_roce_cmd_cleanup(hr_dev);
>+
>+_error_failed_cmd_init:
>+	(void)hns_roce_engine_reset(hr_dev, 0);
>+
>+_error_failed_reset_engine:
>+	hns_roce_free_cfg(hr_dev);
>+
>+_error_failed_get_cfg:
>+	ib_dealloc_device(&hr_dev->ib_dev);
>+
>+	return ret;
>+}