From: "Wei Hu (Xavier)" <xavier.huwei@huawei.com>
To: <leon@kernel.org>
Cc: Lijun Ou <oulijun@huawei.com>, <dledford@redhat.com>,
<sean.hefty@intel.com>, <hal.rosenstock@gmail.com>,
<davem@davemloft.net>, <jeffrey.t.kirsher@intel.com>,
<jiri@mellanox.com>, <ogerlitz@mellanox.com>,
<linux-rdma@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
<netdev@vger.kernel.org>, <gongyangming@huawei.com>,
<xiaokun@huawei.com>, <tangchaofei@huawei.com>,
<haifeng.wei@huawei.com>, <yisen.zhuang@huawei.com>,
<yankejian@huawei.com>, <charles.chenxin@huawei.com>,
<linuxarm@huawei.com>
Subject: Re: [PATCH v10 08/22] IB/hns: Add icm support
Date: Mon, 20 Jun 2016 15:49:24 +0800 [thread overview]
Message-ID: <5767A004.4060808@huawei.com> (raw)
In-Reply-To: <20160620060614.GC1172@leon.nu>
On 2016/6/20 14:06, Leon Romanovsky wrote:
> On Mon, Jun 20, 2016 at 12:37:40PM +0800, Wei Hu (Xavier) wrote:
>>
>> On 2016/6/17 17:58, Leon Romanovsky wrote:
>>> On Thu, Jun 16, 2016 at 10:35:16PM +0800, Lijun Ou wrote:
>>>> This patch mainly added icm support for RoCE. It initializes icm
>>>> which managers the relative memory blocks for RoCE. The data
>>>> structures of RoCE will be located in it. For example, CQ table,
>>>> QP table and MTPT table so on.
>>>>
>>>> Signed-off-by: Wei Hu <xavier.huwei@huawei.com>
>>>> Signed-off-by: Nenglong Zhao <zhaonenglong@hisilicon.com>
>>>> Signed-off-by: Lijun Ou <oulijun@huawei.com>
>>>> ---
>>> <...>
>>>
>>>> +
>>>> +static int hns_roce_alloc_icm_pages(struct scatterlist *mem, int order,
>>>> + gfp_t gfp_mask)
>>>> +{
>>>> + struct page *page;
>>>> +
>>>> + page = alloc_pages(gfp_mask, order);
>>>> + if (!page)
>>>> + return -ENOMEM;
>>>> +
>>>> + sg_set_page(mem, page, PAGE_SIZE << order, 0);
>>>> +
>>>> + return 0;
>>>> +}
>>>> +
>>>> +static int hns_roce_alloc_icm_coherent(struct device *dev,
>>>> + struct scatterlist *mem, int order,
>>>> + gfp_t gfp_mask)
>>>> +{
>>>> + void *buf = dma_alloc_coherent(dev, PAGE_SIZE << order,
>>>> + &sg_dma_address(mem), gfp_mask);
>>>> + if (!buf)
>>>> + return -ENOMEM;
>>>> +
>>>> + sg_set_buf(mem, buf, PAGE_SIZE << order);
>>>> + WARN_ON(mem->offset);
>>>> + sg_dma_len(mem) = PAGE_SIZE << order;
>>>> + return 0;
>>>> +}
>>>> +
>>> <...>
>>>
>>>> +
>>>> +static void hns_roce_free_icm_pages(struct hns_roce_dev *hr_dev,
>>>> + struct hns_roce_icm_chunk *chunk)
>>>> +{
>>>> + int i;
>>>> +
>>>> + if (chunk->nsg > 0)
>>>> + dma_unmap_sg(&hr_dev->pdev->dev, chunk->mem, chunk->npages,
>>>> + DMA_BIDIRECTIONAL);
>>>> +
>>>> + for (i = 0; i < chunk->npages; ++i)
>>>> + __free_pages(sg_page(&chunk->mem[i]),
>>>> + get_order(chunk->mem[i].length));
>>> You used alloc_pages for this allocation, so why are you using
>>> __free_pages instead of free_pages?
>> Hi, Leon
>> The function prototype of these functions as below:
>> static inline struct page * alloc_pages(gfp_t gfp_mask, unsigned int
>> order);
>> void free_pages(unsigned long addr, unsigned int order);
>> void __free_pages(struct page *page, unsigned int order);
>>
>> The type of the first parameter of free_pages is same with the type of
>> return value of alloc_pages.
>> Maybe it is better to call __free_pages to release memory that allocated
>> by calling alloc_pages.
> OK, I see.
>
> Another question which you didn't answer [1].
>
> "I wonder if you have the same needs for ICM as it is in mlx4 device.
> Do you have firmware?"
>
> [1] http://marc.info/?l=linux-rdma&m=146545553104913&w=2
Hi, Leon
Now we haven't firmware.
But hardware still need memory for QPC\CQC\MTPT\mtt etc.
Thanks
Wei Hu
>> Regards
>> Wei Hu
>>
next prev parent reply other threads:[~2016-06-20 7:55 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-06-16 14:35 [PATCH v10 00/22] Add HiSilicon RoCE driver Lijun Ou
2016-06-16 14:35 ` [PATCH v10 01/22] net: hns: Add reset function support for " Lijun Ou
2016-06-24 11:49 ` Leon Romanovsky
2016-06-27 6:41 ` oulijun
2016-06-16 14:35 ` [PATCH v10 02/22] devicetree: bindings: IB: Add binding document for HiSilicon RoCE Lijun Ou
2016-06-16 14:35 ` [PATCH v10 03/22] IB/hns: Add initial main frame driver and get cfg info Lijun Ou
2016-06-24 11:48 ` Leon Romanovsky
[not found] ` <576E5D0B.7010003@huawei.com>
2016-06-27 7:00 ` Leon Romanovsky
2016-06-27 7:29 ` Wei Hu (Xavier)
2016-06-16 14:35 ` [PATCH v10 04/22] IB/hns: Add RoCE engine reset function Lijun Ou
2016-06-24 14:59 ` Leon Romanovsky
[not found] ` <576E5C21.5030904@huawei.com>
2016-06-27 8:01 ` Leon Romanovsky
2016-06-27 8:31 ` oulijun
2016-06-28 6:31 ` Wei Hu (Xavier)
2016-06-28 8:09 ` Leon Romanovsky
2016-06-16 14:35 ` [PATCH v10 05/22] IB/hns: Add initial profile resource Lijun Ou
2016-06-24 15:10 ` Leon Romanovsky
2016-06-28 6:56 ` oulijun
2016-06-16 14:35 ` [PATCH v10 06/22] IB/hns: Add initial cmd operation Lijun Ou
2016-06-20 13:33 ` Leon Romanovsky
2016-06-21 10:50 ` Wei Hu (Xavier)
2016-06-21 11:28 ` Leon Romanovsky
2016-06-21 13:01 ` Wei Hu (Xavier)
2016-06-22 4:54 ` Leon Romanovsky
2016-06-22 6:50 ` Wei Hu (Xavier)
2016-06-16 14:35 ` [PATCH v10 07/22] IB/hns: Add event queue support Lijun Ou
2016-06-24 15:46 ` Leon Romanovsky
2016-06-24 15:56 ` Doug Ledford
2016-06-29 8:53 ` oulijun
2016-06-29 10:41 ` Leon Romanovsky
2016-06-16 14:35 ` [PATCH v10 08/22] IB/hns: Add icm support Lijun Ou
2016-06-17 9:58 ` Leon Romanovsky
[not found] ` <57677314.70909@huawei.com>
2016-06-20 6:06 ` Leon Romanovsky
2016-06-20 7:49 ` Wei Hu (Xavier) [this message]
2016-06-20 9:27 ` Leon Romanovsky
2016-06-20 9:48 ` Wei Hu (Xavier)
2016-06-20 13:04 ` Leon Romanovsky
2016-06-21 4:37 ` Wei Hu (Xavier)
2016-06-21 11:55 ` Leon Romanovsky
2016-06-22 3:53 ` Wei Hu (Xavier)
2016-06-16 14:35 ` [PATCH v10 09/22] IB/hns: Add hca support Lijun Ou
2016-06-24 15:50 ` Leon Romanovsky
2016-06-16 14:35 ` [PATCH v10 10/22] IB/hns: Add process flow to init RoCE engine Lijun Ou
2016-06-16 14:35 ` [PATCH v10 11/22] IB/hns: Add IB device registration Lijun Ou
2016-06-16 14:35 ` [PATCH v10 12/22] IB/hns: Set mtu and gid support Lijun Ou
2016-06-16 14:35 ` [PATCH v10 13/22] IB/hns: Add interface of the protocol stack registration Lijun Ou
2016-06-16 14:35 ` [PATCH v10 14/22] IB/hns: Add operations support for IB device and port Lijun Ou
2016-06-16 14:35 ` [PATCH v10 15/22] IB/hns: Add PD operations support Lijun Ou
2016-06-16 14:35 ` [PATCH v10 16/22] IB/hns: Add ah " Lijun Ou
2016-06-16 14:35 ` [PATCH v10 17/22] IB/hns: Add QP " Lijun Ou
2016-06-16 14:35 ` [PATCH v10 18/22] IB/hns: Add CQ " Lijun Ou
2016-06-16 14:35 ` [PATCH v10 19/22] IB/hns: Add memory region " Lijun Ou
2016-06-16 14:35 ` [PATCH v10 20/22] IB/hns: Add operation for getting immutable port Lijun Ou
2016-06-16 14:35 ` [PATCH v10 21/22] IB/hns: Kconfig and Makefile for RoCE module Lijun Ou
2016-06-16 14:35 ` [PATCH v10 22/22] MAINTAINERS: Add maintainers for HiSilicon RoCE driver Lijun Ou
2016-06-24 15:55 ` [PATCH v10 00/22] Add " Leon Romanovsky
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=5767A004.4060808@huawei.com \
--to=xavier.huwei@huawei.com \
--cc=charles.chenxin@huawei.com \
--cc=davem@davemloft.net \
--cc=dledford@redhat.com \
--cc=gongyangming@huawei.com \
--cc=haifeng.wei@huawei.com \
--cc=hal.rosenstock@gmail.com \
--cc=jeffrey.t.kirsher@intel.com \
--cc=jiri@mellanox.com \
--cc=leon@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rdma@vger.kernel.org \
--cc=linuxarm@huawei.com \
--cc=netdev@vger.kernel.org \
--cc=ogerlitz@mellanox.com \
--cc=oulijun@huawei.com \
--cc=sean.hefty@intel.com \
--cc=tangchaofei@huawei.com \
--cc=xiaokun@huawei.com \
--cc=yankejian@huawei.com \
--cc=yisen.zhuang@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).