public inbox for linux-scsi@vger.kernel.org
 help / color / mirror / Atom feed
From: John Garry <john.garry@huawei.com>
To: Steffen Maier <maier@linux.ibm.com>,
	jejb@linux.vnet.ibm.com, martin.petersen@oracle.com
Cc: linuxarm@huawei.com, linux-kernel@vger.kernel.org,
	linux-scsi@vger.kernel.org,
	Xiang Chen <chenxiang66@hisilicon.com>
Subject: Re: [PATCH 3/4] scsi: hisi_sas: Make sg_tablesize consistent value
Date: Wed, 21 Nov 2018 11:45:10 +0000	[thread overview]
Message-ID: <b4411de6-aba0-4398-e9c0-1c3698ffe1fd@huawei.com> (raw)
In-Reply-To: <64113584-d9cc-6662-bcde-852ec7957517@linux.ibm.com>

On 21/11/2018 11:08, Steffen Maier wrote:
> On 11/21/2018 12:02 PM, Steffen Maier wrote:
>> On 11/20/2018 03:59 PM, John Garry wrote:
>>> From: Xiang Chen <chenxiang66@hisilicon.com>
>>>
>>> Sht->sg_tablesize is set in the driver, and it will be assigned to
>>> shost->sg_tablesize in SCSI mid-layer. So it is not necessary to
>>> assign shost->sg_table one more time in the driver.
>>>
>>> In addition to the change, change each scsi_host_template.sg_tablesize
>>> to HISI_SAS_SGE_PAGE_CNT instead of SG_ALL.
>>
>> Might be completely irrelevant, so just as information: I once had
>> problems due to changing (reducing) SHT.sg_tablesize because block
>> queue limits of BSG devices of Scsi_Host and fc_host (not sure if you
>> have an equivalent bsg device for your transport(s)) inherit from SHT,
>> but don't update (automatically) on later updates of
>> shost->sg_tablesize, which in turn affect scsi_devices allocated after
>> the shost update.
>> Cf.
>> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/drivers/s390/scsi?id=5fea4291deacd80188b996d2f555fc6a1940e5d4
>
>
> Figured, your new constant seems to have the same value so no problem.
>

Right, so previously we were doing similar to what you describe - 
setting the value in the SHT and then setting shost->sg_tablesize after 
the host is allocated.

However the values were the same, so in this patch we're just removing 
setting shost->sg_tablesize (again).

Thanks,
John

> #define SG_CHUNK_SIZE    128
> #define SG_ALL    SG_CHUNK_SIZE
> #define HISI_SAS_SGE_PAGE_CNT SG_CHUNK_SIZE
>
>> ("[SCSI] zfcp: block queue limits with data router")
>> if you need more details.
>>
>>> Signed-off-by: Xiang Chen <chenxiang66@hisilicon.com>
>>> Signed-off-by: John Garry <john.garry@huawei.com>
>>> ---
>>>   drivers/scsi/hisi_sas/hisi_sas_main.c  | 1 -
>>>   drivers/scsi/hisi_sas/hisi_sas_v1_hw.c | 2 +-
>>>   drivers/scsi/hisi_sas/hisi_sas_v2_hw.c | 2 +-
>>>   drivers/scsi/hisi_sas/hisi_sas_v3_hw.c | 3 +--
>>>   4 files changed, 3 insertions(+), 5 deletions(-)
>>>
>>> diff --git a/drivers/scsi/hisi_sas/hisi_sas_main.c
>>> b/drivers/scsi/hisi_sas/hisi_sas_main.c
>>> index d13a662..cbda48e 100644
>>> --- a/drivers/scsi/hisi_sas/hisi_sas_main.c
>>> +++ b/drivers/scsi/hisi_sas/hisi_sas_main.c
>>> @@ -2410,7 +2410,6 @@ int hisi_sas_probe(struct platform_device *pdev,
>>>       shost->max_lun = ~0;
>>>       shost->max_channel = 1;
>>>       shost->max_cmd_len = 16;
>>> -    shost->sg_tablesize = min_t(u16, SG_ALL, HISI_SAS_SGE_PAGE_CNT);
>>>       if (hisi_hba->hw->slot_index_alloc) {
>>>           shost->can_queue = hisi_hba->hw->max_command_entries;
>>>           shost->cmd_per_lun = hisi_hba->hw->max_command_entries;
>>> diff --git a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
>>> b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
>>> index d24342b..2d035cc 100644
>>> --- a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
>>> +++ b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
>>> @@ -1815,7 +1815,7 @@ static int hisi_sas_v1_init(struct hisi_hba
>>> *hisi_hba)
>>>       .change_queue_depth    = sas_change_queue_depth,
>>>       .bios_param        = sas_bios_param,
>>>       .this_id        = -1,
>>> -    .sg_tablesize        = SG_ALL,
>>> +    .sg_tablesize        = HISI_SAS_SGE_PAGE_CNT,
>>>       .max_sectors        = SCSI_DEFAULT_MAX_SECTORS,
>>>       .use_clustering        = ENABLE_CLUSTERING,
>>>       .eh_device_reset_handler = sas_eh_device_reset_handler,
>>> diff --git a/drivers/scsi/hisi_sas/hisi_sas_v2_hw.c
>>> b/drivers/scsi/hisi_sas/hisi_sas_v2_hw.c
>>> index e78a97e..79e58a7 100644
>>> --- a/drivers/scsi/hisi_sas/hisi_sas_v2_hw.c
>>> +++ b/drivers/scsi/hisi_sas/hisi_sas_v2_hw.c
>>> @@ -3570,7 +3570,7 @@ struct device_attribute *host_attrs_v2_hw[] = {
>>>       .change_queue_depth    = sas_change_queue_depth,
>>>       .bios_param        = sas_bios_param,
>>>       .this_id        = -1,
>>> -    .sg_tablesize        = SG_ALL,
>>> +    .sg_tablesize        = HISI_SAS_SGE_PAGE_CNT,
>>>       .max_sectors        = SCSI_DEFAULT_MAX_SECTORS,
>>>       .use_clustering        = ENABLE_CLUSTERING,
>>>       .eh_device_reset_handler = sas_eh_device_reset_handler,
>>> diff --git a/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c
>>> b/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c
>>> index 7e2b020..8a08078 100644
>>> --- a/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c
>>> +++ b/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c
>>> @@ -2224,7 +2224,7 @@ struct device_attribute *host_attrs_v3_hw[] = {
>>>       .change_queue_depth    = sas_change_queue_depth,
>>>       .bios_param        = sas_bios_param,
>>>       .this_id        = -1,
>>> -    .sg_tablesize        = SG_ALL,
>>> +    .sg_tablesize        = HISI_SAS_SGE_PAGE_CNT,
>>>       .max_sectors        = SCSI_DEFAULT_MAX_SECTORS,
>>>       .use_clustering        = ENABLE_CLUSTERING,
>>>       .eh_device_reset_handler = sas_eh_device_reset_handler,
>>> @@ -2366,7 +2366,6 @@ struct device_attribute *host_attrs_v3_hw[] = {
>>>       shost->max_lun = ~0;
>>>       shost->max_channel = 1;
>>>       shost->max_cmd_len = 16;
>>> -    shost->sg_tablesize = min_t(u16, SG_ALL, HISI_SAS_SGE_PAGE_CNT);
>>>       shost->can_queue = hisi_hba->hw->max_command_entries -
>>>           HISI_SAS_RESERVED_IPTT_CNT;
>>>       shost->cmd_per_lun = hisi_hba->hw->max_command_entries -
>>>
>>
>>
>
>

  reply	other threads:[~2018-11-21 11:45 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-11-20 14:59 [PATCH 0/4] hisi_sas: DIF/DIX support John Garry
2018-11-20 14:59 ` [PATCH 1/4] scsi: Add scsi_prot_op_normal() John Garry
2018-11-20 16:19   ` Christoph Hellwig
2018-11-20 16:36     ` John Garry
2018-11-20 14:59 ` [PATCH 2/4] scsi: hisi_sas: Relocate some code to reduce complexity John Garry
2018-11-20 14:59 ` [PATCH 3/4] scsi: hisi_sas: Make sg_tablesize consistent value John Garry
2018-11-21 11:02   ` Steffen Maier
2018-11-21 11:08     ` Steffen Maier
2018-11-21 11:45       ` John Garry [this message]
2018-11-20 14:59 ` [PATCH 4/4] scsi: hisi_sas: Add support for DIF/DIX feature for v3 hw John Garry
2018-11-24 17:50   ` kbuild test robot
2018-11-26 15:42     ` John Garry
2018-12-05  0:50       ` [kbuild-all] " Li, Philip

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=b4411de6-aba0-4398-e9c0-1c3698ffe1fd@huawei.com \
    --to=john.garry@huawei.com \
    --cc=chenxiang66@hisilicon.com \
    --cc=jejb@linux.vnet.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-scsi@vger.kernel.org \
    --cc=linuxarm@huawei.com \
    --cc=maier@linux.ibm.com \
    --cc=martin.petersen@oracle.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